Syntax coding

Change-Id: I6cac24c4f1e44f29ffcc9b87ba1167eeb32d1b69
2015-04-15 16:48:45 -07:00
1110 changed files with 188194 additions and 193322 deletions
--- a/.clang-format
+++ b/.clang-format
@@ -1,68 +0,0 @@
---
-Language:        Cpp
-# BasedOnStyle:  Google
-# Generated with clang-format 3.7.1
-AccessModifierOffset: -1
-AlignAfterOpenBracket: true
-AlignConsecutiveAssignments: false
-AlignEscapedNewlinesLeft: true
-AlignOperands:   true
-AlignTrailingComments: true
-AllowAllParametersOfDeclarationOnNextLine: true
-AllowShortBlocksOnASingleLine: false
-AllowShortCaseLabelsOnASingleLine: true
-AllowShortFunctionsOnASingleLine: All
-AllowShortIfStatementsOnASingleLine: true
-AllowShortLoopsOnASingleLine: true
-AlwaysBreakAfterDefinitionReturnType: None
-AlwaysBreakBeforeMultilineStrings: true
-AlwaysBreakTemplateDeclarations: true
-BinPackArguments: true
-BinPackParameters: true
-BreakBeforeBinaryOperators: None
-BreakBeforeBraces: Attach
-BreakBeforeTernaryOperators: true
-BreakConstructorInitializersBeforeComma: false
-ColumnLimit:     80
-CommentPragmas:  '^ IWYU pragma:'
-ConstructorInitializerAllOnOneLineOrOnePerLine: false
-ConstructorInitializerIndentWidth: 4
-ContinuationIndentWidth: 4
-Cpp11BracedListStyle: false
-DerivePointerAlignment: false
-DisableFormat:   false
-ExperimentalAutoDetectBinPacking: false
-ForEachMacros:   [ foreach, Q_FOREACH, BOOST_FOREACH ]
-IndentCaseLabels: true
-IndentWidth:     2
-IndentWrappedFunctionNames: false
-KeepEmptyLinesAtTheStartOfBlocks: false
-MacroBlockBegin: ''
-MacroBlockEnd:   ''
-MaxEmptyLinesToKeep: 1
-NamespaceIndentation: None
-ObjCBlockIndentWidth: 2
-ObjCSpaceAfterProperty: false
-ObjCSpaceBeforeProtocolList: false
-PenaltyBreakBeforeFirstCallParameter: 1
-PenaltyBreakComment: 300
-PenaltyBreakFirstLessLess: 120
-PenaltyBreakString: 1000
-PenaltyExcessCharacter: 1000000
-PenaltyReturnTypeOnItsOwnLine: 200
-PointerAlignment: Right
-SpaceAfterCStyleCast: false
-SpaceBeforeAssignmentOperators: true
-SpaceBeforeParens: ControlStatements
-SpaceInEmptyParentheses: false
-SpacesBeforeTrailingComments: 2
-SpacesInAngles:  false
-SpacesInContainerLiterals: true
-SpacesInCStyleCastParentheses: false
-SpacesInParentheses: false
-SpacesInSquareBrackets: false
-Standard:        Auto
-TabWidth:        8
-UseTab:          Never
-...
-
--- a/.gitignore
+++ b/.gitignore
@@ -30,17 +30,14 @@
 /examples/decode_with_partial_drops
 /examples/example_xma
 /examples/postproc
-/examples/resize_util
 /examples/set_maps
 /examples/simple_decoder
 /examples/simple_encoder
 /examples/twopass_encoder
 /examples/vp8_multi_resolution_encoder
 /examples/vp8cx_set_ref
-/examples/vp9_lossless_encoder
 /examples/vp9_spatial_scalable_encoder
 /examples/vpx_temporal_scalable_patterns
-/examples/vpx_temporal_svc_encoder
 /ivfdec
 /ivfdec.dox
 /ivfenc
@@ -48,14 +45,12 @@
 /libvpx.so*
 /libvpx.ver
 /samples.dox
-/test_intra_pred_speed
 /test_libvpx
 /vp8_api1_migration.dox
 /vp[89x]_rtcd.h
 /vpx.pc
 /vpx_config.c
 /vpx_config.h
-/vpx_dsp_rtcd.h
 /vpx_scale_rtcd.h
 /vpx_version.h
 /vpxdec
--- a/.mailmap
+++ b/.mailmap
@@ -1,32 +1,18 @@
 Adrian Grange <agrange@google.com>
-Aℓex Converse <aconverse@google.com>
-Aℓex Converse <aconverse@google.com> <alex.converse@gmail.com>
 Alexis Ballier <aballier@gentoo.org> <alexis.ballier@gmail.com>
-Alpha Lam <hclam@google.com> <hclam@chromium.org>
-Deb Mukherjee <debargha@google.com>
-Erik Niemeyer <erik.a.niemeyer@intel.com> <erik.a.niemeyer@gmail.com>
-Guillaume Martres <gmartres@google.com> <smarter3@gmail.com>
 Hangyu Kuang <hkuang@google.com>
-Hui Su <huisu@google.com>
-Jacky Chen <jackychen@google.com>
 Jim Bankoski <jimbankoski@google.com>
+John Koleszar <jkoleszar@google.com>
 Johann Koenig <johannkoenig@google.com>
 Johann Koenig <johannkoenig@google.com> <johann.koenig@duck.com>
-Johann Koenig <johannkoenig@google.com> <johann.koenig@gmail.com>
-John Koleszar <jkoleszar@google.com>
-Joshua Litt <joshualitt@google.com> <joshualitt@chromium.org>
-Marco Paniconi <marpan@google.com>
-Marco Paniconi <marpan@google.com> <marpan@chromium.org>
+Johann Koenig <johannkoenig@google.com> <johannkoenig@dhcp-172-19-7-52.mtv.corp.google.com>
 Pascal Massimino <pascal.massimino@gmail.com>
-Paul Wilkins <paulwilkins@google.com>
-Ralph Giles <giles@xiph.org> <giles@entropywave.com>
-Ralph Giles <giles@xiph.org> <giles@mozilla.com>
-Ronald S. Bultje <rsbultje@gmail.com> <rbultje@google.com>
 Sami Pietilä <samipietila@google.com>
-Tamar Levy <tamar.levy@intel.com>
-Tamar Levy <tamar.levy@intel.com> <levytamar82@gmail.com>
 Tero Rintaluoma <teror@google.com> <tero.rintaluoma@on2.com>
 Timothy B. Terriberry <tterribe@xiph.org> Tim Terriberry <tterriberry@mozilla.com>
 Tom Finegan <tomfinegan@google.com>
-Tom Finegan <tomfinegan@google.com> <tomfinegan@chromium.org>
+Ralph Giles <giles@xiph.org> <giles@entropywave.com>
+Ralph Giles <giles@xiph.org> <giles@mozilla.com>
+Alpha Lam <hclam@google.com> <hclam@chromium.org>
+Deb Mukherjee <debargha@google.com>
 Yaowu Xu <yaowu@google.com> <yaowu@xuyaowu.com>
--- a/40
+++ b/40
@@ -3,11 +3,10 @@

 Aaron Watry <awatry@gmail.com>
 Abo Talib Mahfoodh <ab.mahfoodh@gmail.com>
-Adam Xu <adam@xuyaowu.com>
 Adrian Grange <agrange@google.com>
-Aℓex Converse <aconverse@google.com>
 Ahmad Sharif <asharif@google.com>
 Alexander Voronov <avoronov@graphics.cs.msu.ru>
+Alex Converse <alex.converse@gmail.com>
 Alexis Ballier <aballier@gentoo.org>
 Alok Ahuja <waveletcoeff@gmail.com>
 Alpha Lam <hclam@google.com>
@@ -15,65 +14,44 @@ A.Mahfoodh <ab.mahfoodh@gmail.com>
 Ami Fischman <fischman@chromium.org>
 Andoni Morales Alastruey <ylatuya@gmail.com>
 Andres Mejia <mcitadel@gmail.com>
-Andrew Russell <anrussell@google.com>
-Angie Chiang <angiebird@google.com>
 Aron Rosenberg <arosenberg@logitech.com>
 Attila Nagy <attilanagy@google.com>
-Brion Vibber <bvibber@wikimedia.org>
 changjun.yang <changjun.yang@intel.com>
-Charles 'Buck' Krasic <ckrasic@google.com>
 chm <chm@rock-chips.com>
 Christian Duvivier <cduvivier@google.com>
 Daniel Kang <ddkang@google.com>
 Deb Mukherjee <debargha@google.com>
-Dim Temp <dimtemp0@gmail.com>
 Dmitry Kovalev <dkovalev@google.com>
 Dragan Mrdjan <dmrdjan@mips.com>
-Ed Baker <edward.baker@intel.com>
-Ehsan Akhgari <ehsan.akhgari@gmail.com>
-Erik Niemeyer <erik.a.niemeyer@intel.com>
+Erik Niemeyer <erik.a.niemeyer@gmail.com>
 Fabio Pedretti <fabio.ped@libero.it>
 Frank Galligan <fgalligan@google.com>
 Fredrik Söderquist <fs@opera.com>
 Fritz Koenig <frkoenig@google.com>
 Gaute Strokkenes <gaute.strokkenes@broadcom.com>
-Geza Lore <gezalore@gmail.com>
-Ghislain MARY <ghislainmary2@gmail.com>
 Giuseppe Scrivano <gscrivano@gnu.org>
-Gordana Cmiljanovic <gordana.cmiljanovic@imgtec.com>
 Guillaume Martres <gmartres@google.com>
 Guillermo Ballester Valor <gbvalor@gmail.com>
 Hangyu Kuang <hkuang@google.com>
-Hanno Böck <hanno@hboeck.de>
 Henrik Lundin <hlundin@google.com>
 Hui Su <huisu@google.com>
 Ivan Maltz <ivanmaltz@google.com>
-Jacek Caban <cjacek@gmail.com>
-Jacky Chen <jackychen@google.com>
 James Berry <jamesberry@google.com>
-James Yu <james.yu@linaro.org>
 James Zern <jzern@google.com>
-Jan Gerber <j@mailb.org>
 Jan Kratochvil <jan.kratochvil@redhat.com>
 Janne Salonen <jsalonen@google.com>
 Jeff Faust <jfaust@google.com>
 Jeff Muizelaar <jmuizelaar@mozilla.com>
 Jeff Petkau <jpet@chromium.org>
-Jia Jia <jia.jia@linaro.org>
 Jim Bankoski <jimbankoski@google.com>
 Jingning Han <jingning@google.com>
-Joey Parrish <joeyparrish@google.com>
 Johann Koenig <johannkoenig@google.com>
 John Koleszar <jkoleszar@google.com>
-Johnny Klonaris <google@jawknee.com>
-John Stark <jhnstrk@gmail.com>
 Joshua Bleecher Snyder <josh@treelinelabs.com>
 Joshua Litt <joshualitt@google.com>
-Julia Robson <juliamrobson@gmail.com>
 Justin Clift <justin@salasaga.org>
 Justin Lebar <justin.lebar@gmail.com>
 KO Myung-Hun <komh@chollian.net>
-Lawrence Velázquez <larryv@macports.org>
 Lou Quillio <louquillio@google.com>
 Luca Barbato <lu_zero@gentoo.org>
 Makoto Kato <makoto.kt@gmail.com>
@@ -87,48 +65,36 @@ Michael Kohler <michaelkohler@live.com>
 Mike Frysinger <vapier@chromium.org>
 Mike Hommey <mhommey@mozilla.com>
 Mikhal Shemer <mikhal@google.com>
-Minghai Shang <minghai@google.com>
 Morton Jonuschat <yabawock@gmail.com>
-Nico Weber <thakis@chromium.org>
 Parag Salasakar <img.mips1@gmail.com>
 Pascal Massimino <pascal.massimino@gmail.com>
 Patrik Westin <patrik.westin@gmail.com>
 Paul Wilkins <paulwilkins@google.com>
 Pavol Rusnak <stick@gk2.sk>
 Paweł Hajdan <phajdan@google.com>
-Pengchong Jin <pengchong@google.com>
-Peter de Rivaz <peter.derivaz@gmail.com>
 Philip Jägenstedt <philipj@opera.com>
 Priit Laes <plaes@plaes.org>
 Rafael Ávila de Espíndola <rafael.espindola@gmail.com>
 Rafaël Carré <funman@videolan.org>
 Ralph Giles <giles@xiph.org>
 Rob Bradford <rob@linux.intel.com>
-Ronald S. Bultje <rsbultje@gmail.com>
-Rui Ueyama <ruiu@google.com>
+Ronald S. Bultje <rbultje@google.com>
 Sami Pietilä <samipietila@google.com>
 Scott Graham <scottmg@chromium.org>
 Scott LaVarnway <slavarnway@google.com>
-Sean McGovern <gseanmcg@gmail.com>
-Sergey Ulanov <sergeyu@chromium.org>
 Shimon Doodkin <helpmepro1@gmail.com>
-Shunyao Li <shunyaoli@google.com>
 Stefan Holmer <holmer@google.com>
 Suman Sunkara <sunkaras@google.com>
 Taekhyun Kim <takim@nvidia.com>
 Takanori MATSUURA <t.matsuu@gmail.com>
 Tamar Levy <tamar.levy@intel.com>
-Tao Bai <michaelbai@chromium.org>
 Tero Rintaluoma <teror@google.com>
 Thijs Vermeir <thijsvermeir@gmail.com>
-Tim Kopp <tkopp@google.com>
 Timothy B. Terriberry <tterribe@xiph.org>
 Tom Finegan <tomfinegan@google.com>
 Vignesh Venkatasubramanian <vigneshv@google.com>
 Yaowu Xu <yaowu@google.com>
-Yongzhe Wang <yongzhe@google.com>
 Yunqing Wang <yunqingwang@google.com>
-Zoe Liu <zoeliu@google.com>
 Google Inc.
 The Mozilla Foundation
 The Xiph.Org Foundation
--- a/44
+++ b/44
@@ -1,47 +1,3 @@
-Next Release
-  - Incompatible changes:
-    The VP9 encoder's default keyframe interval changed to 128 from 9999.
-
-2015-11-09 v1.5.0 "Javan Whistling Duck"
-  This release improves upon the VP9 encoder and speeds up the encoding and
-  decoding processes.
-
-  - Upgrading:
-    This release is ABI incompatible with 1.4.0. It drops deprecated VP8
-    controls and adds a variety of VP9 controls for testing.
-
-    The vpxenc utility now prefers VP9 by default.
-
-  - Enhancements:
-    Faster VP9 encoding and decoding
-    Smaller library size by combining functions used by VP8 and VP9
-
-  - Bug Fixes:
-    A variety of fuzzing issues
-
-2015-04-03 v1.4.0 "Indian Runner Duck"
-  This release includes significant improvements to the VP9 codec.
-
-  - Upgrading:
-    This release is ABI incompatible with 1.3.0. It drops the compatibility
-    layer, requiring VPX_IMG_FMT_* instead of IMG_FMT_*, and adds several codec
-    controls for VP9.
-
-  - Enhancements:
-    Faster VP9 encoding and decoding
-    Multithreaded VP9 decoding (tile and frame-based)
-    Multithreaded VP9 encoding - on by default
-    YUV 4:2:2 and 4:4:4 support in VP9
-    10 and 12bit support in VP9
-    64bit ARM support by replacing ARM assembly with intrinsics
-
-  - Bug Fixes:
-    Fixes a VP9 bitstream issue in Profile 1. This only affected non-YUV 4:2:0
-    files.
-
-  - Known Issues:
-    Frame Parallel decoding fails for segmented and non-420 files.
-
 2013-11-15 v1.3.0 "Forest"
  This release introduces the VP9 codec in a backward-compatible way.
  All existing users of VP8 can continue to use the library without
--- a/2
+++ b/2
@@ -17,7 +17,7 @@ or agree to the institution of patent litigation or any other patent
 enforcement activity against any entity (including a cross-claim or
 counterclaim in a lawsuit) alleging that any of these implementations of WebM
 or any code incorporated within any of these implementations of WebM
-constitute direct or contributory patent infringement, or inducement of
+constitutes direct or contributory patent infringement, or inducement of
 patent infringement, then any patent rights granted to you under this License
 for these implementations of WebM shall terminate as of the date such
 litigation is filed.
--- a/26
+++ b/26
@@ -1,4 +1,4 @@
-README - 23 March 2015
+README - 30 May 2014

 Welcome to the WebM VP8/VP9 Codec SDK!

@@ -47,6 +47,7 @@ COMPILING THE APPLICATIONS/LIBRARIES:
  --help output of the configure script. As of this writing, the list of
  available targets is:

+    armv6-darwin-gcc
    armv6-linux-rvct
    armv6-linux-gcc
    armv6-none-rvct
@@ -58,10 +59,15 @@ COMPILING THE APPLICATIONS/LIBRARIES:
    armv7-none-rvct
    armv7-win32-vs11
    armv7-win32-vs12
-    armv7-win32-vs14
    armv7s-darwin-gcc
    mips32-linux-gcc
    mips64-linux-gcc
+    ppc32-darwin8-gcc
+    ppc32-darwin9-gcc
+    ppc32-linux-gcc
+    ppc64-darwin8-gcc
+    ppc64-darwin9-gcc
+    ppc64-linux-gcc
    sparc-solaris-gcc
    x86-android-gcc
    x86-darwin8-gcc
@@ -72,33 +78,39 @@ COMPILING THE APPLICATIONS/LIBRARIES:
    x86-darwin11-gcc
    x86-darwin12-gcc
    x86-darwin13-gcc
-    x86-darwin14-gcc
    x86-iphonesimulator-gcc
    x86-linux-gcc
    x86-linux-icc
    x86-os2-gcc
    x86-solaris-gcc
    x86-win32-gcc
+    x86-win32-vs7
+    x86-win32-vs8
+    x86-win32-vs9
    x86-win32-vs10
    x86-win32-vs11
    x86-win32-vs12
-    x86-win32-vs14
-    x86_64-android-gcc
    x86_64-darwin9-gcc
    x86_64-darwin10-gcc
    x86_64-darwin11-gcc
    x86_64-darwin12-gcc
    x86_64-darwin13-gcc
-    x86_64-darwin14-gcc
    x86_64-iphonesimulator-gcc
    x86_64-linux-gcc
    x86_64-linux-icc
    x86_64-solaris-gcc
    x86_64-win64-gcc
+    x86_64-win64-vs8
+    x86_64-win64-vs9
    x86_64-win64-vs10
    x86_64-win64-vs11
    x86_64-win64-vs12
-    x86_64-win64-vs14
+    universal-darwin8-gcc
+    universal-darwin9-gcc
+    universal-darwin10-gcc
+    universal-darwin11-gcc
+    universal-darwin12-gcc
+    universal-darwin13-gcc
    generic-gnu

  The generic-gnu target, in conjunction with the CROSS environment variable,
--- a/args.c
+++ b/args.c
@@ -14,7 +14,9 @@
 #include <limits.h>
 #include "args.h"

-#include "vpx_ports/msvc.h"
+#ifdef _MSC_VER
+#define snprintf _snprintf
+#endif

 #if defined(__GNUC__) && __GNUC__
 extern void die(const char *fmt, ...) __attribute__((noreturn));
--- a/build/make/Android.mk
+++ b/build/make/Android.mk
@@ -67,8 +67,6 @@ else ifeq  ($(TARGET_ARCH_ABI),arm64-v8a)
  LOCAL_ARM_MODE := arm
 else ifeq ($(TARGET_ARCH_ABI),x86)
  include $(CONFIG_DIR)libs-x86-android-gcc.mk
-else ifeq ($(TARGET_ARCH_ABI),x86_64)
-  include $(CONFIG_DIR)libs-x86_64-android-gcc.mk
 else ifeq ($(TARGET_ARCH_ABI),mips)
  include $(CONFIG_DIR)libs-mips-android-gcc.mk
 else
@@ -160,26 +158,24 @@ LOCAL_CFLAGS += \

 LOCAL_MODULE := libvpx

+LOCAL_LDLIBS := -llog
+
 ifeq ($(CONFIG_RUNTIME_CPU_DETECT),yes)
  LOCAL_STATIC_LIBRARIES := cpufeatures
 endif

 # Add a dependency to force generation of the RTCD files.
-define rtcd_dep_template
-rtcd_dep_template_SRCS := $(addprefix $(LOCAL_PATH)/, $(LOCAL_SRC_FILES))
-rtcd_dep_template_SRCS := $$(rtcd_dep_template_SRCS:.neon=)
-ifeq ($(CONFIG_VP10), yes)
-$$(rtcd_dep_template_SRCS): vp10_rtcd.h
+ifeq ($(CONFIG_VP8), yes)
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vp8_rtcd.h
 endif
-$$(rtcd_dep_template_SRCS): vpx_scale_rtcd.h
-$$(rtcd_dep_template_SRCS): vpx_dsp_rtcd.h
-
-ifneq ($(findstring $(TARGET_ARCH_ABI),x86 x86_64),)
-$$(rtcd_dep_template_SRCS): vpx_config.asm
+ifeq ($(CONFIG_VP9), yes)
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vp9_rtcd.h
 endif
-endef
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vpx_scale_rtcd.h

-$(eval $(call rtcd_dep_template))
+ifeq ($(TARGET_ARCH_ABI),x86)
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vpx_config.asm
+endif

 .PHONY: clean
 clean:
@@ -188,11 +184,7 @@ clean:
 	@$(RM) -r $(ASM_CNV_PATH)
 	@$(RM) $(CLEAN-OBJS)

-ifeq ($(ENABLE_SHARED),1)
-  include $(BUILD_SHARED_LIBRARY)
-else
-  include $(BUILD_STATIC_LIBRARY)
-endif
+include $(BUILD_SHARED_LIBRARY)

 ifeq ($(CONFIG_RUNTIME_CPU_DETECT),yes)
 $(call import-module,cpufeatures)
--- a/build/make/Makefile
+++ b/build/make/Makefile
@@ -22,11 +22,9 @@ clean:: .DEFAULT
 exampletest: .DEFAULT
 install:: .DEFAULT
 test:: .DEFAULT
-test-no-data-check:: .DEFAULT
 testdata:: .DEFAULT
 utiltest: .DEFAULT
-exampletest-no-data-check utiltest-no-data-check: .DEFAULT
-test_%: .DEFAULT ;
+

 # Note: md5sum is not installed on OS X, but openssl is. Openssl may not be
 # installed on cygwin, so we need to autodetect here.
@@ -58,10 +56,13 @@ dist:
        fi
 endif

-# Since we invoke make recursively for multiple targets we need to include the
-# .mk file for the correct target, but only when $(target) is non-empty.
 ifneq ($(target),)
-include $(target)-$(TOOLCHAIN).mk
+# Normally, we want to build the filename from the target and the toolchain.
+# This disambiguates from the $(target).mk file that exists in the source tree.
+# However, the toolchain is part of the target in universal builds, so we
+# don't want to include TOOLCHAIN in that case. FAT_ARCHS is used to test
+# if we're in the universal case.
+include $(target)$(if $(FAT_ARCHS),,-$(TOOLCHAIN)).mk
 endif
 BUILD_ROOT?=.
 VPATH=$(SRC_PATH_BARE)
@@ -115,29 +116,28 @@ test::
 testdata::
 .PHONY: utiltest
 utiltest:
-.PHONY: test-no-data-check exampletest-no-data-check utiltest-no-data-check
-test-no-data-check::
-exampletest-no-data-check utiltest-no-data-check:

-# Force to realign stack always on OS/2
+# Add compiler flags for intrinsic files
 ifeq ($(TOOLCHAIN), x86-os2-gcc)
-CFLAGS += -mstackrealign
+STACKREALIGN=-mstackrealign
+else
+STACKREALIGN=
 endif

 $(BUILD_PFX)%_mmx.c.d: CFLAGS += -mmmx
 $(BUILD_PFX)%_mmx.c.o: CFLAGS += -mmmx
-$(BUILD_PFX)%_sse2.c.d: CFLAGS += -msse2
-$(BUILD_PFX)%_sse2.c.o: CFLAGS += -msse2
-$(BUILD_PFX)%_sse3.c.d: CFLAGS += -msse3
-$(BUILD_PFX)%_sse3.c.o: CFLAGS += -msse3
-$(BUILD_PFX)%_ssse3.c.d: CFLAGS += -mssse3
-$(BUILD_PFX)%_ssse3.c.o: CFLAGS += -mssse3
-$(BUILD_PFX)%_sse4.c.d: CFLAGS += -msse4.1
-$(BUILD_PFX)%_sse4.c.o: CFLAGS += -msse4.1
-$(BUILD_PFX)%_avx.c.d: CFLAGS += -mavx
-$(BUILD_PFX)%_avx.c.o: CFLAGS += -mavx
-$(BUILD_PFX)%_avx2.c.d: CFLAGS += -mavx2
-$(BUILD_PFX)%_avx2.c.o: CFLAGS += -mavx2
+$(BUILD_PFX)%_sse2.c.d: CFLAGS += -msse2 $(STACKREALIGN)
+$(BUILD_PFX)%_sse2.c.o: CFLAGS += -msse2 $(STACKREALIGN)
+$(BUILD_PFX)%_sse3.c.d: CFLAGS += -msse3 $(STACKREALIGN)
+$(BUILD_PFX)%_sse3.c.o: CFLAGS += -msse3 $(STACKREALIGN)
+$(BUILD_PFX)%_ssse3.c.d: CFLAGS += -mssse3 $(STACKREALIGN)
+$(BUILD_PFX)%_ssse3.c.o: CFLAGS += -mssse3 $(STACKREALIGN)
+$(BUILD_PFX)%_sse4.c.d: CFLAGS += -msse4.1 $(STACKREALIGN)
+$(BUILD_PFX)%_sse4.c.o: CFLAGS += -msse4.1 $(STACKREALIGN)
+$(BUILD_PFX)%_avx.c.d: CFLAGS += -mavx $(STACKREALIGN)
+$(BUILD_PFX)%_avx.c.o: CFLAGS += -mavx $(STACKREALIGN)
+$(BUILD_PFX)%_avx2.c.d: CFLAGS += -mavx2 $(STACKREALIGN)
+$(BUILD_PFX)%_avx2.c.o: CFLAGS += -mavx2 $(STACKREALIGN)

 $(BUILD_PFX)%.c.d: %.c
 	$(if $(quiet),@echo "    [DEP] $@")
@@ -283,7 +283,7 @@ define archive_template
 # for creating them.
 $(1):
 	$(if $(quiet),@echo "    [AR] $$@")
-	$(qexec)$$(AR) $$(ARFLAGS) $$@ $$^
+	$(qexec)$$(AR) $$(ARFLAGS) $$@ $$?
 endef

 define so_template
@@ -313,15 +313,18 @@ $(1):
        $$(filter %.o,$$^) $$(extralibs)
 endef

-define dll_template
-# Not using a pattern rule here because we don't want to generate empty
-# archives when they are listed as a dependency in files not responsible
-# for creating them.
-$(1):
-	$(if $(quiet),@echo "    [LD] $$@")
-	$(qexec)$$(LD) -Zdll $$(LDFLAGS) \
-        -o $$@ \
-        $$(filter %.o,$$^) $$(extralibs) $$(EXPORTS_FILE)
+
+
+define lipo_lib_template
+$(1): $(addsuffix /$(1),$(FAT_ARCHS))
+	$(if $(quiet),@echo "    [LIPO] $$@")
+	$(qexec)libtool -static -o $$@ $$?
+endef
+
+define lipo_bin_template
+$(1): $(addsuffix /$(1),$(FAT_ARCHS))
+	$(if $(quiet),@echo "    [LIPO] $$@")
+	$(qexec)lipo -output $$@ -create $$?
 endef


@@ -380,9 +383,8 @@ LIBS=$(call enabled,LIBS)
 .libs: $(LIBS)
 	@touch $@
 $(foreach lib,$(filter %_g.a,$(LIBS)),$(eval $(call archive_template,$(lib))))
-$(foreach lib,$(filter %so.$(SO_VERSION_MAJOR).$(SO_VERSION_MINOR).$(SO_VERSION_PATCH),$(LIBS)),$(eval $(call so_template,$(lib))))
-$(foreach lib,$(filter %$(SO_VERSION_MAJOR).dylib,$(LIBS)),$(eval $(call dl_template,$(lib))))
-$(foreach lib,$(filter %$(SO_VERSION_MAJOR).dll,$(LIBS)),$(eval $(call dll_template,$(lib))))
+$(foreach lib,$(filter %so.$(VERSION_MAJOR).$(VERSION_MINOR).$(VERSION_PATCH),$(LIBS)),$(eval $(call so_template,$(lib))))
+$(foreach lib,$(filter %$(VERSION_MAJOR).dylib,$(LIBS)),$(eval $(call dl_template,$(lib))))

 INSTALL-LIBS=$(call cond_enabled,CONFIG_INSTALL_LIBS,INSTALL-LIBS)
 ifeq ($(MAKECMDGOALS),dist)
@@ -418,6 +420,7 @@ ifneq ($(call enabled,DIST-SRCS),)
    DIST-SRCS-yes            += build/make/gen_asm_deps.sh
    DIST-SRCS-yes            += build/make/Makefile
    DIST-SRCS-$(CONFIG_MSVS)  += build/make/gen_msvs_def.sh
+    DIST-SRCS-$(CONFIG_MSVS)  += build/make/gen_msvs_proj.sh
    DIST-SRCS-$(CONFIG_MSVS)  += build/make/gen_msvs_sln.sh
    DIST-SRCS-$(CONFIG_MSVS)  += build/make/gen_msvs_vcxproj.sh
    DIST-SRCS-$(CONFIG_MSVS)  += build/make/msvs_common.sh
@@ -448,5 +451,3 @@ all: $(BUILD_TARGETS)
 install:: $(INSTALL_TARGETS)
 dist: $(INSTALL_TARGETS)
 test::
-
-.SUFFIXES:  # Delete default suffix rules
--- a/build/make/configure.sh
+++ b/build/make/configure.sh
@@ -73,7 +73,6 @@ Build options:
  --target=TARGET             target platform tuple [generic-gnu]
  --cpu=CPU                   optimize for a specific cpu rather than a family
  --extra-cflags=ECFLAGS      add ECFLAGS to CFLAGS [$CFLAGS]
-  --extra-cxxflags=ECXXFLAGS  add ECXXFLAGS to CXXFLAGS [$CXXFLAGS]
  ${toggle_extra_warnings}    emit harmless warnings (always non-fatal)
  ${toggle_werror}            treat warnings as errors, if possible
                              (not available with all compilers)
@@ -185,7 +184,6 @@ add_extralibs() {
 #
 # Boolean Manipulation Functions
 #
-
 enable_feature(){
  set_all yes $*
 }
@@ -202,24 +200,6 @@ disabled(){
  eval test "x\$$1" = "xno"
 }

-enable_codec(){
-  enabled "${1}" || echo "  enabling ${1}"
-  enable_feature "${1}"
-
-  is_in "${1}" vp8 vp9 vp10 && enable_feature "${1}_encoder" "${1}_decoder"
-}
-
-disable_codec(){
-  disabled "${1}" || echo "  disabling ${1}"
-  disable_feature "${1}"
-
-  is_in "${1}" vp8 vp9 vp10 && disable_feature "${1}_encoder" "${1}_decoder"
-}
-
-# Iterates through positional parameters, checks to confirm the parameter has
-# not been explicitly (force) disabled, and enables the setting controlled by
-# the parameter when the setting is not disabled.
-# Note: Does NOT alter RTCD generation options ($RTCD_OPTIONS).
 soft_enable() {
  for var in $*; do
    if ! disabled $var; then
@@ -229,10 +209,6 @@ soft_enable() {
  done
 }

-# Iterates through positional parameters, checks to confirm the parameter has
-# not been explicitly (force) enabled, and disables the setting controlled by
-# the parameter when the setting is not enabled.
-# Note: Does NOT alter RTCD generation options ($RTCD_OPTIONS).
 soft_disable() {
  for var in $*; do
    if ! enabled $var; then
@@ -361,10 +337,6 @@ check_add_cflags() {
  check_cflags "$@" && add_cflags_only "$@"
 }

-check_add_cxxflags() {
-  check_cxxflags "$@" && add_cxxflags_only "$@"
-}
-
 check_add_asflags() {
  log add_asflags "$@"
  add_asflags "$@"
@@ -418,7 +390,7 @@ write_common_config_banner() {
 write_common_config_targets() {
  for t in ${all_targets}; do
    if enabled ${t}; then
-      if enabled child; then
+      if enabled universal || enabled child; then
        fwrite config.mk "ALL_TARGETS += ${t}-${toolchain}"
      else
        fwrite config.mk "ALL_TARGETS += ${t}"
@@ -456,7 +428,7 @@ NM=${NM}

 CFLAGS  = ${CFLAGS}
 CXXFLAGS  = ${CXXFLAGS}
-ARFLAGS = -crs\$(if \$(quiet),,v)
+ARFLAGS = -rus\$(if \$(quiet),c,v)
 LDFLAGS = ${LDFLAGS}
 ASFLAGS = ${ASFLAGS}
 extralibs = ${extralibs}
@@ -531,25 +503,24 @@ process_common_cmdline() {
      --extra-cflags=*)
        extra_cflags="${optval}"
        ;;
-      --extra-cxxflags=*)
-        extra_cxxflags="${optval}"
-        ;;
      --enable-?*|--disable-?*)
        eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'`
-        if is_in ${option} ${ARCH_EXT_LIST}; then
+        if echo "${ARCH_EXT_LIST}" | grep "^ *$option\$" >/dev/null; then
          [ $action = "disable" ] && RTCD_OPTIONS="${RTCD_OPTIONS}--disable-${option} "
        elif [ $action = "disable" ] && ! disabled $option ; then
-          is_in ${option} ${CMDLINE_SELECT} || die_unknown $opt
+          echo "${CMDLINE_SELECT}" | grep "^ *$option\$" >/dev/null ||
+            die_unknown $opt
          log_echo "  disabling $option"
        elif [ $action = "enable" ] && ! enabled $option ; then
-          is_in ${option} ${CMDLINE_SELECT} || die_unknown $opt
+          echo "${CMDLINE_SELECT}" | grep "^ *$option\$" >/dev/null ||
+            die_unknown $opt
          log_echo "  enabling $option"
        fi
        ${action}_feature $option
        ;;
      --require-?*)
        eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'`
-        if is_in ${option} ${ARCH_EXT_LIST}; then
+        if echo "${ARCH_EXT_LIST}" none | grep "^ *$option\$" >/dev/null; then
            RTCD_OPTIONS="${RTCD_OPTIONS}${opt} "
        else
            die_unknown $opt
@@ -646,44 +617,16 @@ show_darwin_sdk_path() {
    xcodebuild -sdk $1 -version Path 2>/dev/null
 }

-# Print the major version number of the Darwin SDK specified by $1.
-show_darwin_sdk_major_version() {
-  xcrun --sdk $1 --show-sdk-version 2>/dev/null | cut -d. -f1
-}
-
-# Print the Xcode version.
-show_xcode_version() {
-  xcodebuild -version | head -n1 | cut -d' ' -f2
-}
-
-# Fails when Xcode version is less than 6.3.
-check_xcode_minimum_version() {
-  xcode_major=$(show_xcode_version | cut -f1 -d.)
-  xcode_minor=$(show_xcode_version | cut -f2 -d.)
-  xcode_min_major=6
-  xcode_min_minor=3
-  if [ ${xcode_major} -lt ${xcode_min_major} ]; then
-    return 1
-  fi
-  if [ ${xcode_major} -eq ${xcode_min_major} ] \
-    && [ ${xcode_minor} -lt ${xcode_min_minor} ]; then
-    return 1
-  fi
-}
-
 process_common_toolchain() {
  if [ -z "$toolchain" ]; then
    gcctarget="${CHOST:-$(gcc -dumpmachine 2> /dev/null)}"

    # detect tgt_isa
    case "$gcctarget" in
-      aarch64*)
-        tgt_isa=arm64
-        ;;
      armv6*)
        tgt_isa=armv6
        ;;
-      armv7*-hardfloat* | armv7*-gnueabihf | arm-*-gnueabihf)
+      armv7*-hardfloat*)
        tgt_isa=armv7
        float_abi=hard
        ;;
@@ -697,6 +640,12 @@ process_common_toolchain() {
      *i[3456]86*)
        tgt_isa=x86
        ;;
+      *powerpc64*)
+        tgt_isa=ppc64
+        ;;
+      *powerpc*)
+        tgt_isa=ppc32
+        ;;
      *sparc*)
        tgt_isa=sparc
        ;;
@@ -704,6 +653,14 @@ process_common_toolchain() {

    # detect tgt_os
    case "$gcctarget" in
+      *darwin8*)
+        tgt_isa=universal
+        tgt_os=darwin8
+        ;;
+      *darwin9*)
+        tgt_isa=universal
+        tgt_os=darwin9
+        ;;
      *darwin10*)
        tgt_isa=x86_64
        tgt_os=darwin10
@@ -724,10 +681,6 @@ process_common_toolchain() {
        tgt_isa=x86_64
        tgt_os=darwin14
        ;;
-      *darwin15*)
-        tgt_isa=x86_64
-        tgt_os=darwin15
-        ;;
      x86_64*mingw32*)
        tgt_os=win64
        ;;
@@ -784,27 +737,12 @@ process_common_toolchain() {
  enabled shared && soft_enable pic

  # Minimum iOS version for all target platforms (darwin and iphonesimulator).
-  # Shared library framework builds are only possible on iOS 8 and later.
-  if enabled shared; then
-    IOS_VERSION_OPTIONS="--enable-shared"
-    IOS_VERSION_MIN="8.0"
-  else
-    IOS_VERSION_OPTIONS=""
-    IOS_VERSION_MIN="6.0"
-  fi
+  IOS_VERSION_MIN="6.0"

  # Handle darwin variants. Newer SDKs allow targeting older
  # platforms, so use the newest one available.
  case ${toolchain} in
-    arm*-darwin*)
-      add_cflags "-miphoneos-version-min=${IOS_VERSION_MIN}"
-      iphoneos_sdk_dir="$(show_darwin_sdk_path iphoneos)"
-      if [ -d "${iphoneos_sdk_dir}" ]; then
-        add_cflags  "-isysroot ${iphoneos_sdk_dir}"
-        add_ldflags "-isysroot ${iphoneos_sdk_dir}"
-      fi
-      ;;
-    x86*-darwin*)
+    *-darwin*)
      osx_sdk_dir="$(show_darwin_sdk_path macosx)"
      if [ -d "${osx_sdk_dir}" ]; then
        add_cflags  "-isysroot ${osx_sdk_dir}"
@@ -842,10 +780,6 @@ process_common_toolchain() {
      add_cflags  "-mmacosx-version-min=10.10"
      add_ldflags "-mmacosx-version-min=10.10"
      ;;
-    *-darwin15-*)
-      add_cflags  "-mmacosx-version-min=10.11"
-      add_ldflags "-mmacosx-version-min=10.11"
-      ;;
    *-iphonesimulator-*)
      add_cflags  "-miphoneos-version-min=${IOS_VERSION_MIN}"
      add_ldflags "-miphoneos-version-min=${IOS_VERSION_MIN}"
@@ -861,6 +795,7 @@ process_common_toolchain() {
  case ${toolchain} in
    sparc-solaris-*)
      add_extralibs -lposix4
+      disable_feature fast_unaligned
      ;;
    *-solaris-*)
      add_extralibs -lposix4
@@ -883,36 +818,12 @@ process_common_toolchain() {
          if disabled neon && enabled neon_asm; then
            die "Disabling neon while keeping neon-asm is not supported"
          fi
-          case ${toolchain} in
-            # Apple iOS SDKs no longer support armv6 as of the version 9
-            # release (coincides with release of Xcode 7). Only enable media
-            # when using earlier SDK releases.
-            *-darwin*)
-              if [ "$(show_darwin_sdk_major_version iphoneos)" -lt 9 ]; then
-                soft_enable media
-              else
-                soft_disable media
-                RTCD_OPTIONS="${RTCD_OPTIONS}--disable-media "
-              fi
-              ;;
-            *)
-              soft_enable media
-              ;;
-          esac
+          soft_enable media
+          soft_enable fast_unaligned
          ;;
        armv6)
-          case ${toolchain} in
-            *-darwin*)
-              if [ "$(show_darwin_sdk_major_version iphoneos)" -lt 9 ]; then
-                soft_enable media
-              else
-                die "Your iOS SDK does not support armv6."
-              fi
-              ;;
-            *)
-              soft_enable media
-              ;;
-          esac
+          soft_enable media
+          soft_enable fast_unaligned
          ;;
      esac

@@ -920,6 +831,7 @@ process_common_toolchain() {

      case ${tgt_cc} in
        gcc)
+          CROSS=${CROSS:-arm-none-linux-gnueabi-}
          link_with_cc=gcc
          setup_gnu_toolchain
          arch_int=${tgt_isa##armv}
@@ -941,9 +853,6 @@ EOF
              check_add_cflags -mfpu=neon #-ftree-vectorize
              check_add_asflags -mfpu=neon
            fi
-          elif [ ${tgt_isa} = "arm64" ] || [ ${tgt_isa} = "armv8" ]; then
-            check_add_cflags -march=armv8-a
-            check_add_asflags -march=armv8-a
          else
            check_add_cflags -march=${tgt_isa}
            check_add_asflags -march=${tgt_isa}
@@ -1011,10 +920,6 @@ EOF
          ;;

        android*)
-          if [ -z "${sdk_path}" ]; then
-            die "Must specify --sdk-path for Android builds."
-          fi
-
          SDK_PATH=${sdk_path}
          COMPILER_LOCATION=`find "${SDK_PATH}" \
                             -name "arm-linux-androideabi-gcc*" -print -quit`
@@ -1036,10 +941,8 @@ EOF
                awk '{ print $1 }' | tail -1`
          fi

-          if [ -d "${alt_libc}" ]; then
-            add_cflags "--sysroot=${alt_libc}"
-            add_ldflags "--sysroot=${alt_libc}"
-          fi
+          add_cflags "--sysroot=${alt_libc}"
+          add_ldflags "--sysroot=${alt_libc}"

          # linker flag that routes around a CPU bug in some
          # Cortex-A8 implementations (NDK Dev Guide)
@@ -1065,7 +968,18 @@ EOF
          NM="$(${XCRUN_FIND} nm)"
          RANLIB="$(${XCRUN_FIND} ranlib)"
          AS_SFX=.s
-          LD="${CXX:-$(${XCRUN_FIND} ld)}"
+
+          # Special handling of ld for armv6 because libclang_rt.ios.a does
+          # not contain armv6 support in Apple's clang package:
+          #   Apple LLVM version 5.1 (clang-503.0.40) (based on LLVM 3.4svn).
+          # TODO(tomfinegan): Remove this. Our minimum iOS version (6.0)
+          # renders support for armv6 unnecessary because the 3GS and up
+          # support neon.
+          if [ "${tgt_isa}" = "armv6" ]; then
+            LD="$(${XCRUN_FIND} ld)"
+          else
+            LD="${CXX:-$(${XCRUN_FIND} ld)}"
+          fi

          # ASFLAGS is written here instead of using check_add_asflags
          # because we need to overwrite all of ASFLAGS and purge the
@@ -1091,26 +1005,7 @@ EOF
            [ -d "${try_dir}" ] && add_ldflags -L"${try_dir}"
          done

-          case ${tgt_isa} in
-            armv7|armv7s|armv8|arm64)
-              if enabled neon && ! check_xcode_minimum_version; then
-                soft_disable neon
-                log_echo "  neon disabled: upgrade Xcode (need v6.3+)."
-                if enabled neon_asm; then
-                  soft_disable neon_asm
-                  log_echo "  neon_asm disabled: upgrade Xcode (need v6.3+)."
-                fi
-              fi
-              ;;
-          esac
-
          asm_conversion_cmd="${source_path}/build/make/ads2gas_apple.pl"
-
-          if [ "$(show_darwin_sdk_major_version iphoneos)" -gt 8 ]; then
-            check_add_cflags -fembed-bitcode
-            check_add_asflags -fembed-bitcode
-            check_add_ldflags -fembed-bitcode
-          fi
          ;;

        linux*)
@@ -1118,7 +1013,7 @@ EOF
          if enabled rvct; then
            # Check if we have CodeSourcery GCC in PATH. Needed for
            # libraries
-            which arm-none-linux-gnueabi-gcc 2>&- || \
+            hash arm-none-linux-gnueabi-gcc 2>&- || \
              die "Couldn't find CodeSourcery GCC from PATH"

            # Use armcc as a linker to enable translation of
@@ -1144,32 +1039,30 @@ EOF
      tune_cflags="-mtune="
      if enabled dspr2; then
        check_add_cflags -mips32r2 -mdspr2
-      fi
-
-      if enabled runtime_cpu_detect; then
-        disable_feature runtime_cpu_detect
+        disable_feature fast_unaligned
      fi

      if [ -n "${tune_cpu}" ]; then
        case ${tune_cpu} in
          p5600)
-            check_add_cflags -mips32r5 -mload-store-pairs
-            check_add_cflags -msched-weight -mhard-float -mfp64
-            check_add_asflags -mips32r5 -mhard-float -mfp64
-            check_add_ldflags -mfp64
+            add_cflags -mips32r5 -funroll-loops -mload-store-pairs
+            add_cflags -msched-weight -mhard-float
+            add_asflags -mips32r5 -mhard-float
            ;;
-          i6400|p6600)
-            check_add_cflags -mips64r6 -mabi=64 -msched-weight
-            check_add_cflags  -mload-store-pairs -mhard-float -mfp64
-            check_add_asflags -mips64r6 -mabi=64 -mhard-float -mfp64
-            check_add_ldflags -mips64r6 -mabi=64 -mfp64
+          i6400)
+            add_cflags -mips64r6 -mabi=64 -funroll-loops -mload-store-pairs
+            add_cflags -msched-weight -mhard-float
+            add_asflags -mips64r6 -mabi=64 -mhard-float
+            add_ldflags -mips64r6 -mabi=64
            ;;
        esac

        if enabled msa; then
-          add_cflags -mmsa
-          add_asflags -mmsa
-          add_ldflags -mmsa
+          add_cflags -mmsa -mfp64 -flax-vector-conversions
+          add_asflags -mmsa -mfp64 -flax-vector-conversions
+          add_ldflags -mmsa -mfp64 -flax-vector-conversions
+
+          disable_feature fast_unaligned
        fi
      fi

@@ -1177,6 +1070,29 @@ EOF
      check_add_asflags -march=${tgt_isa}
      check_add_asflags -KPIC
      ;;
+    ppc*)
+      enable_feature ppc
+      bits=${tgt_isa##ppc}
+      link_with_cc=gcc
+      setup_gnu_toolchain
+      add_asflags -force_cpusubtype_ALL -I"\$(dir \$<)darwin"
+      soft_enable altivec
+      enabled altivec && add_cflags -maltivec
+
+      case "$tgt_os" in
+        linux*)
+          add_asflags -maltivec -mregnames -I"\$(dir \$<)linux"
+          ;;
+        darwin*)
+          darwin_arch="-arch ppc"
+          enabled ppc64 && darwin_arch="${darwin_arch}64"
+          add_cflags  ${darwin_arch} -m${bits} -fasm-blocks
+          add_asflags ${darwin_arch} -force_cpusubtype_ALL -I"\$(dir \$<)darwin"
+          add_ldflags ${darwin_arch} -m${bits}
+          enabled altivec && add_cflags -faltivec
+          ;;
+      esac
+      ;;
    x86*)
      case  ${tgt_os} in
        win*)
@@ -1186,12 +1102,10 @@ EOF
          CC=${CC:-${CROSS}gcc}
          CXX=${CXX:-${CROSS}g++}
          LD=${LD:-${CROSS}gcc}
-          CROSS=${CROSS-g}
+          CROSS=${CROSS:-g}
          ;;
        os2)
-          disable_feature pic
          AS=${AS:-nasm}
-          add_ldflags -Zhigh-mem
          ;;
      esac

@@ -1239,12 +1153,6 @@ EOF
              soft_disable avx2
              ;;
          esac
-          case $vc_version in
-            7|8|9)
-              echo "${tgt_cc} omits stdint.h, disabling webm-io..."
-              soft_disable webm_io
-              ;;
-          esac
          ;;
      esac

@@ -1265,43 +1173,32 @@ EOF
      soft_enable runtime_cpu_detect
      # We can't use 'check_cflags' until the compiler is configured and CC is
      # populated.
-      for ext in ${ARCH_EXT_LIST_X86}; do
-        # disable higher order extensions to simplify asm dependencies
-        if [ "$disable_exts" = "yes" ]; then
-          if ! disabled $ext; then
-            RTCD_OPTIONS="${RTCD_OPTIONS}--disable-${ext} "
-            disable_feature $ext
-          fi
-        elif disabled $ext; then
-          disable_exts="yes"
-        else
-          # use the shortened version for the flag: sse4_1 -> sse4
-          check_gcc_machine_option ${ext%_*} $ext
-        fi
-      done
+      check_gcc_machine_option mmx
+      check_gcc_machine_option sse
+      check_gcc_machine_option sse2
+      check_gcc_machine_option sse3
+      check_gcc_machine_option ssse3
+      check_gcc_machine_option sse4 sse4_1
+      check_gcc_machine_option avx
+      check_gcc_machine_option avx2

-      if enabled external_build; then
-        log_echo "  skipping assembler detection"
-      else
-        case "${AS}" in
-          auto|"")
-            which nasm >/dev/null 2>&1 && AS=nasm
-            which yasm >/dev/null 2>&1 && AS=yasm
-            if [ "${AS}" = nasm ] ; then
-              # Apple ships version 0.98 of nasm through at least Xcode 6. Revisit
-              # this check if they start shipping a compatible version.
-              apple=`nasm -v | grep "Apple"`
-              [ -n "${apple}" ] \
-                && echo "Unsupported version of nasm: ${apple}" \
-                && AS=""
-            fi
-            [ "${AS}" = auto ] || [ -z "${AS}" ] \
-              && die "Neither yasm nor nasm have been found." \
-                     "See the prerequisites section in the README for more info."
-            ;;
-        esac
-        log_echo "  using $AS"
-      fi
+      case "${AS}" in
+        auto|"")
+          which nasm >/dev/null 2>&1 && AS=nasm
+          which yasm >/dev/null 2>&1 && AS=yasm
+          if [ "${AS}" = nasm ] ; then
+            # Apple ships version 0.98 of nasm through at least Xcode 6. Revisit
+            # this check if they start shipping a compatible version.
+            apple=`nasm -v | grep "Apple"`
+            [ -n "${apple}" ] \
+              && echo "Unsupported version of nasm: ${apple}" \
+              && AS=""
+          fi
+          [ "${AS}" = auto ] || [ -z "${AS}" ] \
+            && die "Neither yasm nor nasm have been found"
+          ;;
+      esac
+      log_echo "  using $AS"
      [ "${AS##*/}" = nasm ] && add_asflags -Ox
      AS_SFX=.asm
      case  ${tgt_os} in
@@ -1337,13 +1234,6 @@ EOF
          enabled x86 && sim_arch="-arch i386" || sim_arch="-arch x86_64"
          add_cflags  ${sim_arch}
          add_ldflags ${sim_arch}
-
-          if [ "$(show_darwin_sdk_major_version iphonesimulator)" -gt 8 ]; then
-            # yasm v1.3.0 doesn't know what -fembed-bitcode means, so turning it
-            # on is pointless (unless building a C-only lib). Warn the user, but
-            # do nothing here.
-            log "Warning: Bitcode embed disabled for simulator targets."
-          fi
          ;;
        os2)
          add_asflags -f aout
@@ -1355,7 +1245,7 @@ EOF
          ;;
      esac
      ;;
-    *-gcc|generic-gnu)
+    universal*|*-gcc|generic-gnu)
      link_with_cc=gcc
      enable_feature gcc
      setup_gnu_toolchain
@@ -1396,6 +1286,10 @@ EOF
    fi
  fi

+  if [ "${tgt_isa}" = "x86_64" ] || [ "${tgt_isa}" = "x86" ]; then
+    soft_enable use_x86inc
+  fi
+
  # Position Independent Code (PIC) support, for building relocatable
  # shared objects
  enabled gcc && enabled pic && check_add_cflags -fPIC
@@ -1435,15 +1329,11 @@ EOF
  # only for MIPS platforms
  case ${toolchain} in
    mips*)
-      if enabled big_endian; then
-        if enabled dspr2; then
+      if enabled dspr2; then
+        if enabled big_endian; then
          echo "dspr2 optimizations are available only for little endian platforms"
          disable_feature dspr2
        fi
-        if enabled msa; then
-          echo "msa optimizations are available only for little endian platforms"
-          disable_feature msa
-        fi
      fi
      ;;
  esac
@@ -1453,6 +1343,12 @@ EOF
    add_cflags -D_LARGEFILE_SOURCE
    add_cflags -D_FILE_OFFSET_BITS=64
  fi
+
+  # append any user defined extra cflags
+  if [ -n "${extra_cflags}" ] ; then
+    check_add_cflags ${extra_cflags} || \
+    die "Requested extra CFLAGS '${extra_cflags}' not supported by compiler"
+  fi
 }

 process_toolchain() {
--- a/build/make/gen_msvs_proj.sh
+++ b/build/make/gen_msvs_proj.sh
@@ -0,0 +1,486 @@
+#!/bin/bash
+##
+##  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
+##
+##  Use of this source code is governed by a BSD-style license
+##  that can be found in the LICENSE file in the root of the source
+##  tree. An additional intellectual property rights grant can be found
+##  in the file PATENTS.  All contributing project authors may
+##  be found in the AUTHORS file in the root of the source tree.
+##
+
+self=$0
+self_basename=${self##*/}
+self_dirname=$(dirname "$0")
+
+. "$self_dirname/msvs_common.sh"|| exit 127
+
+show_help() {
+    cat <<EOF
+Usage: ${self_basename} --name=projname [options] file1 [file2 ...]
+
+This script generates a Visual Studio project file from a list of source
+code files.
+
+Options:
+    --help                      Print this message
+    --exe                       Generate a project for building an Application
+    --lib                       Generate a project for creating a static library
+    --dll                       Generate a project for creating a dll
+    --static-crt                Use the static C runtime (/MT)
+    --target=isa-os-cc          Target specifier (required)
+    --out=filename              Write output to a file [stdout]
+    --name=project_name         Name of the project (required)
+    --proj-guid=GUID            GUID to use for the project
+    --module-def=filename       File containing export definitions (for DLLs)
+    --ver=version               Version (7,8,9) of visual studio to generate for
+    --src-path-bare=dir         Path to root of source tree
+    -Ipath/to/include           Additional include directories
+    -DFLAG[=value]              Preprocessor macros to define
+    -Lpath/to/lib               Additional library search paths
+    -llibname                   Library to link against
+EOF
+    exit 1
+}
+
+generate_filter() {
+    local var=$1
+    local name=$2
+    local pats=$3
+    local file_list_sz
+    local i
+    local f
+    local saveIFS="$IFS"
+    local pack
+    echo "generating filter '$name' from ${#file_list[@]} files" >&2
+    IFS=*
+
+    open_tag Filter \
+        Name=$name \
+        Filter=$pats \
+        UniqueIdentifier=`generate_uuid` \
+
+    file_list_sz=${#file_list[@]}
+    for i in ${!file_list[@]}; do
+        f=${file_list[i]}
+        for pat in ${pats//;/$IFS}; do
+            if [ "${f##*.}" == "$pat" ]; then
+                unset file_list[i]
+
+                objf=$(echo ${f%.*}.obj \
+                       | sed -e "s,$src_path_bare,," \
+                             -e 's/^[\./]\+//g' -e 's,[:/ ],_,g')
+                open_tag File RelativePath="$f"
+
+                if [ "$pat" == "asm" ] && $asm_use_custom_step; then
+                    for plat in "${platforms[@]}"; do
+                        for cfg in Debug Release; do
+                            open_tag FileConfiguration \
+                                Name="${cfg}|${plat}" \
+
+                            tag Tool \
+                                Name="VCCustomBuildTool" \
+                                Description="Assembling \$(InputFileName)" \
+                                CommandLine="$(eval echo \$asm_${cfg}_cmdline) -o \$(IntDir)\\$objf" \
+                                Outputs="\$(IntDir)\\$objf" \
+
+                            close_tag FileConfiguration
+                        done
+                    done
+                fi
+                if [ "$pat" == "c" ] || \
+                   [ "$pat" == "cc" ] || [ "$pat" == "cpp" ]; then
+                    for plat in "${platforms[@]}"; do
+                        for cfg in Debug Release; do
+                            open_tag FileConfiguration \
+                                Name="${cfg}|${plat}" \
+
+                            tag Tool \
+                                Name="VCCLCompilerTool" \
+                                ObjectFile="\$(IntDir)\\$objf" \
+
+                            close_tag FileConfiguration
+                        done
+                    done
+                fi
+                close_tag File
+
+                break
+            fi
+        done
+    done
+
+    close_tag Filter
+    IFS="$saveIFS"
+}
+
+# Process command line
+unset target
+for opt in "$@"; do
+    optval="${opt#*=}"
+    case "$opt" in
+        --help|-h) show_help
+        ;;
+        --target=*) target="${optval}"
+        ;;
+        --out=*) outfile="$optval"
+        ;;
+        --name=*) name="${optval}"
+        ;;
+        --proj-guid=*) guid="${optval}"
+        ;;
+        --module-def=*) link_opts="${link_opts} ModuleDefinitionFile=${optval}"
+        ;;
+        --exe) proj_kind="exe"
+        ;;
+        --dll) proj_kind="dll"
+        ;;
+        --lib) proj_kind="lib"
+        ;;
+        --src-path-bare=*)
+            src_path_bare=$(fix_path "$optval")
+            src_path_bare=${src_path_bare%/}
+        ;;
+        --static-crt) use_static_runtime=true
+        ;;
+        --ver=*)
+            vs_ver="$optval"
+            case "$optval" in
+                [789])
+                ;;
+                *) die Unrecognized Visual Studio Version in $opt
+                ;;
+            esac
+        ;;
+        -I*)
+            opt=${opt##-I}
+            opt=$(fix_path "$opt")
+            opt="${opt%/}"
+            incs="${incs}${incs:+;}&quot;${opt}&quot;"
+            yasmincs="${yasmincs} -I&quot;${opt}&quot;"
+        ;;
+        -D*) defines="${defines}${defines:+;}${opt##-D}"
+        ;;
+        -L*) # fudge . to $(OutDir)
+            if [ "${opt##-L}" == "." ]; then
+                libdirs="${libdirs}${libdirs:+;}&quot;\$(OutDir)&quot;"
+            else
+                 # Also try directories for this platform/configuration
+                 opt=${opt##-L}
+                 opt=$(fix_path "$opt")
+                 libdirs="${libdirs}${libdirs:+;}&quot;${opt}&quot;"
+                 libdirs="${libdirs}${libdirs:+;}&quot;${opt}/\$(PlatformName)/\$(ConfigurationName)&quot;"
+                 libdirs="${libdirs}${libdirs:+;}&quot;${opt}/\$(PlatformName)&quot;"
+            fi
+        ;;
+        -l*) libs="${libs}${libs:+ }${opt##-l}.lib"
+        ;;
+        -*) die_unknown $opt
+        ;;
+        *)
+            # The paths in file_list are fixed outside of the loop.
+            file_list[${#file_list[@]}]="$opt"
+            case "$opt" in
+                 *.asm) uses_asm=true
+                 ;;
+            esac
+        ;;
+    esac
+done
+
+# Make one call to fix_path for file_list to improve performance.
+fix_file_list
+
+outfile=${outfile:-/dev/stdout}
+guid=${guid:-`generate_uuid`}
+asm_use_custom_step=false
+uses_asm=${uses_asm:-false}
+case "${vs_ver:-8}" in
+    7) vs_ver_id="7.10"
+       asm_use_custom_step=$uses_asm
+       warn_64bit='Detect64BitPortabilityProblems=true'
+    ;;
+    8) vs_ver_id="8.00"
+       asm_use_custom_step=$uses_asm
+       warn_64bit='Detect64BitPortabilityProblems=true'
+    ;;
+    9) vs_ver_id="9.00"
+       asm_use_custom_step=$uses_asm
+       warn_64bit='Detect64BitPortabilityProblems=false'
+    ;;
+esac
+
+[ -n "$name" ] || die "Project name (--name) must be specified!"
+[ -n "$target" ] || die "Target (--target) must be specified!"
+
+if ${use_static_runtime:-false}; then
+    release_runtime=0
+    debug_runtime=1
+    lib_sfx=mt
+else
+    release_runtime=2
+    debug_runtime=3
+    lib_sfx=md
+fi
+
+# Calculate debug lib names: If a lib ends in ${lib_sfx}.lib, then rename
+# it to ${lib_sfx}d.lib. This precludes linking to release libs from a
+# debug exe, so this may need to be refactored later.
+for lib in ${libs}; do
+    if [ "$lib" != "${lib%${lib_sfx}.lib}" ]; then
+        lib=${lib%.lib}d.lib
+    fi
+    debug_libs="${debug_libs}${debug_libs:+ }${lib}"
+done
+
+
+# List Keyword for this target
+case "$target" in
+    x86*) keyword="ManagedCProj"
+    ;;
+    *) die "Unsupported target $target!"
+esac
+
+# List of all platforms supported for this target
+case "$target" in
+    x86_64*)
+        platforms[0]="x64"
+        asm_Debug_cmdline="yasm -Xvc -g cv8 -f win64 ${yasmincs} &quot;\$(InputPath)&quot;"
+        asm_Release_cmdline="yasm -Xvc -f win64 ${yasmincs} &quot;\$(InputPath)&quot;"
+    ;;
+    x86*)
+        platforms[0]="Win32"
+        asm_Debug_cmdline="yasm -Xvc -g cv8 -f win32 ${yasmincs} &quot;\$(InputPath)&quot;"
+        asm_Release_cmdline="yasm -Xvc -f win32 ${yasmincs} &quot;\$(InputPath)&quot;"
+    ;;
+    *) die "Unsupported target $target!"
+    ;;
+esac
+
+generate_vcproj() {
+    case "$proj_kind" in
+        exe) vs_ConfigurationType=1
+        ;;
+        dll) vs_ConfigurationType=2
+        ;;
+        *)   vs_ConfigurationType=4
+        ;;
+    esac
+
+    echo "<?xml version=\"1.0\" encoding=\"Windows-1252\"?>"
+    open_tag VisualStudioProject \
+        ProjectType="Visual C++" \
+        Version="${vs_ver_id}" \
+        Name="${name}" \
+        ProjectGUID="{${guid}}" \
+        RootNamespace="${name}" \
+        Keyword="${keyword}" \
+
+    open_tag Platforms
+    for plat in "${platforms[@]}"; do
+        tag Platform Name="$plat"
+    done
+    close_tag Platforms
+
+    open_tag Configurations
+    for plat in "${platforms[@]}"; do
+        plat_no_ws=`echo $plat | sed 's/[^A-Za-z0-9_]/_/g'`
+        open_tag Configuration \
+            Name="Debug|$plat" \
+            OutputDirectory="\$(SolutionDir)$plat_no_ws/\$(ConfigurationName)" \
+            IntermediateDirectory="$plat_no_ws/\$(ConfigurationName)/${name}" \
+            ConfigurationType="$vs_ConfigurationType" \
+            CharacterSet="1" \
+
+        case "$target" in
+            x86*)
+                case "$name" in
+                    vpx)
+                        tag Tool \
+                            Name="VCCLCompilerTool" \
+                            Optimization="0" \
+                            AdditionalIncludeDirectories="$incs" \
+                            PreprocessorDefinitions="WIN32;_DEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
+                            RuntimeLibrary="$debug_runtime" \
+                            UsePrecompiledHeader="0" \
+                            WarningLevel="3" \
+                            DebugInformationFormat="2" \
+                            $warn_64bit \
+
+                        $uses_asm && tag Tool Name="YASM"  IncludePaths="$incs" Debug="true"
+                    ;;
+                    *)
+                        tag Tool \
+                            Name="VCCLCompilerTool" \
+                            Optimization="0" \
+                            AdditionalIncludeDirectories="$incs" \
+                            PreprocessorDefinitions="WIN32;_DEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
+                            RuntimeLibrary="$debug_runtime" \
+                            UsePrecompiledHeader="0" \
+                            WarningLevel="3" \
+                            DebugInformationFormat="2" \
+                            $warn_64bit \
+
+                        $uses_asm && tag Tool Name="YASM"  IncludePaths="$incs" Debug="true"
+                    ;;
+                esac
+            ;;
+        esac
+
+        case "$proj_kind" in
+            exe)
+                case "$target" in
+                    x86*)
+                        case "$name" in
+                            *)
+                                tag Tool \
+                                    Name="VCLinkerTool" \
+                                    AdditionalDependencies="$debug_libs \$(NoInherit)" \
+                                    AdditionalLibraryDirectories="$libdirs" \
+                                    GenerateDebugInformation="true" \
+                                    ProgramDatabaseFile="\$(OutDir)/${name}.pdb" \
+                            ;;
+                        esac
+                    ;;
+                 esac
+            ;;
+            lib)
+                case "$target" in
+                    x86*)
+                        tag Tool \
+                            Name="VCLibrarianTool" \
+                            OutputFile="\$(OutDir)/${name}${lib_sfx}d.lib" \
+
+                    ;;
+                esac
+            ;;
+            dll)
+                tag Tool \
+                    Name="VCLinkerTool" \
+                    AdditionalDependencies="\$(NoInherit)" \
+                    LinkIncremental="2" \
+                    GenerateDebugInformation="true" \
+                    AssemblyDebug="1" \
+                    TargetMachine="1" \
+                    $link_opts \
+
+            ;;
+        esac
+
+        close_tag Configuration
+
+        open_tag Configuration \
+            Name="Release|$plat" \
+            OutputDirectory="\$(SolutionDir)$plat_no_ws/\$(ConfigurationName)" \
+            IntermediateDirectory="$plat_no_ws/\$(ConfigurationName)/${name}" \
+            ConfigurationType="$vs_ConfigurationType" \
+            CharacterSet="1" \
+            WholeProgramOptimization="0" \
+
+        case "$target" in
+            x86*)
+                case "$name" in
+                    vpx)
+                        tag Tool \
+                            Name="VCCLCompilerTool" \
+                            Optimization="2" \
+                            FavorSizeorSpeed="1" \
+                            AdditionalIncludeDirectories="$incs" \
+                            PreprocessorDefinitions="WIN32;NDEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
+                            RuntimeLibrary="$release_runtime" \
+                            UsePrecompiledHeader="0" \
+                            WarningLevel="3" \
+                            DebugInformationFormat="0" \
+                            $warn_64bit \
+
+                        $uses_asm && tag Tool Name="YASM"  IncludePaths="$incs"
+                    ;;
+                    *)
+                        tag Tool \
+                            Name="VCCLCompilerTool" \
+                            AdditionalIncludeDirectories="$incs" \
+                            Optimization="2" \
+                            FavorSizeorSpeed="1" \
+                            PreprocessorDefinitions="WIN32;NDEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
+                            RuntimeLibrary="$release_runtime" \
+                            UsePrecompiledHeader="0" \
+                            WarningLevel="3" \
+                            DebugInformationFormat="0" \
+                            $warn_64bit \
+
+                        $uses_asm && tag Tool Name="YASM"  IncludePaths="$incs"
+                    ;;
+                esac
+            ;;
+        esac
+
+        case "$proj_kind" in
+            exe)
+                case "$target" in
+                    x86*)
+                        case "$name" in
+                            *)
+                                tag Tool \
+                                    Name="VCLinkerTool" \
+                                    AdditionalDependencies="$libs \$(NoInherit)" \
+                                    AdditionalLibraryDirectories="$libdirs" \
+
+                            ;;
+                        esac
+                    ;;
+                 esac
+            ;;
+            lib)
+                case "$target" in
+                    x86*)
+                        tag Tool \
+                            Name="VCLibrarianTool" \
+                            OutputFile="\$(OutDir)/${name}${lib_sfx}.lib" \
+
+                    ;;
+                esac
+            ;;
+            dll) # note differences to debug version: LinkIncremental, AssemblyDebug
+                tag Tool \
+                    Name="VCLinkerTool" \
+                    AdditionalDependencies="\$(NoInherit)" \
+                    LinkIncremental="1" \
+                    GenerateDebugInformation="true" \
+                    TargetMachine="1" \
+                    $link_opts \
+
+            ;;
+        esac
+
+        close_tag Configuration
+    done
+    close_tag Configurations
+
+    open_tag Files
+    generate_filter srcs   "Source Files"   "c;cc;cpp;def;odl;idl;hpj;bat;asm;asmx"
+    generate_filter hdrs   "Header Files"   "h;hm;inl;inc;xsd"
+    generate_filter resrcs "Resource Files" "rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav"
+    generate_filter resrcs "Build Files"    "mk"
+    close_tag Files
+
+    tag       Globals
+    close_tag VisualStudioProject
+
+    # This must be done from within the {} subshell
+    echo "Ignored files list (${#file_list[@]} items) is:" >&2
+    for f in "${file_list[@]}"; do
+        echo "    $f" >&2
+    done
+}
+
+generate_vcproj |
+    sed  -e '/"/s;\([^ "]\)/;\1\\;g' > ${outfile}
+
+exit
+<!--
+TODO: Add any files not captured by filters.
+                <File
+                        RelativePath=".\ReadMe.txt"
+                        >
+                </File>
+-->
--- a/build/make/gen_msvs_sln.sh
+++ b/build/make/gen_msvs_sln.sh
@@ -19,13 +19,13 @@ show_help() {
    cat <<EOF
 Usage: ${self_basename} [options] file1 [file2 ...]

-This script generates a Visual Studio solution file from a list of project
+This script generates a Visual Studio 2005 solution file from a list of project
 files.

 Options:
    --help                      Print this message
    --out=outfile               Redirect output to a file
-    --ver=version               Version (7,8,9,10,11,12,14) of visual studio to generate for
+    --ver=version               Version (7,8,9,10,11) of visual studio to generate for
    --target=isa-os-cc          Target specifier
 EOF
    exit 1
@@ -55,11 +55,16 @@ indent_pop() {

 parse_project() {
    local file=$1
-    local name=`grep RootNamespace "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
-    local guid=`grep ProjectGuid "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
+    if [ "$sfx" = "vcproj" ]; then
+        local name=`grep Name "$file" | awk 'BEGIN {FS="\""}{if (NR==1) print $2}'`
+        local guid=`grep ProjectGUID "$file" | awk 'BEGIN {FS="\""}{if (NR==1) print $2}'`
+    else
+        local name=`grep RootNamespace "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
+        local guid=`grep ProjectGuid "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
+    fi

    # save the project GUID to a varaible, normalizing to the basename of the
-    # vcxproj file without the extension
+    # vcproj file without the extension
    local var
    var=${file##*/}
    var=${var%%.${sfx}}
@@ -67,8 +72,13 @@ parse_project() {
    eval "${var}_name=$name"
    eval "${var}_guid=$guid"

-    cur_config_list=`grep -B1 'Label="Configuration"' $file |
-        grep Condition | cut -d\' -f4`
+    if [ "$sfx" = "vcproj" ]; then
+        cur_config_list=`grep -A1 '<Configuration' $file |
+            grep Name | cut -d\" -f2`
+    else
+        cur_config_list=`grep -B1 'Label="Configuration"' $file |
+            grep Condition | cut -d\' -f4`
+    fi
    new_config_list=$(for i in $config_list $cur_config_list; do
        echo $i
    done | sort | uniq)
@@ -93,6 +103,25 @@ process_project() {
    eval "${var}_guid=$guid"

    echo "Project(\"{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}\") = \"$name\", \"$file\", \"$guid\""
+    indent_push
+
+    eval "local deps=\"\${${var}_deps}\""
+    if [ -n "$deps" ] && [ "$sfx" = "vcproj" ]; then
+        echo "${indent}ProjectSection(ProjectDependencies) = postProject"
+        indent_push
+
+        for dep in $deps; do
+            eval "local dep_guid=\${${dep}_guid}"
+            [ -z "${dep_guid}" ] && die "Unknown GUID for $dep (dependency of $var)"
+            echo "${indent}$dep_guid = $dep_guid"
+        done
+
+        indent_pop
+        echo "${indent}EndProjectSection"
+
+    fi
+
+    indent_pop
    echo "EndProject"
 }

@@ -162,7 +191,11 @@ process_makefile() {
    IFS=$'\r'$'\n'
    local TAB=$'\t'
    cat <<EOF
+ifeq (\$(CONFIG_VS_VERSION),7)
+MSBUILD_TOOL := devenv.com
+else
 MSBUILD_TOOL := msbuild.exe
+endif
 found_devenv := \$(shell which \$(MSBUILD_TOOL) >/dev/null 2>&1 && echo yes)
 .nodevenv.once:
 ${TAB}@echo "  * \$(MSBUILD_TOOL) not found in path."
@@ -171,7 +204,7 @@ ${TAB}@echo "  * You will have to build all configurations manually using the"
 ${TAB}@echo "  * Visual Studio IDE. To allow make to build them automatically,"
 ${TAB}@echo "  * add the Common7/IDE directory of your Visual Studio"
 ${TAB}@echo "  * installation to your path, eg:"
-${TAB}@echo "  *   C:\Program Files\Microsoft Visual Studio 10.0\Common7\IDE"
+${TAB}@echo "  *   C:\Program Files\Microsoft Visual Studio 8\Common7\IDE"
 ${TAB}@echo "  * "
 ${TAB}@touch \$@
 CLEAN-OBJS += \$(if \$(found_devenv),,.nodevenv.once)
@@ -188,9 +221,16 @@ clean::
 ${TAB}rm -rf "$platform"/"$config"
 .PHONY: $nows_sln_config
 ifneq (\$(found_devenv),)
+  ifeq (\$(CONFIG_VS_VERSION),7)
+$nows_sln_config: $outfile
+${TAB}\$(MSBUILD_TOOL) $outfile -build "$config"
+
+  else
 $nows_sln_config: $outfile
 ${TAB}\$(MSBUILD_TOOL) $outfile -m -t:Build \\
 ${TAB}${TAB}-p:Configuration="$config" -p:Platform="$platform"
+
+  endif
 else
 $nows_sln_config: $outfile .nodevenv.once
 ${TAB}@echo "  * Skipping build of $sln_config (\$(MSBUILD_TOOL) not in path)."
@@ -215,12 +255,23 @@ for opt in "$@"; do
    ;;
    --ver=*) vs_ver="$optval"
             case $optval in
-             10|11|12|14)
+             [789]|10|11|12)
             ;;
             *) die Unrecognized Visual Studio Version in $opt
             ;;
             esac
    ;;
+    --ver=*) vs_ver="$optval"
+             case $optval in
+             7) sln_vers="8.00"
+                sln_vers_str="Visual Studio .NET 2003"
+             ;;
+             [89])
+             ;;
+             *) die "Unrecognized Visual Studio Version '$optval' in $opt"
+             ;;
+             esac
+    ;;
    --target=*) target="${optval}"
    ;;
    -*) die_unknown $opt
@@ -230,7 +281,16 @@ for opt in "$@"; do
 done
 outfile=${outfile:-/dev/stdout}
 mkoutfile=${mkoutfile:-/dev/stdout}
-case "${vs_ver:-10}" in
+case "${vs_ver:-8}" in
+    7) sln_vers="8.00"
+       sln_vers_str="Visual Studio .NET 2003"
+    ;;
+    8) sln_vers="9.00"
+       sln_vers_str="Visual Studio 2005"
+    ;;
+    9) sln_vers="10.00"
+       sln_vers_str="Visual Studio 2008"
+    ;;
    10) sln_vers="11.00"
       sln_vers_str="Visual Studio 2010"
    ;;
@@ -240,11 +300,15 @@ case "${vs_ver:-10}" in
    12) sln_vers="12.00"
       sln_vers_str="Visual Studio 2013"
    ;;
-    14) sln_vers="14.00"
-       sln_vers_str="Visual Studio 2015"
+esac
+case "${vs_ver:-8}" in
+    [789])
+    sfx=vcproj
+    ;;
+    10|11|12)
+    sfx=vcxproj
    ;;
 esac
-sfx=vcxproj

 for f in "${file_list[@]}"; do
    parse_project $f
--- a/build/make/gen_msvs_vcxproj.sh
+++ b/build/make/gen_msvs_vcxproj.sh
@@ -34,7 +34,7 @@ Options:
    --name=project_name         Name of the project (required)
    --proj-guid=GUID            GUID to use for the project
    --module-def=filename       File containing export definitions (for DLLs)
-    --ver=version               Version (10,11,12,14) of visual studio to generate for
+    --ver=version               Version (10,11,12) of visual studio to generate for
    --src-path-bare=dir         Path to root of source tree
    -Ipath/to/include           Additional include directories
    -DFLAG[=value]              Preprocessor macros to define
@@ -168,7 +168,7 @@ for opt in "$@"; do
        --ver=*)
            vs_ver="$optval"
            case "$optval" in
-                10|11|12|14)
+                10|11|12)
                ;;
                *) die Unrecognized Visual Studio Version in $opt
                ;;
@@ -211,14 +211,14 @@ for opt in "$@"; do
 done

 # Make one call to fix_path for file_list to improve performance.
-fix_file_list file_list
+fix_file_list

 outfile=${outfile:-/dev/stdout}
 guid=${guid:-`generate_uuid`}
 asm_use_custom_step=false
 uses_asm=${uses_asm:-false}
 case "${vs_ver:-11}" in
-    10|11|12|14)
+    10|11|12)
       asm_use_custom_step=$uses_asm
    ;;
 esac
@@ -263,8 +263,8 @@ case "$target" in
    ;;
    arm*)
        platforms[0]="ARM"
-        asm_Debug_cmdline="armasm -nologo -oldit &quot;%(FullPath)&quot;"
-        asm_Release_cmdline="armasm -nologo -oldit &quot;%(FullPath)&quot;"
+        asm_Debug_cmdline="armasm -nologo &quot;%(FullPath)&quot;"
+        asm_Release_cmdline="armasm -nologo &quot;%(FullPath)&quot;"
    ;;
    *) die "Unsupported target $target!"
    ;;
@@ -344,9 +344,6 @@ generate_vcxproj() {
                # has to enable AppContainerApplication as well.
                tag_content PlatformToolset v120
            fi
-            if [ "$vs_ver" = "14" ]; then
-                tag_content PlatformToolset v140
-            fi
            tag_content CharacterSet Unicode
            if [ "$config" = "Release" ]; then
                tag_content WholeProgramOptimization true
--- a/build/make/ios-Info.plist
+++ b/build/make/ios-Info.plist
@@ -1,37 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
-<plist version="1.0">
-<dict>
-	<key>CFBundleDevelopmentRegion</key>
-	<string>en</string>
-	<key>CFBundleExecutable</key>
-	<string>VPX</string>
-	<key>CFBundleIdentifier</key>
-	<string>org.webmproject.VPX</string>
-	<key>CFBundleInfoDictionaryVersion</key>
-	<string>6.0</string>
-	<key>CFBundleName</key>
-	<string>VPX</string>
-	<key>CFBundlePackageType</key>
-	<string>FMWK</string>
-	<key>CFBundleShortVersionString</key>
-	<string>${VERSION}</string>
-	<key>CFBundleSignature</key>
-	<string>????</string>
-	<key>CFBundleSupportedPlatforms</key>
-	<array>
-		<string>iPhoneOS</string>
-	</array>
-	<key>CFBundleVersion</key>
-	<string>${VERSION}</string>
-	<key>MinimumOSVersion</key>
-	<string>${IOS_VERSION_MIN}</string>
-	<key>UIDeviceFamily</key>
-	<array>
-		<integer>1</integer>
-		<integer>2</integer>
-	</array>
-	<key>VPXFullVersion</key>
-	<string>${FULLVERSION}</string>
-</dict>
-</plist>
--- a/build/make/iosbuild.sh
+++ b/build/make/iosbuild.sh
@@ -24,44 +24,32 @@ CONFIGURE_ARGS="--disable-docs
                --disable-unit-tests"
 DIST_DIR="_dist"
 FRAMEWORK_DIR="VPX.framework"
-FRAMEWORK_LIB="VPX.framework/VPX"
 HEADER_DIR="${FRAMEWORK_DIR}/Headers/vpx"
+MAKE_JOBS=1
 SCRIPT_DIR=$(dirname "$0")
 LIBVPX_SOURCE_DIR=$(cd ${SCRIPT_DIR}/../..; pwd)
 LIPO=$(xcrun -sdk iphoneos${SDK} -find lipo)
 ORIG_PWD="$(pwd)"
-ARM_TARGETS="arm64-darwin-gcc
-             armv7-darwin-gcc
-             armv7s-darwin-gcc"
-SIM_TARGETS="x86-iphonesimulator-gcc
-             x86_64-iphonesimulator-gcc"
-OSX_TARGETS="x86-darwin15-gcc
-             x86_64-darwin15-gcc"
-TARGETS="${ARM_TARGETS} ${SIM_TARGETS}"
+TARGETS="arm64-darwin-gcc
+         armv7-darwin-gcc
+         armv7s-darwin-gcc
+         x86-iphonesimulator-gcc
+         x86_64-iphonesimulator-gcc"

 # Configures for the target specified by $1, and invokes make with the dist
 # target using $DIST_DIR as the distribution output directory.
 build_target() {
  local target="$1"
  local old_pwd="$(pwd)"
-  local target_specific_flags=""

  vlog "***Building target: ${target}***"

-  case "${target}" in
-    x86-*)
-      target_specific_flags="--enable-pic"
-      vlog "Enabled PIC for ${target}"
-      ;;
-  esac
-
  mkdir "${target}"
  cd "${target}"
  eval "${LIBVPX_SOURCE_DIR}/configure" --target="${target}" \
-    ${CONFIGURE_ARGS} ${EXTRA_CONFIGURE_ARGS} ${target_specific_flags} \
-    ${devnull}
+    ${CONFIGURE_ARGS} ${EXTRA_CONFIGURE_ARGS} ${devnull}
  export DIST_DIR
-  eval make dist ${devnull}
+  eval make -j ${MAKE_JOBS} dist ${devnull}
  cd "${old_pwd}"

  vlog "***Done building target: ${target}***"
@@ -138,44 +126,6 @@ create_vpx_framework_config_shim() {
  printf "#endif  // ${include_guard}" >> "${config_file}"
 }

-# Verifies that $FRAMEWORK_LIB fat library contains requested builds.
-verify_framework_targets() {
-  local requested_cpus=""
-  local cpu=""
-
-  # Extract CPU from full target name.
-  for target; do
-    cpu="${target%%-*}"
-    if [ "${cpu}" = "x86" ]; then
-      # lipo -info outputs i386 for libvpx x86 targets.
-      cpu="i386"
-    fi
-    requested_cpus="${requested_cpus}${cpu} "
-  done
-
-  # Get target CPUs present in framework library.
-  local targets_built=$(${LIPO} -info ${FRAMEWORK_LIB})
-
-  # $LIPO -info outputs a string like the following:
-  #   Architectures in the fat file: $FRAMEWORK_LIB <architectures>
-  # Capture only the architecture strings.
-  targets_built=${targets_built##*: }
-
-  # Sort CPU strings to make the next step a simple string compare.
-  local actual=$(echo ${targets_built} | tr " " "\n" | sort | tr "\n" " ")
-  local requested=$(echo ${requested_cpus} | tr " " "\n" | sort | tr "\n" " ")
-
-  vlog "Requested ${FRAMEWORK_LIB} CPUs: ${requested}"
-  vlog "Actual ${FRAMEWORK_LIB} CPUs: ${actual}"
-
-  if [ "${requested}" != "${actual}" ]; then
-    elog "Actual ${FRAMEWORK_LIB} targets do not match requested target list."
-    elog "  Requested target CPUs: ${requested}"
-    elog "  Actual target CPUs: ${actual}"
-    return 1
-  fi
-}
-
 # Configures and builds each target specified by $1, and then builds
 # VPX.framework.
 build_framework() {
@@ -196,12 +146,7 @@ build_framework() {
  for target in ${targets}; do
    build_target "${target}"
    target_dist_dir="${BUILD_ROOT}/${target}/${DIST_DIR}"
-    if [ "${ENABLE_SHARED}" = "yes" ]; then
-      local suffix="dylib"
-    else
-      local suffix="a"
-    fi
-    lib_list="${lib_list} ${target_dist_dir}/lib/libvpx.${suffix}"
+    lib_list="${lib_list} ${target_dist_dir}/lib/libvpx.a"
  done

  cd "${ORIG_PWD}"
@@ -220,25 +165,13 @@ build_framework() {
  # Copy in vpx_version.h.
  cp -p "${BUILD_ROOT}/${target}/vpx_version.h" "${HEADER_DIR}"

-  if [ "${ENABLE_SHARED}" = "yes" ]; then
-    # Adjust the dylib's name so dynamic linking in apps works as expected.
-    install_name_tool -id '@rpath/VPX.framework/VPX' ${FRAMEWORK_DIR}/VPX
-
-    # Copy in Info.plist.
-    cat "${SCRIPT_DIR}/ios-Info.plist" \
-      | sed "s/\${FULLVERSION}/${FULLVERSION}/g" \
-      | sed "s/\${VERSION}/${VERSION}/g" \
-      | sed "s/\${IOS_VERSION_MIN}/${IOS_VERSION_MIN}/g" \
-      > "${FRAMEWORK_DIR}/Info.plist"
-  fi
-
-  # Confirm VPX.framework/VPX contains the targets requested.
-  verify_framework_targets ${targets}
-
-  vlog "Created fat library ${FRAMEWORK_LIB} containing:"
+  vlog "Created fat library ${FRAMEWORK_DIR}/VPX containing:"
  for lib in ${lib_list}; do
    vlog "  $(echo ${lib} | awk -F / '{print $2, $NF}')"
  done
+
+  # TODO(tomfinegan): Verify that expected targets are included within
+  # VPX.framework/VPX via lipo -info.
 }

 # Trap function. Cleans up the subtree used to build all targets contained in
@@ -256,30 +189,16 @@ cleanup() {
  fi
 }

-print_list() {
-  local indent="$1"
-  shift
-  local list="$@"
-  for entry in ${list}; do
-    echo "${indent}${entry}"
-  done
-}
-
 iosbuild_usage() {
 cat << EOF
  Usage: ${0##*/} [arguments]
    --help: Display this message and exit.
-    --enable-shared: Build a dynamic framework for use on iOS 8 or later.
    --extra-configure-args <args>: Extra args to pass when configuring libvpx.
-    --macosx: Uses darwin15 targets instead of iphonesimulator targets for x86
-              and x86_64. Allows linking to framework when builds target MacOSX
-              instead of iOS.
+    --jobs: Number of make jobs.
    --preserve-build-output: Do not delete the build directory.
    --show-build-output: Show output from each library build.
    --targets <targets>: Override default target list. Defaults:
-$(print_list "        " ${TARGETS})
-    --test-link: Confirms all targets can be linked. Functionally identical to
-                 passing --enable-examples via --extra-configure-args.
+         ${TARGETS}
    --verbose: Output information about the environment and each stage of the
               build.
 EOF
@@ -308,8 +227,9 @@ while [ -n "$1" ]; do
      iosbuild_usage
      exit
      ;;
-    --enable-shared)
-      ENABLE_SHARED=yes
+    --jobs)
+      MAKE_JOBS="$2"
+      shift
      ;;
    --preserve-build-output)
      PRESERVE_BUILD_OUTPUT=yes
@@ -317,16 +237,10 @@ while [ -n "$1" ]; do
    --show-build-output)
      devnull=
      ;;
-    --test-link)
-      EXTRA_CONFIGURE_ARGS="${EXTRA_CONFIGURE_ARGS} --enable-examples"
-      ;;
    --targets)
      TARGETS="$2"
      shift
      ;;
-    --macosx)
-      TARGETS="${ARM_TARGETS} ${OSX_TARGETS}"
-      ;;
    --verbose)
      VERBOSE=yes
      ;;
@@ -338,21 +252,6 @@ while [ -n "$1" ]; do
  shift
 done

-if [ "${ENABLE_SHARED}" = "yes" ]; then
-  CONFIGURE_ARGS="--enable-shared ${CONFIGURE_ARGS}"
-fi
-
-FULLVERSION=$("${SCRIPT_DIR}"/version.sh --bare "${LIBVPX_SOURCE_DIR}")
-VERSION=$(echo "${FULLVERSION}" | sed -E 's/^v([0-9]+\.[0-9]+\.[0-9]+).*$/\1/')
-
-if [ "$ENABLE_SHARED" = "yes" ]; then
-  IOS_VERSION_OPTIONS="--enable-shared"
-  IOS_VERSION_MIN="8.0"
-else
-  IOS_VERSION_OPTIONS=""
-  IOS_VERSION_MIN="6.0"
-fi
-
 if [ "${VERBOSE}" = "yes" ]; then
 cat << EOF
  BUILD_ROOT=${BUILD_ROOT}
@@ -360,24 +259,16 @@ cat << EOF
  CONFIGURE_ARGS=${CONFIGURE_ARGS}
  EXTRA_CONFIGURE_ARGS=${EXTRA_CONFIGURE_ARGS}
  FRAMEWORK_DIR=${FRAMEWORK_DIR}
-  FRAMEWORK_LIB=${FRAMEWORK_LIB}
  HEADER_DIR=${HEADER_DIR}
+  MAKE_JOBS=${MAKE_JOBS}
+  PRESERVE_BUILD_OUTPUT=${PRESERVE_BUILD_OUTPUT}
  LIBVPX_SOURCE_DIR=${LIBVPX_SOURCE_DIR}
  LIPO=${LIPO}
-  MAKEFLAGS=${MAKEFLAGS}
  ORIG_PWD=${ORIG_PWD}
-  PRESERVE_BUILD_OUTPUT=${PRESERVE_BUILD_OUTPUT}
-  TARGETS="$(print_list "" ${TARGETS})"
-  ENABLE_SHARED=${ENABLE_SHARED}
-  OSX_TARGETS="${OSX_TARGETS}"
-  SIM_TARGETS="${SIM_TARGETS}"
-  SCRIPT_DIR="${SCRIPT_DIR}"
-  FULLVERSION="${FULLVERSION}"
-  VERSION="${VERSION}"
-  IOS_VERSION_MIN="${IOS_VERSION_MIN}"
+  TARGETS="${TARGETS}"
 EOF
 fi

 build_framework "${TARGETS}"
 echo "Successfully built '${FRAMEWORK_DIR}' for:"
-print_list "" ${TARGETS}
+echo "         ${TARGETS}"
--- a/build/make/msvs_common.sh
+++ b/build/make/msvs_common.sh
@@ -39,12 +39,11 @@ fix_path() {
 }

 # Corrects the paths in file_list in one pass for efficiency.
-# $1 is the name of the array to be modified.
 fix_file_list() {
-    declare -n array_ref=$1
-    files=$(fix_path "${array_ref[@]}")
+    # TODO(jzern): this could be more generic and take the array as a param.
+    files=$(fix_path "${file_list[@]}")
    local IFS=$'\n'
-    array_ref=($files)
+    file_list=($files)
 }

 generate_uuid() {
--- a/build/make/rtcd.pl
+++ b/build/make/rtcd.pl
@@ -319,8 +319,14 @@ EOF

  print <<EOF;
 #if HAVE_DSPR2
-void vpx_dsputil_static_init();
-vpx_dsputil_static_init();
+#if CONFIG_VP8
+void dsputil_static_init();
+dsputil_static_init();
+#endif
+#if CONFIG_VP9
+void vp9_dsputil_static_init();
+vp9_dsputil_static_init();
+#endif
 #endif
 }
 #endif
--- a/build/make/version.sh
+++ b/build/make/version.sh
@@ -24,9 +24,8 @@ out_file=${2}
 id=${3:-VERSION_STRING}

 git_version_id=""
-if [ -e "${source_path}/.git" ]; then
+if [ -d "${source_path}/.git" ]; then
    # Source Path is a git working copy. Check for local modifications.
-    # Note that git submodules may have a file as .git, not a directory.
    export GIT_DIR="${source_path}/.git"
    git_version_id=`git describe --match=v[0-9]* 2>/dev/null`
 fi
--- a/codereview.settings
+++ b/codereview.settings
@@ -1,4 +0,0 @@
-# This file is used by gcl to get repository specific information.
-GERRIT_HOST: chromium-review.googlesource.com
-GERRIT_PORT: 29418
-CODE_REVIEW_SERVER: chromium-review.googlesource.com
--- a/249
+++ b/249
@@ -31,15 +31,16 @@ Advanced options:
  --size-limit=WxH                max size to allow in the decoder
  --as={yasm|nasm|auto}           use specified assembler [auto, yasm preferred]
  --sdk-path=PATH                 path to root of sdk (android builds only)
+  ${toggle_fast_unaligned}        don't use unaligned accesses, even when
+                                  supported by hardware [auto]
  ${toggle_codec_srcs}            in/exclude codec library source code
  ${toggle_debug_libs}            in/exclude debug version of libraries
  ${toggle_static_msvcrt}         use static MSVCRT (VS builds only)
-  ${toggle_vpx_highbitdepth}      use VPX high bit depth (10/12) profiles
-  ${toggle_better_hw_compatibility}
-                                  enable encoder to produce streams with better
-                                  hardware decoder compatibility
-  ${toggle_vp10}                  VP10 codec support
+  ${toggle_vp9_highbitdepth}      use VP9 high bit depth (10/12) profiles
+  ${toggle_vp8}                   VP8 codec support
+  ${toggle_vp9}                   VP9 codec support
  ${toggle_internal_stats}        output of encoder internal stats for debug, if supported (encoders)
+  ${toggle_mem_tracker}           track memory usage
  ${toggle_postproc}              postprocessing
  ${toggle_vp9_postproc}          vp9 specific postprocessing
  ${toggle_multithread}           multithreaded encoding and decoding
@@ -96,11 +97,11 @@ EOF

 # all_platforms is a list of all supported target platforms. Maintain
 # alphabetically by architecture, generic-gnu last.
-all_platforms="${all_platforms} arm64-darwin-gcc"
-all_platforms="${all_platforms} arm64-linux-gcc"
+all_platforms="${all_platforms} armv6-darwin-gcc"
 all_platforms="${all_platforms} armv6-linux-rvct"
 all_platforms="${all_platforms} armv6-linux-gcc"
 all_platforms="${all_platforms} armv6-none-rvct"
+all_platforms="${all_platforms} arm64-darwin-gcc"
 all_platforms="${all_platforms} armv7-android-gcc"   #neon Cortex-A8
 all_platforms="${all_platforms} armv7-darwin-gcc"    #neon Cortex-A8
 all_platforms="${all_platforms} armv7-linux-rvct"    #neon Cortex-A8
@@ -108,11 +109,15 @@ all_platforms="${all_platforms} armv7-linux-gcc"     #neon Cortex-A8
 all_platforms="${all_platforms} armv7-none-rvct"     #neon Cortex-A8
 all_platforms="${all_platforms} armv7-win32-vs11"
 all_platforms="${all_platforms} armv7-win32-vs12"
-all_platforms="${all_platforms} armv7-win32-vs14"
 all_platforms="${all_platforms} armv7s-darwin-gcc"
-all_platforms="${all_platforms} armv8-linux-gcc"
 all_platforms="${all_platforms} mips32-linux-gcc"
 all_platforms="${all_platforms} mips64-linux-gcc"
+all_platforms="${all_platforms} ppc32-darwin8-gcc"
+all_platforms="${all_platforms} ppc32-darwin9-gcc"
+all_platforms="${all_platforms} ppc32-linux-gcc"
+all_platforms="${all_platforms} ppc64-darwin8-gcc"
+all_platforms="${all_platforms} ppc64-darwin9-gcc"
+all_platforms="${all_platforms} ppc64-linux-gcc"
 all_platforms="${all_platforms} sparc-solaris-gcc"
 all_platforms="${all_platforms} x86-android-gcc"
 all_platforms="${all_platforms} x86-darwin8-gcc"
@@ -124,34 +129,41 @@ all_platforms="${all_platforms} x86-darwin11-gcc"
 all_platforms="${all_platforms} x86-darwin12-gcc"
 all_platforms="${all_platforms} x86-darwin13-gcc"
 all_platforms="${all_platforms} x86-darwin14-gcc"
-all_platforms="${all_platforms} x86-darwin15-gcc"
 all_platforms="${all_platforms} x86-iphonesimulator-gcc"
 all_platforms="${all_platforms} x86-linux-gcc"
 all_platforms="${all_platforms} x86-linux-icc"
 all_platforms="${all_platforms} x86-os2-gcc"
 all_platforms="${all_platforms} x86-solaris-gcc"
 all_platforms="${all_platforms} x86-win32-gcc"
+all_platforms="${all_platforms} x86-win32-vs7"
+all_platforms="${all_platforms} x86-win32-vs8"
+all_platforms="${all_platforms} x86-win32-vs9"
 all_platforms="${all_platforms} x86-win32-vs10"
 all_platforms="${all_platforms} x86-win32-vs11"
 all_platforms="${all_platforms} x86-win32-vs12"
-all_platforms="${all_platforms} x86-win32-vs14"
-all_platforms="${all_platforms} x86_64-android-gcc"
 all_platforms="${all_platforms} x86_64-darwin9-gcc"
 all_platforms="${all_platforms} x86_64-darwin10-gcc"
 all_platforms="${all_platforms} x86_64-darwin11-gcc"
 all_platforms="${all_platforms} x86_64-darwin12-gcc"
 all_platforms="${all_platforms} x86_64-darwin13-gcc"
 all_platforms="${all_platforms} x86_64-darwin14-gcc"
-all_platforms="${all_platforms} x86_64-darwin15-gcc"
 all_platforms="${all_platforms} x86_64-iphonesimulator-gcc"
 all_platforms="${all_platforms} x86_64-linux-gcc"
 all_platforms="${all_platforms} x86_64-linux-icc"
 all_platforms="${all_platforms} x86_64-solaris-gcc"
 all_platforms="${all_platforms} x86_64-win64-gcc"
+all_platforms="${all_platforms} x86_64-win64-vs8"
+all_platforms="${all_platforms} x86_64-win64-vs9"
 all_platforms="${all_platforms} x86_64-win64-vs10"
 all_platforms="${all_platforms} x86_64-win64-vs11"
 all_platforms="${all_platforms} x86_64-win64-vs12"
-all_platforms="${all_platforms} x86_64-win64-vs14"
+all_platforms="${all_platforms} universal-darwin8-gcc"
+all_platforms="${all_platforms} universal-darwin9-gcc"
+all_platforms="${all_platforms} universal-darwin10-gcc"
+all_platforms="${all_platforms} universal-darwin11-gcc"
+all_platforms="${all_platforms} universal-darwin12-gcc"
+all_platforms="${all_platforms} universal-darwin13-gcc"
+all_platforms="${all_platforms} universal-darwin14-gcc"
 all_platforms="${all_platforms} generic-gnu"

 # all_targets is a list of all targets that can be configured
@@ -188,9 +200,6 @@ if [ ${doxy_major:-0} -ge 1 ]; then
    [ $doxy_minor -eq 5 ] && [ $doxy_patch -ge 3 ] && enable_feature doxygen
 fi

-# disable codecs when their source directory does not exist
-[ -d "${source_path}/vp10" ] || disable_codec vp10
-
 # install everything except the sources, by default. sources will have
 # to be enabled when doing dist builds, since that's no longer a common
 # case.
@@ -201,34 +210,45 @@ enable_feature install_libs
 enable_feature static
 enable_feature optimizations
 enable_feature dependency_tracking
+enable_feature fast_unaligned #allow unaligned accesses, if supported by hw
 enable_feature spatial_resampling
 enable_feature multithread
 enable_feature os_support
 enable_feature temporal_denoising

-CODECS="
-    vp10_encoder
-    vp10_decoder
-"
-CODEC_FAMILIES="
-    vp10
-"
+[ -d "${source_path}/../include" ] && enable_feature alt_tree_layout
+for d in vp8 vp9; do
+    [ -d "${source_path}/${d}" ] && disable_feature alt_tree_layout;
+done
+
+if ! enabled alt_tree_layout; then
+# development environment
+[ -d "${source_path}/vp8" ] && CODECS="${CODECS} vp8_encoder vp8_decoder"
+[ -d "${source_path}/vp9" ] && CODECS="${CODECS} vp9_encoder vp9_decoder"
+else
+# customer environment
+[ -f "${source_path}/../include/vpx/vp8cx.h" ] && CODECS="${CODECS} vp8_encoder"
+[ -f "${source_path}/../include/vpx/vp8dx.h" ] && CODECS="${CODECS} vp8_decoder"
+[ -f "${source_path}/../include/vpx/vp9cx.h" ] && CODECS="${CODECS} vp9_encoder"
+[ -f "${source_path}/../include/vpx/vp9dx.h" ] && CODECS="${CODECS} vp9_decoder"
+[ -f "${source_path}/../include/vpx/vp8cx.h" ] || disable_feature vp8_encoder
+[ -f "${source_path}/../include/vpx/vp8dx.h" ] || disable_feature vp8_decoder
+[ -f "${source_path}/../include/vpx/vp9cx.h" ] || disable_feature vp9_encoder
+[ -f "${source_path}/../include/vpx/vp9dx.h" ] || disable_feature vp9_decoder
+
+[ -f "${source_path}/../lib/*/*mt.lib" ] && soft_enable static_msvcrt
+fi
+
+CODECS="$(echo ${CODECS} | tr ' ' '\n')"
+CODEC_FAMILIES="$(for c in ${CODECS}; do echo ${c%_*}; done | sort | uniq)"

 ARCH_LIST="
    arm
    mips
    x86
    x86_64
-"
-ARCH_EXT_LIST_X86="
-    mmx
-    sse
-    sse2
-    sse3
-    ssse3
-    sse4_1
-    avx
-    avx2
+    ppc32
+    ppc64
 "
 ARCH_EXT_LIST="
    edsp
@@ -241,39 +261,30 @@ ARCH_EXT_LIST="
    msa
    mips64

-    ${ARCH_EXT_LIST_X86}
+    mmx
+    sse
+    sse2
+    sse3
+    ssse3
+    sse4_1
+    avx
+    avx2
+
+    altivec
 "
 HAVE_LIST="
    ${ARCH_EXT_LIST}
    vpx_ports
+    stdint_h
+    alt_tree_layout
    pthread_h
+    sys_mman_h
    unistd_h
 "
 EXPERIMENT_LIST="
    spatial_svc
    fp_mb_stats
    emulate_hardware
-    var_tx
-    rect_tx
-    ref_mv
-    dual_filter
-    ext_tx
-    ext_intra
-    ext_inter
-    ext_interp
-    ext_refs
-    global_motion
-    new_quant
-    supertx
-    ans
-    loop_restoration
-    ext_partition
-    ext_partition_types
-    ext_tile
-    obmc
-    warped_motion
-    entropy
-    bidir_pred
 "
 CONFIG_LIST="
    dependency_tracking
@@ -282,6 +293,7 @@ CONFIG_LIST="
    install_bins
    install_libs
    install_srcs
+    use_x86inc
    debug
    gprof
    gcov
@@ -293,6 +305,10 @@ CONFIG_LIST="

    codec_srcs
    debug_libs
+    fast_unaligned
+    mem_manager
+    mem_tracker
+    mem_checks

    dequant_tokens
    dc_recon
@@ -324,8 +340,7 @@ CONFIG_LIST="
    temporal_denoising
    vp9_temporal_denoising
    coefficient_range_checking
-    vpx_highbitdepth
-    better_hw_compatibility
+    vp9_highbitdepth
    experimental
    size_limit
    ${EXPERIMENT_LIST}
@@ -343,6 +358,7 @@ CMDLINE_SELECT="
    gprof
    gcov
    pic
+    use_x86inc
    optimizations
    ccache
    runtime_cpu_detect
@@ -354,6 +370,7 @@ CMDLINE_SELECT="
    libc
    as
    size_limit
+    fast_unaligned
    codec_srcs
    debug_libs

@@ -366,6 +383,7 @@ CMDLINE_SELECT="
    ${CODECS}
    ${CODEC_FAMILIES}
    static_msvcrt
+    mem_tracker
    spatial_resampling
    realtime_only
    onthefly_bitpacking
@@ -383,8 +401,7 @@ CMDLINE_SELECT="
    temporal_denoising
    vp9_temporal_denoising
    coefficient_range_checking
-    better_hw_compatibility
-    vpx_highbitdepth
+    vp9_highbitdepth
    experimental
 "

@@ -392,19 +409,15 @@ process_cmdline() {
    for opt do
        optval="${opt#*=}"
        case "$opt" in
-        --disable-codecs)
-          for c in ${CODEC_FAMILIES}; do disable_codec $c; done
-          ;;
+        --disable-codecs) for c in ${CODECS}; do disable_feature $c; done ;;
        --enable-?*|--disable-?*)
        eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'`
-        if is_in ${option} ${EXPERIMENT_LIST}; then
+        if echo "${EXPERIMENT_LIST}" | grep "^ *$option\$" >/dev/null; then
            if enabled experimental; then
                ${action}_feature $option
            else
                log_echo "Ignoring $opt -- not in experimental mode."
            fi
-        elif is_in ${option} "${CODECS} ${CODEC_FAMILIES}"; then
-            ${action}_codec ${option}
        else
            process_common_cmdline $opt
        fi
@@ -418,6 +431,14 @@ process_cmdline() {
 post_process_cmdline() {
    c=""

+    # If the codec family is disabled, disable all components of that family.
+    # If the codec family is enabled, enable all components of that family.
+    log_echo "Configuring selected codecs"
+    for c in ${CODECS}; do
+        disabled ${c%%_*} && disable_feature ${c}
+        enabled ${c%%_*} && enable_feature ${c}
+    done
+
    # Enable all detected codecs, if they haven't been disabled
    for c in ${CODECS}; do soft_enable $c; done

@@ -435,8 +456,22 @@ post_process_cmdline() {

 process_targets() {
    enabled child || write_common_config_banner
-    write_common_target_config_h ${BUILD_PFX}vpx_config.h
+    enabled universal || write_common_target_config_h  ${BUILD_PFX}vpx_config.h
+
+    # For fat binaries, call configure recursively to configure for each
+    # binary architecture to be included.
+    if enabled universal; then
+        # Call configure (ourselves) for each subarchitecture
+        for arch in $fat_bin_archs; do
+            BUILD_PFX=${arch}/ toolchain=${arch} $self --child $cmdline_args || exit $?
+        done
+    fi
+
+    # The write_common_config (config.mk) logic is deferred until after the
+    # recursive calls to configure complete, because we want our universal
+    # targets to be executed last.
    write_common_config_targets
+    enabled universal && echo "FAT_ARCHS=${fat_bin_archs}" >> config.mk

    # Calculate the default distribution name, based on the enabled features
    cf=""
@@ -512,18 +547,13 @@ process_detect() {
        # Can only build shared libs on a subset of platforms. Doing this check
        # here rather than at option parse time because the target auto-detect
        # magic happens after the command line has been parsed.
-        case "${tgt_os}" in
-        linux|os2|darwin*|iphonesimulator*)
-            # Supported platforms
-            ;;
-        *)
+        if ! enabled linux; then
            if enabled gnu; then
                echo "--enable-shared is only supported on ELF; assuming this is OK"
            else
-                die "--enable-shared only supported on ELF, OS/2, and Darwin for now"
+                die "--enable-shared only supported on ELF for now"
            fi
-            ;;
-        esac
+        fi
    fi
    if [ -z "$CC" ] || enabled external_build; then
        echo "Bypassing toolchain for environment detection."
@@ -550,12 +580,16 @@ process_detect() {
            # Specialize windows and POSIX environments.
            case $toolchain in
                *-win*-*)
-                    # Don't check for any headers in Windows builds.
-                    false
-                ;;
+                    case $header-$toolchain in
+                        stdint*-gcc) true;;
+                        *) false;;
+                    esac && enable_feature $var
+                    ;;
                *)
                    case $header in
+                        stdint.h) true;;
                        pthread.h) true;;
+                        sys/mman.h) true;;
                        unistd.h) true;;
                        *) false;;
                    esac && enable_feature $var
@@ -571,7 +605,9 @@ process_detect() {
 int main(void) {return 0;}
 EOF
    # check system headers
+    check_header stdint.h
    check_header pthread.h
+    check_header sys/mman.h
    check_header unistd.h # for sysconf(3) and friends.

    check_header vpx/vpx_integer.h -I${source_path} && enable_feature vpx_ports
@@ -580,6 +616,30 @@ EOF
 process_toolchain() {
    process_common_toolchain

+    # Handle universal binaries for this architecture
+    case $toolchain in
+        universal-darwin*)
+            darwin_ver=${tgt_os##darwin}
+
+            # Snow Leopard (10.6/darwin10) dropped support for PPC
+            # Include PPC support for all prior versions
+            if [ $darwin_ver -lt 10 ]; then
+                fat_bin_archs="$fat_bin_archs ppc32-${tgt_os}-gcc"
+            fi
+
+            # Tiger (10.4/darwin8) brought support for x86
+            if [ $darwin_ver -ge 8 ]; then
+                fat_bin_archs="$fat_bin_archs x86-${tgt_os}-${tgt_cc}"
+            fi
+
+            # Leopard (10.5/darwin9) brought 64 bit support
+            if [ $darwin_ver -ge 9 ]; then
+                fat_bin_archs="$fat_bin_archs x86_64-${tgt_os}-${tgt_cc}"
+            fi
+            ;;
+    esac
+
+
    # Enable some useful compiler flags
    if enabled gcc; then
        enabled werror && check_add_cflags -Werror
@@ -602,11 +662,7 @@ process_toolchain() {
          ;;
          *) check_add_cflags -Wunused-but-set-variable ;;
        esac
-        if enabled mips || [ -z "${INLINE}" ]; then
-          enabled extra_warnings || check_add_cflags -Wno-unused-function
-        else
-          check_add_cflags -Wunused-function
-        fi
+        enabled extra_warnings || check_add_cflags -Wno-unused-function
    fi

    if enabled icc; then
@@ -654,16 +710,24 @@ process_toolchain() {
        vs*) enable_feature msvs
             enable_feature solution
             vs_version=${tgt_cc##vs}
-             VCPROJ_SFX=vcxproj
-             gen_vcproj_cmd=${source_path}/build/make/gen_msvs_vcxproj.sh
-             enabled werror && gen_vcproj_cmd="${gen_vcproj_cmd} --enable-werror"
+             case $vs_version in
+             [789])
+                 VCPROJ_SFX=vcproj
+                 gen_vcproj_cmd=${source_path}/build/make/gen_msvs_proj.sh
+                 ;;
+             10|11|12)
+                 VCPROJ_SFX=vcxproj
+                 gen_vcproj_cmd=${source_path}/build/make/gen_msvs_vcxproj.sh
+                 enabled werror && gen_vcproj_cmd="${gen_vcproj_cmd} --enable-werror"
+                 ;;
+             esac
             all_targets="${all_targets} solution"
             INLINE="__forceinline"
        ;;
    esac

    # Other toolchain specific defaults
-    case $toolchain in x86*) soft_enable postproc;; esac
+    case $toolchain in x86*|ppc*|universal*) soft_enable postproc;; esac

    if enabled postproc_visualizer; then
        enabled postproc || die "postproc_visualizer requires postproc to be enabled"
@@ -717,16 +781,6 @@ EOF
    esac
    # libwebm needs to be linked with C++ standard library
    enabled webm_io && LD=${CXX}
-
-    # append any user defined extra cflags
-    if [ -n "${extra_cflags}" ] ; then
-        check_add_cflags ${extra_cflags} || \
-        die "Requested extra CFLAGS '${extra_cflags}' not supported by compiler"
-    fi
-    if [ -n "${extra_cxxflags}" ]; then
-        check_add_cxxflags ${extra_cxxflags} || \
-        die "Requested extra CXXFLAGS '${extra_cxxflags}' not supported by compiler"
-    fi
 }


@@ -737,7 +791,6 @@ CONFIGURE_ARGS="$@"
 process "$@"
 print_webm_license ${BUILD_PFX}vpx_config.c "/*" " */"
 cat <<EOF >> ${BUILD_PFX}vpx_config.c
-#include "vpx/vpx_codec.h"
 static const char* const cfg = "$CONFIGURE_ARGS";
 const char *vpx_codec_build_config(void) {return cfg;}
 EOF
--- a/examples.mk
+++ b/examples.mk
@@ -22,44 +22,33 @@ LIBYUV_SRCS +=  third_party/libyuv/include/libyuv/basic_types.h  \
                third_party/libyuv/source/planar_functions.cc \
                third_party/libyuv/source/row_any.cc \
                third_party/libyuv/source/row_common.cc \
-                third_party/libyuv/source/row_gcc.cc \
                third_party/libyuv/source/row_mips.cc \
                third_party/libyuv/source/row_neon.cc \
                third_party/libyuv/source/row_neon64.cc \
+                third_party/libyuv/source/row_posix.cc \
                third_party/libyuv/source/row_win.cc \
                third_party/libyuv/source/scale.cc \
-                third_party/libyuv/source/scale_any.cc \
                third_party/libyuv/source/scale_common.cc \
-                third_party/libyuv/source/scale_gcc.cc \
                third_party/libyuv/source/scale_mips.cc \
                third_party/libyuv/source/scale_neon.cc \
                third_party/libyuv/source/scale_neon64.cc \
+                third_party/libyuv/source/scale_posix.cc \
                third_party/libyuv/source/scale_win.cc \

-LIBWEBM_COMMON_SRCS += third_party/libwebm/common/hdr_util.cc \
-                       third_party/libwebm/common/hdr_util.h \
-                       third_party/libwebm/common/webmids.h
-
-LIBWEBM_MUXER_SRCS += third_party/libwebm/mkvmuxer/mkvmuxer.cc \
-                      third_party/libwebm/mkvmuxer/mkvmuxerutil.cc \
-                      third_party/libwebm/mkvmuxer/mkvwriter.cc \
-                      third_party/libwebm/mkvmuxer/mkvmuxer.h \
-                      third_party/libwebm/mkvmuxer/mkvmuxertypes.h \
-                      third_party/libwebm/mkvmuxer/mkvmuxerutil.h \
-                      third_party/libwebm/mkvparser/mkvparser.h \
-                      third_party/libwebm/mkvmuxer/mkvwriter.h
-
-LIBWEBM_PARSER_SRCS = third_party/libwebm/mkvparser/mkvparser.cc \
-                      third_party/libwebm/mkvparser/mkvreader.cc \
-                      third_party/libwebm/mkvparser/mkvparser.h \
-                      third_party/libwebm/mkvparser/mkvreader.h
-
-# Add compile flags and include path for libwebm sources.
-ifeq ($(CONFIG_WEBM_IO),yes)
-  CXXFLAGS     += -D__STDC_CONSTANT_MACROS -D__STDC_LIMIT_MACROS
-  INC_PATH-yes += $(SRC_PATH_BARE)/third_party/libwebm
-endif
+LIBWEBM_MUXER_SRCS += third_party/libwebm/mkvmuxer.cpp \
+                      third_party/libwebm/mkvmuxerutil.cpp \
+                      third_party/libwebm/mkvwriter.cpp \
+                      third_party/libwebm/mkvmuxer.hpp \
+                      third_party/libwebm/mkvmuxertypes.hpp \
+                      third_party/libwebm/mkvmuxerutil.hpp \
+                      third_party/libwebm/mkvparser.hpp \
+                      third_party/libwebm/mkvwriter.hpp \
+                      third_party/libwebm/webmids.hpp

+LIBWEBM_PARSER_SRCS = third_party/libwebm/mkvparser.cpp \
+                      third_party/libwebm/mkvreader.cpp \
+                      third_party/libwebm/mkvparser.hpp \
+                      third_party/libwebm/mkvreader.hpp

 # List of examples to build. UTILS are tools meant for distribution
 # while EXAMPLES demonstrate specific portions of the API.
@@ -67,7 +56,6 @@ UTILS-$(CONFIG_DECODERS)    += vpxdec.c
 vpxdec.SRCS                 += md5_utils.c md5_utils.h
 vpxdec.SRCS                 += vpx_ports/mem_ops.h
 vpxdec.SRCS                 += vpx_ports/mem_ops_aligned.h
-vpxdec.SRCS                 += vpx_ports/msvc.h
 vpxdec.SRCS                 += vpx_ports/vpx_timer.h
 vpxdec.SRCS                 += vpx/vpx_integer.h
 vpxdec.SRCS                 += args.c args.h
@@ -78,8 +66,6 @@ ifeq ($(CONFIG_LIBYUV),yes)
  vpxdec.SRCS                 += $(LIBYUV_SRCS)
 endif
 ifeq ($(CONFIG_WEBM_IO),yes)
-  vpxdec.SRCS                 += $(LIBWEBM_COMMON_SRCS)
-  vpxdec.SRCS                 += $(LIBWEBM_MUXER_SRCS)
  vpxdec.SRCS                 += $(LIBWEBM_PARSER_SRCS)
  vpxdec.SRCS                 += webmdec.cc webmdec.h
 endif
@@ -94,27 +80,38 @@ vpxenc.SRCS                 += tools_common.c tools_common.h
 vpxenc.SRCS                 += warnings.c warnings.h
 vpxenc.SRCS                 += vpx_ports/mem_ops.h
 vpxenc.SRCS                 += vpx_ports/mem_ops_aligned.h
-vpxenc.SRCS                 += vpx_ports/msvc.h
 vpxenc.SRCS                 += vpx_ports/vpx_timer.h
 vpxenc.SRCS                 += vpxstats.c vpxstats.h
 ifeq ($(CONFIG_LIBYUV),yes)
  vpxenc.SRCS                 += $(LIBYUV_SRCS)
 endif
 ifeq ($(CONFIG_WEBM_IO),yes)
-  vpxenc.SRCS                 += $(LIBWEBM_COMMON_SRCS)
  vpxenc.SRCS                 += $(LIBWEBM_MUXER_SRCS)
-  vpxenc.SRCS                 += $(LIBWEBM_PARSER_SRCS)
  vpxenc.SRCS                 += webmenc.cc webmenc.h
 endif
 vpxenc.GUID                  = 548DEC74-7A15-4B2B-AFC3-AA102E7C25C1
 vpxenc.DESCRIPTION           = Full featured encoder
+ifeq ($(CONFIG_SPATIAL_SVC),yes)
+  EXAMPLES-$(CONFIG_VP9_ENCODER)      += vp9_spatial_svc_encoder.c
+  vp9_spatial_svc_encoder.SRCS        += args.c args.h
+  vp9_spatial_svc_encoder.SRCS        += ivfenc.c ivfenc.h
+  vp9_spatial_svc_encoder.SRCS        += tools_common.c tools_common.h
+  vp9_spatial_svc_encoder.SRCS        += video_common.h
+  vp9_spatial_svc_encoder.SRCS        += video_writer.h video_writer.c
+  vp9_spatial_svc_encoder.SRCS        += vpxstats.c vpxstats.h
+  vp9_spatial_svc_encoder.GUID        = 4A38598D-627D-4505-9C7B-D4020C84100D
+  vp9_spatial_svc_encoder.DESCRIPTION = VP9 Spatial SVC Encoder
+endif
+
+ifneq ($(CONFIG_SHARED),yes)
+EXAMPLES-$(CONFIG_VP9_ENCODER)    += resize_util.c
+endif

 EXAMPLES-$(CONFIG_ENCODERS)          += vpx_temporal_svc_encoder.c
 vpx_temporal_svc_encoder.SRCS        += ivfenc.c ivfenc.h
 vpx_temporal_svc_encoder.SRCS        += tools_common.c tools_common.h
 vpx_temporal_svc_encoder.SRCS        += video_common.h
 vpx_temporal_svc_encoder.SRCS        += video_writer.h video_writer.c
-vpx_temporal_svc_encoder.SRCS        += vpx_ports/msvc.h
 vpx_temporal_svc_encoder.GUID        = B18C08F2-A439-4502-A78E-849BE3D60947
 vpx_temporal_svc_encoder.DESCRIPTION = Temporal SVC Encoder
 EXAMPLES-$(CONFIG_DECODERS)        += simple_decoder.c
@@ -125,8 +122,16 @@ simple_decoder.SRCS                += video_common.h
 simple_decoder.SRCS                += video_reader.h video_reader.c
 simple_decoder.SRCS                += vpx_ports/mem_ops.h
 simple_decoder.SRCS                += vpx_ports/mem_ops_aligned.h
-simple_decoder.SRCS                += vpx_ports/msvc.h
 simple_decoder.DESCRIPTION          = Simplified decoder loop
+EXAMPLES-$(CONFIG_DECODERS)        += postproc.c
+postproc.SRCS                      += ivfdec.h ivfdec.c
+postproc.SRCS                      += tools_common.h tools_common.c
+postproc.SRCS                      += video_common.h
+postproc.SRCS                      += video_reader.h video_reader.c
+postproc.SRCS                      += vpx_ports/mem_ops.h
+postproc.SRCS                      += vpx_ports/mem_ops_aligned.h
+postproc.GUID                       = 65E33355-F35E-4088-884D-3FD4905881D7
+postproc.DESCRIPTION                = Decoder postprocessor control
 EXAMPLES-$(CONFIG_DECODERS)        += decode_to_md5.c
 decode_to_md5.SRCS                 += md5_utils.h md5_utils.c
 decode_to_md5.SRCS                 += ivfdec.h ivfdec.c
@@ -135,7 +140,6 @@ decode_to_md5.SRCS                 += video_common.h
 decode_to_md5.SRCS                 += video_reader.h video_reader.c
 decode_to_md5.SRCS                 += vpx_ports/mem_ops.h
 decode_to_md5.SRCS                 += vpx_ports/mem_ops_aligned.h
-decode_to_md5.SRCS                 += vpx_ports/msvc.h
 decode_to_md5.GUID                  = 59120B9B-2735-4BFE-B022-146CA340FE42
 decode_to_md5.DESCRIPTION           = Frame by frame MD5 checksum
 EXAMPLES-$(CONFIG_ENCODERS)     += simple_encoder.c
@@ -143,23 +147,20 @@ simple_encoder.SRCS             += ivfenc.h ivfenc.c
 simple_encoder.SRCS             += tools_common.h tools_common.c
 simple_encoder.SRCS             += video_common.h
 simple_encoder.SRCS             += video_writer.h video_writer.c
-simple_encoder.SRCS             += vpx_ports/msvc.h
 simple_encoder.GUID              = 4607D299-8A71-4D2C-9B1D-071899B6FBFD
 simple_encoder.DESCRIPTION       = Simplified encoder loop
-EXAMPLES-$(CONFIG_VP10_ENCODER) += lossless_encoder.c
-lossless_encoder.SRCS           += ivfenc.h ivfenc.c
-lossless_encoder.SRCS           += tools_common.h tools_common.c
-lossless_encoder.SRCS           += video_common.h
-lossless_encoder.SRCS           += video_writer.h video_writer.c
-lossless_encoder.SRCS           += vpx_ports/msvc.h
-lossless_encoder.GUID            = B63C7C88-5348-46DC-A5A6-CC151EF93366
-lossless_encoder.DESCRIPTION     = Simplified lossless encoder
+EXAMPLES-$(CONFIG_VP9_ENCODER)  += vp9_lossless_encoder.c
+vp9_lossless_encoder.SRCS       += ivfenc.h ivfenc.c
+vp9_lossless_encoder.SRCS       += tools_common.h tools_common.c
+vp9_lossless_encoder.SRCS       += video_common.h
+vp9_lossless_encoder.SRCS       += video_writer.h video_writer.c
+vp9_lossless_encoder.GUID        = B63C7C88-5348-46DC-A5A6-CC151EF93366
+vp9_lossless_encoder.DESCRIPTION = Simplified lossless VP9 encoder
 EXAMPLES-$(CONFIG_ENCODERS)     += twopass_encoder.c
 twopass_encoder.SRCS            += ivfenc.h ivfenc.c
 twopass_encoder.SRCS            += tools_common.h tools_common.c
 twopass_encoder.SRCS            += video_common.h
 twopass_encoder.SRCS            += video_writer.h video_writer.c
-twopass_encoder.SRCS            += vpx_ports/msvc.h
 twopass_encoder.GUID             = 73494FA6-4AF9-4763-8FBB-265C92402FD8
 twopass_encoder.DESCRIPTION      = Two-pass encoder loop
 EXAMPLES-$(CONFIG_DECODERS)     += decode_with_drops.c
@@ -169,7 +170,6 @@ decode_with_drops.SRCS          += video_common.h
 decode_with_drops.SRCS          += video_reader.h video_reader.c
 decode_with_drops.SRCS          += vpx_ports/mem_ops.h
 decode_with_drops.SRCS          += vpx_ports/mem_ops_aligned.h
-decode_with_drops.SRCS          += vpx_ports/msvc.h
 decode_with_drops.GUID           = CE5C53C4-8DDA-438A-86ED-0DDD3CDB8D26
 decode_with_drops.DESCRIPTION    = Drops frames while decoding
 EXAMPLES-$(CONFIG_ENCODERS)        += set_maps.c
@@ -177,33 +177,40 @@ set_maps.SRCS                      += ivfenc.h ivfenc.c
 set_maps.SRCS                      += tools_common.h tools_common.c
 set_maps.SRCS                      += video_common.h
 set_maps.SRCS                      += video_writer.h video_writer.c
-set_maps.SRCS                      += vpx_ports/msvc.h
 set_maps.GUID                       = ECB2D24D-98B8-4015-A465-A4AF3DCC145F
 set_maps.DESCRIPTION                = Set active and ROI maps
+EXAMPLES-$(CONFIG_VP8_ENCODER)     += vp8cx_set_ref.c
+vp8cx_set_ref.SRCS                 += ivfenc.h ivfenc.c
+vp8cx_set_ref.SRCS                 += tools_common.h tools_common.c
+vp8cx_set_ref.SRCS                 += video_common.h
+vp8cx_set_ref.SRCS                 += video_writer.h video_writer.c
+vp8cx_set_ref.GUID                  = C5E31F7F-96F6-48BD-BD3E-10EBF6E8057A
+vp8cx_set_ref.DESCRIPTION           = VP8 set encoder reference frame


-ifeq ($(CONFIG_VP10_ENCODER), yes)
-ifeq ($(CONFIG_DECODERS),yes)
-EXAMPLES-yes                       += vpxcx_set_ref.c
-vpxcx_set_ref.SRCS                 += ivfenc.h ivfenc.c
-vpxcx_set_ref.SRCS                 += tools_common.h tools_common.c
-vpxcx_set_ref.SRCS                 += video_common.h
-vpxcx_set_ref.SRCS                 += video_writer.h video_writer.c
-vpxcx_set_ref.GUID                  = 65D7F14A-2EE6-4293-B958-AB5107A03B55
-vpxcx_set_ref.DESCRIPTION           = VP10 set encoder reference frame
+ifeq ($(CONFIG_MULTI_RES_ENCODING),yes)
+ifeq ($(CONFIG_LIBYUV),yes)
+EXAMPLES-$(CONFIG_VP8_ENCODER)          += vp8_multi_resolution_encoder.c
+vp8_multi_resolution_encoder.SRCS       += ivfenc.h ivfenc.c
+vp8_multi_resolution_encoder.SRCS       += tools_common.h tools_common.c
+vp8_multi_resolution_encoder.SRCS       += video_writer.h video_writer.c
+vp8_multi_resolution_encoder.SRCS       += $(LIBYUV_SRCS)
+vp8_multi_resolution_encoder.GUID        = 04f8738e-63c8-423b-90fa-7c2703a374de
+vp8_multi_resolution_encoder.DESCRIPTION = VP8 Multiple-resolution Encoding
 endif
 endif

-
 # Handle extra library flags depending on codec configuration

 # We should not link to math library (libm) on RVCT
 # when building for bare-metal targets
 ifeq ($(CONFIG_OS_SUPPORT), yes)
-CODEC_EXTRA_LIBS-$(CONFIG_VP10)        += m
+CODEC_EXTRA_LIBS-$(CONFIG_VP8)         += m
+CODEC_EXTRA_LIBS-$(CONFIG_VP9)         += m
 else
    ifeq ($(CONFIG_GCC), yes)
-    CODEC_EXTRA_LIBS-$(CONFIG_VP10)        += m
+    CODEC_EXTRA_LIBS-$(CONFIG_VP8)         += m
+    CODEC_EXTRA_LIBS-$(CONFIG_VP9)         += m
    endif
 endif
 #
@@ -220,8 +227,10 @@ ifeq ($(HAVE_ALT_TREE_LAYOUT),yes)
    INC_PATH-yes := $(SRC_PATH_BARE)/../include
 else
    LIB_PATH-yes                     += $(if $(BUILD_PFX),$(BUILD_PFX),.)
-    INC_PATH-$(CONFIG_VP10_DECODER)   += $(SRC_PATH_BARE)/vp10
-    INC_PATH-$(CONFIG_VP10_ENCODER)   += $(SRC_PATH_BARE)/vp10
+    INC_PATH-$(CONFIG_VP8_DECODER)   += $(SRC_PATH_BARE)/vp8
+    INC_PATH-$(CONFIG_VP8_ENCODER)   += $(SRC_PATH_BARE)/vp8
+    INC_PATH-$(CONFIG_VP9_DECODER)   += $(SRC_PATH_BARE)/vp9
+    INC_PATH-$(CONFIG_VP9_ENCODER)   += $(SRC_PATH_BARE)/vp9
 endif
 INC_PATH-$(CONFIG_LIBYUV) += $(SRC_PATH_BARE)/third_party/libyuv/include
 LIB_PATH := $(call enabled,LIB_PATH)
@@ -245,6 +254,14 @@ CODEC_EXTRA_LIBS=$(sort $(call enabled,CODEC_EXTRA_LIBS))
 $(foreach ex,$(ALL_EXAMPLES),$(eval $(notdir $(ex:.c=)).SRCS += $(ex) examples.mk))


+# If this is a universal (fat) binary, then all the subarchitectures have
+# already been built and our job is to stitch them together. The
+# BUILD_OBJS variable indicates whether we should be building
+# (compiling, linking) the library. The LIPO_OBJS variable indicates
+# that we're stitching.
+$(eval $(if $(filter universal%,$(TOOLCHAIN)),LIPO_OBJS,BUILD_OBJS):=yes)
+
+
 # Create build/install dependencies for all examples. The common case
 # is handled here. The MSVS case is handled below.
 NOT_MSVS = $(if $(CONFIG_MSVS),,yes)
@@ -252,28 +269,24 @@ DIST-BINS-$(NOT_MSVS)      += $(addprefix bin/,$(ALL_EXAMPLES:.c=$(EXE_SFX)))
 INSTALL-BINS-$(NOT_MSVS)   += $(addprefix bin/,$(UTILS:.c=$(EXE_SFX)))
 DIST-SRCS-yes              += $(ALL_SRCS)
 INSTALL-SRCS-yes           += $(UTIL_SRCS)
-OBJS-$(NOT_MSVS)           += $(call objs,$(ALL_SRCS))
+OBJS-$(NOT_MSVS)           += $(if $(BUILD_OBJS),$(call objs,$(ALL_SRCS)))
 BINS-$(NOT_MSVS)           += $(addprefix $(BUILD_PFX),$(ALL_EXAMPLES:.c=$(EXE_SFX)))


 # Instantiate linker template for all examples.
 CODEC_LIB=$(if $(CONFIG_DEBUG_LIBS),vpx_g,vpx)
-ifneq ($(filter darwin%,$(TGT_OS)),)
-SHARED_LIB_SUF=.dylib
-else
-ifneq ($(filter os2%,$(TGT_OS)),)
-SHARED_LIB_SUF=_dll.a
-else
-SHARED_LIB_SUF=.so
-endif
-endif
+SHARED_LIB_SUF=$(if $(filter darwin%,$(TGT_OS)),.dylib,.so)
 CODEC_LIB_SUF=$(if $(CONFIG_SHARED),$(SHARED_LIB_SUF),.a)
 $(foreach bin,$(BINS-yes),\
-    $(eval $(bin):$(LIB_PATH)/lib$(CODEC_LIB)$(CODEC_LIB_SUF))\
-    $(eval $(call linker_template,$(bin),\
+    $(if $(BUILD_OBJS),$(eval $(bin):\
+        $(LIB_PATH)/lib$(CODEC_LIB)$(CODEC_LIB_SUF)))\
+    $(if $(BUILD_OBJS),$(eval $(call linker_template,$(bin),\
        $(call objs,$($(notdir $(bin:$(EXE_SFX)=)).SRCS)) \
        -l$(CODEC_LIB) $(addprefix -l,$(CODEC_EXTRA_LIBS))\
-        )))
+        )))\
+    $(if $(LIPO_OBJS),$(eval $(call lipo_bin_template,$(bin))))\
+    )
+

 # The following pairs define a mapping of locations in the distribution
 # tree to locations in the source/build trees.
@@ -301,8 +314,8 @@ endif
 # the makefiles). We may want to revisit this.
 define vcproj_template
 $(1): $($(1:.$(VCPROJ_SFX)=).SRCS) vpx.$(VCPROJ_SFX)
-	$(if $(quiet),@echo "    [vcproj] $$@")
-	$(qexec)$$(GEN_VCPROJ)\
+	@echo "    [vcproj] $$@"
+	$$(GEN_VCPROJ)\
            --exe\
            --target=$$(TOOLCHAIN)\
            --name=$$(@:.$(VCPROJ_SFX)=)\
--- a/examples/decode_to_md5.c
+++ b/examples/decode_to_md5.c
@@ -71,7 +71,7 @@ static void print_md5(FILE *stream, unsigned char digest[16]) {

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <infile> <outfile>\n", exec_name);
  exit(EXIT_FAILURE);
 }
--- a/examples/decode_with_drops.c
+++ b/examples/decode_with_drops.c
@@ -65,7 +65,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <infile> <outfile> <N-M|N/M>\n", exec_name);
  exit(EXIT_FAILURE);
 }
--- a/examples/postproc.c
+++ b/examples/postproc.c
@@ -0,0 +1,138 @@
+/*
+ *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+// Postprocessing Decoder
+// ======================
+//
+// This example adds postprocessing to the simple decoder loop.
+//
+// Initializing Postprocessing
+// ---------------------------
+// You must inform the codec that you might request postprocessing at
+// initialization time. This is done by passing the VPX_CODEC_USE_POSTPROC
+// flag to `vpx_codec_dec_init`. If the codec does not support
+// postprocessing, this call will return VPX_CODEC_INCAPABLE. For
+// demonstration purposes, we also fall back to default initialization if
+// the codec does not provide support.
+//
+// Using Adaptive Postprocessing
+// -----------------------------
+// VP6 provides "adaptive postprocessing." It will automatically select the
+// best postprocessing filter on a frame by frame basis based on the amount
+// of time remaining before the user's specified deadline expires. The
+// special value 0 indicates that the codec should take as long as
+// necessary to provide the best quality frame. This example gives the
+// codec 15ms (15000us) to return a frame. Remember that this is a soft
+// deadline, and the codec may exceed it doing its regular processing. In
+// these cases, no additional postprocessing will be done.
+//
+// Codec Specific Postprocessing Controls
+// --------------------------------------
+// Some codecs provide fine grained controls over their built-in
+// postprocessors. VP8 is one example. The following sample code toggles
+// postprocessing on and off every 15 frames.
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "vpx/vp8dx.h"
+#include "vpx/vpx_decoder.h"
+
+#include "../tools_common.h"
+#include "../video_reader.h"
+#include "./vpx_config.h"
+
+static const char *exec_name;
+
+void usage_exit() {
+  fprintf(stderr, "Usage: %s <infile> <outfile>\n", exec_name);
+  exit(EXIT_FAILURE);
+}
+
+int main(int argc, char **argv) {
+  int frame_cnt = 0;
+  FILE *outfile = NULL;
+  vpx_codec_ctx_t codec;
+  vpx_codec_err_t res;
+  VpxVideoReader *reader = NULL;
+  const VpxInterface *decoder = NULL;
+  const VpxVideoInfo *info = NULL;
+
+  exec_name = argv[0];
+
+  if (argc != 3)
+    die("Invalid number of arguments.");
+
+  reader = vpx_video_reader_open(argv[1]);
+  if (!reader)
+    die("Failed to open %s for reading.", argv[1]);
+
+  if (!(outfile = fopen(argv[2], "wb")))
+    die("Failed to open %s for writing", argv[2]);
+
+  info = vpx_video_reader_get_info(reader);
+
+  decoder = get_vpx_decoder_by_fourcc(info->codec_fourcc);
+  if (!decoder)
+    die("Unknown input codec.");
+
+  printf("Using %s\n", vpx_codec_iface_name(decoder->codec_interface()));
+
+  res = vpx_codec_dec_init(&codec, decoder->codec_interface(), NULL,
+                           VPX_CODEC_USE_POSTPROC);
+  if (res == VPX_CODEC_INCAPABLE)
+    die_codec(&codec, "Postproc not supported by this decoder.");
+
+  if (res)
+    die_codec(&codec, "Failed to initialize decoder.");
+
+  while (vpx_video_reader_read_frame(reader)) {
+    vpx_codec_iter_t iter = NULL;
+    vpx_image_t *img = NULL;
+    size_t frame_size = 0;
+    const unsigned char *frame = vpx_video_reader_get_frame(reader,
+                                                            &frame_size);
+
+    ++frame_cnt;
+
+    if (frame_cnt % 30 == 1) {
+      vp8_postproc_cfg_t pp = {0, 0, 0};
+
+    if (vpx_codec_control(&codec, VP8_SET_POSTPROC, &pp))
+      die_codec(&codec, "Failed to turn off postproc.");
+    } else if (frame_cnt % 30 == 16) {
+      vp8_postproc_cfg_t pp = {VP8_DEBLOCK | VP8_DEMACROBLOCK | VP8_MFQE,
+                               4, 0};
+      if (vpx_codec_control(&codec, VP8_SET_POSTPROC, &pp))
+        die_codec(&codec, "Failed to turn on postproc.");
+    };
+
+    // Decode the frame with 15ms deadline
+    if (vpx_codec_decode(&codec, frame, (unsigned int)frame_size, NULL, 15000))
+      die_codec(&codec, "Failed to decode frame");
+
+    while ((img = vpx_codec_get_frame(&codec, &iter)) != NULL) {
+      vpx_img_write(img, outfile);
+    }
+  }
+
+  printf("Processed %d frames.\n", frame_cnt);
+  if (vpx_codec_destroy(&codec))
+    die_codec(&codec, "Failed to destroy codec");
+
+  printf("Play: ffplay -f rawvideo -pix_fmt yuv420p -s %dx%d %s\n",
+         info->frame_width, info->frame_height, argv[2]);
+
+  vpx_video_reader_close(reader);
+
+  fclose(outfile);
+  return EXIT_SUCCESS;
+}
--- a/examples/resize_util.c
+++ b/examples/resize_util.c
@@ -15,7 +15,6 @@
 #include <stdlib.h>
 #include <string.h>

-#include "../tools_common.h"
 #include "../vp9/encoder/vp9_resize.h"

 static const char *exec_name = NULL;
@@ -27,7 +26,7 @@ static void usage() {
  printf("<output_yuv> [<frames>]\n");
 }

-void usage_exit(void) {
+void usage_exit() {
  usage();
  exit(EXIT_FAILURE);
 }
--- a/examples/set_maps.c
+++ b/examples/set_maps.c
@@ -55,7 +55,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <codec> <width> <height> <infile> <outfile>\n",
          exec_name);
  exit(EXIT_FAILURE);
--- a/examples/simple_decoder.c
+++ b/examples/simple_decoder.c
@@ -88,7 +88,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <infile> <outfile>\n", exec_name);
  exit(EXIT_FAILURE);
 }
--- a/examples/simple_encoder.c
+++ b/examples/simple_encoder.c
@@ -106,11 +106,11 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr,
          "Usage: %s <codec> <width> <height> <infile> <outfile> "
-              "<keyframe-interval> <error-resilient> <frames to encode>\n"
-              "See comments in simple_encoder.c for more information.\n",
+              "<keyframe-interval> [<error-resilient>]\nSee comments in "
+              "simple_encoder.c for more information.\n",
          exec_name);
  exit(EXIT_FAILURE);
 }
@@ -147,7 +147,6 @@ static int encode_frame(vpx_codec_ctx_t *codec,
  return got_pkts;
 }

-// TODO(tomfinegan): Improve command line parsing and add args for bitrate/fps.
 int main(int argc, char **argv) {
  FILE *infile = NULL;
  vpx_codec_ctx_t codec;
@@ -158,11 +157,12 @@ int main(int argc, char **argv) {
  VpxVideoInfo info = {0};
  VpxVideoWriter *writer = NULL;
  const VpxInterface *encoder = NULL;
-  const int fps = 30;
-  const int bitrate = 200;
+  const int fps = 30;        // TODO(dkovalev) add command line argument
+  const int bitrate = 200;   // kbit/s TODO(dkovalev) add command line argument
  int keyframe_interval = 0;
-  int max_frames = 0;
-  int frames_encoded = 0;
+
+  // TODO(dkovalev): Add some simple command line parsing code to make the
+  // command line more flexible.
  const char *codec_arg = NULL;
  const char *width_arg = NULL;
  const char *height_arg = NULL;
@@ -172,7 +172,7 @@ int main(int argc, char **argv) {

  exec_name = argv[0];

-  if (argc != 9)
+  if (argc < 7)
    die("Invalid number of arguments");

  codec_arg = argv[1];
@@ -181,7 +181,6 @@ int main(int argc, char **argv) {
  infile_arg = argv[4];
  outfile_arg = argv[5];
  keyframe_interval_arg = argv[6];
-  max_frames = strtol(argv[8], NULL, 0);

  encoder = get_vpx_encoder_by_name(codec_arg);
  if (!encoder)
@@ -220,7 +219,7 @@ int main(int argc, char **argv) {
  cfg.g_timebase.num = info.time_base.numerator;
  cfg.g_timebase.den = info.time_base.denominator;
  cfg.rc_target_bitrate = bitrate;
-  cfg.g_error_resilient = strtol(argv[7], NULL, 0);
+  cfg.g_error_resilient = argc > 7 ? strtol(argv[7], NULL, 0) : 0;

  writer = vpx_video_writer_open(outfile_arg, kContainerIVF, &info);
  if (!writer)
@@ -238,9 +237,6 @@ int main(int argc, char **argv) {
    if (keyframe_interval > 0 && frame_count % keyframe_interval == 0)
      flags |= VPX_EFLAG_FORCE_KF;
    encode_frame(&codec, &raw, frame_count++, flags, writer);
-    frames_encoded++;
-    if (max_frames > 0 && frames_encoded >= max_frames)
-      break;
  }

  // Flush encoder.
--- a/examples/twopass_encoder.c
+++ b/examples/twopass_encoder.c
@@ -58,10 +58,8 @@

 static const char *exec_name;

-void usage_exit(void) {
-  fprintf(stderr,
-          "Usage: %s <codec> <width> <height> <infile> <outfile> "
-              "<frame limit>\n",
+void usage_exit() {
+  fprintf(stderr, "Usage: %s <codec> <width> <height> <infile> <outfile>\n",
          exec_name);
  exit(EXIT_FAILURE);
 }
@@ -131,8 +129,7 @@ static int encode_frame(vpx_codec_ctx_t *ctx,
 static vpx_fixed_buf_t pass0(vpx_image_t *raw,
                             FILE *infile,
                             const VpxInterface *encoder,
-                             const vpx_codec_enc_cfg_t *cfg,
-                             int max_frames) {
+                             const vpx_codec_enc_cfg_t *cfg) {
  vpx_codec_ctx_t codec;
  int frame_count = 0;
  vpx_fixed_buf_t stats = {NULL, 0};
@@ -145,8 +142,6 @@ static vpx_fixed_buf_t pass0(vpx_image_t *raw,
    ++frame_count;
    get_frame_stats(&codec, raw, frame_count, 1, 0, VPX_DL_GOOD_QUALITY,
                    &stats);
-    if (max_frames > 0 && frame_count >= max_frames)
-      break;
  }

  // Flush encoder.
@@ -164,8 +159,7 @@ static void pass1(vpx_image_t *raw,
                  FILE *infile,
                  const char *outfile_name,
                  const VpxInterface *encoder,
-                  const vpx_codec_enc_cfg_t *cfg,
-                  int max_frames) {
+                  const vpx_codec_enc_cfg_t *cfg) {
  VpxVideoInfo info = {
    encoder->fourcc,
    cfg->g_w,
@@ -187,9 +181,6 @@ static void pass1(vpx_image_t *raw,
  while (vpx_img_read(raw, infile)) {
    ++frame_count;
    encode_frame(&codec, raw, frame_count, 1, 0, VPX_DL_GOOD_QUALITY, writer);
-
-    if (max_frames > 0 && frame_count >= max_frames)
-      break;
  }

  // Flush encoder.
@@ -222,14 +213,11 @@ int main(int argc, char **argv) {
  const char *const height_arg = argv[3];
  const char *const infile_arg = argv[4];
  const char *const outfile_arg = argv[5];
-  int max_frames = 0;
  exec_name = argv[0];

-  if (argc != 7)
+  if (argc != 6)
    die("Invalid number of arguments.");

-  max_frames = strtol(argv[6], NULL, 0);
-
  encoder = get_vpx_encoder_by_name(codec_arg);
  if (!encoder)
    die("Unsupported codec.");
@@ -261,13 +249,13 @@ int main(int argc, char **argv) {

  // Pass 0
  cfg.g_pass = VPX_RC_FIRST_PASS;
-  stats = pass0(&raw, infile, encoder, &cfg, max_frames);
+  stats = pass0(&raw, infile, encoder, &cfg);

  // Pass 1
  rewind(infile);
  cfg.g_pass = VPX_RC_LAST_PASS;
  cfg.rc_twopass_stats_in = stats;
-  pass1(&raw, infile, outfile_arg, encoder, &cfg, max_frames);
+  pass1(&raw, infile, outfile_arg, encoder, &cfg);
  free(stats.buf);

  vpx_img_free(&raw);
--- a/examples/vp8_multi_resolution_encoder.c
+++ b/examples/vp8_multi_resolution_encoder.c
@@ -0,0 +1,737 @@
+/*
+ *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+/*
+ * This is an example demonstrating multi-resolution encoding in VP8.
+ * High-resolution input video is down-sampled to lower-resolutions. The
+ * encoder then encodes the video and outputs multiple bitstreams with
+ * different resolutions.
+ *
+ * This test also allows for settings temporal layers for each spatial layer.
+ * Different number of temporal layers per spatial stream may be used.
+ * Currently up to 3 temporal layers per spatial stream (encoder) are supported
+ * in this test.
+ */
+
+#include "./vpx_config.h"
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdarg.h>
+#include <string.h>
+#include <math.h>
+#include <assert.h>
+#include <sys/time.h>
+#if USE_POSIX_MMAP
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <sys/mman.h>
+#include <fcntl.h>
+#include <unistd.h>
+#endif
+#include "vpx_ports/vpx_timer.h"
+#define VPX_CODEC_DISABLE_COMPAT 1
+#include "vpx/vpx_encoder.h"
+#include "vpx/vp8cx.h"
+#include "vpx_ports/mem_ops.h"
+#include "./tools_common.h"
+#define interface (vpx_codec_vp8_cx())
+#define fourcc    0x30385056
+
+void usage_exit() {
+  exit(EXIT_FAILURE);
+}
+
+/*
+ * The input video frame is downsampled several times to generate a multi-level
+ * hierarchical structure. NUM_ENCODERS is defined as the number of encoding
+ * levels required. For example, if the size of input video is 1280x720,
+ * NUM_ENCODERS is 3, and down-sampling factor is 2, the encoder outputs 3
+ * bitstreams with resolution of 1280x720(level 0), 640x360(level 1), and
+ * 320x180(level 2) respectively.
+ */
+
+/* Number of encoders (spatial resolutions) used in this test. */
+#define NUM_ENCODERS 3
+
+/* Maximum number of temporal layers allowed for this test. */
+#define MAX_NUM_TEMPORAL_LAYERS 3
+
+/* This example uses the scaler function in libyuv. */
+#include "third_party/libyuv/include/libyuv/basic_types.h"
+#include "third_party/libyuv/include/libyuv/scale.h"
+#include "third_party/libyuv/include/libyuv/cpu_id.h"
+
+int (*read_frame_p)(FILE *f, vpx_image_t *img);
+
+static int read_frame(FILE *f, vpx_image_t *img) {
+    size_t nbytes, to_read;
+    int    res = 1;
+
+    to_read = img->w*img->h*3/2;
+    nbytes = fread(img->planes[0], 1, to_read, f);
+    if(nbytes != to_read) {
+        res = 0;
+        if(nbytes > 0)
+            printf("Warning: Read partial frame. Check your width & height!\n");
+    }
+    return res;
+}
+
+static int read_frame_by_row(FILE *f, vpx_image_t *img) {
+    size_t nbytes, to_read;
+    int    res = 1;
+    int plane;
+
+    for (plane = 0; plane < 3; plane++)
+    {
+        unsigned char *ptr;
+        int w = (plane ? (1 + img->d_w) / 2 : img->d_w);
+        int h = (plane ? (1 + img->d_h) / 2 : img->d_h);
+        int r;
+
+        /* Determine the correct plane based on the image format. The for-loop
+         * always counts in Y,U,V order, but this may not match the order of
+         * the data on disk.
+         */
+        switch (plane)
+        {
+        case 1:
+            ptr = img->planes[img->fmt==VPX_IMG_FMT_YV12? VPX_PLANE_V : VPX_PLANE_U];
+            break;
+        case 2:
+            ptr = img->planes[img->fmt==VPX_IMG_FMT_YV12?VPX_PLANE_U : VPX_PLANE_V];
+            break;
+        default:
+            ptr = img->planes[plane];
+        }
+
+        for (r = 0; r < h; r++)
+        {
+            to_read = w;
+
+            nbytes = fread(ptr, 1, to_read, f);
+            if(nbytes != to_read) {
+                res = 0;
+                if(nbytes > 0)
+                    printf("Warning: Read partial frame. Check your width & height!\n");
+                break;
+            }
+
+            ptr += img->stride[plane];
+        }
+        if (!res)
+            break;
+    }
+
+    return res;
+}
+
+static void write_ivf_file_header(FILE *outfile,
+                                  const vpx_codec_enc_cfg_t *cfg,
+                                  int frame_cnt) {
+    char header[32];
+
+    if(cfg->g_pass != VPX_RC_ONE_PASS && cfg->g_pass != VPX_RC_LAST_PASS)
+        return;
+    header[0] = 'D';
+    header[1] = 'K';
+    header[2] = 'I';
+    header[3] = 'F';
+    mem_put_le16(header+4,  0);                   /* version */
+    mem_put_le16(header+6,  32);                  /* headersize */
+    mem_put_le32(header+8,  fourcc);              /* headersize */
+    mem_put_le16(header+12, cfg->g_w);            /* width */
+    mem_put_le16(header+14, cfg->g_h);            /* height */
+    mem_put_le32(header+16, cfg->g_timebase.den); /* rate */
+    mem_put_le32(header+20, cfg->g_timebase.num); /* scale */
+    mem_put_le32(header+24, frame_cnt);           /* length */
+    mem_put_le32(header+28, 0);                   /* unused */
+
+    (void) fwrite(header, 1, 32, outfile);
+}
+
+static void write_ivf_frame_header(FILE *outfile,
+                                   const vpx_codec_cx_pkt_t *pkt)
+{
+    char             header[12];
+    vpx_codec_pts_t  pts;
+
+    if(pkt->kind != VPX_CODEC_CX_FRAME_PKT)
+        return;
+
+    pts = pkt->data.frame.pts;
+    mem_put_le32(header, pkt->data.frame.sz);
+    mem_put_le32(header+4, pts&0xFFFFFFFF);
+    mem_put_le32(header+8, pts >> 32);
+
+    (void) fwrite(header, 1, 12, outfile);
+}
+
+/* Temporal scaling parameters */
+/* This sets all the temporal layer parameters given |num_temporal_layers|,
+ * including the target bit allocation across temporal layers. Bit allocation
+ * parameters will be passed in as user parameters in another version.
+ */
+static void set_temporal_layer_pattern(int num_temporal_layers,
+                                       vpx_codec_enc_cfg_t *cfg,
+                                       int bitrate,
+                                       int *layer_flags)
+{
+    assert(num_temporal_layers <= MAX_NUM_TEMPORAL_LAYERS);
+    switch (num_temporal_layers)
+    {
+    case 1:
+    {
+        /* 1-layer */
+        cfg->ts_number_layers     = 1;
+        cfg->ts_periodicity       = 1;
+        cfg->ts_rate_decimator[0] = 1;
+        cfg->ts_layer_id[0] = 0;
+        cfg->ts_target_bitrate[0] = bitrate;
+
+        // Update L only.
+        layer_flags[0] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
+        break;
+    }
+
+    case 2:
+    {
+        /* 2-layers, with sync point at first frame of layer 1. */
+        cfg->ts_number_layers     = 2;
+        cfg->ts_periodicity       = 2;
+        cfg->ts_rate_decimator[0] = 2;
+        cfg->ts_rate_decimator[1] = 1;
+        cfg->ts_layer_id[0] = 0;
+        cfg->ts_layer_id[1] = 1;
+        // Use 60/40 bit allocation as example.
+        cfg->ts_target_bitrate[0] = 0.6f * bitrate;
+        cfg->ts_target_bitrate[1] = bitrate;
+
+        /* 0=L, 1=GF */
+        // ARF is used as predictor for all frames, and is only updated on
+        // key frame. Sync point every 8 frames.
+
+        // Layer 0: predict from L and ARF, update L and G.
+        layer_flags[0] = VP8_EFLAG_NO_REF_GF |
+                         VP8_EFLAG_NO_UPD_ARF;
+
+        // Layer 1: sync point: predict from L and ARF, and update G.
+        layer_flags[1] = VP8_EFLAG_NO_REF_GF |
+                         VP8_EFLAG_NO_UPD_LAST |
+                         VP8_EFLAG_NO_UPD_ARF;
+
+        // Layer 0, predict from L and ARF, update L.
+        layer_flags[2] = VP8_EFLAG_NO_REF_GF  |
+                         VP8_EFLAG_NO_UPD_GF  |
+                         VP8_EFLAG_NO_UPD_ARF;
+
+        // Layer 1: predict from L, G and ARF, and update G.
+        layer_flags[3] = VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_UPD_LAST |
+                         VP8_EFLAG_NO_UPD_ENTROPY;
+
+        // Layer 0
+        layer_flags[4] = layer_flags[2];
+
+        // Layer 1
+        layer_flags[5] = layer_flags[3];
+
+        // Layer 0
+        layer_flags[6] = layer_flags[4];
+
+        // Layer 1
+        layer_flags[7] = layer_flags[5];
+        break;
+    }
+
+    case 3:
+    default:
+    {
+        // 3-layers structure where ARF is used as predictor for all frames,
+        // and is only updated on key frame.
+        // Sync points for layer 1 and 2 every 8 frames.
+        cfg->ts_number_layers     = 3;
+        cfg->ts_periodicity       = 4;
+        cfg->ts_rate_decimator[0] = 4;
+        cfg->ts_rate_decimator[1] = 2;
+        cfg->ts_rate_decimator[2] = 1;
+        cfg->ts_layer_id[0] = 0;
+        cfg->ts_layer_id[1] = 2;
+        cfg->ts_layer_id[2] = 1;
+        cfg->ts_layer_id[3] = 2;
+        // Use 40/20/40 bit allocation as example.
+        cfg->ts_target_bitrate[0] = 0.4f * bitrate;
+        cfg->ts_target_bitrate[1] = 0.6f * bitrate;
+        cfg->ts_target_bitrate[2] = bitrate;
+
+        /* 0=L, 1=GF, 2=ARF */
+
+        // Layer 0: predict from L and ARF; update L and G.
+        layer_flags[0] =  VP8_EFLAG_NO_UPD_ARF |
+                          VP8_EFLAG_NO_REF_GF;
+
+        // Layer 2: sync point: predict from L and ARF; update none.
+        layer_flags[1] = VP8_EFLAG_NO_REF_GF |
+                         VP8_EFLAG_NO_UPD_GF |
+                         VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_UPD_LAST |
+                         VP8_EFLAG_NO_UPD_ENTROPY;
+
+        // Layer 1: sync point: predict from L and ARF; update G.
+        layer_flags[2] = VP8_EFLAG_NO_REF_GF |
+                         VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_UPD_LAST;
+
+        // Layer 2: predict from L, G, ARF; update none.
+        layer_flags[3] = VP8_EFLAG_NO_UPD_GF |
+                         VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_UPD_LAST |
+                         VP8_EFLAG_NO_UPD_ENTROPY;
+
+        // Layer 0: predict from L and ARF; update L.
+        layer_flags[4] = VP8_EFLAG_NO_UPD_GF |
+                         VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_REF_GF;
+
+        // Layer 2: predict from L, G, ARF; update none.
+        layer_flags[5] = layer_flags[3];
+
+        // Layer 1: predict from L, G, ARF; update G.
+        layer_flags[6] = VP8_EFLAG_NO_UPD_ARF |
+                         VP8_EFLAG_NO_UPD_LAST;
+
+        // Layer 2: predict from L, G, ARF; update none.
+        layer_flags[7] = layer_flags[3];
+        break;
+    }
+    }
+}
+
+/* The periodicity of the pattern given the number of temporal layers. */
+static int periodicity_to_num_layers[MAX_NUM_TEMPORAL_LAYERS] = {1, 8, 8};
+
+int main(int argc, char **argv)
+{
+    FILE                 *infile, *outfile[NUM_ENCODERS];
+    FILE                 *downsampled_input[NUM_ENCODERS - 1];
+    char                 filename[50];
+    vpx_codec_ctx_t      codec[NUM_ENCODERS];
+    vpx_codec_enc_cfg_t  cfg[NUM_ENCODERS];
+    int                  frame_cnt = 0;
+    vpx_image_t          raw[NUM_ENCODERS];
+    vpx_codec_err_t      res[NUM_ENCODERS];
+
+    int                  i;
+    long                 width;
+    long                 height;
+    int                  length_frame;
+    int                  frame_avail;
+    int                  got_data;
+    int                  flags = 0;
+    int                  layer_id = 0;
+
+    int                  layer_flags[VPX_TS_MAX_PERIODICITY * NUM_ENCODERS]
+                                     = {0};
+    int                  flag_periodicity;
+
+    /*Currently, only realtime mode is supported in multi-resolution encoding.*/
+    int                  arg_deadline = VPX_DL_REALTIME;
+
+    /* Set show_psnr to 1/0 to show/not show PSNR. Choose show_psnr=0 if you
+       don't need to know PSNR, which will skip PSNR calculation and save
+       encoding time. */
+    int                  show_psnr = 0;
+    int                  key_frame_insert = 0;
+    uint64_t             psnr_sse_total[NUM_ENCODERS] = {0};
+    uint64_t             psnr_samples_total[NUM_ENCODERS] = {0};
+    double               psnr_totals[NUM_ENCODERS][4] = {{0,0}};
+    int                  psnr_count[NUM_ENCODERS] = {0};
+
+    double               cx_time = 0;
+    struct  timeval      tv1, tv2, difftv;
+
+    /* Set the required target bitrates for each resolution level.
+     * If target bitrate for highest-resolution level is set to 0,
+     * (i.e. target_bitrate[0]=0), we skip encoding at that level.
+     */
+    unsigned int         target_bitrate[NUM_ENCODERS]={1000, 500, 100};
+
+    /* Enter the frame rate of the input video */
+    int                  framerate = 30;
+
+    /* Set down-sampling factor for each resolution level.
+       dsf[0] controls down sampling from level 0 to level 1;
+       dsf[1] controls down sampling from level 1 to level 2;
+       dsf[2] is not used. */
+    vpx_rational_t dsf[NUM_ENCODERS] = {{2, 1}, {2, 1}, {1, 1}};
+
+    /* Set the number of temporal layers for each encoder/resolution level,
+     * starting from highest resoln down to lowest resoln. */
+    unsigned int         num_temporal_layers[NUM_ENCODERS] = {3, 3, 3};
+
+    if(argc!= (7 + 3 * NUM_ENCODERS))
+        die("Usage: %s <width> <height> <frame_rate>  <infile> <outfile(s)> "
+            "<rate_encoder(s)> <temporal_layer(s)> <key_frame_insert> <output psnr?> \n",
+            argv[0]);
+
+    printf("Using %s\n",vpx_codec_iface_name(interface));
+
+    width = strtol(argv[1], NULL, 0);
+    height = strtol(argv[2], NULL, 0);
+    framerate = strtol(argv[3], NULL, 0);
+
+    if(width < 16 || width%2 || height <16 || height%2)
+        die("Invalid resolution: %ldx%ld", width, height);
+
+    /* Open input video file for encoding */
+    if(!(infile = fopen(argv[4], "rb")))
+        die("Failed to open %s for reading", argv[4]);
+
+    /* Open output file for each encoder to output bitstreams */
+    for (i=0; i< NUM_ENCODERS; i++)
+    {
+        if(!target_bitrate[i])
+        {
+            outfile[i] = NULL;
+            continue;
+        }
+
+        if(!(outfile[i] = fopen(argv[i+5], "wb")))
+            die("Failed to open %s for writing", argv[i+4]);
+    }
+
+    // Bitrates per spatial layer: overwrite default rates above.
+    for (i=0; i< NUM_ENCODERS; i++)
+    {
+        target_bitrate[i] = strtol(argv[NUM_ENCODERS + 5 + i], NULL, 0);
+    }
+
+    // Temporal layers per spatial layers: overwrite default settings above.
+    for (i=0; i< NUM_ENCODERS; i++)
+    {
+        num_temporal_layers[i] = strtol(argv[2 * NUM_ENCODERS + 5 + i], NULL, 0);
+        if (num_temporal_layers[i] < 1 || num_temporal_layers[i] > 3)
+          die("Invalid temporal layers: %d, Must be 1, 2, or 3. \n",
+              num_temporal_layers);
+    }
+
+    /* Open file to write out each spatially downsampled input stream. */
+    for (i=0; i< NUM_ENCODERS - 1; i++)
+    {
+       // Highest resoln is encoder 0.
+        if (sprintf(filename,"ds%d.yuv",NUM_ENCODERS - i) < 0)
+        {
+            return EXIT_FAILURE;
+        }
+        downsampled_input[i] = fopen(filename,"wb");
+    }
+
+    key_frame_insert = strtol(argv[3 * NUM_ENCODERS + 5], NULL, 0);
+
+    show_psnr = strtol(argv[3 * NUM_ENCODERS + 6], NULL, 0);
+
+
+    /* Populate default encoder configuration */
+    for (i=0; i< NUM_ENCODERS; i++)
+    {
+        res[i] = vpx_codec_enc_config_default(interface, &cfg[i], 0);
+        if(res[i]) {
+            printf("Failed to get config: %s\n", vpx_codec_err_to_string(res[i]));
+            return EXIT_FAILURE;
+        }
+    }
+
+    /*
+     * Update the default configuration according to needs of the application.
+     */
+    /* Highest-resolution encoder settings */
+    cfg[0].g_w = width;
+    cfg[0].g_h = height;
+    cfg[0].rc_dropframe_thresh = 0;
+    cfg[0].rc_end_usage = VPX_CBR;
+    cfg[0].rc_resize_allowed = 0;
+    cfg[0].rc_min_quantizer = 2;
+    cfg[0].rc_max_quantizer = 56;
+    cfg[0].rc_undershoot_pct = 100;
+    cfg[0].rc_overshoot_pct = 15;
+    cfg[0].rc_buf_initial_sz = 500;
+    cfg[0].rc_buf_optimal_sz = 600;
+    cfg[0].rc_buf_sz = 1000;
+    cfg[0].g_error_resilient = 1;              /* Enable error resilient mode */
+    cfg[0].g_lag_in_frames   = 0;
+
+    /* Disable automatic keyframe placement */
+    /* Note: These 3 settings are copied to all levels. But, except the lowest
+     * resolution level, all other levels are set to VPX_KF_DISABLED internally.
+     */
+    cfg[0].kf_mode           = VPX_KF_AUTO;
+    cfg[0].kf_min_dist = 3000;
+    cfg[0].kf_max_dist = 3000;
+
+    cfg[0].rc_target_bitrate = target_bitrate[0];       /* Set target bitrate */
+    cfg[0].g_timebase.num = 1;                          /* Set fps */
+    cfg[0].g_timebase.den = framerate;
+
+    /* Other-resolution encoder settings */
+    for (i=1; i< NUM_ENCODERS; i++)
+    {
+        memcpy(&cfg[i], &cfg[0], sizeof(vpx_codec_enc_cfg_t));
+
+        cfg[i].rc_target_bitrate = target_bitrate[i];
+
+        /* Note: Width & height of other-resolution encoders are calculated
+         * from the highest-resolution encoder's size and the corresponding
+         * down_sampling_factor.
+         */
+        {
+            unsigned int iw = cfg[i-1].g_w*dsf[i-1].den + dsf[i-1].num - 1;
+            unsigned int ih = cfg[i-1].g_h*dsf[i-1].den + dsf[i-1].num - 1;
+            cfg[i].g_w = iw/dsf[i-1].num;
+            cfg[i].g_h = ih/dsf[i-1].num;
+        }
+
+        /* Make width & height to be multiplier of 2. */
+        // Should support odd size ???
+        if((cfg[i].g_w)%2)cfg[i].g_w++;
+        if((cfg[i].g_h)%2)cfg[i].g_h++;
+    }
+
+
+    // Set the number of threads per encode/spatial layer.
+    // (1, 1, 1) means no encoder threading.
+    cfg[0].g_threads = 2;
+    cfg[1].g_threads = 1;
+    cfg[2].g_threads = 1;
+
+    /* Allocate image for each encoder */
+    for (i=0; i< NUM_ENCODERS; i++)
+        if(!vpx_img_alloc(&raw[i], VPX_IMG_FMT_I420, cfg[i].g_w, cfg[i].g_h, 32))
+            die("Failed to allocate image", cfg[i].g_w, cfg[i].g_h);
+
+    if (raw[0].stride[VPX_PLANE_Y] == raw[0].d_w)
+        read_frame_p = read_frame;
+    else
+        read_frame_p = read_frame_by_row;
+
+    for (i=0; i< NUM_ENCODERS; i++)
+        if(outfile[i])
+            write_ivf_file_header(outfile[i], &cfg[i], 0);
+
+    /* Temporal layers settings */
+    for ( i=0; i<NUM_ENCODERS; i++)
+    {
+        set_temporal_layer_pattern(num_temporal_layers[i],
+                                   &cfg[i],
+                                   cfg[i].rc_target_bitrate,
+                                   &layer_flags[i * VPX_TS_MAX_PERIODICITY]);
+    }
+
+    /* Initialize multi-encoder */
+    if(vpx_codec_enc_init_multi(&codec[0], interface, &cfg[0], NUM_ENCODERS,
+                                (show_psnr ? VPX_CODEC_USE_PSNR : 0), &dsf[0]))
+        die_codec(&codec[0], "Failed to initialize encoder");
+
+    /* The extra encoding configuration parameters can be set as follows. */
+    /* Set encoding speed */
+    for ( i=0; i<NUM_ENCODERS; i++)
+    {
+        int speed = -6;
+        /* Lower speed for the lowest resolution. */
+        if (i == NUM_ENCODERS - 1) speed = -4;
+        if(vpx_codec_control(&codec[i], VP8E_SET_CPUUSED, speed))
+            die_codec(&codec[i], "Failed to set cpu_used");
+    }
+
+    /* Set static threshold = 1 for all encoders */
+    for ( i=0; i<NUM_ENCODERS; i++)
+    {
+        if(vpx_codec_control(&codec[i], VP8E_SET_STATIC_THRESHOLD, 1))
+            die_codec(&codec[i], "Failed to set static threshold");
+    }
+
+    /* Set NOISE_SENSITIVITY to do TEMPORAL_DENOISING */
+    /* Enable denoising for the highest-resolution encoder. */
+    if(vpx_codec_control(&codec[0], VP8E_SET_NOISE_SENSITIVITY, 1))
+        die_codec(&codec[0], "Failed to set noise_sensitivity");
+    for ( i=1; i< NUM_ENCODERS; i++)
+    {
+        if(vpx_codec_control(&codec[i], VP8E_SET_NOISE_SENSITIVITY, 0))
+            die_codec(&codec[i], "Failed to set noise_sensitivity");
+    }
+
+    /* Set the number of token partitions */
+    for ( i=0; i<NUM_ENCODERS; i++)
+    {
+        if(vpx_codec_control(&codec[i], VP8E_SET_TOKEN_PARTITIONS, 1))
+            die_codec(&codec[i], "Failed to set static threshold");
+    }
+
+    /* Set the max intra target bitrate */
+    for ( i=0; i<NUM_ENCODERS; i++)
+    {
+        unsigned int max_intra_size_pct =
+            (int)(((double)cfg[0].rc_buf_optimal_sz * 0.5) * framerate / 10);
+        if(vpx_codec_control(&codec[i], VP8E_SET_MAX_INTRA_BITRATE_PCT,
+                             max_intra_size_pct))
+            die_codec(&codec[i], "Failed to set static threshold");
+       //printf("%d %d \n",i,max_intra_size_pct);
+    }
+
+    frame_avail = 1;
+    got_data = 0;
+
+    while(frame_avail || got_data)
+    {
+        vpx_codec_iter_t iter[NUM_ENCODERS]={NULL};
+        const vpx_codec_cx_pkt_t *pkt[NUM_ENCODERS];
+
+        flags = 0;
+        frame_avail = read_frame_p(infile, &raw[0]);
+
+        if(frame_avail)
+        {
+            for ( i=1; i<NUM_ENCODERS; i++)
+            {
+                /*Scale the image down a number of times by downsampling factor*/
+                /* FilterMode 1 or 2 give better psnr than FilterMode 0. */
+                I420Scale(raw[i-1].planes[VPX_PLANE_Y], raw[i-1].stride[VPX_PLANE_Y],
+                          raw[i-1].planes[VPX_PLANE_U], raw[i-1].stride[VPX_PLANE_U],
+                          raw[i-1].planes[VPX_PLANE_V], raw[i-1].stride[VPX_PLANE_V],
+                          raw[i-1].d_w, raw[i-1].d_h,
+                          raw[i].planes[VPX_PLANE_Y], raw[i].stride[VPX_PLANE_Y],
+                          raw[i].planes[VPX_PLANE_U], raw[i].stride[VPX_PLANE_U],
+                          raw[i].planes[VPX_PLANE_V], raw[i].stride[VPX_PLANE_V],
+                          raw[i].d_w, raw[i].d_h, 1);
+                /* Write out down-sampled input. */
+                length_frame = cfg[i].g_w *  cfg[i].g_h *3/2;
+                if (fwrite(raw[i].planes[0], 1, length_frame,
+                           downsampled_input[NUM_ENCODERS - i - 1]) !=
+                               length_frame)
+                {
+                    return EXIT_FAILURE;
+                }
+            }
+        }
+
+        /* Set the flags (reference and update) for all the encoders.*/
+        for ( i=0; i<NUM_ENCODERS; i++)
+        {
+            layer_id = cfg[i].ts_layer_id[frame_cnt % cfg[i].ts_periodicity];
+            flags = 0;
+            flag_periodicity = periodicity_to_num_layers
+                [num_temporal_layers[i] - 1];
+            flags = layer_flags[i * VPX_TS_MAX_PERIODICITY +
+                                frame_cnt % flag_periodicity];
+            // Key frame flag for first frame.
+            if (frame_cnt == 0)
+            {
+                flags |= VPX_EFLAG_FORCE_KF;
+            }
+            if (frame_cnt > 0 && frame_cnt == key_frame_insert)
+            {
+                flags = VPX_EFLAG_FORCE_KF;
+            }
+
+            vpx_codec_control(&codec[i], VP8E_SET_FRAME_FLAGS, flags);
+            vpx_codec_control(&codec[i], VP8E_SET_TEMPORAL_LAYER_ID, layer_id);
+        }
+
+        gettimeofday(&tv1, NULL);
+        /* Encode each frame at multi-levels */
+        /* Note the flags must be set to 0 in the encode call if they are set
+           for each frame with the vpx_codec_control(), as done above. */
+        if(vpx_codec_encode(&codec[0], frame_avail? &raw[0] : NULL,
+            frame_cnt, 1, 0, arg_deadline))
+        {
+            die_codec(&codec[0], "Failed to encode frame");
+        }
+        gettimeofday(&tv2, NULL);
+        timersub(&tv2, &tv1, &difftv);
+        cx_time += (double)(difftv.tv_sec * 1000000 + difftv.tv_usec);
+        for (i=NUM_ENCODERS-1; i>=0 ; i--)
+        {
+            got_data = 0;
+            while( (pkt[i] = vpx_codec_get_cx_data(&codec[i], &iter[i])) )
+            {
+                got_data = 1;
+                switch(pkt[i]->kind) {
+                    case VPX_CODEC_CX_FRAME_PKT:
+                        write_ivf_frame_header(outfile[i], pkt[i]);
+                        (void) fwrite(pkt[i]->data.frame.buf, 1,
+                                      pkt[i]->data.frame.sz, outfile[i]);
+                    break;
+                    case VPX_CODEC_PSNR_PKT:
+                        if (show_psnr)
+                        {
+                            int j;
+
+                            psnr_sse_total[i] += pkt[i]->data.psnr.sse[0];
+                            psnr_samples_total[i] += pkt[i]->data.psnr.samples[0];
+                            for (j = 0; j < 4; j++)
+                            {
+                                psnr_totals[i][j] += pkt[i]->data.psnr.psnr[j];
+                            }
+                            psnr_count[i]++;
+                        }
+
+                        break;
+                    default:
+                        break;
+                }
+                printf(pkt[i]->kind == VPX_CODEC_CX_FRAME_PKT
+                       && (pkt[i]->data.frame.flags & VPX_FRAME_IS_KEY)? "K":"");
+                fflush(stdout);
+            }
+        }
+        frame_cnt++;
+    }
+    printf("\n");
+    printf("FPS for encoding %d %f %f \n", frame_cnt, (float)cx_time / 1000000,
+           1000000 * (double)frame_cnt / (double)cx_time);
+
+    fclose(infile);
+
+    printf("Processed %ld frames.\n",(long int)frame_cnt-1);
+    for (i=0; i< NUM_ENCODERS; i++)
+    {
+        /* Calculate PSNR and print it out */
+        if ( (show_psnr) && (psnr_count[i]>0) )
+        {
+            int j;
+            double ovpsnr = sse_to_psnr(psnr_samples_total[i], 255.0,
+                                        psnr_sse_total[i]);
+
+            fprintf(stderr, "\n ENC%d PSNR (Overall/Avg/Y/U/V)", i);
+
+            fprintf(stderr, " %.3lf", ovpsnr);
+            for (j = 0; j < 4; j++)
+            {
+                fprintf(stderr, " %.3lf", psnr_totals[i][j]/psnr_count[i]);
+            }
+        }
+
+        if(vpx_codec_destroy(&codec[i]))
+            die_codec(&codec[i], "Failed to destroy codec");
+
+        vpx_img_free(&raw[i]);
+
+        if(!outfile[i])
+            continue;
+
+        /* Try to rewrite the file header with the actual frame count */
+        if(!fseek(outfile[i], 0, SEEK_SET))
+            write_ivf_file_header(outfile[i], &cfg[i], frame_cnt-1);
+        fclose(outfile[i]);
+    }
+    printf("\n");
+
+    return EXIT_SUCCESS;
+}
--- a/examples/vp8cx_set_ref.c
+++ b/examples/vp8cx_set_ref.c
@@ -58,7 +58,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <width> <height> <infile> <outfile> <frame>\n",
          exec_name);
  exit(EXIT_FAILURE);
--- a/examples/vp9_lossless_encoder.c
+++ b/examples/vp9_lossless_encoder.c
@@ -20,8 +20,8 @@

 static const char *exec_name;

-void usage_exit(void) {
-  fprintf(stderr, "lossless_encoder: Example demonstrating lossless "
+void usage_exit() {
+  fprintf(stderr, "vp9_lossless_encoder: Example demonstrating VP9 lossless "
                  "encoding feature. Supports raw input only.\n");
  fprintf(stderr, "Usage: %s <width> <height> <infile> <outfile>\n", exec_name);
  exit(EXIT_FAILURE);
--- a/examples/vp9_spatial_svc_encoder.c
+++ b/examples/vp9_spatial_svc_encoder.c
@@ -0,0 +1,448 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+/*
+ * This is an example demonstrating how to implement a multi-layer
+ * VP9 encoding scheme based on spatial scalability for video applications
+ * that benefit from a scalable bitstream.
+ */
+
+#include <stdarg.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+
+#include "../args.h"
+#include "../tools_common.h"
+#include "../video_writer.h"
+
+#include "vpx/svc_context.h"
+#include "vpx/vp8cx.h"
+#include "vpx/vpx_encoder.h"
+#include "../vpxstats.h"
+
+static const arg_def_t skip_frames_arg =
+    ARG_DEF("s", "skip-frames", 1, "input frames to skip");
+static const arg_def_t frames_arg =
+    ARG_DEF("f", "frames", 1, "number of frames to encode");
+static const arg_def_t width_arg = ARG_DEF("w", "width", 1, "source width");
+static const arg_def_t height_arg = ARG_DEF("h", "height", 1, "source height");
+static const arg_def_t timebase_arg =
+    ARG_DEF("t", "timebase", 1, "timebase (num/den)");
+static const arg_def_t bitrate_arg = ARG_DEF(
+    "b", "target-bitrate", 1, "encoding bitrate, in kilobits per second");
+static const arg_def_t spatial_layers_arg =
+    ARG_DEF("sl", "spatial-layers", 1, "number of spatial SVC layers");
+static const arg_def_t temporal_layers_arg =
+    ARG_DEF("tl", "temporal-layers", 1, "number of temporal SVC layers");
+static const arg_def_t kf_dist_arg =
+    ARG_DEF("k", "kf-dist", 1, "number of frames between keyframes");
+static const arg_def_t scale_factors_arg =
+    ARG_DEF("r", "scale-factors", 1, "scale factors (lowest to highest layer)");
+static const arg_def_t passes_arg =
+    ARG_DEF("p", "passes", 1, "Number of passes (1/2)");
+static const arg_def_t pass_arg =
+    ARG_DEF(NULL, "pass", 1, "Pass to execute (1/2)");
+static const arg_def_t fpf_name_arg =
+    ARG_DEF(NULL, "fpf", 1, "First pass statistics file name");
+static const arg_def_t min_q_arg =
+    ARG_DEF(NULL, "min-q", 1, "Minimum quantizer");
+static const arg_def_t max_q_arg =
+    ARG_DEF(NULL, "max-q", 1, "Maximum quantizer");
+static const arg_def_t min_bitrate_arg =
+    ARG_DEF(NULL, "min-bitrate", 1, "Minimum bitrate");
+static const arg_def_t max_bitrate_arg =
+    ARG_DEF(NULL, "max-bitrate", 1, "Maximum bitrate");
+static const arg_def_t lag_in_frame_arg =
+    ARG_DEF(NULL, "lag-in-frames", 1, "Number of frame to input before "
+        "generating any outputs");
+static const arg_def_t rc_end_usage_arg =
+    ARG_DEF(NULL, "rc-end-usage", 1, "0 - 3: VBR, CBR, CQ, Q");
+
+#if CONFIG_VP9_HIGHBITDEPTH
+static const struct arg_enum_list bitdepth_enum[] = {
+  {"8",  VPX_BITS_8},
+  {"10", VPX_BITS_10},
+  {"12", VPX_BITS_12},
+  {NULL, 0}
+};
+
+static const arg_def_t bitdepth_arg =
+    ARG_DEF_ENUM("d", "bit-depth", 1, "Bit depth for codec 8, 10 or 12. ",
+                 bitdepth_enum);
+#endif  // CONFIG_VP9_HIGHBITDEPTH
+
+
+static const arg_def_t *svc_args[] = {
+  &frames_arg,        &width_arg,         &height_arg,
+  &timebase_arg,      &bitrate_arg,       &skip_frames_arg, &spatial_layers_arg,
+  &kf_dist_arg,       &scale_factors_arg, &passes_arg,      &pass_arg,
+  &fpf_name_arg,      &min_q_arg,         &max_q_arg,       &min_bitrate_arg,
+  &max_bitrate_arg,   &temporal_layers_arg,                 &lag_in_frame_arg,
+#if CONFIG_VP9_HIGHBITDEPTH
+  &bitdepth_arg,
+#endif
+  &rc_end_usage_arg,  NULL
+};
+
+static const uint32_t default_frames_to_skip = 0;
+static const uint32_t default_frames_to_code = 60 * 60;
+static const uint32_t default_width = 1920;
+static const uint32_t default_height = 1080;
+static const uint32_t default_timebase_num = 1;
+static const uint32_t default_timebase_den = 60;
+static const uint32_t default_bitrate = 1000;
+static const uint32_t default_spatial_layers = 5;
+static const uint32_t default_temporal_layers = 1;
+static const uint32_t default_kf_dist = 100;
+
+typedef struct {
+  const char *input_filename;
+  const char *output_filename;
+  uint32_t frames_to_code;
+  uint32_t frames_to_skip;
+  struct VpxInputContext input_ctx;
+  stats_io_t rc_stats;
+  int passes;
+  int pass;
+} AppInput;
+
+static const char *exec_name;
+
+void usage_exit() {
+  fprintf(stderr, "Usage: %s <options> input_filename output_filename\n",
+          exec_name);
+  fprintf(stderr, "Options:\n");
+  arg_show_usage(stderr, svc_args);
+  exit(EXIT_FAILURE);
+}
+
+static void parse_command_line(int argc, const char **argv_,
+                               AppInput *app_input, SvcContext *svc_ctx,
+                               vpx_codec_enc_cfg_t *enc_cfg) {
+  struct arg arg = {0};
+  char **argv = NULL;
+  char **argi = NULL;
+  char **argj = NULL;
+  vpx_codec_err_t res;
+  int passes = 0;
+  int pass = 0;
+  const char *fpf_file_name = NULL;
+  unsigned int min_bitrate = 0;
+  unsigned int max_bitrate = 0;
+  char string_options[1024] = {0};
+
+  // initialize SvcContext with parameters that will be passed to vpx_svc_init
+  svc_ctx->log_level = SVC_LOG_DEBUG;
+  svc_ctx->spatial_layers = default_spatial_layers;
+  svc_ctx->temporal_layers = default_temporal_layers;
+
+  // start with default encoder configuration
+  res = vpx_codec_enc_config_default(vpx_codec_vp9_cx(), enc_cfg, 0);
+  if (res) {
+    die("Failed to get config: %s\n", vpx_codec_err_to_string(res));
+  }
+  // update enc_cfg with app default values
+  enc_cfg->g_w = default_width;
+  enc_cfg->g_h = default_height;
+  enc_cfg->g_timebase.num = default_timebase_num;
+  enc_cfg->g_timebase.den = default_timebase_den;
+  enc_cfg->rc_target_bitrate = default_bitrate;
+  enc_cfg->kf_min_dist = default_kf_dist;
+  enc_cfg->kf_max_dist = default_kf_dist;
+  enc_cfg->rc_end_usage = VPX_CQ;
+
+  // initialize AppInput with default values
+  app_input->frames_to_code = default_frames_to_code;
+  app_input->frames_to_skip = default_frames_to_skip;
+
+  // process command line options
+  argv = argv_dup(argc - 1, argv_ + 1);
+  for (argi = argj = argv; (*argj = *argi); argi += arg.argv_step) {
+    arg.argv_step = 1;
+
+    if (arg_match(&arg, &frames_arg, argi)) {
+      app_input->frames_to_code = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &width_arg, argi)) {
+      enc_cfg->g_w = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &height_arg, argi)) {
+      enc_cfg->g_h = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &timebase_arg, argi)) {
+      enc_cfg->g_timebase = arg_parse_rational(&arg);
+    } else if (arg_match(&arg, &bitrate_arg, argi)) {
+      enc_cfg->rc_target_bitrate = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &skip_frames_arg, argi)) {
+      app_input->frames_to_skip = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &spatial_layers_arg, argi)) {
+      svc_ctx->spatial_layers = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &temporal_layers_arg, argi)) {
+      svc_ctx->temporal_layers = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &kf_dist_arg, argi)) {
+      enc_cfg->kf_min_dist = arg_parse_uint(&arg);
+      enc_cfg->kf_max_dist = enc_cfg->kf_min_dist;
+    } else if (arg_match(&arg, &scale_factors_arg, argi)) {
+      snprintf(string_options, sizeof(string_options), "%s scale-factors=%s",
+               string_options, arg.val);
+    } else if (arg_match(&arg, &passes_arg, argi)) {
+      passes = arg_parse_uint(&arg);
+      if (passes < 1 || passes > 2) {
+        die("Error: Invalid number of passes (%d)\n", passes);
+      }
+    } else if (arg_match(&arg, &pass_arg, argi)) {
+      pass = arg_parse_uint(&arg);
+      if (pass < 1 || pass > 2) {
+        die("Error: Invalid pass selected (%d)\n", pass);
+      }
+    } else if (arg_match(&arg, &fpf_name_arg, argi)) {
+      fpf_file_name = arg.val;
+    } else if (arg_match(&arg, &min_q_arg, argi)) {
+      snprintf(string_options, sizeof(string_options), "%s min-quantizers=%s",
+               string_options, arg.val);
+    } else if (arg_match(&arg, &max_q_arg, argi)) {
+      snprintf(string_options, sizeof(string_options), "%s max-quantizers=%s",
+               string_options, arg.val);
+    } else if (arg_match(&arg, &min_bitrate_arg, argi)) {
+      min_bitrate = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &max_bitrate_arg, argi)) {
+      max_bitrate = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &lag_in_frame_arg, argi)) {
+      enc_cfg->g_lag_in_frames = arg_parse_uint(&arg);
+    } else if (arg_match(&arg, &rc_end_usage_arg, argi)) {
+      enc_cfg->rc_end_usage = arg_parse_uint(&arg);
+#if CONFIG_VP9_HIGHBITDEPTH
+    } else if (arg_match(&arg, &bitdepth_arg, argi)) {
+      enc_cfg->g_bit_depth = arg_parse_enum_or_int(&arg);
+      switch (enc_cfg->g_bit_depth) {
+        case VPX_BITS_8:
+          enc_cfg->g_input_bit_depth = 8;
+          enc_cfg->g_profile = 0;
+          break;
+        case VPX_BITS_10:
+          enc_cfg->g_input_bit_depth = 10;
+          enc_cfg->g_profile = 2;
+          break;
+         case VPX_BITS_12:
+          enc_cfg->g_input_bit_depth = 12;
+          enc_cfg->g_profile = 2;
+          break;
+        default:
+          die("Error: Invalid bit depth selected (%d)\n", enc_cfg->g_bit_depth);
+          break;
+      }
+#endif  // CONFIG_VP9_HIGHBITDEPTH
+    } else {
+      ++argj;
+    }
+  }
+
+  // There will be a space in front of the string options
+  if (strlen(string_options) > 0)
+    vpx_svc_set_options(svc_ctx, string_options + 1);
+
+  if (passes == 0 || passes == 1) {
+    if (pass) {
+      fprintf(stderr, "pass is ignored since there's only one pass\n");
+    }
+    enc_cfg->g_pass = VPX_RC_ONE_PASS;
+  } else {
+    if (pass == 0) {
+      die("pass must be specified when passes is 2\n");
+    }
+
+    if (fpf_file_name == NULL) {
+      die("fpf must be specified when passes is 2\n");
+    }
+
+    if (pass == 1) {
+      enc_cfg->g_pass = VPX_RC_FIRST_PASS;
+      if (!stats_open_file(&app_input->rc_stats, fpf_file_name, 0)) {
+        fatal("Failed to open statistics store");
+      }
+    } else {
+      enc_cfg->g_pass = VPX_RC_LAST_PASS;
+      if (!stats_open_file(&app_input->rc_stats, fpf_file_name, 1)) {
+        fatal("Failed to open statistics store");
+      }
+      enc_cfg->rc_twopass_stats_in = stats_get(&app_input->rc_stats);
+    }
+    app_input->passes = passes;
+    app_input->pass = pass;
+  }
+
+  if (enc_cfg->rc_target_bitrate > 0) {
+    if (min_bitrate > 0) {
+      enc_cfg->rc_2pass_vbr_minsection_pct =
+          min_bitrate * 100 / enc_cfg->rc_target_bitrate;
+    }
+    if (max_bitrate > 0) {
+      enc_cfg->rc_2pass_vbr_maxsection_pct =
+          max_bitrate * 100 / enc_cfg->rc_target_bitrate;
+    }
+  }
+
+  // Check for unrecognized options
+  for (argi = argv; *argi; ++argi)
+    if (argi[0][0] == '-' && strlen(argi[0]) > 1)
+      die("Error: Unrecognized option %s\n", *argi);
+
+  if (argv[0] == NULL || argv[1] == 0) {
+    usage_exit();
+  }
+  app_input->input_filename = argv[0];
+  app_input->output_filename = argv[1];
+  free(argv);
+
+  if (enc_cfg->g_w < 16 || enc_cfg->g_w % 2 || enc_cfg->g_h < 16 ||
+      enc_cfg->g_h % 2)
+    die("Invalid resolution: %d x %d\n", enc_cfg->g_w, enc_cfg->g_h);
+
+  printf(
+      "Codec %s\nframes: %d, skip: %d\n"
+      "layers: %d\n"
+      "width %d, height: %d,\n"
+      "num: %d, den: %d, bitrate: %d,\n"
+      "gop size: %d\n",
+      vpx_codec_iface_name(vpx_codec_vp9_cx()), app_input->frames_to_code,
+      app_input->frames_to_skip,
+      svc_ctx->spatial_layers, enc_cfg->g_w, enc_cfg->g_h,
+      enc_cfg->g_timebase.num, enc_cfg->g_timebase.den,
+      enc_cfg->rc_target_bitrate, enc_cfg->kf_max_dist);
+}
+
+int main(int argc, const char **argv) {
+  AppInput app_input = {0};
+  VpxVideoWriter *writer = NULL;
+  VpxVideoInfo info = {0};
+  vpx_codec_ctx_t codec;
+  vpx_codec_enc_cfg_t enc_cfg;
+  SvcContext svc_ctx;
+  uint32_t i;
+  uint32_t frame_cnt = 0;
+  vpx_image_t raw;
+  vpx_codec_err_t res;
+  int pts = 0;            /* PTS starts at 0 */
+  int frame_duration = 1; /* 1 timebase tick per frame */
+  FILE *infile = NULL;
+  int end_of_stream = 0;
+  int frames_received = 0;
+
+  memset(&svc_ctx, 0, sizeof(svc_ctx));
+  svc_ctx.log_print = 1;
+  exec_name = argv[0];
+  parse_command_line(argc, argv, &app_input, &svc_ctx, &enc_cfg);
+
+  // Allocate image buffer
+#if CONFIG_VP9_HIGHBITDEPTH
+  if (!vpx_img_alloc(&raw, enc_cfg.g_input_bit_depth == 8 ?
+                         VPX_IMG_FMT_I420 : VPX_IMG_FMT_I42016,
+                     enc_cfg.g_w, enc_cfg.g_h, 32)) {
+    die("Failed to allocate image %dx%d\n", enc_cfg.g_w, enc_cfg.g_h);
+  }
+#else
+  if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, enc_cfg.g_w, enc_cfg.g_h, 32)) {
+    die("Failed to allocate image %dx%d\n", enc_cfg.g_w, enc_cfg.g_h);
+  }
+#endif  // CONFIG_VP9_HIGHBITDEPTH
+
+  if (!(infile = fopen(app_input.input_filename, "rb")))
+    die("Failed to open %s for reading\n", app_input.input_filename);
+
+  // Initialize codec
+  if (vpx_svc_init(&svc_ctx, &codec, vpx_codec_vp9_cx(), &enc_cfg) !=
+      VPX_CODEC_OK)
+    die("Failed to initialize encoder\n");
+
+  info.codec_fourcc = VP9_FOURCC;
+  info.time_base.numerator = enc_cfg.g_timebase.num;
+  info.time_base.denominator = enc_cfg.g_timebase.den;
+
+  if (!(app_input.passes == 2 && app_input.pass == 1)) {
+    // We don't save the bitstream for the 1st pass on two pass rate control
+    writer = vpx_video_writer_open(app_input.output_filename, kContainerIVF,
+                                   &info);
+    if (!writer)
+      die("Failed to open %s for writing\n", app_input.output_filename);
+  }
+
+  // skip initial frames
+  for (i = 0; i < app_input.frames_to_skip; ++i)
+    vpx_img_read(&raw, infile);
+
+  // Encode frames
+  while (!end_of_stream) {
+    vpx_codec_iter_t iter = NULL;
+    const vpx_codec_cx_pkt_t *cx_pkt;
+    if (frame_cnt >= app_input.frames_to_code || !vpx_img_read(&raw, infile)) {
+      // We need one extra vpx_svc_encode call at end of stream to flush
+      // encoder and get remaining data
+      end_of_stream = 1;
+    }
+
+    res = vpx_svc_encode(&svc_ctx, &codec, (end_of_stream ? NULL : &raw),
+                         pts, frame_duration, VPX_DL_GOOD_QUALITY);
+    printf("%s", vpx_svc_get_message(&svc_ctx));
+    if (res != VPX_CODEC_OK) {
+      die_codec(&codec, "Failed to encode frame");
+    }
+
+    while ((cx_pkt = vpx_codec_get_cx_data(&codec, &iter)) != NULL) {
+      switch (cx_pkt->kind) {
+        case VPX_CODEC_CX_FRAME_PKT: {
+          if (cx_pkt->data.frame.sz > 0)
+            vpx_video_writer_write_frame(writer,
+                                         cx_pkt->data.frame.buf,
+                                         cx_pkt->data.frame.sz,
+                                         cx_pkt->data.frame.pts);
+
+          printf("SVC frame: %d, kf: %d, size: %d, pts: %d\n", frames_received,
+                 !!(cx_pkt->data.frame.flags & VPX_FRAME_IS_KEY),
+                 (int)cx_pkt->data.frame.sz, (int)cx_pkt->data.frame.pts);
+          ++frames_received;
+          break;
+        }
+        case VPX_CODEC_STATS_PKT: {
+          stats_write(&app_input.rc_stats,
+                      cx_pkt->data.twopass_stats.buf,
+                      cx_pkt->data.twopass_stats.sz);
+          break;
+        }
+        default: {
+          break;
+        }
+      }
+    }
+
+    if (!end_of_stream) {
+      ++frame_cnt;
+      pts += frame_duration;
+    }
+  }
+
+  printf("Processed %d frames\n", frame_cnt);
+
+  fclose(infile);
+  if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec");
+
+  if (app_input.passes == 2)
+    stats_close(&app_input.rc_stats, 1);
+
+  if (writer) {
+    vpx_video_writer_close(writer);
+  }
+
+  vpx_img_free(&raw);
+
+  // display average size, psnr
+  printf("%s", vpx_svc_dump_statistics(&svc_ctx));
+
+  vpx_svc_release(&svc_ctx);
+
+  return EXIT_SUCCESS;
+}
--- a/examples/vpx_temporal_svc_encoder.c
+++ b/examples/vpx_temporal_svc_encoder.c
@@ -28,7 +28,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  exit(EXIT_FAILURE);
 }

@@ -41,7 +41,7 @@ enum denoiserState {
  kDenoiserOnAdaptive
 };

-static int mode_to_num_layers[13] = {1, 2, 2, 3, 3, 3, 3, 5, 2, 3, 3, 3, 3};
+static int mode_to_num_layers[12] = {1, 2, 2, 3, 3, 3, 3, 5, 2, 3, 3, 3};

 // For rate control encoding stats.
 struct RateControlMetrics {
@@ -70,7 +70,6 @@ struct RateControlMetrics {
  int window_size;
  // Number of window measurements.
  int window_count;
-  int layer_target_bitrate[VPX_MAX_LAYERS];
 };

 // Note: these rate control metrics assume only 1 key frame in the
@@ -86,13 +85,13 @@ static void set_rate_control_metrics(struct RateControlMetrics *rc,
  // per-frame-bandwidth, for the rate control encoding stats below.
  const double framerate = cfg->g_timebase.den / cfg->g_timebase.num;
  rc->layer_framerate[0] = framerate / cfg->ts_rate_decimator[0];
-  rc->layer_pfb[0] = 1000.0 * rc->layer_target_bitrate[0] /
+  rc->layer_pfb[0] = 1000.0 * cfg->ts_target_bitrate[0] /
      rc->layer_framerate[0];
  for (i = 0; i < cfg->ts_number_layers; ++i) {
    if (i > 0) {
      rc->layer_framerate[i] = framerate / cfg->ts_rate_decimator[i];
      rc->layer_pfb[i] = 1000.0 *
-          (rc->layer_target_bitrate[i] - rc->layer_target_bitrate[i - 1]) /
+          (cfg->ts_target_bitrate[i] - cfg->ts_target_bitrate[i - 1]) /
          (rc->layer_framerate[i] - rc->layer_framerate[i - 1]);
    }
    rc->layer_input_frames[i] = 0;
@@ -129,7 +128,7 @@ static void printout_rate_control_summary(struct RateControlMetrics *rc,
    rc->layer_avg_rate_mismatch[i] = 100.0 * rc->layer_avg_rate_mismatch[i] /
        rc->layer_enc_frames[i];
    printf("For layer#: %d \n", i);
-    printf("Bitrate (target vs actual): %d %f \n", rc->layer_target_bitrate[i],
+    printf("Bitrate (target vs actual): %d %f \n", cfg->ts_target_bitrate[i],
           rc->layer_encoding_bitrate[i]);
    printf("Average frame size (target vs actual): %f %f \n", rc->layer_pfb[i],
           rc->layer_avg_frame_size[i]);
@@ -432,32 +431,7 @@ static void set_temporal_layer_pattern(int layering_mode,
      layer_flags[7] = layer_flags[3];
      break;
    }
-    case 11: {
-      // 3-layers structure with one reference frame.
-      // This works same as temporal_layering_mode 3.
-      // This was added to compare with vp9_spatial_svc_encoder.
-
-      // 3-layers, 4-frame period.
-      int ids[4] = {0, 2, 1, 2};
-      cfg->ts_periodicity = 4;
-      *flag_periodicity = 4;
-      cfg->ts_number_layers = 3;
-      cfg->ts_rate_decimator[0] = 4;
-      cfg->ts_rate_decimator[1] = 2;
-      cfg->ts_rate_decimator[2] = 1;
-      memcpy(cfg->ts_layer_id, ids, sizeof(ids));
-      // 0=L, 1=GF, 2=ARF, Intra-layer prediction disabled.
-      layer_flags[0] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
-          VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
-      layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
-          VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
-      layer_flags[1] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
-          VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
-      layer_flags[3] = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_REF_ARF |
-          VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
-      break;
-    }
-    case 12:
+    case 11:
    default: {
      // 3-layers structure as in case 10, but no sync/refresh points for
      // layer 1 and 2.
@@ -517,13 +491,13 @@ int main(int argc, char **argv) {
  struct RateControlMetrics rc;
  int64_t cx_time = 0;
  const int min_args_base = 11;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  vpx_bit_depth_t bit_depth = VPX_BITS_8;
  int input_bit_depth = 8;
  const int min_args = min_args_base + 1;
 #else
  const int min_args = min_args_base;
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  double sum_bitrate = 0.0;
  double sum_bitrate2 = 0.0;
  double framerate  = 30.0;
@@ -531,7 +505,7 @@ int main(int argc, char **argv) {
  exec_name = argv[0];
  // Check usage and arguments.
  if (argc < min_args) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    die("Usage: %s <infile> <outfile> <codec_type(vp8/vp9)> <width> <height> "
        "<rate_num> <rate_den> <speed> <frame_drop_threshold> <mode> "
        "<Rate_0> ... <Rate_nlayers-1> <bit-depth> \n", argv[0]);
@@ -539,7 +513,7 @@ int main(int argc, char **argv) {
    die("Usage: %s <infile> <outfile> <codec_type(vp8/vp9)> <width> <height> "
        "<rate_num> <rate_den> <speed> <frame_drop_threshold> <mode> "
        "<Rate_0> ... <Rate_nlayers-1> \n", argv[0]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  }

  encoder = get_vpx_encoder_by_name(argv[3]);
@@ -555,7 +529,7 @@ int main(int argc, char **argv) {
  }

  layering_mode = strtol(argv[10], NULL, 0);
-  if (layering_mode < 0 || layering_mode > 13) {
+  if (layering_mode < 0 || layering_mode > 12) {
    die("Invalid layering mode (0..12) %s", argv[10]);
  }

@@ -563,7 +537,7 @@ int main(int argc, char **argv) {
    die("Invalid number of arguments");
  }

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  switch (strtol(argv[argc-1], NULL, 0)) {
    case 8:
      bit_depth = VPX_BITS_8;
@@ -590,7 +564,7 @@ int main(int argc, char **argv) {
  if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, width, height, 32)) {
    die("Failed to allocate image", width, height);
  }
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

  // Populate encoder configuration.
  res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
@@ -603,13 +577,13 @@ int main(int argc, char **argv) {
  cfg.g_w = width;
  cfg.g_h = height;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  if (bit_depth != VPX_BITS_8) {
    cfg.g_bit_depth = bit_depth;
    cfg.g_input_bit_depth = input_bit_depth;
    cfg.g_profile = 2;
  }
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

  // Timebase format e.g. 30fps: numerator=1, demoninator = 30.
  cfg.g_timebase.num = strtol(argv[6], NULL, 0);
@@ -623,16 +597,13 @@ int main(int argc, char **argv) {
  for (i = min_args_base;
       (int)i < min_args_base + mode_to_num_layers[layering_mode];
       ++i) {
-    rc.layer_target_bitrate[i - 11] = strtol(argv[i], NULL, 0);
-    if (strncmp(encoder->name, "vp8", 3) == 0)
-      cfg.ts_target_bitrate[i - 11] = rc.layer_target_bitrate[i - 11];
-    else if (strncmp(encoder->name, "vp9", 3) == 0)
-      cfg.layer_target_bitrate[i - 11] = rc.layer_target_bitrate[i - 11];
+    cfg.ts_target_bitrate[i - 11] = strtol(argv[i], NULL, 0);
  }

  // Real time parameters.
  cfg.rc_dropframe_thresh = strtol(argv[9], NULL, 0);
  cfg.rc_end_usage = VPX_CBR;
+  cfg.rc_resize_allowed = 0;
  cfg.rc_min_quantizer = 2;
  cfg.rc_max_quantizer = 56;
  if (strncmp(encoder->name, "vp9", 3) == 0)
@@ -643,9 +614,6 @@ int main(int argc, char **argv) {
  cfg.rc_buf_optimal_sz = 600;
  cfg.rc_buf_sz = 1000;

-  // Disable dynamic resizing by default.
-  cfg.rc_resize_allowed = 0;
-
  // Use 1 thread as default.
  cfg.g_threads = 1;

@@ -657,8 +625,6 @@ int main(int argc, char **argv) {
  // Disable automatic keyframe placement.
  cfg.kf_min_dist = cfg.kf_max_dist = 3000;

-  cfg.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
-
  set_temporal_layer_pattern(layering_mode,
                             &cfg,
                             layer_flags,
@@ -667,8 +633,8 @@ int main(int argc, char **argv) {
  set_rate_control_metrics(&rc, &cfg);

  // Target bandwidth for the whole stream.
-  // Set to layer_target_bitrate for highest layer (total bitrate).
-  cfg.rc_target_bitrate = rc.layer_target_bitrate[cfg.ts_number_layers - 1];
+  // Set to ts_target_bitrate for highest layer (total bitrate).
+  cfg.rc_target_bitrate = cfg.ts_target_bitrate[cfg.ts_number_layers - 1];

  // Open input file.
  if (!(infile = fopen(argv[1], "rb"))) {
@@ -697,37 +663,29 @@ int main(int argc, char **argv) {
  cfg.ss_number_layers = 1;

  // Initialize codec.
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  if (vpx_codec_enc_init(
          &codec, encoder->codec_interface(), &cfg,
          bit_depth == VPX_BITS_8 ? 0 : VPX_CODEC_USE_HIGHBITDEPTH))
 #else
  if (vpx_codec_enc_init(&codec, encoder->codec_interface(), &cfg, 0))
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
    die_codec(&codec, "Failed to initialize encoder");

  if (strncmp(encoder->name, "vp8", 3) == 0) {
    vpx_codec_control(&codec, VP8E_SET_CPUUSED, -speed);
-    vpx_codec_control(&codec, VP8E_SET_NOISE_SENSITIVITY, kDenoiserOff);
+    vpx_codec_control(&codec, VP8E_SET_NOISE_SENSITIVITY, kDenoiserOnYOnly);
    vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
  } else if (strncmp(encoder->name, "vp9", 3) == 0) {
-    vpx_svc_extra_cfg_t svc_params;
-    vpx_codec_control(&codec, VP8E_SET_CPUUSED, speed);
-    vpx_codec_control(&codec, VP9E_SET_AQ_MODE, 3);
-    vpx_codec_control(&codec, VP9E_SET_FRAME_PERIODIC_BOOST, 0);
-    vpx_codec_control(&codec, VP9E_SET_NOISE_SENSITIVITY, kDenoiserOff);
-    vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
-    vpx_codec_control(&codec, VP9E_SET_TUNE_CONTENT, 0);
-    vpx_codec_control(&codec, VP9E_SET_TILE_COLUMNS, (cfg.g_threads >> 1));
-    if (vpx_codec_control(&codec, VP9E_SET_SVC, layering_mode > 0 ? 1: 0))
-      die_codec(&codec, "Failed to set SVC");
-    for (i = 0; i < cfg.ts_number_layers; ++i) {
-      svc_params.max_quantizers[i] = cfg.rc_max_quantizer;
-      svc_params.min_quantizers[i] = cfg.rc_min_quantizer;
+      vpx_codec_control(&codec, VP8E_SET_CPUUSED, speed);
+      vpx_codec_control(&codec, VP9E_SET_AQ_MODE, 3);
+      vpx_codec_control(&codec, VP9E_SET_FRAME_PERIODIC_BOOST, 0);
+      vpx_codec_control(&codec, VP9E_SET_NOISE_SENSITIVITY, 0);
+      vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
+      vpx_codec_control(&codec, VP9E_SET_TILE_COLUMNS, (cfg.g_threads >> 1));
+      if (vpx_codec_control(&codec, VP9E_SET_SVC, layering_mode > 0 ? 1: 0)) {
+        die_codec(&codec, "Failed to set SVC");
    }
-    svc_params.scaling_factor_num[0] = cfg.g_h;
-    svc_params.scaling_factor_den[0] = cfg.g_h;
-    vpx_codec_control(&codec, VP9E_SET_SVC_PARAMETERS, &svc_params);
  }
  if (strncmp(encoder->name, "vp8", 3) == 0) {
    vpx_codec_control(&codec, VP8E_SET_SCREEN_CONTENT_MODE, 0);
--- a/examples/vpxcx_set_ref.c
+++ b/examples/vpxcx_set_ref.c
@@ -1,455 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-
-// VP10 Set Reference Frame
-// ============================
-//
-// This is an example demonstrating how to overwrite the VP10 encoder's
-// internal reference frame. In the sample we set the last frame to the
-// current frame. This technique could be used to bounce between two cameras.
-//
-// The decoder would also have to set the reference frame to the same value
-// on the same frame, or the video will become corrupt. The 'test_decode'
-// variable is set to 1 in this example that tests if the encoder and decoder
-// results are matching.
-//
-// Usage
-// -----
-// This example encodes a raw video. And the last argument passed in specifies
-// the frame number to update the reference frame on. For example, run
-// examples/vpx_cx_set_ref vp10 352 288 in.yuv out.ivf 4 30
-// The parameter is parsed as follows:
-//
-//
-// Extra Variables
-// ---------------
-// This example maintains the frame number passed on the command line
-// in the `update_frame_num` variable.
-//
-//
-// Configuration
-// -------------
-//
-// The reference frame is updated on the frame specified on the command
-// line.
-//
-// Observing The Effects
-// ---------------------
-// The encoder and decoder results should be matching when the same reference
-// frame setting operation is done in both encoder and decoder. Otherwise,
-// the encoder/decoder mismatch would be seen.
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "vpx/vp8cx.h"
-#include "vpx/vpx_decoder.h"
-#include "vpx/vpx_encoder.h"
-
-#include "./tools_common.h"
-#include "./video_writer.h"
-
-static const char *exec_name;
-
-void usage_exit() {
-  fprintf(stderr, "Usage: %s <codec> <width> <height> <infile> <outfile> "
-          "<frame> <limit(optional)>\n",
-          exec_name);
-  exit(EXIT_FAILURE);
-}
-
-static int compare_img(const vpx_image_t *const img1,
-                       const vpx_image_t *const img2) {
-  uint32_t l_w = img1->d_w;
-  uint32_t c_w =
-      (img1->d_w + img1->x_chroma_shift) >> img1->x_chroma_shift;
-  const uint32_t c_h =
-      (img1->d_h + img1->y_chroma_shift) >> img1->y_chroma_shift;
-  uint32_t i;
-  int match = 1;
-
-  match &= (img1->fmt == img2->fmt);
-  match &= (img1->d_w == img2->d_w);
-  match &= (img1->d_h == img2->d_h);
-
-  for (i = 0; i < img1->d_h; ++i)
-    match &= (memcmp(img1->planes[VPX_PLANE_Y] + i * img1->stride[VPX_PLANE_Y],
-                     img2->planes[VPX_PLANE_Y] + i * img2->stride[VPX_PLANE_Y],
-                     l_w) == 0);
-
-  for (i = 0; i < c_h; ++i)
-    match &= (memcmp(img1->planes[VPX_PLANE_U] + i * img1->stride[VPX_PLANE_U],
-                     img2->planes[VPX_PLANE_U] + i * img2->stride[VPX_PLANE_U],
-                     c_w) == 0);
-
-  for (i = 0; i < c_h; ++i)
-    match &= (memcmp(img1->planes[VPX_PLANE_V] + i * img1->stride[VPX_PLANE_V],
-                     img2->planes[VPX_PLANE_V] + i * img2->stride[VPX_PLANE_V],
-                     c_w) == 0);
-
-  return match;
-}
-
-#define mmin(a, b)  ((a) < (b) ? (a) : (b))
-static void find_mismatch(const vpx_image_t *const img1,
-                          const vpx_image_t *const img2,
-                          int yloc[4], int uloc[4], int vloc[4]) {
-  const uint32_t bsize = 64;
-  const uint32_t bsizey = bsize >> img1->y_chroma_shift;
-  const uint32_t bsizex = bsize >> img1->x_chroma_shift;
-  const uint32_t c_w =
-      (img1->d_w + img1->x_chroma_shift) >> img1->x_chroma_shift;
-  const uint32_t c_h =
-      (img1->d_h + img1->y_chroma_shift) >> img1->y_chroma_shift;
-  int match = 1;
-  uint32_t i, j;
-  yloc[0] = yloc[1] = yloc[2] = yloc[3] = -1;
-  for (i = 0, match = 1; match && i < img1->d_h; i += bsize) {
-    for (j = 0; match && j < img1->d_w; j += bsize) {
-      int k, l;
-      const int si = mmin(i + bsize, img1->d_h) - i;
-      const int sj = mmin(j + bsize, img1->d_w) - j;
-      for (k = 0; match && k < si; ++k) {
-        for (l = 0; match && l < sj; ++l) {
-          if (*(img1->planes[VPX_PLANE_Y] +
-                (i + k) * img1->stride[VPX_PLANE_Y] + j + l) !=
-              *(img2->planes[VPX_PLANE_Y] +
-                (i + k) * img2->stride[VPX_PLANE_Y] + j + l)) {
-            yloc[0] = i + k;
-            yloc[1] = j + l;
-            yloc[2] = *(img1->planes[VPX_PLANE_Y] +
-                        (i + k) * img1->stride[VPX_PLANE_Y] + j + l);
-            yloc[3] = *(img2->planes[VPX_PLANE_Y] +
-                        (i + k) * img2->stride[VPX_PLANE_Y] + j + l);
-            match = 0;
-            break;
-          }
-        }
-      }
-    }
-  }
-
-  uloc[0] = uloc[1] = uloc[2] = uloc[3] = -1;
-  for (i = 0, match = 1; match && i < c_h; i += bsizey) {
-    for (j = 0; match && j < c_w; j += bsizex) {
-      int k, l;
-      const int si = mmin(i + bsizey, c_h - i);
-      const int sj = mmin(j + bsizex, c_w - j);
-      for (k = 0; match && k < si; ++k) {
-        for (l = 0; match && l < sj; ++l) {
-          if (*(img1->planes[VPX_PLANE_U] +
-                (i + k) * img1->stride[VPX_PLANE_U] + j + l) !=
-              *(img2->planes[VPX_PLANE_U] +
-                (i + k) * img2->stride[VPX_PLANE_U] + j + l)) {
-            uloc[0] = i + k;
-            uloc[1] = j + l;
-            uloc[2] = *(img1->planes[VPX_PLANE_U] +
-                        (i + k) * img1->stride[VPX_PLANE_U] + j + l);
-            uloc[3] = *(img2->planes[VPX_PLANE_U] +
-                        (i + k) * img2->stride[VPX_PLANE_U] + j + l);
-            match = 0;
-            break;
-          }
-        }
-      }
-    }
-  }
-  vloc[0] = vloc[1] = vloc[2] = vloc[3] = -1;
-  for (i = 0, match = 1; match && i < c_h; i += bsizey) {
-    for (j = 0; match && j < c_w; j += bsizex) {
-      int k, l;
-      const int si = mmin(i + bsizey, c_h - i);
-      const int sj = mmin(j + bsizex, c_w - j);
-      for (k = 0; match && k < si; ++k) {
-        for (l = 0; match && l < sj; ++l) {
-          if (*(img1->planes[VPX_PLANE_V] +
-                (i + k) * img1->stride[VPX_PLANE_V] + j + l) !=
-              *(img2->planes[VPX_PLANE_V] +
-                (i + k) * img2->stride[VPX_PLANE_V] + j + l)) {
-            vloc[0] = i + k;
-            vloc[1] = j + l;
-            vloc[2] = *(img1->planes[VPX_PLANE_V] +
-                        (i + k) * img1->stride[VPX_PLANE_V] + j + l);
-            vloc[3] = *(img2->planes[VPX_PLANE_V] +
-                        (i + k) * img2->stride[VPX_PLANE_V] + j + l);
-            match = 0;
-            break;
-          }
-        }
-      }
-    }
-  }
-}
-
-static void testing_decode(vpx_codec_ctx_t *encoder,
-                           vpx_codec_ctx_t *decoder,
-                           vpx_codec_enc_cfg_t *cfg,
-                           unsigned int frame_out,
-                           int *mismatch_seen) {
-  vpx_image_t enc_img, dec_img;
-  struct vp9_ref_frame ref_enc, ref_dec;
-
-  if (*mismatch_seen)
-    return;
-
-  ref_enc.idx = 0;
-  ref_dec.idx = 0;
-  if (vpx_codec_control(encoder, VP9_GET_REFERENCE, &ref_enc))
-    die_codec(encoder,  "Failed to get encoder reference frame");
-  enc_img = ref_enc.img;
-  if (vpx_codec_control(decoder, VP9_GET_REFERENCE, &ref_dec))
-    die_codec(decoder, "Failed to get decoder reference frame");
-  dec_img = ref_dec.img;
-
-  if (!compare_img(&enc_img, &dec_img)) {
-    int y[4], u[4], v[4];
-
-    *mismatch_seen = 1;
-
-    find_mismatch(&enc_img, &dec_img, y, u, v);
-    printf("Encode/decode mismatch on frame %d at"
-           " Y[%d, %d] {%d/%d},"
-           " U[%d, %d] {%d/%d},"
-           " V[%d, %d] {%d/%d}",
-           frame_out,
-           y[0], y[1], y[2], y[3],
-           u[0], u[1], u[2], u[3],
-           v[0], v[1], v[2], v[3]);
-  }
-
-  vpx_img_free(&enc_img);
-  vpx_img_free(&dec_img);
-}
-
-static int encode_frame(vpx_codec_ctx_t *ecodec,
-                        vpx_codec_enc_cfg_t *cfg,
-                        vpx_image_t *img,
-                        unsigned int frame_in,
-                        VpxVideoWriter *writer,
-                        int test_decode,
-                        vpx_codec_ctx_t *dcodec,
-                        unsigned int *frame_out,
-                        int *mismatch_seen) {
-  int got_pkts = 0;
-  vpx_codec_iter_t iter = NULL;
-  const vpx_codec_cx_pkt_t *pkt = NULL;
-  int got_data;
-  const vpx_codec_err_t res = vpx_codec_encode(ecodec, img, frame_in, 1,
-                                               0, VPX_DL_GOOD_QUALITY);
-  if (res != VPX_CODEC_OK)
-    die_codec(ecodec, "Failed to encode frame");
-
-  got_data = 0;
-
-  while ((pkt = vpx_codec_get_cx_data(ecodec, &iter)) != NULL) {
-    got_pkts = 1;
-
-    if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
-      const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
-
-      if (!(pkt->data.frame.flags & VPX_FRAME_IS_FRAGMENT)) {
-                *frame_out += 1;
-        }
-
-      if (!vpx_video_writer_write_frame(writer,
-                                        pkt->data.frame.buf,
-                                        pkt->data.frame.sz,
-                                        pkt->data.frame.pts)) {
-        die_codec(ecodec, "Failed to write compressed frame");
-      }
-      printf(keyframe ? "K" : ".");
-      fflush(stdout);
-      got_data = 1;
-
-      // Decode 1 frame.
-      if (test_decode) {
-        if (vpx_codec_decode(dcodec, pkt->data.frame.buf,
-                             (unsigned int)pkt->data.frame.sz, NULL, 0))
-          die_codec(dcodec, "Failed to decode frame.");
-      }
-    }
-  }
-
-  // Mismatch checking
-  if (got_data && test_decode) {
-    testing_decode(ecodec, dcodec, cfg, *frame_out, mismatch_seen);
-  }
-
-  return got_pkts;
-}
-
-int main(int argc, char **argv) {
-  FILE *infile = NULL;
-  // Encoder
-  vpx_codec_ctx_t ecodec = {0};
-  vpx_codec_enc_cfg_t cfg = {0};
-  unsigned int frame_in = 0;
-  vpx_image_t raw;
-  vpx_codec_err_t res;
-  VpxVideoInfo info = {0};
-  VpxVideoWriter *writer = NULL;
-  const VpxInterface *encoder = NULL;
-
-  // Test encoder/decoder mismatch.
-  int test_decode = 1;
-  // Decoder
-  vpx_codec_ctx_t dcodec;
-  unsigned int frame_out = 0;
-
-  // The frame number to set reference frame on
-  unsigned int update_frame_num = 0;
-  int mismatch_seen = 0;
-
-  const int fps = 30;
-  const int bitrate = 500;
-
-  const char *codec_arg = NULL;
-  const char *width_arg = NULL;
-  const char *height_arg = NULL;
-  const char *infile_arg = NULL;
-  const char *outfile_arg = NULL;
-  unsigned int limit = 0;
-  exec_name = argv[0];
-
-  if (argc < 7)
-    die("Invalid number of arguments");
-
-  codec_arg = argv[1];
-  width_arg = argv[2];
-  height_arg = argv[3];
-  infile_arg = argv[4];
-  outfile_arg = argv[5];
-
-  encoder = get_vpx_encoder_by_name(codec_arg);
-  if (!encoder)
-    die("Unsupported codec.");
-
-  update_frame_num = atoi(argv[6]);
-  // In VP10, the reference buffers (cm->buffer_pool->frame_bufs[i].buf) are
-  // allocated while calling vpx_codec_encode(), thus, setting reference for
-  // 1st frame isn't supported.
-  if (update_frame_num <= 1)
-    die("Couldn't parse frame number '%s'\n", argv[6]);
-
-  if (argc > 7) {
-    limit = atoi(argv[7]);
-    if (update_frame_num > limit)
-      die("Update frame number couldn't larger than limit\n");
-  }
-
-  info.codec_fourcc = encoder->fourcc;
-  info.frame_width = strtol(width_arg, NULL, 0);
-  info.frame_height = strtol(height_arg, NULL, 0);
-  info.time_base.numerator = 1;
-  info.time_base.denominator = fps;
-
-  if (info.frame_width <= 0 ||
-      info.frame_height <= 0 ||
-      (info.frame_width % 2) != 0 ||
-      (info.frame_height % 2) != 0) {
-    die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
-  }
-
-  if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
-                                             info.frame_height, 1)) {
-    die("Failed to allocate image.");
-  }
-
-  printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
-
-  res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
-  if (res)
-    die_codec(&ecodec, "Failed to get default codec config.");
-
-  cfg.g_w = info.frame_width;
-  cfg.g_h = info.frame_height;
-  cfg.g_timebase.num = info.time_base.numerator;
-  cfg.g_timebase.den = info.time_base.denominator;
-  cfg.rc_target_bitrate = bitrate;
-  cfg.g_lag_in_frames = 3;
-
-  writer = vpx_video_writer_open(outfile_arg, kContainerIVF, &info);
-  if (!writer)
-    die("Failed to open %s for writing.", outfile_arg);
-
-  if (!(infile = fopen(infile_arg, "rb")))
-    die("Failed to open %s for reading.", infile_arg);
-
-  if (vpx_codec_enc_init(&ecodec, encoder->codec_interface(), &cfg, 0))
-    die_codec(&ecodec, "Failed to initialize encoder");
-
-  // Disable alt_ref.
-  if (vpx_codec_control(&ecodec, VP8E_SET_ENABLEAUTOALTREF, 0))
-    die_codec(&ecodec, "Failed to set enable auto alt ref");
-
-  if (test_decode) {
-      const VpxInterface *decoder = get_vpx_decoder_by_name(codec_arg);
-      if (vpx_codec_dec_init(&dcodec, decoder->codec_interface(), NULL, 0))
-        die_codec(&dcodec, "Failed to initialize decoder.");
-  }
-
-  // Encode frames.
-  while (vpx_img_read(&raw, infile)) {
-    if (limit && frame_in >= limit)
-      break;
-    if (update_frame_num > 1 && frame_out + 1 == update_frame_num) {
-      vpx_ref_frame_t ref;
-      ref.frame_type = VP8_LAST_FRAME;
-      ref.img = raw;
-      // Set reference frame in encoder.
-      if (vpx_codec_control(&ecodec, VP8_SET_REFERENCE, &ref))
-        die_codec(&ecodec, "Failed to set reference frame");
-      printf(" <SET_REF>");
-
-      // If set_reference in decoder is commented out, the enc/dec mismatch
-      // would be seen.
-      if (test_decode) {
-        if (vpx_codec_control(&dcodec, VP8_SET_REFERENCE, &ref))
-          die_codec(&dcodec, "Failed to set reference frame");
-      }
-    }
-
-    encode_frame(&ecodec, &cfg, &raw, frame_in, writer, test_decode,
-                 &dcodec, &frame_out, &mismatch_seen);
-    frame_in++;
-    if (mismatch_seen)
-      break;
-  }
-
-  // Flush encoder.
-  if (!mismatch_seen)
-    while (encode_frame(&ecodec, &cfg, NULL, frame_in, writer, test_decode,
-                        &dcodec, &frame_out, &mismatch_seen)) {}
-
-  printf("\n");
-  fclose(infile);
-  printf("Processed %d frames.\n", frame_out);
-
-  if (test_decode) {
-    if (!mismatch_seen)
-      printf("Encoder/decoder results are matching.\n");
-    else
-      printf("Encoder/decoder results are NOT matching.\n");
-  }
-
-  if (test_decode)
-    if (vpx_codec_destroy(&dcodec))
-      die_codec(&dcodec, "Failed to destroy decoder");
-
-  vpx_img_free(&raw);
-  if (vpx_codec_destroy(&ecodec))
-    die_codec(&ecodec, "Failed to destroy encoder.");
-
-  vpx_video_writer_close(writer);
-
-  return EXIT_SUCCESS;
-}
--- a/ivfdec.c
+++ b/ivfdec.c
@@ -23,7 +23,7 @@ static void fix_framerate(int *num, int *den) {
  // we can guess the framerate using only the timebase in this
  // case. Other files would require reading ahead to guess the
  // timebase, like we do for webm.
-  if (*den > 0 && *den < 1000000000 && *num > 0 && *num < 1000) {
+  if (*num < 1000) {
    // Correct for the factor of 2 applied to the timebase in the encoder.
    if (*num & 1)
      *den *= 2;
--- a/libs.mk
+++ b/libs.mk
@@ -25,7 +25,7 @@ $$(BUILD_PFX)$(1).h: $$(SRC_PATH_BARE)/$(2)
 	@echo "    [CREATE] $$@"
 	$$(qexec)$$(SRC_PATH_BARE)/build/make/rtcd.pl --arch=$$(TGT_ISA) \
          --sym=$(1) \
-          --config=$$(CONFIG_DIR)$$(target)-$$(TOOLCHAIN).mk \
+          --config=$$(CONFIG_DIR)$$(target)$$(if $$(FAT_ARCHS),,-$$(TOOLCHAIN)).mk \
          $$(RTCD_OPTIONS) $$^ > $$@
 CLEAN-OBJS += $$(BUILD_PFX)$(1).h
 RTCD += $$(BUILD_PFX)$(1).h
@@ -34,6 +34,13 @@ endef
 CODEC_SRCS-yes += CHANGELOG
 CODEC_SRCS-yes += libs.mk

+# If this is a universal (fat) binary, then all the subarchitectures have
+# already been built and our job is to stitch them together. The
+# BUILD_LIBVPX variable indicates whether we should be building
+# (compiling, linking) the library. The LIPO_LIBVPX variable indicates
+# that we're stitching.
+$(eval $(if $(filter universal%,$(TOOLCHAIN)),LIPO_LIBVPX,BUILD_LIBVPX):=yes)
+
 include $(SRC_PATH_BARE)/vpx/vpx_codec.mk
 CODEC_SRCS-yes += $(addprefix vpx/,$(call enabled,API_SRCS))
 CODEC_DOC_SRCS += $(addprefix vpx/,$(call enabled,API_DOC_SRCS))
@@ -47,45 +54,61 @@ CODEC_SRCS-yes += $(addprefix vpx_scale/,$(call enabled,SCALE_SRCS))
 include $(SRC_PATH_BARE)/vpx_ports/vpx_ports.mk
 CODEC_SRCS-yes += $(addprefix vpx_ports/,$(call enabled,PORTS_SRCS))

-include $(SRC_PATH_BARE)/vpx_dsp/vpx_dsp.mk
-CODEC_SRCS-yes += $(addprefix vpx_dsp/,$(call enabled,DSP_SRCS))
-
-include $(SRC_PATH_BARE)/vpx_util/vpx_util.mk
-CODEC_SRCS-yes += $(addprefix vpx_util/,$(call enabled,UTIL_SRCS))
-
-#  VP10 make file
-ifeq ($(CONFIG_VP10),yes)
-  VP10_PREFIX=vp10/
-  include $(SRC_PATH_BARE)/$(VP10_PREFIX)vp10_common.mk
+ifneq ($(CONFIG_VP8_ENCODER)$(CONFIG_VP8_DECODER),)
+  VP8_PREFIX=vp8/
+  include $(SRC_PATH_BARE)/$(VP8_PREFIX)vp8_common.mk
 endif

-ifeq ($(CONFIG_VP10_ENCODER),yes)
-  VP10_PREFIX=vp10/
-  include $(SRC_PATH_BARE)/$(VP10_PREFIX)vp10cx.mk
-  CODEC_SRCS-yes += $(addprefix $(VP10_PREFIX),$(call enabled,VP10_CX_SRCS))
-  CODEC_EXPORTS-yes += $(addprefix $(VP10_PREFIX),$(VP10_CX_EXPORTS))
-  CODEC_SRCS-yes += $(VP10_PREFIX)vp10cx.mk vpx/vp8.h vpx/vp8cx.h
+ifeq ($(CONFIG_VP8_ENCODER),yes)
+  include $(SRC_PATH_BARE)/$(VP8_PREFIX)vp8cx.mk
+  CODEC_SRCS-yes += $(addprefix $(VP8_PREFIX),$(call enabled,VP8_CX_SRCS))
+  CODEC_EXPORTS-yes += $(addprefix $(VP8_PREFIX),$(VP8_CX_EXPORTS))
+  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8cx.h
+  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP8_PREFIX)/%
+  CODEC_DOC_SECTIONS += vp8 vp8_encoder
+endif
+
+ifeq ($(CONFIG_VP8_DECODER),yes)
+  include $(SRC_PATH_BARE)/$(VP8_PREFIX)vp8dx.mk
+  CODEC_SRCS-yes += $(addprefix $(VP8_PREFIX),$(call enabled,VP8_DX_SRCS))
+  CODEC_EXPORTS-yes += $(addprefix $(VP8_PREFIX),$(VP8_DX_EXPORTS))
+  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8dx.h
+  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP8_PREFIX)/%
+  CODEC_DOC_SECTIONS += vp8 vp8_decoder
+endif
+
+ifneq ($(CONFIG_VP9_ENCODER)$(CONFIG_VP9_DECODER),)
+  VP9_PREFIX=vp9/
+  include $(SRC_PATH_BARE)/$(VP9_PREFIX)vp9_common.mk
+endif
+
+ifeq ($(CONFIG_VP9_ENCODER),yes)
+  VP9_PREFIX=vp9/
+  include $(SRC_PATH_BARE)/$(VP9_PREFIX)vp9cx.mk
+  CODEC_SRCS-yes += $(addprefix $(VP9_PREFIX),$(call enabled,VP9_CX_SRCS))
+  CODEC_EXPORTS-yes += $(addprefix $(VP9_PREFIX),$(VP9_CX_EXPORTS))
+  CODEC_SRCS-yes += $(VP9_PREFIX)vp9cx.mk vpx/vp8.h vpx/vp8cx.h
  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8cx.h
  INSTALL-LIBS-$(CONFIG_SPATIAL_SVC) += include/vpx/svc_context.h
-  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP10_PREFIX)/%
+  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP9_PREFIX)/%
  CODEC_DOC_SRCS += vpx/vp8.h vpx/vp8cx.h
  CODEC_DOC_SECTIONS += vp9 vp9_encoder
 endif

-ifeq ($(CONFIG_VP10_DECODER),yes)
-  VP10_PREFIX=vp10/
-  include $(SRC_PATH_BARE)/$(VP10_PREFIX)vp10dx.mk
-  CODEC_SRCS-yes += $(addprefix $(VP10_PREFIX),$(call enabled,VP10_DX_SRCS))
-  CODEC_EXPORTS-yes += $(addprefix $(VP10_PREFIX),$(VP10_DX_EXPORTS))
-  CODEC_SRCS-yes += $(VP10_PREFIX)vp10dx.mk vpx/vp8.h vpx/vp8dx.h
+ifeq ($(CONFIG_VP9_DECODER),yes)
+  VP9_PREFIX=vp9/
+  include $(SRC_PATH_BARE)/$(VP9_PREFIX)vp9dx.mk
+  CODEC_SRCS-yes += $(addprefix $(VP9_PREFIX),$(call enabled,VP9_DX_SRCS))
+  CODEC_EXPORTS-yes += $(addprefix $(VP9_PREFIX),$(VP9_DX_EXPORTS))
+  CODEC_SRCS-yes += $(VP9_PREFIX)vp9dx.mk vpx/vp8.h vpx/vp8dx.h
  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8dx.h
-  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP10_PREFIX)/%
+  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP9_PREFIX)/%
  CODEC_DOC_SRCS += vpx/vp8.h vpx/vp8dx.h
  CODEC_DOC_SECTIONS += vp9 vp9_decoder
 endif

-VP10_PREFIX=vp10/
-$(BUILD_PFX)$(VP10_PREFIX)%.c.o: CFLAGS += -Wextra
+VP9_PREFIX=vp9/
+$(BUILD_PFX)$(VP9_PREFIX)%.c.o: CFLAGS += -Wextra

 ifeq ($(CONFIG_ENCODERS),yes)
  CODEC_DOC_SECTIONS += encoder
@@ -114,22 +137,19 @@ INSTALL_MAPS += $(foreach p,$(VS_PLATFORMS),$(LIBSUBDIR)/$(p)/%  $(p)/Release/%)
 INSTALL_MAPS += $(foreach p,$(VS_PLATFORMS),$(LIBSUBDIR)/$(p)/%  $(p)/Debug/%)
 endif

-CODEC_SRCS-yes += build/make/version.sh
-CODEC_SRCS-yes += build/make/rtcd.pl
-CODEC_SRCS-yes += vpx_ports/emmintrin_compat.h
-CODEC_SRCS-yes += vpx_ports/mem_ops.h
-CODEC_SRCS-yes += vpx_ports/mem_ops_aligned.h
-CODEC_SRCS-yes += vpx_ports/vpx_once.h
-CODEC_SRCS-yes += $(BUILD_PFX)vpx_config.c
+CODEC_SRCS-$(BUILD_LIBVPX) += build/make/version.sh
+CODEC_SRCS-$(BUILD_LIBVPX) += build/make/rtcd.pl
+CODEC_SRCS-$(BUILD_LIBVPX) += vpx_ports/emmintrin_compat.h
+CODEC_SRCS-$(BUILD_LIBVPX) += vpx_ports/mem_ops.h
+CODEC_SRCS-$(BUILD_LIBVPX) += vpx_ports/mem_ops_aligned.h
+CODEC_SRCS-$(BUILD_LIBVPX) += vpx_ports/vpx_once.h
+CODEC_SRCS-$(BUILD_LIBVPX) += $(BUILD_PFX)vpx_config.c
 INSTALL-SRCS-no += $(BUILD_PFX)vpx_config.c
 ifeq ($(ARCH_X86)$(ARCH_X86_64),yes)
 INSTALL-SRCS-$(CONFIG_CODEC_SRCS) += third_party/x86inc/x86inc.asm
 endif
-CODEC_EXPORTS-yes += vpx/exports_com
+CODEC_EXPORTS-$(BUILD_LIBVPX) += vpx/exports_com
 CODEC_EXPORTS-$(CONFIG_ENCODERS) += vpx/exports_enc
-ifeq ($(CONFIG_SPATIAL_SVC),yes)
-CODEC_EXPORTS-$(CONFIG_ENCODERS) += vpx/exports_spatial_svc
-endif
 CODEC_EXPORTS-$(CONFIG_DECODERS) += vpx/exports_dec

 INSTALL-LIBS-yes += include/vpx/vpx_codec.h
@@ -195,7 +215,7 @@ vpx.$(VCPROJ_SFX): $(CODEC_SRCS) vpx.def
            $(filter-out $(addprefix %, $(ASM_INCLUDES)), $^) \
            --src-path-bare="$(SRC_PATH_BARE)" \

-PROJECTS-yes += vpx.$(VCPROJ_SFX)
+PROJECTS-$(BUILD_LIBVPX) += vpx.$(VCPROJ_SFX)

 vpx.$(VCPROJ_SFX): vpx_config.asm
 vpx.$(VCPROJ_SFX): $(RTCD)
@@ -203,49 +223,32 @@ vpx.$(VCPROJ_SFX): $(RTCD)
 endif
 else
 LIBVPX_OBJS=$(call objs,$(CODEC_SRCS))
-OBJS-yes += $(LIBVPX_OBJS)
-LIBS-$(if yes,$(CONFIG_STATIC)) += $(BUILD_PFX)libvpx.a $(BUILD_PFX)libvpx_g.a
+OBJS-$(BUILD_LIBVPX) += $(LIBVPX_OBJS)
+LIBS-$(if $(BUILD_LIBVPX),$(CONFIG_STATIC)) += $(BUILD_PFX)libvpx.a $(BUILD_PFX)libvpx_g.a
 $(BUILD_PFX)libvpx_g.a: $(LIBVPX_OBJS)

-SO_VERSION_MAJOR := 3
-SO_VERSION_MINOR := 0
-SO_VERSION_PATCH := 0
+
+BUILD_LIBVPX_SO         := $(if $(BUILD_LIBVPX),$(CONFIG_SHARED))
+
 ifeq ($(filter darwin%,$(TGT_OS)),$(TGT_OS))
-LIBVPX_SO               := libvpx.$(SO_VERSION_MAJOR).dylib
-SHARED_LIB_SUF          := .dylib
+LIBVPX_SO               := libvpx.$(VERSION_MAJOR).dylib
 EXPORT_FILE             := libvpx.syms
 LIBVPX_SO_SYMLINKS      := $(addprefix $(LIBSUBDIR)/, \
                             libvpx.dylib  )
 else
-ifeq ($(filter iphonesimulator%,$(TGT_OS)),$(TGT_OS))
-LIBVPX_SO               := libvpx.$(SO_VERSION_MAJOR).dylib
-SHARED_LIB_SUF          := .dylib
-EXPORT_FILE             := libvpx.syms
-LIBVPX_SO_SYMLINKS      := $(addprefix $(LIBSUBDIR)/, libvpx.dylib)
-else
-ifeq ($(filter os2%,$(TGT_OS)),$(TGT_OS))
-LIBVPX_SO               := libvpx$(SO_VERSION_MAJOR).dll
-SHARED_LIB_SUF          := _dll.a
-EXPORT_FILE             := libvpx.def
-LIBVPX_SO_SYMLINKS      :=
-LIBVPX_SO_IMPLIB        := libvpx_dll.a
-else
-LIBVPX_SO               := libvpx.so.$(SO_VERSION_MAJOR).$(SO_VERSION_MINOR).$(SO_VERSION_PATCH)
-SHARED_LIB_SUF          := .so
+LIBVPX_SO               := libvpx.so.$(VERSION_MAJOR).$(VERSION_MINOR).$(VERSION_PATCH)
 EXPORT_FILE             := libvpx.ver
+SYM_LINK                := libvpx.so
 LIBVPX_SO_SYMLINKS      := $(addprefix $(LIBSUBDIR)/, \
-                             libvpx.so libvpx.so.$(SO_VERSION_MAJOR) \
-                             libvpx.so.$(SO_VERSION_MAJOR).$(SO_VERSION_MINOR))
-endif
-endif
+                             libvpx.so libvpx.so.$(VERSION_MAJOR) \
+                             libvpx.so.$(VERSION_MAJOR).$(VERSION_MINOR))
 endif

-LIBS-$(CONFIG_SHARED) += $(BUILD_PFX)$(LIBVPX_SO)\
-                           $(notdir $(LIBVPX_SO_SYMLINKS)) \
-                           $(if $(LIBVPX_SO_IMPLIB), $(BUILD_PFX)$(LIBVPX_SO_IMPLIB))
+LIBS-$(BUILD_LIBVPX_SO) += $(BUILD_PFX)$(LIBVPX_SO)\
+                           $(notdir $(LIBVPX_SO_SYMLINKS))
 $(BUILD_PFX)$(LIBVPX_SO): $(LIBVPX_OBJS) $(EXPORT_FILE)
 $(BUILD_PFX)$(LIBVPX_SO): extralibs += -lm
-$(BUILD_PFX)$(LIBVPX_SO): SONAME = libvpx.so.$(SO_VERSION_MAJOR)
+$(BUILD_PFX)$(LIBVPX_SO): SONAME = libvpx.so.$(VERSION_MAJOR)
 $(BUILD_PFX)$(LIBVPX_SO): EXPORTS_FILE = $(EXPORT_FILE)

 libvpx.ver: $(call enabled,CODEC_EXPORTS)
@@ -260,19 +263,6 @@ libvpx.syms: $(call enabled,CODEC_EXPORTS)
 	$(qexec)awk '{print "_"$$2}' $^ >$@
 CLEAN-OBJS += libvpx.syms

-libvpx.def: $(call enabled,CODEC_EXPORTS)
-	@echo "    [CREATE] $@"
-	$(qexec)echo LIBRARY $(LIBVPX_SO:.dll=) INITINSTANCE TERMINSTANCE > $@
-	$(qexec)echo "DATA MULTIPLE NONSHARED" >> $@
-	$(qexec)echo "EXPORTS" >> $@
-	$(qexec)awk '!/vpx_svc_*/ {print "_"$$2}' $^ >>$@
-CLEAN-OBJS += libvpx.def
-
-libvpx_dll.a: $(LIBVPX_SO)
-	@echo "    [IMPLIB] $@"
-	$(qexec)emximp -o $@ $<
-CLEAN-OBJS += libvpx_dll.a
-
 define libvpx_symlink_template
 $(1): $(2)
 	@echo "    [LN]     $(2) $$@"
@@ -288,12 +278,11 @@ $(eval $(call libvpx_symlink_template,\
    $(LIBVPX_SO)))


-INSTALL-LIBS-$(CONFIG_SHARED) += $(LIBVPX_SO_SYMLINKS)
-INSTALL-LIBS-$(CONFIG_SHARED) += $(LIBSUBDIR)/$(LIBVPX_SO)
-INSTALL-LIBS-$(CONFIG_SHARED) += $(if $(LIBVPX_SO_IMPLIB),$(LIBSUBDIR)/$(LIBVPX_SO_IMPLIB))
+INSTALL-LIBS-$(BUILD_LIBVPX_SO) += $(LIBVPX_SO_SYMLINKS)
+INSTALL-LIBS-$(BUILD_LIBVPX_SO) += $(LIBSUBDIR)/$(LIBVPX_SO)


-LIBS-yes += vpx.pc
+LIBS-$(BUILD_LIBVPX) += vpx.pc
 vpx.pc: config.mk libs.mk
 	@echo "    [CREATE] $@"
 	$(qexec)echo '# pkg-config file from libvpx $(VERSION_STRING)' > $@
@@ -319,6 +308,9 @@ INSTALL_MAPS += $(LIBSUBDIR)/pkgconfig/%.pc %.pc
 CLEAN-OBJS += vpx.pc
 endif

+LIBS-$(LIPO_LIBVPX) += libvpx.a
+$(eval $(if $(LIPO_LIBVPX),$(call lipo_lib_template,libvpx.a)))
+
 #
 # Rule to make assembler configuration file from C configuration file
 #
@@ -348,12 +340,6 @@ $(filter %$(ASM).o,$(OBJS-yes)): $(BUILD_PFX)vpx_config.asm
 $(shell $(SRC_PATH_BARE)/build/make/version.sh "$(SRC_PATH_BARE)" $(BUILD_PFX)vpx_version.h)
 CLEAN-OBJS += $(BUILD_PFX)vpx_version.h

-#
-# Add include path for libwebm sources.
-#
-ifeq ($(CONFIG_WEBM_IO),yes)
-  CXXFLAGS += -I$(SRC_PATH_BARE)/third_party/libwebm
-endif

 ##
 ## libvpx test directives
@@ -363,15 +349,11 @@ LIBVPX_TEST_DATA_PATH ?= .

 include $(SRC_PATH_BARE)/test/test.mk
 LIBVPX_TEST_SRCS=$(addprefix test/,$(call enabled,LIBVPX_TEST_SRCS))
-LIBVPX_TEST_BIN=./test_libvpx$(EXE_SFX)
+LIBVPX_TEST_BINS=./test_libvpx$(EXE_SFX)
 LIBVPX_TEST_DATA=$(addprefix $(LIBVPX_TEST_DATA_PATH)/,\
                     $(call enabled,LIBVPX_TEST_DATA))
 libvpx_test_data_url=http://downloads.webmproject.org/test_data/libvpx/$(1)

-TEST_INTRA_PRED_SPEED_BIN=./test_intra_pred_speed$(EXE_SFX)
-TEST_INTRA_PRED_SPEED_SRCS=$(addprefix test/,$(call enabled,TEST_INTRA_PRED_SPEED_SRCS))
-TEST_INTRA_PRED_SPEED_OBJS := $(sort $(call objs,$(TEST_INTRA_PRED_SPEED_SRCS)))
-
 libvpx_test_srcs.txt:
 	@echo "    [CREATE] $@"
 	@echo $(LIBVPX_TEST_SRCS) | xargs -n1 echo | LC_ALL=C sort -u > $@
@@ -389,10 +371,12 @@ testdata:: $(LIBVPX_TEST_DATA)
          if [ -n "$${sha1sum}" ]; then\
            set -e;\
            echo "Checking test data:";\
-            for f in $(call enabled,LIBVPX_TEST_DATA); do\
-                grep $$f $(SRC_PATH_BARE)/test/test-data.sha1 |\
-                    (cd $(LIBVPX_TEST_DATA_PATH); $${sha1sum} -c);\
-            done; \
+            if [ -n "$(LIBVPX_TEST_DATA)" ]; then\
+                for f in $(call enabled,LIBVPX_TEST_DATA); do\
+                    grep $$f $(SRC_PATH_BARE)/test/test-data.sha1 |\
+                        (cd $(LIBVPX_TEST_DATA_PATH); $${sha1sum} -c);\
+                done; \
+            fi; \
        else\
            echo "Skipping test data integrity check, sha1sum not found.";\
        fi
@@ -429,30 +413,11 @@ test_libvpx.$(VCPROJ_SFX): $(LIBVPX_TEST_SRCS) vpx.$(VCPROJ_SFX) gtest.$(VCPROJ_
            $(if $(CONFIG_STATIC_MSVCRT),--static-crt) \
            --out=$@ $(INTERNAL_CFLAGS) $(CFLAGS) \
            -I. -I"$(SRC_PATH_BARE)/third_party/googletest/src/include" \
-            $(if $(CONFIG_WEBM_IO),-I"$(SRC_PATH_BARE)/third_party/libwebm") \
            -L. -l$(CODEC_LIB) -l$(GTEST_LIB) $^

 PROJECTS-$(CONFIG_MSVS) += test_libvpx.$(VCPROJ_SFX)

-LIBVPX_TEST_BIN := $(addprefix $(TGT_OS:win64=x64)/Release/,$(notdir $(LIBVPX_TEST_BIN)))
-
-ifneq ($(strip $(TEST_INTRA_PRED_SPEED_OBJS)),)
-PROJECTS-$(CONFIG_MSVS) += test_intra_pred_speed.$(VCPROJ_SFX)
-test_intra_pred_speed.$(VCPROJ_SFX): $(TEST_INTRA_PRED_SPEED_SRCS) vpx.$(VCPROJ_SFX) gtest.$(VCPROJ_SFX)
-	@echo "    [CREATE] $@"
-	$(qexec)$(GEN_VCPROJ) \
-            --exe \
-            --target=$(TOOLCHAIN) \
-            --name=test_intra_pred_speed \
-            -D_VARIADIC_MAX=10 \
-            --proj-guid=CD837F5F-52D8-4314-A370-895D614166A7 \
-            --ver=$(CONFIG_VS_VERSION) \
-            --src-path-bare="$(SRC_PATH_BARE)" \
-            $(if $(CONFIG_STATIC_MSVCRT),--static-crt) \
-            --out=$@ $(INTERNAL_CFLAGS) $(CFLAGS) \
-            -I. -I"$(SRC_PATH_BARE)/third_party/googletest/src/include" \
-            -L. -l$(CODEC_LIB) -l$(GTEST_LIB) $^
-endif  # TEST_INTRA_PRED_SPEED
+LIBVPX_TEST_BINS := $(addprefix $(TGT_OS:win64=x64)/Release/,$(notdir $(LIBVPX_TEST_BINS)))
 endif
 else

@@ -463,54 +428,45 @@ ifeq ($(filter win%,$(TGT_OS)),$(TGT_OS))
 # Disabling pthreads globally will cause issues on darwin and possibly elsewhere
 $(GTEST_OBJS) $(GTEST_OBJS:.o=.d): CXXFLAGS += -DGTEST_HAS_PTHREAD=0
 endif
-GTEST_INCLUDES := -I$(SRC_PATH_BARE)/third_party/googletest/src
-GTEST_INCLUDES += -I$(SRC_PATH_BARE)/third_party/googletest/src/include
-$(GTEST_OBJS) $(GTEST_OBJS:.o=.d): CXXFLAGS += $(GTEST_INCLUDES)
-OBJS-yes += $(GTEST_OBJS)
-LIBS-yes += $(BUILD_PFX)libgtest.a $(BUILD_PFX)libgtest_g.a
+$(GTEST_OBJS) $(GTEST_OBJS:.o=.d): CXXFLAGS += -I$(SRC_PATH_BARE)/third_party/googletest/src
+$(GTEST_OBJS) $(GTEST_OBJS:.o=.d): CXXFLAGS += -I$(SRC_PATH_BARE)/third_party/googletest/src/include
+OBJS-$(BUILD_LIBVPX) += $(GTEST_OBJS)
+LIBS-$(BUILD_LIBVPX) += $(BUILD_PFX)libgtest.a $(BUILD_PFX)libgtest_g.a
 $(BUILD_PFX)libgtest_g.a: $(GTEST_OBJS)

 LIBVPX_TEST_OBJS=$(sort $(call objs,$(LIBVPX_TEST_SRCS)))
-$(LIBVPX_TEST_OBJS) $(LIBVPX_TEST_OBJS:.o=.d): CXXFLAGS += $(GTEST_INCLUDES)
-OBJS-yes += $(LIBVPX_TEST_OBJS)
-BINS-yes += $(LIBVPX_TEST_BIN)
+$(LIBVPX_TEST_OBJS) $(LIBVPX_TEST_OBJS:.o=.d): CXXFLAGS += -I$(SRC_PATH_BARE)/third_party/googletest/src
+$(LIBVPX_TEST_OBJS) $(LIBVPX_TEST_OBJS:.o=.d): CXXFLAGS += -I$(SRC_PATH_BARE)/third_party/googletest/src/include
+OBJS-$(BUILD_LIBVPX) += $(LIBVPX_TEST_OBJS)
+BINS-$(BUILD_LIBVPX) += $(LIBVPX_TEST_BINS)

 CODEC_LIB=$(if $(CONFIG_DEBUG_LIBS),vpx_g,vpx)
-CODEC_LIB_SUF=$(if $(CONFIG_SHARED),$(SHARED_LIB_SUF),.a)
-TEST_LIBS := lib$(CODEC_LIB)$(CODEC_LIB_SUF) libgtest.a
-$(LIBVPX_TEST_BIN): $(TEST_LIBS)
-$(eval $(call linkerxx_template,$(LIBVPX_TEST_BIN), \
-              $(LIBVPX_TEST_OBJS) \
-              -L. -lvpx -lgtest $(extralibs) -lm))
+CODEC_LIB_SUF=$(if $(CONFIG_SHARED),.so,.a)
+$(foreach bin,$(LIBVPX_TEST_BINS),\
+    $(if $(BUILD_LIBVPX),$(eval $(bin): \
+        lib$(CODEC_LIB)$(CODEC_LIB_SUF) libgtest.a ))\
+    $(if $(BUILD_LIBVPX),$(eval $(call linkerxx_template,$(bin),\
+        $(LIBVPX_TEST_OBJS) \
+        -L. -lvpx -lgtest $(extralibs) -lm)\
+        )))\
+    $(if $(LIPO_LIBS),$(eval $(call lipo_bin_template,$(bin))))\

-ifneq ($(strip $(TEST_INTRA_PRED_SPEED_OBJS)),)
-$(TEST_INTRA_PRED_SPEED_OBJS) $(TEST_INTRA_PRED_SPEED_OBJS:.o=.d): CXXFLAGS += $(GTEST_INCLUDES)
-OBJS-yes += $(TEST_INTRA_PRED_SPEED_OBJS)
-BINS-yes += $(TEST_INTRA_PRED_SPEED_BIN)
-
-$(TEST_INTRA_PRED_SPEED_BIN): $(TEST_LIBS)
-$(eval $(call linkerxx_template,$(TEST_INTRA_PRED_SPEED_BIN), \
-              $(TEST_INTRA_PRED_SPEED_OBJS) \
-              -L. -lvpx -lgtest $(extralibs) -lm))
-endif  # TEST_INTRA_PRED_SPEED
-
-endif  # CONFIG_UNIT_TESTS
+endif

 # Install test sources only if codec source is included
 INSTALL-SRCS-$(CONFIG_CODEC_SRCS) += $(patsubst $(SRC_PATH_BARE)/%,%,\
    $(shell find $(SRC_PATH_BARE)/third_party/googletest -type f))
 INSTALL-SRCS-$(CONFIG_CODEC_SRCS) += $(LIBVPX_TEST_SRCS)
-INSTALL-SRCS-$(CONFIG_CODEC_SRCS) += $(TEST_INTRA_PRED_SPEED_SRCS)

 define test_shard_template
 test:: test_shard.$(1)
-test-no-data-check:: test_shard_ndc.$(1)
-test_shard.$(1) test_shard_ndc.$(1): $(LIBVPX_TEST_BIN)
+test_shard.$(1): $(LIBVPX_TEST_BINS) testdata
 	@set -e; \
-	 export GTEST_SHARD_INDEX=$(1); \
-	 export GTEST_TOTAL_SHARDS=$(2); \
-	 $(LIBVPX_TEST_BIN)
-test_shard.$(1): testdata
+	 for t in $(LIBVPX_TEST_BINS); do \
+	   export GTEST_SHARD_INDEX=$(1); \
+	   export GTEST_TOTAL_SHARDS=$(2); \
+	   $$$$t; \
+	 done
 .PHONY: test_shard.$(1)
 endef

@@ -555,16 +511,15 @@ ifeq ($(CONFIG_MSVS),yes)
 # TODO(tomfinegan): Support running the debug versions of tools?
 TEST_BIN_PATH := $(addsuffix /$(TGT_OS:win64=x64)/Release, $(TEST_BIN_PATH))
 endif
-utiltest utiltest-no-data-check:
+utiltest: testdata
 	$(qexec)$(SRC_PATH_BARE)/test/vpxdec.sh \
 		--test-data-path $(LIBVPX_TEST_DATA_PATH) \
 		--bin-path $(TEST_BIN_PATH)
 	$(qexec)$(SRC_PATH_BARE)/test/vpxenc.sh \
 		--test-data-path $(LIBVPX_TEST_DATA_PATH) \
 		--bin-path $(TEST_BIN_PATH)
-utiltest: testdata
 else
-utiltest utiltest-no-data-check:
+utiltest:
 	@echo Unit tests must be enabled to make the utiltest target.
 endif

@@ -582,12 +537,11 @@ ifeq ($(CONFIG_MSVS),yes)
 # TODO(tomfinegan): Support running the debug versions of tools?
 EXAMPLES_BIN_PATH := $(TGT_OS:win64=x64)/Release
 endif
-exampletest exampletest-no-data-check: examples
+exampletest: examples testdata
 	$(qexec)$(SRC_PATH_BARE)/test/examples.sh \
 		--test-data-path $(LIBVPX_TEST_DATA_PATH) \
 		--bin-path $(EXAMPLES_BIN_PATH)
-exampletest: testdata
 else
-exampletest exampletest-no-data-check:
+exampletest:
 	@echo Unit tests must be enabled to make the exampletest target.
 endif
--- a/md5_utils.c
+++ b/md5_utils.c
@@ -24,7 +24,7 @@

 #include "md5_utils.h"

-static void
+void
 byteSwap(UWORD32 *buf, unsigned words) {
  md5byte *p;

@@ -150,23 +150,12 @@ MD5Final(md5byte digest[16], struct MD5Context *ctx) {
 #define MD5STEP(f,w,x,y,z,in,s) \
  (w += f(x,y,z) + in, w = (w<<s | w>>(32-s)) + x)

-#if defined(__clang__) && defined(__has_attribute)
-#if __has_attribute(no_sanitize)
-#define VPX_NO_UNSIGNED_OVERFLOW_CHECK \
-  __attribute__((no_sanitize("unsigned-integer-overflow")))
-#endif
-#endif
-
-#ifndef VPX_NO_UNSIGNED_OVERFLOW_CHECK
-#define VPX_NO_UNSIGNED_OVERFLOW_CHECK
-#endif
-
 /*
 * The core of the MD5 algorithm, this alters an existing MD5 hash to
 * reflect the addition of 16 longwords of new data.  MD5Update blocks
 * the data and converts bytes into longwords for this routine.
 */
-VPX_NO_UNSIGNED_OVERFLOW_CHECK void
+void
 MD5Transform(UWORD32 buf[4], UWORD32 const in[16]) {
  register UWORD32 a, b, c, d;

@@ -249,6 +238,4 @@ MD5Transform(UWORD32 buf[4], UWORD32 const in[16]) {
  buf[3] += d;
 }

-#undef VPX_NO_UNSIGNED_OVERFLOW_CHECK
-
 #endif
--- a/rate_hist.c
+++ b/rate_hist.c
@@ -88,9 +88,6 @@ void update_rate_histogram(struct rate_hist *hist,
  if (now < cfg->rc_buf_initial_sz)
    return;

-  if (!cfg->rc_target_bitrate)
-    return;
-
  then = now;

  /* Sum the size over the past rc_buf_sz ms */
--- a/test/acm_random.h
+++ b/test/acm_random.h
@@ -32,12 +32,6 @@ class ACMRandom {
    return (value >> 15) & 0xffff;
  }

-  int16_t Rand9Signed(void) {
-    // Use 9 bits: values between 255 (0x0FF) and -256 (0x100).
-    const uint32_t value = random_.Generate(512);
-    return static_cast<int16_t>(value) - 256;
-  }
-
  uint8_t Rand8(void) {
    const uint32_t value =
        random_.Generate(testing::internal::Random::kMaxRange);
--- a/test/active_map_refresh_test.cc
+++ b/test/active_map_refresh_test.cc
@@ -1,134 +0,0 @@
-/*
- *  Copyright (c) 2015 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-#include <algorithm>
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/util.h"
-#include "test/y4m_video_source.h"
-
-namespace {
-
-// Check if any pixel in a 16x16 macroblock varies between frames.
-int CheckMb(const vpx_image_t &current, const vpx_image_t &previous,
-            int mb_r, int mb_c) {
-  for (int plane = 0; plane < 3; plane++) {
-    int r = 16 * mb_r;
-    int c0 = 16 * mb_c;
-    int r_top = std::min(r + 16, static_cast<int>(current.d_h));
-    int c_top = std::min(c0 + 16, static_cast<int>(current.d_w));
-    r = std::max(r, 0);
-    c0 = std::max(c0, 0);
-    if (plane > 0 && current.x_chroma_shift) {
-      c_top = (c_top + 1) >> 1;
-      c0 >>= 1;
-    }
-    if (plane > 0 && current.y_chroma_shift) {
-      r_top = (r_top + 1) >> 1;
-      r >>= 1;
-    }
-    for (; r < r_top; ++r) {
-      for (int c = c0; c < c_top; ++c) {
-        if (current.planes[plane][current.stride[plane] * r + c] !=
-            previous.planes[plane][previous.stride[plane] * r + c])
-          return 1;
-      }
-    }
-  }
-  return 0;
-}
-
-void GenerateMap(int mb_rows, int mb_cols, const vpx_image_t &current,
-                 const vpx_image_t &previous, uint8_t *map) {
-  for (int mb_r = 0; mb_r < mb_rows; ++mb_r) {
-    for (int mb_c = 0; mb_c < mb_cols; ++mb_c) {
-      map[mb_r * mb_cols + mb_c] = CheckMb(current, previous, mb_r, mb_c);
-    }
-  }
-}
-
-const int kAqModeCyclicRefresh = 3;
-
-class ActiveMapRefreshTest
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- protected:
-  ActiveMapRefreshTest() : EncoderTest(GET_PARAM(0)) {}
-  virtual ~ActiveMapRefreshTest() {}
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(GET_PARAM(1));
-    cpu_used_ = GET_PARAM(2);
-  }
-
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    ::libvpx_test::Y4mVideoSource *y4m_video =
-        static_cast<libvpx_test::Y4mVideoSource *>(video);
-    if (video->frame() == 1) {
-      encoder->Control(VP8E_SET_CPUUSED, cpu_used_);
-      encoder->Control(VP9E_SET_AQ_MODE, kAqModeCyclicRefresh);
-    } else if (video->frame() >= 2 && video->img()) {
-      vpx_image_t *current = video->img();
-      vpx_image_t *previous = y4m_holder_->img();
-      ASSERT_TRUE(previous != NULL);
-      vpx_active_map_t map = vpx_active_map_t();
-      const int width = static_cast<int>(current->d_w);
-      const int height = static_cast<int>(current->d_h);
-      const int mb_width = (width + 15) / 16;
-      const int mb_height = (height + 15) / 16;
-      uint8_t *active_map = new uint8_t[mb_width * mb_height];
-      GenerateMap(mb_height, mb_width, *current, *previous, active_map);
-      map.cols = mb_width;
-      map.rows = mb_height;
-      map.active_map = active_map;
-      encoder->Control(VP8E_SET_ACTIVEMAP, &map);
-      delete[] active_map;
-    }
-    if (video->img()) {
-      y4m_video->SwapBuffers(y4m_holder_);
-    }
-  }
-
-  int cpu_used_;
-  ::libvpx_test::Y4mVideoSource *y4m_holder_;
-};
-
-TEST_P(ActiveMapRefreshTest, Test) {
-  cfg_.g_lag_in_frames = 0;
-  cfg_.g_profile = 1;
-  cfg_.rc_target_bitrate = 600;
-  cfg_.rc_resize_allowed = 0;
-  cfg_.rc_min_quantizer = 8;
-  cfg_.rc_max_quantizer = 30;
-  cfg_.g_pass = VPX_RC_ONE_PASS;
-  cfg_.rc_end_usage = VPX_CBR;
-  cfg_.kf_max_dist = 90000;
-
-#if CONFIG_VP10
-  const int nframes = codec_ == &libvpx_test::kVP10 ? 10 : 30;
-#else
-  const int nframes = 30;
-#endif  // CONFIG_VP10
-  ::libvpx_test::Y4mVideoSource video("desktop_credits.y4m", 0, nframes);
-  ::libvpx_test::Y4mVideoSource video_holder("desktop_credits.y4m", 0, nframes);
-  video_holder.Begin();
-  y4m_holder_ = &video_holder;
-
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-}
-
-#if CONFIG_VP10
-VP10_INSTANTIATE_TEST_CASE(ActiveMapRefreshTest,
-                           ::testing::Values(::libvpx_test::kRealTime),
-                           ::testing::Range(5, 6));
-#endif  // CONFIG_VP10
-}  // namespace
--- a/test/active_map_test.cc
+++ b/test/active_map_test.cc
@@ -65,39 +65,25 @@ class ActiveMapTest
    }
  }

-  void DoTest() {
-    // Validate that this non multiple of 64 wide clip encodes
-    cfg_.g_lag_in_frames = 0;
-    cfg_.rc_target_bitrate = 400;
-    cfg_.rc_resize_allowed = 0;
-    cfg_.g_pass = VPX_RC_ONE_PASS;
-    cfg_.rc_end_usage = VPX_CBR;
-    cfg_.kf_max_dist = 90000;
-    ::libvpx_test::I420VideoSource video("hantro_odd.yuv", kWidth, kHeight, 30,
-                                         1, 0, 20);
-
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  }
-
  int cpu_used_;
 };

 TEST_P(ActiveMapTest, Test) {
-  DoTest();
+  // Validate that this non multiple of 64 wide clip encodes
+  cfg_.g_lag_in_frames = 0;
+  cfg_.rc_target_bitrate = 400;
+  cfg_.rc_resize_allowed = 0;
+  cfg_.g_pass = VPX_RC_ONE_PASS;
+  cfg_.rc_end_usage = VPX_CBR;
+  cfg_.kf_max_dist = 90000;
+
+  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", kWidth, kHeight, 30,
+                                       1, 0, 20);
+
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

-class ActiveMapTestLarge : public ActiveMapTest {};
-
-TEST_P(ActiveMapTestLarge, Test) {
-  DoTest();
-}
-
-VP10_INSTANTIATE_TEST_CASE(ActiveMapTestLarge,
-                           ::testing::Values(::libvpx_test::kRealTime),
-                           ::testing::Range(0, 5));
-
-VP10_INSTANTIATE_TEST_CASE(ActiveMapTest,
-                           ::testing::Values(::libvpx_test::kRealTime),
-                           ::testing::Range(5, 9));
-
+VP9_INSTANTIATE_TEST_CASE(ActiveMapTest,
+                          ::testing::Values(::libvpx_test::kRealTime),
+                          ::testing::Range(0, 6));
 }  // namespace
--- a/test/add_noise_test.cc
+++ b/test/add_noise_test.cc
@@ -1,155 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-#include <math.h>
-#include "test/clear_system_state.h"
-#include "test/register_state_check.h"
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-#include "vpx_dsp/postproc.h"
-#include "vpx_mem/vpx_mem.h"
-
-namespace {
-
-// TODO(jimbankoski): make width and height integers not unsigned.
-typedef void (*AddNoiseFunc)(unsigned char *start, char *noise,
-                             char blackclamp[16], char whiteclamp[16],
-                             char bothclamp[16], unsigned int width,
-                             unsigned int height, int pitch);
-
-class AddNoiseTest
-    : public ::testing::TestWithParam<AddNoiseFunc> {
- public:
-  virtual void TearDown() {
-    libvpx_test::ClearSystemState();
-  }
-  virtual ~AddNoiseTest() {}
-};
-
-double stddev6(char a, char b, char c, char d, char e, char f) {
-  const double n = (a + b + c + d + e + f) / 6.0;
-  const double v = ((a - n) * (a - n) + (b - n) * (b - n) + (c - n) * (c - n) +
-                    (d - n) * (d - n) + (e - n) * (e - n) + (f - n) * (f - n)) /
-                   6.0;
-  return sqrt(v);
-}
-
-TEST_P(AddNoiseTest, CheckNoiseAdded) {
-  DECLARE_ALIGNED(16, char, blackclamp[16]);
-  DECLARE_ALIGNED(16, char, whiteclamp[16]);
-  DECLARE_ALIGNED(16, char, bothclamp[16]);
-  const int width  = 64;
-  const int height = 64;
-  const int image_size = width * height;
-  char noise[3072];
-  const int clamp = vpx_setup_noise(4.4, sizeof(noise), noise);
-
-  for (int i = 0; i < 16; i++) {
-    blackclamp[i] = clamp;
-    whiteclamp[i] = clamp;
-    bothclamp[i] = 2 * clamp;
-  }
-
-  uint8_t *const s = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
-  memset(s, 99, image_size);
-
-  ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
-                                      bothclamp, width, height, width));
-
-  // Check to make sure we don't end up having either the same or no added
-  // noise either vertically or horizontally.
-  for (int i = 0; i < image_size - 6 * width - 6; ++i) {
-    const double hd = stddev6(s[i] - 99, s[i + 1] - 99, s[i + 2] - 99,
-                              s[i + 3] - 99, s[i + 4] - 99, s[i + 5] - 99);
-    const double vd = stddev6(s[i] - 99, s[i + width] - 99,
-                              s[i + 2 * width] - 99, s[i + 3 * width] - 99,
-                              s[i + 4 * width] - 99, s[i + 5 * width] - 99);
-
-    EXPECT_NE(hd, 0);
-    EXPECT_NE(vd, 0);
-  }
-
-  // Initialize pixels in the image to 255 and check for roll over.
-  memset(s, 255, image_size);
-
-  ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
-                                      bothclamp, width, height, width));
-
-  // Check to make sure don't roll over.
-  for (int i = 0; i < image_size; ++i) {
-    EXPECT_GT(static_cast<int>(s[i]), clamp) << "i = " << i;
-  }
-
-  // Initialize pixels in the image to 0 and check for roll under.
-  memset(s, 0, image_size);
-
-  ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
-                                      bothclamp, width, height, width));
-
-  // Check to make sure don't roll under.
-  for (int i = 0; i < image_size; ++i) {
-    EXPECT_LT(static_cast<int>(s[i]), 255 - clamp) << "i = " << i;
-  }
-
-  vpx_free(s);
-}
-
-TEST_P(AddNoiseTest, CheckCvsAssembly) {
-  DECLARE_ALIGNED(16, char, blackclamp[16]);
-  DECLARE_ALIGNED(16, char, whiteclamp[16]);
-  DECLARE_ALIGNED(16, char, bothclamp[16]);
-  const int width  = 64;
-  const int height = 64;
-  const int image_size = width * height;
-  char noise[3072];
-
-  const int clamp = vpx_setup_noise(4.4, sizeof(noise), noise);
-
-  for (int i = 0; i < 16; i++) {
-    blackclamp[i] = clamp;
-    whiteclamp[i] = clamp;
-    bothclamp[i] = 2 * clamp;
-  }
-
-  uint8_t *const s = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
-  uint8_t *const d = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
-
-  memset(s, 99, image_size);
-  memset(d, 99, image_size);
-
-  srand(0);
-  ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
-                                      bothclamp, width, height, width));
-  srand(0);
-  ASM_REGISTER_STATE_CHECK(vpx_plane_add_noise_c(d, noise, blackclamp,
-                                                 whiteclamp, bothclamp,
-                                                 width, height, width));
-
-  for (int i = 0; i < image_size; ++i) {
-    EXPECT_EQ(static_cast<int>(s[i]), static_cast<int>(d[i])) << "i = " << i;
-  }
-
-  vpx_free(d);
-  vpx_free(s);
-}
-
-INSTANTIATE_TEST_CASE_P(C, AddNoiseTest,
-                        ::testing::Values(vpx_plane_add_noise_c));
-
-#if HAVE_SSE2
-INSTANTIATE_TEST_CASE_P(SSE2, AddNoiseTest,
-                        ::testing::Values(vpx_plane_add_noise_sse2));
-#endif
-
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(MSA, AddNoiseTest,
-                        ::testing::Values(vpx_plane_add_noise_msa));
-#endif
-}  // namespace
--- a/test/altref_test.cc
+++ b/test/altref_test.cc
@@ -14,86 +14,56 @@
 #include "test/util.h"
 namespace {

-class AltRefForcedKeyTestLarge
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
+// lookahead range: [kLookAheadMin, kLookAheadMax).
+const int kLookAheadMin = 5;
+const int kLookAheadMax = 26;
+
+class AltRefTest : public ::libvpx_test::EncoderTest,
+    public ::libvpx_test::CodecTestWithParam<int> {
 protected:
-  AltRefForcedKeyTestLarge()
-      : EncoderTest(GET_PARAM(0)),
-        encoding_mode_(GET_PARAM(1)),
-        cpu_used_(GET_PARAM(2)),
-        forced_kf_frame_num_(1),
-        frame_num_(0) {}
-  virtual ~AltRefForcedKeyTestLarge() {}
+  AltRefTest() : EncoderTest(GET_PARAM(0)), altref_count_(0) {}
+  virtual ~AltRefTest() {}

  virtual void SetUp() {
    InitializeConfig();
-    SetMode(encoding_mode_);
-    cfg_.rc_end_usage = VPX_VBR;
-    cfg_.g_threads = 0;
+    SetMode(libvpx_test::kTwoPassGood);
  }

-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (video->frame() == 0) {
-      encoder->Control(VP8E_SET_CPUUSED, cpu_used_);
+  virtual void BeginPassHook(unsigned int pass) {
+    altref_count_ = 0;
+  }
+
+  virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
+                                  libvpx_test::Encoder *encoder) {
+    if (video->frame() == 1) {
      encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
-      // override test default for tile columns if necessary.
-#if CONFIG_VP10_ENCODER
-      if (GET_PARAM(0) == &libvpx_test::kVP10) {
-        encoder->Control(VP9E_SET_TILE_COLUMNS, 6);
-      }
-#endif
+      encoder->Control(VP8E_SET_CPUUSED, 3);
    }
-    frame_flags_ =
-        (video->frame() == forced_kf_frame_num_) ? VPX_EFLAG_FORCE_KF : 0;
  }

  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
-    if (frame_num_ == forced_kf_frame_num_) {
-      ASSERT_TRUE(!!(pkt->data.frame.flags & VPX_FRAME_IS_KEY))
-          << "Frame #" << frame_num_ << " isn't a keyframe!";
-    }
-    ++frame_num_;
+    if (pkt->data.frame.flags & VPX_FRAME_IS_INVISIBLE) ++altref_count_;
  }

-  ::libvpx_test::TestMode encoding_mode_;
-  int cpu_used_;
-  unsigned int forced_kf_frame_num_;
-  unsigned int frame_num_;
+  int altref_count() const { return altref_count_; }
+
+ private:
+  int altref_count_;
 };

-TEST_P(AltRefForcedKeyTestLarge, Frame1IsKey) {
-  const vpx_rational timebase = { 1, 30 };
-  const int lag_values[] = { 3, 15, 25, -1 };
+TEST_P(AltRefTest, MonotonicTimestamps) {
+  const vpx_rational timebase = { 33333333, 1000000000 };
+  cfg_.g_timebase = timebase;
+  cfg_.rc_target_bitrate = 1000;
+  cfg_.g_lag_in_frames = GET_PARAM(1);

-  forced_kf_frame_num_ = 1;
-  for (int i = 0; lag_values[i] != -1; ++i) {
-    frame_num_ = 0;
-    cfg_.g_lag_in_frames = lag_values[i];
-    libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       timebase.den, timebase.num, 0, 30);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  }
+  libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                     timebase.den, timebase.num, 0, 30);
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+  EXPECT_GE(altref_count(), 1);
 }

-TEST_P(AltRefForcedKeyTestLarge, ForcedFrameIsKey) {
-  const vpx_rational timebase = { 1, 30 };
-  const int lag_values[] = { 3, 15, 25, -1 };
-
-  for (int i = 0; lag_values[i] != -1; ++i) {
-    frame_num_ = 0;
-    forced_kf_frame_num_ = lag_values[i] - 1;
-    cfg_.g_lag_in_frames = lag_values[i];
-    libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       timebase.den, timebase.num, 0, 30);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  }
-}
-
-VP10_INSTANTIATE_TEST_CASE(
-    AltRefForcedKeyTestLarge,
-    ::testing::Values(::libvpx_test::kOnePassGood),
-    ::testing::Range(0, 9));

+VP8_INSTANTIATE_TEST_CASE(AltRefTest,
+                          ::testing::Range(kLookAheadMin, kLookAheadMax));
 }  // namespace
--- a/test/android/Android.mk
+++ b/test/android/Android.mk
@@ -40,17 +40,9 @@ include $(CLEAR_VARS)
 LOCAL_ARM_MODE := arm
 LOCAL_MODULE := libvpx_test
 LOCAL_STATIC_LIBRARIES := gtest libwebm
-
-ifeq ($(ENABLE_SHARED),1)
-  LOCAL_SHARED_LIBRARIES := vpx
-else
-  LOCAL_STATIC_LIBRARIES += vpx
-endif
-
+LOCAL_SHARED_LIBRARIES := vpx
 include $(LOCAL_PATH)/test/test.mk
 LOCAL_C_INCLUDES := $(BINDINGS_DIR)
 FILTERED_SRC := $(sort $(filter %.cc %.c, $(LIBVPX_TEST_SRCS-yes)))
 LOCAL_SRC_FILES := $(addprefix ./test/, $(FILTERED_SRC))
-# some test files depend on *_rtcd.h, ensure they're generated first.
-$(eval $(call rtcd_dep_template))
 include $(BUILD_EXECUTABLE)
--- a/test/aq_segment_test.cc
+++ b/test/aq_segment_test.cc
@@ -38,22 +38,6 @@ class AqSegmentTest
    }
  }

-  void DoTest(int aq_mode) {
-    aq_mode_ = aq_mode;
-    cfg_.kf_max_dist = 12;
-    cfg_.rc_min_quantizer = 8;
-    cfg_.rc_max_quantizer = 56;
-    cfg_.rc_end_usage = VPX_CBR;
-    cfg_.g_lag_in_frames = 6;
-    cfg_.rc_buf_initial_sz = 500;
-    cfg_.rc_buf_optimal_sz = 500;
-    cfg_.rc_buf_sz = 1000;
-    cfg_.rc_target_bitrate = 300;
-    ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv",
-                                         352, 288, 30, 1, 0, 15);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  }
-
  int set_cpu_used_;
  int aq_mode_;
 };
@@ -61,42 +45,65 @@ class AqSegmentTest
 // Validate that this AQ segmentation mode (AQ=1, variance_ap)
 // encodes and decodes without a mismatch.
 TEST_P(AqSegmentTest, TestNoMisMatchAQ1) {
-  DoTest(1);
+  cfg_.rc_min_quantizer = 8;
+  cfg_.rc_max_quantizer = 56;
+  cfg_.rc_end_usage = VPX_CBR;
+  cfg_.g_lag_in_frames = 0;
+  cfg_.rc_buf_initial_sz = 500;
+  cfg_.rc_buf_optimal_sz = 500;
+  cfg_.rc_buf_sz = 1000;
+  cfg_.rc_target_bitrate = 300;
+
+  aq_mode_ = 1;
+
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                        30, 1, 0, 100);
+
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

 // Validate that this AQ segmentation mode (AQ=2, complexity_aq)
 // encodes and decodes without a mismatch.
 TEST_P(AqSegmentTest, TestNoMisMatchAQ2) {
-  DoTest(2);
+  cfg_.rc_min_quantizer = 8;
+  cfg_.rc_max_quantizer = 56;
+  cfg_.rc_end_usage = VPX_CBR;
+  cfg_.g_lag_in_frames = 0;
+  cfg_.rc_buf_initial_sz = 500;
+  cfg_.rc_buf_optimal_sz = 500;
+  cfg_.rc_buf_sz = 1000;
+  cfg_.rc_target_bitrate = 300;
+
+  aq_mode_ = 2;
+
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                        30, 1, 0, 100);
+
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

 // Validate that this AQ segmentation mode (AQ=3, cyclic_refresh_aq)
 // encodes and decodes without a mismatch.
 TEST_P(AqSegmentTest, TestNoMisMatchAQ3) {
-  DoTest(3);
+  cfg_.rc_min_quantizer = 8;
+  cfg_.rc_max_quantizer = 56;
+  cfg_.rc_end_usage = VPX_CBR;
+  cfg_.g_lag_in_frames = 0;
+  cfg_.rc_buf_initial_sz = 500;
+  cfg_.rc_buf_optimal_sz = 500;
+  cfg_.rc_buf_sz = 1000;
+  cfg_.rc_target_bitrate = 300;
+
+  aq_mode_ = 3;
+
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                        30, 1, 0, 100);
+
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

-class AqSegmentTestLarge : public AqSegmentTest {};
-
-TEST_P(AqSegmentTestLarge, TestNoMisMatchAQ1) {
-  DoTest(1);
-}
-
-TEST_P(AqSegmentTestLarge, TestNoMisMatchAQ2) {
-  DoTest(2);
-}
-
-TEST_P(AqSegmentTestLarge, TestNoMisMatchAQ3) {
-  DoTest(3);
-}
-
-
-VP10_INSTANTIATE_TEST_CASE(AqSegmentTest,
-                           ::testing::Values(::libvpx_test::kRealTime,
-                                             ::libvpx_test::kOnePassGood),
-                           ::testing::Range(5, 9));
-VP10_INSTANTIATE_TEST_CASE(AqSegmentTestLarge,
-                           ::testing::Values(::libvpx_test::kRealTime,
-                                             ::libvpx_test::kOnePassGood),
-                           ::testing::Range(3, 5));
+VP9_INSTANTIATE_TEST_CASE(AqSegmentTest,
+                          ::testing::Values(::libvpx_test::kRealTime,
+                                            ::libvpx_test::kOnePassGood),
+                          ::testing::Range(3, 9));
 }  // namespace
--- a/test/arf_freq_test.cc
+++ b/test/arf_freq_test.cc
@@ -1,249 +0,0 @@
-/*
- *  Copyright (c) 2015 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/util.h"
-#include "test/y4m_video_source.h"
-#include "test/yuv_video_source.h"
-#include "vp10/encoder/ratectrl.h"
-
-namespace {
-
-const unsigned int kFrames = 100;
-const int kBitrate = 500;
-
-#define ARF_NOT_SEEN               1000001
-#define ARF_SEEN_ONCE              1000000
-
-typedef struct {
-  const char *filename;
-  unsigned int width;
-  unsigned int height;
-  unsigned int framerate_num;
-  unsigned int framerate_den;
-  unsigned int input_bit_depth;
-  vpx_img_fmt fmt;
-  vpx_bit_depth_t bit_depth;
-  unsigned int profile;
-} TestVideoParam;
-
-typedef struct {
-  libvpx_test::TestMode mode;
-  int cpu_used;
-} TestEncodeParam;
-
-const TestVideoParam kTestVectors[] = {
-  // artificially increase framerate to trigger default check
-  {"hantro_collage_w352h288.yuv", 352, 288, 5000, 1,
-    8, VPX_IMG_FMT_I420, VPX_BITS_8, 0},
-  {"hantro_collage_w352h288.yuv", 352, 288, 30, 1,
-    8, VPX_IMG_FMT_I420, VPX_BITS_8, 0},
-  {"rush_hour_444.y4m", 352, 288, 30, 1,
-    8, VPX_IMG_FMT_I444, VPX_BITS_8, 1},
-#if CONFIG_VPX_HIGHBITDEPTH
-  // Add list of profile 2/3 test videos here ...
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-};
-
-const TestEncodeParam kEncodeVectors[] = {
-  {::libvpx_test::kOnePassGood, 2},
-  {::libvpx_test::kOnePassGood, 5},
-  {::libvpx_test::kTwoPassGood, 1},
-  {::libvpx_test::kTwoPassGood, 2},
-  {::libvpx_test::kTwoPassGood, 5},
-  {::libvpx_test::kRealTime, 5},
-};
-
-const int kMinArfVectors[] = {
-  // NOTE: 0 refers to the default built-in logic in:
-  //       vp9_rc_get_default_min_gf_interval(...)
-  0, 4, 8, 12, 15
-};
-
-int is_extension_y4m(const char *filename) {
-  const char *dot = strrchr(filename, '.');
-  if (!dot || dot == filename)
-    return 0;
-  else
-    return !strcmp(dot, ".y4m");
-}
-
-class ArfFreqTestLarge
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith3Params<TestVideoParam, \
-                                                 TestEncodeParam, int> {
- protected:
-  ArfFreqTestLarge()
-      : EncoderTest(GET_PARAM(0)),
-        test_video_param_(GET_PARAM(1)),
-        test_encode_param_(GET_PARAM(2)),
-        min_arf_requested_(GET_PARAM(3)) {
-  }
-
-  virtual ~ArfFreqTestLarge() {}
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(test_encode_param_.mode);
-    if (test_encode_param_.mode != ::libvpx_test::kRealTime) {
-      cfg_.g_lag_in_frames = 25;
-      cfg_.rc_end_usage = VPX_VBR;
-    } else {
-      cfg_.g_lag_in_frames = 0;
-      cfg_.rc_end_usage = VPX_CBR;
-      cfg_.rc_buf_sz = 1000;
-      cfg_.rc_buf_initial_sz = 500;
-      cfg_.rc_buf_optimal_sz = 600;
-    }
-    dec_cfg_.threads = 4;
-  }
-
-  virtual void BeginPassHook(unsigned int) {
-    min_run_ = ARF_NOT_SEEN;
-    run_of_visible_frames_ = 0;
-  }
-
-  int GetNumFramesInPkt(const vpx_codec_cx_pkt_t *pkt) {
-    const uint8_t *buffer = reinterpret_cast<uint8_t*>(pkt->data.frame.buf);
-    const uint8_t marker = buffer[pkt->data.frame.sz - 1];
-    const int mag = ((marker >> 3) & 3) + 1;
-    int frames = (marker & 0x7) + 1;
-    const unsigned int index_sz = 2 + mag  * frames;
-    // Check for superframe or not.
-    // Assume superframe has only one visible frame, the rest being
-    // invisible. If superframe index is not found, then there is only
-    // one frame.
-    if (!((marker & 0xe0) == 0xc0 &&
-          pkt->data.frame.sz >= index_sz &&
-          buffer[pkt->data.frame.sz - index_sz] == marker)) {
-      frames = 1;
-    }
-    return frames;
-  }
-
-  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
-    if (pkt->kind != VPX_CODEC_CX_FRAME_PKT)
-      return;
-    const int frames = GetNumFramesInPkt(pkt);
-    if (frames == 1) {
-      run_of_visible_frames_++;
-    } else if (frames == 2) {
-      if (min_run_ == ARF_NOT_SEEN) {
-        min_run_ = ARF_SEEN_ONCE;
-      } else if (min_run_ == ARF_SEEN_ONCE ||
-                 run_of_visible_frames_ < min_run_) {
-        min_run_ = run_of_visible_frames_;
-      }
-      run_of_visible_frames_ = 1;
-    } else {
-      min_run_ = 0;
-      run_of_visible_frames_ = 1;
-    }
-  }
-
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (video->frame() == 0) {
-      encoder->Control(VP9E_SET_FRAME_PARALLEL_DECODING, 1);
-      encoder->Control(VP9E_SET_TILE_COLUMNS, 4);
-      encoder->Control(VP8E_SET_CPUUSED, test_encode_param_.cpu_used);
-      encoder->Control(VP9E_SET_MIN_GF_INTERVAL, min_arf_requested_);
-      if (test_encode_param_.mode != ::libvpx_test::kRealTime) {
-        encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
-        encoder->Control(VP8E_SET_ARNR_MAXFRAMES, 7);
-        encoder->Control(VP8E_SET_ARNR_STRENGTH, 5);
-        encoder->Control(VP8E_SET_ARNR_TYPE, 3);
-      }
-    }
-  }
-
-  int GetMinVisibleRun() const {
-    return min_run_;
-  }
-
-  int GetMinArfDistanceRequested() const {
-    if (min_arf_requested_)
-      return min_arf_requested_;
-    else
-      return vp10_rc_get_default_min_gf_interval(
-          test_video_param_.width, test_video_param_.height,
-          (double)test_video_param_.framerate_num /
-          test_video_param_.framerate_den);
-  }
-
-  TestVideoParam test_video_param_;
-  TestEncodeParam test_encode_param_;
-
- private:
-  int min_arf_requested_;
-  int min_run_;
-  int run_of_visible_frames_;
-};
-
-TEST_P(ArfFreqTestLarge, MinArfFreqTest) {
-  cfg_.rc_target_bitrate = kBitrate;
-  cfg_.g_error_resilient = 0;
-  cfg_.g_profile = test_video_param_.profile;
-  cfg_.g_input_bit_depth = test_video_param_.input_bit_depth;
-  cfg_.g_bit_depth = test_video_param_.bit_depth;
-  init_flags_ = VPX_CODEC_USE_PSNR;
-  if (cfg_.g_bit_depth > 8)
-    init_flags_ |= VPX_CODEC_USE_HIGHBITDEPTH;
-
-  testing::internal::scoped_ptr<libvpx_test::VideoSource> video;
-  if (is_extension_y4m(test_video_param_.filename)) {
-    video.reset(new libvpx_test::Y4mVideoSource(test_video_param_.filename,
-                                                0, kFrames));
-  } else {
-    video.reset(new libvpx_test::YUVVideoSource(test_video_param_.filename,
-                                                test_video_param_.fmt,
-                                                test_video_param_.width,
-                                                test_video_param_.height,
-                                                test_video_param_.framerate_num,
-                                                test_video_param_.framerate_den,
-                                                0, kFrames));
-  }
-
-  ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
-  const int min_run = GetMinVisibleRun();
-  const int min_arf_dist_requested = GetMinArfDistanceRequested();
-  if (min_run != ARF_NOT_SEEN && min_run != ARF_SEEN_ONCE) {
-    const int min_arf_dist = min_run + 1;
-    EXPECT_GE(min_arf_dist, min_arf_dist_requested);
-  }
-}
-
-#if CONFIG_VPX_HIGHBITDEPTH || CONFIG_EXT_REFS
-#if CONFIG_VP10_ENCODER
-// TODO(angiebird): 25-29 fail in high bitdepth mode.
-// TODO(zoeliu): This ArfFreqTest does not work with BWDREF_FRAME, as
-// BWDREF_FRAME is also a non-show frame, and the minimum run between two
-// consecutive BWDREF_FRAME's may vary between 1 and any arbitrary positive
-// number as long as it does not exceed the gf_group interval.
-INSTANTIATE_TEST_CASE_P(
-    DISABLED_VP10, ArfFreqTestLarge,
-    ::testing::Combine(
-        ::testing::Values(static_cast<const libvpx_test::CodecFactory *>(
-            &libvpx_test::kVP10)),
-        ::testing::ValuesIn(kTestVectors),
-        ::testing::ValuesIn(kEncodeVectors),
-        ::testing::ValuesIn(kMinArfVectors)));
-#endif  // CONFIG_VP10_ENCODER
-#else
-VP10_INSTANTIATE_TEST_CASE(
-    ArfFreqTestLarge,
-    ::testing::ValuesIn(kTestVectors),
-    ::testing::ValuesIn(kEncodeVectors),
-    ::testing::ValuesIn(kMinArfVectors));
-#endif  // CONFIG_VPX_HIGHBITDEPTH || CONFIG_EXT_REFS
-}  // namespace
--- a/test/avg_test.cc
+++ b/test/avg_test.cc
@@ -1,411 +0,0 @@
-/*
- *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <limits.h>
-#include <stdio.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-
-#include "test/acm_random.h"
-#include "test/clear_system_state.h"
-#include "test/register_state_check.h"
-#include "test/util.h"
-#include "vpx_mem/vpx_mem.h"
-
-using libvpx_test::ACMRandom;
-
-namespace {
-class AverageTestBase : public ::testing::Test {
- public:
-  AverageTestBase(int width, int height) : width_(width), height_(height) {}
-
-  static void SetUpTestCase() {
-    source_data_ = reinterpret_cast<uint8_t*>(
-        vpx_memalign(kDataAlignment, kDataBlockSize));
-  }
-
-  static void TearDownTestCase() {
-    vpx_free(source_data_);
-    source_data_ = NULL;
-  }
-
-  virtual void TearDown() {
-    libvpx_test::ClearSystemState();
-  }
-
- protected:
-  // Handle blocks up to 4 blocks 64x64 with stride up to 128
-  static const int kDataAlignment = 16;
-  static const int kDataBlockSize = 64 * 128;
-
-  virtual void SetUp() {
-    source_stride_ = (width_ + 31) & ~31;
-    rnd_.Reset(ACMRandom::DeterministicSeed());
-  }
-
-  // Sum Pixels
-  unsigned int ReferenceAverage8x8(const uint8_t* source, int pitch) {
-    unsigned int average = 0;
-    for (int h = 0; h < 8; ++h)
-      for (int w = 0; w < 8; ++w)
-        average += source[h * pitch + w];
-    return ((average + 32) >> 6);
-  }
-
-  unsigned int ReferenceAverage4x4(const uint8_t* source, int pitch) {
-    unsigned int average = 0;
-    for (int h = 0; h < 4; ++h)
-      for (int w = 0; w < 4; ++w)
-        average += source[h * pitch + w];
-    return ((average + 8) >> 4);
-  }
-
-  void FillConstant(uint8_t fill_constant) {
-    for (int i = 0; i < width_ * height_; ++i) {
-        source_data_[i] = fill_constant;
-    }
-  }
-
-  void FillRandom() {
-    for (int i = 0; i < width_ * height_; ++i) {
-        source_data_[i] = rnd_.Rand8();
-    }
-  }
-
-  int width_, height_;
-  static uint8_t* source_data_;
-  int source_stride_;
-
-  ACMRandom rnd_;
-};
-typedef unsigned int (*AverageFunction)(const uint8_t* s, int pitch);
-
-typedef std::tr1::tuple<int, int, int, int, AverageFunction> AvgFunc;
-
-class AverageTest
-    : public AverageTestBase,
-      public ::testing::WithParamInterface<AvgFunc>{
- public:
-  AverageTest() : AverageTestBase(GET_PARAM(0), GET_PARAM(1)) {}
-
- protected:
-  void CheckAverages() {
-    unsigned int expected = 0;
-    if (GET_PARAM(3) == 8) {
-      expected = ReferenceAverage8x8(source_data_+ GET_PARAM(2),
-                                     source_stride_);
-    } else  if (GET_PARAM(3) == 4) {
-      expected = ReferenceAverage4x4(source_data_+ GET_PARAM(2),
-                                     source_stride_);
-    }
-
-    ASM_REGISTER_STATE_CHECK(GET_PARAM(4)(source_data_+ GET_PARAM(2),
-                                          source_stride_));
-    unsigned int actual = GET_PARAM(4)(source_data_+ GET_PARAM(2),
-                                       source_stride_);
-
-    EXPECT_EQ(expected, actual);
-  }
-};
-
-typedef void (*IntProRowFunc)(int16_t hbuf[16], uint8_t const *ref,
-                              const int ref_stride, const int height);
-
-typedef std::tr1::tuple<int, IntProRowFunc, IntProRowFunc> IntProRowParam;
-
-class IntProRowTest
-    : public AverageTestBase,
-      public ::testing::WithParamInterface<IntProRowParam> {
- public:
-  IntProRowTest()
-    : AverageTestBase(16, GET_PARAM(0)),
-      hbuf_asm_(NULL),
-      hbuf_c_(NULL) {
-    asm_func_ = GET_PARAM(1);
-    c_func_ = GET_PARAM(2);
-  }
-
- protected:
-  virtual void SetUp() {
-    hbuf_asm_ = reinterpret_cast<int16_t*>(
-        vpx_memalign(kDataAlignment, sizeof(*hbuf_asm_) * 16));
-    hbuf_c_ = reinterpret_cast<int16_t*>(
-        vpx_memalign(kDataAlignment, sizeof(*hbuf_c_) * 16));
-  }
-
-  virtual void TearDown() {
-    vpx_free(hbuf_c_);
-    hbuf_c_ = NULL;
-    vpx_free(hbuf_asm_);
-    hbuf_asm_ = NULL;
-  }
-
-  void RunComparison() {
-    ASM_REGISTER_STATE_CHECK(c_func_(hbuf_c_, source_data_, 0, height_));
-    ASM_REGISTER_STATE_CHECK(asm_func_(hbuf_asm_, source_data_, 0, height_));
-    EXPECT_EQ(0, memcmp(hbuf_c_, hbuf_asm_, sizeof(*hbuf_c_) * 16))
-        << "Output mismatch";
-  }
-
- private:
-  IntProRowFunc asm_func_;
-  IntProRowFunc c_func_;
-  int16_t *hbuf_asm_;
-  int16_t *hbuf_c_;
-};
-
-typedef int16_t (*IntProColFunc)(uint8_t const *ref, const int width);
-
-typedef std::tr1::tuple<int, IntProColFunc, IntProColFunc> IntProColParam;
-
-class IntProColTest
-    : public AverageTestBase,
-      public ::testing::WithParamInterface<IntProColParam> {
- public:
-  IntProColTest() : AverageTestBase(GET_PARAM(0), 1), sum_asm_(0), sum_c_(0) {
-    asm_func_ = GET_PARAM(1);
-    c_func_ = GET_PARAM(2);
-  }
-
- protected:
-  void RunComparison() {
-    ASM_REGISTER_STATE_CHECK(sum_c_ = c_func_(source_data_, width_));
-    ASM_REGISTER_STATE_CHECK(sum_asm_ = asm_func_(source_data_, width_));
-    EXPECT_EQ(sum_c_, sum_asm_) << "Output mismatch";
-  }
-
- private:
-  IntProColFunc asm_func_;
-  IntProColFunc c_func_;
-  int16_t sum_asm_;
-  int16_t sum_c_;
-};
-
-typedef int (*SatdFunc)(const int16_t *coeffs, int length);
-typedef std::tr1::tuple<int, SatdFunc> SatdTestParam;
-
-class SatdTest
-    : public ::testing::Test,
-      public ::testing::WithParamInterface<SatdTestParam> {
- protected:
-  virtual void SetUp() {
-    satd_size_ = GET_PARAM(0);
-    satd_func_ = GET_PARAM(1);
-    rnd_.Reset(ACMRandom::DeterministicSeed());
-    src_ = reinterpret_cast<int16_t*>(
-        vpx_memalign(16, sizeof(*src_) * satd_size_));
-    ASSERT_TRUE(src_ != NULL);
-  }
-
-  virtual void TearDown() {
-    libvpx_test::ClearSystemState();
-    vpx_free(src_);
-  }
-
-  void FillConstant(const int16_t val) {
-    for (int i = 0; i < satd_size_; ++i) src_[i] = val;
-  }
-
-  void FillRandom() {
-    for (int i = 0; i < satd_size_; ++i) src_[i] = rnd_.Rand16();
-  }
-
-  void Check(const int expected) {
-    int total;
-    ASM_REGISTER_STATE_CHECK(total = satd_func_(src_, satd_size_));
-    EXPECT_EQ(expected, total);
-  }
-
-  int satd_size_;
-
- private:
-  int16_t *src_;
-  SatdFunc satd_func_;
-  ACMRandom rnd_;
-};
-
-uint8_t* AverageTestBase::source_data_ = NULL;
-
-TEST_P(AverageTest, MinValue) {
-  FillConstant(0);
-  CheckAverages();
-}
-
-TEST_P(AverageTest, MaxValue) {
-  FillConstant(255);
-  CheckAverages();
-}
-
-TEST_P(AverageTest, Random) {
-  // The reference frame, but not the source frame, may be unaligned for
-  // certain types of searches.
-  for (int i = 0; i < 1000; i++) {
-    FillRandom();
-    CheckAverages();
-  }
-}
-
-TEST_P(IntProRowTest, MinValue) {
-  FillConstant(0);
-  RunComparison();
-}
-
-TEST_P(IntProRowTest, MaxValue) {
-  FillConstant(255);
-  RunComparison();
-}
-
-TEST_P(IntProRowTest, Random) {
-  FillRandom();
-  RunComparison();
-}
-
-TEST_P(IntProColTest, MinValue) {
-  FillConstant(0);
-  RunComparison();
-}
-
-TEST_P(IntProColTest, MaxValue) {
-  FillConstant(255);
-  RunComparison();
-}
-
-TEST_P(IntProColTest, Random) {
-  FillRandom();
-  RunComparison();
-}
-
-
-TEST_P(SatdTest, MinValue) {
-  const int kMin = -32640;
-  const int expected = -kMin * satd_size_;
-  FillConstant(kMin);
-  Check(expected);
-}
-
-TEST_P(SatdTest, MaxValue) {
-  const int kMax = 32640;
-  const int expected = kMax * satd_size_;
-  FillConstant(kMax);
-  Check(expected);
-}
-
-TEST_P(SatdTest, Random) {
-  int expected;
-  switch (satd_size_) {
-    case 16: expected = 205298; break;
-    case 64: expected = 1113950; break;
-    case 256: expected = 4268415; break;
-    case 1024: expected = 16954082; break;
-    default:
-      FAIL() << "Invalid satd size (" << satd_size_
-             << ") valid: 16/64/256/1024";
-  }
-  FillRandom();
-  Check(expected);
-}
-
-using std::tr1::make_tuple;
-
-INSTANTIATE_TEST_CASE_P(
-    C, AverageTest,
-    ::testing::Values(
-        make_tuple(16, 16, 1, 8, &vpx_avg_8x8_c),
-        make_tuple(16, 16, 1, 4, &vpx_avg_4x4_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    C, SatdTest,
-    ::testing::Values(
-        make_tuple(16, &vpx_satd_c),
-        make_tuple(64, &vpx_satd_c),
-        make_tuple(256, &vpx_satd_c),
-        make_tuple(1024, &vpx_satd_c)));
-
-#if HAVE_SSE2
-INSTANTIATE_TEST_CASE_P(
-    SSE2, AverageTest,
-    ::testing::Values(
-        make_tuple(16, 16, 0, 8, &vpx_avg_8x8_sse2),
-        make_tuple(16, 16, 5, 8, &vpx_avg_8x8_sse2),
-        make_tuple(32, 32, 15, 8, &vpx_avg_8x8_sse2),
-        make_tuple(16, 16, 0, 4, &vpx_avg_4x4_sse2),
-        make_tuple(16, 16, 5, 4, &vpx_avg_4x4_sse2),
-        make_tuple(32, 32, 15, 4, &vpx_avg_4x4_sse2)));
-
-INSTANTIATE_TEST_CASE_P(
-    SSE2, IntProRowTest, ::testing::Values(
-        make_tuple(16, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
-        make_tuple(32, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
-        make_tuple(64, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    SSE2, IntProColTest, ::testing::Values(
-        make_tuple(16, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
-        make_tuple(32, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
-        make_tuple(64, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    SSE2, SatdTest,
-    ::testing::Values(
-        make_tuple(16, &vpx_satd_sse2),
-        make_tuple(64, &vpx_satd_sse2),
-        make_tuple(256, &vpx_satd_sse2),
-        make_tuple(1024, &vpx_satd_sse2)));
-#endif
-
-#if HAVE_NEON
-INSTANTIATE_TEST_CASE_P(
-    NEON, AverageTest,
-    ::testing::Values(
-        make_tuple(16, 16, 0, 8, &vpx_avg_8x8_neon),
-        make_tuple(16, 16, 5, 8, &vpx_avg_8x8_neon),
-        make_tuple(32, 32, 15, 8, &vpx_avg_8x8_neon),
-        make_tuple(16, 16, 0, 4, &vpx_avg_4x4_neon),
-        make_tuple(16, 16, 5, 4, &vpx_avg_4x4_neon),
-        make_tuple(32, 32, 15, 4, &vpx_avg_4x4_neon)));
-
-INSTANTIATE_TEST_CASE_P(
-    NEON, IntProRowTest, ::testing::Values(
-        make_tuple(16, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
-        make_tuple(32, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
-        make_tuple(64, &vpx_int_pro_row_neon, &vpx_int_pro_row_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    NEON, IntProColTest, ::testing::Values(
-        make_tuple(16, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
-        make_tuple(32, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
-        make_tuple(64, &vpx_int_pro_col_neon, &vpx_int_pro_col_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    NEON, SatdTest,
-    ::testing::Values(
-        make_tuple(16, &vpx_satd_neon),
-        make_tuple(64, &vpx_satd_neon),
-        make_tuple(256, &vpx_satd_neon),
-        make_tuple(1024, &vpx_satd_neon)));
-#endif
-
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(
-    MSA, AverageTest,
-    ::testing::Values(
-        make_tuple(16, 16, 0, 8, &vpx_avg_8x8_msa),
-        make_tuple(16, 16, 5, 8, &vpx_avg_8x8_msa),
-        make_tuple(32, 32, 15, 8, &vpx_avg_8x8_msa),
-        make_tuple(16, 16, 0, 4, &vpx_avg_4x4_msa),
-        make_tuple(16, 16, 5, 4, &vpx_avg_4x4_msa),
-        make_tuple(32, 32, 15, 4, &vpx_avg_4x4_msa)));
-#endif
-
-}  // namespace
--- a/test/blend_a64_mask_1d_test.cc
+++ b/test/blend_a64_mask_1d_test.cc
@@ -1,367 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <math.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/register_state_check.h"
-
-#include "test/function_equivalence_test.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-#include "./vp10_rtcd.h"
-
-#include "vp10/common/enums.h"
-
-#include "vpx_dsp/blend.h"
-
-using libvpx_test::FunctionEquivalenceTest;
-
-namespace {
-
-template<typename F, typename T>
-class BlendA64Mask1DTest : public FunctionEquivalenceTest<F> {
- public:
-  static const int kIterations = 10000;
-  static const int kMaxWidth = MAX_SB_SIZE * 5;  // * 5 to cover longer strides
-  static const int kMaxHeight = MAX_SB_SIZE;
-  static const int kBufSize = kMaxWidth * kMaxHeight;
-  static const int kMaxMaskWidth = 2 * MAX_SB_SIZE;
-  static const int kMaxMaskSize = kMaxMaskWidth;
-
-  virtual ~BlendA64Mask1DTest() {}
-
-  virtual void Execute(const T *p_src0, const T *p_src1) = 0;
-
-  void Common() {
-    w_ = 1 << this->rng_(MAX_SB_SIZE_LOG2 + 1);
-    h_ = 1 << this->rng_(MAX_SB_SIZE_LOG2 + 1);
-
-    dst_offset_ = this->rng_(33);
-    dst_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    src0_offset_ = this->rng_(33);
-    src0_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    src1_offset_ = this->rng_(33);
-    src1_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    T *p_src0;
-    T *p_src1;
-
-    switch (this->rng_(3)) {
-      case 0:   // Separate sources
-        p_src0 = src0_;
-        p_src1 = src1_;
-        break;
-      case 1:   // src0 == dst
-        p_src0 = dst_tst_;
-        src0_stride_ = dst_stride_;
-        src0_offset_ = dst_offset_;
-        p_src1 = src1_;
-        break;
-      case 2:   // src1 == dst
-        p_src0 = src0_;
-        p_src1 = dst_tst_;
-        src1_stride_ = dst_stride_;
-        src1_offset_ = dst_offset_;
-        break;
-      default:
-        FAIL();
-    }
-
-    Execute(p_src0, p_src1);
-
-    for (int r = 0 ; r < h_ ; ++r) {
-      for (int c = 0 ; c < w_ ; ++c) {
-        ASSERT_EQ(dst_ref_[dst_offset_ + r * dst_stride_ + c],
-                  dst_tst_[dst_offset_ + r * dst_stride_ + c]);
-      }
-    }
-  }
-
-  T dst_ref_[kBufSize];
-  T dst_tst_[kBufSize];
-  size_t dst_stride_;
-  size_t dst_offset_;
-
-  T src0_[kBufSize];
-  size_t src0_stride_;
-  size_t src0_offset_;
-
-  T src1_[kBufSize];
-  size_t src1_stride_;
-  size_t src1_offset_;
-
-  uint8_t mask_[kMaxMaskSize];
-
-  int w_;
-  int h_;
-};
-
-//////////////////////////////////////////////////////////////////////////////
-// 8 bit version
-//////////////////////////////////////////////////////////////////////////////
-
-typedef void (*F8B)(uint8_t *dst, uint32_t dst_stride,
-                    const uint8_t *src0, uint32_t src0_stride,
-                    const uint8_t *src1, uint32_t src1_stride,
-                    const uint8_t *mask, int h, int w);
-typedef libvpx_test::FuncParam<F8B> TestFuncs;
-
-class BlendA64Mask1DTest8B : public BlendA64Mask1DTest<F8B, uint8_t> {
- protected:
-  void Execute(const uint8_t *p_src0, const uint8_t *p_src1) {
-    params_.ref_func(dst_ref_ + dst_offset_, dst_stride_,
-                     p_src0 + src0_offset_, src0_stride_,
-                     p_src1 + src1_offset_, src1_stride_, mask_, h_, w_);
-    ASM_REGISTER_STATE_CHECK(
-        params_.tst_func(dst_tst_ + dst_offset_, dst_stride_,
-                         p_src0 + src0_offset_, src0_stride_,
-                         p_src1 + src1_offset_, src1_stride_, mask_, h_, w_));
-  }
-};
-
-TEST_P(BlendA64Mask1DTest8B, RandomValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_.Rand8();
-      dst_tst_[i] = rng_.Rand8();
-
-      src0_[i] = rng_.Rand8();
-      src1_[i] = rng_.Rand8();
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(VPX_BLEND_A64_MAX_ALPHA + 1);
-
-    Common();
-  }
-}
-
-TEST_P(BlendA64Mask1DTest8B, ExtremeValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(2) + 254;
-      dst_tst_[i] = rng_(2) + 254;
-      src0_[i] = rng_(2) + 254;
-      src1_[i] = rng_(2) + 254;
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(2) + VPX_BLEND_A64_MAX_ALPHA - 1;
-
-    Common();
-  }
-}
-
-static void blend_a64_hmask_ref(
-    uint8_t *dst, uint32_t dst_stride,
-    const uint8_t *src0, uint32_t src0_stride,
-    const uint8_t *src1, uint32_t src1_stride,
-    const uint8_t *mask, int h, int w) {
-  uint8_t mask2d[BlendA64Mask1DTest8B::kMaxMaskSize]
-                [BlendA64Mask1DTest8B::kMaxMaskSize];
-
-  for (int row = 0 ; row < h ; ++row)
-    for (int col = 0 ; col < w ; ++col)
-      mask2d[row][col] = mask[col];
-
-  vpx_blend_a64_mask_c(dst, dst_stride,
-                       src0, src0_stride,
-                       src1, src1_stride,
-                       &mask2d[0][0], BlendA64Mask1DTest8B::kMaxMaskSize,
-                       h, w, 0, 0);
-}
-
-static void blend_a64_vmask_ref(
-    uint8_t *dst, uint32_t dst_stride,
-    const uint8_t *src0, uint32_t src0_stride,
-    const uint8_t *src1, uint32_t src1_stride,
-    const uint8_t *mask, int h, int w) {
-  uint8_t mask2d[BlendA64Mask1DTest8B::kMaxMaskSize]
-                [BlendA64Mask1DTest8B::kMaxMaskSize];
-
-  for (int row = 0 ; row < h ; ++row)
-    for (int col = 0 ; col < w ; ++col)
-      mask2d[row][col] = mask[row];
-
-  vpx_blend_a64_mask_c(dst, dst_stride,
-                       src0, src0_stride,
-                       src1, src1_stride,
-                       &mask2d[0][0], BlendA64Mask1DTest8B::kMaxMaskSize,
-                       h, w, 0, 0);
-}
-
-INSTANTIATE_TEST_CASE_P(
-  C, BlendA64Mask1DTest8B,
-  ::testing::Values(TestFuncs(blend_a64_hmask_ref, vpx_blend_a64_hmask_c),
-                    TestFuncs(blend_a64_vmask_ref, vpx_blend_a64_vmask_c)));
-
-#if HAVE_SSE4_1
-INSTANTIATE_TEST_CASE_P(
-  SSE4_1, BlendA64Mask1DTest8B,
-  ::testing::Values(
-      TestFuncs(blend_a64_hmask_ref, vpx_blend_a64_hmask_sse4_1),
-      TestFuncs(blend_a64_vmask_ref, vpx_blend_a64_vmask_sse4_1)));
-#endif  // HAVE_SSE4_1
-
-#if CONFIG_VPX_HIGHBITDEPTH
-//////////////////////////////////////////////////////////////////////////////
-// High bit-depth version
-//////////////////////////////////////////////////////////////////////////////
-
-typedef void (*FHBD)(uint8_t *dst, uint32_t dst_stride,
-                     const uint8_t *src0, uint32_t src0_stride,
-                     const uint8_t *src1, uint32_t src1_stride,
-                     const uint8_t *mask, int h, int w, int bd);
-typedef libvpx_test::FuncParam<FHBD> TestFuncsHBD;
-
-class BlendA64Mask1DTestHBD : public BlendA64Mask1DTest<FHBD, uint16_t> {
- protected:
-  void Execute(const uint16_t *p_src0, const uint16_t *p_src1) {
-    params_.ref_func(CONVERT_TO_BYTEPTR(dst_ref_ + dst_offset_), dst_stride_,
-                     CONVERT_TO_BYTEPTR(p_src0 + src0_offset_), src0_stride_,
-                     CONVERT_TO_BYTEPTR(p_src1 + src1_offset_), src1_stride_,
-                     mask_, h_, w_, bit_depth_);
-    ASM_REGISTER_STATE_CHECK(params_.tst_func(
-        CONVERT_TO_BYTEPTR(dst_tst_ + dst_offset_), dst_stride_,
-        CONVERT_TO_BYTEPTR(p_src0 + src0_offset_), src0_stride_,
-        CONVERT_TO_BYTEPTR(p_src1 + src1_offset_), src1_stride_,
-        mask_, h_, w_, bit_depth_));
-  }
-
-  int bit_depth_;
-};
-
-TEST_P(BlendA64Mask1DTestHBD, RandomValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    switch (rng_(3)) {
-    case 0:
-      bit_depth_ = 8;
-      break;
-    case 1:
-      bit_depth_ = 10;
-      break;
-    default:
-      bit_depth_ = 12;
-      break;
-    }
-
-    const int hi = 1 << bit_depth_;
-
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(hi);
-      dst_tst_[i] = rng_(hi);
-      src0_[i] = rng_(hi);
-      src1_[i] = rng_(hi);
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(VPX_BLEND_A64_MAX_ALPHA + 1);
-
-    Common();
-  }
-}
-
-TEST_P(BlendA64Mask1DTestHBD, ExtremeValues) {
-  for (int iter = 0 ; iter < 1000 && !HasFatalFailure(); ++iter) {
-    switch (rng_(3)) {
-    case 0:
-      bit_depth_ = 8;
-      break;
-    case 1:
-      bit_depth_ = 10;
-      break;
-    default:
-      bit_depth_ = 12;
-      break;
-    }
-
-    const int hi = 1 << bit_depth_;
-    const int lo = hi - 2;
-
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(hi - lo) + lo;
-      dst_tst_[i] = rng_(hi - lo) + lo;
-      src0_[i] = rng_(hi - lo) + lo;
-      src1_[i] = rng_(hi - lo) + lo;
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(2) + VPX_BLEND_A64_MAX_ALPHA - 1;
-
-    Common();
-  }
-}
-
-static void highbd_blend_a64_hmask_ref(
-    uint8_t *dst, uint32_t dst_stride,
-    const uint8_t *src0, uint32_t src0_stride,
-    const uint8_t *src1, uint32_t src1_stride,
-    const uint8_t *mask, int h, int w, int bd) {
-  uint8_t mask2d[BlendA64Mask1DTestHBD::kMaxMaskSize]
-                [BlendA64Mask1DTestHBD::kMaxMaskSize];
-
-  for (int row = 0 ; row < h ; ++row)
-    for (int col = 0 ; col < w ; ++col)
-      mask2d[row][col] = mask[col];
-
-  vpx_highbd_blend_a64_mask_c(dst, dst_stride,
-                              src0, src0_stride,
-                              src1, src1_stride,
-                              &mask2d[0][0],
-                              BlendA64Mask1DTestHBD::kMaxMaskSize,
-                              h, w, 0, 0, bd);
-}
-
-static void highbd_blend_a64_vmask_ref(
-    uint8_t *dst, uint32_t dst_stride,
-    const uint8_t *src0, uint32_t src0_stride,
-    const uint8_t *src1, uint32_t src1_stride,
-    const uint8_t *mask, int h, int w, int bd) {
-  uint8_t mask2d[BlendA64Mask1DTestHBD::kMaxMaskSize]
-                [BlendA64Mask1DTestHBD::kMaxMaskSize];
-
-  for (int row = 0 ; row < h ; ++row)
-    for (int col = 0 ; col < w ; ++col)
-      mask2d[row][col] = mask[row];
-
-  vpx_highbd_blend_a64_mask_c(dst, dst_stride,
-                              src0, src0_stride,
-                              src1, src1_stride,
-                              &mask2d[0][0],
-                              BlendA64Mask1DTestHBD::kMaxMaskSize,
-                              h, w, 0, 0, bd);
-}
-
-INSTANTIATE_TEST_CASE_P(
-  C, BlendA64Mask1DTestHBD,
-  ::testing::Values(
-      TestFuncsHBD(highbd_blend_a64_hmask_ref, vpx_highbd_blend_a64_hmask_c),
-      TestFuncsHBD(highbd_blend_a64_vmask_ref, vpx_highbd_blend_a64_vmask_c)));
-
-#if HAVE_SSE4_1
-INSTANTIATE_TEST_CASE_P(
-  SSE4_1, BlendA64Mask1DTestHBD,
-  ::testing::Values(
-      TestFuncsHBD(highbd_blend_a64_hmask_ref,
-                   vpx_highbd_blend_a64_hmask_sse4_1),
-      TestFuncsHBD(highbd_blend_a64_vmask_ref,
-                   vpx_highbd_blend_a64_vmask_sse4_1)));
-#endif  // HAVE_SSE4_1
-
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-}  // namespace
--- a/test/blend_a64_mask_test.cc
+++ b/test/blend_a64_mask_test.cc
@@ -1,288 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <math.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/register_state_check.h"
-
-#include "test/function_equivalence_test.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-#include "./vp10_rtcd.h"
-
-#include "vp10/common/enums.h"
-
-#include "vpx_dsp/blend.h"
-
-using libvpx_test::FunctionEquivalenceTest;
-
-namespace {
-
-template<typename F, typename T>
-class BlendA64MaskTest : public FunctionEquivalenceTest<F> {
- protected:
-  static const int kIterations = 10000;
-  static const int kMaxWidth = MAX_SB_SIZE * 5;  // * 5 to cover longer strides
-  static const int kMaxHeight = MAX_SB_SIZE;
-  static const int kBufSize = kMaxWidth * kMaxHeight;
-  static const int kMaxMaskWidth = 2 * MAX_SB_SIZE;
-  static const int kMaxMaskSize = kMaxMaskWidth * kMaxMaskWidth;
-
-  virtual ~BlendA64MaskTest() {}
-
-  virtual void Execute(const T *p_src0, const T *p_src1) = 0;
-
-  void Common() {
-    w_ = 1 << this->rng_(MAX_SB_SIZE_LOG2 + 1);
-    h_ = 1 << this->rng_(MAX_SB_SIZE_LOG2 + 1);
-
-    subx_ = this->rng_(2);
-    suby_ = this->rng_(2);
-
-    dst_offset_ = this->rng_(33);
-    dst_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    src0_offset_ = this->rng_(33);
-    src0_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    src1_offset_ = this->rng_(33);
-    src1_stride_ = this->rng_(kMaxWidth + 1 - w_) + w_;
-
-    mask_stride_ = this->rng_(kMaxWidth + 1 - w_ * (subx_ ? 2 : 1)) +
-                   w_ * (subx_ ? 2 : 1);
-
-    T *p_src0;
-    T *p_src1;
-
-    switch (this->rng_(3)) {
-      case 0:   // Separate sources
-        p_src0 = src0_;
-        p_src1 = src1_;
-        break;
-      case 1:   // src0 == dst
-        p_src0 = dst_tst_;
-        src0_stride_ = dst_stride_;
-        src0_offset_ = dst_offset_;
-        p_src1 = src1_;
-        break;
-      case 2:   // src1 == dst
-        p_src0 = src0_;
-        p_src1 = dst_tst_;
-        src1_stride_ = dst_stride_;
-        src1_offset_ = dst_offset_;
-        break;
-      default:
-        FAIL();
-    }
-
-    Execute(p_src0, p_src1);
-
-    for (int r = 0 ; r < h_ ; ++r) {
-      for (int c = 0 ; c < w_ ; ++c) {
-        ASSERT_EQ(dst_ref_[dst_offset_ + r * dst_stride_ + c],
-                  dst_tst_[dst_offset_ + r * dst_stride_ + c]);
-      }
-    }
-  }
-
-  T dst_ref_[kBufSize];
-  T dst_tst_[kBufSize];
-  size_t dst_stride_;
-  size_t dst_offset_;
-
-  T src0_[kBufSize];
-  size_t src0_stride_;
-  size_t src0_offset_;
-
-  T src1_[kBufSize];
-  size_t src1_stride_;
-  size_t src1_offset_;
-
-  uint8_t mask_[kMaxMaskSize];
-  size_t mask_stride_;
-
-  int w_;
-  int h_;
-
-  bool suby_;
-  bool subx_;
-};
-
-//////////////////////////////////////////////////////////////////////////////
-// 8 bit version
-//////////////////////////////////////////////////////////////////////////////
-
-typedef void (*F8B)(uint8_t *dst, uint32_t dst_stride,
-                    const uint8_t *src0, uint32_t src0_stride,
-                    const uint8_t *src1, uint32_t src1_stride,
-                    const uint8_t *mask, uint32_t mask_stride,
-                    int h, int w, int suby, int subx);
-typedef libvpx_test::FuncParam<F8B> TestFuncs;
-
-class BlendA64MaskTest8B : public BlendA64MaskTest<F8B, uint8_t> {
- protected:
-  void Execute(const uint8_t *p_src0, const uint8_t *p_src1) {
-    params_.ref_func(dst_ref_ + dst_offset_, dst_stride_,
-                     p_src0 + src0_offset_, src0_stride_,
-                     p_src1 + src1_offset_, src1_stride_,
-                     mask_, kMaxMaskWidth, h_, w_, suby_, subx_);
-    ASM_REGISTER_STATE_CHECK(
-        params_.tst_func(dst_tst_ + dst_offset_, dst_stride_,
-                         p_src0 + src0_offset_, src0_stride_,
-                         p_src1 + src1_offset_, src1_stride_,
-                         mask_, kMaxMaskWidth, h_, w_, suby_, subx_));
-  }
-};
-
-TEST_P(BlendA64MaskTest8B, RandomValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_.Rand8();
-      dst_tst_[i] = rng_.Rand8();
-
-      src0_[i] = rng_.Rand8();
-      src1_[i] = rng_.Rand8();
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(VPX_BLEND_A64_MAX_ALPHA + 1);
-
-    Common();
-  }
-}
-
-TEST_P(BlendA64MaskTest8B, ExtremeValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(2) + 254;
-      dst_tst_[i] = rng_(2) + 254;
-      src0_[i] = rng_(2) + 254;
-      src1_[i] = rng_(2) + 254;
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(2) + VPX_BLEND_A64_MAX_ALPHA - 1;
-
-    Common();
-  }
-}
-
-#if HAVE_SSE4_1
-INSTANTIATE_TEST_CASE_P(
-  SSE4_1_C_COMPARE, BlendA64MaskTest8B,
-  ::testing::Values(
-      TestFuncs(vpx_blend_a64_mask_c, vpx_blend_a64_mask_sse4_1)));
-#endif  // HAVE_SSE4_1
-
-#if CONFIG_VPX_HIGHBITDEPTH
-//////////////////////////////////////////////////////////////////////////////
-// High bit-depth version
-//////////////////////////////////////////////////////////////////////////////
-
-typedef void (*FHBD)(uint8_t *dst, uint32_t dst_stride,
-                     const uint8_t *src0, uint32_t src0_stride,
-                     const uint8_t *src1, uint32_t src1_stride,
-                     const uint8_t *mask, uint32_t mask_stride,
-                     int h, int w, int suby, int subx, int bd);
-typedef libvpx_test::FuncParam<FHBD> TestFuncsHBD;
-
-class BlendA64MaskTestHBD : public BlendA64MaskTest<FHBD, uint16_t> {
- protected:
-  void Execute(const uint16_t *p_src0, const uint16_t *p_src1) {
-    params_.ref_func(CONVERT_TO_BYTEPTR(dst_ref_ + dst_offset_), dst_stride_,
-                     CONVERT_TO_BYTEPTR(p_src0 + src0_offset_), src0_stride_,
-                     CONVERT_TO_BYTEPTR(p_src1 + src1_offset_), src1_stride_,
-                     mask_, kMaxMaskWidth, h_, w_, suby_, subx_, bit_depth_);
-    ASM_REGISTER_STATE_CHECK(params_.tst_func(
-        CONVERT_TO_BYTEPTR(dst_tst_ + dst_offset_), dst_stride_,
-        CONVERT_TO_BYTEPTR(p_src0 + src0_offset_), src0_stride_,
-        CONVERT_TO_BYTEPTR(p_src1 + src1_offset_), src1_stride_,
-        mask_, kMaxMaskWidth, h_, w_, suby_, subx_, bit_depth_));
-  }
-
-  int bit_depth_;
-};
-
-TEST_P(BlendA64MaskTestHBD, RandomValues) {
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure(); ++iter) {
-    switch (rng_(3)) {
-    case 0:
-      bit_depth_ = 8;
-      break;
-    case 1:
-      bit_depth_ = 10;
-      break;
-    default:
-      bit_depth_ = 12;
-      break;
-    }
-
-    const int hi = 1 << bit_depth_;
-
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(hi);
-      dst_tst_[i] = rng_(hi);
-      src0_[i] = rng_(hi);
-      src1_[i] = rng_(hi);
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(VPX_BLEND_A64_MAX_ALPHA + 1);
-
-    Common();
-  }
-}
-
-TEST_P(BlendA64MaskTestHBD, ExtremeValues) {
-  for (int iter = 0 ; iter < 1000 && !HasFatalFailure(); ++iter) {
-    switch (rng_(3)) {
-    case 0:
-      bit_depth_ = 8;
-      break;
-    case 1:
-      bit_depth_ = 10;
-      break;
-    default:
-      bit_depth_ = 12;
-      break;
-    }
-
-    const int hi = 1 << bit_depth_;
-    const int lo = hi - 2;
-
-    for (int i = 0 ; i < kBufSize ; ++i) {
-      dst_ref_[i] = rng_(hi - lo) + lo;
-      dst_tst_[i] = rng_(hi - lo) + lo;
-      src0_[i] = rng_(hi - lo) + lo;
-      src1_[i] = rng_(hi - lo) + lo;
-    }
-
-    for (int i = 0 ; i < kMaxMaskSize ; ++i)
-      mask_[i] = rng_(2) + VPX_BLEND_A64_MAX_ALPHA - 1;
-
-    Common();
-  }
-}
-
-#if HAVE_SSE4_1
-INSTANTIATE_TEST_CASE_P(
-  SSE4_1_C_COMPARE, BlendA64MaskTestHBD,
-  ::testing::Values(
-      TestFuncsHBD(vpx_highbd_blend_a64_mask_c,
-                   vpx_highbd_blend_a64_mask_sse4_1)));
-#endif  // HAVE_SSE4_1
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-}  // namespace
--- a/test/borders_test.cc
+++ b/test/borders_test.cc
@@ -52,7 +52,7 @@ TEST_P(BordersTest, TestEncodeHighBitrate) {
  // extend into the border and test the border condition.
  cfg_.g_lag_in_frames = 25;
  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_maxsection_pct = 2000;
+  cfg_.rc_2pass_vbr_minsection_pct = 2000;
  cfg_.rc_target_bitrate = 2000;
  cfg_.rc_max_quantizer = 10;

@@ -78,6 +78,6 @@ TEST_P(BordersTest, TestLowBitrate) {
  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

-VP10_INSTANTIATE_TEST_CASE(BordersTest, ::testing::Values(
+VP9_INSTANTIATE_TEST_CASE(BordersTest, ::testing::Values(
    ::libvpx_test::kTwoPassGood));
 }  // namespace
--- a/test/byte_alignment_test.cc
+++ b/test/byte_alignment_test.cc
@@ -0,0 +1,189 @@
+/*
+ *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include <string>
+
+#include "./vpx_config.h"
+#include "test/codec_factory.h"
+#include "test/decode_test_driver.h"
+#include "test/md5_helper.h"
+#include "test/util.h"
+#if CONFIG_WEBM_IO
+#include "test/webm_video_source.h"
+#endif
+
+namespace {
+
+const int kLegacyByteAlignment = 0;
+const int kLegacyYPlaneByteAlignment = 32;
+const int kNumPlanesToCheck = 3;
+const char kVP9TestFile[] = "vp90-2-02-size-lf-1920x1080.webm";
+const char kVP9Md5File[] = "vp90-2-02-size-lf-1920x1080.webm.md5";
+
+#if CONFIG_WEBM_IO
+
+struct ByteAlignmentTestParam {
+  int byte_alignment;
+  vpx_codec_err_t expected_value;
+  bool decode_remaining;
+};
+
+const ByteAlignmentTestParam kBaTestParams[] = {
+  {kLegacyByteAlignment, VPX_CODEC_OK, true},
+  {32, VPX_CODEC_OK, true},
+  {64, VPX_CODEC_OK, true},
+  {128, VPX_CODEC_OK, true},
+  {256, VPX_CODEC_OK, true},
+  {512, VPX_CODEC_OK, true},
+  {1024, VPX_CODEC_OK, true},
+  {1, VPX_CODEC_INVALID_PARAM, false},
+  {-2, VPX_CODEC_INVALID_PARAM, false},
+  {4, VPX_CODEC_INVALID_PARAM, false},
+  {16, VPX_CODEC_INVALID_PARAM, false},
+  {255, VPX_CODEC_INVALID_PARAM, false},
+  {2048, VPX_CODEC_INVALID_PARAM, false},
+};
+
+// Class for testing byte alignment of reference buffers.
+class ByteAlignmentTest
+    : public ::testing::TestWithParam<ByteAlignmentTestParam> {
+ protected:
+  ByteAlignmentTest()
+      : video_(NULL),
+        decoder_(NULL),
+        md5_file_(NULL) {}
+
+  virtual void SetUp() {
+    video_ = new libvpx_test::WebMVideoSource(kVP9TestFile);
+    ASSERT_TRUE(video_ != NULL);
+    video_->Init();
+    video_->Begin();
+
+    const vpx_codec_dec_cfg_t cfg = vpx_codec_dec_cfg_t();
+    decoder_ = new libvpx_test::VP9Decoder(cfg, 0);
+    ASSERT_TRUE(decoder_ != NULL);
+
+    OpenMd5File(kVP9Md5File);
+  }
+
+  virtual void TearDown() {
+    if (md5_file_ != NULL)
+      fclose(md5_file_);
+
+    delete decoder_;
+    delete video_;
+  }
+
+  void SetByteAlignment(int byte_alignment, vpx_codec_err_t expected_value) {
+    decoder_->Control(VP9_SET_BYTE_ALIGNMENT, byte_alignment, expected_value);
+  }
+
+  vpx_codec_err_t DecodeOneFrame(int byte_alignment_to_check) {
+    const vpx_codec_err_t res =
+        decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
+    CheckDecodedFrames(byte_alignment_to_check);
+    if (res == VPX_CODEC_OK)
+      video_->Next();
+    return res;
+  }
+
+  vpx_codec_err_t DecodeRemainingFrames(int byte_alignment_to_check) {
+    for (; video_->cxdata() != NULL; video_->Next()) {
+      const vpx_codec_err_t res =
+          decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
+      if (res != VPX_CODEC_OK)
+        return res;
+      CheckDecodedFrames(byte_alignment_to_check);
+    }
+    return VPX_CODEC_OK;
+  }
+
+ private:
+  // Check if |data| is aligned to |byte_alignment_to_check|.
+  // |byte_alignment_to_check| must be a power of 2.
+  void CheckByteAlignment(const uint8_t *data, int byte_alignment_to_check) {
+    ASSERT_EQ(0u, reinterpret_cast<size_t>(data) % byte_alignment_to_check);
+  }
+
+  // Iterate through the planes of the decoded frames and check for
+  // alignment based off |byte_alignment_to_check|.
+  void CheckDecodedFrames(int byte_alignment_to_check) {
+    libvpx_test::DxDataIterator dec_iter = decoder_->GetDxData();
+    const vpx_image_t *img;
+
+    // Get decompressed data
+    while ((img = dec_iter.Next()) != NULL) {
+      if (byte_alignment_to_check == kLegacyByteAlignment) {
+        CheckByteAlignment(img->planes[0], kLegacyYPlaneByteAlignment);
+      } else {
+        for (int i = 0; i < kNumPlanesToCheck; ++i) {
+          CheckByteAlignment(img->planes[i], byte_alignment_to_check);
+        }
+      }
+      CheckMd5(*img);
+    }
+  }
+
+  // TODO(fgalligan): Move the MD5 testing code into another class.
+  void OpenMd5File(const std::string &md5_file_name_) {
+    md5_file_ = libvpx_test::OpenTestDataFile(md5_file_name_);
+    ASSERT_TRUE(md5_file_ != NULL) << "MD5 file open failed. Filename: "
+        << md5_file_name_;
+  }
+
+  void CheckMd5(const vpx_image_t &img) {
+    ASSERT_TRUE(md5_file_ != NULL);
+    char expected_md5[33];
+    char junk[128];
+
+    // Read correct md5 checksums.
+    const int res = fscanf(md5_file_, "%s  %s", expected_md5, junk);
+    ASSERT_NE(EOF, res) << "Read md5 data failed";
+    expected_md5[32] = '\0';
+
+    ::libvpx_test::MD5 md5_res;
+    md5_res.Add(&img);
+    const char *const actual_md5 = md5_res.Get();
+
+    // Check md5 match.
+    ASSERT_STREQ(expected_md5, actual_md5) << "MD5 checksums don't match";
+  }
+
+  libvpx_test::WebMVideoSource *video_;
+  libvpx_test::VP9Decoder *decoder_;
+  FILE *md5_file_;
+};
+
+TEST_F(ByteAlignmentTest, SwitchByteAlignment) {
+  const int num_elements = 14;
+  const int byte_alignments[] = { 0, 32, 64, 128, 256, 512, 1024,
+                                  0, 1024, 32, 512, 64, 256, 128 };
+
+  for (int i = 0; i < num_elements; ++i) {
+    SetByteAlignment(byte_alignments[i], VPX_CODEC_OK);
+    ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame(byte_alignments[i]));
+  }
+  SetByteAlignment(byte_alignments[0], VPX_CODEC_OK);
+  ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames(byte_alignments[0]));
+}
+
+TEST_P(ByteAlignmentTest, TestAlignment) {
+  const ByteAlignmentTestParam t = GetParam();
+  SetByteAlignment(t.byte_alignment, t.expected_value);
+  if (t.decode_remaining)
+    ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames(t.byte_alignment));
+}
+
+INSTANTIATE_TEST_CASE_P(Alignments, ByteAlignmentTest,
+                        ::testing::ValuesIn(kBaTestParams));
+
+#endif  // CONFIG_WEBM_IO
+
+}  // namespace
--- a/test/codec_factory.h
+++ b/test/codec_factory.h
@@ -13,10 +13,10 @@
 #include "./vpx_config.h"
 #include "vpx/vpx_decoder.h"
 #include "vpx/vpx_encoder.h"
-#if CONFIG_VP10_ENCODER
+#if CONFIG_VP8_ENCODER || CONFIG_VP9_ENCODER
 #include "vpx/vp8cx.h"
 #endif
-#if CONFIG_VP10_DECODER
+#if CONFIG_VP8_DECODER || CONFIG_VP9_DECODER
 #include "vpx/vp8dx.h"
 #endif

@@ -69,47 +69,47 @@ class CodecTestWith3Params : public ::testing::TestWithParam<
 };

 /*
- * VP10 Codec Definitions
+ * VP8 Codec Definitions
 */
-#if CONFIG_VP10
-class VP10Decoder : public Decoder {
+#if CONFIG_VP8
+class VP8Decoder : public Decoder {
 public:
-  VP10Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
+  VP8Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
      : Decoder(cfg, deadline) {}

-  VP10Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
-              unsigned long deadline)  // NOLINT
+  VP8Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
+             unsigned long deadline)  // NOLINT
      : Decoder(cfg, flag, deadline) {}

 protected:
  virtual vpx_codec_iface_t* CodecInterface() const {
-#if CONFIG_VP10_DECODER
-    return &vpx_codec_vp10_dx_algo;
+#if CONFIG_VP8_DECODER
+    return &vpx_codec_vp8_dx_algo;
 #else
    return NULL;
 #endif
  }
 };

-class VP10Encoder : public Encoder {
+class VP8Encoder : public Encoder {
 public:
-  VP10Encoder(vpx_codec_enc_cfg_t cfg, unsigned long deadline,
-              const unsigned long init_flags, TwopassStatsStore *stats)
+  VP8Encoder(vpx_codec_enc_cfg_t cfg, unsigned long deadline,
+             const unsigned long init_flags, TwopassStatsStore *stats)
      : Encoder(cfg, deadline, init_flags, stats) {}

 protected:
  virtual vpx_codec_iface_t* CodecInterface() const {
-#if CONFIG_VP10_ENCODER
-    return &vpx_codec_vp10_cx_algo;
+#if CONFIG_VP8_ENCODER
+    return &vpx_codec_vp8_cx_algo;
 #else
    return NULL;
 #endif
  }
 };

-class VP10CodecFactory : public CodecFactory {
+class VP8CodecFactory : public CodecFactory {
 public:
-  VP10CodecFactory() : CodecFactory() {}
+  VP8CodecFactory() : CodecFactory() {}

  virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
                                 unsigned long deadline) const {
@@ -119,8 +119,8 @@ class VP10CodecFactory : public CodecFactory {
  virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
                                 const vpx_codec_flags_t flags,
                                 unsigned long deadline) const {  // NOLINT
-#if CONFIG_VP10_DECODER
-    return new VP10Decoder(cfg, flags, deadline);
+#if CONFIG_VP8_DECODER
+    return new VP8Decoder(cfg, flags, deadline);
 #else
    return NULL;
 #endif
@@ -130,8 +130,8 @@ class VP10CodecFactory : public CodecFactory {
                                 unsigned long deadline,
                                 const unsigned long init_flags,
                                 TwopassStatsStore *stats) const {
-#if CONFIG_VP10_ENCODER
-    return new VP10Encoder(cfg, deadline, init_flags, stats);
+#if CONFIG_VP8_ENCODER
+    return new VP8Encoder(cfg, deadline, init_flags, stats);
 #else
    return NULL;
 #endif
@@ -139,25 +139,119 @@ class VP10CodecFactory : public CodecFactory {

  virtual vpx_codec_err_t DefaultEncoderConfig(vpx_codec_enc_cfg_t *cfg,
                                               int usage) const {
-#if CONFIG_VP10_ENCODER
-    return vpx_codec_enc_config_default(&vpx_codec_vp10_cx_algo, cfg, usage);
+#if CONFIG_VP8_ENCODER
+    return vpx_codec_enc_config_default(&vpx_codec_vp8_cx_algo, cfg, usage);
 #else
    return VPX_CODEC_INCAPABLE;
 #endif
  }
 };

-const libvpx_test::VP10CodecFactory kVP10;
+const libvpx_test::VP8CodecFactory kVP8;

-#define VP10_INSTANTIATE_TEST_CASE(test, ...)\
-  INSTANTIATE_TEST_CASE_P(VP10, test, \
+#define VP8_INSTANTIATE_TEST_CASE(test, ...)\
+  INSTANTIATE_TEST_CASE_P(VP8, test, \
      ::testing::Combine( \
          ::testing::Values(static_cast<const libvpx_test::CodecFactory*>( \
-               &libvpx_test::kVP10)), \
+              &libvpx_test::kVP8)), \
          __VA_ARGS__))
 #else
-#define VP10_INSTANTIATE_TEST_CASE(test, ...)
-#endif  // CONFIG_VP10
+#define VP8_INSTANTIATE_TEST_CASE(test, ...)
+#endif  // CONFIG_VP8
+
+
+/*
+ * VP9 Codec Definitions
+ */
+#if CONFIG_VP9
+class VP9Decoder : public Decoder {
+ public:
+  VP9Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
+      : Decoder(cfg, deadline) {}
+
+  VP9Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
+             unsigned long deadline)  // NOLINT
+      : Decoder(cfg, flag, deadline) {}
+
+ protected:
+  virtual vpx_codec_iface_t* CodecInterface() const {
+#if CONFIG_VP9_DECODER
+    return &vpx_codec_vp9_dx_algo;
+#else
+    return NULL;
+#endif
+  }
+};
+
+class VP9Encoder : public Encoder {
+ public:
+  VP9Encoder(vpx_codec_enc_cfg_t cfg, unsigned long deadline,
+             const unsigned long init_flags, TwopassStatsStore *stats)
+      : Encoder(cfg, deadline, init_flags, stats) {}
+
+ protected:
+  virtual vpx_codec_iface_t* CodecInterface() const {
+#if CONFIG_VP9_ENCODER
+    return &vpx_codec_vp9_cx_algo;
+#else
+    return NULL;
+#endif
+  }
+};
+
+class VP9CodecFactory : public CodecFactory {
+ public:
+  VP9CodecFactory() : CodecFactory() {}
+
+  virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
+                                 unsigned long deadline) const {
+    return CreateDecoder(cfg, 0, deadline);
+  }
+
+  virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
+                                 const vpx_codec_flags_t flags,
+                                 unsigned long deadline) const {  // NOLINT
+#if CONFIG_VP9_DECODER
+    return new VP9Decoder(cfg, flags, deadline);
+#else
+    return NULL;
+#endif
+  }
+
+  virtual Encoder* CreateEncoder(vpx_codec_enc_cfg_t cfg,
+                                 unsigned long deadline,
+                                 const unsigned long init_flags,
+                                 TwopassStatsStore *stats) const {
+#if CONFIG_VP9_ENCODER
+    return new VP9Encoder(cfg, deadline, init_flags, stats);
+#else
+    return NULL;
+#endif
+  }
+
+  virtual vpx_codec_err_t DefaultEncoderConfig(vpx_codec_enc_cfg_t *cfg,
+                                               int usage) const {
+#if CONFIG_VP9_ENCODER
+    return vpx_codec_enc_config_default(&vpx_codec_vp9_cx_algo, cfg, usage);
+#else
+    return VPX_CODEC_INCAPABLE;
+#endif
+  }
+};
+
+const libvpx_test::VP9CodecFactory kVP9;
+
+#define VP9_INSTANTIATE_TEST_CASE(test, ...)\
+  INSTANTIATE_TEST_CASE_P(VP9, test, \
+      ::testing::Combine( \
+          ::testing::Values(static_cast<const libvpx_test::CodecFactory*>( \
+               &libvpx_test::kVP9)), \
+          __VA_ARGS__))
+#else
+#define VP9_INSTANTIATE_TEST_CASE(test, ...)
+#endif  // CONFIG_VP9
+

 }  // namespace libvpx_test
+
 #endif  // TEST_CODEC_FACTORY_H_
--- a/test/config_test.cc
+++ b/test/config_test.cc
@@ -0,0 +1,60 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "test/codec_factory.h"
+#include "test/encode_test_driver.h"
+#include "test/util.h"
+#include "test/video_source.h"
+
+namespace {
+
+class ConfigTest : public ::libvpx_test::EncoderTest,
+    public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
+ protected:
+  ConfigTest() : EncoderTest(GET_PARAM(0)),
+                 frame_count_in_(0), frame_count_out_(0), frame_count_max_(0) {}
+  virtual ~ConfigTest() {}
+
+  virtual void SetUp() {
+    InitializeConfig();
+    SetMode(GET_PARAM(1));
+  }
+
+  virtual void BeginPassHook(unsigned int /*pass*/) {
+    frame_count_in_ = 0;
+    frame_count_out_ = 0;
+  }
+
+  virtual void PreEncodeFrameHook(libvpx_test::VideoSource* /*video*/) {
+    ++frame_count_in_;
+    abort_ |= (frame_count_in_ >= frame_count_max_);
+  }
+
+  virtual void FramePktHook(const vpx_codec_cx_pkt_t* /*pkt*/) {
+    ++frame_count_out_;
+  }
+
+  unsigned int frame_count_in_;
+  unsigned int frame_count_out_;
+  unsigned int frame_count_max_;
+};
+
+TEST_P(ConfigTest, LagIsDisabled) {
+  frame_count_max_ = 2;
+  cfg_.g_lag_in_frames = 15;
+
+  libvpx_test::DummyVideoSource video;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  EXPECT_EQ(frame_count_in_, frame_count_out_);
+}
+
+VP8_INSTANTIATE_TEST_CASE(ConfigTest, ONE_PASS_TEST_MODES);
+}  // namespace
--- a/test/convolve_test.cc
+++ b/test/convolve_test.cc
--- a/test/cpu_speed_test.cc
+++ b/test/cpu_speed_test.cc
@@ -26,8 +26,7 @@ class CpuSpeedTest
      : EncoderTest(GET_PARAM(0)),
        encoding_mode_(GET_PARAM(1)),
        set_cpu_used_(GET_PARAM(2)),
-        min_psnr_(kMaxPSNR),
-        tune_content_(VPX_CONTENT_DEFAULT) {}
+        min_psnr_(kMaxPSNR) {}
  virtual ~CpuSpeedTest() {}

  virtual void SetUp() {
@@ -50,7 +49,6 @@ class CpuSpeedTest
                                  ::libvpx_test::Encoder *encoder) {
    if (video->frame() == 1) {
      encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
-      encoder->Control(VP9E_SET_TUNE_CONTENT, tune_content_);
      if (encoding_mode_ != ::libvpx_test::kRealTime) {
        encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
        encoder->Control(VP8E_SET_ARNR_MAXFRAMES, 7);
@@ -68,7 +66,6 @@ class CpuSpeedTest
  ::libvpx_test::TestMode encoding_mode_;
  int set_cpu_used_;
  double min_psnr_;
-  int tune_content_;
 };

 TEST_P(CpuSpeedTest, TestQ0) {
@@ -77,13 +74,13 @@ TEST_P(CpuSpeedTest, TestQ0) {
  // the encoder to producing lots of big partitions which will likely
  // extend into the border and test the border condition.
  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_maxsection_pct = 2000;
+  cfg_.rc_2pass_vbr_minsection_pct = 2000;
  cfg_.rc_target_bitrate = 400;
  cfg_.rc_max_quantizer = 0;
  cfg_.rc_min_quantizer = 0;

  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
-                                       10);
+                                       20);

  init_flags_ = VPX_CODEC_USE_PSNR;

@@ -92,10 +89,10 @@ TEST_P(CpuSpeedTest, TestQ0) {
 }

 TEST_P(CpuSpeedTest, TestScreencastQ0) {
-  ::libvpx_test::Y4mVideoSource video("screendata.y4m", 0, 10);
+  ::libvpx_test::Y4mVideoSource video("screendata.y4m", 0, 25);
  cfg_.g_timebase = video.timebase();
  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_maxsection_pct = 2000;
+  cfg_.rc_2pass_vbr_minsection_pct = 2000;
  cfg_.rc_target_bitrate = 400;
  cfg_.rc_max_quantizer = 0;
  cfg_.rc_min_quantizer = 0;
@@ -106,34 +103,19 @@ TEST_P(CpuSpeedTest, TestScreencastQ0) {
  EXPECT_GE(min_psnr_, kMaxPSNR);
 }

-TEST_P(CpuSpeedTest, TestTuneScreen) {
-  ::libvpx_test::Y4mVideoSource video("screendata.y4m", 0, 10);
-  cfg_.g_timebase = video.timebase();
-  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_minsection_pct = 2000;
-  cfg_.rc_target_bitrate = 2000;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.rc_min_quantizer = 0;
-  tune_content_ = VPX_CONTENT_SCREEN;
-
-  init_flags_ = VPX_CODEC_USE_PSNR;
-
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-}
-
 TEST_P(CpuSpeedTest, TestEncodeHighBitrate) {
  // Validate that this non multiple of 64 wide clip encodes and decodes
  // without a mismatch when passing in a very low max q.  This pushes
  // the encoder to producing lots of big partitions which will likely
  // extend into the border and test the border condition.
  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_maxsection_pct = 2000;
+  cfg_.rc_2pass_vbr_minsection_pct = 2000;
  cfg_.rc_target_bitrate = 12000;
  cfg_.rc_max_quantizer = 10;
  cfg_.rc_min_quantizer = 0;

  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
-                                       10);
+                                       20);

  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }
@@ -143,18 +125,19 @@ TEST_P(CpuSpeedTest, TestLowBitrate) {
  // when passing in a very high min q.  This pushes the encoder to producing
  // lots of small partitions which might will test the other condition.
  cfg_.rc_2pass_vbr_minsection_pct = 5;
-  cfg_.rc_2pass_vbr_maxsection_pct = 2000;
+  cfg_.rc_2pass_vbr_minsection_pct = 2000;
  cfg_.rc_target_bitrate = 200;
  cfg_.rc_min_quantizer = 40;

  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
-                                       10);
+                                       20);

  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

-VP10_INSTANTIATE_TEST_CASE(
+VP9_INSTANTIATE_TEST_CASE(
    CpuSpeedTest,
-    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood),
-    ::testing::Range(0, 3));
+    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood,
+                      ::libvpx_test::kRealTime),
+    ::testing::Range(0, 9));
 }  // namespace
--- a/test/cq_test.cc
+++ b/test/cq_test.cc
@@ -0,0 +1,134 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+#include <cmath>
+#include <map>
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "test/codec_factory.h"
+#include "test/encode_test_driver.h"
+#include "test/i420_video_source.h"
+#include "test/util.h"
+
+namespace {
+
+// CQ level range: [kCQLevelMin, kCQLevelMax).
+const int kCQLevelMin = 4;
+const int kCQLevelMax = 63;
+const int kCQLevelStep = 8;
+const unsigned int kCQTargetBitrate = 2000;
+
+class CQTest : public ::libvpx_test::EncoderTest,
+    public ::libvpx_test::CodecTestWithParam<int> {
+ public:
+  // maps the cqlevel to the bitrate produced.
+  typedef std::map<int, uint32_t> BitrateMap;
+
+  static void SetUpTestCase() {
+    bitrates_.clear();
+  }
+
+  static void TearDownTestCase() {
+    ASSERT_TRUE(!HasFailure())
+        << "skipping bitrate validation due to earlier failure.";
+    uint32_t prev_actual_bitrate = kCQTargetBitrate;
+    for (BitrateMap::const_iterator iter = bitrates_.begin();
+         iter != bitrates_.end(); ++iter) {
+      const uint32_t cq_actual_bitrate = iter->second;
+      EXPECT_LE(cq_actual_bitrate, prev_actual_bitrate)
+          << "cq_level: " << iter->first
+          << ", bitrate should decrease with increase in CQ level.";
+      prev_actual_bitrate = cq_actual_bitrate;
+    }
+  }
+
+ protected:
+  CQTest() : EncoderTest(GET_PARAM(0)), cq_level_(GET_PARAM(1)) {
+    init_flags_ = VPX_CODEC_USE_PSNR;
+  }
+
+  virtual ~CQTest() {}
+
+  virtual void SetUp() {
+    InitializeConfig();
+    SetMode(libvpx_test::kTwoPassGood);
+  }
+
+  virtual void BeginPassHook(unsigned int /*pass*/) {
+    file_size_ = 0;
+    psnr_ = 0.0;
+    n_frames_ = 0;
+  }
+
+  virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
+                                  libvpx_test::Encoder *encoder) {
+    if (video->frame() == 1) {
+      if (cfg_.rc_end_usage == VPX_CQ) {
+        encoder->Control(VP8E_SET_CQ_LEVEL, cq_level_);
+      }
+      encoder->Control(VP8E_SET_CPUUSED, 3);
+    }
+  }
+
+  virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
+    psnr_ += pow(10.0, pkt->data.psnr.psnr[0] / 10.0);
+    n_frames_++;
+  }
+
+  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
+    file_size_ += pkt->data.frame.sz;
+  }
+
+  double GetLinearPSNROverBitrate() const {
+    double avg_psnr = log10(psnr_ / n_frames_) * 10.0;
+    return pow(10.0, avg_psnr / 10.0) / file_size_;
+  }
+
+  int cq_level() const { return cq_level_; }
+  size_t file_size() const { return file_size_; }
+  int n_frames() const { return n_frames_; }
+
+  static BitrateMap bitrates_;
+
+ private:
+  int cq_level_;
+  size_t file_size_;
+  double psnr_;
+  int n_frames_;
+};
+
+CQTest::BitrateMap CQTest::bitrates_;
+
+TEST_P(CQTest, LinearPSNRIsHigherForCQLevel) {
+  const vpx_rational timebase = { 33333333, 1000000000 };
+  cfg_.g_timebase = timebase;
+  cfg_.rc_target_bitrate = kCQTargetBitrate;
+  cfg_.g_lag_in_frames = 25;
+
+  cfg_.rc_end_usage = VPX_CQ;
+  libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                     timebase.den, timebase.num, 0, 30);
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+  const double cq_psnr_lin = GetLinearPSNROverBitrate();
+  const unsigned int cq_actual_bitrate =
+      static_cast<unsigned int>(file_size()) * 8 * 30 / (n_frames() * 1000);
+  EXPECT_LE(cq_actual_bitrate, kCQTargetBitrate);
+  bitrates_[cq_level()] = cq_actual_bitrate;
+
+  // try targeting the approximate same bitrate with VBR mode
+  cfg_.rc_end_usage = VPX_VBR;
+  cfg_.rc_target_bitrate = cq_actual_bitrate;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+  const double vbr_psnr_lin = GetLinearPSNROverBitrate();
+  EXPECT_GE(cq_psnr_lin, vbr_psnr_lin);
+}
+
+VP8_INSTANTIATE_TEST_CASE(CQTest,
+                          ::testing::Range(kCQLevelMin, kCQLevelMax,
+                                           kCQLevelStep));
+}  // namespace
--- a/test/cx_set_ref.sh
+++ b/test/cx_set_ref.sh
@@ -1,77 +0,0 @@
-#!/bin/sh
-##
-##  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
-##
-##  Use of this source code is governed by a BSD-style license
-##  that can be found in the LICENSE file in the root of the source
-##  tree. An additional intellectual property rights grant can be found
-##  in the file PATENTS.  All contributing project authors may
-##  be found in the AUTHORS file in the root of the source tree.
-##
-##  This file tests the libvpx cx_set_ref example. To add new tests to this
-##  file, do the following:
-##    1. Write a shell function (this is your test).
-##    2. Add the function to cx_set_ref_tests (on a new line).
-##
-. $(dirname $0)/tools_common.sh
-
-# Environment check: $YUV_RAW_INPUT is required.
-cx_set_ref_verify_environment() {
-  if [ ! -e "${YUV_RAW_INPUT}" ]; then
-    echo "Libvpx test data must exist in LIBVPX_TEST_DATA_PATH."
-    return 1
-  fi
-}
-
-# Runs cx_set_ref and updates the reference frame before encoding frame 90.
-# $1 is the codec name.
-vpx_set_ref() {
-  local codec="$1"
-  local encoder="${LIBVPX_BIN_PATH}/vpxcx_set_ref${VPX_TEST_EXE_SUFFIX}"
-
-  if [ "$codec" = "vp8" ]; then
-    encoder="${LIBVPX_BIN_PATH}/vp8cx_set_ref${VPX_TEST_EXE_SUFFIX}"
-  fi
-
-  local output_file="${VPX_TEST_OUTPUT_DIR}/${codec}cx_set_ref_${codec}.ivf"
-  local ref_frame_num=90
-
-  if [ ! -x "${encoder}" ]; then
-    elog "${encoder} does not exist or is not executable."
-    return 1
-  fi
-
-  if [ "$codec" = "vp8" ]; then
-    eval "${VPX_TEST_PREFIX}" "${encoder}" "${YUV_RAW_INPUT_WIDTH}" \
-        "${YUV_RAW_INPUT_HEIGHT}" "${YUV_RAW_INPUT}" "${output_file}" \
-        "${ref_frame_num}" ${devnull}
-  else
-    eval "${VPX_TEST_PREFIX}" "${encoder}" "${codec}" "${YUV_RAW_INPUT_WIDTH}" \
-        "${YUV_RAW_INPUT_HEIGHT}" "${YUV_RAW_INPUT}" "${output_file}" \
-        "${ref_frame_num}" ${devnull}
-  fi
-
-  [ -e "${output_file}" ] || return 1
-}
-
-cx_set_ref_vp8() {
-  if [ "$(vp8_encode_available)" = "yes" ]; then
-    vpx_set_ref vp8 || return 1
-  fi
-}
-
-cx_set_ref_vp9() {
-  if [ "$(vp9_encode_available)" = "yes" ]; then
-    vpx_set_ref vp9 || return 1
-  fi
-}
-
-cx_set_ref_vp10() {
-  if [ "$(vp10_encode_available)" = "yes" ]; then
-    vpx_set_ref vp10 || return 1
-  fi
-}
-
-cx_set_ref_tests="cx_set_ref_vp8 cx_set_ref_vp9 cx_set_ref_vp10"
-
-run_tests cx_set_ref_verify_environment "${cx_set_ref_tests}"
--- a/test/datarate_test.cc
+++ b/test/datarate_test.cc
@@ -14,7 +14,6 @@
 #include "test/i420_video_source.h"
 #include "test/util.h"
 #include "test/y4m_video_source.h"
-#include "vpx/vpx_codec.h"

 namespace {

@@ -90,7 +89,7 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
          << pkt->data.frame.pts;
    }

-    const int64_t frame_size_in_bits = pkt->data.frame.sz * 8;
+    const size_t frame_size_in_bits = pkt->data.frame.sz * 8;

    // Subtract from the buffer the bits associated with a played back frame.
    bits_in_buffer_model_ -= frame_size_in_bits;
@@ -135,12 +134,65 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
  double duration_;
  double file_datarate_;
  double effective_datarate_;
-  int64_t bits_in_last_frame_;
+  size_t bits_in_last_frame_;
  int denoiser_on_;
  int denoiser_offon_test_;
  int denoiser_offon_period_;
 };

+#if CONFIG_TEMPORAL_DENOISING
+// Check basic datarate targeting, for a single bitrate, but loop over the
+// various denoiser settings.
+TEST_P(DatarateTestLarge, DenoiserLevels) {
+  cfg_.rc_buf_initial_sz = 500;
+  cfg_.rc_dropframe_thresh = 1;
+  cfg_.rc_max_quantizer = 56;
+  cfg_.rc_end_usage = VPX_CBR;
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                       30, 1, 0, 140);
+  for (int j = 1; j < 5; ++j) {
+    // Run over the denoiser levels.
+    // For the temporal denoiser (#if CONFIG_TEMPORAL_DENOISING) the level j
+    // refers to the 4 denoiser modes: denoiserYonly, denoiserOnYUV,
+    // denoiserOnAggressive, and denoiserOnAdaptive.
+    // For the spatial denoiser (if !CONFIG_TEMPORAL_DENOISING), the level j
+    // refers to the blur thresholds: 20, 40, 60 80.
+    // The j = 0 case (denoiser off) is covered in the tests below.
+    denoiser_on_ = j;
+    cfg_.rc_target_bitrate = 300;
+    ResetModel();
+    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+    ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
+        << " The datarate for the file exceeds the target!";
+
+    ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.3)
+        << " The datarate for the file missed the target!";
+  }
+}
+
+// Check basic datarate targeting, for a single bitrate, when denoiser is off
+// and on.
+TEST_P(DatarateTestLarge, DenoiserOffOn) {
+  cfg_.rc_buf_initial_sz = 500;
+  cfg_.rc_dropframe_thresh = 1;
+  cfg_.rc_max_quantizer = 56;
+  cfg_.rc_end_usage = VPX_CBR;
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                       30, 1, 0, 299);
+  cfg_.rc_target_bitrate = 300;
+  ResetModel();
+  // The denoiser is off by default.
+  denoiser_on_ = 0;
+  // Set the offon test flag.
+  denoiser_offon_test_ = 1;
+  denoiser_offon_period_ = 100;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+  ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
+      << " The datarate for the file exceeds the target!";
+  ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.3)
+      << " The datarate for the file missed the target!";
+}
+#endif  // CONFIG_TEMPORAL_DENOISING

 TEST_P(DatarateTestLarge, BasicBufferModel) {
  denoiser_on_ = 0;
@@ -319,7 +371,9 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
        encoder->Control(VP9E_SET_SVC, 1);
      }
      vpx_svc_layer_id_t layer_id;
+#if VPX_ENCODER_ABI_VERSION > (4 + VPX_CODEC_ABI_VERSION)
      layer_id.spatial_layer_id = 0;
+#endif
      frame_flags_ = SetFrameFlags(video->frame(), cfg_.ts_number_layers);
      layer_id.temporal_layer_id = SetLayerId(video->frame(),
                                              cfg_.ts_number_layers);
@@ -397,28 +451,7 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
  int denoiser_offon_period_;
 };

-// Check basic rate targeting for VBR mode.
-TEST_P(DatarateTestVP9Large, BasicRateTargetingVBR) {
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.g_error_resilient = 0;
-  cfg_.rc_end_usage = VPX_VBR;
-  cfg_.g_lag_in_frames = 0;
-
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 300);
-  for (int i = 400; i <= 800; i += 400) {
-    cfg_.rc_target_bitrate = i;
-    ResetModel();
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.75)
-        << " The datarate for the file is lower than target by too much!";
-    ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.25)
-        << " The datarate for the file is greater than target by too much!";
-  }
-}
-
-// Check basic rate targeting for CBR,
+// Check basic rate targeting,
 TEST_P(DatarateTestVP9Large, BasicRateTargeting) {
  cfg_.rc_buf_initial_sz = 500;
  cfg_.rc_buf_optimal_sz = 500;
@@ -442,7 +475,7 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting) {
  }
 }

-// Check basic rate targeting for CBR.
+// Check basic rate targeting,
 TEST_P(DatarateTestVP9Large, BasicRateTargeting444) {
  ::libvpx_test::Y4mVideoSource video("rush_hour_444.y4m", 0, 140);

@@ -487,9 +520,6 @@ TEST_P(DatarateTestVP9Large, ChangingDropFrameThresh) {
  cfg_.rc_end_usage = VPX_CBR;
  cfg_.rc_target_bitrate = 200;
  cfg_.g_lag_in_frames = 0;
-  // TODO(marpan): Investigate datarate target failures with a smaller keyframe
-  // interval (128).
-  cfg_.kf_max_dist = 9999;

  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
                                       30, 1, 0, 140);
@@ -509,7 +539,7 @@ TEST_P(DatarateTestVP9Large, ChangingDropFrameThresh) {
        << " The first dropped frame for drop_thresh " << i
        << " > first dropped frame for drop_thresh "
        << i - kDropFrameThreshTestStep;
-    ASSERT_GE(num_drops_, last_num_drops * 0.85)
+    ASSERT_GE(num_drops_, last_num_drops * 0.90)
        << " The number of dropped frames for drop_thresh " << i
        << " < number of dropped frames for drop_thresh "
        << i - kDropFrameThreshTestStep;
@@ -535,8 +565,6 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting2TemporalLayers) {
  cfg_.ts_rate_decimator[0] = 2;
  cfg_.ts_rate_decimator[1] = 1;

-  cfg_.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
-
  if (deadline_ == VPX_DL_REALTIME)
    cfg_.g_error_resilient = 1;

@@ -546,14 +574,14 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting2TemporalLayers) {
    cfg_.rc_target_bitrate = i;
    ResetModel();
    // 60-40 bitrate allocation for 2 temporal layers.
-    cfg_.layer_target_bitrate[0] = 60 * cfg_.rc_target_bitrate / 100;
-    cfg_.layer_target_bitrate[1] = cfg_.rc_target_bitrate;
+    cfg_.ts_target_bitrate[0] = 60 * cfg_.rc_target_bitrate / 100;
+    cfg_.ts_target_bitrate[1] = cfg_.rc_target_bitrate;
    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
    for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
-      ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.85)
+      ASSERT_GE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 0.85)
          << " The datarate for the file is lower than target by too much, "
              "for layer: " << j;
-      ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.15)
+      ASSERT_LE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 1.15)
          << " The datarate for the file is greater than target by too much, "
              "for layer: " << j;
    }
@@ -578,27 +606,25 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayers) {
  cfg_.ts_rate_decimator[1] = 2;
  cfg_.ts_rate_decimator[2] = 1;

-  cfg_.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
-
  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
                                       30, 1, 0, 200);
  for (int i = 200; i <= 800; i += 200) {
    cfg_.rc_target_bitrate = i;
    ResetModel();
    // 40-20-40 bitrate allocation for 3 temporal layers.
-    cfg_.layer_target_bitrate[0] = 40 * cfg_.rc_target_bitrate / 100;
-    cfg_.layer_target_bitrate[1] = 60 * cfg_.rc_target_bitrate / 100;
-    cfg_.layer_target_bitrate[2] = cfg_.rc_target_bitrate;
+    cfg_.ts_target_bitrate[0] = 40 * cfg_.rc_target_bitrate / 100;
+    cfg_.ts_target_bitrate[1] = 60 * cfg_.rc_target_bitrate / 100;
+    cfg_.ts_target_bitrate[2] = cfg_.rc_target_bitrate;
    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
    for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
      // TODO(yaowu): Work out more stable rc control strategy and
      //              Adjust the thresholds to be tighter than .75.
-      ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.75)
+      ASSERT_GE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 0.75)
          << " The datarate for the file is lower than target by too much, "
              "for layer: " << j;
      // TODO(yaowu): Work out more stable rc control strategy and
      //              Adjust the thresholds to be tighter than 1.25.
-      ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.25)
+      ASSERT_LE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 1.25)
          << " The datarate for the file is greater than target by too much, "
              "for layer: " << j;
    }
@@ -626,22 +652,20 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayersFrameDropping) {
  cfg_.ts_rate_decimator[1] = 2;
  cfg_.ts_rate_decimator[2] = 1;

-  cfg_.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
-
  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
                                       30, 1, 0, 200);
  cfg_.rc_target_bitrate = 200;
  ResetModel();
  // 40-20-40 bitrate allocation for 3 temporal layers.
-  cfg_.layer_target_bitrate[0] = 40 * cfg_.rc_target_bitrate / 100;
-  cfg_.layer_target_bitrate[1] = 60 * cfg_.rc_target_bitrate / 100;
-  cfg_.layer_target_bitrate[2] = cfg_.rc_target_bitrate;
+  cfg_.ts_target_bitrate[0] = 40 * cfg_.rc_target_bitrate / 100;
+  cfg_.ts_target_bitrate[1] = 60 * cfg_.rc_target_bitrate / 100;
+  cfg_.ts_target_bitrate[2] = cfg_.rc_target_bitrate;
  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
  for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
-    ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.85)
+    ASSERT_GE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 0.85)
        << " The datarate for the file is lower than target by too much, "
            "for layer: " << j;
-    ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.15)
+    ASSERT_LE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 1.15)
        << " The datarate for the file is greater than target by too much, "
            "for layer: " << j;
    // Expect some frame drops in this test: for this 200 frames test,
@@ -651,388 +675,71 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayersFrameDropping) {
  }
 }

-class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
-    public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- public:
-  DatarateOnePassCbrSvc() : EncoderTest(GET_PARAM(0)) {}
-  virtual ~DatarateOnePassCbrSvc() {}
- protected:
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(GET_PARAM(1));
-    speed_setting_ = GET_PARAM(2);
-    ResetModel();
-  }
-  virtual void ResetModel() {
-    last_pts_ = 0;
-    bits_in_buffer_model_ = cfg_.rc_target_bitrate * cfg_.rc_buf_initial_sz;
-    frame_number_ = 0;
-    first_drop_ = 0;
-    bits_total_ = 0;
-    duration_ = 0.0;
-    mismatch_psnr_ = 0.0;
-    mismatch_nframes_ = 0;
-  }
-  virtual void BeginPassHook(unsigned int /*pass*/) {
-  }
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (video->frame() == 0) {
-      int i;
-      for (i = 0; i < VPX_MAX_LAYERS; ++i) {
-        svc_params_.max_quantizers[i] = 63;
-        svc_params_.min_quantizers[i] = 0;
-      }
-      encoder->Control(VP9E_SET_SVC, 1);
-      encoder->Control(VP9E_SET_SVC_PARAMETERS, &svc_params_);
-      encoder->Control(VP8E_SET_CPUUSED, speed_setting_);
-      encoder->Control(VP9E_SET_TILE_COLUMNS, 0);
-      encoder->Control(VP8E_SET_MAX_INTRA_BITRATE_PCT, 300);
-      encoder->Control(VP9E_SET_TILE_COLUMNS, (cfg_.g_threads >> 1));
-    }
-    const vpx_rational_t tb = video->timebase();
-    timebase_ = static_cast<double>(tb.num) / tb.den;
-    duration_ = 0;
-  }
-  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
-    vpx_codec_pts_t duration = pkt->data.frame.pts - last_pts_;
-    if (last_pts_ == 0)
-      duration = 1;
-    bits_in_buffer_model_ += static_cast<int64_t>(
-        duration * timebase_ * cfg_.rc_target_bitrate * 1000);
-    const bool key_frame = (pkt->data.frame.flags & VPX_FRAME_IS_KEY)
-                         ? true: false;
-    if (!key_frame) {
-      ASSERT_GE(bits_in_buffer_model_, 0) << "Buffer Underrun at frame "
-          << pkt->data.frame.pts;
-    }
-    const size_t frame_size_in_bits = pkt->data.frame.sz * 8;
-    bits_in_buffer_model_ -= frame_size_in_bits;
-    bits_total_ += frame_size_in_bits;
-    if (!first_drop_ && duration > 1)
-      first_drop_ = last_pts_ + 1;
-    last_pts_ = pkt->data.frame.pts;
-    bits_in_last_frame_ = frame_size_in_bits;
-    ++frame_number_;
-  }
-  virtual void EndPassHook(void) {
-    if (bits_total_) {
-      const double file_size_in_kb = bits_total_ / 1000.;  // bits per kilobit
-      duration_ = (last_pts_ + 1) * timebase_;
-      file_datarate_ = file_size_in_kb / duration_;
-    }
-  }
-
-  virtual void MismatchHook(const vpx_image_t *img1,
-                            const vpx_image_t *img2) {
-    double mismatch_psnr = compute_psnr(img1, img2);
-    mismatch_psnr_ += mismatch_psnr;
-    ++mismatch_nframes_;
-  }
-
-  unsigned int GetMismatchFrames() {
-    return mismatch_nframes_;
-  }
-
-  vpx_codec_pts_t last_pts_;
-  int64_t bits_in_buffer_model_;
-  double timebase_;
-  int frame_number_;
-  vpx_codec_pts_t first_drop_;
-  int64_t bits_total_;
-  double duration_;
-  double file_datarate_;
-  size_t bits_in_last_frame_;
-  vpx_svc_extra_cfg_t svc_params_;
-  int speed_setting_;
-  double mismatch_psnr_;
-  int mismatch_nframes_;
-};
-static void assign_layer_bitrates(vpx_codec_enc_cfg_t *const enc_cfg,
-    const vpx_svc_extra_cfg_t *svc_params,
-    int spatial_layers,
-    int temporal_layers,
-    int temporal_layering_mode) {
-  int sl, spatial_layer_target;
-  float total = 0;
-  float alloc_ratio[VPX_MAX_LAYERS] = {0};
-  for (sl = 0; sl < spatial_layers; ++sl) {
-    if (svc_params->scaling_factor_den[sl] > 0) {
-      alloc_ratio[sl] = (float)(svc_params->scaling_factor_num[sl] *
-          1.0 / svc_params->scaling_factor_den[sl]);
-      total += alloc_ratio[sl];
-    }
-  }
-  for (sl = 0; sl < spatial_layers; ++sl) {
-    enc_cfg->ss_target_bitrate[sl] = spatial_layer_target =
-        (unsigned int)(enc_cfg->rc_target_bitrate *
-            alloc_ratio[sl] / total);
-    const int index = sl * temporal_layers;
-    if (temporal_layering_mode == 3) {
-      enc_cfg->layer_target_bitrate[index] =
-          spatial_layer_target >> 1;
-      enc_cfg->layer_target_bitrate[index + 1] =
-          (spatial_layer_target >> 1) + (spatial_layer_target >> 2);
-      enc_cfg->layer_target_bitrate[index + 2] =
-          spatial_layer_target;
-    } else if (temporal_layering_mode == 2) {
-      enc_cfg->layer_target_bitrate[index] =
-          spatial_layer_target * 2 / 3;
-      enc_cfg->layer_target_bitrate[index + 1] =
-          spatial_layer_target;
-    }
-  }
-}
-
-// Check basic rate targeting for 1 pass CBR SVC: 2 spatial layers and
-// 3 temporal layers. Run CIF clip with 1 thread.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayers) {
+#if CONFIG_VP9_TEMPORAL_DENOISING
+// Check basic datarate targeting, for a single bitrate, when denoiser is on.
+TEST_P(DatarateTestVP9Large, DenoiserLevels) {
  cfg_.rc_buf_initial_sz = 500;
  cfg_.rc_buf_optimal_sz = 500;
  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
+  cfg_.rc_dropframe_thresh = 1;
+  cfg_.rc_min_quantizer = 2;
+  cfg_.rc_max_quantizer = 56;
  cfg_.rc_end_usage = VPX_CBR;
  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 2;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 1;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 144;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 288;
-  svc_params_.scaling_factor_den[1] = 288;
-  cfg_.rc_dropframe_thresh = 10;
-  cfg_.kf_max_dist = 9999;
+
  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 200);
-  // TODO(wonkap/marpan): Check that effective_datarate for each layer hits the
-  // layer target_bitrate.
-  for (int i = 200; i <= 800; i += 200) {
-    cfg_.rc_target_bitrate = i;
-    ResetModel();
-    assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-        cfg_.ts_number_layers, cfg_.temporal_layering_mode);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-            << " The datarate for the file exceeds the target by too much!";
-    ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
-        << " The datarate for the file is lower than the target by too much!";
-    EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-  }
+                                       30, 1, 0, 140);
+
+  // For the temporal denoiser (#if CONFIG_VP9_TEMPORAL_DENOISING),
+  // there is only one denoiser mode: denoiserYonly(which is 1),
+  // but may add more modes in the future.
+  cfg_.rc_target_bitrate = 300;
+  ResetModel();
+  // Turn on the denoiser.
+  denoiser_on_ = 1;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+  ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.85)
+      << " The datarate for the file is lower than target by too much!";
+  ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.15)
+      << " The datarate for the file is greater than target by too much!";
 }

-// Check basic rate targeting for 1 pass CBR SVC: 2 spatial layers and 3
-// temporal layers. Run CIF clip with 1 thread, and few short key frame periods.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayersSmallKf) {
+// Check basic datarate targeting, for a single bitrate, when denoiser is off
+// and on.
+TEST_P(DatarateTestVP9Large, DenoiserOffOn) {
  cfg_.rc_buf_initial_sz = 500;
  cfg_.rc_buf_optimal_sz = 500;
  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
+  cfg_.rc_dropframe_thresh = 1;
+  cfg_.rc_min_quantizer = 2;
+  cfg_.rc_max_quantizer = 56;
  cfg_.rc_end_usage = VPX_CBR;
  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 2;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 1;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 144;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 288;
-  svc_params_.scaling_factor_den[1] = 288;
-  cfg_.rc_dropframe_thresh = 10;
+
  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 200);
-  cfg_.rc_target_bitrate = 400;
-  // For this 3 temporal layer case, pattern repeats every 4 frames, so choose
-  // 4 key neighboring key frame periods (so key frame will land on 0-2-1-2).
-  for (int j = 64; j <= 67; j++) {
-    cfg_.kf_max_dist = j;
-    ResetModel();
-    assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-        cfg_.ts_number_layers, cfg_.temporal_layering_mode);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-            << " The datarate for the file exceeds the target by too much!";
-    ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
-        << " The datarate for the file is lower than the target by too much!";
-    EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-  }
-}
+                                       30, 1, 0, 299);

-// Check basic rate targeting for 1 pass CBR SVC: 2 spatial layers and
-// 3 temporal layers. Run HD clip with 4 threads.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayers4threads) {
-  cfg_.rc_buf_initial_sz = 500;
-  cfg_.rc_buf_optimal_sz = 500;
-  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.rc_end_usage = VPX_CBR;
-  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 2;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 4;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 144;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 288;
-  svc_params_.scaling_factor_den[1] = 288;
-  cfg_.rc_dropframe_thresh = 10;
-  cfg_.kf_max_dist = 9999;
-  ::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
-                                       30, 1, 0, 300);
-  cfg_.rc_target_bitrate = 800;
+  // For the temporal denoiser (#if CONFIG_VP9_TEMPORAL_DENOISING),
+  // there is only one denoiser mode: denoiserYonly(which is 1),
+  // but may add more modes in the future.
+  cfg_.rc_target_bitrate = 300;
  ResetModel();
-  assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-      cfg_.ts_number_layers, cfg_.temporal_layering_mode);
+  // The denoiser is off by default.
+  denoiser_on_ = 0;
+  // Set the offon test flag.
+  denoiser_offon_test_ = 1;
+  denoiser_offon_period_ = 100;
  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-          << " The datarate for the file exceeds the target by too much!";
-  ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
-      << " The datarate for the file is lower than the target by too much!";
-  EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
+  ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.85)
+      << " The datarate for the file is lower than target by too much!";
+  ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.15)
+      << " The datarate for the file is greater than target by too much!";
 }
+#endif  // CONFIG_VP9_TEMPORAL_DENOISING

-// Check basic rate targeting for 1 pass CBR SVC: 3 spatial layers and
-// 3 temporal layers. Run CIF clip with 1 thread.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayers) {
-  cfg_.rc_buf_initial_sz = 500;
-  cfg_.rc_buf_optimal_sz = 500;
-  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.rc_end_usage = VPX_CBR;
-  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 3;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 1;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 72;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 144;
-  svc_params_.scaling_factor_den[1] = 288;
-  svc_params_.scaling_factor_num[2] = 288;
-  svc_params_.scaling_factor_den[2] = 288;
-  cfg_.rc_dropframe_thresh = 10;
-  cfg_.kf_max_dist = 9999;
-  ::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
-                                       30, 1, 0, 300);
-  cfg_.rc_target_bitrate = 800;
-  ResetModel();
-  assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-     cfg_.ts_number_layers, cfg_.temporal_layering_mode);
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-          << " The datarate for the file exceeds the target by too much!";
-  ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.22)
-      << " The datarate for the file is lower than the target by too much!";
-  EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-}
-
-// Check basic rate targeting for 1 pass CBR SVC: 3 spatial layers and 3
-// temporal layers. Run CIF clip with 1 thread, and few short key frame periods.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayersSmallKf) {
-  cfg_.rc_buf_initial_sz = 500;
-  cfg_.rc_buf_optimal_sz = 500;
-  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.rc_end_usage = VPX_CBR;
-  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 3;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 1;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 72;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 144;
-  svc_params_.scaling_factor_den[1] = 288;
-  svc_params_.scaling_factor_num[2] = 288;
-  svc_params_.scaling_factor_den[2] = 288;
-  cfg_.rc_dropframe_thresh = 10;
-  ::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
-                                       30, 1, 0, 300);
-  cfg_.rc_target_bitrate = 800;
-  // For this 3 temporal layer case, pattern repeats every 4 frames, so choose
-  // 4 key neighboring key frame periods (so key frame will land on 0-2-1-2).
-  for (int j = 32; j <= 35; j++) {
-    cfg_.kf_max_dist = j;
-    ResetModel();
-    assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-       cfg_.ts_number_layers, cfg_.temporal_layering_mode);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-            << " The datarate for the file exceeds the target by too much!";
-    ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.30)
-        << " The datarate for the file is lower than the target by too much!";
-    EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-  }
-}
-
-// Check basic rate targeting for 1 pass CBR SVC: 3 spatial layers and
-// 3 temporal layers. Run HD clip with 4 threads.
-TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayers4threads) {
-  cfg_.rc_buf_initial_sz = 500;
-  cfg_.rc_buf_optimal_sz = 500;
-  cfg_.rc_buf_sz = 1000;
-  cfg_.rc_min_quantizer = 0;
-  cfg_.rc_max_quantizer = 63;
-  cfg_.rc_end_usage = VPX_CBR;
-  cfg_.g_lag_in_frames = 0;
-  cfg_.ss_number_layers = 3;
-  cfg_.ts_number_layers = 3;
-  cfg_.ts_rate_decimator[0] = 4;
-  cfg_.ts_rate_decimator[1] = 2;
-  cfg_.ts_rate_decimator[2] = 1;
-  cfg_.g_error_resilient = 1;
-  cfg_.g_threads = 4;
-  cfg_.temporal_layering_mode = 3;
-  svc_params_.scaling_factor_num[0] = 72;
-  svc_params_.scaling_factor_den[0] = 288;
-  svc_params_.scaling_factor_num[1] = 144;
-  svc_params_.scaling_factor_den[1] = 288;
-  svc_params_.scaling_factor_num[2] = 288;
-  svc_params_.scaling_factor_den[2] = 288;
-  cfg_.rc_dropframe_thresh = 10;
-  cfg_.kf_max_dist = 9999;
-  ::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
-                                       30, 1, 0, 300);
-  cfg_.rc_target_bitrate = 800;
-  ResetModel();
-  assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
-      cfg_.ts_number_layers, cfg_.temporal_layering_mode);
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
-          << " The datarate for the file exceeds the target by too much!";
-  ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.22)
-      << " The datarate for the file is lower than the target by too much!";
-  EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-}
-
-/* VP10 does not support multiple layers yet.
-VP10_INSTANTIATE_TEST_CASE(DatarateOnePassCbrSvc,
-                           ::testing::Values(::libvpx_test::kRealTime),
-                           ::testing::Range(5, 8));
-                           */
+VP8_INSTANTIATE_TEST_CASE(DatarateTestLarge, ALL_TEST_MODES);
+VP9_INSTANTIATE_TEST_CASE(DatarateTestVP9Large,
+                          ::testing::Values(::libvpx_test::kOnePassGood,
+                                            ::libvpx_test::kRealTime),
+                          ::testing::Range(2, 7));
 }  // namespace
--- a/test/dct16x16_test.cc
+++ b/test/dct16x16_test.cc
@@ -13,25 +13,54 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp10_rtcd.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vp10/common/entropy.h"
-#include "vp10/common/scan.h"
+
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_entropy.h"
 #include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"
-#include "vpx_ports/msvc.h"  // for round()

 using libvpx_test::ACMRandom;

 namespace {

+#ifdef _MSC_VER
+static int round(double x) {
+  if (x < 0)
+    return static_cast<int>(ceil(x - 0.5));
+  else
+    return static_cast<int>(floor(x + 0.5));
+}
+#endif
+
 const int kNumCoeffs = 256;
+const double PI = 3.1415926535898;
+void reference2_16x16_idct_2d(double *input, double *output) {
+  double x;
+  for (int l = 0; l < 16; ++l) {
+    for (int k = 0; k < 16; ++k) {
+      double s = 0;
+      for (int i = 0; i < 16; ++i) {
+        for (int j = 0; j < 16; ++j) {
+          x = cos(PI * j * (l + 0.5) / 16.0) *
+              cos(PI * i * (k + 0.5) / 16.0) *
+              input[i * 16 + j] / 256;
+          if (i != 0)
+            x *= sqrt(2.0);
+          if (j != 0)
+            x *= sqrt(2.0);
+          s += x;
+        }
+      }
+      output[k*16+l] = s;
+    }
+  }
+}
+
+
 const double C1 = 0.995184726672197;
 const double C2 = 0.98078528040323;
 const double C3 = 0.956940335732209;
@@ -240,31 +269,31 @@ typedef std::tr1::tuple<IdctFunc, IdctFunc, int, vpx_bit_depth_t>

 void fdct16x16_ref(const int16_t *in, tran_low_t *out, int stride,
                   int /*tx_type*/) {
-  vpx_fdct16x16_c(in, out, stride);
+  vp9_fdct16x16_c(in, out, stride);
 }

 void idct16x16_ref(const tran_low_t *in, uint8_t *dest, int stride,
                   int /*tx_type*/) {
-  vpx_idct16x16_256_add_c(in, dest, stride);
+  vp9_idct16x16_256_add_c(in, dest, stride);
 }

 void fht16x16_ref(const int16_t *in, tran_low_t *out, int stride,
                  int tx_type) {
-  vp10_fht16x16_c(in, out, stride, tx_type);
+  vp9_fht16x16_c(in, out, stride, tx_type);
 }

 void iht16x16_ref(const tran_low_t *in, uint8_t *dest, int stride,
                  int tx_type) {
-  vp10_iht16x16_256_add_c(in, dest, stride, tx_type);
+  vp9_iht16x16_256_add_c(in, dest, stride, tx_type);
 }

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 void idct16x16_10(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_256_add_c(in, out, stride, 10);
+  vp9_highbd_idct16x16_256_add_c(in, out, stride, 10);
 }

 void idct16x16_12(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_256_add_c(in, out, stride, 12);
+  vp9_highbd_idct16x16_256_add_c(in, out, stride, 12);
 }

 void idct16x16_10_ref(const tran_low_t *in, uint8_t *out, int stride,
@@ -278,39 +307,39 @@ void idct16x16_12_ref(const tran_low_t *in, uint8_t *out, int stride,
 }

 void iht16x16_10(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht16x16_256_add_c(in, out, stride, tx_type, 10);
+  vp9_highbd_iht16x16_256_add_c(in, out, stride, tx_type, 10);
 }

 void iht16x16_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht16x16_256_add_c(in, out, stride, tx_type, 12);
+  vp9_highbd_iht16x16_256_add_c(in, out, stride, tx_type, 12);
 }

-#if HAVE_SSE2
 void idct16x16_10_add_10_c(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_10_add_c(in, out, stride, 10);
+  vp9_highbd_idct16x16_10_add_c(in, out, stride, 10);
 }

 void idct16x16_10_add_12_c(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_10_add_c(in, out, stride, 12);
+  vp9_highbd_idct16x16_10_add_c(in, out, stride, 12);
 }

+#if HAVE_SSE2
 void idct16x16_256_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_256_add_sse2(in, out, stride, 10);
+  vp9_highbd_idct16x16_256_add_sse2(in, out, stride, 10);
 }

 void idct16x16_256_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_256_add_sse2(in, out, stride, 12);
+  vp9_highbd_idct16x16_256_add_sse2(in, out, stride, 12);
 }

 void idct16x16_10_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_10_add_sse2(in, out, stride, 10);
+  vp9_highbd_idct16x16_10_add_sse2(in, out, stride, 10);
 }

 void idct16x16_10_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct16x16_10_add_sse2(in, out, stride, 12);
+  vp9_highbd_idct16x16_10_add_sse2(in, out, stride, 12);
 }
 #endif  // HAVE_SSE2
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

 class Trans16x16TestBase {
 public:
@@ -327,13 +356,13 @@ class Trans16x16TestBase {
    int64_t total_error = 0;
    const int count_test_block = 10000;
    for (int i = 0; i < count_test_block; ++i) {
-      DECLARE_ALIGNED(16, int16_t, test_input_block[kNumCoeffs]);
-      DECLARE_ALIGNED(16, tran_low_t, test_temp_block[kNumCoeffs]);
-      DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-      DECLARE_ALIGNED(16, uint8_t, src[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-      DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-      DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+      DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_temp_block, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+      DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
 #endif

      // Initialize a test block with input range [-mask_, mask_].
@@ -342,7 +371,7 @@ class Trans16x16TestBase {
          src[j] = rnd.Rand8();
          dst[j] = rnd.Rand8();
          test_input_block[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          src16[j] = rnd.Rand16() & mask_;
          dst16[j] = rnd.Rand16() & mask_;
@@ -356,7 +385,7 @@ class Trans16x16TestBase {
      if (bit_depth_ == VPX_BITS_8) {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, CONVERT_TO_BYTEPTR(dst16), pitch_));
@@ -364,11 +393,11 @@ class Trans16x16TestBase {
      }

      for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
-        const int32_t diff =
+#if CONFIG_VP9_HIGHBITDEPTH
+        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ?  dst[j] - src[j] : dst16[j] - src16[j];
 #else
-        const int32_t diff = dst[j] - src[j];
+        const uint32_t diff = dst[j] - src[j];
 #endif
        const uint32_t error = diff * diff;
        if (max_error < error)
@@ -387,9 +416,9 @@ class Trans16x16TestBase {
  void RunCoeffCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 1000;
-    DECLARE_ALIGNED(16, int16_t, input_block[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, output_block[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);

    for (int i = 0; i < count_test_block; ++i) {
      // Initialize a test block with input range [-mask_, mask_].
@@ -408,13 +437,15 @@ class Trans16x16TestBase {
  void RunMemCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 1000;
-    DECLARE_ALIGNED(16, int16_t, input_extreme_block[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, output_block[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_extreme_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);

    for (int i = 0; i < count_test_block; ++i) {
      // Initialize a test block with input range [-mask_, mask_].
      for (int j = 0; j < kNumCoeffs; ++j) {
+        input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
        input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
      }
      if (i == 0) {
@@ -441,19 +472,24 @@ class Trans16x16TestBase {
  void RunQuantCheck(int dc_thred, int ac_thred) {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 100000;
-    DECLARE_ALIGNED(16, int16_t, input_extreme_block[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_extreme_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);

-    DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, ref[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, ref, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, ref16, kNumCoeffs);
 #endif

    for (int i = 0; i < count_test_block; ++i) {
      // Initialize a test block with input range [-mask_, mask_].
      for (int j = 0; j < kNumCoeffs; ++j) {
+        if (bit_depth_ == VPX_BITS_8)
+          input_block[j] = rnd.Rand8() - rnd.Rand8();
+        else
+          input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
        input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
      }
      if (i == 0)
@@ -466,11 +502,11 @@ class Trans16x16TestBase {
      fwd_txfm_ref(input_extreme_block, output_ref_block, pitch_, tx_type_);

      // clear reconstructed pixel buffers
-      memset(dst, 0, kNumCoeffs * sizeof(uint8_t));
-      memset(ref, 0, kNumCoeffs * sizeof(uint8_t));
-#if CONFIG_VPX_HIGHBITDEPTH
-      memset(dst16, 0, kNumCoeffs * sizeof(uint16_t));
-      memset(ref16, 0, kNumCoeffs * sizeof(uint16_t));
+      vpx_memset(dst, 0, kNumCoeffs * sizeof(uint8_t));
+      vpx_memset(ref, 0, kNumCoeffs * sizeof(uint8_t));
+#if CONFIG_VP9_HIGHBITDEPTH
+      vpx_memset(dst16, 0, kNumCoeffs * sizeof(uint16_t));
+      vpx_memset(ref16, 0, kNumCoeffs * sizeof(uint16_t));
 #endif

      // quantization with maximum allowed step sizes
@@ -480,7 +516,7 @@ class Trans16x16TestBase {
      if (bit_depth_ == VPX_BITS_8) {
        inv_txfm_ref(output_ref_block, ref, pitch_, tx_type_);
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(output_ref_block, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        inv_txfm_ref(output_ref_block, CONVERT_TO_BYTEPTR(ref16), pitch_,
                     tx_type_);
@@ -491,7 +527,7 @@ class Trans16x16TestBase {
      if (bit_depth_ == VPX_BITS_8) {
        for (int j = 0; j < kNumCoeffs; ++j)
          EXPECT_EQ(ref[j], dst[j]);
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        for (int j = 0; j < kNumCoeffs; ++j)
          EXPECT_EQ(ref16[j], dst16[j]);
@@ -503,14 +539,14 @@ class Trans16x16TestBase {
  void RunInvAccuracyCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 1000;
-    DECLARE_ALIGNED(16, int16_t, in[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, src[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, int16_t, in, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH

    for (int i = 0; i < count_test_block; ++i) {
      double out_r[kNumCoeffs];
@@ -521,12 +557,12 @@ class Trans16x16TestBase {
          src[j] = rnd.Rand8();
          dst[j] = rnd.Rand8();
          in[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          src16[j] = rnd.Rand16() & mask_;
          dst16[j] = rnd.Rand16() & mask_;
          in[j] = src16[j] - dst16[j];
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
        }
      }

@@ -536,20 +572,20 @@ class Trans16x16TestBase {

      if (bit_depth_ == VPX_BITS_8) {
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, 16));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
                                            16));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
      }

      for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
        const uint32_t diff = dst[j] - src[j];
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
        const uint32_t error = diff * diff;
        EXPECT_GE(1u, error)
            << "Error: 16x16 IDCT has error " << error
@@ -562,14 +598,14 @@ class Trans16x16TestBase {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 10000;
    const int eob = 10;
-    const int16_t *scan = vp10_default_scan_orders[TX_16X16].scan;
-    DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, ref[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+    const int16_t *scan = vp9_default_scan_orders[TX_16X16].scan;
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, ref, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, ref16, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH

    for (int i = 0; i < count_test_block; ++i) {
      for (int j = 0; j < kNumCoeffs; ++j) {
@@ -582,31 +618,31 @@ class Trans16x16TestBase {
        if (bit_depth_ == VPX_BITS_8) {
          dst[j] = 0;
          ref[j] = 0;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          dst16[j] = 0;
          ref16[j] = 0;
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
        }
      }
      if (bit_depth_ == VPX_BITS_8) {
        ref_txfm(coeff, ref, pitch_);
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
      } else {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        ref_txfm(coeff, CONVERT_TO_BYTEPTR(ref16), pitch_);
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
                                 pitch_));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
      }

      for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - ref[j] : dst16[j] - ref16[j];
 #else
        const uint32_t diff = dst[j] - ref[j];
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
        const uint32_t error = diff * diff;
        EXPECT_EQ(0u, error)
            << "Error: 16x16 IDCT Comparison has error " << error
@@ -638,7 +674,7 @@ class Trans16x16DCT
    fwd_txfm_ref = fdct16x16_ref;
    inv_txfm_ref = idct16x16_ref;
    mask_ = (1 << bit_depth_) - 1;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    switch (bit_depth_) {
      case VPX_BITS_10:
        inv_txfm_ref = idct16x16_10_ref;
@@ -705,7 +741,7 @@ class Trans16x16HT
    fwd_txfm_ref = fht16x16_ref;
    inv_txfm_ref = iht16x16_ref;
    mask_ = (1 << bit_depth_) - 1;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    switch (bit_depth_) {
      case VPX_BITS_10:
        inv_txfm_ref = iht16x16_10;
@@ -784,171 +820,104 @@ TEST_P(InvTrans16x16DCT, CompareReference) {
  CompareInvReference(ref_txfm_, thresh_);
 }

-class PartialTrans16x16Test
-    : public ::testing::TestWithParam<
-          std::tr1::tuple<FdctFunc, vpx_bit_depth_t> > {
- public:
-  virtual ~PartialTrans16x16Test() {}
-  virtual void SetUp() {
-    fwd_txfm_ = GET_PARAM(0);
-    bit_depth_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  vpx_bit_depth_t bit_depth_;
-  FdctFunc fwd_txfm_;
-};
-
-TEST_P(PartialTrans16x16Test, Extremes) {
-#if CONFIG_VPX_HIGHBITDEPTH
-  const int16_t maxval =
-      static_cast<int16_t>(clip_pixel_highbd(1 << 30, bit_depth_));
-#else
-  const int16_t maxval = 255;
-#endif
-  const int minval = -maxval;
-  DECLARE_ALIGNED(16, int16_t, input[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output[kNumCoeffs]);
-
-  for (int i = 0; i < kNumCoeffs; ++i) input[i] = maxval;
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 16));
-  EXPECT_EQ((maxval * kNumCoeffs) >> 1, output[0]);
-
-  for (int i = 0; i < kNumCoeffs; ++i) input[i] = minval;
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 16));
-  EXPECT_EQ((minval * kNumCoeffs) >> 1, output[0]);
-}
-
-TEST_P(PartialTrans16x16Test, Random) {
-#if CONFIG_VPX_HIGHBITDEPTH
-  const int16_t maxval =
-      static_cast<int16_t>(clip_pixel_highbd(1 << 30, bit_depth_));
-#else
-  const int16_t maxval = 255;
-#endif
-  DECLARE_ALIGNED(16, int16_t, input[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output[kNumCoeffs]);
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-
-  int sum = 0;
-  for (int i = 0; i < kNumCoeffs; ++i) {
-    const int val = (i & 1) ? -rnd(maxval + 1) : rnd(maxval + 1);
-    input[i] = val;
-    sum += val;
-  }
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 16));
-  EXPECT_EQ(sum >> 1, output[0]);
-}
-
 using std::tr1::make_tuple;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct16x16_c, &idct16x16_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct16x16_c, &idct16x16_12, 0, VPX_BITS_12),
-        make_tuple(&vpx_fdct16x16_c, &vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
+        make_tuple(&vp9_highbd_fdct16x16_c, &idct16x16_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct16x16_c, &idct16x16_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_fdct16x16_c, &vp9_idct16x16_256_add_c, 0, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct16x16_c, &vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fdct16x16_c, &vp9_idct16x16_256_add_c, 0, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans16x16HT,
    ::testing::Values(
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_10, 0, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_10, 1, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_10, 2, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_10, 3, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_12, 0, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_12, 1, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_12, 2, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht16x16_c, &iht16x16_12, 3, VPX_BITS_12),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 3, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(
-    C, PartialTrans16x16Test,
-    ::testing::Values(make_tuple(&vpx_highbd_fdct16x16_1_c, VPX_BITS_8),
-                      make_tuple(&vpx_highbd_fdct16x16_1_c, VPX_BITS_10),
-                      make_tuple(&vpx_highbd_fdct16x16_1_c, VPX_BITS_12)));
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_10, 1, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_10, 2, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_10, 3, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_12, 1, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_12, 2, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_c, &iht16x16_12, 3, VPX_BITS_12),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 3, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans16x16HT,
    ::testing::Values(
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_c, &vp10_iht16x16_256_add_c, 3, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(C, PartialTrans16x16Test,
-                        ::testing::Values(make_tuple(&vpx_fdct16x16_1_c,
-                                                     VPX_BITS_8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_c, &vp9_iht16x16_256_add_c, 3, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if HAVE_NEON_ASM && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    NEON, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_neon, 0, VPX_BITS_8)));
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_neon, 0, VPX_BITS_8)));
 #endif

-#if HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct16x16_sse2,
-                   &vpx_idct16x16_256_add_sse2, 0, VPX_BITS_8)));
+        make_tuple(&vp9_fdct16x16_sse2,
+                   &vp9_idct16x16_256_add_sse2, 0, VPX_BITS_8)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans16x16HT,
    ::testing::Values(
-        make_tuple(&vp10_fht16x16_sse2, &vp10_iht16x16_256_add_sse2, 0,
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 0,
                   VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2, &vp10_iht16x16_256_add_sse2, 1,
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 1,
                   VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2, &vp10_iht16x16_256_add_sse2, 2,
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 2,
                   VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2, &vp10_iht16x16_256_add_sse2, 3,
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 3,
                   VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans16x16Test,
-                        ::testing::Values(make_tuple(&vpx_fdct16x16_1_sse2,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct16x16_sse2,
+        make_tuple(&vp9_highbd_fdct16x16_sse2,
                   &idct16x16_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct16x16_c,
+        make_tuple(&vp9_highbd_fdct16x16_c,
                   &idct16x16_256_add_10_sse2, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct16x16_sse2,
+        make_tuple(&vp9_highbd_fdct16x16_sse2,
                   &idct16x16_12, 0, VPX_BITS_12),
-        make_tuple(&vpx_highbd_fdct16x16_c,
+        make_tuple(&vp9_highbd_fdct16x16_c,
                   &idct16x16_256_add_12_sse2, 0, VPX_BITS_12),
-        make_tuple(&vpx_fdct16x16_sse2,
-                   &vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
+        make_tuple(&vp9_fdct16x16_sse2,
+                   &vp9_idct16x16_256_add_c, 0, VPX_BITS_8)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans16x16HT,
    ::testing::Values(
-        make_tuple(&vp10_fht16x16_sse2,
-                   &vp10_iht16x16_256_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2,
-                   &vp10_iht16x16_256_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2,
-                   &vp10_iht16x16_256_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_sse2, &vp10_iht16x16_256_add_c, 3,
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_10, 1, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_10, 2, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_10, 3, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_12, 1, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_12, 2, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht16x16_sse2, &iht16x16_12, 3, VPX_BITS_12),
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_c, 3,
                   VPX_BITS_8)));
 // Optimizations take effect at a threshold of 3155, so we use a value close to
 // that to test both branches.
@@ -963,30 +932,13 @@ INSTANTIATE_TEST_CASE_P(
                   &idct16x16_10_add_12_sse2, 3167, VPX_BITS_12),
        make_tuple(&idct16x16_12,
                   &idct16x16_256_add_12_sse2, 3167, VPX_BITS_12)));
-INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans16x16Test,
-                        ::testing::Values(make_tuple(&vpx_fdct16x16_1_sse2,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSSE3 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
-    MSA, Trans16x16DCT,
+    SSSE3, Trans16x16DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct16x16_msa,
-                   &vpx_idct16x16_256_add_msa, 0, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(
-    MSA, Trans16x16HT,
-    ::testing::Values(
-        make_tuple(&vp10_fht16x16_msa,
-                   &vp10_iht16x16_256_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_msa,
-                   &vp10_iht16x16_256_add_msa, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_msa,
-                   &vp10_iht16x16_256_add_msa, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht16x16_msa,
-                   &vp10_iht16x16_256_add_msa, 3, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(MSA, PartialTrans16x16Test,
-                        ::testing::Values(make_tuple(&vpx_fdct16x16_1_msa,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fdct16x16_c, &vp9_idct16x16_256_add_ssse3, 0,
+                   VPX_BITS_8)));
+#endif  // HAVE_SSSE3 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/dct32x32_test.cc
+++ b/test/dct32x32_test.cc
@@ -13,23 +13,28 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp10_rtcd.h"
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vp10/common/entropy.h"
+
+#include "./vpx_config.h"
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_entropy.h"
 #include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"
-#include "vpx_ports/msvc.h"  // for round()

 using libvpx_test::ACMRandom;

 namespace {
+#ifdef _MSC_VER
+static int round(double x) {
+  if (x < 0)
+    return static_cast<int>(ceil(x - 0.5));
+  else
+    return static_cast<int>(floor(x + 0.5));
+}
+#endif

 const int kNumCoeffs = 1024;
 const double kPi = 3.141592653589793238462643383279502884;
@@ -73,15 +78,19 @@ typedef void (*InvTxfmFunc)(const tran_low_t *in, uint8_t *out, int stride);
 typedef std::tr1::tuple<FwdTxfmFunc, InvTxfmFunc, int, vpx_bit_depth_t>
    Trans32x32Param;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
+void idct32x32_8(const tran_low_t *in, uint8_t *out, int stride) {
+  vp9_highbd_idct32x32_1024_add_c(in, out, stride, 8);
+}
+
 void idct32x32_10(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct32x32_1024_add_c(in, out, stride, 10);
+  vp9_highbd_idct32x32_1024_add_c(in, out, stride, 10);
 }

 void idct32x32_12(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct32x32_1024_add_c(in, out, stride, 12);
+  vp9_highbd_idct32x32_1024_add_c(in, out, stride, 12);
 }
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

 class Trans32x32Test : public ::testing::TestWithParam<Trans32x32Param> {
 public:
@@ -110,13 +119,13 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
  uint32_t max_error = 0;
  int64_t total_error = 0;
  const int count_test_block = 10000;
-  DECLARE_ALIGNED(16, int16_t, test_input_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, test_temp_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, src[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-  DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_temp_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
 #endif

  for (int i = 0; i < count_test_block; ++i) {
@@ -126,7 +135,7 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
        src[j] = rnd.Rand8();
        dst[j] = rnd.Rand8();
        test_input_block[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        src16[j] = rnd.Rand16() & mask_;
        dst16[j] = rnd.Rand16() & mask_;
@@ -138,7 +147,7 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
    ASM_REGISTER_STATE_CHECK(fwd_txfm_(test_input_block, test_temp_block, 32));
    if (bit_depth_ == VPX_BITS_8) {
      ASM_REGISTER_STATE_CHECK(inv_txfm_(test_temp_block, dst, 32));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    } else {
      ASM_REGISTER_STATE_CHECK(inv_txfm_(test_temp_block,
                                         CONVERT_TO_BYTEPTR(dst16), 32));
@@ -146,11 +155,11 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
    }

    for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
-      const int32_t diff =
+#if CONFIG_VP9_HIGHBITDEPTH
+      const uint32_t diff =
          bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
-      const int32_t diff = dst[j] - src[j];
+      const uint32_t diff = dst[j] - src[j];
 #endif
      const uint32_t error = diff * diff;
      if (max_error < error)
@@ -175,16 +184,16 @@ TEST_P(Trans32x32Test, CoeffCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = 1000;

-  DECLARE_ALIGNED(16, int16_t, input_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output_block[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);

  for (int i = 0; i < count_test_block; ++i) {
    for (int j = 0; j < kNumCoeffs; ++j)
      input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);

    const int stride = 32;
-    vpx_fdct32x32_c(input_block, output_ref_block, stride);
+    vp9_fdct32x32_c(input_block, output_ref_block, stride);
    ASM_REGISTER_STATE_CHECK(fwd_txfm_(input_block, output_block, stride));

    if (version_ == 0) {
@@ -203,13 +212,15 @@ TEST_P(Trans32x32Test, MemCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = 2000;

-  DECLARE_ALIGNED(16, int16_t, input_extreme_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output_block[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, input_extreme_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);

  for (int i = 0; i < count_test_block; ++i) {
    // Initialize a test block with input range [-mask_, mask_].
    for (int j = 0; j < kNumCoeffs; ++j) {
+      input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
      input_extreme_block[j] = rnd.Rand8() & 1 ? mask_ : -mask_;
    }
    if (i == 0) {
@@ -221,7 +232,7 @@ TEST_P(Trans32x32Test, MemCheck) {
    }

    const int stride = 32;
-    vpx_fdct32x32_c(input_extreme_block, output_ref_block, stride);
+    vp9_fdct32x32_c(input_extreme_block, output_ref_block, stride);
    ASM_REGISTER_STATE_CHECK(
        fwd_txfm_(input_extreme_block, output_block, stride));

@@ -246,13 +257,13 @@ TEST_P(Trans32x32Test, MemCheck) {
 TEST_P(Trans32x32Test, InverseAccuracy) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = 1000;
-  DECLARE_ALIGNED(16, int16_t, in[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, src[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-  DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, in, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
 #endif

  for (int i = 0; i < count_test_block; ++i) {
@@ -264,7 +275,7 @@ TEST_P(Trans32x32Test, InverseAccuracy) {
        src[j] = rnd.Rand8();
        dst[j] = rnd.Rand8();
        in[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        src16[j] = rnd.Rand16() & mask_;
        dst16[j] = rnd.Rand16() & mask_;
@@ -278,13 +289,13 @@ TEST_P(Trans32x32Test, InverseAccuracy) {
      coeff[j] = static_cast<tran_low_t>(round(out_r[j]));
    if (bit_depth_ == VPX_BITS_8) {
      ASM_REGISTER_STATE_CHECK(inv_txfm_(coeff, dst, 32));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    } else {
      ASM_REGISTER_STATE_CHECK(inv_txfm_(coeff, CONVERT_TO_BYTEPTR(dst16), 32));
 #endif
    }
    for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      const int diff =
          bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
@@ -298,165 +309,77 @@ TEST_P(Trans32x32Test, InverseAccuracy) {
  }
 }

-class PartialTrans32x32Test
-    : public ::testing::TestWithParam<
-          std::tr1::tuple<FwdTxfmFunc, vpx_bit_depth_t> > {
- public:
-  virtual ~PartialTrans32x32Test() {}
-  virtual void SetUp() {
-    fwd_txfm_ = GET_PARAM(0);
-    bit_depth_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  vpx_bit_depth_t bit_depth_;
-  FwdTxfmFunc fwd_txfm_;
-};
-
-TEST_P(PartialTrans32x32Test, Extremes) {
-#if CONFIG_VPX_HIGHBITDEPTH
-  const int16_t maxval =
-      static_cast<int16_t>(clip_pixel_highbd(1 << 30, bit_depth_));
-#else
-  const int16_t maxval = 255;
-#endif
-  const int minval = -maxval;
-  DECLARE_ALIGNED(16, int16_t, input[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output[kNumCoeffs]);
-
-  for (int i = 0; i < kNumCoeffs; ++i) input[i] = maxval;
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 32));
-  EXPECT_EQ((maxval * kNumCoeffs) >> 3, output[0]);
-
-  for (int i = 0; i < kNumCoeffs; ++i) input[i] = minval;
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 32));
-  EXPECT_EQ((minval * kNumCoeffs) >> 3, output[0]);
-}
-
-TEST_P(PartialTrans32x32Test, Random) {
-#if CONFIG_VPX_HIGHBITDEPTH
-  const int16_t maxval =
-      static_cast<int16_t>(clip_pixel_highbd(1 << 30, bit_depth_));
-#else
-  const int16_t maxval = 255;
-#endif
-  DECLARE_ALIGNED(16, int16_t, input[kNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output[kNumCoeffs]);
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-
-  int sum = 0;
-  for (int i = 0; i < kNumCoeffs; ++i) {
-    const int val = (i & 1) ? -rnd(maxval + 1) : rnd(maxval + 1);
-    input[i] = val;
-    sum += val;
-  }
-  output[0] = 0;
-  ASM_REGISTER_STATE_CHECK(fwd_txfm_(input, output, 32));
-  EXPECT_EQ(sum >> 3, output[0]);
-}
-
 using std::tr1::make_tuple;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct32x32_c,
+        make_tuple(&vp9_highbd_fdct32x32_c,
                   &idct32x32_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct32x32_rd_c,
+        make_tuple(&vp9_highbd_fdct32x32_rd_c,
                   &idct32x32_10, 1, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct32x32_c,
+        make_tuple(&vp9_highbd_fdct32x32_c,
                   &idct32x32_12, 0, VPX_BITS_12),
-        make_tuple(&vpx_highbd_fdct32x32_rd_c,
+        make_tuple(&vp9_highbd_fdct32x32_rd_c,
                   &idct32x32_12, 1, VPX_BITS_12),
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_c,
-                   &vpx_idct32x32_1024_add_c, 1, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(
-    C, PartialTrans32x32Test,
-    ::testing::Values(make_tuple(&vpx_highbd_fdct32x32_1_c, VPX_BITS_8),
-                      make_tuple(&vpx_highbd_fdct32x32_1_c, VPX_BITS_10),
-                      make_tuple(&vpx_highbd_fdct32x32_1_c, VPX_BITS_12)));
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fdct32x32_rd_c,
+                   &vp9_idct32x32_1024_add_c, 1, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_c,
-                   &vpx_idct32x32_1024_add_c, 1, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(C, PartialTrans32x32Test,
-                        ::testing::Values(make_tuple(&vpx_fdct32x32_1_c,
-                                                     VPX_BITS_8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fdct32x32_rd_c,
+                   &vp9_idct32x32_1024_add_c, 1, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    NEON, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_neon, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_c,
-                   &vpx_idct32x32_1024_add_neon, 1, VPX_BITS_8)));
-#endif  // HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_neon, 0, VPX_BITS_8),
+        make_tuple(&vp9_fdct32x32_rd_c,
+                   &vp9_idct32x32_1024_add_neon, 1, VPX_BITS_8)));
+#endif  // HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_sse2,
-                   &vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_sse2,
-                   &vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans32x32Test,
-                        ::testing::Values(make_tuple(&vpx_fdct32x32_1_sse2,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fdct32x32_sse2,
+                   &vp9_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
+        make_tuple(&vp9_fdct32x32_rd_sse2,
+                   &vp9_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
+#endif  // HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct32x32_sse2, &idct32x32_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct32x32_rd_sse2, &idct32x32_10, 1,
+        make_tuple(&vp9_highbd_fdct32x32_sse2, &idct32x32_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct32x32_rd_sse2, &idct32x32_10, 1,
                   VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct32x32_sse2, &idct32x32_12, 0, VPX_BITS_12),
-        make_tuple(&vpx_highbd_fdct32x32_rd_sse2, &idct32x32_12, 1,
+        make_tuple(&vp9_highbd_fdct32x32_sse2, &idct32x32_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fdct32x32_rd_sse2, &idct32x32_12, 1,
                   VPX_BITS_12),
-        make_tuple(&vpx_fdct32x32_sse2, &vpx_idct32x32_1024_add_c, 0,
+        make_tuple(&vp9_fdct32x32_sse2, &vp9_idct32x32_1024_add_c, 0,
                   VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_sse2, &vpx_idct32x32_1024_add_c, 1,
+        make_tuple(&vp9_fdct32x32_rd_sse2, &vp9_idct32x32_1024_add_c, 1,
                   VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans32x32Test,
-                        ::testing::Values(make_tuple(&vpx_fdct32x32_1_sse2,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_AVX2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_AVX2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    AVX2, Trans32x32Test,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_avx2,
-                   &vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_avx2,
-                   &vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
-#endif  // HAVE_AVX2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-
-#if HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    MSA, Trans32x32Test,
-    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_msa,
-                   &vpx_idct32x32_1024_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vpx_fdct32x32_rd_msa,
-                   &vpx_idct32x32_1024_add_msa, 1, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(MSA, PartialTrans32x32Test,
-                        ::testing::Values(make_tuple(&vpx_fdct32x32_1_msa,
-                                                     VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fdct32x32_avx2,
+                   &vp9_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
+        make_tuple(&vp9_fdct32x32_rd_avx2,
+                   &vp9_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
+#endif  // HAVE_AVX2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/decode_api_test.cc
+++ b/test/decode_api_test.cc
@@ -7,11 +7,10 @@
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */
-
 #include "third_party/googletest/src/include/gtest/gtest.h"

-#include "./vpx_config.h"
 #include "test/ivf_video_source.h"
+#include "./vpx_config.h"
 #include "vpx/vp8dx.h"
 #include "vpx/vpx_decoder.h"

@@ -21,8 +20,11 @@ namespace {

 TEST(DecodeAPI, InvalidParams) {
  static const vpx_codec_iface_t *kCodecs[] = {
-#if CONFIG_VP10_DECODER
-    &vpx_codec_vp10_dx_algo,
+#if CONFIG_VP8_DECODER
+    &vpx_codec_vp8_dx_algo,
+#endif
+#if CONFIG_VP9_DECODER
+    &vpx_codec_vp9_dx_algo,
 #endif
  };
  uint8_t buf[1] = {0};
@@ -55,4 +57,86 @@ TEST(DecodeAPI, InvalidParams) {
  }
 }

+#if CONFIG_VP8_DECODER
+TEST(DecodeAPI, OptionalParams) {
+  vpx_codec_ctx_t dec;
+
+#if CONFIG_ERROR_CONCEALMENT
+  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_dec_init(&dec, &vpx_codec_vp8_dx_algo, NULL,
+                                             VPX_CODEC_USE_ERROR_CONCEALMENT));
+#else
+  EXPECT_EQ(VPX_CODEC_INCAPABLE,
+            vpx_codec_dec_init(&dec, &vpx_codec_vp8_dx_algo, NULL,
+                               VPX_CODEC_USE_ERROR_CONCEALMENT));
+#endif  // CONFIG_ERROR_CONCEALMENT
+}
+#endif  // CONFIG_VP8_DECODER
+
+#if CONFIG_VP9_DECODER
+// Test VP9 codec controls after a decode error to ensure the code doesn't
+// misbehave.
+void TestVp9Controls(vpx_codec_ctx_t *dec) {
+  static const int kControls[] = {
+    VP8D_GET_LAST_REF_UPDATES,
+    VP8D_GET_FRAME_CORRUPTED,
+    VP9D_GET_DISPLAY_SIZE,
+    VP9D_GET_FRAME_SIZE
+  };
+  int val[2];
+
+  for (int i = 0; i < NELEMENTS(kControls); ++i) {
+    const vpx_codec_err_t res = vpx_codec_control_(dec, kControls[i], val);
+    switch (kControls[i]) {
+      case VP8D_GET_FRAME_CORRUPTED:
+        EXPECT_EQ(VPX_CODEC_ERROR, res) << kControls[i];
+        break;
+      default:
+        EXPECT_EQ(VPX_CODEC_OK, res) << kControls[i];
+        break;
+    }
+    EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
+              vpx_codec_control_(dec, kControls[i], NULL));
+  }
+
+  vp9_ref_frame_t ref;
+  ref.idx = 0;
+  EXPECT_EQ(VPX_CODEC_ERROR, vpx_codec_control(dec, VP9_GET_REFERENCE, &ref));
+  EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
+            vpx_codec_control(dec, VP9_GET_REFERENCE, NULL));
+
+  vpx_ref_frame_t ref_copy;
+  const int width = 352;
+  const int height = 288;
+  ASSERT_TRUE(
+      vpx_img_alloc(&ref_copy.img, VPX_IMG_FMT_I420, width, height, 1) != NULL);
+  ref_copy.frame_type = VP8_LAST_FRAME;
+  EXPECT_EQ(VPX_CODEC_ERROR,
+            vpx_codec_control(dec, VP8_COPY_REFERENCE, &ref_copy));
+  EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
+            vpx_codec_control(dec, VP8_COPY_REFERENCE, NULL));
+  vpx_img_free(&ref_copy.img);
+}
+
+TEST(DecodeAPI, Vp9InvalidDecode) {
+  const vpx_codec_iface_t *const codec = &vpx_codec_vp9_dx_algo;
+  const char filename[] =
+      "invalid-vp90-2-00-quantizer-00.webm.ivf.s5861_r01-05_b6-.v2.ivf";
+  libvpx_test::IVFVideoSource video(filename);
+  video.Init();
+  video.Begin();
+  ASSERT_TRUE(!HasFailure());
+
+  vpx_codec_ctx_t dec;
+  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_dec_init(&dec, codec, NULL, 0));
+  const uint32_t frame_size = static_cast<uint32_t>(video.frame_size());
+  EXPECT_EQ(VPX_CODEC_MEM_ERROR,
+            vpx_codec_decode(&dec, video.cxdata(), frame_size, NULL, 0));
+  vpx_codec_iter_t iter = NULL;
+  EXPECT_EQ(NULL, vpx_codec_get_frame(&dec, &iter));
+
+  TestVp9Controls(&dec);
+  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_destroy(&dec));
+}
+#endif  // CONFIG_VP9_DECODER
+
 }  // namespace
--- a/test/decode_perf_test.cc
+++ b/test/decode_perf_test.cc
@@ -268,6 +268,6 @@ TEST_P(VP9NewEncodeDecodePerfTest, PerfTest) {
  printf("}\n");
 }

-VP10_INSTANTIATE_TEST_CASE(
+VP9_INSTANTIATE_TEST_CASE(
  VP9NewEncodeDecodePerfTest, ::testing::Values(::libvpx_test::kTwoPassGood));
 }  // namespace
--- a/test/decode_test_driver.cc
+++ b/test/decode_test_driver.cc
@@ -7,18 +7,15 @@
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
 #include "test/codec_factory.h"
 #include "test/decode_test_driver.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
 #include "test/register_state_check.h"
 #include "test/video_source.h"

 namespace libvpx_test {

 const char kVP8Name[] = "WebM Project VP8";
-const char kVP10Name[] = "WebM Project VP10";

 vpx_codec_err_t Decoder::PeekStream(const uint8_t *cxdata, size_t size,
                                    vpx_codec_stream_info_t *stream_info) {
@@ -47,11 +44,6 @@ bool Decoder::IsVP8() const {
  return strncmp(kVP8Name, codec_name, sizeof(kVP8Name) - 1) == 0;
 }

-bool Decoder::IsVP10() const {
-  const char *codec_name = GetDecoderName();
-  return strncmp(kVP10Name, codec_name, sizeof(kVP10Name) - 1) == 0;
-}
-
 void DecoderTest::HandlePeekResult(Decoder *const decoder,
                                   CompressedVideoSource *video,
                                   const vpx_codec_err_t res_peek) {
--- a/test/decode_test_driver.h
+++ b/test/decode_test_driver.h
@@ -107,8 +107,6 @@ class Decoder {

  bool IsVP8() const;

-  bool IsVP10() const;
-
  vpx_codec_ctx_t * GetDecoder() {
    return &decoder_;
  }
--- a/test/encode_api_test.cc
+++ b/test/encode_api_test.cc
@@ -1,62 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "vpx/vp8cx.h"
-#include "vpx/vpx_encoder.h"
-
-namespace {
-
-#define NELEMENTS(x) static_cast<int>(sizeof(x) / sizeof(x[0]))
-
-TEST(EncodeAPI, InvalidParams) {
-  static const vpx_codec_iface_t *kCodecs[] = {
-#if CONFIG_VP10_ENCODER
-    &vpx_codec_vp10_cx_algo,
-#endif
-  };
-  uint8_t buf[1] = {0};
-  vpx_image_t img;
-  vpx_codec_ctx_t enc;
-  vpx_codec_enc_cfg_t cfg;
-
-  EXPECT_EQ(&img, vpx_img_wrap(&img, VPX_IMG_FMT_I420, 1, 1, 1, buf));
-
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_enc_init(NULL, NULL, NULL, 0));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_enc_init(&enc, NULL, NULL, 0));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_encode(NULL, NULL, 0, 0, 0, 0));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_encode(NULL, &img, 0, 0, 0, 0));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_destroy(NULL));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-            vpx_codec_enc_config_default(NULL, NULL, 0));
-  EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-            vpx_codec_enc_config_default(NULL, &cfg, 0));
-  EXPECT_TRUE(vpx_codec_error(NULL) != NULL);
-
-  for (int i = 0; i < NELEMENTS(kCodecs); ++i) {
-    SCOPED_TRACE(vpx_codec_iface_name(kCodecs[i]));
-    EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-              vpx_codec_enc_init(NULL, kCodecs[i], NULL, 0));
-    EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-              vpx_codec_enc_init(&enc, kCodecs[i], NULL, 0));
-    EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-              vpx_codec_enc_config_default(kCodecs[i], &cfg, 1));
-
-    EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_config_default(kCodecs[i], &cfg, 0));
-    EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_init(&enc, kCodecs[i], &cfg, 0));
-    EXPECT_EQ(VPX_CODEC_OK, vpx_codec_encode(&enc, NULL, 0, 0, 0, 0));
-
-    EXPECT_EQ(VPX_CODEC_OK, vpx_codec_destroy(&enc));
-  }
-}
-
-}  // namespace
--- a/test/encode_perf_test.cc
+++ b/test/encode_perf_test.cc
@@ -197,6 +197,6 @@ TEST_P(VP9EncodePerfTest, PerfTest) {
  }
 }

-VP10_INSTANTIATE_TEST_CASE(
+VP9_INSTANTIATE_TEST_CASE(
    VP9EncodePerfTest, ::testing::Values(::libvpx_test::kRealTime));
 }  // namespace
--- a/test/encode_test_driver.cc
+++ b/test/encode_test_driver.cc
@@ -10,15 +10,13 @@

 #include <string>

-#include "third_party/googletest/src/include/gtest/gtest.h"
-
 #include "./vpx_config.h"
-#include "vpx_ports/mem.h"
 #include "test/codec_factory.h"
-#include "test/decode_test_driver.h"
 #include "test/encode_test_driver.h"
+#include "test/decode_test_driver.h"
 #include "test/register_state_check.h"
 #include "test/video_source.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"

 namespace libvpx_test {
 void Encoder::InitEncoder(VideoSource *video) {
@@ -31,23 +29,26 @@ void Encoder::InitEncoder(VideoSource *video) {
    cfg_.g_timebase = video->timebase();
    cfg_.rc_twopass_stats_in = stats_->buf();

+    // Default to 1 thread.
+    cfg_.g_threads = 1;
    res = vpx_codec_enc_init(&encoder_, CodecInterface(), &cfg_,
                             init_flags_);
    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();

-#if CONFIG_VP10_ENCODER
-    if (CodecInterface() == &vpx_codec_vp10_cx_algo) {
-      // Default to 1 tile column for VP10. With CONFIG_EXT_TILE, the
-      // default is already the largest possible tile size
-#if !CONFIG_EXT_TILE
+#if CONFIG_VP9_ENCODER
+    if (CodecInterface() == &vpx_codec_vp9_cx_algo) {
+      // Default to 1 tile column for VP9.
      const int log2_tile_columns = 0;
      res = vpx_codec_control_(&encoder_, VP9E_SET_TILE_COLUMNS,
                               log2_tile_columns);
      ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
-#endif  // !CONFIG_EXT_TILE
    } else
 #endif
    {
+#if CONFIG_VP8_ENCODER
+      ASSERT_EQ(&vpx_codec_vp8_cx_algo, CodecInterface())
+          << "Unknown Codec Interface";
+#endif
    }
  }
 }
@@ -129,120 +130,38 @@ void EncoderTest::SetMode(TestMode mode) {
  else
    passes_ = 1;
 }
-
-static bool compare_plane(const uint8_t *const buf1, const int stride1,
-                          const uint8_t *const buf2, const int stride2,
-                          const int w, const int h,
-                          int *const mismatch_row,
-                          int *const mismatch_col,
-                          int *const mismatch_pix1,
-                          int *const mismatch_pix2) {
-  int r, c;
-
-  for (r = 0; r < h; ++r) {
-    for (c = 0; c < w; ++c) {
-      const int pix1 = buf1[r * stride1 + c];
-      const int pix2 = buf2[r * stride2 + c];
-
-      if (pix1 != pix2) {
-        if (mismatch_row != NULL)
-          *mismatch_row = r;
-        if (mismatch_col != NULL)
-          *mismatch_col = c;
-        if (mismatch_pix1 != NULL)
-          *mismatch_pix1 = pix1;
-        if (mismatch_pix2 != NULL)
-          *mismatch_pix2 = pix2;
-        return false;
-      }
-    }
-  }
-
-  return true;
-}
-
 // The function should return "true" most of the time, therefore no early
 // break-out is implemented within the match checking process.
 static bool compare_img(const vpx_image_t *img1,
-                        const vpx_image_t *img2,
-                        int *const mismatch_row,
-                        int *const mismatch_col,
-                        int *const mismatch_plane,
-                        int *const mismatch_pix1,
-                        int *const mismatch_pix2) {
+                        const vpx_image_t *img2) {
+  bool match = (img1->fmt == img2->fmt) &&
+               (img1->cs == img2->cs) &&
+               (img1->d_w == img2->d_w) &&
+               (img1->d_h == img2->d_h);

-  const unsigned int w_y = img1->d_w;
-  const unsigned int h_y = img1->d_h;
-  const unsigned int w_uv = ROUND_POWER_OF_TWO(w_y, img1->x_chroma_shift);
-  const unsigned int h_uv = ROUND_POWER_OF_TWO(h_y, img1->y_chroma_shift);
-
-  if (img1->fmt != img2->fmt
-      || img1->cs != img2->cs
-      || img1->d_w != img2->d_w
-      || img1->d_h != img2->d_h) {
-    if (mismatch_row != NULL)
-      *mismatch_row = -1;
-    if (mismatch_col != NULL)
-      *mismatch_col = -1;
-    return false;
-  }
-
-  if (!compare_plane(img1->planes[VPX_PLANE_Y],  img1->stride[VPX_PLANE_Y],
-                     img2->planes[VPX_PLANE_Y],  img2->stride[VPX_PLANE_Y],
-                     w_y, h_y,
-                     mismatch_row, mismatch_col,
-                     mismatch_pix1, mismatch_pix2)) {
-    if (mismatch_plane != NULL)
-      *mismatch_plane = VPX_PLANE_Y;
-    return false;
-  }
-
-  if (!compare_plane(img1->planes[VPX_PLANE_U],  img1->stride[VPX_PLANE_U],
-                     img2->planes[VPX_PLANE_U],  img2->stride[VPX_PLANE_U],
-                     w_uv, h_uv,
-                     mismatch_row, mismatch_col,
-                     mismatch_pix1, mismatch_pix2)) {
-    if (mismatch_plane != NULL)
-      *mismatch_plane = VPX_PLANE_U;
-    return false;
-  }
-
-  if (!compare_plane(img1->planes[VPX_PLANE_V],  img1->stride[VPX_PLANE_V],
-                     img2->planes[VPX_PLANE_V],  img2->stride[VPX_PLANE_V],
-                     w_uv, h_uv,
-                     mismatch_row, mismatch_col,
-                     mismatch_pix1, mismatch_pix2)) {
-    if (mismatch_plane != NULL)
-      *mismatch_plane = VPX_PLANE_U;
-    return false;
-  }
-
-  return true;
+  const unsigned int width_y  = img1->d_w;
+  const unsigned int height_y = img1->d_h;
+  unsigned int i;
+  for (i = 0; i < height_y; ++i)
+    match = (memcmp(img1->planes[VPX_PLANE_Y] + i * img1->stride[VPX_PLANE_Y],
+                    img2->planes[VPX_PLANE_Y] + i * img2->stride[VPX_PLANE_Y],
+                    width_y) == 0) && match;
+  const unsigned int width_uv  = (img1->d_w + 1) >> 1;
+  const unsigned int height_uv = (img1->d_h + 1) >> 1;
+  for (i = 0; i <  height_uv; ++i)
+    match = (memcmp(img1->planes[VPX_PLANE_U] + i * img1->stride[VPX_PLANE_U],
+                    img2->planes[VPX_PLANE_U] + i * img2->stride[VPX_PLANE_U],
+                    width_uv) == 0) && match;
+  for (i = 0; i < height_uv; ++i)
+    match = (memcmp(img1->planes[VPX_PLANE_V] + i * img1->stride[VPX_PLANE_V],
+                    img2->planes[VPX_PLANE_V] + i * img2->stride[VPX_PLANE_V],
+                    width_uv) == 0) && match;
+  return match;
 }

-void EncoderTest::MismatchHook(const vpx_image_t* img_enc,
-                               const vpx_image_t* img_dec) {
-  int mismatch_row = 0;
-  int mismatch_col = 0;
-  int mismatch_plane = 0;
-  int mismatch_pix_enc = 0;
-  int mismatch_pix_dec = 0;
-
-  ASSERT_FALSE(compare_img(img_enc, img_dec,
-                           &mismatch_row, &mismatch_col,
-                           &mismatch_plane,
-                           &mismatch_pix_enc,
-                           &mismatch_pix_dec));
-
-  GTEST_FAIL()
-    << "Encode/Decode mismatch found:"
-    << std::endl
-    << "  pixel value enc/dec: "  << mismatch_pix_enc << "/" << mismatch_pix_dec
-    << std::endl
-    << "                plane: " << mismatch_plane
-    << std::endl
-    << "              row/col: " << mismatch_row << "/" << mismatch_col
-    << std::endl;
+void EncoderTest::MismatchHook(const vpx_image_t* /*img1*/,
+                               const vpx_image_t* /*img2*/) {
+  ASSERT_TRUE(0) << "Encode/Decode mismatch found";
 }

 void EncoderTest::RunLoop(VideoSource *video) {
@@ -262,36 +181,25 @@ void EncoderTest::RunLoop(VideoSource *video) {
      cfg_.g_pass = VPX_RC_LAST_PASS;

    BeginPassHook(pass);
-    testing::internal::scoped_ptr<Encoder> encoder(
-        codec_->CreateEncoder(cfg_, deadline_, init_flags_, &stats_));
-    ASSERT_TRUE(encoder.get() != NULL);
+    Encoder* const encoder = codec_->CreateEncoder(cfg_, deadline_, init_flags_,
+                                                   &stats_);
+    ASSERT_TRUE(encoder != NULL);

-    ASSERT_NO_FATAL_FAILURE(video->Begin());
+    video->Begin();
    encoder->InitEncoder(video);
-    ASSERT_FALSE(::testing::Test::HasFatalFailure());

    unsigned long dec_init_flags = 0;  // NOLINT
    // Use fragment decoder if encoder outputs partitions.
    // NOTE: fragment decoder and partition encoder are only supported by VP8.
    if (init_flags_ & VPX_CODEC_USE_OUTPUT_PARTITION)
      dec_init_flags |= VPX_CODEC_USE_INPUT_FRAGMENTS;
-    testing::internal::scoped_ptr<Decoder> decoder(
-        codec_->CreateDecoder(dec_cfg, dec_init_flags, 0));
-#if CONFIG_VP10 && CONFIG_EXT_TILE
-    if (decoder->IsVP10()) {
-      // Set dec_cfg.tile_row = -1 and dec_cfg.tile_col = -1 so that the whole
-      // frame is decoded.
-      decoder->Control(VP10_SET_DECODE_TILE_ROW, -1);
-      decoder->Control(VP10_SET_DECODE_TILE_COL, -1);
-    }
-#endif
-
+    Decoder* const decoder = codec_->CreateDecoder(dec_cfg, dec_init_flags, 0);
    bool again;
    for (again = true; again; video->Next()) {
      again = (video->img() != NULL);

      PreEncodeFrameHook(video);
-      PreEncodeFrameHook(video, encoder.get());
+      PreEncodeFrameHook(video, encoder);
      encoder->EncodeFrame(video, frame_flags_);

      CxDataIterator iter = encoder->GetCxData();
@@ -304,11 +212,11 @@ void EncoderTest::RunLoop(VideoSource *video) {
        switch (pkt->kind) {
          case VPX_CODEC_CX_FRAME_PKT:
            has_cxdata = true;
-            if (decoder.get() != NULL && DoDecode()) {
+            if (decoder && DoDecode()) {
              vpx_codec_err_t res_dec = decoder->DecodeFrame(
                  (const uint8_t*)pkt->data.frame.buf, pkt->data.frame.sz);

-              if (!HandleDecodeResult(res_dec, *video, decoder.get()))
+              if (!HandleDecodeResult(res_dec, *video, decoder))
                break;

              has_dxdata = true;
@@ -330,7 +238,7 @@ void EncoderTest::RunLoop(VideoSource *video) {
      // Flush the decoder when there are no more fragments.
      if ((init_flags_ & VPX_CODEC_USE_OUTPUT_PARTITION) && has_dxdata) {
        const vpx_codec_err_t res_dec = decoder->DecodeFrame(NULL, 0);
-        if (!HandleDecodeResult(res_dec, *video, decoder.get()))
+        if (!HandleDecodeResult(res_dec, *video, decoder))
          break;
      }

@@ -339,8 +247,7 @@ void EncoderTest::RunLoop(VideoSource *video) {
        DxDataIterator dec_iter = decoder->GetDxData();
        const vpx_image_t *img_dec = dec_iter.Next();
        if (img_enc && img_dec) {
-          const bool res = compare_img(img_enc, img_dec,
-                                       NULL, NULL, NULL, NULL, NULL);
+          const bool res = compare_img(img_enc, img_dec);
          if (!res) {  // Mismatch
            MismatchHook(img_enc, img_dec);
          }
@@ -354,6 +261,10 @@ void EncoderTest::RunLoop(VideoSource *video) {

    EndPassHook();

+    if (decoder)
+      delete decoder;
+    delete encoder;
+
    if (!Continue())
      break;
  }
--- a/test/encode_test_driver.h
+++ b/test/encode_test_driver.h
@@ -13,13 +13,12 @@
 #include <string>
 #include <vector>

-#include "third_party/googletest/src/include/gtest/gtest.h"
-
 #include "./vpx_config.h"
-#if CONFIG_VP10_ENCODER
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "vpx/vpx_encoder.h"
+#if CONFIG_VP8_ENCODER || CONFIG_VP9_ENCODER
 #include "vpx/vp8cx.h"
 #endif
-#include "vpx/vpx_encoder.h"

 namespace libvpx_test {

@@ -124,11 +123,6 @@ class Encoder {
    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
  }

-  void Control(int ctrl_id, int *arg) {
-    const vpx_codec_err_t res = vpx_codec_control_(&encoder_, ctrl_id, arg);
-    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
-  }
-
  void Control(int ctrl_id, struct vpx_scaling_mode *arg) {
    const vpx_codec_err_t res = vpx_codec_control_(&encoder_, ctrl_id, arg);
    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
@@ -139,11 +133,7 @@ class Encoder {
    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
  }

-  void Control(int ctrl_id, struct vpx_svc_parameters *arg) {
-    const vpx_codec_err_t res = vpx_codec_control_(&encoder_, ctrl_id, arg);
-    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
-  }
-#if CONFIG_VP10_ENCODER
+#if CONFIG_VP8_ENCODER || CONFIG_VP9_ENCODER
  void Control(int ctrl_id, vpx_active_map_t *arg) {
    const vpx_codec_err_t res = vpx_codec_control_(&encoder_, ctrl_id, arg);
    ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
@@ -193,10 +183,7 @@ class EncoderTest {
 protected:
  explicit EncoderTest(const CodecFactory *codec)
      : codec_(codec), abort_(false), init_flags_(0), frame_flags_(0),
-        last_pts_(0) {
-    // Default to 1 thread.
-    cfg_.g_threads = 1;
-  }
+        last_pts_(0) {}

  virtual ~EncoderTest() {}

--- a/test/encoder_parms_get_to_decoder.cc
+++ b/test/encoder_parms_get_to_decoder.cc
@@ -1,149 +0,0 @@
-/*
- *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/util.h"
-#include "test/y4m_video_source.h"
-#include "vp10/vp10_dx_iface.c"
-
-namespace {
-
-const int kCpuUsed = 2;
-
-struct EncodePerfTestVideo {
-  const char *name;
-  uint32_t width;
-  uint32_t height;
-  uint32_t bitrate;
-  int frames;
-};
-
-const EncodePerfTestVideo kVP9EncodePerfTestVectors[] = {
-  {"niklas_1280_720_30.y4m", 1280, 720, 600, 10},
-};
-
-struct EncodeParameters {
-  int32_t tile_rows;
-  int32_t tile_cols;
-  int32_t lossless;
-  int32_t error_resilient;
-  int32_t frame_parallel;
-  vpx_color_range_t color_range;
-  vpx_color_space_t cs;
-  int render_size[2];
-  // TODO(JBB): quantizers / bitrate
-};
-
-const EncodeParameters kVP9EncodeParameterSet[] = {
-  {0, 0, 0, 1, 0, VPX_CR_STUDIO_RANGE, VPX_CS_BT_601, { 0, 0 }},
-  {0, 0, 0, 0, 0, VPX_CR_FULL_RANGE, VPX_CS_BT_709, { 0, 0 }},
-  {0, 0, 1, 0, 0, VPX_CR_FULL_RANGE, VPX_CS_BT_2020, { 0, 0 }},
-  {0, 2, 0, 0, 1, VPX_CR_STUDIO_RANGE, VPX_CS_UNKNOWN, { 640, 480 }},
-  // TODO(JBB): Test profiles (requires more work).
-};
-
-class VpxEncoderParmsGetToDecoder
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<EncodeParameters,
-                                                 EncodePerfTestVideo> {
- protected:
-  VpxEncoderParmsGetToDecoder()
-      : EncoderTest(GET_PARAM(0)), encode_parms(GET_PARAM(1)) {}
-
-  virtual ~VpxEncoderParmsGetToDecoder() {}
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(::libvpx_test::kTwoPassGood);
-    cfg_.g_lag_in_frames = 25;
-    cfg_.g_error_resilient = encode_parms.error_resilient;
-    dec_cfg_.threads = 4;
-    test_video_ = GET_PARAM(2);
-    cfg_.rc_target_bitrate = test_video_.bitrate;
-  }
-
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (video->frame() == 1) {
-      encoder->Control(VP9E_SET_COLOR_SPACE, encode_parms.cs);
-      encoder->Control(VP9E_SET_COLOR_RANGE, encode_parms.color_range);
-      encoder->Control(VP9E_SET_LOSSLESS, encode_parms.lossless);
-      encoder->Control(VP9E_SET_FRAME_PARALLEL_DECODING,
-                       encode_parms.frame_parallel);
-      encoder->Control(VP9E_SET_TILE_ROWS, encode_parms.tile_rows);
-      encoder->Control(VP9E_SET_TILE_COLUMNS, encode_parms.tile_cols);
-      encoder->Control(VP8E_SET_CPUUSED, kCpuUsed);
-      encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
-      encoder->Control(VP8E_SET_ARNR_MAXFRAMES, 7);
-      encoder->Control(VP8E_SET_ARNR_STRENGTH, 5);
-      encoder->Control(VP8E_SET_ARNR_TYPE, 3);
-      if (encode_parms.render_size[0] > 0 && encode_parms.render_size[1] > 0)
-        encoder->Control(VP9E_SET_RENDER_SIZE, encode_parms.render_size);
-    }
-  }
-
-  virtual bool HandleDecodeResult(const vpx_codec_err_t res_dec,
-                                  const libvpx_test::VideoSource & /*video*/,
-                                  libvpx_test::Decoder *decoder) {
-    vpx_codec_ctx_t *const vp9_decoder = decoder->GetDecoder();
-    vpx_codec_alg_priv_t *const priv =
-        reinterpret_cast<vpx_codec_alg_priv_t *>(vp9_decoder->priv);
-    FrameWorkerData *const worker_data =
-        reinterpret_cast<FrameWorkerData *>(priv->frame_workers[0].data1);
-    VP10_COMMON *const common = &worker_data->pbi->common;
-
-    if (encode_parms.lossless) {
-      EXPECT_EQ(0, common->base_qindex);
-      EXPECT_EQ(0, common->y_dc_delta_q);
-      EXPECT_EQ(0, common->uv_dc_delta_q);
-      EXPECT_EQ(0, common->uv_ac_delta_q);
-      EXPECT_EQ(ONLY_4X4, common->tx_mode);
-    }
-    EXPECT_EQ(encode_parms.error_resilient, common->error_resilient_mode);
-    if (encode_parms.error_resilient) {
-      EXPECT_EQ(0, common->use_prev_frame_mvs);
-    }
-    EXPECT_EQ(encode_parms.color_range, common->color_range);
-    EXPECT_EQ(encode_parms.cs, common->color_space);
-    if (encode_parms.render_size[0] > 0 && encode_parms.render_size[1] > 0) {
-      EXPECT_EQ(encode_parms.render_size[0], common->render_width);
-      EXPECT_EQ(encode_parms.render_size[1], common->render_height);
-    }
-    EXPECT_EQ(encode_parms.tile_cols, common->log2_tile_cols);
-    EXPECT_EQ(encode_parms.tile_rows, common->log2_tile_rows);
-
-    EXPECT_EQ(VPX_CODEC_OK, res_dec) << decoder->DecodeError();
-    return VPX_CODEC_OK == res_dec;
-  }
-
-  EncodePerfTestVideo test_video_;
-
- private:
-  EncodeParameters encode_parms;
-};
-
-TEST_P(VpxEncoderParmsGetToDecoder, BitstreamParms) {
-  init_flags_ = VPX_CODEC_USE_PSNR;
-
-  libvpx_test::VideoSource *const video =
-      new libvpx_test::Y4mVideoSource(test_video_.name, 0, test_video_.frames);
-  ASSERT_TRUE(video != NULL);
-
-  ASSERT_NO_FATAL_FAILURE(RunLoop(video));
-  delete video;
-}
-
-VP10_INSTANTIATE_TEST_CASE(VpxEncoderParmsGetToDecoder,
-                          ::testing::ValuesIn(kVP9EncodeParameterSet),
-                          ::testing::ValuesIn(kVP9EncodePerfTestVectors));
-}  // namespace
--- a/test/error_resilience_test.cc
+++ b/test/error_resilience_test.cc
@@ -20,11 +20,10 @@ const int kMaxErrorFrames = 12;
 const int kMaxDroppableFrames = 12;

 class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
-    public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, bool> {
+    public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
 protected:
  ErrorResilienceTestLarge()
      : EncoderTest(GET_PARAM(0)),
-        svc_support_(GET_PARAM(2)),
        psnr_(0.0),
        nframes_(0),
        mismatch_psnr_(0.0),
@@ -100,7 +99,7 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
  }

  virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder * /*encoder*/) {
+                                  ::libvpx_test::Encoder *encoder) {
    frame_flags_ &= ~(VP8_EFLAG_NO_UPD_LAST |
                      VP8_EFLAG_NO_UPD_GF |
                      VP8_EFLAG_NO_UPD_ARF);
@@ -164,7 +163,6 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
    mismatch_psnr_ += mismatch_psnr;
    ++mismatch_nframes_;
    // std::cout << "Mismatch frame psnr: " << mismatch_psnr << "\n";
-    ::libvpx_test::EncoderTest::MismatchHook(img1, img2);
  }

  void SetErrorFrames(int num, unsigned int *list) {
@@ -195,8 +193,6 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
     pattern_switch_ = frame_switch;
   }

-  bool svc_support_;
-
 private:
  double psnr_;
  unsigned int nframes_;
@@ -306,10 +302,6 @@ TEST_P(ErrorResilienceTestLarge, DropFramesWithoutRecovery) {
 // two layer temporal pattern. The base layer does not predict from the top
 // layer, so successful decoding is expected.
 TEST_P(ErrorResilienceTestLarge, 2LayersDropEnhancement) {
-  // This test doesn't run if SVC is not supported.
-  if (!svc_support_)
-    return;
-
  const vpx_rational timebase = { 33333333, 1000000000 };
  cfg_.g_timebase = timebase;
  cfg_.rc_target_bitrate = 500;
@@ -355,10 +347,6 @@ TEST_P(ErrorResilienceTestLarge, 2LayersDropEnhancement) {
 // for a two layer temporal pattern, where at some point in the
 // sequence, the LAST ref is not used anymore.
 TEST_P(ErrorResilienceTestLarge, 2LayersNoRefLast) {
-  // This test doesn't run if SVC is not supported.
-  if (!svc_support_)
-    return;
-
  const vpx_rational timebase = { 33333333, 1000000000 };
  cfg_.g_timebase = timebase;
  cfg_.rc_target_bitrate = 500;
@@ -591,7 +579,8 @@ TEST_P(ErrorResilienceTestLargeCodecControls, CodecControl3TemporalLayers) {
  }
 }

-// SVC-related tests don't run for VP10 since SVC is not supported.
-VP10_INSTANTIATE_TEST_CASE(ErrorResilienceTestLarge, ONE_PASS_TEST_MODES,
-                           ::testing::Values(false));
+VP8_INSTANTIATE_TEST_CASE(ErrorResilienceTestLarge, ONE_PASS_TEST_MODES);
+VP8_INSTANTIATE_TEST_CASE(ErrorResilienceTestLargeCodecControls,
+                          ONE_PASS_TEST_MODES);
+VP9_INSTANTIATE_TEST_CASE(ErrorResilienceTestLarge, ONE_PASS_TEST_MODES);
 }  // namespace
--- a/test/ethread_test.cc
+++ b/test/ethread_test.cc
@@ -1,192 +0,0 @@
-/*
- *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <string>
-#include <vector>
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/md5_helper.h"
-#include "test/util.h"
-#include "test/y4m_video_source.h"
-
-namespace {
-class VPxEncoderThreadTest
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- protected:
-  VPxEncoderThreadTest()
-      : EncoderTest(GET_PARAM(0)),
-        encoder_initialized_(false),
-        encoding_mode_(GET_PARAM(1)),
-        set_cpu_used_(GET_PARAM(2)) {
-    init_flags_ = VPX_CODEC_USE_PSNR;
-    vpx_codec_dec_cfg_t cfg = vpx_codec_dec_cfg_t();
-    cfg.w = 1280;
-    cfg.h = 720;
-    decoder_ = codec_->CreateDecoder(cfg, 0);
-#if CONFIG_VP10 && CONFIG_EXT_TILE
-    if (decoder_->IsVP10()) {
-      decoder_->Control(VP10_SET_DECODE_TILE_ROW, -1);
-      decoder_->Control(VP10_SET_DECODE_TILE_COL, -1);
-    }
-#endif
-
-    size_enc_.clear();
-    md5_dec_.clear();
-    md5_enc_.clear();
-  }
-  virtual ~VPxEncoderThreadTest() {
-    delete decoder_;
-  }
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(encoding_mode_);
-
-    if (encoding_mode_ != ::libvpx_test::kRealTime) {
-      cfg_.g_lag_in_frames = 3;
-      cfg_.rc_end_usage = VPX_VBR;
-      cfg_.rc_2pass_vbr_minsection_pct = 5;
-      cfg_.rc_2pass_vbr_maxsection_pct = 2000;
-    } else {
-      cfg_.g_lag_in_frames = 0;
-      cfg_.rc_end_usage = VPX_CBR;
-      cfg_.g_error_resilient = 1;
-    }
-    cfg_.rc_max_quantizer = 56;
-    cfg_.rc_min_quantizer = 0;
-  }
-
-  virtual void BeginPassHook(unsigned int /*pass*/) {
-    encoder_initialized_ = false;
-  }
-
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource * /*video*/,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (!encoder_initialized_) {
-#if CONFIG_VP10 && CONFIG_EXT_TILE
-      encoder->Control(VP9E_SET_TILE_COLUMNS, 1);
-      if (codec_ == &libvpx_test::kVP10) {
-        // TODO(geza): Start using multiple tile rows when the multi-threaded
-        // encoder can handle them
-        encoder->Control(VP9E_SET_TILE_ROWS, 32);
-      } else {
-        encoder->Control(VP9E_SET_TILE_ROWS, 0);
-      }
-#else
-      // Encode 4 tile columns.
-      encoder->Control(VP9E_SET_TILE_COLUMNS, 2);
-      encoder->Control(VP9E_SET_TILE_ROWS, 0);
-#endif  // CONFIG_VP10 && CONFIG_EXT_TILE
-      encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
-      if (encoding_mode_ != ::libvpx_test::kRealTime) {
-        encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
-        encoder->Control(VP8E_SET_ARNR_MAXFRAMES, 7);
-        encoder->Control(VP8E_SET_ARNR_STRENGTH, 5);
-        encoder->Control(VP8E_SET_ARNR_TYPE, 3);
-      } else {
-        encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 0);
-        encoder->Control(VP9E_SET_AQ_MODE, 3);
-      }
-      encoder_initialized_ = true;
-    }
-  }
-
-  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
-    size_enc_.push_back(pkt->data.frame.sz);
-
-    ::libvpx_test::MD5 md5_enc;
-    md5_enc.Add(reinterpret_cast<uint8_t*>(pkt->data.frame.buf),
-                pkt->data.frame.sz);
-    md5_enc_.push_back(md5_enc.Get());
-
-    const vpx_codec_err_t res = decoder_->DecodeFrame(
-        reinterpret_cast<uint8_t*>(pkt->data.frame.buf), pkt->data.frame.sz);
-    if (res != VPX_CODEC_OK) {
-      abort_ = true;
-      ASSERT_EQ(VPX_CODEC_OK, res);
-    }
-    const vpx_image_t *img = decoder_->GetDxData().Next();
-
-    if (img) {
-      ::libvpx_test::MD5 md5_res;
-      md5_res.Add(img);
-      md5_dec_.push_back(md5_res.Get());
-    }
-  }
-
-  void DoTest() {
-    ::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 15, 18);
-    cfg_.rc_target_bitrate = 1000;
-
-    // Encode using single thread.
-    cfg_.g_threads = 1;
-    init_flags_ = VPX_CODEC_USE_PSNR;
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    std::vector<size_t> single_thr_size_enc;
-    std::vector<std::string> single_thr_md5_enc;
-    std::vector<std::string> single_thr_md5_dec;
-    single_thr_size_enc = size_enc_;
-    single_thr_md5_enc = md5_enc_;
-    single_thr_md5_dec = md5_dec_;
-    size_enc_.clear();
-    md5_enc_.clear();
-    md5_dec_.clear();
-
-    // Encode using multiple threads.
-    cfg_.g_threads = 4;
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    std::vector<size_t> multi_thr_size_enc;
-    std::vector<std::string> multi_thr_md5_enc;
-    std::vector<std::string> multi_thr_md5_dec;
-    multi_thr_size_enc = size_enc_;
-    multi_thr_md5_enc = md5_enc_;
-    multi_thr_md5_dec = md5_dec_;
-    size_enc_.clear();
-    md5_enc_.clear();
-    md5_dec_.clear();
-
-    // Check that the vectors are equal.
-    ASSERT_EQ(single_thr_size_enc, multi_thr_size_enc);
-    ASSERT_EQ(single_thr_md5_enc, multi_thr_md5_enc);
-    ASSERT_EQ(single_thr_md5_dec, multi_thr_md5_dec);
-  }
-
-  bool encoder_initialized_;
-  ::libvpx_test::TestMode encoding_mode_;
-  int set_cpu_used_;
-  ::libvpx_test::Decoder *decoder_;
-  std::vector<size_t> size_enc_;
-  std::vector<std::string> md5_enc_;
-  std::vector<std::string> md5_dec_;
-};
-
-TEST_P(VPxEncoderThreadTest, EncoderResultTest) {
-  DoTest();
-}
-
-class VPxEncoderThreadTestLarge : public VPxEncoderThreadTest {};
-
-TEST_P(VPxEncoderThreadTestLarge, EncoderResultTest) {
-  DoTest();
-}
-
-
-VP10_INSTANTIATE_TEST_CASE(
-    VPxEncoderThreadTest,
-    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood),
-    ::testing::Range(3, 9));
-
-VP10_INSTANTIATE_TEST_CASE(
-    VPxEncoderThreadTestLarge,
-    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood),
-    ::testing::Range(1, 3));
-}  // namespace
--- a/test/external_frame_buffer_test.cc
+++ b/test/external_frame_buffer_test.cc
@@ -0,0 +1,488 @@
+/*
+ *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include <string>
+
+#include "./vpx_config.h"
+#include "test/codec_factory.h"
+#include "test/decode_test_driver.h"
+#include "test/ivf_video_source.h"
+#include "test/md5_helper.h"
+#include "test/test_vectors.h"
+#include "test/util.h"
+#if CONFIG_WEBM_IO
+#include "test/webm_video_source.h"
+#endif
+
+namespace {
+
+const int kVideoNameParam = 1;
+const char kVP9TestFile[] = "vp90-2-02-size-lf-1920x1080.webm";
+
+struct ExternalFrameBuffer {
+  uint8_t *data;
+  size_t size;
+  int in_use;
+};
+
+// Class to manipulate a list of external frame buffers.
+class ExternalFrameBufferList {
+ public:
+  ExternalFrameBufferList()
+      : num_buffers_(0),
+        ext_fb_list_(NULL) {}
+
+  virtual ~ExternalFrameBufferList() {
+    for (int i = 0; i < num_buffers_; ++i) {
+      delete [] ext_fb_list_[i].data;
+    }
+    delete [] ext_fb_list_;
+  }
+
+  // Creates the list to hold the external buffers. Returns true on success.
+  bool CreateBufferList(int num_buffers) {
+    if (num_buffers < 0)
+      return false;
+
+    num_buffers_ = num_buffers;
+    ext_fb_list_ = new ExternalFrameBuffer[num_buffers_];
+    EXPECT_TRUE(ext_fb_list_ != NULL);
+    memset(ext_fb_list_, 0, sizeof(ext_fb_list_[0]) * num_buffers_);
+    return true;
+  }
+
+  // Searches the frame buffer list for a free frame buffer. Makes sure
+  // that the frame buffer is at least |min_size| in bytes. Marks that the
+  // frame buffer is in use by libvpx. Finally sets |fb| to point to the
+  // external frame buffer. Returns < 0 on an error.
+  int GetFreeFrameBuffer(size_t min_size, vpx_codec_frame_buffer_t *fb) {
+    EXPECT_TRUE(fb != NULL);
+    const int idx = FindFreeBufferIndex();
+    if (idx == num_buffers_)
+      return -1;
+
+    if (ext_fb_list_[idx].size < min_size) {
+      delete [] ext_fb_list_[idx].data;
+      ext_fb_list_[idx].data = new uint8_t[min_size];
+      memset(ext_fb_list_[idx].data, 0, min_size);
+      ext_fb_list_[idx].size = min_size;
+    }
+
+    SetFrameBuffer(idx, fb);
+    return 0;
+  }
+
+  // Test function that will not allocate any data for the frame buffer.
+  // Returns < 0 on an error.
+  int GetZeroFrameBuffer(size_t min_size, vpx_codec_frame_buffer_t *fb) {
+    EXPECT_TRUE(fb != NULL);
+    const int idx = FindFreeBufferIndex();
+    if (idx == num_buffers_)
+      return -1;
+
+    if (ext_fb_list_[idx].size < min_size) {
+      delete [] ext_fb_list_[idx].data;
+      ext_fb_list_[idx].data = NULL;
+      ext_fb_list_[idx].size = min_size;
+    }
+
+    SetFrameBuffer(idx, fb);
+    return 0;
+  }
+
+  // Marks the external frame buffer that |fb| is pointing to as free.
+  // Returns < 0 on an error.
+  int ReturnFrameBuffer(vpx_codec_frame_buffer_t *fb) {
+    if (fb == NULL) {
+      EXPECT_TRUE(fb != NULL);
+      return -1;
+    }
+    ExternalFrameBuffer *const ext_fb =
+        reinterpret_cast<ExternalFrameBuffer*>(fb->priv);
+    if (ext_fb == NULL) {
+      EXPECT_TRUE(ext_fb != NULL);
+      return -1;
+    }
+    EXPECT_EQ(1, ext_fb->in_use);
+    ext_fb->in_use = 0;
+    return 0;
+  }
+
+  // Checks that the ximage data is contained within the external frame buffer
+  // private data passed back in the ximage.
+  void CheckXImageFrameBuffer(const vpx_image_t *img) {
+    if (img->fb_priv != NULL) {
+      const struct ExternalFrameBuffer *const ext_fb =
+          reinterpret_cast<ExternalFrameBuffer*>(img->fb_priv);
+
+      ASSERT_TRUE(img->planes[0] >= ext_fb->data &&
+                  img->planes[0] < (ext_fb->data + ext_fb->size));
+    }
+  }
+
+ private:
+  // Returns the index of the first free frame buffer. Returns |num_buffers_|
+  // if there are no free frame buffers.
+  int FindFreeBufferIndex() {
+    int i;
+    // Find a free frame buffer.
+    for (i = 0; i < num_buffers_; ++i) {
+      if (!ext_fb_list_[i].in_use)
+        break;
+    }
+    return i;
+  }
+
+  // Sets |fb| to an external frame buffer. idx is the index into the frame
+  // buffer list.
+  void SetFrameBuffer(int idx, vpx_codec_frame_buffer_t *fb) {
+    ASSERT_TRUE(fb != NULL);
+    fb->data = ext_fb_list_[idx].data;
+    fb->size = ext_fb_list_[idx].size;
+    ASSERT_EQ(0, ext_fb_list_[idx].in_use);
+    ext_fb_list_[idx].in_use = 1;
+    fb->priv = &ext_fb_list_[idx];
+  }
+
+  int num_buffers_;
+  ExternalFrameBuffer *ext_fb_list_;
+};
+
+// Callback used by libvpx to request the application to return a frame
+// buffer of at least |min_size| in bytes.
+int get_vp9_frame_buffer(void *user_priv, size_t min_size,
+                         vpx_codec_frame_buffer_t *fb) {
+  ExternalFrameBufferList *const fb_list =
+      reinterpret_cast<ExternalFrameBufferList*>(user_priv);
+  return fb_list->GetFreeFrameBuffer(min_size, fb);
+}
+
+// Callback used by libvpx to tell the application that |fb| is not needed
+// anymore.
+int release_vp9_frame_buffer(void *user_priv,
+                             vpx_codec_frame_buffer_t *fb) {
+  ExternalFrameBufferList *const fb_list =
+      reinterpret_cast<ExternalFrameBufferList*>(user_priv);
+  return fb_list->ReturnFrameBuffer(fb);
+}
+
+// Callback will not allocate data for frame buffer.
+int get_vp9_zero_frame_buffer(void *user_priv, size_t min_size,
+                              vpx_codec_frame_buffer_t *fb) {
+  ExternalFrameBufferList *const fb_list =
+      reinterpret_cast<ExternalFrameBufferList*>(user_priv);
+  return fb_list->GetZeroFrameBuffer(min_size, fb);
+}
+
+// Callback will allocate one less byte than |min_size|.
+int get_vp9_one_less_byte_frame_buffer(void *user_priv, size_t min_size,
+                                       vpx_codec_frame_buffer_t *fb) {
+  ExternalFrameBufferList *const fb_list =
+      reinterpret_cast<ExternalFrameBufferList*>(user_priv);
+  return fb_list->GetFreeFrameBuffer(min_size - 1, fb);
+}
+
+// Callback will not release the external frame buffer.
+int do_not_release_vp9_frame_buffer(void *user_priv,
+                                    vpx_codec_frame_buffer_t *fb) {
+  (void)user_priv;
+  (void)fb;
+  return 0;
+}
+
+// Class for testing passing in external frame buffers to libvpx.
+class ExternalFrameBufferMD5Test
+    : public ::libvpx_test::DecoderTest,
+      public ::libvpx_test::CodecTestWithParam<const char*> {
+ protected:
+  ExternalFrameBufferMD5Test()
+      : DecoderTest(GET_PARAM(::libvpx_test::kCodecFactoryParam)),
+        md5_file_(NULL),
+        num_buffers_(0) {}
+
+  virtual ~ExternalFrameBufferMD5Test() {
+    if (md5_file_ != NULL)
+      fclose(md5_file_);
+  }
+
+  virtual void PreDecodeFrameHook(
+      const libvpx_test::CompressedVideoSource &video,
+      libvpx_test::Decoder *decoder) {
+    if (num_buffers_ > 0 && video.frame_number() == 0) {
+      // Have libvpx use frame buffers we create.
+      ASSERT_TRUE(fb_list_.CreateBufferList(num_buffers_));
+      ASSERT_EQ(VPX_CODEC_OK,
+                decoder->SetFrameBufferFunctions(
+                    GetVP9FrameBuffer, ReleaseVP9FrameBuffer, this));
+    }
+  }
+
+  void OpenMD5File(const std::string &md5_file_name_) {
+    md5_file_ = libvpx_test::OpenTestDataFile(md5_file_name_);
+    ASSERT_TRUE(md5_file_ != NULL) << "Md5 file open failed. Filename: "
+        << md5_file_name_;
+  }
+
+  virtual void DecompressedFrameHook(const vpx_image_t &img,
+                                     const unsigned int frame_number) {
+    ASSERT_TRUE(md5_file_ != NULL);
+    char expected_md5[33];
+    char junk[128];
+
+    // Read correct md5 checksums.
+    const int res = fscanf(md5_file_, "%s  %s", expected_md5, junk);
+    ASSERT_NE(EOF, res) << "Read md5 data failed";
+    expected_md5[32] = '\0';
+
+    ::libvpx_test::MD5 md5_res;
+    md5_res.Add(&img);
+    const char *const actual_md5 = md5_res.Get();
+
+    // Check md5 match.
+    ASSERT_STREQ(expected_md5, actual_md5)
+        << "Md5 checksums don't match: frame number = " << frame_number;
+  }
+
+  // Callback to get a free external frame buffer. Return value < 0 is an
+  // error.
+  static int GetVP9FrameBuffer(void *user_priv, size_t min_size,
+                               vpx_codec_frame_buffer_t *fb) {
+    ExternalFrameBufferMD5Test *const md5Test =
+        reinterpret_cast<ExternalFrameBufferMD5Test*>(user_priv);
+    return md5Test->fb_list_.GetFreeFrameBuffer(min_size, fb);
+  }
+
+  // Callback to release an external frame buffer. Return value < 0 is an
+  // error.
+  static int ReleaseVP9FrameBuffer(void *user_priv,
+                                   vpx_codec_frame_buffer_t *fb) {
+    ExternalFrameBufferMD5Test *const md5Test =
+        reinterpret_cast<ExternalFrameBufferMD5Test*>(user_priv);
+    return md5Test->fb_list_.ReturnFrameBuffer(fb);
+  }
+
+  void set_num_buffers(int num_buffers) { num_buffers_ = num_buffers; }
+  int num_buffers() const { return num_buffers_; }
+
+ private:
+  FILE *md5_file_;
+  int num_buffers_;
+  ExternalFrameBufferList fb_list_;
+};
+
+#if CONFIG_WEBM_IO
+// Class for testing passing in external frame buffers to libvpx.
+class ExternalFrameBufferTest : public ::testing::Test {
+ protected:
+  ExternalFrameBufferTest()
+      : video_(NULL),
+        decoder_(NULL),
+        num_buffers_(0) {}
+
+  virtual void SetUp() {
+    video_ = new libvpx_test::WebMVideoSource(kVP9TestFile);
+    ASSERT_TRUE(video_ != NULL);
+    video_->Init();
+    video_->Begin();
+
+    vpx_codec_dec_cfg_t cfg = vpx_codec_dec_cfg_t();
+    decoder_ = new libvpx_test::VP9Decoder(cfg, 0);
+    ASSERT_TRUE(decoder_ != NULL);
+  }
+
+  virtual void TearDown() {
+    delete decoder_;
+    delete video_;
+  }
+
+  // Passes the external frame buffer information to libvpx.
+  vpx_codec_err_t SetFrameBufferFunctions(
+      int num_buffers,
+      vpx_get_frame_buffer_cb_fn_t cb_get,
+      vpx_release_frame_buffer_cb_fn_t cb_release) {
+    if (num_buffers > 0) {
+      num_buffers_ = num_buffers;
+      EXPECT_TRUE(fb_list_.CreateBufferList(num_buffers_));
+    }
+
+    return decoder_->SetFrameBufferFunctions(cb_get, cb_release, &fb_list_);
+  }
+
+  vpx_codec_err_t DecodeOneFrame() {
+    const vpx_codec_err_t res =
+        decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
+    CheckDecodedFrames();
+    if (res == VPX_CODEC_OK)
+      video_->Next();
+    return res;
+  }
+
+  vpx_codec_err_t DecodeRemainingFrames() {
+    for (; video_->cxdata() != NULL; video_->Next()) {
+      const vpx_codec_err_t res =
+          decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
+      if (res != VPX_CODEC_OK)
+        return res;
+      CheckDecodedFrames();
+    }
+    return VPX_CODEC_OK;
+  }
+
+ private:
+  void CheckDecodedFrames() {
+    libvpx_test::DxDataIterator dec_iter = decoder_->GetDxData();
+    const vpx_image_t *img = NULL;
+
+    // Get decompressed data
+    while ((img = dec_iter.Next()) != NULL) {
+      fb_list_.CheckXImageFrameBuffer(img);
+    }
+  }
+
+  libvpx_test::WebMVideoSource *video_;
+  libvpx_test::VP9Decoder *decoder_;
+  int num_buffers_;
+  ExternalFrameBufferList fb_list_;
+};
+#endif  // CONFIG_WEBM_IO
+
+// This test runs through the set of test vectors, and decodes them.
+// Libvpx will call into the application to allocate a frame buffer when
+// needed. The md5 checksums are computed for each frame in the video file.
+// If md5 checksums match the correct md5 data, then the test is passed.
+// Otherwise, the test failed.
+TEST_P(ExternalFrameBufferMD5Test, ExtFBMD5Match) {
+  const std::string filename = GET_PARAM(kVideoNameParam);
+  libvpx_test::CompressedVideoSource *video = NULL;
+
+  // Number of buffers equals #VP9_MAXIMUM_REF_BUFFERS +
+  // #VPX_MAXIMUM_WORK_BUFFERS + four jitter buffers.
+  const int jitter_buffers = 4;
+  const int num_buffers =
+      VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS + jitter_buffers;
+  set_num_buffers(num_buffers);
+
+#if CONFIG_VP8_DECODER
+  // Tell compiler we are not using kVP8TestVectors.
+  (void)libvpx_test::kVP8TestVectors;
+#endif
+
+  // Open compressed video file.
+  if (filename.substr(filename.length() - 3, 3) == "ivf") {
+    video = new libvpx_test::IVFVideoSource(filename);
+  } else {
+#if CONFIG_WEBM_IO
+    video = new libvpx_test::WebMVideoSource(filename);
+#else
+    fprintf(stderr, "WebM IO is disabled, skipping test vector %s\n",
+            filename.c_str());
+    return;
+#endif
+  }
+  ASSERT_TRUE(video != NULL);
+  video->Init();
+
+  // Construct md5 file name.
+  const std::string md5_filename = filename + ".md5";
+  OpenMD5File(md5_filename);
+
+  // Decode frame, and check the md5 matching.
+  ASSERT_NO_FATAL_FAILURE(RunLoop(video));
+  delete video;
+}
+
+#if CONFIG_WEBM_IO
+TEST_F(ExternalFrameBufferTest, MinFrameBuffers) {
+  // Minimum number of external frame buffers for VP9 is
+  // #VP9_MAXIMUM_REF_BUFFERS + #VPX_MAXIMUM_WORK_BUFFERS.
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(
+                num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames());
+}
+
+TEST_F(ExternalFrameBufferTest, EightJitterBuffers) {
+  // Number of buffers equals #VP9_MAXIMUM_REF_BUFFERS +
+  // #VPX_MAXIMUM_WORK_BUFFERS + eight jitter buffers.
+  const int jitter_buffers = 8;
+  const int num_buffers =
+      VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS + jitter_buffers;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(
+                num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames());
+}
+
+TEST_F(ExternalFrameBufferTest, NotEnoughBuffers) {
+  // Minimum number of external frame buffers for VP9 is
+  // #VP9_MAXIMUM_REF_BUFFERS + #VPX_MAXIMUM_WORK_BUFFERS. Most files will
+  // only use 5 frame buffers at one time.
+  const int num_buffers = 2;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(
+                num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame());
+  ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeRemainingFrames());
+}
+
+TEST_F(ExternalFrameBufferTest, NoRelease) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer,
+                                    do_not_release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame());
+  ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeRemainingFrames());
+}
+
+TEST_F(ExternalFrameBufferTest, NullRealloc) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(num_buffers, get_vp9_zero_frame_buffer,
+                                    release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeOneFrame());
+}
+
+TEST_F(ExternalFrameBufferTest, ReallocOneLessByte) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_OK,
+            SetFrameBufferFunctions(
+                num_buffers, get_vp9_one_less_byte_frame_buffer,
+                release_vp9_frame_buffer));
+  ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeOneFrame());
+}
+
+TEST_F(ExternalFrameBufferTest, NullGetFunction) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_INVALID_PARAM,
+            SetFrameBufferFunctions(num_buffers, NULL,
+                                    release_vp9_frame_buffer));
+}
+
+TEST_F(ExternalFrameBufferTest, NullReleaseFunction) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_INVALID_PARAM,
+            SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer, NULL));
+}
+
+TEST_F(ExternalFrameBufferTest, SetAfterDecode) {
+  const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
+  ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame());
+  ASSERT_EQ(VPX_CODEC_ERROR,
+            SetFrameBufferFunctions(
+                num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
+}
+#endif  // CONFIG_WEBM_IO
+
+VP9_INSTANTIATE_TEST_CASE(ExternalFrameBufferMD5Test,
+                          ::testing::ValuesIn(libvpx_test::kVP9TestVectors,
+                                              libvpx_test::kVP9TestVectors +
+                                              libvpx_test::kNumVP9TestVectors));
+}  // namespace
--- a/test/fdct4x4_test.cc
+++ b/test/fdct4x4_test.cc
@@ -13,86 +13,272 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp10_rtcd.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
-#include "test/transform_test_base.h"
 #include "test/util.h"
-#include "vp10/common/entropy.h"
+
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_entropy.h"
 #include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"

 using libvpx_test::ACMRandom;

 namespace {
+const int kNumCoeffs = 16;
 typedef void (*FdctFunc)(const int16_t *in, tran_low_t *out, int stride);
 typedef void (*IdctFunc)(const tran_low_t *in, uint8_t *out, int stride);
+typedef void (*FhtFunc)(const int16_t *in, tran_low_t *out, int stride,
+                        int tx_type);
 typedef void (*IhtFunc)(const tran_low_t *in, uint8_t *out, int stride,
                        int tx_type);
-using libvpx_test::FhtFunc;

-typedef std::tr1::tuple<FdctFunc, IdctFunc, int, vpx_bit_depth_t, int>
-Dct4x4Param;
-typedef std::tr1::tuple<FhtFunc, IhtFunc, int, vpx_bit_depth_t, int>
-Ht4x4Param;
+typedef std::tr1::tuple<FdctFunc, IdctFunc, int, vpx_bit_depth_t> Dct4x4Param;
+typedef std::tr1::tuple<FhtFunc, IhtFunc, int, vpx_bit_depth_t> Ht4x4Param;

 void fdct4x4_ref(const int16_t *in, tran_low_t *out, int stride,
-                 int /*tx_type*/) {
-  vpx_fdct4x4_c(in, out, stride);
+                 int tx_type) {
+  vp9_fdct4x4_c(in, out, stride);
 }

 void fht4x4_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
-  vp10_fht4x4_c(in, out, stride, tx_type);
+  vp9_fht4x4_c(in, out, stride, tx_type);
 }

 void fwht4x4_ref(const int16_t *in, tran_low_t *out, int stride,
-                 int /*tx_type*/) {
-  vp10_fwht4x4_c(in, out, stride);
+                 int tx_type) {
+  vp9_fwht4x4_c(in, out, stride);
 }

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 void idct4x4_10(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct4x4_16_add_c(in, out, stride, 10);
+  vp9_highbd_idct4x4_16_add_c(in, out, stride, 10);
 }

 void idct4x4_12(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct4x4_16_add_c(in, out, stride, 12);
+  vp9_highbd_idct4x4_16_add_c(in, out, stride, 12);
 }

 void iht4x4_10(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht4x4_16_add_c(in, out, stride, tx_type, 10);
+  vp9_highbd_iht4x4_16_add_c(in, out, stride, tx_type, 10);
 }

 void iht4x4_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht4x4_16_add_c(in, out, stride, tx_type, 12);
+  vp9_highbd_iht4x4_16_add_c(in, out, stride, tx_type, 12);
 }

 void iwht4x4_10(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_iwht4x4_16_add_c(in, out, stride, 10);
+  vp9_highbd_iwht4x4_16_add_c(in, out, stride, 10);
 }

 void iwht4x4_12(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_iwht4x4_16_add_c(in, out, stride, 12);
+  vp9_highbd_iwht4x4_16_add_c(in, out, stride, 12);
 }

 #if HAVE_SSE2
 void idct4x4_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct4x4_16_add_sse2(in, out, stride, 10);
+  vp9_highbd_idct4x4_16_add_sse2(in, out, stride, 10);
 }

 void idct4x4_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct4x4_16_add_sse2(in, out, stride, 12);
+  vp9_highbd_idct4x4_16_add_sse2(in, out, stride, 12);
 }
 #endif  // HAVE_SSE2
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

+class Trans4x4TestBase {
+ public:
+  virtual ~Trans4x4TestBase() {}
+
+ protected:
+  virtual void RunFwdTxfm(const int16_t *in, tran_low_t *out, int stride) = 0;
+
+  virtual void RunInvTxfm(const tran_low_t *out, uint8_t *dst, int stride) = 0;
+
+  void RunAccuracyCheck(int limit) {
+    ACMRandom rnd(ACMRandom::DeterministicSeed());
+    uint32_t max_error = 0;
+    int64_t total_error = 0;
+    const int count_test_block = 10000;
+    for (int i = 0; i < count_test_block; ++i) {
+      DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_temp_block, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+      DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+      DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
+#endif
+
+      // Initialize a test block with input range [-255, 255].
+      for (int j = 0; j < kNumCoeffs; ++j) {
+        if (bit_depth_ == VPX_BITS_8) {
+          src[j] = rnd.Rand8();
+          dst[j] = rnd.Rand8();
+          test_input_block[j] = src[j] - dst[j];
+#if CONFIG_VP9_HIGHBITDEPTH
+        } else {
+          src16[j] = rnd.Rand16() & mask_;
+          dst16[j] = rnd.Rand16() & mask_;
+          test_input_block[j] = src16[j] - dst16[j];
+#endif
+        }
+      }
+
+      ASM_REGISTER_STATE_CHECK(RunFwdTxfm(test_input_block,
+                                          test_temp_block, pitch_));
+      if (bit_depth_ == VPX_BITS_8) {
+        ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block, dst, pitch_));
+#if CONFIG_VP9_HIGHBITDEPTH
+      } else {
+        ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block,
+                                            CONVERT_TO_BYTEPTR(dst16), pitch_));
+#endif
+      }
+
+      for (int j = 0; j < kNumCoeffs; ++j) {
+#if CONFIG_VP9_HIGHBITDEPTH
+        const uint32_t diff =
+            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
+#else
+        const uint32_t diff = dst[j] - src[j];
+#endif
+        const uint32_t error = diff * diff;
+        if (max_error < error)
+          max_error = error;
+        total_error += error;
+      }
+    }
+
+    EXPECT_GE(static_cast<uint32_t>(limit), max_error)
+        << "Error: 4x4 FHT/IHT has an individual round trip error > "
+        << limit;
+
+    EXPECT_GE(count_test_block * limit, total_error)
+        << "Error: 4x4 FHT/IHT has average round trip error > " << limit
+        << " per block";
+  }
+
+  void RunCoeffCheck() {
+    ACMRandom rnd(ACMRandom::DeterministicSeed());
+    const int count_test_block = 5000;
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);
+
+    for (int i = 0; i < count_test_block; ++i) {
+      // Initialize a test block with input range [-mask_, mask_].
+      for (int j = 0; j < kNumCoeffs; ++j)
+        input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
+
+      fwd_txfm_ref(input_block, output_ref_block, pitch_, tx_type_);
+      ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_block, output_block, pitch_));
+
+      // The minimum quant value is 4.
+      for (int j = 0; j < kNumCoeffs; ++j)
+        EXPECT_EQ(output_block[j], output_ref_block[j]);
+    }
+  }
+
+  void RunMemCheck() {
+    ACMRandom rnd(ACMRandom::DeterministicSeed());
+    const int count_test_block = 5000;
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, input_extreme_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_block, kNumCoeffs);
+
+    for (int i = 0; i < count_test_block; ++i) {
+      // Initialize a test block with input range [-mask_, mask_].
+      for (int j = 0; j < kNumCoeffs; ++j) {
+        input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
+        input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
+      }
+      if (i == 0) {
+        for (int j = 0; j < kNumCoeffs; ++j)
+          input_extreme_block[j] = mask_;
+      } else if (i == 1) {
+        for (int j = 0; j < kNumCoeffs; ++j)
+          input_extreme_block[j] = -mask_;
+      }
+
+      fwd_txfm_ref(input_extreme_block, output_ref_block, pitch_, tx_type_);
+      ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_extreme_block,
+                                          output_block, pitch_));
+
+      // The minimum quant value is 4.
+      for (int j = 0; j < kNumCoeffs; ++j) {
+        EXPECT_EQ(output_block[j], output_ref_block[j]);
+        EXPECT_GE(4 * DCT_MAX_VALUE << (bit_depth_ - 8), abs(output_block[j]))
+            << "Error: 4x4 FDCT has coefficient larger than 4*DCT_MAX_VALUE";
+      }
+    }
+  }
+
+  void RunInvAccuracyCheck(int limit) {
+    ACMRandom rnd(ACMRandom::DeterministicSeed());
+    const int count_test_block = 1000;
+    DECLARE_ALIGNED_ARRAY(16, int16_t, in, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
+#endif
+
+    for (int i = 0; i < count_test_block; ++i) {
+      // Initialize a test block with input range [-mask_, mask_].
+      for (int j = 0; j < kNumCoeffs; ++j) {
+        if (bit_depth_ == VPX_BITS_8) {
+          src[j] = rnd.Rand8();
+          dst[j] = rnd.Rand8();
+          in[j] = src[j] - dst[j];
+#if CONFIG_VP9_HIGHBITDEPTH
+        } else {
+          src16[j] = rnd.Rand16() & mask_;
+          dst16[j] = rnd.Rand16() & mask_;
+          in[j] = src16[j] - dst16[j];
+#endif
+        }
+      }
+
+      fwd_txfm_ref(in, coeff, pitch_, tx_type_);
+
+      if (bit_depth_ == VPX_BITS_8) {
+        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
+#if CONFIG_VP9_HIGHBITDEPTH
+      } else {
+        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
+                                            pitch_));
+#endif
+      }
+
+      for (int j = 0; j < kNumCoeffs; ++j) {
+#if CONFIG_VP9_HIGHBITDEPTH
+        const uint32_t diff =
+            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
+#else
+        const uint32_t diff = dst[j] - src[j];
+#endif
+        const uint32_t error = diff * diff;
+        EXPECT_GE(static_cast<uint32_t>(limit), error)
+            << "Error: 4x4 IDCT has error " << error
+            << " at index " << j;
+      }
+    }
+  }
+
+  int pitch_;
+  int tx_type_;
+  FhtFunc fwd_txfm_ref;
+  vpx_bit_depth_t bit_depth_;
+  int mask_;
+};

 class Trans4x4DCT
-    : public libvpx_test::TransformTestBase,
+    : public Trans4x4TestBase,
      public ::testing::TestWithParam<Dct4x4Param> {
 public:
  virtual ~Trans4x4DCT() {}
@@ -105,7 +291,6 @@ class Trans4x4DCT
    fwd_txfm_ref = fdct4x4_ref;
    bit_depth_ = GET_PARAM(3);
    mask_ = (1 << bit_depth_) - 1;
-    num_coeffs_ = GET_PARAM(4);
  }
  virtual void TearDown() { libvpx_test::ClearSystemState(); }

@@ -138,7 +323,7 @@ TEST_P(Trans4x4DCT, InvAccuracyCheck) {
 }

 class Trans4x4HT
-    : public libvpx_test::TransformTestBase,
+    : public Trans4x4TestBase,
      public ::testing::TestWithParam<Ht4x4Param> {
 public:
  virtual ~Trans4x4HT() {}
@@ -151,7 +336,6 @@ class Trans4x4HT
    fwd_txfm_ref = fht4x4_ref;
    bit_depth_ = GET_PARAM(3);
    mask_ = (1 << bit_depth_) - 1;
-    num_coeffs_ = GET_PARAM(4);
  }
  virtual void TearDown() { libvpx_test::ClearSystemState(); }

@@ -185,7 +369,7 @@ TEST_P(Trans4x4HT, InvAccuracyCheck) {
 }

 class Trans4x4WHT
-    : public libvpx_test::TransformTestBase,
+    : public Trans4x4TestBase,
      public ::testing::TestWithParam<Dct4x4Param> {
 public:
  virtual ~Trans4x4WHT() {}
@@ -198,7 +382,6 @@ class Trans4x4WHT
    fwd_txfm_ref = fwht4x4_ref;
    bit_depth_ = GET_PARAM(3);
    mask_ = (1 << bit_depth_) - 1;
-    num_coeffs_ = GET_PARAM(4);
  }
  virtual void TearDown() { libvpx_test::ClearSystemState(); }

@@ -231,155 +414,126 @@ TEST_P(Trans4x4WHT, InvAccuracyCheck) {
 }
 using std::tr1::make_tuple;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_10, 0, VPX_BITS_10, 16),
-        make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_12, 0, VPX_BITS_12, 16),
-        make_tuple(&vpx_fdct4x4_c, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8, 16)));
+        make_tuple(&vp9_highbd_fdct4x4_c, &idct4x4_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct4x4_c, &idct4x4_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_fdct4x4_c, &vp9_idct4x4_16_add_c, 0, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4DCT,
    ::testing::Values(
-         make_tuple(&vpx_fdct4x4_c, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8, 16)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fdct4x4_c, &vp9_idct4x4_16_add_c, 0, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4HT,
    ::testing::Values(
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_10, 0, VPX_BITS_10, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_10, 1, VPX_BITS_10, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_10, 2, VPX_BITS_10, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_10, 3, VPX_BITS_10, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_12, 0, VPX_BITS_12, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_12, 1, VPX_BITS_12, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_12, 2, VPX_BITS_12, 16),
-        make_tuple(&vp10_highbd_fht4x4_c, &iht4x4_12, 3, VPX_BITS_12, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 0, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 1, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 2, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 3, VPX_BITS_8, 16)));
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_10, 1, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_10, 2, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_10, 3, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_12, 1, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_12, 2, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_c, &iht4x4_12, 3, VPX_BITS_12),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 3, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4HT,
    ::testing::Values(
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 0, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 1, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 2, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_c, 3, VPX_BITS_8, 16)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_c, 3, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4WHT,
    ::testing::Values(
-        make_tuple(&vp10_highbd_fwht4x4_c, &iwht4x4_10, 0, VPX_BITS_10, 16),
-        make_tuple(&vp10_highbd_fwht4x4_c, &iwht4x4_12, 0, VPX_BITS_12, 16),
-        make_tuple(&vp10_fwht4x4_c, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8, 16)));
+        make_tuple(&vp9_highbd_fwht4x4_c, &iwht4x4_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fwht4x4_c, &iwht4x4_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_fwht4x4_c, &vp9_iwht4x4_16_add_c, 0, VPX_BITS_8)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4WHT,
    ::testing::Values(
-        make_tuple(&vp10_fwht4x4_c, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8, 16)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fwht4x4_c, &vp9_iwht4x4_16_add_c, 0, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if HAVE_NEON_ASM && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    NEON, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct4x4_c,
-                   &vpx_idct4x4_16_add_neon, 0, VPX_BITS_8, 16)));
-#endif  // HAVE_NEON_ASM && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fdct4x4_c,
+                   &vp9_idct4x4_16_add_neon, 0, VPX_BITS_8)));
+#endif  // HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    NEON, Trans4x4HT,
    ::testing::Values(
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_neon,
-                   0, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_neon,
-                   1, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_neon,
-                   2, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_c, &vp10_iht4x4_16_add_neon,
-                   3, VPX_BITS_8, 16)));
-#endif  // HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 3, VPX_BITS_8)));
+#endif  // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE
+#if CONFIG_USE_X86INC && HAVE_MMX && !CONFIG_VP9_HIGHBITDEPTH && \
+    !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
-    SSE2, Trans4x4WHT,
+    MMX, Trans4x4WHT,
    ::testing::Values(
-        make_tuple(&vp10_fwht4x4_c, &vpx_iwht4x4_16_add_c,
-                   0, VPX_BITS_8, 16),
-        make_tuple(&vp10_fwht4x4_c, &vpx_iwht4x4_16_add_sse2,
-                   0, VPX_BITS_8, 16)));
+        make_tuple(&vp9_fwht4x4_mmx, &vp9_iwht4x4_16_add_c, 0, VPX_BITS_8)));
 #endif

-#if HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct4x4_sse2,
-                   &vpx_idct4x4_16_add_sse2, 0, VPX_BITS_8, 16)));
+        make_tuple(&vp9_fdct4x4_sse2,
+                   &vp9_idct4x4_16_add_sse2, 0, VPX_BITS_8)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4HT,
    ::testing::Values(
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_sse2, 0,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_sse2, 1,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_sse2, 2,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_sse2, 3,
-                   VPX_BITS_8, 16)));
-#endif  // HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_sse2, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_sse2, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_sse2, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_sse2, 3, VPX_BITS_8)));
+#endif  // HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct4x4_c,    &idct4x4_10_sse2, 0,
-                   VPX_BITS_10, 16),
-        make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_10_sse2, 0,
-                   VPX_BITS_10, 16),
-        make_tuple(&vpx_highbd_fdct4x4_c,    &idct4x4_12_sse2, 0,
-                   VPX_BITS_12, 16),
-        make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_12_sse2, 0,
-                   VPX_BITS_12, 16),
-        make_tuple(&vpx_fdct4x4_sse2,      &vpx_idct4x4_16_add_c, 0,
-                   VPX_BITS_8, 16)));
+        make_tuple(&vp9_highbd_fdct4x4_c,    &idct4x4_10_sse2, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct4x4_sse2, &idct4x4_10_sse2, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct4x4_c,    &idct4x4_12_sse2, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fdct4x4_sse2, &idct4x4_12_sse2, 0, VPX_BITS_12),
+        make_tuple(&vp9_fdct4x4_sse2,      &vp9_idct4x4_16_add_c, 0,
+                   VPX_BITS_8)));

 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4HT,
    ::testing::Values(
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_c,
-                   0, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_c,
-                   1, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_c,
-                   2, VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_sse2, &vp10_iht4x4_16_add_c,
-                   3, VPX_BITS_8, 16)));
-#endif  // HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-
-#if HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    MSA, Trans4x4DCT,
-    ::testing::Values(
-        make_tuple(&vpx_fdct4x4_msa, &vpx_idct4x4_16_add_msa, 0,
-                   VPX_BITS_8, 16)));
-INSTANTIATE_TEST_CASE_P(
-    MSA, Trans4x4HT,
-    ::testing::Values(
-        make_tuple(&vp10_fht4x4_msa, &vp10_iht4x4_16_add_msa, 0,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_msa, &vp10_iht4x4_16_add_msa, 1,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_msa, &vp10_iht4x4_16_add_msa, 2,
-                   VPX_BITS_8, 16),
-        make_tuple(&vp10_fht4x4_msa, &vp10_iht4x4_16_add_msa, 3,
-                   VPX_BITS_8, 16)));
-#endif  // HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_10, 1, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_10, 2, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_10, 3, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_12, 1, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_12, 2, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht4x4_sse2, &iht4x4_12, 3, VPX_BITS_12),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht4x4_sse2, &vp9_iht4x4_16_add_c, 3, VPX_BITS_8)));
+#endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/fdct8x8_test.cc
+++ b/test/fdct8x8_test.cc
@@ -13,41 +13,19 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp10_rtcd.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vp10/common/entropy.h"
-#include "vp10/common/scan.h"
+
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_entropy.h"
 #include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"
-
-using libvpx_test::ACMRandom;
-
-namespace {

 const int kNumCoeffs = 64;
 const double kPi = 3.141592653589793238462643383279502884;
-
-const int kSignBiasMaxDiff255 = 1500;
-const int kSignBiasMaxDiff15 = 10000;
-
-typedef void (*FdctFunc)(const int16_t *in, tran_low_t *out, int stride);
-typedef void (*IdctFunc)(const tran_low_t *in, uint8_t *out, int stride);
-typedef void (*FhtFunc)(const int16_t *in, tran_low_t *out, int stride,
-                        int tx_type);
-typedef void (*IhtFunc)(const tran_low_t *in, uint8_t *out, int stride,
-                        int tx_type);
-
-typedef std::tr1::tuple<FdctFunc, IdctFunc, int, vpx_bit_depth_t> Dct8x8Param;
-typedef std::tr1::tuple<FhtFunc, IhtFunc, int, vpx_bit_depth_t> Ht8x8Param;
-typedef std::tr1::tuple<IdctFunc, IdctFunc, int, vpx_bit_depth_t> Idct8x8Param;
-
-void reference_8x8_dct_1d(const double in[8], double out[8]) {
+void reference_8x8_dct_1d(const double in[8], double out[8], int stride) {
  const double kInvSqrt2 = 0.707106781186547524400844362104;
  for (int k = 0; k < 8; k++) {
    out[k] = 0.0;
@@ -65,7 +43,7 @@ void reference_8x8_dct_2d(const int16_t input[kNumCoeffs],
    double temp_in[8], temp_out[8];
    for (int j = 0; j < 8; ++j)
      temp_in[j] = input[j*8 + i];
-    reference_8x8_dct_1d(temp_in, temp_out);
+    reference_8x8_dct_1d(temp_in, temp_out, 1);
    for (int j = 0; j < 8; ++j)
      output[j * 8 + i] = temp_out[j];
  }
@@ -74,67 +52,82 @@ void reference_8x8_dct_2d(const int16_t input[kNumCoeffs],
    double temp_in[8], temp_out[8];
    for (int j = 0; j < 8; ++j)
      temp_in[j] = output[j + i*8];
-    reference_8x8_dct_1d(temp_in, temp_out);
+    reference_8x8_dct_1d(temp_in, temp_out, 1);
    // Scale by some magic number
    for (int j = 0; j < 8; ++j)
      output[j + i * 8] = temp_out[j] * 2;
  }
 }

+using libvpx_test::ACMRandom;

-void fdct8x8_ref(const int16_t *in, tran_low_t *out, int stride,
-                 int /*tx_type*/) {
-  vpx_fdct8x8_c(in, out, stride);
+namespace {
+
+const int kSignBiasMaxDiff255 = 1500;
+const int kSignBiasMaxDiff15 = 10000;
+
+typedef void (*FdctFunc)(const int16_t *in, tran_low_t *out, int stride);
+typedef void (*IdctFunc)(const tran_low_t *in, uint8_t *out, int stride);
+typedef void (*FhtFunc)(const int16_t *in, tran_low_t *out, int stride,
+                        int tx_type);
+typedef void (*IhtFunc)(const tran_low_t *in, uint8_t *out, int stride,
+                        int tx_type);
+
+typedef std::tr1::tuple<FdctFunc, IdctFunc, int, vpx_bit_depth_t> Dct8x8Param;
+typedef std::tr1::tuple<FhtFunc, IhtFunc, int, vpx_bit_depth_t> Ht8x8Param;
+typedef std::tr1::tuple<IdctFunc, IdctFunc, int, vpx_bit_depth_t> Idct8x8Param;
+
+void fdct8x8_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
+  vp9_fdct8x8_c(in, out, stride);
 }

 void fht8x8_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
-  vp10_fht8x8_c(in, out, stride, tx_type);
+  vp9_fht8x8_c(in, out, stride, tx_type);
 }

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 void idct8x8_10(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_64_add_c(in, out, stride, 10);
+  vp9_highbd_idct8x8_64_add_c(in, out, stride, 10);
 }

 void idct8x8_12(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_64_add_c(in, out, stride, 12);
+  vp9_highbd_idct8x8_64_add_c(in, out, stride, 12);
 }

 void iht8x8_10(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht8x8_64_add_c(in, out, stride, tx_type, 10);
+  vp9_highbd_iht8x8_64_add_c(in, out, stride, tx_type, 10);
 }

 void iht8x8_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
-  vp10_highbd_iht8x8_64_add_c(in, out, stride, tx_type, 12);
+  vp9_highbd_iht8x8_64_add_c(in, out, stride, tx_type, 12);
 }

-#if HAVE_SSE2
-
 void idct8x8_10_add_10_c(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_10_add_c(in, out, stride, 10);
+  vp9_highbd_idct8x8_10_add_c(in, out, stride, 10);
 }

 void idct8x8_10_add_12_c(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_10_add_c(in, out, stride, 12);
+  vp9_highbd_idct8x8_10_add_c(in, out, stride, 12);
 }

+#if HAVE_SSE2
 void idct8x8_10_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_10_add_sse2(in, out, stride, 10);
+  vp9_highbd_idct8x8_10_add_sse2(in, out, stride, 10);
 }

 void idct8x8_10_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_10_add_sse2(in, out, stride, 12);
+  vp9_highbd_idct8x8_10_add_sse2(in, out, stride, 12);
 }

 void idct8x8_64_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_64_add_sse2(in, out, stride, 10);
+  vp9_highbd_idct8x8_64_add_sse2(in, out, stride, 10);
 }

 void idct8x8_64_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct8x8_64_add_sse2(in, out, stride, 12);
+  vp9_highbd_idct8x8_64_add_sse2(in, out, stride, 12);
 }
 #endif  // HAVE_SSE2
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

 class FwdTrans8x8TestBase {
 public:
@@ -146,8 +139,8 @@ class FwdTrans8x8TestBase {

  void RunSignBiasCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
-    DECLARE_ALIGNED(16, int16_t, test_input_block[64]);
-    DECLARE_ALIGNED(16, tran_low_t, test_output_block[64]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_output_block, 64);
    int count_sign_block[64][2];
    const int count_test_block = 100000;

@@ -217,13 +210,13 @@ class FwdTrans8x8TestBase {
    int max_error = 0;
    int total_error = 0;
    const int count_test_block = 100000;
-    DECLARE_ALIGNED(16, int16_t, test_input_block[64]);
-    DECLARE_ALIGNED(16, tran_low_t, test_temp_block[64]);
-    DECLARE_ALIGNED(16, uint8_t, dst[64]);
-    DECLARE_ALIGNED(16, uint8_t, src[64]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[64]);
-    DECLARE_ALIGNED(16, uint16_t, src16[64]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_temp_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, src, 64);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, 64);
 #endif

    for (int i = 0; i < count_test_block; ++i) {
@@ -233,7 +226,7 @@ class FwdTrans8x8TestBase {
          src[j] = rnd.Rand8();
          dst[j] = rnd.Rand8();
          test_input_block[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          src16[j] = rnd.Rand16() & mask_;
          dst16[j] = rnd.Rand16() & mask_;
@@ -258,7 +251,7 @@ class FwdTrans8x8TestBase {
      if (bit_depth_ == VPX_BITS_8) {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, CONVERT_TO_BYTEPTR(dst16), pitch_));
@@ -266,7 +259,7 @@ class FwdTrans8x8TestBase {
      }

      for (int j = 0; j < 64; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        const int diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
@@ -294,14 +287,14 @@ class FwdTrans8x8TestBase {
    int total_error = 0;
    int total_coeff_error = 0;
    const int count_test_block = 100000;
-    DECLARE_ALIGNED(16, int16_t, test_input_block[64]);
-    DECLARE_ALIGNED(16, tran_low_t, test_temp_block[64]);
-    DECLARE_ALIGNED(16, tran_low_t, ref_temp_block[64]);
-    DECLARE_ALIGNED(16, uint8_t, dst[64]);
-    DECLARE_ALIGNED(16, uint8_t, src[64]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[64]);
-    DECLARE_ALIGNED(16, uint16_t, src16[64]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, test_input_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_temp_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_temp_block, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, src, 64);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, 64);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, 64);
 #endif

    for (int i = 0; i < count_test_block; ++i) {
@@ -319,7 +312,7 @@ class FwdTrans8x8TestBase {
            dst[j] = rnd.Rand8() % 2 ? 255 : 0;
          }
          test_input_block[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          if (i == 0) {
            src16[j] = mask_;
@@ -343,7 +336,7 @@ class FwdTrans8x8TestBase {
      if (bit_depth_ == VPX_BITS_8) {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ASM_REGISTER_STATE_CHECK(
            RunInvTxfm(test_temp_block, CONVERT_TO_BYTEPTR(dst16), pitch_));
@@ -351,7 +344,7 @@ class FwdTrans8x8TestBase {
      }

      for (int j = 0; j < 64; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        const int diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
@@ -383,13 +376,13 @@ class FwdTrans8x8TestBase {
  void RunInvAccuracyCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 1000;
-    DECLARE_ALIGNED(16, int16_t, in[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, src[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, in, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, src, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, src16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
 #endif

    for (int i = 0; i < count_test_block; ++i) {
@@ -401,7 +394,7 @@ class FwdTrans8x8TestBase {
          src[j] = rnd.Rand8() % 2 ? 255 : 0;
          dst[j] = src[j] > 0 ? 0 : 255;
          in[j] = src[j] - dst[j];
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          src16[j] = rnd.Rand8() % 2 ? mask_ : 0;
          dst16[j] = src16[j] > 0 ? 0 : mask_;
@@ -416,7 +409,7 @@ class FwdTrans8x8TestBase {

      if (bit_depth_ == VPX_BITS_8) {
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
                                            pitch_));
@@ -424,11 +417,11 @@ class FwdTrans8x8TestBase {
      }

      for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
-        const int diff =
+#if CONFIG_VP9_HIGHBITDEPTH
+        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
-        const int diff = dst[j] - src[j];
+        const uint32_t diff = dst[j] - src[j];
 #endif
        const uint32_t error = diff * diff;
        EXPECT_GE(1u << 2 * (bit_depth_ - 8), error)
@@ -441,9 +434,9 @@ class FwdTrans8x8TestBase {
  void RunFwdAccuracyCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 1000;
-    DECLARE_ALIGNED(16, int16_t, in[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, coeff_r[kNumCoeffs]);
-    DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, int16_t, in, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff_r, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);

    for (int i = 0; i < count_test_block; ++i) {
      double out_r[kNumCoeffs];
@@ -458,7 +451,7 @@ class FwdTrans8x8TestBase {
        coeff_r[j] = static_cast<tran_low_t>(round(out_r[j]));

      for (int j = 0; j < kNumCoeffs; ++j) {
-        const int32_t diff = coeff[j] - coeff_r[j];
+        const uint32_t diff = coeff[j] - coeff_r[j];
        const uint32_t error = diff * diff;
        EXPECT_GE(9u << 2 * (bit_depth_ - 8), error)
            << "Error: 8x8 DCT has error " << error
@@ -471,14 +464,14 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 10000;
    const int eob = 12;
-    DECLARE_ALIGNED(16, tran_low_t, coeff[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, dst[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint8_t, ref[kNumCoeffs]);
-#if CONFIG_VPX_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, ref, kNumCoeffs);
+#if CONFIG_VP9_HIGHBITDEPTH
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, ref16, kNumCoeffs);
 #endif
-    const int16_t *scan = vp10_default_scan_orders[TX_8X8].scan;
+    const int16_t *scan = vp9_default_scan_orders[TX_8X8].scan;

    for (int i = 0; i < count_test_block; ++i) {
      for (int j = 0; j < kNumCoeffs; ++j) {
@@ -491,7 +484,7 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
        if (bit_depth_ == VPX_BITS_8) {
          dst[j] = 0;
          ref[j] = 0;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
        } else {
          dst16[j] = 0;
          ref16[j] = 0;
@@ -501,7 +494,7 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
      if (bit_depth_ == VPX_BITS_8) {
        ref_txfm(coeff, ref, pitch_);
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
      } else {
        ref_txfm(coeff, CONVERT_TO_BYTEPTR(ref16), pitch_);
        ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
@@ -510,11 +503,11 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
      }

      for (int j = 0; j < kNumCoeffs; ++j) {
-#if CONFIG_VPX_HIGHBITDEPTH
-        const int diff =
+#if CONFIG_VP9_HIGHBITDEPTH
+        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - ref[j] : dst16[j] - ref16[j];
 #else
-        const int diff = dst[j] - ref[j];
+        const uint32_t diff = dst[j] - ref[j];
 #endif
        const uint32_t error = diff * diff;
        EXPECT_EQ(0u, error)
@@ -643,7 +636,7 @@ class InvTrans8x8DCT
  void RunInvTxfm(tran_low_t *out, uint8_t *dst, int stride) {
    inv_txfm_(out, dst, stride);
  }
-  void RunFwdTxfm(int16_t * /*out*/, tran_low_t * /*dst*/, int /*stride*/) {}
+  void RunFwdTxfm(int16_t *out, tran_low_t *dst, int stride) {}

  IdctFunc ref_txfm_;
  IdctFunc inv_txfm_;
@@ -656,104 +649,108 @@ TEST_P(InvTrans8x8DCT, CompareReference) {

 using std::tr1::make_tuple;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_c, &vpx_idct8x8_64_add_c, 0, VPX_BITS_8),
-        make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_12, 0, VPX_BITS_12)));
+        make_tuple(&vp9_fdct8x8_c, &vp9_idct8x8_64_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_highbd_fdct8x8_c, &idct8x8_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fdct8x8_c, &idct8x8_12, 0, VPX_BITS_12)));
 #else
 INSTANTIATE_TEST_CASE_P(
    C, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_c, &vpx_idct8x8_64_add_c, 0, VPX_BITS_8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fdct8x8_c, &vp9_idct8x8_64_add_c, 0, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    C, FwdTrans8x8HT,
    ::testing::Values(
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_10, 0, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_10, 1, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_10, 2, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_10, 3, VPX_BITS_10),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_12, 0, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_12, 1, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_12, 2, VPX_BITS_12),
-        make_tuple(&vp10_highbd_fht8x8_c, &iht8x8_12, 3, VPX_BITS_12),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 3, VPX_BITS_8)));
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_10, 0, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_10, 1, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_10, 2, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_10, 3, VPX_BITS_10),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_12, 0, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_12, 1, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_12, 2, VPX_BITS_12),
+        make_tuple(&vp9_highbd_fht8x8_c, &iht8x8_12, 3, VPX_BITS_12),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 3, VPX_BITS_8)));
 #else
+// TODO(jingning): re-enable after this handles the expanded range [0, 65535]
+// returned from Rand16().
 INSTANTIATE_TEST_CASE_P(
    C, FwdTrans8x8HT,
    ::testing::Values(
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_c, 3, VPX_BITS_8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 3, VPX_BITS_8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

-#if HAVE_NEON_ASM && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+// TODO(jingning): re-enable after this handles the expanded range [0, 65535]
+// returned from Rand16().
 INSTANTIATE_TEST_CASE_P(
    NEON, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_neon, &vpx_idct8x8_64_add_neon, 0,
+        make_tuple(&vp9_fdct8x8_neon, &vp9_idct8x8_64_add_neon, 0,
                   VPX_BITS_8)));
-#endif  // HAVE_NEON_ASM && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    NEON, FwdTrans8x8HT,
    ::testing::Values(
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_neon, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_neon, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_neon, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_c, &vp10_iht8x8_64_add_neon, 3, VPX_BITS_8)));
-#endif  // HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 3, VPX_BITS_8)));
+#endif  // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+// TODO(jingning): re-enable after these handle the expanded range [0, 65535]
+// returned from Rand16().
 INSTANTIATE_TEST_CASE_P(
    SSE2, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_sse2, &vpx_idct8x8_64_add_sse2, 0,
+        make_tuple(&vp9_fdct8x8_sse2, &vp9_idct8x8_64_add_sse2, 0,
                   VPX_BITS_8)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, FwdTrans8x8HT,
    ::testing::Values(
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_sse2,
-                   0, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_sse2,
-                   1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_sse2,
-                   2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_sse2,
-                   3, VPX_BITS_8)));
-#endif  // HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_sse2, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_sse2, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_sse2, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_sse2, 3, VPX_BITS_8)));
+#endif  // HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_sse2, &vpx_idct8x8_64_add_c, 0, VPX_BITS_8),
-        make_tuple(&vpx_highbd_fdct8x8_c,
+        make_tuple(&vp9_fdct8x8_sse2, &vp9_idct8x8_64_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_highbd_fdct8x8_c,
                   &idct8x8_64_add_10_sse2, 12, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct8x8_sse2,
+        make_tuple(&vp9_highbd_fdct8x8_sse2,
                   &idct8x8_64_add_10_sse2, 12, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct8x8_c,
+        make_tuple(&vp9_highbd_fdct8x8_c,
                   &idct8x8_64_add_12_sse2, 12, VPX_BITS_12),
-        make_tuple(&vpx_highbd_fdct8x8_sse2,
+        make_tuple(&vp9_highbd_fdct8x8_sse2,
                   &idct8x8_64_add_12_sse2, 12, VPX_BITS_12)));

+// TODO(jingning): re-enable after these handle the expanded range [0, 65535]
+// returned from Rand16().
 INSTANTIATE_TEST_CASE_P(
    SSE2, FwdTrans8x8HT,
    ::testing::Values(
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_c, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_c, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_c, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_sse2, &vp10_iht8x8_64_add_c, 3, VPX_BITS_8)));
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_c, 0, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_c, 1, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_c, 2, VPX_BITS_8),
+        make_tuple(&vp9_fht8x8_sse2, &vp9_iht8x8_64_add_c, 3, VPX_BITS_8)));

 // Optimizations take effect at a threshold of 6201, so we use a value close to
 // that to test both branches.
@@ -768,28 +765,16 @@ INSTANTIATE_TEST_CASE_P(
                   &idct8x8_10_add_12_sse2, 6225, VPX_BITS_12),
        make_tuple(&idct8x8_12,
                   &idct8x8_64_add_12_sse2, 6225, VPX_BITS_12)));
-#endif  // HAVE_SSE2 && CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSSE3 && ARCH_X86_64 && \
-    !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_VP9_HIGHBITDEPTH && \
+    !CONFIG_EMULATE_HARDWARE
+// TODO(jingning): re-enable after this handles the expanded range [0, 65535]
+// returned from Rand16().
 INSTANTIATE_TEST_CASE_P(
    SSSE3, FwdTrans8x8DCT,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_ssse3, &vpx_idct8x8_64_add_ssse3, 0,
+        make_tuple(&vp9_fdct8x8_ssse3, &vp9_idct8x8_64_add_ssse3, 0,
                   VPX_BITS_8)));
 #endif
-
-#if HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    MSA, FwdTrans8x8DCT,
-    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_msa, &vpx_idct8x8_64_add_msa, 0, VPX_BITS_8)));
-INSTANTIATE_TEST_CASE_P(
-    MSA, FwdTrans8x8HT,
-    ::testing::Values(
-        make_tuple(&vp10_fht8x8_msa, &vp10_iht8x8_64_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_msa, &vp10_iht8x8_64_add_msa, 1, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_msa, &vp10_iht8x8_64_add_msa, 2, VPX_BITS_8),
-        make_tuple(&vp10_fht8x8_msa, &vp10_iht8x8_64_add_msa, 3, VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/frame_size_tests.cc
+++ b/test/frame_size_tests.cc
@@ -17,7 +17,7 @@ class VP9FrameSizeTestsLarge
    : public ::libvpx_test::EncoderTest,
      public ::testing::Test {
 protected:
-  VP9FrameSizeTestsLarge() : EncoderTest(&::libvpx_test::kVP10),
+  VP9FrameSizeTestsLarge() : EncoderTest(&::libvpx_test::kVP9),
                             expected_res_(VPX_CODEC_OK) {}
  virtual ~VP9FrameSizeTestsLarge() {}

@@ -74,7 +74,7 @@ TEST_F(VP9FrameSizeTestsLarge, ValidSizes) {
  // size or almost 1 gig of memory.
  // In total the allocations will exceed 2GiB which may cause a failure with
  // mingw + wine, use a smaller size in that case.
-#if defined(_WIN32) && !defined(_WIN64) || defined(__OS2__)
+#if defined(_WIN32) && !defined(_WIN64)
  video.SetSize(4096, 3072);
 #else
  video.SetSize(4096, 4096);
--- a/test/function_equivalence_test.h
+++ b/test/function_equivalence_test.h
@@ -1,65 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#ifndef TEST_FUNCTION_EQUIVALENCE_TEST_H_
-#define TEST_FUNCTION_EQUIVALENCE_TEST_H_
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/acm_random.h"
-#include "test/clear_system_state.h"
-#include "test/util.h"
-
-using libvpx_test::ACMRandom;
-
-namespace libvpx_test {
-// Base class for tests that compare 2 implementations of the same function
-// for equivalence. The template parameter should be pointer to a function
-// that is being tested.
-//
-// The test takes a 3-parameters encapsulating struct 'FuncParam', containing:
-//   - Pointer to reference function
-//   - Pointer to tested function
-//   - Integer bit depth (default to 0).
-//
-// These values are then accessible in the tests as member of params_:
-// params_.ref_func, params_.tst_func, and params_.bit_depth.
-//
-
-template <typename T>
-struct FuncParam {
-  FuncParam(T ref = NULL, T tst = NULL, int bit_depth = 0)
-      : ref_func(ref), tst_func(tst), bit_depth(bit_depth) {}
-  T ref_func;
-  T tst_func;
-  int bit_depth;
-};
-
-template <typename T>
-class FunctionEquivalenceTest : public ::testing::TestWithParam<FuncParam<T> > {
- public:
-  FunctionEquivalenceTest() : rng_(ACMRandom::DeterministicSeed()) {}
-
-  virtual ~FunctionEquivalenceTest() {}
-
-  virtual void SetUp() {
-    params_ = this->GetParam();
-  }
-
-  virtual void TearDown() {
-    libvpx_test::ClearSystemState();
-  }
-
- protected:
-  ACMRandom rng_;
-  FuncParam<T> params_;
-};
-
-}   // namespace libvpx_test
-#endif  // TEST_FUNCTION_EQUIVALENCE_TEST_H_
--- a/test/hadamard_test.cc
+++ b/test/hadamard_test.cc
@@ -1,220 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <algorithm>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_dsp_rtcd.h"
-
-#include "test/acm_random.h"
-#include "test/register_state_check.h"
-
-namespace {
-
-using ::libvpx_test::ACMRandom;
-
-typedef void (*HadamardFunc)(const int16_t *a, int a_stride, int16_t *b);
-
-void hadamard_loop(const int16_t *a, int a_stride, int16_t *out) {
-  int16_t b[8];
-  for (int i = 0; i < 8; i += 2) {
-    b[i + 0] = a[i * a_stride] + a[(i + 1) * a_stride];
-    b[i + 1] = a[i * a_stride] - a[(i + 1) * a_stride];
-  }
-  int16_t c[8];
-  for (int i = 0; i < 8; i += 4) {
-    c[i + 0] = b[i + 0] + b[i + 2];
-    c[i + 1] = b[i + 1] + b[i + 3];
-    c[i + 2] = b[i + 0] - b[i + 2];
-    c[i + 3] = b[i + 1] - b[i + 3];
-  }
-  out[0] = c[0] + c[4];
-  out[7] = c[1] + c[5];
-  out[3] = c[2] + c[6];
-  out[4] = c[3] + c[7];
-  out[2] = c[0] - c[4];
-  out[6] = c[1] - c[5];
-  out[1] = c[2] - c[6];
-  out[5] = c[3] - c[7];
-}
-
-void reference_hadamard8x8(const int16_t *a, int a_stride, int16_t *b) {
-  int16_t buf[64];
-  for (int i = 0; i < 8; ++i) {
-    hadamard_loop(a + i, a_stride, buf + i * 8);
-  }
-
-  for (int i = 0; i < 8; ++i) {
-    hadamard_loop(buf + i, 8, b + i * 8);
-  }
-}
-
-void reference_hadamard16x16(const int16_t *a, int a_stride, int16_t *b) {
-  /* The source is a 16x16 block. The destination is rearranged to 8x32.
-   * Input is 9 bit. */
-  reference_hadamard8x8(a + 0 + 0 * a_stride, a_stride, b + 0);
-  reference_hadamard8x8(a + 8 + 0 * a_stride, a_stride, b + 64);
-  reference_hadamard8x8(a + 0 + 8 * a_stride, a_stride, b + 128);
-  reference_hadamard8x8(a + 8 + 8 * a_stride, a_stride, b + 192);
-
-  /* Overlay the 8x8 blocks and combine. */
-  for (int i = 0; i < 64; ++i) {
-    /* 8x8 steps the range up to 15 bits. */
-    const int16_t a0 = b[0];
-    const int16_t a1 = b[64];
-    const int16_t a2 = b[128];
-    const int16_t a3 = b[192];
-
-    /* Prevent the result from escaping int16_t. */
-    const int16_t b0 = (a0 + a1) >> 1;
-    const int16_t b1 = (a0 - a1) >> 1;
-    const int16_t b2 = (a2 + a3) >> 1;
-    const int16_t b3 = (a2 - a3) >> 1;
-
-    /* Store a 16 bit value. */
-    b[  0] = b0 + b2;
-    b[ 64] = b1 + b3;
-    b[128] = b0 - b2;
-    b[192] = b1 - b3;
-
-    ++b;
-  }
-}
-
-class HadamardTestBase : public ::testing::TestWithParam<HadamardFunc> {
- public:
-  virtual void SetUp() {
-    h_func_ = GetParam();
-    rnd_.Reset(ACMRandom::DeterministicSeed());
-  }
-
- protected:
-  HadamardFunc h_func_;
-  ACMRandom rnd_;
-};
-
-class Hadamard8x8Test : public HadamardTestBase {};
-
-TEST_P(Hadamard8x8Test, CompareReferenceRandom) {
-  DECLARE_ALIGNED(16, int16_t, a[64]);
-  DECLARE_ALIGNED(16, int16_t, b[64]);
-  int16_t b_ref[64];
-  for (int i = 0; i < 64; ++i) {
-    a[i] = rnd_.Rand9Signed();
-  }
-  memset(b, 0, sizeof(b));
-  memset(b_ref, 0, sizeof(b_ref));
-
-  reference_hadamard8x8(a, 8, b_ref);
-  ASM_REGISTER_STATE_CHECK(h_func_(a, 8, b));
-
-  // The order of the output is not important. Sort before checking.
-  std::sort(b, b + 64);
-  std::sort(b_ref, b_ref + 64);
-  EXPECT_EQ(0, memcmp(b, b_ref, sizeof(b)));
-}
-
-TEST_P(Hadamard8x8Test, VaryStride) {
-  DECLARE_ALIGNED(16, int16_t, a[64 * 8]);
-  DECLARE_ALIGNED(16, int16_t, b[64]);
-  int16_t b_ref[64];
-  for (int i = 0; i < 64 * 8; ++i) {
-    a[i] = rnd_.Rand9Signed();
-  }
-
-  for (int i = 8; i < 64; i += 8) {
-    memset(b, 0, sizeof(b));
-    memset(b_ref, 0, sizeof(b_ref));
-
-    reference_hadamard8x8(a, i, b_ref);
-    ASM_REGISTER_STATE_CHECK(h_func_(a, i, b));
-
-    // The order of the output is not important. Sort before checking.
-    std::sort(b, b + 64);
-    std::sort(b_ref, b_ref + 64);
-    EXPECT_EQ(0, memcmp(b, b_ref, sizeof(b)));
-  }
-}
-
-INSTANTIATE_TEST_CASE_P(C, Hadamard8x8Test,
-                        ::testing::Values(&vpx_hadamard_8x8_c));
-
-#if HAVE_SSE2
-INSTANTIATE_TEST_CASE_P(SSE2, Hadamard8x8Test,
-                        ::testing::Values(&vpx_hadamard_8x8_sse2));
-#endif  // HAVE_SSE2
-
-#if HAVE_SSSE3 && ARCH_X86_64
-INSTANTIATE_TEST_CASE_P(SSSE3, Hadamard8x8Test,
-                        ::testing::Values(&vpx_hadamard_8x8_ssse3));
-#endif  // HAVE_SSSE3 && ARCH_X86_64
-
-#if HAVE_NEON
-INSTANTIATE_TEST_CASE_P(NEON, Hadamard8x8Test,
-                        ::testing::Values(&vpx_hadamard_8x8_neon));
-#endif  // HAVE_NEON
-
-class Hadamard16x16Test : public HadamardTestBase {};
-
-TEST_P(Hadamard16x16Test, CompareReferenceRandom) {
-  DECLARE_ALIGNED(16, int16_t, a[16 * 16]);
-  DECLARE_ALIGNED(16, int16_t, b[16 * 16]);
-  int16_t b_ref[16 * 16];
-  for (int i = 0; i < 16 * 16; ++i) {
-    a[i] = rnd_.Rand9Signed();
-  }
-  memset(b, 0, sizeof(b));
-  memset(b_ref, 0, sizeof(b_ref));
-
-  reference_hadamard16x16(a, 16, b_ref);
-  ASM_REGISTER_STATE_CHECK(h_func_(a, 16, b));
-
-  // The order of the output is not important. Sort before checking.
-  std::sort(b, b + 16 * 16);
-  std::sort(b_ref, b_ref + 16 * 16);
-  EXPECT_EQ(0, memcmp(b, b_ref, sizeof(b)));
-}
-
-TEST_P(Hadamard16x16Test, VaryStride) {
-  DECLARE_ALIGNED(16, int16_t, a[16 * 16 * 8]);
-  DECLARE_ALIGNED(16, int16_t, b[16 * 16]);
-  int16_t b_ref[16 * 16];
-  for (int i = 0; i < 16 * 16 * 8; ++i) {
-    a[i] = rnd_.Rand9Signed();
-  }
-
-  for (int i = 8; i < 64; i += 8) {
-    memset(b, 0, sizeof(b));
-    memset(b_ref, 0, sizeof(b_ref));
-
-    reference_hadamard16x16(a, i, b_ref);
-    ASM_REGISTER_STATE_CHECK(h_func_(a, i, b));
-
-    // The order of the output is not important. Sort before checking.
-    std::sort(b, b + 16 * 16);
-    std::sort(b_ref, b_ref + 16 * 16);
-    EXPECT_EQ(0, memcmp(b, b_ref, sizeof(b)));
-  }
-}
-
-INSTANTIATE_TEST_CASE_P(C, Hadamard16x16Test,
-                        ::testing::Values(&vpx_hadamard_16x16_c));
-
-#if HAVE_SSE2
-INSTANTIATE_TEST_CASE_P(SSE2, Hadamard16x16Test,
-                        ::testing::Values(&vpx_hadamard_16x16_sse2));
-#endif  // HAVE_SSE2
-
-#if HAVE_NEON
-INSTANTIATE_TEST_CASE_P(NEON, Hadamard16x16Test,
-                        ::testing::Values(&vpx_hadamard_16x16_neon));
-#endif  // HAVE_NEON
-}  // namespace
--- a/test/hbd_metrics_test.cc
+++ b/test/hbd_metrics_test.cc
@@ -1,251 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <math.h>
-#include <stdlib.h>
-#include <new>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/acm_random.h"
-#include "test/util.h"
-#include "./vpx_config.h"
-#include "vpx_dsp/psnr.h"
-#include "vpx_dsp/ssim.h"
-#include "vpx_ports/mem.h"
-#include "vpx_ports/msvc.h"
-#include "vpx_scale/yv12config.h"
-
-
-using libvpx_test::ACMRandom;
-
-namespace {
-
-typedef double (*LBDMetricFunc)(const YV12_BUFFER_CONFIG *source,
-                                const YV12_BUFFER_CONFIG *dest);
-typedef double (*HBDMetricFunc)(const YV12_BUFFER_CONFIG *source,
-                                const YV12_BUFFER_CONFIG *dest,
-                                uint32_t in_bd, uint32_t bd);
-
-double compute_hbd_psnr(const YV12_BUFFER_CONFIG *source,
-                        const YV12_BUFFER_CONFIG *dest,
-                        uint32_t in_bd, uint32_t bd) {
-  PSNR_STATS psnr;
-  vpx_calc_highbd_psnr(source, dest, &psnr, bd, in_bd);
-  return psnr.psnr[0];
-}
-
-double compute_psnr(const YV12_BUFFER_CONFIG *source,
-  const YV12_BUFFER_CONFIG *dest) {
-  PSNR_STATS psnr;
-  vpx_calc_psnr(source, dest, &psnr);
-  return psnr.psnr[0];
-}
-
-double compute_hbd_psnrhvs(const YV12_BUFFER_CONFIG *source,
-                           const YV12_BUFFER_CONFIG *dest,
-                           uint32_t in_bd, uint32_t bd) {
-  double tempy, tempu, tempv;
-  return vpx_psnrhvs(source, dest,
-                     &tempy, &tempu, &tempv, bd, in_bd);
-}
-
-double compute_psnrhvs(const YV12_BUFFER_CONFIG *source,
-  const YV12_BUFFER_CONFIG *dest) {
-  double tempy, tempu, tempv;
-  return vpx_psnrhvs(source, dest,
-                     &tempy, &tempu, &tempv, 8, 8);
-}
-
-double compute_hbd_fastssim(const YV12_BUFFER_CONFIG *source,
-                            const YV12_BUFFER_CONFIG *dest,
-                            uint32_t in_bd, uint32_t bd) {
-  double tempy, tempu, tempv;
-  return vpx_calc_fastssim(source, dest,
-                               &tempy, &tempu, &tempv, bd, in_bd);
-}
-
-double compute_fastssim(const YV12_BUFFER_CONFIG *source,
-                        const YV12_BUFFER_CONFIG *dest) {
-  double tempy, tempu, tempv;
-  return vpx_calc_fastssim(source, dest,
-                           &tempy, &tempu, &tempv, 8, 8);
-}
-
-double compute_hbd_vpxssim(const YV12_BUFFER_CONFIG *source,
-                           const YV12_BUFFER_CONFIG *dest,
-                           uint32_t in_bd, uint32_t bd) {
-  double ssim, weight;
-  ssim = vpx_highbd_calc_ssim(source, dest, &weight, bd, in_bd);
-  return 100 * pow(ssim / weight, 8.0);
-}
-
-double compute_vpxssim(const YV12_BUFFER_CONFIG *source,
-  const YV12_BUFFER_CONFIG *dest) {
-  double ssim, weight;
-  ssim = vpx_calc_ssim(source, dest, &weight);
-  return 100 * pow(ssim / weight, 8.0);
-}
-
-
-class HBDMetricsTestBase {
- public:
-  virtual ~HBDMetricsTestBase() {}
-
- protected:
-  void RunAccuracyCheck() {
-    const int width = 1920;
-    const int height = 1080;
-    int i = 0;
-    const uint8_t kPixFiller = 128;
-    YV12_BUFFER_CONFIG lbd_src, lbd_dst;
-    YV12_BUFFER_CONFIG hbd_src, hbd_dst;
-    ACMRandom rnd(ACMRandom::DeterministicSeed());
-    double lbd_db, hbd_db;
-
-    memset(&lbd_src, 0, sizeof(lbd_src));
-    memset(&lbd_dst, 0, sizeof(lbd_dst));
-    memset(&hbd_src, 0, sizeof(hbd_src));
-    memset(&hbd_dst, 0, sizeof(hbd_dst));
-
-    vpx_alloc_frame_buffer(&lbd_src, width, height, 1, 1, 0, 32, 16);
-    vpx_alloc_frame_buffer(&lbd_dst, width, height, 1, 1, 0, 32, 16);
-    vpx_alloc_frame_buffer(&hbd_src, width, height, 1, 1, 1, 32, 16);
-    vpx_alloc_frame_buffer(&hbd_dst, width, height, 1, 1, 1, 32, 16);
-
-    memset(lbd_src.buffer_alloc, kPixFiller, lbd_src.buffer_alloc_sz);
-    while (i < lbd_src.buffer_alloc_sz) {
-      uint16_t spel, dpel;
-      spel = lbd_src.buffer_alloc[i];
-      // Create some distortion for dst buffer.
-      dpel = rnd.Rand8();
-      lbd_dst.buffer_alloc[i] = (uint8_t)dpel;
-      ((uint16_t*)(hbd_src.buffer_alloc))[i] = spel << (bit_depth_ - 8);
-      ((uint16_t*)(hbd_dst.buffer_alloc))[i] = dpel << (bit_depth_ - 8);
-      i++;
-    }
-
-    lbd_db = lbd_metric_(&lbd_src, &lbd_dst);
-    hbd_db = hbd_metric_(&hbd_src, &hbd_dst, input_bit_depth_, bit_depth_);
-    EXPECT_LE(fabs(lbd_db - hbd_db), threshold_);
-
-    i = 0;
-    while (i < lbd_src.buffer_alloc_sz) {
-      uint16_t dpel;
-      // Create some small distortion for dst buffer.
-      dpel = 120 + (rnd.Rand8() >> 4);
-      lbd_dst.buffer_alloc[i] = (uint8_t)dpel;
-      ((uint16_t*)(hbd_dst.buffer_alloc))[i] = dpel << (bit_depth_ - 8);
-      i++;
-    }
-
-    lbd_db = lbd_metric_(&lbd_src, &lbd_dst);
-    hbd_db = hbd_metric_(&hbd_src, &hbd_dst, input_bit_depth_, bit_depth_);
-    EXPECT_LE(fabs(lbd_db - hbd_db), threshold_);
-
-    i = 0;
-    while (i < lbd_src.buffer_alloc_sz) {
-      uint16_t dpel;
-      // Create some small distortion for dst buffer.
-      dpel = 126 + (rnd.Rand8() >> 6);
-      lbd_dst.buffer_alloc[i] = (uint8_t)dpel;
-      ((uint16_t*)(hbd_dst.buffer_alloc))[i] = dpel << (bit_depth_ - 8);
-      i++;
-    }
-
-    lbd_db = lbd_metric_(&lbd_src, &lbd_dst);
-    hbd_db = hbd_metric_(&hbd_src, &hbd_dst, input_bit_depth_, bit_depth_);
-    EXPECT_LE(fabs(lbd_db - hbd_db), threshold_);
-
-    vpx_free_frame_buffer(&lbd_src);
-    vpx_free_frame_buffer(&lbd_dst);
-    vpx_free_frame_buffer(&hbd_src);
-    vpx_free_frame_buffer(&hbd_dst);
-  }
-
-  int input_bit_depth_;
-  int bit_depth_;
-  double threshold_;
-  LBDMetricFunc lbd_metric_;
-  HBDMetricFunc hbd_metric_;
-};
-
-typedef std::tr1::tuple<LBDMetricFunc,
-                        HBDMetricFunc, int, int, double> MetricTestTParam;
-class HBDMetricsTest
-    : public HBDMetricsTestBase,
-      public ::testing::TestWithParam<MetricTestTParam> {
- public:
-  virtual void SetUp() {
-    lbd_metric_ = GET_PARAM(0);
-    hbd_metric_ = GET_PARAM(1);
-    input_bit_depth_ = GET_PARAM(2);
-    bit_depth_ = GET_PARAM(3);
-    threshold_ = GET_PARAM(4);
-  }
-  virtual void TearDown() {}
-};
-
-TEST_P(HBDMetricsTest, RunAccuracyCheck) {
-  RunAccuracyCheck();
-}
-
-// Allow small variation due to floating point operations.
-static const double kSsim_thresh = 0.001;
-// Allow some additional errors accumulated in floating point operations.
-static const double kFSsim_thresh = 0.03;
-// Allow some extra variation due to rounding error accumulated in dct.
-static const double kPhvs_thresh = 0.3;
-
-INSTANTIATE_TEST_CASE_P(
-    VPXSSIM, HBDMetricsTest,
-    ::testing::Values(
-        MetricTestTParam(&compute_vpxssim, &compute_hbd_vpxssim, 8, 10,
-                         kSsim_thresh),
-        MetricTestTParam(&compute_vpxssim, &compute_hbd_vpxssim, 10, 10,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_vpxssim, &compute_hbd_vpxssim, 8, 12,
-                         kSsim_thresh),
-        MetricTestTParam(&compute_vpxssim, &compute_hbd_vpxssim, 12, 12,
-                         kPhvs_thresh)));
-INSTANTIATE_TEST_CASE_P(
-    FASTSSIM, HBDMetricsTest,
-    ::testing::Values(
-        MetricTestTParam(&compute_fastssim, &compute_hbd_fastssim, 8, 10,
-                         kFSsim_thresh),
-        MetricTestTParam(&compute_fastssim, &compute_hbd_fastssim, 10, 10,
-                         kFSsim_thresh),
-        MetricTestTParam(&compute_fastssim, &compute_hbd_fastssim, 8, 12,
-                         kFSsim_thresh),
-        MetricTestTParam(&compute_fastssim, &compute_hbd_fastssim, 12, 12,
-                         kFSsim_thresh)));
-INSTANTIATE_TEST_CASE_P(
-    PSNRHVS, HBDMetricsTest,
-    ::testing::Values(
-        MetricTestTParam(&compute_psnrhvs, &compute_hbd_psnrhvs, 8, 10,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnrhvs, &compute_hbd_psnrhvs, 10, 10,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnrhvs, &compute_hbd_psnrhvs, 8, 12,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnrhvs, &compute_hbd_psnrhvs, 12, 12,
-                         kPhvs_thresh)));
-INSTANTIATE_TEST_CASE_P(
-    PSNR, HBDMetricsTest,
-    ::testing::Values(
-        MetricTestTParam(&compute_psnr, &compute_hbd_psnr, 8, 10,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnr, &compute_hbd_psnr, 10, 10,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnr, &compute_hbd_psnr, 8, 12,
-                         kPhvs_thresh),
-        MetricTestTParam(&compute_psnr, &compute_hbd_psnr, 12, 12,
-                         kPhvs_thresh)));
-}  // namespace
-
--- a/test/idct8x8_test.cc
+++ b/test/idct8x8_test.cc
@@ -14,15 +14,24 @@

 #include "third_party/googletest/src/include/gtest/gtest.h"

-#include "./vpx_dsp_rtcd.h"
+#include "./vp9_rtcd.h"
+
 #include "test/acm_random.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/msvc.h"  // for round()

 using libvpx_test::ACMRandom;

 namespace {

+#ifdef _MSC_VER
+static int round(double x) {
+  if (x < 0)
+    return static_cast<int>(ceil(x - 0.5));
+  else
+    return static_cast<int>(floor(x + 0.5));
+}
+#endif
+
 void reference_dct_1d(double input[8], double output[8]) {
  const double kPi = 3.141592653589793238462643383279502884;
  const double kInvSqrt2 = 0.707106781186547524400844362104;
@@ -59,6 +68,43 @@ void reference_dct_2d(int16_t input[64], double output[64]) {
    output[i] *= 2;
 }

+void reference_idct_1d(double input[8], double output[8]) {
+  const double kPi = 3.141592653589793238462643383279502884;
+  const double kSqrt2 = 1.414213562373095048801688724209698;
+  for (int k = 0; k < 8; k++) {
+    output[k] = 0.0;
+    for (int n = 0; n < 8; n++) {
+      output[k] += input[n]*cos(kPi*(2*k+1)*n/16.0);
+      if (n == 0)
+        output[k] = output[k]/kSqrt2;
+    }
+  }
+}
+
+void reference_idct_2d(double input[64], int16_t output[64]) {
+  double out[64], out2[64];
+  // First transform rows
+  for (int i = 0; i < 8; ++i) {
+    double temp_in[8], temp_out[8];
+    for (int j = 0; j < 8; ++j)
+      temp_in[j] = input[j + i*8];
+    reference_idct_1d(temp_in, temp_out);
+    for (int j = 0; j < 8; ++j)
+      out[j + i*8] = temp_out[j];
+  }
+  // Then transform columns
+  for (int i = 0; i < 8; ++i) {
+    double temp_in[8], temp_out[8];
+    for (int j = 0; j < 8; ++j)
+      temp_in[j] = out[j*8 + i];
+    reference_idct_1d(temp_in, temp_out);
+    for (int j = 0; j < 8; ++j)
+      out2[j*8 + i] = temp_out[j];
+  }
+  for (int i = 0; i < 64; ++i)
+    output[i] = round(out2[i]/32);
+}
+
 TEST(VP9Idct8x8Test, AccuracyCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = 10000;
@@ -78,8 +124,8 @@ TEST(VP9Idct8x8Test, AccuracyCheck) {

    reference_dct_2d(input, output_r);
    for (int j = 0; j < 64; ++j)
-      coeff[j] = static_cast<tran_low_t>(round(output_r[j]));
-    vpx_idct8x8_64_add_c(coeff, dst, 8);
+      coeff[j] = round(output_r[j]);
+    vp9_idct8x8_64_add_c(coeff, dst, 8);
    for (int j = 0; j < 64; ++j) {
      const int diff = dst[j] - src[j];
      const int error = diff * diff;
--- a/test/idct_test.cc
+++ b/test/idct_test.cc
@@ -9,11 +9,11 @@
 */

 #include "./vpx_config.h"
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
+#include "./vp8_rtcd.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
+
 #include "vpx/vpx_integer.h"

 typedef void (*IdctFunc)(int16_t *input, unsigned char *pred_ptr,
@@ -113,8 +113,4 @@ INSTANTIATE_TEST_CASE_P(C, IDCTTest, ::testing::Values(vp8_short_idct4x4llm_c));
 INSTANTIATE_TEST_CASE_P(MMX, IDCTTest,
                        ::testing::Values(vp8_short_idct4x4llm_mmx));
 #endif
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(MSA, IDCTTest,
-                        ::testing::Values(vp8_short_idct4x4llm_msa));
-#endif
 }
--- a/test/intrapred_test.cc
+++ b/test/intrapred_test.cc
@@ -1,5 +1,5 @@
 /*
- *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
 *
 *  Use of this source code is governed by a BSD-style license
 *  that can be found in the LICENSE file in the root of the source
@@ -8,205 +8,389 @@
 *  be found in the AUTHORS file in the root of the source tree.
 */

-#include <string>

-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
+#include <string.h>
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
-#include "test/util.h"
-#include "vp10/common/blockd.h"
-#include "vp10/common/pred_common.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
+
+#include "./vpx_config.h"
+#include "./vp8_rtcd.h"
+#include "vp8/common/blockd.h"
 #include "vpx_mem/vpx_mem.h"

 namespace {

 using libvpx_test::ACMRandom;

-const int count_test_block = 100000;
+class IntraPredBase {
+ public:
+  virtual ~IntraPredBase() { libvpx_test::ClearSystemState(); }

-typedef void (*IntraPred)(uint16_t* dst, ptrdiff_t stride,
-                          const uint16_t* above, const uint16_t* left,
-                          int bps);
+ protected:
+  void SetupMacroblock(MACROBLOCKD *mbptr,
+                       MODE_INFO *miptr,
+                       uint8_t *data,
+                       int block_size,
+                       int stride,
+                       int num_planes) {
+    mbptr_ = mbptr;
+    miptr_ = miptr;
+    mbptr_->up_available = 1;
+    mbptr_->left_available = 1;
+    mbptr_->mode_info_context = miptr_;
+    stride_ = stride;
+    block_size_ = block_size;
+    num_planes_ = num_planes;
+    for (int p = 0; p < num_planes; p++)
+      data_ptr_[p] = data + stride * (block_size + 1) * p +
+                     stride + block_size;
+  }

-struct IntraPredFunc {
-  IntraPredFunc(IntraPred pred = NULL, IntraPred ref = NULL,
-                int block_size_value = 0, int bit_depth_value = 0)
-      : pred_fn(pred), ref_fn(ref),
-        block_size(block_size_value), bit_depth(bit_depth_value) {}
+  void FillRandom() {
+    // Fill edges with random data
+    ACMRandom rnd(ACMRandom::DeterministicSeed());
+    for (int p = 0; p < num_planes_; p++) {
+      for (int x = -1 ; x <= block_size_; x++)
+        data_ptr_[p][x - stride_] = rnd.Rand8();
+      for (int y = 0; y < block_size_; y++)
+        data_ptr_[p][y * stride_ - 1] = rnd.Rand8();
+    }
+  }

-  IntraPred pred_fn;
-  IntraPred ref_fn;
-  int block_size;
-  int bit_depth;
+  virtual void Predict(MB_PREDICTION_MODE mode) = 0;
+
+  void SetLeftUnavailable() {
+    mbptr_->left_available = 0;
+    for (int p = 0; p < num_planes_; p++)
+      for (int i = -1; i < block_size_; ++i)
+        data_ptr_[p][stride_ * i - 1] = 129;
+  }
+
+  void SetTopUnavailable() {
+    mbptr_->up_available = 0;
+    for (int p = 0; p < num_planes_; p++)
+      memset(&data_ptr_[p][-1 - stride_], 127, block_size_ + 2);
+  }
+
+  void SetTopLeftUnavailable() {
+    SetLeftUnavailable();
+    SetTopUnavailable();
+  }
+
+  int BlockSizeLog2Min1() const {
+    switch (block_size_) {
+      case 16:
+        return 3;
+      case 8:
+        return 2;
+      default:
+        return 0;
+    }
+  }
+
+  // check DC prediction output against a reference
+  void CheckDCPrediction() const {
+    for (int p = 0; p < num_planes_; p++) {
+      // calculate expected DC
+      int expected;
+      if (mbptr_->up_available || mbptr_->left_available) {
+        int sum = 0, shift = BlockSizeLog2Min1() + mbptr_->up_available +
+                             mbptr_->left_available;
+        if (mbptr_->up_available)
+          for (int x = 0; x < block_size_; x++)
+            sum += data_ptr_[p][x - stride_];
+        if (mbptr_->left_available)
+          for (int y = 0; y < block_size_; y++)
+            sum += data_ptr_[p][y * stride_ - 1];
+        expected = (sum + (1 << (shift - 1))) >> shift;
+      } else {
+        expected = 0x80;
+      }
+      // check that all subsequent lines are equal to the first
+      for (int y = 1; y < block_size_; ++y)
+        ASSERT_EQ(0, memcmp(data_ptr_[p], &data_ptr_[p][y * stride_],
+                            block_size_));
+      // within the first line, ensure that each pixel has the same value
+      for (int x = 1; x < block_size_; ++x)
+        ASSERT_EQ(data_ptr_[p][0], data_ptr_[p][x]);
+      // now ensure that that pixel has the expected (DC) value
+      ASSERT_EQ(expected, data_ptr_[p][0]);
+    }
+  }
+
+  // check V prediction output against a reference
+  void CheckVPrediction() const {
+    // check that all lines equal the top border
+    for (int p = 0; p < num_planes_; p++)
+      for (int y = 0; y < block_size_; y++)
+        ASSERT_EQ(0, memcmp(&data_ptr_[p][-stride_],
+                            &data_ptr_[p][y * stride_], block_size_));
+  }
+
+  // check H prediction output against a reference
+  void CheckHPrediction() const {
+    // for each line, ensure that each pixel is equal to the left border
+    for (int p = 0; p < num_planes_; p++)
+      for (int y = 0; y < block_size_; y++)
+        for (int x = 0; x < block_size_; x++)
+          ASSERT_EQ(data_ptr_[p][-1 + y * stride_],
+                    data_ptr_[p][x + y * stride_]);
+  }
+
+  static int ClipByte(int value) {
+    if (value > 255)
+      return 255;
+    else if (value < 0)
+      return 0;
+    return value;
+  }
+
+  // check TM prediction output against a reference
+  void CheckTMPrediction() const {
+    for (int p = 0; p < num_planes_; p++)
+      for (int y = 0; y < block_size_; y++)
+        for (int x = 0; x < block_size_; x++) {
+          const int expected = ClipByte(data_ptr_[p][x - stride_]
+                                      + data_ptr_[p][stride_ * y - 1]
+                                      - data_ptr_[p][-1 - stride_]);
+          ASSERT_EQ(expected, data_ptr_[p][y * stride_ + x]);
+       }
+  }
+
+  // Actual test
+  void RunTest() {
+    {
+      SCOPED_TRACE("DC_PRED");
+      FillRandom();
+      Predict(DC_PRED);
+      CheckDCPrediction();
+    }
+    {
+      SCOPED_TRACE("DC_PRED LEFT");
+      FillRandom();
+      SetLeftUnavailable();
+      Predict(DC_PRED);
+      CheckDCPrediction();
+    }
+    {
+      SCOPED_TRACE("DC_PRED TOP");
+      FillRandom();
+      SetTopUnavailable();
+      Predict(DC_PRED);
+      CheckDCPrediction();
+    }
+    {
+      SCOPED_TRACE("DC_PRED TOP_LEFT");
+      FillRandom();
+      SetTopLeftUnavailable();
+      Predict(DC_PRED);
+      CheckDCPrediction();
+    }
+    {
+      SCOPED_TRACE("H_PRED");
+      FillRandom();
+      Predict(H_PRED);
+      CheckHPrediction();
+    }
+    {
+      SCOPED_TRACE("V_PRED");
+      FillRandom();
+      Predict(V_PRED);
+      CheckVPrediction();
+    }
+    {
+      SCOPED_TRACE("TM_PRED");
+      FillRandom();
+      Predict(TM_PRED);
+      CheckTMPrediction();
+    }
+  }
+
+  MACROBLOCKD *mbptr_;
+  MODE_INFO *miptr_;
+  uint8_t *data_ptr_[2];  // in the case of Y, only [0] is used
+  int stride_;
+  int block_size_;
+  int num_planes_;
 };

-class VP9IntraPredTest : public ::testing::TestWithParam<IntraPredFunc> {
+typedef void (*IntraPredYFunc)(MACROBLOCKD *x,
+                               uint8_t *yabove_row,
+                               uint8_t *yleft,
+                               int left_stride,
+                               uint8_t *ypred_ptr,
+                               int y_stride);
+
+class IntraPredYTest
+    : public IntraPredBase,
+      public ::testing::TestWithParam<IntraPredYFunc> {
 public:
-  void RunTest(uint16_t* left_col, uint16_t* above_data,
-               uint16_t* dst, uint16_t* ref_dst) {
-    ACMRandom rnd(ACMRandom::DeterministicSeed());
-    const int block_size = params_.block_size;
-    above_row_ = above_data + 16;
-    left_col_ = left_col;
-    dst_ = dst;
-    ref_dst_ = ref_dst;
-    int error_count = 0;
-    for (int i = 0; i < count_test_block; ++i) {
-      // Fill edges with random data, try first with saturated values.
-      for (int x = -1; x <= block_size * 2; x++) {
-        if (i == 0) {
-          above_row_[x] = mask_;
-        } else {
-          above_row_[x] = rnd.Rand16() & mask_;
-        }
-      }
-      for (int y = 0; y < block_size; y++) {
-        if (i == 0) {
-          left_col_[y] = mask_;
-        } else {
-          left_col_[y] = rnd.Rand16() & mask_;
-        }
-      }
-      Predict();
-      CheckPrediction(i, &error_count);
-    }
-    ASSERT_EQ(0, error_count);
+  static void SetUpTestCase() {
+    mb_ = reinterpret_cast<MACROBLOCKD*>(
+        vpx_memalign(32, sizeof(MACROBLOCKD)));
+    mi_ = reinterpret_cast<MODE_INFO*>(
+        vpx_memalign(32, sizeof(MODE_INFO)));
+    data_array_ = reinterpret_cast<uint8_t*>(
+        vpx_memalign(kDataAlignment, kDataBufferSize));
+  }
+
+  static void TearDownTestCase() {
+    vpx_free(data_array_);
+    vpx_free(mi_);
+    vpx_free(mb_);
+    data_array_ = NULL;
  }

 protected:
+  static const int kBlockSize = 16;
+  static const int kDataAlignment = 16;
+  static const int kStride = kBlockSize * 3;
+  // We use 48 so that the data pointer of the first pixel in each row of
+  // each macroblock is 16-byte aligned, and this gives us access to the
+  // top-left and top-right corner pixels belonging to the top-left/right
+  // macroblocks.
+  // We use 17 lines so we have one line above us for top-prediction.
+  static const int kDataBufferSize = kStride * (kBlockSize + 1);
+
  virtual void SetUp() {
-    params_ = GetParam();
-    stride_ = params_.block_size * 3;
-    mask_   = (1 << params_.bit_depth) - 1;
+    pred_fn_ = GetParam();
+    SetupMacroblock(mb_, mi_, data_array_, kBlockSize, kStride, 1);
  }

-  void Predict() {
-    const int bit_depth = params_.bit_depth;
-    params_.ref_fn(ref_dst_, stride_, above_row_, left_col_, bit_depth);
-    ASM_REGISTER_STATE_CHECK(params_.pred_fn(dst_, stride_,
-                                             above_row_, left_col_, bit_depth));
+  virtual void Predict(MB_PREDICTION_MODE mode) {
+    mbptr_->mode_info_context->mbmi.mode = mode;
+    ASM_REGISTER_STATE_CHECK(pred_fn_(mbptr_,
+                                      data_ptr_[0] - kStride,
+                                      data_ptr_[0] - 1, kStride,
+                                      data_ptr_[0], kStride));
  }

-  void CheckPrediction(int test_case_number, int *error_count) const {
-    // For each pixel ensure that the calculated value is the same as reference.
-    const int block_size = params_.block_size;
-    for (int y = 0; y < block_size; y++) {
-      for (int x = 0; x < block_size; x++) {
-        *error_count += ref_dst_[x + y * stride_] != dst_[x + y * stride_];
-        if (*error_count == 1) {
-          ASSERT_EQ(ref_dst_[x + y * stride_], dst_[x + y * stride_])
-              << " Failed on Test Case Number "<< test_case_number;
-        }
-      }
-    }
-  }
-
-  uint16_t *above_row_;
-  uint16_t *left_col_;
-  uint16_t *dst_;
-  uint16_t *ref_dst_;
-  ptrdiff_t stride_;
-  int mask_;
-
-  IntraPredFunc params_;
+  IntraPredYFunc pred_fn_;
+  static uint8_t* data_array_;
+  static MACROBLOCKD * mb_;
+  static MODE_INFO *mi_;
 };

-TEST_P(VP9IntraPredTest, IntraPredTests) {
-  // max block size is 32
-  DECLARE_ALIGNED(16, uint16_t, left_col[2*32]);
-  DECLARE_ALIGNED(16, uint16_t, above_data[2*32+32]);
-  DECLARE_ALIGNED(16, uint16_t, dst[3 * 32 * 32]);
-  DECLARE_ALIGNED(16, uint16_t, ref_dst[3 * 32 * 32]);
-  RunTest(left_col, above_data, dst, ref_dst);
+MACROBLOCKD* IntraPredYTest::mb_ = NULL;
+MODE_INFO* IntraPredYTest::mi_ = NULL;
+uint8_t* IntraPredYTest::data_array_ = NULL;
+
+TEST_P(IntraPredYTest, IntraPredTests) {
+  RunTest();
 }

+INSTANTIATE_TEST_CASE_P(C, IntraPredYTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mby_s_c));
 #if HAVE_SSE2
-#if CONFIG_VPX_HIGHBITDEPTH
-INSTANTIATE_TEST_CASE_P(SSE2_TO_C_8, VP9IntraPredTest,
-    ::testing::Values(
-      IntraPredFunc(&vpx_highbd_dc_predictor_32x32_sse2,
-                    &vpx_highbd_dc_predictor_32x32_c, 32, 8),
-      IntraPredFunc(&vpx_highbd_tm_predictor_16x16_sse2,
-                    &vpx_highbd_tm_predictor_16x16_c, 16, 8),
-      IntraPredFunc(&vpx_highbd_tm_predictor_32x32_sse2,
-                    &vpx_highbd_tm_predictor_32x32_c, 32, 8),
-      IntraPredFunc(&vpx_highbd_dc_predictor_4x4_sse2,
-                    &vpx_highbd_dc_predictor_4x4_c, 4, 8),
-      IntraPredFunc(&vpx_highbd_dc_predictor_8x8_sse2,
-                    &vpx_highbd_dc_predictor_8x8_c, 8, 8),
-      IntraPredFunc(&vpx_highbd_dc_predictor_16x16_sse2,
-                    &vpx_highbd_dc_predictor_16x16_c, 16, 8),
-      IntraPredFunc(&vpx_highbd_v_predictor_4x4_sse2,
-                    &vpx_highbd_v_predictor_4x4_c, 4, 8),
-      IntraPredFunc(&vpx_highbd_v_predictor_8x8_sse2,
-                    &vpx_highbd_v_predictor_8x8_c, 8, 8),
-      IntraPredFunc(&vpx_highbd_v_predictor_16x16_sse2,
-                    &vpx_highbd_v_predictor_16x16_c, 16, 8),
-      IntraPredFunc(&vpx_highbd_v_predictor_32x32_sse2,
-                    &vpx_highbd_v_predictor_32x32_c, 32, 8),
-      IntraPredFunc(&vpx_highbd_tm_predictor_4x4_sse2,
-                    &vpx_highbd_tm_predictor_4x4_c, 4, 8),
-      IntraPredFunc(&vpx_highbd_tm_predictor_8x8_sse2,
-                    &vpx_highbd_tm_predictor_8x8_c, 8, 8)));
+INSTANTIATE_TEST_CASE_P(SSE2, IntraPredYTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mby_s_sse2));
+#endif
+#if HAVE_SSSE3
+INSTANTIATE_TEST_CASE_P(SSSE3, IntraPredYTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mby_s_ssse3));
+#endif
+#if HAVE_NEON
+INSTANTIATE_TEST_CASE_P(NEON, IntraPredYTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mby_s_neon));
+#endif

-INSTANTIATE_TEST_CASE_P(SSE2_TO_C_10, VP9IntraPredTest,
-    ::testing::Values(
-      IntraPredFunc(&vpx_highbd_dc_predictor_32x32_sse2,
-                    &vpx_highbd_dc_predictor_32x32_c, 32, 10),
-      IntraPredFunc(&vpx_highbd_tm_predictor_16x16_sse2,
-                    &vpx_highbd_tm_predictor_16x16_c, 16, 10),
-      IntraPredFunc(&vpx_highbd_tm_predictor_32x32_sse2,
-                    &vpx_highbd_tm_predictor_32x32_c, 32, 10),
-      IntraPredFunc(&vpx_highbd_dc_predictor_4x4_sse2,
-                    &vpx_highbd_dc_predictor_4x4_c, 4, 10),
-      IntraPredFunc(&vpx_highbd_dc_predictor_8x8_sse2,
-                    &vpx_highbd_dc_predictor_8x8_c, 8, 10),
-      IntraPredFunc(&vpx_highbd_dc_predictor_16x16_sse2,
-                    &vpx_highbd_dc_predictor_16x16_c, 16, 10),
-      IntraPredFunc(&vpx_highbd_v_predictor_4x4_sse2,
-                    &vpx_highbd_v_predictor_4x4_c, 4, 10),
-      IntraPredFunc(&vpx_highbd_v_predictor_8x8_sse2,
-                    &vpx_highbd_v_predictor_8x8_c, 8, 10),
-      IntraPredFunc(&vpx_highbd_v_predictor_16x16_sse2,
-                    &vpx_highbd_v_predictor_16x16_c, 16, 10),
-      IntraPredFunc(&vpx_highbd_v_predictor_32x32_sse2,
-                    &vpx_highbd_v_predictor_32x32_c, 32, 10),
-      IntraPredFunc(&vpx_highbd_tm_predictor_4x4_sse2,
-                    &vpx_highbd_tm_predictor_4x4_c, 4, 10),
-      IntraPredFunc(&vpx_highbd_tm_predictor_8x8_sse2,
-                    &vpx_highbd_tm_predictor_8x8_c, 8, 10)));
+typedef void (*IntraPredUvFunc)(MACROBLOCKD *x,
+                                uint8_t *uabove_row,
+                                uint8_t *vabove_row,
+                                uint8_t *uleft,
+                                uint8_t *vleft,
+                                int left_stride,
+                                uint8_t *upred_ptr,
+                                uint8_t *vpred_ptr,
+                                int pred_stride);

-INSTANTIATE_TEST_CASE_P(SSE2_TO_C_12, VP9IntraPredTest,
-    ::testing::Values(
-      IntraPredFunc(&vpx_highbd_dc_predictor_32x32_sse2,
-                    &vpx_highbd_dc_predictor_32x32_c, 32, 12),
-      IntraPredFunc(&vpx_highbd_tm_predictor_16x16_sse2,
-                    &vpx_highbd_tm_predictor_16x16_c, 16, 12),
-      IntraPredFunc(&vpx_highbd_tm_predictor_32x32_sse2,
-                    &vpx_highbd_tm_predictor_32x32_c, 32, 12),
-      IntraPredFunc(&vpx_highbd_dc_predictor_4x4_sse2,
-                    &vpx_highbd_dc_predictor_4x4_c, 4, 12),
-      IntraPredFunc(&vpx_highbd_dc_predictor_8x8_sse2,
-                    &vpx_highbd_dc_predictor_8x8_c, 8, 12),
-      IntraPredFunc(&vpx_highbd_dc_predictor_16x16_sse2,
-                    &vpx_highbd_dc_predictor_16x16_c, 16, 12),
-      IntraPredFunc(&vpx_highbd_v_predictor_4x4_sse2,
-                    &vpx_highbd_v_predictor_4x4_c, 4, 12),
-      IntraPredFunc(&vpx_highbd_v_predictor_8x8_sse2,
-                    &vpx_highbd_v_predictor_8x8_c, 8, 12),
-      IntraPredFunc(&vpx_highbd_v_predictor_16x16_sse2,
-                    &vpx_highbd_v_predictor_16x16_c, 16, 12),
-      IntraPredFunc(&vpx_highbd_v_predictor_32x32_sse2,
-                    &vpx_highbd_v_predictor_32x32_c, 32, 12),
-      IntraPredFunc(&vpx_highbd_tm_predictor_4x4_sse2,
-                    &vpx_highbd_tm_predictor_4x4_c, 4, 12),
-      IntraPredFunc(&vpx_highbd_tm_predictor_8x8_sse2,
-                    &vpx_highbd_tm_predictor_8x8_c, 8, 12)));
+class IntraPredUVTest
+    : public IntraPredBase,
+      public ::testing::TestWithParam<IntraPredUvFunc> {
+ public:
+  static void SetUpTestCase() {
+    mb_ = reinterpret_cast<MACROBLOCKD*>(
+        vpx_memalign(32, sizeof(MACROBLOCKD)));
+    mi_ = reinterpret_cast<MODE_INFO*>(
+        vpx_memalign(32, sizeof(MODE_INFO)));
+    data_array_ = reinterpret_cast<uint8_t*>(
+        vpx_memalign(kDataAlignment, kDataBufferSize));
+  }
+
+  static void TearDownTestCase() {
+    vpx_free(data_array_);
+    vpx_free(mi_);
+    vpx_free(mb_);
+    data_array_ = NULL;
+  }
+
+ protected:
+  static const int kBlockSize = 8;
+  static const int kDataAlignment = 8;
+  static const int kStride = kBlockSize * 3;
+  // We use 24 so that the data pointer of the first pixel in each row of
+  // each macroblock is 8-byte aligned, and this gives us access to the
+  // top-left and top-right corner pixels belonging to the top-left/right
+  // macroblocks.
+  // We use 9 lines so we have one line above us for top-prediction.
+  // [0] = U, [1] = V
+  static const int kDataBufferSize = 2 * kStride * (kBlockSize + 1);
+
+  virtual void SetUp() {
+    pred_fn_ = GetParam();
+    SetupMacroblock(mb_, mi_, data_array_, kBlockSize, kStride, 2);
+  }
+
+  virtual void Predict(MB_PREDICTION_MODE mode) {
+    mbptr_->mode_info_context->mbmi.uv_mode = mode;
+    pred_fn_(mbptr_, data_ptr_[0] - kStride, data_ptr_[1] - kStride,
+             data_ptr_[0] - 1, data_ptr_[1] - 1, kStride,
+             data_ptr_[0], data_ptr_[1], kStride);
+  }
+
+  IntraPredUvFunc pred_fn_;
+  // We use 24 so that the data pointer of the first pixel in each row of
+  // each macroblock is 8-byte aligned, and this gives us access to the
+  // top-left and top-right corner pixels belonging to the top-left/right
+  // macroblocks.
+  // We use 9 lines so we have one line above us for top-prediction.
+  // [0] = U, [1] = V
+  static uint8_t* data_array_;
+  static MACROBLOCKD* mb_;
+  static MODE_INFO* mi_;
+};
+
+MACROBLOCKD* IntraPredUVTest::mb_ = NULL;
+MODE_INFO* IntraPredUVTest::mi_ = NULL;
+uint8_t* IntraPredUVTest::data_array_ = NULL;
+
+TEST_P(IntraPredUVTest, IntraPredTests) {
+  RunTest();
+}
+
+INSTANTIATE_TEST_CASE_P(C, IntraPredUVTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mbuv_s_c));
+#if HAVE_SSE2
+INSTANTIATE_TEST_CASE_P(SSE2, IntraPredUVTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mbuv_s_sse2));
+#endif
+#if HAVE_SSSE3
+INSTANTIATE_TEST_CASE_P(SSSE3, IntraPredUVTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mbuv_s_ssse3));
+#endif
+#if HAVE_NEON
+INSTANTIATE_TEST_CASE_P(NEON, IntraPredUVTest,
+                        ::testing::Values(
+                            vp8_build_intra_predictors_mbuv_s_neon));
+#endif

-#endif  // CONFIG_VPX_HIGHBITDEPTH
-#endif  // HAVE_SSE2
 }  // namespace
--- a/test/invalid_file_test.cc
+++ b/test/invalid_file_test.cc
@@ -0,0 +1,167 @@
+/*
+ *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include <cstdio>
+#include <cstdlib>
+#include <string>
+#include <vector>
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "./vpx_config.h"
+#include "test/codec_factory.h"
+#include "test/decode_test_driver.h"
+#include "test/ivf_video_source.h"
+#include "test/util.h"
+#if CONFIG_WEBM_IO
+#include "test/webm_video_source.h"
+#endif
+#include "vpx_mem/vpx_mem.h"
+
+namespace {
+
+struct DecodeParam {
+  int threads;
+  const char *filename;
+};
+
+std::ostream &operator<<(std::ostream &os, const DecodeParam &dp) {
+  return os << "threads: " << dp.threads << " file: " << dp.filename;
+}
+
+class InvalidFileTest
+    : public ::libvpx_test::DecoderTest,
+      public ::libvpx_test::CodecTestWithParam<DecodeParam> {
+ protected:
+  InvalidFileTest() : DecoderTest(GET_PARAM(0)), res_file_(NULL) {}
+
+  virtual ~InvalidFileTest() {
+    if (res_file_ != NULL)
+      fclose(res_file_);
+  }
+
+  void OpenResFile(const std::string &res_file_name_) {
+    res_file_ = libvpx_test::OpenTestDataFile(res_file_name_);
+    ASSERT_TRUE(res_file_ != NULL) << "Result file open failed. Filename: "
+        << res_file_name_;
+  }
+
+  virtual bool HandleDecodeResult(
+      const vpx_codec_err_t res_dec,
+      const libvpx_test::CompressedVideoSource &video,
+      libvpx_test::Decoder *decoder) {
+    EXPECT_TRUE(res_file_ != NULL);
+    int expected_res_dec;
+
+    // Read integer result.
+    const int res = fscanf(res_file_, "%d", &expected_res_dec);
+    EXPECT_NE(res, EOF) << "Read result data failed";
+
+    // Check results match.
+    EXPECT_EQ(expected_res_dec, res_dec)
+        << "Results don't match: frame number = " << video.frame_number()
+        << ". (" << decoder->DecodeError() << ")";
+
+    return !HasFailure();
+  }
+
+  void RunTest() {
+    const DecodeParam input = GET_PARAM(1);
+    libvpx_test::CompressedVideoSource *video = NULL;
+    vpx_codec_dec_cfg_t cfg = vpx_codec_dec_cfg_t();
+    cfg.threads = input.threads;
+    const std::string filename = input.filename;
+
+    // Open compressed video file.
+    if (filename.substr(filename.length() - 3, 3) == "ivf") {
+      video = new libvpx_test::IVFVideoSource(filename);
+    } else if (filename.substr(filename.length() - 4, 4) == "webm") {
+#if CONFIG_WEBM_IO
+      video = new libvpx_test::WebMVideoSource(filename);
+#else
+      fprintf(stderr, "WebM IO is disabled, skipping test vector %s\n",
+              filename.c_str());
+      return;
+#endif
+    }
+    video->Init();
+
+    // Construct result file name. The file holds a list of expected integer
+    // results, one for each decoded frame.  Any result that doesn't match
+    // the files list will cause a test failure.
+    const std::string res_filename = filename + ".res";
+    OpenResFile(res_filename);
+
+    // Decode frame, and check the md5 matching.
+    ASSERT_NO_FATAL_FAILURE(RunLoop(video, cfg));
+    delete video;
+  }
+
+ private:
+  FILE *res_file_;
+};
+
+TEST_P(InvalidFileTest, ReturnCode) {
+  RunTest();
+}
+
+const DecodeParam kVP9InvalidFileTests[] = {
+  {1, "invalid-vp90-02-v2.webm"},
+  {1, "invalid-vp90-2-00-quantizer-00.webm.ivf.s5861_r01-05_b6-.v2.ivf"},
+  {1, "invalid-vp90-03-v3.webm"},
+  {1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-.ivf"},
+  {1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-z.ivf"},
+  {1, "invalid-vp90-2-12-droppable_1.ivf.s3676_r01-05_b6-.ivf"},
+  {1, "invalid-vp90-2-05-resize.ivf.s59293_r01-05_b6-.ivf"},
+  {1, "invalid-vp90-2-09-subpixel-00.ivf.s20492_r01-05_b6-.v2.ivf"},
+  {1, "invalid-vp91-2-mixedrefcsp-444to420.ivf"},
+  {1, "invalid-vp90-2-12-droppable_1.ivf.s73804_r01-05_b6-.ivf"},
+  {1, "invalid-vp90-2-03-size-224x196.webm.ivf.s44156_r01-05_b6-.ivf"},
+  {1, "invalid-vp90-2-03-size-202x210.webm.ivf.s113306_r01-05_b6-.ivf"},
+};
+
+VP9_INSTANTIATE_TEST_CASE(InvalidFileTest,
+                          ::testing::ValuesIn(kVP9InvalidFileTests));
+
+// This class will include test vectors that are expected to fail
+// peek. However they are still expected to have no fatal failures.
+class InvalidFileInvalidPeekTest : public InvalidFileTest {
+ protected:
+  InvalidFileInvalidPeekTest() : InvalidFileTest() {}
+  virtual void HandlePeekResult(libvpx_test::Decoder *const /*decoder*/,
+                                libvpx_test::CompressedVideoSource* /*video*/,
+                                const vpx_codec_err_t /*res_peek*/) {}
+};
+
+TEST_P(InvalidFileInvalidPeekTest, ReturnCode) {
+  RunTest();
+}
+
+const DecodeParam kVP9InvalidFileInvalidPeekTests[] = {
+  {1, "invalid-vp90-01-v2.webm"},
+};
+
+VP9_INSTANTIATE_TEST_CASE(InvalidFileInvalidPeekTest,
+                          ::testing::ValuesIn(kVP9InvalidFileInvalidPeekTests));
+
+const DecodeParam kMultiThreadedVP9InvalidFileTests[] = {
+  {4, "invalid-vp90-2-08-tile_1x4_frame_parallel_all_key.webm"},
+  {4, "invalid-"
+      "vp90-2-08-tile_1x2_frame_parallel.webm.ivf.s47039_r01-05_b6-.ivf"},
+  {4, "invalid-vp90-2-08-tile_1x8_frame_parallel.webm.ivf.s288_r01-05_b6-.ivf"},
+  {2, "invalid-vp90-2-09-aq2.webm.ivf.s3984_r01-05_b6-.v2.ivf"},
+  {4, "invalid-vp90-2-09-subpixel-00.ivf.s19552_r01-05_b6-.v2.ivf"},
+};
+
+INSTANTIATE_TEST_CASE_P(
+    VP9MultiThreaded, InvalidFileTest,
+    ::testing::Combine(
+        ::testing::Values(
+            static_cast<const libvpx_test::CodecFactory*>(&libvpx_test::kVP9)),
+        ::testing::ValuesIn(kMultiThreadedVP9InvalidFileTests)));
+}  // namespace
--- a/test/keyframe_test.cc
+++ b/test/keyframe_test.cc
@@ -0,0 +1,145 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+#include <climits>
+#include <vector>
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "test/codec_factory.h"
+#include "test/encode_test_driver.h"
+#include "test/i420_video_source.h"
+#include "test/util.h"
+
+namespace {
+
+class KeyframeTest : public ::libvpx_test::EncoderTest,
+    public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
+ protected:
+  KeyframeTest() : EncoderTest(GET_PARAM(0)) {}
+  virtual ~KeyframeTest() {}
+
+  virtual void SetUp() {
+    InitializeConfig();
+    SetMode(GET_PARAM(1));
+    kf_count_ = 0;
+    kf_count_max_ = INT_MAX;
+    kf_do_force_kf_ = false;
+    set_cpu_used_ = 0;
+  }
+
+  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
+                                  ::libvpx_test::Encoder *encoder) {
+    if (kf_do_force_kf_)
+      frame_flags_ = (video->frame() % 3) ? 0 : VPX_EFLAG_FORCE_KF;
+    if (set_cpu_used_ && video->frame() == 1)
+      encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
+  }
+
+  virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
+    if (pkt->data.frame.flags & VPX_FRAME_IS_KEY) {
+      kf_pts_list_.push_back(pkt->data.frame.pts);
+      kf_count_++;
+      abort_ |= kf_count_ > kf_count_max_;
+    }
+  }
+
+  bool kf_do_force_kf_;
+  int kf_count_;
+  int kf_count_max_;
+  std::vector<vpx_codec_pts_t> kf_pts_list_;
+  int set_cpu_used_;
+};
+
+TEST_P(KeyframeTest, TestRandomVideoSource) {
+  // Validate that encoding the RandomVideoSource produces multiple keyframes.
+  // This validates the results of the TestDisableKeyframes test.
+  kf_count_max_ = 2;  // early exit successful tests.
+
+  ::libvpx_test::RandomVideoSource video;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  // In realtime mode - auto placed keyframes are exceedingly rare,  don't
+  // bother with this check   if(GetParam() > 0)
+  if (GET_PARAM(1) > 0)
+    EXPECT_GT(kf_count_, 1);
+}
+
+TEST_P(KeyframeTest, TestDisableKeyframes) {
+  cfg_.kf_mode = VPX_KF_DISABLED;
+  kf_count_max_ = 1;  // early exit failed tests.
+
+  ::libvpx_test::RandomVideoSource video;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  EXPECT_EQ(1, kf_count_);
+}
+
+TEST_P(KeyframeTest, TestForceKeyframe) {
+  cfg_.kf_mode = VPX_KF_DISABLED;
+  kf_do_force_kf_ = true;
+
+  ::libvpx_test::DummyVideoSource video;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  // verify that every third frame is a keyframe.
+  for (std::vector<vpx_codec_pts_t>::const_iterator iter = kf_pts_list_.begin();
+       iter != kf_pts_list_.end(); ++iter) {
+    ASSERT_EQ(0, *iter % 3) << "Unexpected keyframe at frame " << *iter;
+  }
+}
+
+TEST_P(KeyframeTest, TestKeyframeMaxDistance) {
+  cfg_.kf_max_dist = 25;
+
+  ::libvpx_test::DummyVideoSource video;
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  // verify that keyframe interval matches kf_max_dist
+  for (std::vector<vpx_codec_pts_t>::const_iterator iter = kf_pts_list_.begin();
+       iter != kf_pts_list_.end(); ++iter) {
+    ASSERT_EQ(0, *iter % 25) << "Unexpected keyframe at frame " << *iter;
+  }
+}
+
+TEST_P(KeyframeTest, TestAutoKeyframe) {
+  cfg_.kf_mode = VPX_KF_AUTO;
+  kf_do_force_kf_ = false;
+
+  // Force a deterministic speed step in Real Time mode, as the faster modes
+  // may not produce a keyframe like we expect. This is necessary when running
+  // on very slow environments (like Valgrind). The step -11 was determined
+  // experimentally as the fastest mode that still throws the keyframe.
+  if (deadline_ == VPX_DL_REALTIME)
+    set_cpu_used_ = -11;
+
+  // This clip has a cut scene every 30 frames -> Frame 0, 30, 60, 90, 120.
+  // I check only the first 40 frames to make sure there's a keyframe at frame
+  // 0 and 30.
+  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
+                                       30, 1, 0, 40);
+
+  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
+
+  // In realtime mode - auto placed keyframes are exceedingly rare,  don't
+  // bother with this check
+  if (GET_PARAM(1) > 0)
+    EXPECT_EQ(2u, kf_pts_list_.size()) << " Not the right number of keyframes ";
+
+  // Verify that keyframes match the file keyframes in the file.
+  for (std::vector<vpx_codec_pts_t>::const_iterator iter = kf_pts_list_.begin();
+       iter != kf_pts_list_.end(); ++iter) {
+    if (deadline_ == VPX_DL_REALTIME && *iter > 0)
+      EXPECT_EQ(0, (*iter - 1) % 30) << "Unexpected keyframe at frame "
+        << *iter;
+    else
+      EXPECT_EQ(0, *iter % 30) << "Unexpected keyframe at frame " << *iter;
+  }
+}
+
+VP8_INSTANTIATE_TEST_CASE(KeyframeTest, ALL_TEST_MODES);
+}  // namespace
--- a/test/level_test.cc
+++ b/test/level_test.cc
@@ -1,119 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/i420_video_source.h"
-#include "test/util.h"
-
-namespace {
-class LevelTest
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- protected:
-  LevelTest()
-     : EncoderTest(GET_PARAM(0)),
-       encoding_mode_(GET_PARAM(1)),
-       cpu_used_(GET_PARAM(2)),
-       min_gf_internal_(24),
-       target_level_(0),
-       level_(0) {}
-  virtual ~LevelTest() {}
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(encoding_mode_);
-    if (encoding_mode_ != ::libvpx_test::kRealTime) {
-      cfg_.g_lag_in_frames = 25;
-      cfg_.rc_end_usage = VPX_VBR;
-    } else {
-      cfg_.g_lag_in_frames = 0;
-      cfg_.rc_end_usage = VPX_CBR;
-    }
-    cfg_.rc_2pass_vbr_minsection_pct = 5;
-    cfg_.rc_2pass_vbr_maxsection_pct = 2000;
-    cfg_.rc_target_bitrate = 400;
-    cfg_.rc_max_quantizer = 63;
-    cfg_.rc_min_quantizer = 0;
-  }
-
-  virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
-                                  ::libvpx_test::Encoder *encoder) {
-    if (video->frame() == 0) {
-      encoder->Control(VP8E_SET_CPUUSED, cpu_used_);
-      encoder->Control(VP9E_SET_TARGET_LEVEL, target_level_);
-      encoder->Control(VP9E_SET_MIN_GF_INTERVAL, min_gf_internal_);
-      if (encoding_mode_ != ::libvpx_test::kRealTime) {
-        encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
-        encoder->Control(VP8E_SET_ARNR_MAXFRAMES, 7);
-        encoder->Control(VP8E_SET_ARNR_STRENGTH, 5);
-        encoder->Control(VP8E_SET_ARNR_TYPE, 3);
-      }
-    }
-    encoder->Control(VP9E_GET_LEVEL, &level_);
-    ASSERT_LE(level_, 51);
-    ASSERT_GE(level_, 0);
-  }
-
-  ::libvpx_test::TestMode encoding_mode_;
-  int cpu_used_;
-  int min_gf_internal_;
-  int target_level_;
-  int level_;
-};
-
-// Test for keeping level stats only
-TEST_P(LevelTest, TestTargetLevel0) {
-  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
-                                       40);
-  target_level_ = 0;
-  min_gf_internal_ = 4;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_EQ(11, level_);
-
-  cfg_.rc_target_bitrate = 1600;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_EQ(20, level_);
-}
-
-// Test for level control being turned off
-TEST_P(LevelTest, TestTargetLevel255) {
-  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
-                                       30);
-  target_level_ = 255;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-}
-
-TEST_P(LevelTest, TestTargetLevelApi) {
-  ::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0, 1);
-  static const vpx_codec_iface_t *codec = &vpx_codec_vp9_cx_algo;
-  vpx_codec_ctx_t enc;
-  vpx_codec_enc_cfg_t cfg;
-  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_config_default(codec, &cfg, 0));
-  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_init(&enc, codec, &cfg, 0));
-  for (int level = 0; level <= 256; ++level) {
-    if (level == 10 || level == 11 || level == 20 || level == 21 ||
-        level == 30 || level == 31 || level == 40 || level == 41 ||
-        level == 50 || level == 51 || level == 52 || level == 60 ||
-        level == 61 || level == 62 || level == 0 || level == 255)
-      EXPECT_EQ(VPX_CODEC_OK,
-                vpx_codec_control(&enc, VP9E_SET_TARGET_LEVEL, level));
-    else
-      EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
-                vpx_codec_control(&enc, VP9E_SET_TARGET_LEVEL, level));
-  }
-  EXPECT_EQ(VPX_CODEC_OK, vpx_codec_destroy(&enc));
-}
-
-VP9_INSTANTIATE_TEST_CASE(LevelTest,
-                          ::testing::Values(::libvpx_test::kTwoPassGood,
-                                            ::libvpx_test::kOnePassGood),
-                          ::testing::Range(0, 9));
-}  // namespace
--- a/test/lpf_8_test.cc
+++ b/test/lpf_8_test.cc
@@ -13,15 +13,15 @@
 #include <string>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vp10/common/entropy.h"
-#include "vp10/common/loopfilter.h"
+
+#include "./vpx_config.h"
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_entropy.h"
+#include "vp9/common/vp9_loopfilter.h"
 #include "vpx/vpx_integer.h"

 using libvpx_test::ACMRandom;
@@ -34,26 +34,109 @@ const int kNumCoeffs = 1024;

 const int number_of_iterations = 10000;

-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 typedef void (*loop_op_t)(uint16_t *s, int p, const uint8_t *blimit,
                          const uint8_t *limit, const uint8_t *thresh,
-                          int bd);
+                          int count, int bd);
 typedef void (*dual_loop_op_t)(uint16_t *s, int p, const uint8_t *blimit0,
                               const uint8_t *limit0, const uint8_t *thresh0,
                               const uint8_t *blimit1, const uint8_t *limit1,
                               const uint8_t *thresh1, int bd);
 #else
 typedef void (*loop_op_t)(uint8_t *s, int p, const uint8_t *blimit,
-                          const uint8_t *limit, const uint8_t *thresh);
+                          const uint8_t *limit, const uint8_t *thresh,
+                          int count);
 typedef void (*dual_loop_op_t)(uint8_t *s, int p, const uint8_t *blimit0,
                               const uint8_t *limit0, const uint8_t *thresh0,
                               const uint8_t *blimit1, const uint8_t *limit1,
                               const uint8_t *thresh1);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH

 typedef std::tr1::tuple<loop_op_t, loop_op_t, int> loop8_param_t;
 typedef std::tr1::tuple<dual_loop_op_t, dual_loop_op_t, int> dualloop8_param_t;

+#if HAVE_SSE2
+#if CONFIG_VP9_HIGHBITDEPTH
+void wrapper_vertical_16_sse2(uint16_t *s, int p, const uint8_t *blimit,
+                              const uint8_t *limit, const uint8_t *thresh,
+                              int count, int bd) {
+  vp9_highbd_lpf_vertical_16_sse2(s, p, blimit, limit, thresh, bd);
+}
+
+void wrapper_vertical_16_c(uint16_t *s, int p, const uint8_t *blimit,
+                           const uint8_t *limit, const uint8_t *thresh,
+                           int count, int bd) {
+  vp9_highbd_lpf_vertical_16_c(s, p, blimit, limit, thresh, bd);
+}
+
+void wrapper_vertical_16_dual_sse2(uint16_t *s, int p, const uint8_t *blimit,
+                                   const uint8_t *limit, const uint8_t *thresh,
+                                   int count, int bd) {
+  vp9_highbd_lpf_vertical_16_dual_sse2(s, p, blimit, limit, thresh, bd);
+}
+
+void wrapper_vertical_16_dual_c(uint16_t *s, int p, const uint8_t *blimit,
+                                const uint8_t *limit, const uint8_t *thresh,
+                                int count, int bd) {
+  vp9_highbd_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh, bd);
+}
+#else
+void wrapper_vertical_16_sse2(uint8_t *s, int p, const uint8_t *blimit,
+                              const uint8_t *limit, const uint8_t *thresh,
+                              int count) {
+  vp9_lpf_vertical_16_sse2(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_c(uint8_t *s, int p, const uint8_t *blimit,
+                           const uint8_t *limit, const uint8_t *thresh,
+                           int count) {
+  vp9_lpf_vertical_16_c(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_dual_sse2(uint8_t *s, int p, const uint8_t *blimit,
+                                   const uint8_t *limit, const uint8_t *thresh,
+                                   int count) {
+  vp9_lpf_vertical_16_dual_sse2(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_dual_c(uint8_t *s, int p, const uint8_t *blimit,
+                                const uint8_t *limit, const uint8_t *thresh,
+                                int count) {
+  vp9_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
+}
+#endif  // CONFIG_VP9_HIGHBITDEPTH
+#endif  // HAVE_SSE2
+
+#if HAVE_NEON_ASM
+#if CONFIG_VP9_HIGHBITDEPTH
+// No neon high bitdepth functions.
+#else
+void wrapper_vertical_16_neon(uint8_t *s, int p, const uint8_t *blimit,
+                              const uint8_t *limit, const uint8_t *thresh,
+                              int count) {
+  vp9_lpf_vertical_16_neon(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_c(uint8_t *s, int p, const uint8_t *blimit,
+                           const uint8_t *limit, const uint8_t *thresh,
+                           int count) {
+  vp9_lpf_vertical_16_c(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_dual_neon(uint8_t *s, int p, const uint8_t *blimit,
+                                   const uint8_t *limit, const uint8_t *thresh,
+                                   int count) {
+  vp9_lpf_vertical_16_dual_neon(s, p, blimit, limit, thresh);
+}
+
+void wrapper_vertical_16_dual_c(uint8_t *s, int p, const uint8_t *blimit,
+                                const uint8_t *limit, const uint8_t *thresh,
+                                int count) {
+  vp9_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
+}
+#endif  // CONFIG_VP9_HIGHBITDEPTH
+#endif  // HAVE_NEON_ASM
+
 class Loop8Test6Param : public ::testing::TestWithParam<loop8_param_t> {
 public:
  virtual ~Loop8Test6Param() {}
@@ -95,14 +178,14 @@ class Loop8Test9Param : public ::testing::TestWithParam<dualloop8_param_t> {
 TEST_P(Loop8Test6Param, OperationCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = number_of_iterations;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  int32_t bd = bit_depth_;
-  DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_s, kNumCoeffs);
 #else
-  DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(8, uint8_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(8, uint8_t, ref_s, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < count_test_block; ++i) {
@@ -123,6 +206,7 @@ TEST_P(Loop8Test6Param, OperationCheck) {
        tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
    };
    int32_t p = kNumCoeffs/32;
+    int count = 1;

    uint16_t tmp_s[kNumCoeffs];
    int j = 0;
@@ -153,15 +237,15 @@ TEST_P(Loop8Test6Param, OperationCheck) {
      }
      ref_s[j] = s[j];
    }
-#if CONFIG_VPX_HIGHBITDEPTH
-    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bd);
+#if CONFIG_VP9_HIGHBITDEPTH
+    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, count, bd);
    ASM_REGISTER_STATE_CHECK(
-        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bd));
+        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, count, bd));
 #else
-    ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh);
+    ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh, count);
    ASM_REGISTER_STATE_CHECK(
-        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, count));
+#endif  // CONFIG_VP9_HIGHBITDEPTH

    for (int j = 0; j < kNumCoeffs; ++j) {
      err_count += ref_s[j] != s[j];
@@ -180,14 +264,14 @@ TEST_P(Loop8Test6Param, OperationCheck) {
 TEST_P(Loop8Test6Param, ValueCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = number_of_iterations;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  const int32_t bd = bit_depth_;
-  DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_s, kNumCoeffs);
 #else
-  DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(8, uint8_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(8, uint8_t, ref_s, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  int err_count_total = 0;
  int first_failure = -1;

@@ -195,8 +279,8 @@ TEST_P(Loop8Test6Param, ValueCheck) {
  // function of sharpness_lvl and the loopfilter lvl as:
  // block_inside_limit = lvl >> ((sharpness_lvl > 0) + (sharpness_lvl > 4));
  // ...
-  // memset(lfi->lfthr[lvl].mblim, (2 * (lvl + 2) + block_inside_limit),
-  //        SIMD_WIDTH);
+  // vpx_memset(lfi->lfthr[lvl].mblim, (2 * (lvl + 2) + block_inside_limit),
+  //            SIMD_WIDTH);
  // This means that the largest value for mblim will occur when sharpness_lvl
  // is equal to 0, and lvl is equal to its greatest value (MAX_LOOP_FILTER).
  // In this case block_inside_limit will be equal to MAX_LOOP_FILTER and
@@ -221,19 +305,20 @@ TEST_P(Loop8Test6Param, ValueCheck) {
        tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
    };
    int32_t p = kNumCoeffs / 32;
+    int count = 1;
    for (int j = 0; j < kNumCoeffs; ++j) {
      s[j] = rnd.Rand16() & mask_;
      ref_s[j] = s[j];
    }
-#if CONFIG_VPX_HIGHBITDEPTH
-    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bd);
+#if CONFIG_VP9_HIGHBITDEPTH
+    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, count, bd);
    ASM_REGISTER_STATE_CHECK(
-        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bd));
+        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, count, bd));
 #else
-    ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh);
+    ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh, count);
    ASM_REGISTER_STATE_CHECK(
-        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, count));
+#endif  // CONFIG_VP9_HIGHBITDEPTH
    for (int j = 0; j < kNumCoeffs; ++j) {
      err_count += ref_s[j] != s[j];
    }
@@ -251,14 +336,14 @@ TEST_P(Loop8Test6Param, ValueCheck) {
 TEST_P(Loop8Test9Param, OperationCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = number_of_iterations;
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
  const int32_t bd = bit_depth_;
-  DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_s, kNumCoeffs);
 #else
-  DECLARE_ALIGNED(8,  uint8_t,  s[kNumCoeffs]);
-  DECLARE_ALIGNED(8,  uint8_t,  ref_s[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(8,  uint8_t,  s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(8,  uint8_t,  ref_s, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < count_test_block; ++i) {
@@ -324,7 +409,7 @@ TEST_P(Loop8Test9Param, OperationCheck) {
      }
      ref_s[j] = s[j];
    }
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
                       blimit1, limit1, thresh1, bd);
    ASM_REGISTER_STATE_CHECK(
@@ -336,7 +421,7 @@ TEST_P(Loop8Test9Param, OperationCheck) {
    ASM_REGISTER_STATE_CHECK(
        loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0, thresh0,
                       blimit1, limit1, thresh1));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
    for (int j = 0; j < kNumCoeffs; ++j) {
      err_count += ref_s[j] != s[j];
    }
@@ -354,13 +439,13 @@ TEST_P(Loop8Test9Param, OperationCheck) {
 TEST_P(Loop8Test9Param, ValueCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = number_of_iterations;
-#if CONFIG_VPX_HIGHBITDEPTH
-  DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
+#if CONFIG_VP9_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_s, kNumCoeffs);
 #else
-  DECLARE_ALIGNED(8,  uint8_t, s[kNumCoeffs]);
-  DECLARE_ALIGNED(8,  uint8_t, ref_s[kNumCoeffs]);
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+  DECLARE_ALIGNED_ARRAY(8,  uint8_t, s, kNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(8,  uint8_t, ref_s, kNumCoeffs);
+#endif  // CONFIG_VP9_HIGHBITDEPTH
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < count_test_block; ++i) {
@@ -400,7 +485,7 @@ TEST_P(Loop8Test9Param, ValueCheck) {
      s[j] = rnd.Rand16() & mask_;
      ref_s[j] = s[j];
    }
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
    const int32_t bd = bit_depth_;
    ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
                       blimit1, limit1, thresh1, bd);
@@ -413,7 +498,7 @@ TEST_P(Loop8Test9Param, ValueCheck) {
    ASM_REGISTER_STATE_CHECK(
        loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0, thresh0,
                       blimit1, limit1, thresh1));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+#endif  // CONFIG_VP9_HIGHBITDEPTH
    for (int j = 0; j < kNumCoeffs; ++j) {
      err_count += ref_s[j] != s[j];
    }
@@ -431,137 +516,116 @@ TEST_P(Loop8Test9Param, ValueCheck) {
 using std::tr1::make_tuple;

 #if HAVE_SSE2
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_8_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_16_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
-                   &vpx_highbd_lpf_vertical_16_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_8_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_16_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
-                   &vpx_highbd_lpf_vertical_16_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 12),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 12),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_8_c, 12),
-        make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
-                   &vpx_highbd_lpf_horizontal_edge_16_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
-                   &vpx_highbd_lpf_vertical_16_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
-                   &vpx_highbd_lpf_vertical_16_dual_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
-                   &vpx_highbd_lpf_vertical_16_dual_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
-                   &vpx_highbd_lpf_vertical_16_dual_c, 12)));
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 8),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 8),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 8),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 8),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 8),
+        make_tuple(&wrapper_vertical_16_sse2,
+                   &wrapper_vertical_16_c, 8),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 10),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 10),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 10),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 10),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 10),
+        make_tuple(&wrapper_vertical_16_sse2,
+                   &wrapper_vertical_16_c, 10),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 12),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 12),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 12),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 12),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 12),
+        make_tuple(&wrapper_vertical_16_sse2,
+                   &wrapper_vertical_16_c, 12),
+        make_tuple(&wrapper_vertical_16_dual_sse2,
+                   &wrapper_vertical_16_dual_c, 8),
+        make_tuple(&wrapper_vertical_16_dual_sse2,
+                   &wrapper_vertical_16_dual_c, 10),
+        make_tuple(&wrapper_vertical_16_dual_sse2,
+                   &wrapper_vertical_16_dual_c, 12)));
 #else
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_sse2,
-                   &vpx_lpf_horizontal_4_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_sse2,
-                   &vpx_lpf_horizontal_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_8_sse2,
-                   &vpx_lpf_horizontal_edge_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_16_sse2,
-                   &vpx_lpf_horizontal_edge_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_sse2,
-                   &vpx_lpf_vertical_4_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_sse2,
-                   &vpx_lpf_vertical_8_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_sse2,
-                   &vpx_lpf_vertical_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_dual_sse2,
-                   &vpx_lpf_vertical_16_dual_c, 8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_lpf_horizontal_8_sse2, &vp9_lpf_horizontal_8_c, 8),
+        make_tuple(&vp9_lpf_horizontal_16_sse2, &vp9_lpf_horizontal_16_c, 8),
+        make_tuple(&vp9_lpf_vertical_8_sse2, &vp9_lpf_vertical_8_c, 8),
+        make_tuple(&wrapper_vertical_16_sse2, &wrapper_vertical_16_c, 8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif

-#if HAVE_AVX2 && (!CONFIG_VPX_HIGHBITDEPTH)
+#if HAVE_AVX2 && (!CONFIG_VP9_HIGHBITDEPTH)
 INSTANTIATE_TEST_CASE_P(
    AVX2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_edge_8_avx2,
-                   &vpx_lpf_horizontal_edge_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_16_avx2,
-                   &vpx_lpf_horizontal_edge_16_c, 8)));
+        make_tuple(&vp9_lpf_horizontal_16_avx2, &vp9_lpf_horizontal_16_c, 8)));
 #endif

 #if HAVE_SSE2
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test9Param,
    ::testing::Values(
-        make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_4_dual_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_8_dual_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
-                   &vpx_highbd_lpf_vertical_4_dual_c, 8),
-        make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
-                   &vpx_highbd_lpf_vertical_8_dual_c, 8),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_4_dual_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_8_dual_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
-                   &vpx_highbd_lpf_vertical_4_dual_c, 10),
-        make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
-                   &vpx_highbd_lpf_vertical_8_dual_c, 10),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_4_dual_c, 12),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
-                   &vpx_highbd_lpf_horizontal_8_dual_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
-                   &vpx_highbd_lpf_vertical_4_dual_c, 12),
-        make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
-                   &vpx_highbd_lpf_vertical_8_dual_c, 12)));
+        make_tuple(&vp9_highbd_lpf_horizontal_4_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_4_dual_c, 8),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_8_dual_c, 8),
+        make_tuple(&vp9_highbd_lpf_vertical_4_dual_sse2,
+                   &vp9_highbd_lpf_vertical_4_dual_c, 8),
+        make_tuple(&vp9_highbd_lpf_vertical_8_dual_sse2,
+                   &vp9_highbd_lpf_vertical_8_dual_c, 8),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_4_dual_c, 10),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_8_dual_c, 10),
+        make_tuple(&vp9_highbd_lpf_vertical_4_dual_sse2,
+                   &vp9_highbd_lpf_vertical_4_dual_c, 10),
+        make_tuple(&vp9_highbd_lpf_vertical_8_dual_sse2,
+                   &vp9_highbd_lpf_vertical_8_dual_c, 10),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_4_dual_c, 12),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_dual_sse2,
+                   &vp9_highbd_lpf_horizontal_8_dual_c, 12),
+        make_tuple(&vp9_highbd_lpf_vertical_4_dual_sse2,
+                   &vp9_highbd_lpf_vertical_4_dual_c, 12),
+        make_tuple(&vp9_highbd_lpf_vertical_8_dual_sse2,
+                   &vp9_highbd_lpf_vertical_8_dual_c, 12)));
 #else
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test9Param,
    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_dual_sse2,
-                   &vpx_lpf_horizontal_4_dual_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_dual_sse2,
-                   &vpx_lpf_horizontal_8_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_dual_sse2,
-                   &vpx_lpf_vertical_4_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_dual_sse2,
-                   &vpx_lpf_vertical_8_dual_c, 8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_lpf_horizontal_4_dual_sse2,
+                   &vp9_lpf_horizontal_4_dual_c, 8),
+        make_tuple(&vp9_lpf_horizontal_8_dual_sse2,
+                   &vp9_lpf_horizontal_8_dual_c, 8),
+        make_tuple(&vp9_lpf_vertical_4_dual_sse2,
+                   &vp9_lpf_vertical_4_dual_c, 8),
+        make_tuple(&vp9_lpf_vertical_8_dual_sse2,
+                   &vp9_lpf_vertical_8_dual_c, 8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif

 #if HAVE_NEON
-#if CONFIG_VPX_HIGHBITDEPTH
+#if CONFIG_VP9_HIGHBITDEPTH
 // No neon high bitdepth functions.
 #else
 INSTANTIATE_TEST_CASE_P(
@@ -570,103 +634,35 @@ INSTANTIATE_TEST_CASE_P(
 #if HAVE_NEON_ASM
 // Using #if inside the macro is unsupported on MSVS but the tests are not
 // currently built for MSVS with ARM and NEON.
-        make_tuple(&vpx_lpf_horizontal_edge_8_neon,
-                   &vpx_lpf_horizontal_edge_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_16_neon,
-                   &vpx_lpf_horizontal_edge_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_neon,
-                   &vpx_lpf_vertical_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_dual_neon,
-                   &vpx_lpf_vertical_16_dual_c, 8),
+        make_tuple(&vp9_lpf_horizontal_16_neon,
+                   &vp9_lpf_horizontal_16_c, 8),
+        make_tuple(&wrapper_vertical_16_neon,
+                   &wrapper_vertical_16_c, 8),
+        make_tuple(&wrapper_vertical_16_dual_neon,
+                   &wrapper_vertical_16_dual_c, 8),
+        make_tuple(&vp9_lpf_horizontal_8_neon,
+                   &vp9_lpf_horizontal_8_c, 8),
+        make_tuple(&vp9_lpf_vertical_8_neon,
+                   &vp9_lpf_vertical_8_c, 8),
 #endif  // HAVE_NEON_ASM
-        make_tuple(&vpx_lpf_horizontal_8_neon,
-                   &vpx_lpf_horizontal_8_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_neon,
-                   &vpx_lpf_vertical_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_4_neon,
-                   &vpx_lpf_horizontal_4_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_neon,
-                   &vpx_lpf_vertical_4_c, 8)));
+        make_tuple(&vp9_lpf_horizontal_4_neon,
+                   &vp9_lpf_horizontal_4_c, 8),
+        make_tuple(&vp9_lpf_vertical_4_neon,
+                   &vp9_lpf_vertical_4_c, 8)));
 INSTANTIATE_TEST_CASE_P(
    NEON, Loop8Test9Param,
    ::testing::Values(
 #if HAVE_NEON_ASM
-        make_tuple(&vpx_lpf_horizontal_8_dual_neon,
-                   &vpx_lpf_horizontal_8_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_dual_neon,
-                   &vpx_lpf_vertical_8_dual_c, 8),
+        make_tuple(&vp9_lpf_horizontal_8_dual_neon,
+                   &vp9_lpf_horizontal_8_dual_c, 8),
+        make_tuple(&vp9_lpf_vertical_8_dual_neon,
+                   &vp9_lpf_vertical_8_dual_c, 8),
 #endif  // HAVE_NEON_ASM
-        make_tuple(&vpx_lpf_horizontal_4_dual_neon,
-                   &vpx_lpf_horizontal_4_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_dual_neon,
-                   &vpx_lpf_vertical_4_dual_c, 8)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
+        make_tuple(&vp9_lpf_horizontal_4_dual_neon,
+                   &vp9_lpf_horizontal_4_dual_c, 8),
+        make_tuple(&vp9_lpf_vertical_4_dual_neon,
+                   &vp9_lpf_vertical_4_dual_c, 8)));
+#endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif  // HAVE_NEON

-#if HAVE_DSPR2 && !CONFIG_VPX_HIGHBITDEPTH
-INSTANTIATE_TEST_CASE_P(
-    DSPR2, Loop8Test6Param,
-    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_dspr2,
-                   &vpx_lpf_horizontal_4_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_dspr2,
-                   &vpx_lpf_horizontal_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_8,
-                   &vpx_lpf_horizontal_edge_8, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_16,
-                   &vpx_lpf_horizontal_edge_16, 8),
-        make_tuple(&vpx_lpf_vertical_4_dspr2,
-                   &vpx_lpf_vertical_4_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_dspr2,
-                   &vpx_lpf_vertical_8_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_dspr2,
-                   &vpx_lpf_vertical_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_dual_dspr2,
-                   &vpx_lpf_vertical_16_dual_c, 8)));
-
-INSTANTIATE_TEST_CASE_P(
-    DSPR2, Loop8Test9Param,
-    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_dual_dspr2,
-                   &vpx_lpf_horizontal_4_dual_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_dual_dspr2,
-                   &vpx_lpf_horizontal_8_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_dual_dspr2,
-                   &vpx_lpf_vertical_4_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_dual_dspr2,
-                   &vpx_lpf_vertical_8_dual_c, 8)));
-#endif  // HAVE_DSPR2 && !CONFIG_VPX_HIGHBITDEPTH
-
-#if HAVE_MSA && (!CONFIG_VPX_HIGHBITDEPTH)
-INSTANTIATE_TEST_CASE_P(
-    MSA, Loop8Test6Param,
-    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_msa,
-                   &vpx_lpf_horizontal_4_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_msa,
-                   &vpx_lpf_horizontal_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_8_msa,
-                   &vpx_lpf_horizontal_edge_8_c, 8),
-        make_tuple(&vpx_lpf_horizontal_edge_16_msa,
-                   &vpx_lpf_horizontal_edge_16_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_msa,
-                   &vpx_lpf_vertical_4_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_msa,
-                   &vpx_lpf_vertical_8_c, 8),
-        make_tuple(&vpx_lpf_vertical_16_msa,
-                   &vpx_lpf_vertical_16_c, 8)));
-
-INSTANTIATE_TEST_CASE_P(
-    MSA, Loop8Test9Param,
-    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_4_dual_msa,
-                   &vpx_lpf_horizontal_4_dual_c, 8),
-        make_tuple(&vpx_lpf_horizontal_8_dual_msa,
-                   &vpx_lpf_horizontal_8_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_4_dual_msa,
-                   &vpx_lpf_vertical_4_dual_c, 8),
-        make_tuple(&vpx_lpf_vertical_8_dual_msa,
-                   &vpx_lpf_vertical_8_dual_c, 8)));
-#endif  // HAVE_MSA && (!CONFIG_VPX_HIGHBITDEPTH)
-
 }  // namespace
--- a/test/masked_sad_test.cc
+++ b/test/masked_sad_test.cc
@@ -1,225 +0,0 @@
-/*
- *  Copyright (c) 2015 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <math.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/acm_random.h"
-#include "test/clear_system_state.h"
-#include "test/register_state_check.h"
-#include "test/util.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-using libvpx_test::ACMRandom;
-
-namespace {
-const int number_of_iterations = 500;
-
-typedef unsigned int (*MaskedSADFunc)(const uint8_t *a, int a_stride,
-                                      const uint8_t *b, int b_stride,
-                                      const uint8_t *m, int m_stride);
-typedef std::tr1::tuple<MaskedSADFunc, MaskedSADFunc> MaskedSADParam;
-
-class MaskedSADTest : public ::testing::TestWithParam<MaskedSADParam> {
- public:
-  virtual ~MaskedSADTest() {}
-  virtual void SetUp() {
-    maskedSAD_op_   = GET_PARAM(0);
-    ref_maskedSAD_op_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  MaskedSADFunc maskedSAD_op_;
-  MaskedSADFunc ref_maskedSAD_op_;
-};
-
-TEST_P(MaskedSADTest, OperationCheck) {
-  unsigned int ref_ret, ret;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint8_t, src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t, ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t, msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-  for (int i = 0; i < number_of_iterations; ++i) {
-    for (int j = 0; j < MAX_SB_SIZE*MAX_SB_SIZE; j++) {
-      src_ptr[j] = rnd.Rand8();
-      ref_ptr[j] = rnd.Rand8();
-      msk_ptr[j] = ((rnd.Rand8()&0x7f) > 64) ? rnd.Rand8()&0x3f : 64;
-      assert(msk_ptr[j] <= 64);
-    }
-
-    ref_ret = ref_maskedSAD_op_(src_ptr, src_stride, ref_ptr, ref_stride,
-                                msk_ptr, msk_stride);
-    ASM_REGISTER_STATE_CHECK(ret = maskedSAD_op_(src_ptr, src_stride,
-                                                 ref_ptr, ref_stride,
-                                                 msk_ptr, msk_stride));
-    if (ret != ref_ret) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-  EXPECT_EQ(0, err_count)
-    << "Error: Masked SAD Test, C output doesn't match SSSE3 output. "
-    << "First failed at test case " << first_failure;
-}
-
-#if CONFIG_VPX_HIGHBITDEPTH
-typedef unsigned int (*HighbdMaskedSADFunc)(const uint8_t *a, int a_stride,
-                                            const uint8_t *b, int b_stride,
-                                            const uint8_t *m, int m_stride);
-typedef std::tr1::tuple<HighbdMaskedSADFunc, HighbdMaskedSADFunc>
-    HighbdMaskedSADParam;
-
-class HighbdMaskedSADTest : public ::testing::
-        TestWithParam<HighbdMaskedSADParam> {
- public:
-  virtual ~HighbdMaskedSADTest() {}
-  virtual void SetUp() {
-    maskedSAD_op_   = GET_PARAM(0);
-    ref_maskedSAD_op_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  HighbdMaskedSADFunc maskedSAD_op_;
-  HighbdMaskedSADFunc ref_maskedSAD_op_;
-};
-
-TEST_P(HighbdMaskedSADTest, OperationCheck) {
-  unsigned int ref_ret, ret;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint16_t, src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint16_t, ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t, msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  uint8_t* src8_ptr = CONVERT_TO_BYTEPTR(src_ptr);
-  uint8_t* ref8_ptr = CONVERT_TO_BYTEPTR(ref_ptr);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-  for (int i = 0; i < number_of_iterations; ++i) {
-    for (int j = 0; j < MAX_SB_SIZE*MAX_SB_SIZE; j++) {
-      src_ptr[j] = rnd.Rand16()&0xfff;
-      ref_ptr[j] = rnd.Rand16()&0xfff;
-      msk_ptr[j] = ((rnd.Rand8()&0x7f) > 64) ? rnd.Rand8()&0x3f : 64;
-    }
-
-    ref_ret = ref_maskedSAD_op_(src8_ptr, src_stride, ref8_ptr, ref_stride,
-                                msk_ptr, msk_stride);
-    ASM_REGISTER_STATE_CHECK(ret = maskedSAD_op_(src8_ptr, src_stride,
-                                                 ref8_ptr, ref_stride,
-                                                 msk_ptr, msk_stride));
-    if (ret != ref_ret) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-  EXPECT_EQ(0, err_count)
-    << "Error: High BD Masked SAD Test, C output doesn't match SSSE3 output. "
-    << "First failed at test case " << first_failure;
-}
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-
-using std::tr1::make_tuple;
-
-#if HAVE_SSSE3
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, MaskedSADTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_sad128x128_ssse3,
-               &vpx_masked_sad128x128_c),
-    make_tuple(&vpx_masked_sad128x64_ssse3,
-               &vpx_masked_sad128x64_c),
-    make_tuple(&vpx_masked_sad64x128_ssse3,
-               &vpx_masked_sad64x128_c),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_sad64x64_ssse3,
-               &vpx_masked_sad64x64_c),
-    make_tuple(&vpx_masked_sad64x32_ssse3,
-               &vpx_masked_sad64x32_c),
-    make_tuple(&vpx_masked_sad32x64_ssse3,
-               &vpx_masked_sad32x64_c),
-    make_tuple(&vpx_masked_sad32x32_ssse3,
-               &vpx_masked_sad32x32_c),
-    make_tuple(&vpx_masked_sad32x16_ssse3,
-               &vpx_masked_sad32x16_c),
-    make_tuple(&vpx_masked_sad16x32_ssse3,
-               &vpx_masked_sad16x32_c),
-    make_tuple(&vpx_masked_sad16x16_ssse3,
-               &vpx_masked_sad16x16_c),
-    make_tuple(&vpx_masked_sad16x8_ssse3,
-               &vpx_masked_sad16x8_c),
-    make_tuple(&vpx_masked_sad8x16_ssse3,
-               &vpx_masked_sad8x16_c),
-    make_tuple(&vpx_masked_sad8x8_ssse3,
-               &vpx_masked_sad8x8_c),
-    make_tuple(&vpx_masked_sad8x4_ssse3,
-               &vpx_masked_sad8x4_c),
-    make_tuple(&vpx_masked_sad4x8_ssse3,
-               &vpx_masked_sad4x8_c),
-    make_tuple(&vpx_masked_sad4x4_ssse3,
-               &vpx_masked_sad4x4_c)));
-#if CONFIG_VPX_HIGHBITDEPTH
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, HighbdMaskedSADTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_sad128x128_ssse3,
-               &vpx_highbd_masked_sad128x128_c),
-    make_tuple(&vpx_highbd_masked_sad128x64_ssse3,
-               &vpx_highbd_masked_sad128x64_c),
-    make_tuple(&vpx_highbd_masked_sad64x128_ssse3,
-               &vpx_highbd_masked_sad64x128_c),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_sad64x64_ssse3,
-               &vpx_highbd_masked_sad64x64_c),
-    make_tuple(&vpx_highbd_masked_sad64x32_ssse3,
-               &vpx_highbd_masked_sad64x32_c),
-    make_tuple(&vpx_highbd_masked_sad32x64_ssse3,
-               &vpx_highbd_masked_sad32x64_c),
-    make_tuple(&vpx_highbd_masked_sad32x32_ssse3,
-               &vpx_highbd_masked_sad32x32_c),
-    make_tuple(&vpx_highbd_masked_sad32x16_ssse3,
-               &vpx_highbd_masked_sad32x16_c),
-    make_tuple(&vpx_highbd_masked_sad16x32_ssse3,
-               &vpx_highbd_masked_sad16x32_c),
-    make_tuple(&vpx_highbd_masked_sad16x16_ssse3,
-               &vpx_highbd_masked_sad16x16_c),
-    make_tuple(&vpx_highbd_masked_sad16x8_ssse3,
-               &vpx_highbd_masked_sad16x8_c),
-    make_tuple(&vpx_highbd_masked_sad8x16_ssse3,
-               &vpx_highbd_masked_sad8x16_c),
-    make_tuple(&vpx_highbd_masked_sad8x8_ssse3,
-               &vpx_highbd_masked_sad8x8_c),
-    make_tuple(&vpx_highbd_masked_sad8x4_ssse3,
-               &vpx_highbd_masked_sad8x4_c),
-    make_tuple(&vpx_highbd_masked_sad4x8_ssse3,
-               &vpx_highbd_masked_sad4x8_c),
-    make_tuple(&vpx_highbd_masked_sad4x4_ssse3,
-               &vpx_highbd_masked_sad4x4_c)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-#endif  // HAVE_SSSE3
-}  // namespace
--- a/test/masked_variance_test.cc
+++ b/test/masked_variance_test.cc
@@ -1,816 +0,0 @@
-/*
- *  Copyright (c) 2015 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <math.h>
-#include <stdlib.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/acm_random.h"
-#include "test/clear_system_state.h"
-#include "test/register_state_check.h"
-#include "test/util.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_codec.h"
-#include "vpx/vpx_integer.h"
-#include "vpx_dsp/vpx_filter.h"
-#include "vpx_mem/vpx_mem.h"
-
-using libvpx_test::ACMRandom;
-
-namespace {
-const int number_of_iterations = 500;
-
-typedef unsigned int (*MaskedVarianceFunc)(const uint8_t *a, int a_stride,
-                                           const uint8_t *b, int b_stride,
-                                           const uint8_t *m, int m_stride,
-                                           unsigned int *sse);
-
-typedef std::tr1::tuple<MaskedVarianceFunc,
-                        MaskedVarianceFunc> MaskedVarianceParam;
-
-class MaskedVarianceTest :
-  public ::testing::TestWithParam<MaskedVarianceParam> {
- public:
-  virtual ~MaskedVarianceTest() {}
-  virtual void SetUp() {
-    opt_func_ = GET_PARAM(0);
-    ref_func_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  MaskedVarianceFunc opt_func_;
-  MaskedVarianceFunc ref_func_;
-};
-
-TEST_P(MaskedVarianceTest, OperationCheck) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint8_t,  src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-
-  for (int i = 0; i < number_of_iterations; ++i) {
-    for (int j = 0; j < MAX_SB_SIZE*MAX_SB_SIZE; j++) {
-      src_ptr[j] = rnd.Rand8();
-      ref_ptr[j] = rnd.Rand8();
-      msk_ptr[j] = rnd(65);
-    }
-
-    ref_ret = ref_func_(src_ptr, src_stride,
-                        ref_ptr, ref_stride,
-                        msk_ptr, msk_stride,
-                        &ref_sse);
-    ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src_ptr, src_stride,
-                                                 ref_ptr, ref_stride,
-                                                 msk_ptr, msk_stride,
-                                                 &opt_sse));
-
-    if (opt_ret != ref_ret || opt_sse != ref_sse) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test OperationCheck,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure;
-}
-
-TEST_P(MaskedVarianceTest, ExtremeValues) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint8_t,  src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-
-  for (int i = 0; i < 8; ++i) {
-    memset(src_ptr, (i & 0x1) ? 255 : 0, MAX_SB_SIZE*MAX_SB_SIZE);
-    memset(ref_ptr, (i & 0x2) ? 255 : 0, MAX_SB_SIZE*MAX_SB_SIZE);
-    memset(msk_ptr, (i & 0x4) ?  64 : 0, MAX_SB_SIZE*MAX_SB_SIZE);
-
-    ref_ret = ref_func_(src_ptr, src_stride,
-                        ref_ptr, ref_stride,
-                        msk_ptr, msk_stride,
-                        &ref_sse);
-    ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src_ptr, src_stride,
-                                                 ref_ptr, ref_stride,
-                                                 msk_ptr, msk_stride,
-                                                 &opt_sse));
-
-    if (opt_ret != ref_ret || opt_sse != ref_sse) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test ExtremeValues,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure;
-}
-
-typedef unsigned int (*MaskedSubPixelVarianceFunc)(
-    const uint8_t *a, int a_stride,
-    int xoffset, int  yoffset,
-    const uint8_t *b, int b_stride,
-    const uint8_t *m, int m_stride,
-    unsigned int *sse);
-
-typedef std::tr1::tuple<MaskedSubPixelVarianceFunc,
-                        MaskedSubPixelVarianceFunc> MaskedSubPixelVarianceParam;
-
-class MaskedSubPixelVarianceTest :
-  public ::testing::TestWithParam<MaskedSubPixelVarianceParam> {
- public:
-  virtual ~MaskedSubPixelVarianceTest() {}
-  virtual void SetUp() {
-    opt_func_ = GET_PARAM(0);
-    ref_func_ = GET_PARAM(1);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  MaskedSubPixelVarianceFunc opt_func_;
-  MaskedSubPixelVarianceFunc ref_func_;
-};
-
-TEST_P(MaskedSubPixelVarianceTest, OperationCheck) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint8_t,  src_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  ref_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = (MAX_SB_SIZE+1);
-  int ref_stride = (MAX_SB_SIZE+1);
-  int msk_stride = (MAX_SB_SIZE+1);
-  int xoffset;
-  int yoffset;
-
-  for (int i = 0; i < number_of_iterations; ++i) {
-    int xoffsets[] = {0, 4, rnd(BIL_SUBPEL_SHIFTS)};
-    int yoffsets[] = {0, 4, rnd(BIL_SUBPEL_SHIFTS)};
-    for (int j = 0; j < (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1); j++) {
-      src_ptr[j] = rnd.Rand8();
-      ref_ptr[j] = rnd.Rand8();
-      msk_ptr[j] = rnd(65);
-    }
-    for (int k = 0; k < 3; k++) {
-      xoffset = xoffsets[k];
-      for (int l = 0; l < 3; l++) {
-        xoffset = xoffsets[k];
-        yoffset = yoffsets[l];
-
-        ref_ret = ref_func_(src_ptr, src_stride,
-                            xoffset, yoffset,
-                            ref_ptr, ref_stride,
-                            msk_ptr, msk_stride,
-                            &ref_sse);
-        ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src_ptr, src_stride,
-                                                     xoffset, yoffset,
-                                                     ref_ptr, ref_stride,
-                                                     msk_ptr, msk_stride,
-                                                     &opt_sse));
-
-        if (opt_ret != ref_ret || opt_sse != ref_sse) {
-        err_count++;
-        if (first_failure == -1)
-            first_failure = i;
-        }
-      }
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-    << "Error: Masked Sub Pixel Variance Test OperationCheck,"
-    << "C output doesn't match SSSE3 output. "
-    << "First failed at test case " << first_failure;
-}
-
-TEST_P(MaskedSubPixelVarianceTest, ExtremeValues) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint8_t,  src_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  ref_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  int first_failure_x = -1;
-  int first_failure_y = -1;
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = (MAX_SB_SIZE+1);
-  int ref_stride = (MAX_SB_SIZE+1);
-  int msk_stride = (MAX_SB_SIZE+1);
-
-  for (int xoffset = 0 ; xoffset < BIL_SUBPEL_SHIFTS ; xoffset++) {
-    for (int yoffset = 0 ; yoffset < BIL_SUBPEL_SHIFTS ; yoffset++) {
-      for (int i = 0; i < 8; ++i) {
-        memset(src_ptr, (i & 0x1) ? 255 : 0, (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-        memset(ref_ptr, (i & 0x2) ? 255 : 0, (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-        memset(msk_ptr, (i & 0x4) ?  64 : 0, (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-
-        ref_ret = ref_func_(src_ptr, src_stride,
-                            xoffset, yoffset,
-                            ref_ptr, ref_stride,
-                            msk_ptr, msk_stride,
-                            &ref_sse);
-        ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src_ptr, src_stride,
-                                                     xoffset, yoffset,
-                                                     ref_ptr, ref_stride,
-                                                     msk_ptr, msk_stride,
-                                                     &opt_sse));
-
-        if (opt_ret != ref_ret || opt_sse != ref_sse) {
-          err_count++;
-          if (first_failure == -1) {
-            first_failure = i;
-            first_failure_x = xoffset;
-            first_failure_y = yoffset;
-          }
-        }
-      }
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test ExtremeValues,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure
-  << " x_offset = " << first_failure_x
-  << " y_offset = " << first_failure_y;
-}
-
-#if CONFIG_VPX_HIGHBITDEPTH
-typedef std::tr1::tuple<MaskedVarianceFunc,
-                        MaskedVarianceFunc,
-                        vpx_bit_depth_t> HighbdMaskedVarianceParam;
-
-class HighbdMaskedVarianceTest :
-  public ::testing::TestWithParam<HighbdMaskedVarianceParam> {
- public:
-  virtual ~HighbdMaskedVarianceTest() {}
-  virtual void SetUp() {
-    opt_func_ = GET_PARAM(0);
-    ref_func_ = GET_PARAM(1);
-    bit_depth_ = GET_PARAM(2);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  MaskedVarianceFunc opt_func_;
-  MaskedVarianceFunc ref_func_;
-  vpx_bit_depth_t bit_depth_;
-};
-
-TEST_P(HighbdMaskedVarianceTest, OperationCheck) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint16_t, src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint16_t, ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  uint8_t* src8_ptr = CONVERT_TO_BYTEPTR(src_ptr);
-  uint8_t* ref8_ptr = CONVERT_TO_BYTEPTR(ref_ptr);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-
-  for (int i = 0; i < number_of_iterations; ++i) {
-    for (int j = 0; j < MAX_SB_SIZE*MAX_SB_SIZE; j++) {
-      src_ptr[j] = rnd.Rand16() & ((1 << bit_depth_) - 1);
-      ref_ptr[j] = rnd.Rand16() & ((1 << bit_depth_) - 1);
-      msk_ptr[j] = rnd(65);
-    }
-
-    ref_ret = ref_func_(src8_ptr, src_stride,
-                        ref8_ptr, ref_stride,
-                        msk_ptr, msk_stride,
-                        &ref_sse);
-    ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src8_ptr, src_stride,
-                                                 ref8_ptr, ref_stride,
-                                                 msk_ptr, msk_stride,
-                                                 &opt_sse));
-
-    if (opt_ret != ref_ret || opt_sse != ref_sse) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test OperationCheck,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure;
-}
-
-TEST_P(HighbdMaskedVarianceTest, ExtremeValues) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint16_t, src_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint16_t, ref_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[MAX_SB_SIZE*MAX_SB_SIZE]);
-  uint8_t* src8_ptr = CONVERT_TO_BYTEPTR(src_ptr);
-  uint8_t* ref8_ptr = CONVERT_TO_BYTEPTR(ref_ptr);
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = MAX_SB_SIZE;
-  int ref_stride = MAX_SB_SIZE;
-  int msk_stride = MAX_SB_SIZE;
-
-  for (int i = 0; i < 8; ++i) {
-    vpx_memset16(src_ptr, (i & 0x1) ? ((1 << bit_depth_) - 1) : 0,
-                 MAX_SB_SIZE*MAX_SB_SIZE);
-    vpx_memset16(ref_ptr, (i & 0x2) ? ((1 << bit_depth_) - 1) : 0,
-                 MAX_SB_SIZE*MAX_SB_SIZE);
-    memset(msk_ptr, (i & 0x4) ?  64 : 0, MAX_SB_SIZE*MAX_SB_SIZE);
-
-    ref_ret = ref_func_(src8_ptr, src_stride,
-                        ref8_ptr, ref_stride,
-                        msk_ptr, msk_stride,
-                        &ref_sse);
-    ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src8_ptr, src_stride,
-                                                 ref8_ptr, ref_stride,
-                                                 msk_ptr, msk_stride,
-                                                 &opt_sse));
-
-    if (opt_ret != ref_ret || opt_sse != ref_sse) {
-      err_count++;
-      if (first_failure == -1)
-        first_failure = i;
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test ExtremeValues,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure;
-}
-
-typedef std::tr1::tuple<MaskedSubPixelVarianceFunc,
-                        MaskedSubPixelVarianceFunc,
-                        vpx_bit_depth_t> HighbdMaskedSubPixelVarianceParam;
-
-class HighbdMaskedSubPixelVarianceTest :
-  public ::testing::TestWithParam<HighbdMaskedSubPixelVarianceParam> {
- public:
-  virtual ~HighbdMaskedSubPixelVarianceTest() {}
-  virtual void SetUp() {
-    opt_func_ = GET_PARAM(0);
-    ref_func_ = GET_PARAM(1);
-    bit_depth_ = GET_PARAM(2);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  MaskedSubPixelVarianceFunc opt_func_;
-  MaskedSubPixelVarianceFunc ref_func_;
-  vpx_bit_depth_t bit_depth_;
-};
-
-TEST_P(HighbdMaskedSubPixelVarianceTest, OperationCheck) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint16_t, src_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint16_t, ref_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  uint8_t* src8_ptr = CONVERT_TO_BYTEPTR(src_ptr);
-  uint8_t* ref8_ptr = CONVERT_TO_BYTEPTR(ref_ptr);
-  int err_count = 0;
-  int first_failure = -1;
-  int first_failure_x = -1;
-  int first_failure_y = -1;
-  int src_stride = (MAX_SB_SIZE+1);
-  int ref_stride = (MAX_SB_SIZE+1);
-  int msk_stride = (MAX_SB_SIZE+1);
-  int xoffset, yoffset;
-
-  for (int i = 0; i < number_of_iterations; ++i) {
-    for (xoffset = 0; xoffset < BIL_SUBPEL_SHIFTS; xoffset++) {
-      for (yoffset = 0; yoffset < BIL_SUBPEL_SHIFTS; yoffset++) {
-        for (int j = 0; j < (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1); j++) {
-          src_ptr[j] = rnd.Rand16() & ((1 << bit_depth_) - 1);
-          ref_ptr[j] = rnd.Rand16() & ((1 << bit_depth_) - 1);
-          msk_ptr[j] = rnd(65);
-        }
-
-        ref_ret = ref_func_(src8_ptr, src_stride,
-                            xoffset, yoffset,
-                            ref8_ptr, ref_stride,
-                            msk_ptr, msk_stride,
-                            &ref_sse);
-        ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src8_ptr, src_stride,
-                                                     xoffset, yoffset,
-                                                     ref8_ptr, ref_stride,
-                                                     msk_ptr, msk_stride,
-                                                     &opt_sse));
-
-        if (opt_ret != ref_ret || opt_sse != ref_sse) {
-          err_count++;
-          if (first_failure == -1) {
-            first_failure = i;
-            first_failure_x = xoffset;
-            first_failure_y = yoffset;
-          }
-        }
-      }
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-    << "Error: Masked Sub Pixel Variance Test OperationCheck,"
-    << "C output doesn't match SSSE3 output. "
-    << "First failed at test case " << first_failure
-    << " x_offset = " << first_failure_x
-    << " y_offset = " << first_failure_y;
-}
-
-TEST_P(HighbdMaskedSubPixelVarianceTest, ExtremeValues) {
-  unsigned int ref_ret, opt_ret;
-  unsigned int ref_sse, opt_sse;
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, uint16_t, src_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint16_t, ref_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  DECLARE_ALIGNED(16, uint8_t,  msk_ptr[(MAX_SB_SIZE+1)*(MAX_SB_SIZE+1)]);
-  uint8_t* src8_ptr = CONVERT_TO_BYTEPTR(src_ptr);
-  uint8_t* ref8_ptr = CONVERT_TO_BYTEPTR(ref_ptr);
-  int first_failure_x = -1;
-  int first_failure_y = -1;
-  int err_count = 0;
-  int first_failure = -1;
-  int src_stride = (MAX_SB_SIZE+1);
-  int ref_stride = (MAX_SB_SIZE+1);
-  int msk_stride = (MAX_SB_SIZE+1);
-
-  for (int xoffset = 0 ; xoffset < BIL_SUBPEL_SHIFTS ; xoffset++) {
-    for (int yoffset = 0 ; yoffset < BIL_SUBPEL_SHIFTS ; yoffset++) {
-      for (int i = 0; i < 8; ++i) {
-        vpx_memset16(src_ptr, (i & 0x1) ? ((1 << bit_depth_) - 1) : 0,
-                     (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-        vpx_memset16(ref_ptr, (i & 0x2) ? ((1 << bit_depth_) - 1) : 0,
-                     (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-        memset(msk_ptr, (i & 0x4) ?   64 : 0, (MAX_SB_SIZE+1)*(MAX_SB_SIZE+1));
-
-        ref_ret = ref_func_(src8_ptr, src_stride,
-                            xoffset, yoffset,
-                            ref8_ptr, ref_stride,
-                            msk_ptr, msk_stride,
-                            &ref_sse);
-        ASM_REGISTER_STATE_CHECK(opt_ret = opt_func_(src8_ptr, src_stride,
-                                                     xoffset, yoffset,
-                                                     ref8_ptr, ref_stride,
-                                                     msk_ptr, msk_stride,
-                                                     &opt_sse));
-
-        if (opt_ret != ref_ret || opt_sse != ref_sse) {
-          err_count++;
-          if (first_failure == -1) {
-            first_failure = i;
-            first_failure_x = xoffset;
-            first_failure_y = yoffset;
-          }
-        }
-      }
-    }
-  }
-
-  EXPECT_EQ(0, err_count)
-  << "Error: Masked Variance Test ExtremeValues,"
-  << "C output doesn't match SSSE3 output. "
-  << "First failed at test case " << first_failure
-  << " x_offset = " << first_failure_x
-  << " y_offset = " << first_failure_y;
-}
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-
-using std::tr1::make_tuple;
-
-#if HAVE_SSSE3
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, MaskedVarianceTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_variance128x128_ssse3,
-               &vpx_masked_variance128x128_c),
-    make_tuple(&vpx_masked_variance128x64_ssse3,
-               &vpx_masked_variance128x64_c),
-    make_tuple(&vpx_masked_variance64x128_ssse3,
-               &vpx_masked_variance64x128_c),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_variance64x64_ssse3,
-               &vpx_masked_variance64x64_c),
-    make_tuple(&vpx_masked_variance64x32_ssse3,
-               &vpx_masked_variance64x32_c),
-    make_tuple(&vpx_masked_variance32x64_ssse3,
-               &vpx_masked_variance32x64_c),
-    make_tuple(&vpx_masked_variance32x32_ssse3,
-               &vpx_masked_variance32x32_c),
-    make_tuple(&vpx_masked_variance32x16_ssse3,
-               &vpx_masked_variance32x16_c),
-    make_tuple(&vpx_masked_variance16x32_ssse3,
-               &vpx_masked_variance16x32_c),
-    make_tuple(&vpx_masked_variance16x16_ssse3,
-               &vpx_masked_variance16x16_c),
-    make_tuple(&vpx_masked_variance16x8_ssse3,
-               &vpx_masked_variance16x8_c),
-    make_tuple(&vpx_masked_variance8x16_ssse3,
-               &vpx_masked_variance8x16_c),
-    make_tuple(&vpx_masked_variance8x8_ssse3,
-               &vpx_masked_variance8x8_c),
-    make_tuple(&vpx_masked_variance8x4_ssse3,
-               &vpx_masked_variance8x4_c),
-    make_tuple(&vpx_masked_variance4x8_ssse3,
-               &vpx_masked_variance4x8_c),
-    make_tuple(&vpx_masked_variance4x4_ssse3,
-               &vpx_masked_variance4x4_c)));
-
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, MaskedSubPixelVarianceTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_sub_pixel_variance128x128_ssse3,
-               &vpx_masked_sub_pixel_variance128x128_c),
-    make_tuple(&vpx_masked_sub_pixel_variance128x64_ssse3,
-               &vpx_masked_sub_pixel_variance128x64_c),
-    make_tuple(&vpx_masked_sub_pixel_variance64x128_ssse3,
-               &vpx_masked_sub_pixel_variance64x128_c),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_masked_sub_pixel_variance64x64_ssse3,
-               &vpx_masked_sub_pixel_variance64x64_c),
-    make_tuple(&vpx_masked_sub_pixel_variance64x32_ssse3,
-               &vpx_masked_sub_pixel_variance64x32_c),
-    make_tuple(&vpx_masked_sub_pixel_variance32x64_ssse3,
-               &vpx_masked_sub_pixel_variance32x64_c),
-    make_tuple(&vpx_masked_sub_pixel_variance32x32_ssse3,
-               &vpx_masked_sub_pixel_variance32x32_c),
-    make_tuple(&vpx_masked_sub_pixel_variance32x16_ssse3,
-               &vpx_masked_sub_pixel_variance32x16_c),
-    make_tuple(&vpx_masked_sub_pixel_variance16x32_ssse3,
-               &vpx_masked_sub_pixel_variance16x32_c),
-    make_tuple(&vpx_masked_sub_pixel_variance16x16_ssse3,
-               &vpx_masked_sub_pixel_variance16x16_c),
-    make_tuple(&vpx_masked_sub_pixel_variance16x8_ssse3,
-               &vpx_masked_sub_pixel_variance16x8_c),
-    make_tuple(&vpx_masked_sub_pixel_variance8x16_ssse3,
-               &vpx_masked_sub_pixel_variance8x16_c),
-    make_tuple(&vpx_masked_sub_pixel_variance8x8_ssse3,
-               &vpx_masked_sub_pixel_variance8x8_c),
-    make_tuple(&vpx_masked_sub_pixel_variance8x4_ssse3,
-               &vpx_masked_sub_pixel_variance8x4_c),
-    make_tuple(&vpx_masked_sub_pixel_variance4x8_ssse3,
-               &vpx_masked_sub_pixel_variance4x8_c),
-    make_tuple(&vpx_masked_sub_pixel_variance4x4_ssse3,
-               &vpx_masked_sub_pixel_variance4x4_c)));
-
-#if CONFIG_VPX_HIGHBITDEPTH
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, HighbdMaskedVarianceTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_variance128x128_ssse3,
-               &vpx_highbd_masked_variance128x128_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance128x64_ssse3,
-               &vpx_highbd_masked_variance128x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance64x128_ssse3,
-               &vpx_highbd_masked_variance64x128_c, VPX_BITS_8),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_variance64x64_ssse3,
-               &vpx_highbd_masked_variance64x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance64x32_ssse3,
-               &vpx_highbd_masked_variance64x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance32x64_ssse3,
-               &vpx_highbd_masked_variance32x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance32x32_ssse3,
-               &vpx_highbd_masked_variance32x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance32x16_ssse3,
-               &vpx_highbd_masked_variance32x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance16x32_ssse3,
-               &vpx_highbd_masked_variance16x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance16x16_ssse3,
-               &vpx_highbd_masked_variance16x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance16x8_ssse3,
-               &vpx_highbd_masked_variance16x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance8x16_ssse3,
-               &vpx_highbd_masked_variance8x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance8x8_ssse3,
-               &vpx_highbd_masked_variance8x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance8x4_ssse3,
-               &vpx_highbd_masked_variance8x4_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance4x8_ssse3,
-               &vpx_highbd_masked_variance4x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_variance4x4_ssse3,
-               &vpx_highbd_masked_variance4x4_c, VPX_BITS_8),
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_10_masked_variance128x128_ssse3,
-               &vpx_highbd_10_masked_variance128x128_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance128x64_ssse3,
-               &vpx_highbd_10_masked_variance128x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance64x128_ssse3,
-               &vpx_highbd_10_masked_variance64x128_c, VPX_BITS_10),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_10_masked_variance64x64_ssse3,
-               &vpx_highbd_10_masked_variance64x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance64x32_ssse3,
-               &vpx_highbd_10_masked_variance64x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance32x64_ssse3,
-               &vpx_highbd_10_masked_variance32x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance32x32_ssse3,
-               &vpx_highbd_10_masked_variance32x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance32x16_ssse3,
-               &vpx_highbd_10_masked_variance32x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance16x32_ssse3,
-               &vpx_highbd_10_masked_variance16x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance16x16_ssse3,
-               &vpx_highbd_10_masked_variance16x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance16x8_ssse3,
-               &vpx_highbd_10_masked_variance16x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance8x16_ssse3,
-               &vpx_highbd_10_masked_variance8x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance8x8_ssse3,
-               &vpx_highbd_10_masked_variance8x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance8x4_ssse3,
-               &vpx_highbd_10_masked_variance8x4_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance4x8_ssse3,
-               &vpx_highbd_10_masked_variance4x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_variance4x4_ssse3,
-               &vpx_highbd_10_masked_variance4x4_c, VPX_BITS_10),
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_12_masked_variance128x128_ssse3,
-               &vpx_highbd_12_masked_variance128x128_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance128x64_ssse3,
-               &vpx_highbd_12_masked_variance128x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance64x128_ssse3,
-               &vpx_highbd_12_masked_variance64x128_c, VPX_BITS_12),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_12_masked_variance64x64_ssse3,
-               &vpx_highbd_12_masked_variance64x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance64x32_ssse3,
-               &vpx_highbd_12_masked_variance64x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance32x64_ssse3,
-               &vpx_highbd_12_masked_variance32x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance32x32_ssse3,
-               &vpx_highbd_12_masked_variance32x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance32x16_ssse3,
-               &vpx_highbd_12_masked_variance32x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance16x32_ssse3,
-               &vpx_highbd_12_masked_variance16x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance16x16_ssse3,
-               &vpx_highbd_12_masked_variance16x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance16x8_ssse3,
-               &vpx_highbd_12_masked_variance16x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance8x16_ssse3,
-               &vpx_highbd_12_masked_variance8x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance8x8_ssse3,
-               &vpx_highbd_12_masked_variance8x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance8x4_ssse3,
-               &vpx_highbd_12_masked_variance8x4_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance4x8_ssse3,
-               &vpx_highbd_12_masked_variance4x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_variance4x4_ssse3,
-               &vpx_highbd_12_masked_variance4x4_c, VPX_BITS_12)));
-
-INSTANTIATE_TEST_CASE_P(
-  SSSE3_C_COMPARE, HighbdMaskedSubPixelVarianceTest,
-  ::testing::Values(
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance128x128_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance128x128_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance128x64_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance128x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance64x128_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance64x128_c, VPX_BITS_8),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance64x64_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance64x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance64x32_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance64x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance32x64_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance32x64_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance32x32_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance32x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance32x16_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance32x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance16x32_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance16x32_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance16x16_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance16x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance16x8_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance16x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance8x16_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance8x16_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance8x8_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance8x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance8x4_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance8x4_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance4x8_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance4x8_c, VPX_BITS_8),
-    make_tuple(&vpx_highbd_masked_sub_pixel_variance4x4_ssse3,
-               &vpx_highbd_masked_sub_pixel_variance4x4_c, VPX_BITS_8),
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance128x128_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance128x128_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance128x64_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance128x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance64x128_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance64x128_c, VPX_BITS_10),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance64x64_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance64x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance64x32_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance64x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance32x64_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance32x64_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance32x32_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance32x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance32x16_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance32x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance16x32_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance16x32_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance16x16_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance16x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance16x8_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance16x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance8x16_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance8x16_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance8x8_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance8x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance8x4_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance8x4_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance4x8_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance4x8_c, VPX_BITS_10),
-    make_tuple(&vpx_highbd_10_masked_sub_pixel_variance4x4_ssse3,
-               &vpx_highbd_10_masked_sub_pixel_variance4x4_c, VPX_BITS_10),
-#if CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance128x128_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance128x128_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance128x64_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance128x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance64x128_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance64x128_c, VPX_BITS_12),
-#endif  // CONFIG_EXT_PARTITION
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance64x64_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance64x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance64x32_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance64x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance32x64_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance32x64_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance32x32_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance32x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance32x16_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance32x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance16x32_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance16x32_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance16x16_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance16x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance16x8_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance16x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance8x16_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance8x16_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance8x8_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance8x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance8x4_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance8x4_c, VPX_BITS_12) ,
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance4x8_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance4x8_c, VPX_BITS_12),
-    make_tuple(&vpx_highbd_12_masked_sub_pixel_variance4x4_ssse3,
-               &vpx_highbd_12_masked_sub_pixel_variance4x4_c, VPX_BITS_12)));
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-
-#endif  // HAVE_SSSE3
-}  // namespace
--- a/test/md5_helper.h
+++ b/test/md5_helper.h
@@ -42,10 +42,6 @@ class MD5 {
    }
  }

-  void Add(const uint8_t *data, size_t size) {
-    MD5Update(&md5_, data, static_cast<uint32_t>(size));
-  }
-
  const char *Get(void) {
    static const char hex[16] = {
      '0', '1', '2', '3', '4', '5', '6', '7',
--- a/test/minmax_test.cc
+++ b/test/minmax_test.cc
@@ -1,132 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include <stdlib.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-#include "test/acm_random.h"
-#include "test/register_state_check.h"
-
-namespace {
-
-using ::libvpx_test::ACMRandom;
-
-typedef void (*MinMaxFunc)(const uint8_t *a, int a_stride,
-                           const uint8_t *b, int b_stride,
-                           int *min, int *max);
-
-class MinMaxTest : public ::testing::TestWithParam<MinMaxFunc> {
- public:
-  virtual void SetUp() {
-    mm_func_ = GetParam();
-    rnd_.Reset(ACMRandom::DeterministicSeed());
-  }
-
- protected:
-  MinMaxFunc mm_func_;
-  ACMRandom rnd_;
-};
-
-void reference_minmax(const uint8_t *a, int a_stride,
-                      const uint8_t *b, int b_stride,
-                      int *min_ret, int *max_ret) {
-  int min = 255;
-  int max = 0;
-  for (int i = 0; i < 8; i++) {
-    for (int j = 0; j < 8; j++) {
-      const int diff = abs(a[i * a_stride + j] - b[i * b_stride + j]);
-      if (min > diff) min = diff;
-      if (max < diff) max = diff;
-    }
-  }
-
-  *min_ret = min;
-  *max_ret = max;
-}
-
-TEST_P(MinMaxTest, MinValue) {
-  for (int i = 0; i < 64; i++) {
-    uint8_t a[64], b[64];
-    memset(a, 0, sizeof(a));
-    memset(b, 255, sizeof(b));
-    b[i] = i;  // Set a minimum difference of i.
-
-    int min, max;
-    ASM_REGISTER_STATE_CHECK(mm_func_(a, 8, b, 8, &min, &max));
-    EXPECT_EQ(255, max);
-    EXPECT_EQ(i, min);
-  }
-}
-
-TEST_P(MinMaxTest, MaxValue) {
-  for (int i = 0; i < 64; i++) {
-    uint8_t a[64], b[64];
-    memset(a, 0, sizeof(a));
-    memset(b, 0, sizeof(b));
-    b[i] = i;  // Set a maximum difference of i.
-
-    int min, max;
-    ASM_REGISTER_STATE_CHECK(mm_func_(a, 8, b, 8, &min, &max));
-    EXPECT_EQ(i, max);
-    EXPECT_EQ(0, min);
-  }
-}
-
-TEST_P(MinMaxTest, CompareReference) {
-  uint8_t a[64], b[64];
-  for (int j = 0; j < 64; j++) {
-    a[j] = rnd_.Rand8();
-    b[j] = rnd_.Rand8();
-  }
-
-  int min_ref, max_ref, min, max;
-  reference_minmax(a, 8, b, 8, &min_ref, &max_ref);
-  ASM_REGISTER_STATE_CHECK(mm_func_(a, 8, b, 8, &min, &max));
-  EXPECT_EQ(max_ref, max);
-  EXPECT_EQ(min_ref, min);
-}
-
-TEST_P(MinMaxTest, CompareReferenceAndVaryStride) {
-  uint8_t a[8 * 64], b[8 * 64];
-  for (int i = 0; i < 8 * 64; i++) {
-    a[i] = rnd_.Rand8();
-    b[i] = rnd_.Rand8();
-  }
-  for (int a_stride = 8; a_stride <= 64; a_stride += 8) {
-    for (int b_stride = 8; b_stride <= 64; b_stride += 8) {
-      int min_ref, max_ref, min, max;
-      reference_minmax(a, a_stride, b, b_stride, &min_ref, &max_ref);
-      ASM_REGISTER_STATE_CHECK(mm_func_(a, a_stride, b, b_stride, &min, &max));
-      EXPECT_EQ(max_ref, max) << "when a_stride = " << a_stride
-                              << " and b_stride = " << b_stride;;
-      EXPECT_EQ(min_ref, min) << "when a_stride = " << a_stride
-                              << " and b_stride = " << b_stride;;
-    }
-  }
-}
-
-INSTANTIATE_TEST_CASE_P(C, MinMaxTest, ::testing::Values(&vpx_minmax_8x8_c));
-
-#if HAVE_SSE2
-INSTANTIATE_TEST_CASE_P(SSE2, MinMaxTest,
-                        ::testing::Values(&vpx_minmax_8x8_sse2));
-#endif
-
-#if HAVE_NEON
-INSTANTIATE_TEST_CASE_P(NEON, MinMaxTest,
-                        ::testing::Values(&vpx_minmax_8x8_neon));
-#endif
-
-}  // namespace
--- a/test/obmc_sad_test.cc
+++ b/test/obmc_sad_test.cc
@@ -1,192 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "test/function_equivalence_test.h"
-#include "test/register_state_check.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-#define MAX_SB_SQUARE (MAX_SB_SIZE * MAX_SB_SIZE)
-
-using libvpx_test::FunctionEquivalenceTest;
-
-namespace {
-
-static const int kIterations = 1000;
-static const int kMaskMax = 64;
-
-typedef unsigned int (*ObmcSadF)(const uint8_t *pre, int pre_stride,
-                                 const int32_t *wsrc, const int32_t *mask);
-typedef libvpx_test::FuncParam<ObmcSadF> TestFuncs;
-
-////////////////////////////////////////////////////////////////////////////////
-// 8 bit
-////////////////////////////////////////////////////////////////////////////////
-
-class ObmcSadTest : public FunctionEquivalenceTest<ObmcSadF> {};
-
-TEST_P(ObmcSadTest, RandomValues) {
-  DECLARE_ALIGNED(32, uint8_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = rng_(MAX_SB_SIZE + 1);
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = rng_.Rand8();
-      wsrc[i] = rng_.Rand8() * rng_(kMaskMax * kMaskMax + 1);
-      mask[i] = rng_(kMaskMax * kMaskMax + 1);
-    }
-
-    const unsigned int ref_res = params_.ref_func(pre, pre_stride, wsrc, mask);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(tst_res =
-        params_.tst_func(pre, pre_stride, wsrc, mask));
-
-    ASSERT_EQ(ref_res, tst_res);
-  }
-}
-
-TEST_P(ObmcSadTest, ExtremeValues) {
-  DECLARE_ALIGNED(32, uint8_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < MAX_SB_SIZE && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = iter;
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = UINT8_MAX;
-      wsrc[i] = UINT8_MAX * kMaskMax * kMaskMax;
-      mask[i] = kMaskMax * kMaskMax;
-    }
-
-    const unsigned int ref_res = params_.ref_func(pre, pre_stride, wsrc, mask);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(tst_res =
-        params_.tst_func(pre, pre_stride, wsrc, mask));
-
-    ASSERT_EQ(ref_res, tst_res);
-  }
-}
-
-#if HAVE_SSE4_1
-const ObmcSadTest::ParamType sse4_functions[] = {
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_obmc_sad128x128_c, vpx_obmc_sad128x128_sse4_1),
-  TestFuncs(vpx_obmc_sad128x64_c, vpx_obmc_sad128x64_sse4_1),
-  TestFuncs(vpx_obmc_sad64x128_c, vpx_obmc_sad64x128_sse4_1),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_obmc_sad64x64_c, vpx_obmc_sad64x64_sse4_1),
-  TestFuncs(vpx_obmc_sad64x32_c, vpx_obmc_sad64x32_sse4_1),
-  TestFuncs(vpx_obmc_sad32x64_c, vpx_obmc_sad32x64_sse4_1),
-  TestFuncs(vpx_obmc_sad32x32_c, vpx_obmc_sad32x32_sse4_1),
-  TestFuncs(vpx_obmc_sad32x16_c, vpx_obmc_sad32x16_sse4_1),
-  TestFuncs(vpx_obmc_sad16x32_c, vpx_obmc_sad16x32_sse4_1),
-  TestFuncs(vpx_obmc_sad16x16_c, vpx_obmc_sad16x16_sse4_1),
-  TestFuncs(vpx_obmc_sad16x8_c, vpx_obmc_sad16x8_sse4_1),
-  TestFuncs(vpx_obmc_sad8x16_c, vpx_obmc_sad8x16_sse4_1),
-  TestFuncs(vpx_obmc_sad8x8_c, vpx_obmc_sad8x8_sse4_1),
-  TestFuncs(vpx_obmc_sad8x4_c, vpx_obmc_sad8x4_sse4_1),
-  TestFuncs(vpx_obmc_sad4x8_c, vpx_obmc_sad4x8_sse4_1),
-  TestFuncs(vpx_obmc_sad4x4_c, vpx_obmc_sad4x4_sse4_1)
-};
-
-INSTANTIATE_TEST_CASE_P(SSE4_1_C_COMPARE, ObmcSadTest,
-                        ::testing::ValuesIn(sse4_functions));
-#endif  // HAVE_SSE4_1
-
-////////////////////////////////////////////////////////////////////////////////
-// High bit-depth
-////////////////////////////////////////////////////////////////////////////////
-
-#if CONFIG_VPX_HIGHBITDEPTH
-class ObmcSadHBDTest : public FunctionEquivalenceTest<ObmcSadF> {};
-
-TEST_P(ObmcSadHBDTest, RandomValues) {
-  DECLARE_ALIGNED(32, uint16_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = rng_(MAX_SB_SIZE + 1);
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = rng_(1<<12);
-      wsrc[i] = rng_(1<<12) * rng_(kMaskMax * kMaskMax + 1);
-      mask[i] = rng_(kMaskMax * kMaskMax + 1);
-    }
-
-    const unsigned int ref_res =
-        params_.ref_func(CONVERT_TO_BYTEPTR(pre), pre_stride, wsrc, mask);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(tst_res =
-        params_.tst_func(CONVERT_TO_BYTEPTR(pre), pre_stride, wsrc, mask));
-
-    ASSERT_EQ(ref_res, tst_res);
-  }
-}
-
-TEST_P(ObmcSadHBDTest, ExtremeValues) {
-  DECLARE_ALIGNED(32, uint16_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < MAX_SB_SIZE && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = iter;
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = (1 << 12) - 1;
-      wsrc[i] = ((1 << 12) - 1) * kMaskMax * kMaskMax;
-      mask[i] = kMaskMax * kMaskMax;
-    }
-
-    const unsigned int ref_res =
-        params_.ref_func(CONVERT_TO_BYTEPTR(pre), pre_stride, wsrc, mask);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(tst_res =
-        params_.tst_func(CONVERT_TO_BYTEPTR(pre), pre_stride, wsrc, mask));
-
-    ASSERT_EQ(ref_res, tst_res);
-  }
-}
-
-#if HAVE_SSE4_1
-ObmcSadHBDTest::ParamType sse4_functions_hbd[] = {
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_obmc_sad128x128_c, vpx_highbd_obmc_sad128x128_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad128x64_c, vpx_highbd_obmc_sad128x64_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad64x128_c, vpx_highbd_obmc_sad64x128_sse4_1),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_obmc_sad64x64_c, vpx_highbd_obmc_sad64x64_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad64x32_c, vpx_highbd_obmc_sad64x32_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad32x64_c, vpx_highbd_obmc_sad32x64_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad32x32_c, vpx_highbd_obmc_sad32x32_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad32x16_c, vpx_highbd_obmc_sad32x16_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad16x32_c, vpx_highbd_obmc_sad16x32_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad16x16_c, vpx_highbd_obmc_sad16x16_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad16x8_c, vpx_highbd_obmc_sad16x8_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad8x16_c, vpx_highbd_obmc_sad8x16_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad8x8_c, vpx_highbd_obmc_sad8x8_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad8x4_c, vpx_highbd_obmc_sad8x4_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad4x8_c, vpx_highbd_obmc_sad4x8_sse4_1),
-  TestFuncs(vpx_highbd_obmc_sad4x4_c, vpx_highbd_obmc_sad4x4_sse4_1)
-};
-
-INSTANTIATE_TEST_CASE_P(SSE4_1_C_COMPARE, ObmcSadHBDTest,
-                        ::testing::ValuesIn(sse4_functions_hbd));
-#endif  // HAVE_SSE4_1
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-}  // namespace
--- a/test/obmc_variance_test.cc
+++ b/test/obmc_variance_test.cc
@@ -1,294 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-#include "test/acm_random.h"
-
-#include "test/function_equivalence_test.h"
-#include "test/register_state_check.h"
-
-#include "./vpx_config.h"
-#include "./vpx_dsp_rtcd.h"
-#include "vpx/vpx_integer.h"
-
-#define MAX_SB_SQUARE (MAX_SB_SIZE * MAX_SB_SIZE)
-
-using libvpx_test::ACMRandom;
-using libvpx_test::FunctionEquivalenceTest;
-
-namespace {
-
-static const int kIterations = 1000;
-static const int kMaskMax = 64;
-
-typedef unsigned int (*ObmcVarF)(const uint8_t *pre, int pre_stride,
-                                 const int32_t *wsrc, const int32_t *mask,
-                                 unsigned int *sse);
-typedef libvpx_test::FuncParam<ObmcVarF> TestFuncs;
-
-////////////////////////////////////////////////////////////////////////////////
-// 8 bit
-////////////////////////////////////////////////////////////////////////////////
-
-class ObmcVarianceTest : public FunctionEquivalenceTest<ObmcVarF> {};
-
-TEST_P(ObmcVarianceTest, RandomValues) {
-  DECLARE_ALIGNED(32, uint8_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = this->rng_(MAX_SB_SIZE + 1);
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = this->rng_.Rand8();
-      wsrc[i] = this->rng_.Rand8() * this->rng_(kMaskMax * kMaskMax + 1);
-      mask[i] = this->rng_(kMaskMax * kMaskMax + 1);
-    }
-
-    unsigned int ref_sse, tst_sse;
-    const unsigned int ref_res = params_.ref_func(pre, pre_stride, wsrc, mask,
-                                                  &ref_sse);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(
-        tst_res = params_.tst_func(pre, pre_stride, wsrc, mask, &tst_sse));
-
-    ASSERT_EQ(ref_res, tst_res);
-    ASSERT_EQ(ref_sse, tst_sse);
-  }
-}
-
-TEST_P(ObmcVarianceTest, ExtremeValues) {
-  DECLARE_ALIGNED(32, uint8_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < MAX_SB_SIZE && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = iter;
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = UINT8_MAX;
-      wsrc[i] = UINT8_MAX * kMaskMax * kMaskMax;
-      mask[i] = kMaskMax * kMaskMax;
-    }
-
-    unsigned int ref_sse, tst_sse;
-    const unsigned int ref_res = params_.ref_func(pre, pre_stride, wsrc, mask,
-                                                  &ref_sse);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(
-        tst_res = params_.tst_func(pre, pre_stride, wsrc, mask, &tst_sse));
-
-    ASSERT_EQ(ref_res, tst_res);
-    ASSERT_EQ(ref_sse, tst_sse);
-  }
-}
-
-#if HAVE_SSE4_1
-const ObmcVarianceTest::ParamType sse4_functions[] = {
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_obmc_variance128x128_c, vpx_obmc_variance128x128_sse4_1),
-  TestFuncs(vpx_obmc_variance128x64_c, vpx_obmc_variance128x64_sse4_1),
-  TestFuncs(vpx_obmc_variance64x128_c, vpx_obmc_variance64x128_sse4_1),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_obmc_variance64x64_c, vpx_obmc_variance64x64_sse4_1),
-  TestFuncs(vpx_obmc_variance64x32_c, vpx_obmc_variance64x32_sse4_1),
-  TestFuncs(vpx_obmc_variance32x64_c, vpx_obmc_variance32x64_sse4_1),
-  TestFuncs(vpx_obmc_variance32x32_c, vpx_obmc_variance32x32_sse4_1),
-  TestFuncs(vpx_obmc_variance32x16_c, vpx_obmc_variance32x16_sse4_1),
-  TestFuncs(vpx_obmc_variance16x32_c, vpx_obmc_variance16x32_sse4_1),
-  TestFuncs(vpx_obmc_variance16x16_c, vpx_obmc_variance16x16_sse4_1),
-  TestFuncs(vpx_obmc_variance16x8_c, vpx_obmc_variance16x8_sse4_1),
-  TestFuncs(vpx_obmc_variance8x16_c, vpx_obmc_variance8x16_sse4_1),
-  TestFuncs(vpx_obmc_variance8x8_c, vpx_obmc_variance8x8_sse4_1),
-  TestFuncs(vpx_obmc_variance8x4_c, vpx_obmc_variance8x4_sse4_1),
-  TestFuncs(vpx_obmc_variance4x8_c, vpx_obmc_variance4x8_sse4_1),
-  TestFuncs(vpx_obmc_variance4x4_c, vpx_obmc_variance4x4_sse4_1)
-};
-
-INSTANTIATE_TEST_CASE_P(SSE4_1_C_COMPARE, ObmcVarianceTest,
-                        ::testing::ValuesIn(sse4_functions));
-#endif  // HAVE_SSE4_1
-
-////////////////////////////////////////////////////////////////////////////////
-// High bit-depth
-////////////////////////////////////////////////////////////////////////////////
-
-#if CONFIG_VPX_HIGHBITDEPTH
-class ObmcVarianceHBDTest : public FunctionEquivalenceTest<ObmcVarF> {};
-
-TEST_P(ObmcVarianceHBDTest, RandomValues) {
-  DECLARE_ALIGNED(32, uint16_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < kIterations && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = this->rng_(MAX_SB_SIZE + 1);
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = this->rng_(1 << params_.bit_depth);
-      wsrc[i] = this->rng_(1 << params_.bit_depth) *
-                this->rng_(kMaskMax * kMaskMax + 1);
-      mask[i] = this->rng_(kMaskMax * kMaskMax + 1);
-    }
-
-    unsigned int ref_sse, tst_sse;
-    const unsigned int ref_res = params_.ref_func(CONVERT_TO_BYTEPTR(pre),
-                                                  pre_stride,
-                                                  wsrc, mask, &ref_sse);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(
-        tst_res = params_.tst_func(CONVERT_TO_BYTEPTR(pre),
-                                   pre_stride, wsrc, mask, &tst_sse));
-
-    ASSERT_EQ(ref_res, tst_res);
-    ASSERT_EQ(ref_sse, tst_sse);
-  }
-}
-
-TEST_P(ObmcVarianceHBDTest, ExtremeValues) {
-  DECLARE_ALIGNED(32, uint16_t, pre[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, wsrc[MAX_SB_SQUARE]);
-  DECLARE_ALIGNED(32, int32_t, mask[MAX_SB_SQUARE]);
-
-  for (int iter = 0 ; iter < MAX_SB_SIZE && !HasFatalFailure() ; ++iter) {
-    const int pre_stride = iter;
-
-    for (int i = 0 ; i < MAX_SB_SQUARE ; ++i) {
-      pre[i] = (1 << params_.bit_depth) - 1;
-      wsrc[i] = ((1 << params_.bit_depth) - 1) * kMaskMax * kMaskMax;
-      mask[i] = kMaskMax * kMaskMax;
-    }
-
-    unsigned int ref_sse, tst_sse;
-    const unsigned int ref_res = params_.ref_func(CONVERT_TO_BYTEPTR(pre),
-                                                  pre_stride,
-                                                  wsrc, mask, &ref_sse);
-    unsigned int tst_res;
-    ASM_REGISTER_STATE_CHECK(
-        tst_res = params_.tst_func(CONVERT_TO_BYTEPTR(pre), pre_stride,
-                                   wsrc, mask, &tst_sse));
-
-    ASSERT_EQ(ref_res, tst_res);
-    ASSERT_EQ(ref_sse, tst_sse);
-  }
-}
-
-#if HAVE_SSE4_1
-ObmcVarianceHBDTest::ParamType sse4_functions_hbd[] = {
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_obmc_variance128x128_c,
-            vpx_highbd_obmc_variance128x128_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance128x64_c,
-            vpx_highbd_obmc_variance128x64_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance64x128_c,
-            vpx_highbd_obmc_variance64x128_sse4_1, 8),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_obmc_variance64x64_c,
-            vpx_highbd_obmc_variance64x64_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance64x32_c,
-            vpx_highbd_obmc_variance64x32_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance32x64_c,
-            vpx_highbd_obmc_variance32x64_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance32x32_c,
-            vpx_highbd_obmc_variance32x32_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance32x16_c,
-            vpx_highbd_obmc_variance32x16_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance16x32_c,
-            vpx_highbd_obmc_variance16x32_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance16x16_c,
-            vpx_highbd_obmc_variance16x16_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance16x8_c,
-            vpx_highbd_obmc_variance16x8_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance8x16_c,
-            vpx_highbd_obmc_variance8x16_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance8x8_c,
-            vpx_highbd_obmc_variance8x8_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance8x4_c,
-            vpx_highbd_obmc_variance8x4_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance4x8_c,
-            vpx_highbd_obmc_variance4x8_sse4_1, 8),
-  TestFuncs(vpx_highbd_obmc_variance4x4_c,
-            vpx_highbd_obmc_variance4x4_sse4_1, 8),
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_10_obmc_variance128x128_c,
-            vpx_highbd_10_obmc_variance128x128_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance128x64_c,
-            vpx_highbd_10_obmc_variance128x64_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance64x128_c,
-            vpx_highbd_10_obmc_variance64x128_sse4_1, 10),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_10_obmc_variance64x64_c,
-            vpx_highbd_10_obmc_variance64x64_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance64x32_c,
-            vpx_highbd_10_obmc_variance64x32_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance32x64_c,
-            vpx_highbd_10_obmc_variance32x64_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance32x32_c,
-            vpx_highbd_10_obmc_variance32x32_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance32x16_c,
-            vpx_highbd_10_obmc_variance32x16_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance16x32_c,
-            vpx_highbd_10_obmc_variance16x32_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance16x16_c,
-            vpx_highbd_10_obmc_variance16x16_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance16x8_c,
-            vpx_highbd_10_obmc_variance16x8_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance8x16_c,
-            vpx_highbd_10_obmc_variance8x16_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance8x8_c,
-            vpx_highbd_10_obmc_variance8x8_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance8x4_c,
-            vpx_highbd_10_obmc_variance8x4_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance4x8_c,
-            vpx_highbd_10_obmc_variance4x8_sse4_1, 10),
-  TestFuncs(vpx_highbd_10_obmc_variance4x4_c,
-            vpx_highbd_10_obmc_variance4x4_sse4_1, 10),
-#if CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_12_obmc_variance128x128_c,
-            vpx_highbd_12_obmc_variance128x128_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance128x64_c,
-            vpx_highbd_12_obmc_variance128x64_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance64x128_c,
-            vpx_highbd_12_obmc_variance64x128_sse4_1, 12),
-#endif  // CONFIG_EXT_PARTITION
-  TestFuncs(vpx_highbd_12_obmc_variance64x64_c,
-            vpx_highbd_12_obmc_variance64x64_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance64x32_c,
-            vpx_highbd_12_obmc_variance64x32_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance32x64_c,
-            vpx_highbd_12_obmc_variance32x64_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance32x32_c,
-            vpx_highbd_12_obmc_variance32x32_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance32x16_c,
-            vpx_highbd_12_obmc_variance32x16_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance16x32_c,
-            vpx_highbd_12_obmc_variance16x32_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance16x16_c,
-            vpx_highbd_12_obmc_variance16x16_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance16x8_c,
-            vpx_highbd_12_obmc_variance16x8_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance8x16_c,
-            vpx_highbd_12_obmc_variance8x16_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance8x8_c,
-            vpx_highbd_12_obmc_variance8x8_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance8x4_c,
-            vpx_highbd_12_obmc_variance8x4_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance4x8_c,
-            vpx_highbd_12_obmc_variance4x8_sse4_1, 12),
-  TestFuncs(vpx_highbd_12_obmc_variance4x4_c,
-            vpx_highbd_12_obmc_variance4x4_sse4_1, 12)
-};
-
-INSTANTIATE_TEST_CASE_P(SSE4_1_C_COMPARE, ObmcVarianceHBDTest,
-                        ::testing::ValuesIn(sse4_functions_hbd));
-#endif  // HAVE_SSE4_1
-#endif  // CONFIG_VPX_HIGHBITDEPTH
-}  // namespace
--- a/test/partial_idct_test.cc
+++ b/test/partial_idct_test.cc
@@ -13,15 +13,14 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp10_rtcd.h"
-#include "./vpx_dsp_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vp10/common/blockd.h"
-#include "vp10/common/scan.h"
+
+#include "./vp9_rtcd.h"
+#include "vp9/common/vp9_blockd.h"
+#include "vp9/common/vp9_scan.h"
 #include "vpx/vpx_integer.h"

 using libvpx_test::ACMRandom;
@@ -75,16 +74,16 @@ TEST_P(PartialIDctTest, RunQuantCheck) {
      FAIL() << "Wrong Size!";
      break;
  }
-  DECLARE_ALIGNED(16, tran_low_t, test_coef_block1[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, test_coef_block2[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst1[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst2[kMaxNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_coef_block1, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_coef_block2, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst1, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst2, kMaxNumCoeffs);

  const int count_test_block = 1000;
  const int block_size = size * size;

-  DECLARE_ALIGNED(16, int16_t, input_extreme_block[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kMaxNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, input_extreme_block, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, output_ref_block, kMaxNumCoeffs);

  int max_error = 0;
  for (int i = 0; i < count_test_block; ++i) {
@@ -115,7 +114,7 @@ TEST_P(PartialIDctTest, RunQuantCheck) {
      // quantization with maximum allowed step sizes
      test_coef_block1[0] = (output_ref_block[0] / 1336) * 1336;
      for (int j = 1; j < last_nonzero_; ++j)
-        test_coef_block1[vp10_default_scan_orders[tx_size_].scan[j]]
+        test_coef_block1[vp9_default_scan_orders[tx_size_].scan[j]]
                         = (output_ref_block[j] / 1828) * 1828;
    }

@@ -154,10 +153,10 @@ TEST_P(PartialIDctTest, ResultsMatch) {
      FAIL() << "Wrong Size!";
      break;
  }
-  DECLARE_ALIGNED(16, tran_low_t, test_coef_block1[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, tran_low_t, test_coef_block2[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst1[kMaxNumCoeffs]);
-  DECLARE_ALIGNED(16, uint8_t, dst2[kMaxNumCoeffs]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_coef_block1, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, test_coef_block2, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst1, kMaxNumCoeffs);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, dst2, kMaxNumCoeffs);
  const int count_test_block = 1000;
  const int max_coeff = 32766 / 4;
  const int block_size = size * size;
@@ -177,7 +176,7 @@ TEST_P(PartialIDctTest, ResultsMatch) {
        max_energy_leftover = 0;
        coef = 0;
      }
-      test_coef_block1[vp10_default_scan_orders[tx_size_].scan[j]] = coef;
+      test_coef_block1[vp9_default_scan_orders[tx_size_].scan[j]] = coef;
    }

    memcpy(test_coef_block2, test_coef_block1,
@@ -202,142 +201,117 @@ using std::tr1::make_tuple;
 INSTANTIATE_TEST_CASE_P(
    C, PartialIDctTest,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_34_add_c,
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c,
+                   &vp9_idct32x32_34_add_c,
                   TX_32X32, 34),
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_1_add_c,
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c,
+                   &vp9_idct32x32_1_add_c,
                   TX_32X32, 1),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_10_add_c,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_10_add_c,
                   TX_16X16, 10),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_1_add_c,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_1_add_c,
                   TX_16X16, 1),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_12_add_c,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_12_add_c,
                   TX_8X8, 12),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_1_add_c,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_1_add_c,
                   TX_8X8, 1),
-        make_tuple(&vpx_fdct4x4_c,
-                   &vpx_idct4x4_16_add_c,
-                   &vpx_idct4x4_1_add_c,
+        make_tuple(&vp9_fdct4x4_c,
+                   &vp9_idct4x4_16_add_c,
+                   &vp9_idct4x4_1_add_c,
                   TX_4X4, 1)));

-#if HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_NEON
 INSTANTIATE_TEST_CASE_P(
    NEON, PartialIDctTest,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_1_add_neon,
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c,
+                   &vp9_idct32x32_1_add_neon,
                   TX_32X32, 1),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_10_add_neon,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_10_add_neon,
                   TX_16X16, 10),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_1_add_neon,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_1_add_neon,
                   TX_16X16, 1),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_12_add_neon,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_12_add_neon,
                   TX_8X8, 12),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_1_add_neon,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_1_add_neon,
                   TX_8X8, 1),
-        make_tuple(&vpx_fdct4x4_c,
-                   &vpx_idct4x4_16_add_c,
-                   &vpx_idct4x4_1_add_neon,
+        make_tuple(&vp9_fdct4x4_c,
+                   &vp9_idct4x4_16_add_c,
+                   &vp9_idct4x4_1_add_neon,
                   TX_4X4, 1)));
-#endif  // HAVE_NEON && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#endif  // HAVE_NEON

-#if HAVE_SSE2 && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSE2, PartialIDctTest,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_34_add_sse2,
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c,
+                   &vp9_idct32x32_34_add_sse2,
                   TX_32X32, 34),
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_1_add_sse2,
+        make_tuple(&vp9_fdct32x32_c,
+                   &vp9_idct32x32_1024_add_c,
+                   &vp9_idct32x32_1_add_sse2,
                   TX_32X32, 1),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_10_add_sse2,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_10_add_sse2,
                   TX_16X16, 10),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_1_add_sse2,
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_1_add_sse2,
                   TX_16X16, 1),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_12_add_sse2,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_12_add_sse2,
                   TX_8X8, 12),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_1_add_sse2,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_1_add_sse2,
                   TX_8X8, 1),
-        make_tuple(&vpx_fdct4x4_c,
-                   &vpx_idct4x4_16_add_c,
-                   &vpx_idct4x4_1_add_sse2,
+        make_tuple(&vp9_fdct4x4_c,
+                   &vp9_idct4x4_16_add_c,
+                   &vp9_idct4x4_1_add_sse2,
                   TX_4X4, 1)));
 #endif

-#if HAVE_SSSE3 && ARCH_X86_64 && \
-    !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_VP9_HIGHBITDEPTH && \
+    !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
    SSSE3_64, PartialIDctTest,
    ::testing::Values(
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_12_add_ssse3,
+        make_tuple(&vp9_fdct8x8_c,
+                   &vp9_idct8x8_64_add_c,
+                   &vp9_idct8x8_12_add_ssse3,
                   TX_8X8, 12)));
 #endif

-#if HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
+#if HAVE_SSSE3 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 INSTANTIATE_TEST_CASE_P(
-    MSA, PartialIDctTest,
+    SSSE3, PartialIDctTest,
    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_34_add_msa,
-                   TX_32X32, 34),
-        make_tuple(&vpx_fdct32x32_c,
-                   &vpx_idct32x32_1024_add_c,
-                   &vpx_idct32x32_1_add_msa,
-                   TX_32X32, 1),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_10_add_msa,
-                   TX_16X16, 10),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vpx_idct16x16_256_add_c,
-                   &vpx_idct16x16_1_add_msa,
-                   TX_16X16, 1),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_12_add_msa,
-                   TX_8X8, 10),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vpx_idct8x8_64_add_c,
-                   &vpx_idct8x8_1_add_msa,
-                   TX_8X8, 1),
-        make_tuple(&vpx_fdct4x4_c,
-                   &vpx_idct4x4_16_add_c,
-                   &vpx_idct4x4_1_add_msa,
-                   TX_4X4, 1)));
-#endif  // HAVE_MSA && !CONFIG_VPX_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-
+        make_tuple(&vp9_fdct16x16_c,
+                   &vp9_idct16x16_256_add_c,
+                   &vp9_idct16x16_10_add_ssse3,
+                   TX_16X16, 10)));
+#endif
 }  // namespace
--- a/test/postproc.sh
+++ b/test/postproc.sh
@@ -0,0 +1,63 @@
+#!/bin/sh
+##
+##  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
+##
+##  Use of this source code is governed by a BSD-style license
+##  that can be found in the LICENSE file in the root of the source
+##  tree. An additional intellectual property rights grant can be found
+##  in the file PATENTS.  All contributing project authors may
+##  be found in the AUTHORS file in the root of the source tree.
+##
+##  This file tests the libvpx postproc example code. To add new tests to this
+##  file, do the following:
+##    1. Write a shell function (this is your test).
+##    2. Add the function to postproc_tests (on a new line).
+##
+. $(dirname $0)/tools_common.sh
+
+# Environment check: Make sure input is available:
+#   $VP8_IVF_FILE and $VP9_IVF_FILE are required.
+postproc_verify_environment() {
+  if [ ! -e "${VP8_IVF_FILE}" ] || [ ! -e "${VP9_IVF_FILE}" ]; then
+    echo "Libvpx test data must exist in LIBVPX_TEST_DATA_PATH."
+    return 1
+  fi
+}
+
+# Runs postproc using $1 as input file. $2 is the codec name, and is used
+# solely to name the output file.
+postproc() {
+  local decoder="${LIBVPX_BIN_PATH}/postproc${VPX_TEST_EXE_SUFFIX}"
+  local input_file="$1"
+  local codec="$2"
+  local output_file="${VPX_TEST_OUTPUT_DIR}/postproc_${codec}.raw"
+
+  if [ ! -x "${decoder}" ]; then
+    elog "${decoder} does not exist or is not executable."
+    return 1
+  fi
+
+  eval "${VPX_TEST_PREFIX}" "${decoder}" "${input_file}" "${output_file}" \
+      ${devnull}
+
+  [ -e "${output_file}" ] || return 1
+}
+
+postproc_vp8() {
+  if [ "$(vp8_decode_available)" = "yes" ]; then
+    postproc "${VP8_IVF_FILE}" vp8 || return 1
+  fi
+}
+
+postproc_vp9() {
+  if [ "$(vpx_config_option_enabled CONFIG_VP9_POSTPROC)" = "yes" ]; then
+    if [ "$(vp9_decode_available)" = "yes" ]; then
+      postproc "${VP9_IVF_FILE}" vp9 || return 1
+    fi
+  fi
+}
+
+postproc_tests="postproc_vp8
+                postproc_vp9"
+
+run_tests postproc_verify_environment "${postproc_tests}"
--- a/test/pp_filter_test.cc
+++ b/test/pp_filter_test.cc
@@ -0,0 +1,113 @@
+/*
+ *  Copyright (c) 2012 The WebM project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+#include "test/clear_system_state.h"
+#include "test/register_state_check.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "./vpx_config.h"
+#include "./vp8_rtcd.h"
+#include "vpx/vpx_integer.h"
+#include "vpx_mem/vpx_mem.h"
+
+typedef void (*PostProcFunc)(unsigned char *src_ptr,
+                             unsigned char *dst_ptr,
+                             int src_pixels_per_line,
+                             int dst_pixels_per_line,
+                             int cols,
+                             unsigned char *flimit,
+                             int size);
+
+namespace {
+
+class VP8PostProcessingFilterTest
+    : public ::testing::TestWithParam<PostProcFunc> {
+ public:
+  virtual void TearDown() {
+    libvpx_test::ClearSystemState();
+  }
+};
+
+// Test routine for the VP8 post-processing function
+// vp8_post_proc_down_and_across_mb_row_c.
+
+TEST_P(VP8PostProcessingFilterTest, FilterOutputCheck) {
+  // Size of the underlying data block that will be filtered.
+  const int block_width  = 16;
+  const int block_height = 16;
+
+  // 5-tap filter needs 2 padding rows above and below the block in the input.
+  const int input_width = block_width;
+  const int input_height = block_height + 4;
+  const int input_stride = input_width;
+  const int input_size = input_width * input_height;
+
+  // Filter extends output block by 8 samples at left and right edges.
+  const int output_width = block_width + 16;
+  const int output_height = block_height;
+  const int output_stride = output_width;
+  const int output_size = output_width * output_height;
+
+  uint8_t *const src_image =
+      reinterpret_cast<uint8_t*>(vpx_calloc(input_size, 1));
+  uint8_t *const dst_image =
+      reinterpret_cast<uint8_t*>(vpx_calloc(output_size, 1));
+
+  // Pointers to top-left pixel of block in the input and output images.
+  uint8_t *const src_image_ptr = src_image + (input_stride << 1);
+  uint8_t *const dst_image_ptr = dst_image + 8;
+  uint8_t *const flimits =
+      reinterpret_cast<uint8_t *>(vpx_memalign(16, block_width));
+  (void)vpx_memset(flimits, 255, block_width);
+
+  // Initialize pixels in the input:
+  //   block pixels to value 1,
+  //   border pixels to value 10.
+  (void)vpx_memset(src_image, 10, input_size);
+  uint8_t *pixel_ptr = src_image_ptr;
+  for (int i = 0; i < block_height; ++i) {
+    for (int j = 0; j < block_width; ++j) {
+      pixel_ptr[j] = 1;
+    }
+    pixel_ptr += input_stride;
+  }
+
+  // Initialize pixels in the output to 99.
+  (void)vpx_memset(dst_image, 99, output_size);
+
+  ASM_REGISTER_STATE_CHECK(
+      GetParam()(src_image_ptr, dst_image_ptr, input_stride,
+                 output_stride, block_width, flimits, 16));
+
+  static const uint8_t expected_data[block_height] = {
+    4, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 4
+  };
+
+  pixel_ptr = dst_image_ptr;
+  for (int i = 0; i < block_height; ++i) {
+    for (int j = 0; j < block_width; ++j) {
+      EXPECT_EQ(expected_data[i], pixel_ptr[j])
+          << "VP8PostProcessingFilterTest failed with invalid filter output";
+    }
+    pixel_ptr += output_stride;
+  }
+
+  vpx_free(src_image);
+  vpx_free(dst_image);
+  vpx_free(flimits);
+};
+
+INSTANTIATE_TEST_CASE_P(C, VP8PostProcessingFilterTest,
+    ::testing::Values(vp8_post_proc_down_and_across_mb_row_c));
+
+#if HAVE_SSE2
+INSTANTIATE_TEST_CASE_P(SSE2, VP8PostProcessingFilterTest,
+    ::testing::Values(vp8_post_proc_down_and_across_mb_row_sse2));
+#endif
+
+}  // namespace
--- a/test/quantize_test.cc
+++ b/test/quantize_test.cc
@@ -11,12 +11,13 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
+
+#include "./vpx_config.h"
+#include "./vp8_rtcd.h"
 #include "vp8/common/blockd.h"
 #include "vp8/common/onyx.h"
 #include "vp8/encoder/block.h"
@@ -55,7 +56,7 @@ class QuantizeTestBase {

    // The full configuration is necessary to generate the quantization tables.
    VP8_CONFIG vp8_config;
-    memset(&vp8_config, 0, sizeof(vp8_config));
+    vpx_memset(&vp8_config, 0, sizeof(vp8_config));

    vp8_comp_ = vp8_create_compressor(&vp8_config);

@@ -68,7 +69,8 @@ class QuantizeTestBase {
    // Copy macroblockd from the reference to get pre-set-up dequant values.
    macroblockd_dst_ = reinterpret_cast<MACROBLOCKD *>(
        vpx_memalign(32, sizeof(*macroblockd_dst_)));
-    memcpy(macroblockd_dst_, &vp8_comp_->mb.e_mbd, sizeof(*macroblockd_dst_));
+    vpx_memcpy(macroblockd_dst_, &vp8_comp_->mb.e_mbd,
+               sizeof(*macroblockd_dst_));
    // Fix block pointers - currently they point to the blocks in the reference
    // structure.
    vp8_setup_block_dptrs(macroblockd_dst_);
@@ -77,7 +79,8 @@ class QuantizeTestBase {
  void UpdateQuantizer(int q) {
    vp8_set_quantizer(vp8_comp_, q);

-    memcpy(macroblockd_dst_, &vp8_comp_->mb.e_mbd, sizeof(*macroblockd_dst_));
+    vpx_memcpy(macroblockd_dst_, &vp8_comp_->mb.e_mbd,
+               sizeof(*macroblockd_dst_));
    vp8_setup_block_dptrs(macroblockd_dst_);
  }

@@ -191,12 +194,4 @@ INSTANTIATE_TEST_CASE_P(NEON, QuantizeTest,
                        ::testing::Values(make_tuple(&vp8_fast_quantize_b_neon,
                                                     &vp8_fast_quantize_b_c)));
 #endif  // HAVE_NEON
-
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(
-    MSA, QuantizeTest,
-    ::testing::Values(
-        make_tuple(&vp8_fast_quantize_b_msa, &vp8_fast_quantize_b_c),
-        make_tuple(&vp8_regular_quantize_b_msa, &vp8_regular_quantize_b_c)));
-#endif  // HAVE_MSA
 }  // namespace
--- a/test/realtime_test.cc
+++ b/test/realtime_test.cc
@@ -1,64 +0,0 @@
-/*
- *  Copyright (c) 2016 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-#include "test/codec_factory.h"
-#include "test/encode_test_driver.h"
-#include "test/util.h"
-#include "test/video_source.h"
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-namespace {
-
-const int kVideoSourceWidth = 320;
-const int kVideoSourceHeight = 240;
-const int kFramesToEncode = 2;
-
-class RealtimeTest
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
- protected:
-  RealtimeTest()
-      : EncoderTest(GET_PARAM(0)), frame_packets_(0) {}
-  virtual ~RealtimeTest() {}
-
-  virtual void SetUp() {
-    InitializeConfig();
-    cfg_.g_lag_in_frames = 0;
-    SetMode(::libvpx_test::kRealTime);
-  }
-
-  virtual void BeginPassHook(unsigned int /*pass*/) {
-    // TODO(tomfinegan): We're changing the pass value here to make sure
-    // we get frames when real time mode is combined with |g_pass| set to
-    // VPX_RC_FIRST_PASS. This is necessary because EncoderTest::RunLoop() sets
-    // the pass value based on the mode passed into EncoderTest::SetMode(),
-    // which overrides the one specified in SetUp() above.
-    cfg_.g_pass = VPX_RC_FIRST_PASS;
-  }
-  virtual void FramePktHook(const vpx_codec_cx_pkt_t * /*pkt*/) {
-    frame_packets_++;
-  }
-
-  int frame_packets_;
-};
-
-TEST_P(RealtimeTest, RealtimeFirstPassProducesFrames) {
-  ::libvpx_test::RandomVideoSource video;
-  video.SetSize(kVideoSourceWidth, kVideoSourceHeight);
-  video.set_limit(kFramesToEncode);
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  EXPECT_EQ(kFramesToEncode, frame_packets_);
-}
-
-VP8_INSTANTIATE_TEST_CASE(RealtimeTest,
-                          ::testing::Values(::libvpx_test::kRealTime));
-VP9_INSTANTIATE_TEST_CASE(RealtimeTest,
-                          ::testing::Values(::libvpx_test::kRealTime));
-
-}  // namespace
--- a/test/register_state_check.h
+++ b/test/register_state_check.h
@@ -30,16 +30,20 @@

 #if defined(_WIN64)

-#undef NOMINMAX
-#define NOMINMAX
-#define WIN32_LEAN_AND_MEAN
+#define _WIN32_LEAN_AND_MEAN
 #include <windows.h>
 #include <winnt.h>

+namespace testing {
+namespace internal {
+
 inline bool operator==(const M128A& lhs, const M128A& rhs) {
  return (lhs.Low == rhs.Low && lhs.High == rhs.High);
 }

+}  // namespace internal
+}  // namespace testing
+
 namespace libvpx_test {

 // Compares the state of xmm[6-15] at construction with their state at
@@ -87,12 +91,12 @@ class RegisterStateCheck {

 }  // namespace libvpx_test

-#elif defined(CONFIG_SHARED) && defined(HAVE_NEON_ASM) && defined(CONFIG_VP10) \
-      && !CONFIG_SHARED && HAVE_NEON_ASM && CONFIG_VP10
+#elif defined(CONFIG_SHARED) && defined(HAVE_NEON_ASM) && defined(CONFIG_VP9) \
+      && !CONFIG_SHARED && HAVE_NEON_ASM && CONFIG_VP9

 extern "C" {
 // Save the d8-d15 registers into store.
-void vpx_push_neon(int64_t *store);
+void vp9_push_neon(int64_t *store);
 }

 namespace libvpx_test {
@@ -107,7 +111,7 @@ class RegisterStateCheck {

 private:
  static bool StoreRegisters(int64_t store[8]) {
-    vpx_push_neon(store);
+    vp9_push_neon(store);
    return true;
  }

@@ -115,7 +119,7 @@ class RegisterStateCheck {
  bool Check() const {
    if (!initialized_) return false;
    int64_t post_store[8];
-    vpx_push_neon(post_store);
+    vp9_push_neon(post_store);
    for (int i = 0; i < 8; ++i) {
      EXPECT_EQ(pre_store_[i], post_store[i]) << "d"
          << i + 8 << " has been modified";
--- a/test/resize_test.cc
+++ b/test/resize_test.cc
@@ -7,8 +7,6 @@
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */
-#include <stdio.h>
-
 #include <climits>
 #include <vector>
 #include "third_party/googletest/src/include/gtest/gtest.h"
@@ -83,187 +81,34 @@ static void write_ivf_frame_header(const vpx_codec_cx_pkt_t *const pkt,
 const unsigned int kInitialWidth = 320;
 const unsigned int kInitialHeight = 240;

-struct FrameInfo {
-  FrameInfo(vpx_codec_pts_t _pts, unsigned int _w, unsigned int _h)
-      : pts(_pts), w(_w), h(_h) {}
-
-  vpx_codec_pts_t pts;
-  unsigned int w;
-  unsigned int h;
-};
-
-void ScaleForFrameNumber(unsigned int frame,
-                         unsigned int initial_w,
-                         unsigned int initial_h,
-                         unsigned int *w,
-                         unsigned int *h,
-                         int flag_codec) {
-  if (frame < 10) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 20) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 30) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 40) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 50) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 60) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 70) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 80) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 90) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 100) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 110) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 120) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 130) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 140) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 150) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 160) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 170) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 180) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 190) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 200) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 210) {
-    *w = initial_w / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 220) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 230) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  if (frame < 240) {
-    *w = initial_w * 3 / 4;
-    *h = initial_h * 3 / 4;
-    return;
-  }
-  if (frame < 250) {
-    *w = initial_w  / 2;
-    *h = initial_h / 2;
-    return;
-  }
-  if (frame < 260) {
-    *w = initial_w;
-    *h = initial_h;
-    return;
-  }
-  // Go down very low.
-  if (frame < 270) {
-    *w = initial_w / 4;
-    *h = initial_h / 4;
-    return;
-  }
-  if (flag_codec == 1) {
-    // Cases that only works for VP9.
-    // For VP9: Swap width and height of original.
-    if (frame < 320) {
-      *w = initial_h;
-      *h = initial_w;
-      return;
-    }
-  }
-  *w = initial_w;
-  *h = initial_h;
+unsigned int ScaleForFrameNumber(unsigned int frame, unsigned int val) {
+  if (frame < 10)
+    return val;
+  if (frame < 20)
+    return val / 2;
+  if (frame < 30)
+    return val * 2 / 3;
+  if (frame < 40)
+    return val / 4;
+  if (frame < 50)
+    return val * 7 / 8;
+  return val;
 }

 class ResizingVideoSource : public ::libvpx_test::DummyVideoSource {
 public:
  ResizingVideoSource() {
    SetSize(kInitialWidth, kInitialHeight);
-    limit_ = 350;
+    limit_ = 60;
  }
-  int flag_codec_;
+
  virtual ~ResizingVideoSource() {}

 protected:
  virtual void Next() {
    ++frame_;
-    unsigned int width;
-    unsigned int height;
-    ScaleForFrameNumber(frame_, kInitialWidth, kInitialHeight, &width, &height,
-                        flag_codec_);
-    SetSize(width, height);
+    SetSize(ScaleForFrameNumber(frame_, kInitialWidth),
+            ScaleForFrameNumber(frame_, kInitialHeight));
    FillFrame();
  }
 };
@@ -275,6 +120,15 @@ class ResizeTest : public ::libvpx_test::EncoderTest,

  virtual ~ResizeTest() {}

+  struct FrameInfo {
+    FrameInfo(vpx_codec_pts_t _pts, unsigned int _w, unsigned int _h)
+        : pts(_pts), w(_w), h(_h) {}
+
+    vpx_codec_pts_t pts;
+    unsigned int w;
+    unsigned int h;
+  };
+
  virtual void SetUp() {
    InitializeConfig();
    SetMode(GET_PARAM(1));
@@ -290,17 +144,15 @@ class ResizeTest : public ::libvpx_test::EncoderTest,

 TEST_P(ResizeTest, TestExternalResizeWorks) {
  ResizingVideoSource video;
-  video.flag_codec_ = 0;
  cfg_.g_lag_in_frames = 0;
  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));

  for (std::vector<FrameInfo>::const_iterator info = frame_info_list_.begin();
       info != frame_info_list_.end(); ++info) {
    const unsigned int frame = static_cast<unsigned>(info->pts);
-    unsigned int expected_w;
-    unsigned int expected_h;
-    ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight,
-                        &expected_w, &expected_h, 0);
+    const unsigned int expected_w = ScaleForFrameNumber(frame, kInitialWidth);
+    const unsigned int expected_h = ScaleForFrameNumber(frame, kInitialHeight);
+
    EXPECT_EQ(expected_w, info->w)
        << "Frame " << frame << " had unexpected width";
    EXPECT_EQ(expected_h, info->h)
@@ -344,32 +196,18 @@ class ResizeInternalTest : public ResizeTest {

  virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
                                  libvpx_test::Encoder *encoder) {
-    if (change_config_) {
-      int new_q = 60;
-      if (video->frame() == 0) {
-        struct vpx_scaling_mode mode = {VP8E_ONETWO, VP8E_ONETWO};
-        encoder->Control(VP8E_SET_SCALEMODE, &mode);
-      }
-      if (video->frame() == 1) {
-        struct vpx_scaling_mode mode = {VP8E_NORMAL, VP8E_NORMAL};
-        encoder->Control(VP8E_SET_SCALEMODE, &mode);
-        cfg_.rc_min_quantizer = cfg_.rc_max_quantizer = new_q;
-        encoder->Config(&cfg_);
-      }
-    } else {
-      if (video->frame() == kStepDownFrame) {
-        struct vpx_scaling_mode mode = {VP8E_FOURFIVE, VP8E_THREEFIVE};
-        encoder->Control(VP8E_SET_SCALEMODE, &mode);
-      }
-      if (video->frame() == kStepUpFrame) {
-        struct vpx_scaling_mode mode = {VP8E_NORMAL, VP8E_NORMAL};
-        encoder->Control(VP8E_SET_SCALEMODE, &mode);
-      }
+    if (video->frame() == kStepDownFrame) {
+      struct vpx_scaling_mode mode = {VP8E_FOURFIVE, VP8E_THREEFIVE};
+      encoder->Control(VP8E_SET_SCALEMODE, &mode);
+    }
+    if (video->frame() == kStepUpFrame) {
+      struct vpx_scaling_mode mode = {VP8E_NORMAL, VP8E_NORMAL};
+      encoder->Control(VP8E_SET_SCALEMODE, &mode);
    }
  }

  virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
-    if (frame0_psnr_ == 0.)
+    if (!frame0_psnr_)
      frame0_psnr_ = pkt->data.psnr.psnr[0];
    EXPECT_NEAR(pkt->data.psnr.psnr[0], frame0_psnr_, 2.0);
  }
@@ -389,7 +227,6 @@ class ResizeInternalTest : public ResizeTest {
 #endif

  double frame0_psnr_;
-  bool change_config_;
 #if WRITE_COMPRESSED_STREAM
  FILE *outfile_;
  unsigned int out_frames_;
@@ -400,7 +237,6 @@ TEST_P(ResizeInternalTest, TestInternalResizeWorks) {
  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
                                       30, 1, 0, 10);
  init_flags_ = VPX_CODEC_USE_PSNR;
-  change_config_ = false;

  // q picked such that initial keyframe on this clip is ~30dB PSNR
  cfg_.rc_min_quantizer = cfg_.rc_max_quantizer = 48;
@@ -425,198 +261,6 @@ TEST_P(ResizeInternalTest, TestInternalResizeWorks) {
  }
 }

-TEST_P(ResizeInternalTest, TestInternalResizeChangeConfig) {
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 10);
-  cfg_.g_w = 352;
-  cfg_.g_h = 288;
-  change_config_ = true;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-}
-
-class ResizeRealtimeTest : public ::libvpx_test::EncoderTest,
-  public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- protected:
-  ResizeRealtimeTest() : EncoderTest(GET_PARAM(0)) {}
-  virtual ~ResizeRealtimeTest() {}
-
-  virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
-                                  libvpx_test::Encoder *encoder) {
-    if (video->frame() == 0) {
-      encoder->Control(VP9E_SET_AQ_MODE, 3);
-      encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
-    }
-
-    if (change_bitrate_ && video->frame() == 120) {
-      change_bitrate_ = false;
-      cfg_.rc_target_bitrate = 500;
-      encoder->Config(&cfg_);
-    }
-  }
-
-  virtual void SetUp() {
-    InitializeConfig();
-    SetMode(GET_PARAM(1));
-    set_cpu_used_ = GET_PARAM(2);
-  }
-
-  virtual void DecompressedFrameHook(const vpx_image_t &img,
-                                     vpx_codec_pts_t pts) {
-    frame_info_list_.push_back(FrameInfo(pts, img.d_w, img.d_h));
-  }
-
-  virtual void MismatchHook(const vpx_image_t *img1,
-                             const vpx_image_t *img2) {
-    double mismatch_psnr = compute_psnr(img1, img2);
-    mismatch_psnr_ += mismatch_psnr;
-    ++mismatch_nframes_;
-  }
-
-  unsigned int GetMismatchFrames() {
-      return mismatch_nframes_;
-  }
-
-  void DefaultConfig() {
-    cfg_.rc_buf_initial_sz = 500;
-    cfg_.rc_buf_optimal_sz = 600;
-    cfg_.rc_buf_sz = 1000;
-    cfg_.rc_min_quantizer = 2;
-    cfg_.rc_max_quantizer = 56;
-    cfg_.rc_undershoot_pct = 50;
-    cfg_.rc_overshoot_pct = 50;
-    cfg_.rc_end_usage = VPX_CBR;
-    cfg_.kf_mode = VPX_KF_AUTO;
-    cfg_.g_lag_in_frames = 0;
-    cfg_.kf_min_dist = cfg_.kf_max_dist = 3000;
-    // Enable dropped frames.
-    cfg_.rc_dropframe_thresh = 1;
-    // Enable error_resilience mode.
-    cfg_.g_error_resilient  = 1;
-    // Enable dynamic resizing.
-    cfg_.rc_resize_allowed = 1;
-    // Run at low bitrate.
-    cfg_.rc_target_bitrate = 200;
-  }
-
-  std::vector< FrameInfo > frame_info_list_;
-  int set_cpu_used_;
-  bool change_bitrate_;
-  double mismatch_psnr_;
-  int mismatch_nframes_;
-};
-
-TEST_P(ResizeRealtimeTest, TestExternalResizeWorks) {
-  ResizingVideoSource video;
-  video.flag_codec_ = 1;
-  DefaultConfig();
-  // Disable internal resize for this test.
-  cfg_.rc_resize_allowed = 0;
-  change_bitrate_ = false;
-  mismatch_psnr_ = 0.0;
-  mismatch_nframes_ = 0;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-
-  for (std::vector<FrameInfo>::const_iterator info = frame_info_list_.begin();
-       info != frame_info_list_.end(); ++info) {
-    const unsigned int frame = static_cast<unsigned>(info->pts);
-    unsigned int expected_w;
-    unsigned int expected_h;
-    ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight,
-                        &expected_w, &expected_h, 1);
-    EXPECT_EQ(expected_w, info->w)
-        << "Frame " << frame << " had unexpected width";
-    EXPECT_EQ(expected_h, info->h)
-        << "Frame " << frame << " had unexpected height";
-    EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-  }
-}
-
-// Verify the dynamic resizer behavior for real time, 1 pass CBR mode.
-// Run at low bitrate, with resize_allowed = 1, and verify that we get
-// one resize down event.
-TEST_P(ResizeRealtimeTest, TestInternalResizeDown) {
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 299);
-  DefaultConfig();
-  cfg_.g_w = 352;
-  cfg_.g_h = 288;
-  change_bitrate_ = false;
-  mismatch_psnr_ = 0.0;
-  mismatch_nframes_ = 0;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-
-  unsigned int last_w = cfg_.g_w;
-  unsigned int last_h = cfg_.g_h;
-  int resize_count = 0;
-  for (std::vector<FrameInfo>::const_iterator info = frame_info_list_.begin();
-       info != frame_info_list_.end(); ++info) {
-    if (info->w != last_w || info->h != last_h) {
-      // Verify that resize down occurs.
-      ASSERT_LT(info->w, last_w);
-      ASSERT_LT(info->h, last_h);
-      last_w = info->w;
-      last_h = info->h;
-      resize_count++;
-    }
-  }
-
-#if CONFIG_VP9_DECODER
-  // Verify that we get 1 resize down event in this test.
-  ASSERT_EQ(1, resize_count) << "Resizing should occur.";
-  EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-#else
-  printf("Warning: VP9 decoder unavailable, unable to check resize count!\n");
-#endif
-}
-
-// Verify the dynamic resizer behavior for real time, 1 pass CBR mode.
-// Start at low target bitrate, raise the bitrate in the middle of the clip,
-// scaling-up should occur after bitrate changed.
-TEST_P(ResizeRealtimeTest, TestInternalResizeDownUpChangeBitRate) {
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 359);
-  DefaultConfig();
-  cfg_.g_w = 352;
-  cfg_.g_h = 288;
-  change_bitrate_ = true;
-  mismatch_psnr_ = 0.0;
-  mismatch_nframes_ = 0;
-  // Disable dropped frames.
-  cfg_.rc_dropframe_thresh = 0;
-  // Starting bitrate low.
-  cfg_.rc_target_bitrate = 80;
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-
-  unsigned int last_w = cfg_.g_w;
-  unsigned int last_h = cfg_.g_h;
-  int resize_count = 0;
-  for (std::vector<FrameInfo>::const_iterator info = frame_info_list_.begin();
-       info != frame_info_list_.end(); ++info) {
-    if (info->w != last_w || info->h != last_h) {
-      resize_count++;
-      if (resize_count == 1) {
-        // Verify that resize down occurs.
-        ASSERT_LT(info->w, last_w);
-        ASSERT_LT(info->h, last_h);
-      } else if (resize_count == 2) {
-        // Verify that resize up occurs.
-        ASSERT_GT(info->w, last_w);
-        ASSERT_GT(info->h, last_h);
-      }
-      last_w = info->w;
-      last_h = info->h;
-    }
-  }
-
-#if CONFIG_VP9_DECODER
-  // Verify that we get 2 resize events in this test.
-  ASSERT_EQ(resize_count, 2) << "Resizing should occur twice.";
-  EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
-#else
-  printf("Warning: VP9 decoder unavailable, unable to check resize count!\n");
-#endif
-}
-
 vpx_img_fmt_t CspForFrameNumber(int frame) {
  if (frame < 10)
    return VPX_IMG_FMT_I420;
@@ -671,7 +315,7 @@ class ResizeCspTest : public ResizeTest {
  }

  virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
-    if (frame0_psnr_ == 0.)
+    if (!frame0_psnr_)
      frame0_psnr_ = pkt->data.psnr.psnr[0];
    EXPECT_NEAR(pkt->data.psnr.psnr[0], frame0_psnr_, 2.0);
  }
@@ -722,13 +366,11 @@ TEST_P(ResizeCspTest, TestResizeCspWorks) {
  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
 }

-VP10_INSTANTIATE_TEST_CASE(ResizeTest,
+VP8_INSTANTIATE_TEST_CASE(ResizeTest, ONE_PASS_TEST_MODES);
+VP9_INSTANTIATE_TEST_CASE(ResizeTest,
                          ::testing::Values(::libvpx_test::kRealTime));
-VP10_INSTANTIATE_TEST_CASE(ResizeInternalTest,
+VP9_INSTANTIATE_TEST_CASE(ResizeInternalTest,
                          ::testing::Values(::libvpx_test::kOnePassBest));
-VP10_INSTANTIATE_TEST_CASE(ResizeRealtimeTest,
-                          ::testing::Values(::libvpx_test::kRealTime),
-                          ::testing::Range(5, 9));
-VP10_INSTANTIATE_TEST_CASE(ResizeCspTest,
+VP9_INSTANTIATE_TEST_CASE(ResizeCspTest,
                          ::testing::Values(::libvpx_test::kRealTime));
 }  // namespace
--- a/Show More
+++ b/Show More