Remove a line.

Change-Id: I2965d25efafeee487b4de75fddfa22e4907525ed
2015-05-01 11:09:30 -07:00
671 changed files with 56235 additions and 80471 deletions
--- 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
@@ -1,21 +1,14 @@
 Adrian Grange <agrange@google.com>
-Adrian Grange <agrange@google.com> <agrange@agrange-macbookpro.roam.corp.google.com>
-Aℓex Converse <aconverse@google.com>
-Aℓex Converse <aconverse@google.com> <alex.converse@gmail.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>
-Hangyu Kuang <hkuang@google.com> <hkuang@hkuang-macbookpro.roam.corp.google.com>
-Hui Su <huisu@google.com>
-Jacky Chen <jackychen@google.com>
 Jim Bankoski <jimbankoski@google.com>
 Johann Koenig <johannkoenig@google.com>
 Johann Koenig <johannkoenig@google.com> <johann.koenig@duck.com>
-Johann Koenig <johannkoenig@google.com> <johannkoenig@dhcp-172-19-7-52.mtv.corp.google.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>
@@ -24,13 +17,10 @@ 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>
 Yaowu Xu <yaowu@google.com> <yaowu@xuyaowu.com>
-Yaowu Xu <yaowu@google.com> <yaowu@YAOWU2-W.ad.corp.google.com>
--- a/15
+++ b/15
@@ -5,9 +5,9 @@ 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 <aconverse@google.com>
 Alexis Ballier <aballier@gentoo.org>
 Alok Ahuja <waveletcoeff@gmail.com>
 Alpha Lam <hclam@google.com>
@@ -16,10 +16,8 @@ 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>
@@ -29,7 +27,6 @@ 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>
 Fabio Pedretti <fabio.ped@libero.it>
@@ -37,8 +34,6 @@ 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>
@@ -49,7 +44,7 @@ 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>
+JackyChen <jackychen@google.com>
 James Berry <jamesberry@google.com>
 James Yu <james.yu@linaro.org>
 James Zern <jzern@google.com>
@@ -65,11 +60,9 @@ 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>
@@ -89,7 +82,6 @@ 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>
@@ -104,7 +96,7 @@ 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>
+Ronald S. Bultje <rbultje@google.com>
 Rui Ueyama <ruiu@google.com>
 Sami Pietilä <samipietila@google.com>
 Scott Graham <scottmg@chromium.org>
@@ -112,7 +104,6 @@ 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>
--- a/17
+++ b/17
@@ -1,20 +1,3 @@
-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.

--- a/11
+++ b/11
@@ -59,7 +59,6 @@ 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
@@ -86,8 +85,6 @@ COMPILING THE APPLICATIONS/LIBRARIES:
    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
@@ -104,7 +101,13 @@ COMPILING THE APPLICATIONS/LIBRARIES:
    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
+    universal-darwin14-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 $(CONFIG_DIR)libs-mips-android-gcc.mk
 else
@@ -165,27 +163,17 @@ ifeq ($(CONFIG_RUNTIME_CPU_DETECT),yes)
 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 := $$(rtcd_dep_template_SRCS:.neon=)
 ifeq ($(CONFIG_VP8), yes)
-$$(rtcd_dep_template_SRCS): vp8_rtcd.h
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vp8_rtcd.h
 endif
 ifeq ($(CONFIG_VP9), yes)
-$$(rtcd_dep_template_SRCS): vp9_rtcd.h
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vp9_rtcd.h
 endif
-ifeq ($(CONFIG_VP10), yes)
-$$(rtcd_dep_template_SRCS): vp10_rtcd.h
-endif
-$$(rtcd_dep_template_SRCS): vpx_scale_rtcd.h
-$$(rtcd_dep_template_SRCS): vpx_dsp_rtcd.h
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vpx_scale_rtcd.h

-ifneq ($(findstring $(TARGET_ARCH_ABI),x86 x86_64),)
-$$(rtcd_dep_template_SRCS): vpx_config.asm
+ifeq ($(TARGET_ARCH_ABI),x86)
+$(foreach file, $(LOCAL_SRC_FILES), $(LOCAL_PATH)/$(file)): vpx_config.asm
 endif
-endef
-
-$(eval $(call rtcd_dep_template))

 .PHONY: clean
 clean:
--- a/build/make/Makefile
+++ b/build/make/Makefile
@@ -22,10 +22,8 @@ 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


 # Note: md5sum is not installed on OS X, but openssl is. Openssl may not be
@@ -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)
@@ -115,9 +116,6 @@ 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:

 # Add compiler flags for intrinsic files
 ifeq ($(TOOLCHAIN), x86-os2-gcc)
@@ -140,8 +138,6 @@ $(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)%vp9_reconintra.c.d: CFLAGS += $(STACKREALIGN)
-$(BUILD_PFX)%vp9_reconintra.c.o: CFLAGS += $(STACKREALIGN)

 $(BUILD_PFX)%.c.d: %.c
 	$(if $(quiet),@echo "    [DEP] $@")
@@ -287,7 +283,7 @@ define archive_template
 # for creating them.
 $(1):
 	$(if $(quiet),@echo "    [AR] $$@")
-	$(qexec)$$(AR) $$(ARFLAGS) $$@ $$^
+	$(qexec)$$(AR) $$(ARFLAGS) $$@ $$?
 endef

 define so_template
@@ -317,15 +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


@@ -386,7 +385,6 @@ LIBS=$(call enabled,LIBS)
 $(foreach lib,$(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))))

 INSTALL-LIBS=$(call cond_enabled,CONFIG_INSTALL_LIBS,INSTALL-LIBS)
 ifeq ($(MAKECMDGOALS),dist)
--- 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)
@@ -201,10 +200,6 @@ disabled(){
  eval test "x\$$1" = "xno"
 }

-# 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
@@ -214,10 +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 $var; then
@@ -346,10 +337,6 @@ check_add_cflags() {
  check_cflags "$@" && add_cflags_only "$@"
 }

-check_add_cxxflags() {
-  check_cxxflags "$@" && add_cxxflags_only "$@"
-}
-
 check_add_asflags() {
  log add_asflags "$@"
  add_asflags "$@"
@@ -403,7 +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 += ${t}"
@@ -441,7 +428,7 @@ NM=${NM}

 CFLAGS  = ${CFLAGS}
 CXXFLAGS  = ${CXXFLAGS}
-ARFLAGS = -crs\$(if \$(quiet),,v)
+ARFLAGS = -rus\$(if \$(quiet),c,v)
 LDFLAGS = ${LDFLAGS}
 ASFLAGS = ${ASFLAGS}
 extralibs = ${extralibs}
@@ -516,9 +503,6 @@ process_common_cmdline() {
      --extra-cflags=*)
        extra_cflags="${optval}"
        ;;
-      --extra-cxxflags=*)
-        extra_cxxflags="${optval}"
-        ;;
      --enable-?*|--disable-?*)
        eval `echo "$opt" | sed 's/--/action=/;s/-/ option=/;s/-/_/g'`
        if echo "${ARCH_EXT_LIST}" | grep "^ *$option\$" >/dev/null; then
@@ -633,11 +617,6 @@ 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 $1 --show-sdk-version 2>/dev/null | cut -d. -f1
-}
-
 process_common_toolchain() {
  if [ -z "$toolchain" ]; then
    gcctarget="${CHOST:-$(gcc -dumpmachine 2> /dev/null)}"
@@ -668,6 +647,14 @@ 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
@@ -749,15 +736,7 @@ process_common_toolchain() {
  # 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}"
@@ -810,6 +789,7 @@ process_common_toolchain() {
  case ${toolchain} in
    sparc-solaris-*)
      add_extralibs -lposix4
+      disable_feature fast_unaligned
      ;;
    *-solaris-*)
      add_extralibs -lposix4
@@ -832,36 +812,12 @@ 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

@@ -1044,12 +1000,6 @@ EOF
          done

          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*)
@@ -1083,6 +1033,7 @@ EOF
      tune_cflags="-mtune="
      if enabled dspr2; then
        check_add_cflags -mips32r2 -mdspr2
+        disable_feature fast_unaligned
      fi

      if enabled runtime_cpu_detect; then
@@ -1109,6 +1060,8 @@ EOF
          add_cflags -mmsa
          add_asflags -mmsa
          add_ldflags -mmsa
+
+          disable_feature fast_unaligned
        fi
      fi

@@ -1128,9 +1081,7 @@ EOF
          CROSS=${CROSS:-g}
          ;;
        os2)
-          disable_feature pic
          AS=${AS:-nasm}
-          add_ldflags -Zhigh-mem
          ;;
      esac

@@ -1220,8 +1171,7 @@ EOF
              && AS=""
          fi
          [ "${AS}" = auto ] || [ -z "${AS}" ] \
-            && die "Neither yasm nor nasm have been found." \
-                   "See the prerequisites section in the README for more info."
+            && die "Neither yasm nor nasm have been found"
          ;;
      esac
      log_echo "  using $AS"
@@ -1260,13 +1210,6 @@ 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
@@ -1278,7 +1221,7 @@ EOF
          ;;
      esac
      ;;
-    *-gcc|generic-gnu)
+    universal*|*-gcc|generic-gnu)
      link_with_cc=gcc
      enable_feature gcc
      setup_gnu_toolchain
@@ -1380,6 +1323,12 @@ 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
@@ -73,10 +73,6 @@ generate_filter() {
                open_tag File RelativePath="$f"

                if [ "$pat" == "asm" ] && $asm_use_custom_step; then
-                    # Avoid object file name collisions, i.e. vpx_config.c and
-                    # vpx_config.asm produce the same object file without
-                    # this additional suffix.
-                    objf=${objf%.obj}_asm.obj
                    for plat in "${platforms[@]}"; do
                        for cfg in Debug Release; do
                            open_tag FileConfiguration \
--- 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
@@ -255,7 +255,7 @@ for opt in "$@"; do
    ;;
    --ver=*) vs_ver="$optval"
             case $optval in
-             [789]|10|11|12|14)
+             [789]|10|11|12)
             ;;
             *) die Unrecognized Visual Studio Version in $opt
             ;;
@@ -300,15 +300,12 @@ case "${vs_ver:-8}" 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 "${vs_ver:-8}" in
    [789])
    sfx=vcproj
    ;;
-    10|11|12|14)
+    10|11|12)
    sfx=vcxproj
    ;;
 esac
--- a/build/make/gen_msvs_vcxproj.sh
+++ b/build/make/gen_msvs_vcxproj.sh
@@ -34,7 +34,7 @@ Options:
    --name=project_name         Name of the project (required)
    --proj-guid=GUID            GUID to use for the project
    --module-def=filename       File containing export definitions (for DLLs)
-    --ver=version               Version (10,11,12,14) of visual studio to generate for
+    --ver=version               Version (10,11,12) of visual studio to generate for
    --src-path-bare=dir         Path to root of source tree
    -Ipath/to/include           Additional include directories
    -DFLAG[=value]              Preprocessor macros to define
@@ -168,7 +168,7 @@ for opt in "$@"; do
        --ver=*)
            vs_ver="$optval"
            case "$optval" in
-                10|11|12|14)
+                10|11|12)
                ;;
                *) die Unrecognized Visual Studio Version in $opt
                ;;
@@ -218,7 +218,7 @@ guid=${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 [ "$config" = "Release" ]; then
                tag_content WholeProgramOptimization true
--- a/build/make/iosbuild.sh
+++ b/build/make/iosbuild.sh
@@ -25,6 +25,7 @@ CONFIGURE_ARGS="--disable-docs
 DIST_DIR="_dist"
 FRAMEWORK_DIR="VPX.framework"
 HEADER_DIR="${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)
@@ -40,24 +41,15 @@ TARGETS="arm64-darwin-gcc
 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}***"
@@ -202,12 +194,11 @@ cat << EOF
  Usage: ${0##*/} [arguments]
    --help: Display this message and exit.
    --extra-configure-args <args>: Extra args to pass when configuring libvpx.
+    --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:
         ${TARGETS}
-    --test-link: Confirms all targets can be linked. Functionally identical to
-                 passing --enable-examples via --extra-configure-args.
    --verbose: Output information about the environment and each stage of the
               build.
 EOF
@@ -236,15 +227,16 @@ while [ -n "$1" ]; do
      iosbuild_usage
      exit
      ;;
+    --jobs)
+      MAKE_JOBS="$2"
+      shift
+      ;;
    --preserve-build-output)
      PRESERVE_BUILD_OUTPUT=yes
      ;;
    --show-build-output)
      devnull=
      ;;
-    --test-link)
-      EXTRA_CONFIGURE_ARGS="${EXTRA_CONFIGURE_ARGS} --enable-examples"
-      ;;
    --targets)
      TARGETS="$2"
      shift
@@ -268,11 +260,11 @@ cat << EOF
  EXTRA_CONFIGURE_ARGS=${EXTRA_CONFIGURE_ARGS}
  FRAMEWORK_DIR=${FRAMEWORK_DIR}
  HEADER_DIR=${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=${PRESERVE_BUILD_OUTPUT}
  TARGETS="${TARGETS}"
 EOF
 fi
--- a/build/make/rtcd.pl
+++ b/build/make/rtcd.pl
@@ -319,15 +319,14 @@ EOF

  print <<EOF;
 #if HAVE_DSPR2
-void vpx_dsputil_static_init();
 #if CONFIG_VP8
 void dsputil_static_init();
-#endif
-
-vpx_dsputil_static_init();
-#if CONFIG_VP8
 dsputil_static_init();
 #endif
+#if CONFIG_VP9
+void vp9_dsputil_static_init();
+vp9_dsputil_static_init();
+#endif
 #endif
 }
 #endif
--- 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/118
+++ b/118
@@ -31,13 +31,14 @@ 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_vp9_highbitdepth}      use VP9 high bit depth (10/12) profiles
  ${toggle_vp8}                   VP8 codec support
  ${toggle_vp9}                   VP9 codec support
-  ${toggle_vp10}                  VP10 codec support
  ${toggle_internal_stats}        output of encoder internal stats for debug, if supported (encoders)
  ${toggle_postproc}              postprocessing
  ${toggle_vp9_postproc}          vp9 specific postprocessing
@@ -107,7 +108,6 @@ all_platforms="${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} mips32-linux-gcc"
 all_platforms="${all_platforms} mips64-linux-gcc"
@@ -134,8 +134,6 @@ all_platforms="${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"
@@ -152,7 +150,13 @@ 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
@@ -189,14 +193,6 @@ 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 "${source_path}/vp8" ] || disable_feature vp8
-[ -d "${source_path}/vp9" ] || disable_feature vp9
-[ -d "${source_path}/vp10" ] || disable_feature vp10
-
-# disable vp10 codec by default
-disable_feature 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.
@@ -207,24 +203,37 @@ 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="
-    vp8_encoder
-    vp8_decoder
-    vp9_encoder
-    vp9_decoder
-    vp10_encoder
-    vp10_decoder
-"
-CODEC_FAMILIES="
-    vp8
-    vp9
-    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
@@ -256,6 +265,7 @@ HAVE_LIST="
    ${ARCH_EXT_LIST}
    vpx_ports
    stdint_h
+    alt_tree_layout
    pthread_h
    sys_mman_h
    unistd_h
@@ -264,7 +274,6 @@ EXPERIMENT_LIST="
    spatial_svc
    fp_mb_stats
    emulate_hardware
-    misc_fixes
 "
 CONFIG_LIST="
    dependency_tracking
@@ -285,6 +294,7 @@ CONFIG_LIST="

    codec_srcs
    debug_libs
+    fast_unaligned

    dequant_tokens
    dc_recon
@@ -346,6 +356,7 @@ CMDLINE_SELECT="
    libc
    as
    size_limit
+    fast_unaligned
    codec_srcs
    debug_libs

@@ -430,8 +441,22 @@ 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=""
@@ -507,11 +532,11 @@ 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.
-        if ! enabled linux && ! enabled os2; then
+        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 and OS/2 for now"
+                die "--enable-shared only supported on ELF for now"
            fi
        fi
    fi
@@ -576,6 +601,24 @@ EOF
 process_toolchain() {
    process_common_toolchain

+    # Handle universal binaries for this architecture
+    case $toolchain in
+        universal-darwin*)
+            darwin_ver=${tgt_os##darwin}
+
+            # 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
@@ -651,7 +694,7 @@ process_toolchain() {
                 VCPROJ_SFX=vcproj
                 gen_vcproj_cmd=${source_path}/build/make/gen_msvs_proj.sh
                 ;;
-             10|11|12|14)
+             10|11|12)
                 VCPROJ_SFX=vcxproj
                 gen_vcproj_cmd=${source_path}/build/make/gen_msvs_vcxproj.sh
                 enabled werror && gen_vcproj_cmd="${gen_vcproj_cmd} --enable-werror"
@@ -663,7 +706,7 @@ process_toolchain() {
    esac

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

    if enabled postproc_visualizer; then
        enabled postproc || die "postproc_visualizer requires postproc to be enabled"
@@ -717,16 +760,6 @@ 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 +770,6 @@ 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
@@ -22,22 +22,19 @@ 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/webmids.hpp
-
 LIBWEBM_MUXER_SRCS += third_party/libwebm/mkvmuxer.cpp \
                      third_party/libwebm/mkvmuxerutil.cpp \
                      third_party/libwebm/mkvwriter.cpp \
@@ -45,7 +42,8 @@ LIBWEBM_MUXER_SRCS += third_party/libwebm/mkvmuxer.cpp \
                      third_party/libwebm/mkvmuxertypes.hpp \
                      third_party/libwebm/mkvmuxerutil.hpp \
                      third_party/libwebm/mkvparser.hpp \
-                      third_party/libwebm/mkvwriter.hpp
+                      third_party/libwebm/mkvwriter.hpp \
+                      third_party/libwebm/webmids.hpp

 LIBWEBM_PARSER_SRCS = third_party/libwebm/mkvparser.cpp \
                      third_party/libwebm/mkvreader.cpp \
@@ -58,7 +56,6 @@ UTILS-$(CONFIG_DECODERS)    += vpxdec.c
 vpxdec.SRCS                 += md5_utils.c md5_utils.h
 vpxdec.SRCS                 += vpx_ports/mem_ops.h
 vpxdec.SRCS                 += vpx_ports/mem_ops_aligned.h
-vpxdec.SRCS                 += vpx_ports/msvc.h
 vpxdec.SRCS                 += vpx_ports/vpx_timer.h
 vpxdec.SRCS                 += vpx/vpx_integer.h
 vpxdec.SRCS                 += args.c args.h
@@ -69,7 +66,6 @@ ifeq ($(CONFIG_LIBYUV),yes)
  vpxdec.SRCS                 += $(LIBYUV_SRCS)
 endif
 ifeq ($(CONFIG_WEBM_IO),yes)
-  vpxdec.SRCS                 += $(LIBWEBM_COMMON_SRCS)
  vpxdec.SRCS                 += $(LIBWEBM_PARSER_SRCS)
  vpxdec.SRCS                 += webmdec.cc webmdec.h
 endif
@@ -84,14 +80,12 @@ vpxenc.SRCS                 += tools_common.c tools_common.h
 vpxenc.SRCS                 += warnings.c warnings.h
 vpxenc.SRCS                 += vpx_ports/mem_ops.h
 vpxenc.SRCS                 += vpx_ports/mem_ops_aligned.h
-vpxenc.SRCS                 += vpx_ports/msvc.h
 vpxenc.SRCS                 += vpx_ports/vpx_timer.h
 vpxenc.SRCS                 += vpxstats.c vpxstats.h
 ifeq ($(CONFIG_LIBYUV),yes)
  vpxenc.SRCS                 += $(LIBYUV_SRCS)
 endif
 ifeq ($(CONFIG_WEBM_IO),yes)
-  vpxenc.SRCS                 += $(LIBWEBM_COMMON_SRCS)
  vpxenc.SRCS                 += $(LIBWEBM_MUXER_SRCS)
  vpxenc.SRCS                 += webmenc.cc webmenc.h
 endif
@@ -104,7 +98,6 @@ ifeq ($(CONFIG_SPATIAL_SVC),yes)
  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        += vpx_ports/msvc.h
  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
@@ -119,7 +112,6 @@ 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-$(CONFIG_DECODERS)        += simple_decoder.c
@@ -130,7 +122,6 @@ simple_decoder.SRCS                += video_common.h
 simple_decoder.SRCS                += video_reader.h video_reader.c
 simple_decoder.SRCS                += vpx_ports/mem_ops.h
 simple_decoder.SRCS                += vpx_ports/mem_ops_aligned.h
-simple_decoder.SRCS                += vpx_ports/msvc.h
 simple_decoder.DESCRIPTION          = Simplified decoder loop
 EXAMPLES-$(CONFIG_DECODERS)        += postproc.c
 postproc.SRCS                      += ivfdec.h ivfdec.c
@@ -139,7 +130,6 @@ 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.SRCS                      += vpx_ports/msvc.h
 postproc.GUID                       = 65E33355-F35E-4088-884D-3FD4905881D7
 postproc.DESCRIPTION                = Decoder postprocessor control
 EXAMPLES-$(CONFIG_DECODERS)        += decode_to_md5.c
@@ -150,7 +140,6 @@ decode_to_md5.SRCS                 += video_common.h
 decode_to_md5.SRCS                 += video_reader.h video_reader.c
 decode_to_md5.SRCS                 += vpx_ports/mem_ops.h
 decode_to_md5.SRCS                 += vpx_ports/mem_ops_aligned.h
-decode_to_md5.SRCS                 += vpx_ports/msvc.h
 decode_to_md5.GUID                  = 59120B9B-2735-4BFE-B022-146CA340FE42
 decode_to_md5.DESCRIPTION           = Frame by frame MD5 checksum
 EXAMPLES-$(CONFIG_ENCODERS)     += simple_encoder.c
@@ -158,7 +147,6 @@ 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-$(CONFIG_VP9_ENCODER)  += vp9_lossless_encoder.c
@@ -166,7 +154,6 @@ vp9_lossless_encoder.SRCS       += ivfenc.h ivfenc.c
 vp9_lossless_encoder.SRCS       += tools_common.h tools_common.c
 vp9_lossless_encoder.SRCS       += video_common.h
 vp9_lossless_encoder.SRCS       += video_writer.h video_writer.c
-vp9_lossless_encoder.SRCS       += vpx_ports/msvc.h
 vp9_lossless_encoder.GUID        = B63C7C88-5348-46DC-A5A6-CC151EF93366
 vp9_lossless_encoder.DESCRIPTION = Simplified lossless VP9 encoder
 EXAMPLES-$(CONFIG_ENCODERS)     += twopass_encoder.c
@@ -174,7 +161,6 @@ 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-$(CONFIG_DECODERS)     += decode_with_drops.c
@@ -184,7 +170,6 @@ decode_with_drops.SRCS          += video_common.h
 decode_with_drops.SRCS          += video_reader.h video_reader.c
 decode_with_drops.SRCS          += vpx_ports/mem_ops.h
 decode_with_drops.SRCS          += vpx_ports/mem_ops_aligned.h
-decode_with_drops.SRCS          += vpx_ports/msvc.h
 decode_with_drops.GUID           = CE5C53C4-8DDA-438A-86ED-0DDD3CDB8D26
 decode_with_drops.DESCRIPTION    = Drops frames while decoding
 EXAMPLES-$(CONFIG_ENCODERS)        += set_maps.c
@@ -192,7 +177,6 @@ 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
@@ -200,7 +184,6 @@ 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.SRCS                 += vpx_ports/msvc.h
 vp8cx_set_ref.GUID                  = C5E31F7F-96F6-48BD-BD3E-10EBF6E8057A
 vp8cx_set_ref.DESCRIPTION           = VP8 set encoder reference frame

@@ -211,7 +194,6 @@ 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       += vpx_ports/msvc.h
 vp8_multi_resolution_encoder.SRCS       += $(LIBYUV_SRCS)
 vp8_multi_resolution_encoder.GUID        = 04f8738e-63c8-423b-90fa-7c2703a374de
 vp8_multi_resolution_encoder.DESCRIPTION = VP8 Multiple-resolution Encoding
@@ -272,6 +254,14 @@ CODEC_EXTRA_LIBS=$(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 $(CONFIG_MSVS),,yes)
@@ -279,28 +269,24 @@ DIST-BINS-$(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.
@@ -328,8 +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
@@ -52,7 +52,7 @@

 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/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
@@ -106,7 +106,7 @@

 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>]\nSee comments in "
--- a/examples/twopass_encoder.c
+++ b/examples/twopass_encoder.c
@@ -58,7 +58,7 @@

 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/vp8_multi_resolution_encoder.c
+++ b/examples/vp8_multi_resolution_encoder.c
@@ -37,14 +37,15 @@
 #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"
+#include "./tools_common.h"
 #define interface (vpx_codec_vp8_cx())
 #define fourcc    0x30385056

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

--- 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
@@ -20,7 +20,7 @@

 static const char *exec_name;

-void usage_exit(void) {
+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);
--- a/examples/vp9_spatial_svc_encoder.c
+++ b/examples/vp9_spatial_svc_encoder.c
@@ -14,34 +14,24 @@
 * that benefit from a scalable bitstream.
 */

-#include <math.h>
 #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_ports/vpx_timer.h"
 #include "vpx/svc_context.h"
 #include "vpx/vp8cx.h"
 #include "vpx/vpx_encoder.h"
 #include "../vpxstats.h"
-#define OUTPUT_RC_STATS 1

 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 threads_arg =
-    ARG_DEF("th", "threads", 1, "number of threads to use");
-#if OUTPUT_RC_STATS
-static const arg_def_t output_rc_stats_arg =
-    ARG_DEF("rcstat", "output_rc_stats", 1, "output rc stats");
-#endif
 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 =
@@ -52,9 +42,6 @@ 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 temporal_layering_mode_arg =
-    ARG_DEF("tlm", "temporal-layering-mode", 1, "temporal layering scheme."
-        "VP9E_TEMPORAL_LAYERING_MODE");
 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 =
@@ -78,10 +65,6 @@ static const arg_def_t lag_in_frame_arg =
        "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");
-static const arg_def_t speed_arg =
-    ARG_DEF("sp", "speed", 1, "speed configuration");
-static const arg_def_t aqmode_arg =
-    ARG_DEF("aq", "aqmode", 1, "aq-mode off/on");

 #if CONFIG_VP9_HIGHBITDEPTH
 static const struct arg_enum_list bitdepth_enum[] = {
@@ -102,16 +85,10 @@ static const arg_def_t *svc_args[] = {
  &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, &temporal_layering_mode_arg,
-  &lag_in_frame_arg,  &threads_arg,       &aqmode_arg,
-#if OUTPUT_RC_STATS
-  &output_rc_stats_arg,
-#endif
-
+  &max_bitrate_arg,   &temporal_layers_arg,                 &lag_in_frame_arg,
 #if CONFIG_VP9_HIGHBITDEPTH
  &bitdepth_arg,
 #endif
-  &speed_arg,
  &rc_end_usage_arg,  NULL
 };

@@ -125,10 +102,6 @@ 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;
-static const uint32_t default_temporal_layering_mode = 0;
-static const uint32_t default_output_rc_stats = 0;
-static const int32_t default_speed = -1;  // -1 means use library default.
-static const uint32_t default_threads = 0;  // zero means use library default.

 typedef struct {
  const char *input_filename;
@@ -143,7 +116,7 @@ typedef struct {

 static const char *exec_name;

-void usage_exit(void) {
+void usage_exit() {
  fprintf(stderr, "Usage: %s <options> input_filename output_filename\n",
          exec_name);
  fprintf(stderr, "Options:\n");
@@ -170,12 +143,6 @@ static void parse_command_line(int argc, const char **argv_,
  svc_ctx->log_level = SVC_LOG_DEBUG;
  svc_ctx->spatial_layers = default_spatial_layers;
  svc_ctx->temporal_layers = default_temporal_layers;
-  svc_ctx->temporal_layering_mode = default_temporal_layering_mode;
-#if OUTPUT_RC_STATS
-  svc_ctx->output_rc_stat = default_output_rc_stats;
-#endif
-  svc_ctx->speed = default_speed;
-  svc_ctx->threads = default_threads;

  // start with default encoder configuration
  res = vpx_codec_enc_config_default(vpx_codec_vp9_cx(), enc_cfg, 0);
@@ -217,22 +184,6 @@ static void parse_command_line(int argc, const char **argv_,
      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);
-#if OUTPUT_RC_STATS
-    } else if (arg_match(&arg, &output_rc_stats_arg, argi)) {
-      svc_ctx->output_rc_stat = arg_parse_uint(&arg);
-#endif
-    } else if (arg_match(&arg, &speed_arg, argi)) {
-      svc_ctx->speed = arg_parse_uint(&arg);
-    } else if (arg_match(&arg, &aqmode_arg, argi)) {
-      svc_ctx->aqmode = arg_parse_uint(&arg);
-    } else if (arg_match(&arg, &threads_arg, argi)) {
-      svc_ctx->threads = arg_parse_uint(&arg);
-    } else if (arg_match(&arg, &temporal_layering_mode_arg, argi)) {
-      svc_ctx->temporal_layering_mode =
-          enc_cfg->temporal_layering_mode = arg_parse_int(&arg);
-      if (svc_ctx->temporal_layering_mode) {
-        enc_cfg->g_error_resilient = 1;
-      }
    } 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;
@@ -365,238 +316,6 @@ static void parse_command_line(int argc, const char **argv_,
      enc_cfg->rc_target_bitrate, enc_cfg->kf_max_dist);
 }

-#if OUTPUT_RC_STATS
-// For rate control encoding stats.
-struct RateControlStats {
-  // Number of input frames per layer.
-  int layer_input_frames[VPX_MAX_LAYERS];
-  // Total (cumulative) number of encoded frames per layer.
-  int layer_tot_enc_frames[VPX_MAX_LAYERS];
-  // Number of encoded non-key frames per layer.
-  int layer_enc_frames[VPX_MAX_LAYERS];
-  // Framerate per layer (cumulative).
-  double layer_framerate[VPX_MAX_LAYERS];
-  // Target average frame size per layer (per-frame-bandwidth per layer).
-  double layer_pfb[VPX_MAX_LAYERS];
-  // Actual average frame size per layer.
-  double layer_avg_frame_size[VPX_MAX_LAYERS];
-  // Average rate mismatch per layer (|target - actual| / target).
-  double layer_avg_rate_mismatch[VPX_MAX_LAYERS];
-  // Actual encoding bitrate per layer (cumulative).
-  double layer_encoding_bitrate[VPX_MAX_LAYERS];
-  // Average of the short-time encoder actual bitrate.
-  // TODO(marpan): Should we add these short-time stats for each layer?
-  double avg_st_encoding_bitrate;
-  // Variance of the short-time encoder actual bitrate.
-  double variance_st_encoding_bitrate;
-  // Window (number of frames) for computing short-time encoding bitrate.
-  int window_size;
-  // Number of window measurements.
-  int window_count;
-};
-
-// Note: these rate control stats assume only 1 key frame in the
-// sequence (i.e., first frame only).
-static void set_rate_control_stats(struct RateControlStats *rc,
-                                     vpx_codec_enc_cfg_t *cfg) {
-  unsigned int sl, tl;
-  // Set the layer (cumulative) framerate and the target layer (non-cumulative)
-  // per-frame-bandwidth, for the rate control encoding stats below.
-  const double framerate = cfg->g_timebase.den / cfg->g_timebase.num;
-
-  for (sl = 0; sl < cfg->ss_number_layers; ++sl) {
-    for (tl = 0; tl < cfg->ts_number_layers; ++tl) {
-      const int layer = sl * cfg->ts_number_layers + tl;
-      const int tlayer0 = sl * cfg->ts_number_layers;
-      rc->layer_framerate[layer] =
-          framerate / cfg->ts_rate_decimator[tl];
-      if (tl > 0) {
-        rc->layer_pfb[layer] = 1000.0 *
-            (cfg->layer_target_bitrate[layer] -
-                cfg->layer_target_bitrate[layer - 1]) /
-            (rc->layer_framerate[layer] -
-                rc->layer_framerate[layer - 1]);
-      } else {
-        rc->layer_pfb[tlayer0] = 1000.0 *
-            cfg->layer_target_bitrate[tlayer0] /
-            rc->layer_framerate[tlayer0];
-      }
-      rc->layer_input_frames[layer] = 0;
-      rc->layer_enc_frames[layer] = 0;
-      rc->layer_tot_enc_frames[layer] = 0;
-      rc->layer_encoding_bitrate[layer] = 0.0;
-      rc->layer_avg_frame_size[layer] = 0.0;
-      rc->layer_avg_rate_mismatch[layer] = 0.0;
-    }
-  }
-  rc->window_count = 0;
-  rc->window_size = 15;
-  rc->avg_st_encoding_bitrate = 0.0;
-  rc->variance_st_encoding_bitrate = 0.0;
-}
-
-static void printout_rate_control_summary(struct RateControlStats *rc,
-                                          vpx_codec_enc_cfg_t *cfg,
-                                          int frame_cnt) {
-  unsigned int sl, tl;
-  int tot_num_frames = 0;
-  double perc_fluctuation = 0.0;
-  printf("Total number of processed frames: %d\n\n", frame_cnt - 1);
-  printf("Rate control layer stats for sl%d tl%d layer(s):\n\n",
-      cfg->ss_number_layers, cfg->ts_number_layers);
-  for (sl = 0; sl < cfg->ss_number_layers; ++sl) {
-    for (tl = 0; tl < cfg->ts_number_layers; ++tl) {
-      const int layer = sl * cfg->ts_number_layers + tl;
-      const int num_dropped = (tl > 0) ?
-          (rc->layer_input_frames[layer] - rc->layer_enc_frames[layer]) :
-          (rc->layer_input_frames[layer] - rc->layer_enc_frames[layer] - 1);
-      if (!sl)
-        tot_num_frames += rc->layer_input_frames[layer];
-      rc->layer_encoding_bitrate[layer] = 0.001 * rc->layer_framerate[layer] *
-          rc->layer_encoding_bitrate[layer] / tot_num_frames;
-      rc->layer_avg_frame_size[layer] = rc->layer_avg_frame_size[layer] /
-          rc->layer_enc_frames[layer];
-      rc->layer_avg_rate_mismatch[layer] =
-          100.0 * rc->layer_avg_rate_mismatch[layer] /
-          rc->layer_enc_frames[layer];
-      printf("For layer#: sl%d tl%d \n", sl, tl);
-      printf("Bitrate (target vs actual): %d %f.0 kbps\n",
-             cfg->layer_target_bitrate[layer],
-             rc->layer_encoding_bitrate[layer]);
-      printf("Average frame size (target vs actual): %f %f bits\n",
-             rc->layer_pfb[layer], rc->layer_avg_frame_size[layer]);
-      printf("Average rate_mismatch: %f\n",
-             rc->layer_avg_rate_mismatch[layer]);
-      printf("Number of input frames, encoded (non-key) frames, "
-          "and percent dropped frames: %d %d %f.0 \n",
-          rc->layer_input_frames[layer], rc->layer_enc_frames[layer],
-          100.0 * num_dropped / rc->layer_input_frames[layer]);
-      printf("\n");
-    }
-  }
-  rc->avg_st_encoding_bitrate = rc->avg_st_encoding_bitrate / rc->window_count;
-  rc->variance_st_encoding_bitrate =
-      rc->variance_st_encoding_bitrate / rc->window_count -
-      (rc->avg_st_encoding_bitrate * rc->avg_st_encoding_bitrate);
-  perc_fluctuation = 100.0 * sqrt(rc->variance_st_encoding_bitrate) /
-      rc->avg_st_encoding_bitrate;
-  printf("Short-time stats, for window of %d frames: \n", rc->window_size);
-  printf("Average, rms-variance, and percent-fluct: %f %f %f \n",
-         rc->avg_st_encoding_bitrate,
-         sqrt(rc->variance_st_encoding_bitrate),
-         perc_fluctuation);
-  if (frame_cnt != tot_num_frames)
-    die("Error: Number of input frames not equal to output encoded frames != "
-        "%d tot_num_frames = %d\n", frame_cnt, tot_num_frames);
-}
-
-vpx_codec_err_t parse_superframe_index(const uint8_t *data,
-                                       size_t data_sz,
-                                       uint32_t sizes[8], int *count) {
-  // A chunk ending with a byte matching 0xc0 is an invalid chunk unless
-  // it is a super frame index. If the last byte of real video compression
-  // data is 0xc0 the encoder must add a 0 byte. If we have the marker but
-  // not the associated matching marker byte at the front of the index we have
-  // an invalid bitstream and need to return an error.
-
-  uint8_t marker;
-
-  marker = *(data + data_sz - 1);
-  *count = 0;
-
-
-  if ((marker & 0xe0) == 0xc0) {
-    const uint32_t frames = (marker & 0x7) + 1;
-    const uint32_t mag = ((marker >> 3) & 0x3) + 1;
-    const size_t index_sz = 2 + mag * frames;
-
-    // This chunk is marked as having a superframe index but doesn't have
-    // enough data for it, thus it's an invalid superframe index.
-    if (data_sz < index_sz)
-      return VPX_CODEC_CORRUPT_FRAME;
-
-    {
-      const uint8_t marker2 = *(data + data_sz - index_sz);
-
-      // This chunk is marked as having a superframe index but doesn't have
-      // the matching marker byte at the front of the index therefore it's an
-      // invalid chunk.
-      if (marker != marker2)
-        return VPX_CODEC_CORRUPT_FRAME;
-    }
-
-    {
-      // Found a valid superframe index.
-      uint32_t i, j;
-      const uint8_t *x = &data[data_sz - index_sz + 1];
-
-      for (i = 0; i < frames; ++i) {
-        uint32_t this_sz = 0;
-
-        for (j = 0; j < mag; ++j)
-          this_sz |= (*x++) << (j * 8);
-        sizes[i] = this_sz;
-      }
-      *count = frames;
-    }
-  }
-  return VPX_CODEC_OK;
-}
-#endif
-
-// Example pattern for spatial layers and 2 temporal layers used in the
-// bypass/flexible mode. The pattern corresponds to the pattern
-// VP9E_TEMPORAL_LAYERING_MODE_0101 (temporal_layering_mode == 2) used in
-// non-flexible mode.
-void set_frame_flags_bypass_mode(int sl, int tl, int num_spatial_layers,
-                                 int is_key_frame,
-                                 vpx_svc_ref_frame_config_t *ref_frame_config) {
-  for (sl = 0; sl < num_spatial_layers; ++sl) {
-    if (!tl) {
-      if (!sl) {
-        ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_GF |
-                                            VP8_EFLAG_NO_REF_ARF |
-                                            VP8_EFLAG_NO_UPD_GF |
-                                            VP8_EFLAG_NO_UPD_ARF;
-      } else {
-        if (is_key_frame) {
-          ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_LAST |
-                                              VP8_EFLAG_NO_REF_ARF |
-                                              VP8_EFLAG_NO_UPD_GF |
-                                              VP8_EFLAG_NO_UPD_ARF;
-        } else {
-        ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_ARF |
-                                            VP8_EFLAG_NO_UPD_GF |
-                                            VP8_EFLAG_NO_UPD_ARF;
-        }
-      }
-    } else if (tl == 1) {
-      if (!sl) {
-        ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_GF |
-                                            VP8_EFLAG_NO_REF_ARF |
-                                            VP8_EFLAG_NO_UPD_LAST |
-                                            VP8_EFLAG_NO_UPD_GF;
-      } else {
-        ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_ARF |
-                                            VP8_EFLAG_NO_UPD_LAST |
-                                            VP8_EFLAG_NO_UPD_GF;
-      }
-    }
-    if (tl == 0) {
-      ref_frame_config->lst_fb_idx[sl] = sl;
-      if (sl)
-        ref_frame_config->gld_fb_idx[sl] = sl - 1;
-      else
-        ref_frame_config->gld_fb_idx[sl] = 0;
-      ref_frame_config->alt_fb_idx[sl] = 0;
-    } else if (tl == 1) {
-      ref_frame_config->lst_fb_idx[sl] = sl;
-      ref_frame_config->gld_fb_idx[sl] = num_spatial_layers + sl - 1;
-      ref_frame_config->alt_fb_idx[sl] = num_spatial_layers + sl;
-    }
-  }
-}
-
 int main(int argc, const char **argv) {
  AppInput app_input = {0};
  VpxVideoWriter *writer = NULL;
@@ -613,18 +332,7 @@ int main(int argc, const char **argv) {
  FILE *infile = NULL;
  int end_of_stream = 0;
  int frames_received = 0;
-#if OUTPUT_RC_STATS
-  VpxVideoWriter *outfile[VPX_TS_MAX_LAYERS] = {NULL};
-  struct RateControlStats rc;
-  vpx_svc_layer_id_t layer_id;
-  vpx_svc_ref_frame_config_t ref_frame_config;
-  int sl, tl;
-  double sum_bitrate = 0.0;
-  double sum_bitrate2 = 0.0;
-  double framerate  = 30.0;
-#endif
-  struct vpx_usec_timer timer;
-  int64_t cx_time = 0;
+
  memset(&svc_ctx, 0, sizeof(svc_ctx));
  svc_ctx.log_print = 1;
  exec_name = argv[0];
@@ -651,13 +359,6 @@ int main(int argc, const char **argv) {
      VPX_CODEC_OK)
    die("Failed to initialize encoder\n");

-#if OUTPUT_RC_STATS
-  if (svc_ctx.output_rc_stat) {
-    set_rate_control_stats(&rc, &enc_cfg);
-    framerate = enc_cfg.g_timebase.den / enc_cfg.g_timebase.num;
-  }
-#endif
-
  info.codec_fourcc = VP9_FOURCC;
  info.time_base.numerator = enc_cfg.g_timebase.num;
  info.time_base.denominator = enc_cfg.g_timebase.den;
@@ -669,34 +370,11 @@ int main(int argc, const char **argv) {
    if (!writer)
      die("Failed to open %s for writing\n", app_input.output_filename);
  }
-#if OUTPUT_RC_STATS
-  // For now, just write temporal layer streams.
-  // TODO(wonkap): do spatial by re-writing superframe.
-  if (svc_ctx.output_rc_stat) {
-    for (tl = 0; tl < enc_cfg.ts_number_layers; ++tl) {
-      char file_name[PATH_MAX];
-
-      snprintf(file_name, sizeof(file_name), "%s_t%d.ivf",
-               app_input.output_filename, tl);
-      outfile[tl] = vpx_video_writer_open(file_name, kContainerIVF, &info);
-      if (!outfile[tl])
-        die("Failed to open %s for writing", file_name);
-    }
-  }
-#endif

  // skip initial frames
  for (i = 0; i < app_input.frames_to_skip; ++i)
    vpx_img_read(&raw, infile);

-  if (svc_ctx.speed != -1)
-    vpx_codec_control(&codec, VP8E_SET_CPUUSED, svc_ctx.speed);
-  if (svc_ctx.threads)
-    vpx_codec_control(&codec, VP9E_SET_TILE_COLUMNS, (svc_ctx.threads >> 1));
-  if (svc_ctx.speed >= 5 && svc_ctx.aqmode == 1)
-    vpx_codec_control(&codec, VP9E_SET_AQ_MODE, 3);
-
-
  // Encode frames
  while (!end_of_stream) {
    vpx_codec_iter_t iter = NULL;
@@ -707,37 +385,8 @@ int main(int argc, const char **argv) {
      end_of_stream = 1;
    }

-    // For BYPASS/FLEXIBLE mode, set the frame flags (reference and updates)
-    // and the buffer indices for each spatial layer of the current
-    // (super)frame to be encoded. The temporal layer_id for the current frame
-    // also needs to be set.
-    // TODO(marpan): Should rename the "VP9E_TEMPORAL_LAYERING_MODE_BYPASS"
-    // mode to "VP9E_LAYERING_MODE_BYPASS".
-    if (svc_ctx.temporal_layering_mode == VP9E_TEMPORAL_LAYERING_MODE_BYPASS) {
-      // Example for 2 temporal layers.
-      if (frame_cnt % 2 == 0)
-        layer_id.temporal_layer_id = 0;
-      else
-        layer_id.temporal_layer_id = 1;
-      // Note that we only set the temporal layer_id, since we are calling
-      // the encode for the whole superframe. The encoder will internally loop
-      // over all the spatial layers for the current superframe.
-      vpx_codec_control(&codec, VP9E_SET_SVC_LAYER_ID, &layer_id);
-      set_frame_flags_bypass_mode(sl, layer_id.temporal_layer_id,
-                                  svc_ctx.spatial_layers,
-                                  frame_cnt == 0,
-                                  &ref_frame_config);
-      vpx_codec_control(&codec, VP9E_SET_SVC_REF_FRAME_CONFIG,
-                        &ref_frame_config);
-    }
-
-    vpx_usec_timer_start(&timer);
    res = vpx_svc_encode(&svc_ctx, &codec, (end_of_stream ? NULL : &raw),
-                         pts, frame_duration, svc_ctx.speed >= 5 ?
-                         VPX_DL_REALTIME : VPX_DL_GOOD_QUALITY);
-    vpx_usec_timer_mark(&timer);
-    cx_time += vpx_usec_timer_elapsed(&timer);
-
+                         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");
@@ -746,90 +395,11 @@ int main(int argc, const char **argv) {
    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) {
-#if OUTPUT_RC_STATS
-            uint32_t sizes[8];
-            int count = 0;
-#endif
+          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);
-#if OUTPUT_RC_STATS
-            // TODO(marpan/wonkap): Put this (to line728) in separate function.
-            if (svc_ctx.output_rc_stat) {
-              vpx_codec_control(&codec, VP9E_GET_SVC_LAYER_ID, &layer_id);
-              parse_superframe_index(cx_pkt->data.frame.buf,
-                                     cx_pkt->data.frame.sz, sizes, &count);
-              for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
-                ++rc.layer_input_frames[sl * enc_cfg.ts_number_layers +
-                                        layer_id.temporal_layer_id];
-              }
-              for (tl = layer_id.temporal_layer_id;
-                  tl < enc_cfg.ts_number_layers; ++tl) {
-                vpx_video_writer_write_frame(outfile[tl],
-                                             cx_pkt->data.frame.buf,
-                                             cx_pkt->data.frame.sz,
-                                             cx_pkt->data.frame.pts);
-              }
-
-              for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
-                for (tl = layer_id.temporal_layer_id;
-                    tl < enc_cfg.ts_number_layers; ++tl) {
-                  const int layer = sl * enc_cfg.ts_number_layers + tl;
-                  ++rc.layer_tot_enc_frames[layer];
-                  rc.layer_encoding_bitrate[layer] += 8.0 * sizes[sl];
-                  // Keep count of rate control stats per layer, for non-key
-                  // frames.
-                  if (tl == layer_id.temporal_layer_id &&
-                      !(cx_pkt->data.frame.flags & VPX_FRAME_IS_KEY)) {
-                    rc.layer_avg_frame_size[layer] += 8.0 * sizes[sl];
-                    rc.layer_avg_rate_mismatch[layer] +=
-                        fabs(8.0 * sizes[sl] - rc.layer_pfb[layer]) /
-                        rc.layer_pfb[layer];
-                    ++rc.layer_enc_frames[layer];
-                  }
-                }
-              }
-
-              // Update for short-time encoding bitrate states, for moving
-              // window of size rc->window, shifted by rc->window / 2.
-              // Ignore first window segment, due to key frame.
-              if (frame_cnt > rc.window_size) {
-                tl = layer_id.temporal_layer_id;
-                for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
-                  sum_bitrate += 0.001 * 8.0 * sizes[sl] * framerate;
-                }
-                if (frame_cnt % rc.window_size == 0) {
-                  rc.window_count += 1;
-                  rc.avg_st_encoding_bitrate += sum_bitrate / rc.window_size;
-                  rc.variance_st_encoding_bitrate +=
-                      (sum_bitrate / rc.window_size) *
-                      (sum_bitrate / rc.window_size);
-                  sum_bitrate = 0.0;
-                }
-              }
-
-              // Second shifted window.
-              if (frame_cnt > rc.window_size + rc.window_size / 2) {
-               tl = layer_id.temporal_layer_id;
-               for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
-                 sum_bitrate2 += 0.001 * 8.0 * sizes[sl] * framerate;
-               }
-
-               if (frame_cnt > 2 * rc.window_size &&
-                  frame_cnt % rc.window_size == 0) {
-                 rc.window_count += 1;
-                 rc.avg_st_encoding_bitrate += sum_bitrate2 / rc.window_size;
-                 rc.variance_st_encoding_bitrate +=
-                    (sum_bitrate2 / rc.window_size) *
-                    (sum_bitrate2 / rc.window_size);
-                 sum_bitrate2 = 0.0;
-               }
-              }
-            }
-#endif
-          }

          printf("SVC frame: %d, kf: %d, size: %d, pts: %d\n", frames_received,
                 !!(cx_pkt->data.frame.flags & VPX_FRAME_IS_KEY),
@@ -854,34 +424,25 @@ int main(int argc, const char **argv) {
      pts += frame_duration;
    }
  }
+
  printf("Processed %d frames\n", frame_cnt);
+
  fclose(infile);
-#if OUTPUT_RC_STATS
-  if (svc_ctx.output_rc_stat) {
-    printout_rate_control_summary(&rc, &enc_cfg, frame_cnt);
-    printf("\n");
-  }
-#endif
  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);
  }
-#if OUTPUT_RC_STATS
-  if (svc_ctx.output_rc_stat) {
-    for (tl = 0; tl < enc_cfg.ts_number_layers; ++tl) {
-      vpx_video_writer_close(outfile[tl]);
-    }
-  }
-#endif
-  printf("Frame cnt and encoding time/FPS stats for encoding: %d %f %f \n",
-         frame_cnt,
-         1000 * (float)cx_time / (double)(frame_cnt * 1000000),
-         1000000 * (double)frame_cnt / (double)cx_time);
+
  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);
 }

@@ -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]);
@@ -598,16 +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)
@@ -618,9 +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;

@@ -632,8 +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,
@@ -642,8 +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"))) {
@@ -684,25 +675,17 @@ int main(int argc, char **argv) {
  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_STATIC_THRESHOLD, 1);
+    vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 0);
  } 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, 0);
-    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, 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");
    }
-    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/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))
@@ -47,13 +54,7 @@ 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))
-
-ifeq ($(CONFIG_VP8),yes)
+ifneq ($(CONFIG_VP8_ENCODER)$(CONFIG_VP8_DECODER),)
  VP8_PREFIX=vp8/
  include $(SRC_PATH_BARE)/$(VP8_PREFIX)vp8_common.mk
 endif
@@ -76,7 +77,7 @@ ifeq ($(CONFIG_VP8_DECODER),yes)
  CODEC_DOC_SECTIONS += vp8 vp8_decoder
 endif

-ifeq ($(CONFIG_VP9),yes)
+ifneq ($(CONFIG_VP9_ENCODER)$(CONFIG_VP9_DECODER),)
  VP9_PREFIX=vp9/
  include $(SRC_PATH_BARE)/$(VP9_PREFIX)vp9_common.mk
 endif
@@ -109,40 +110,6 @@ endif
 VP9_PREFIX=vp9/
 $(BUILD_PFX)$(VP9_PREFIX)%.c.o: CFLAGS += -Wextra

-#  VP10 make file
-ifeq ($(CONFIG_VP10),yes)
-  VP10_PREFIX=vp10/
-  include $(SRC_PATH_BARE)/$(VP10_PREFIX)vp10_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 += $(VP10_PREFIX)vp10cx.mk vpx/vp8.h vpx/vp8cx.h
-  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8cx.h
-  INSTALL-LIBS-$(CONFIG_SPATIAL_SVC) += include/vpx/svc_context.h
-  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP10_PREFIX)/%
-  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
-  INSTALL-LIBS-yes += include/vpx/vp8.h include/vpx/vp8dx.h
-  INSTALL_MAPS += include/vpx/% $(SRC_PATH_BARE)/$(VP10_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
-
 ifeq ($(CONFIG_ENCODERS),yes)
  CODEC_DOC_SECTIONS += encoder
 endif
@@ -170,18 +137,18 @@ 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
 CODEC_EXPORTS-$(CONFIG_DECODERS) += vpx/exports_dec

@@ -248,7 +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): $(RTCD)
@@ -256,39 +223,31 @@ vpx.$(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
+
+BUILD_LIBVPX_SO         := $(if $(BUILD_LIBVPX),$(CONFIG_SHARED))
+
+SO_VERSION_MAJOR := 2
 SO_VERSION_MINOR := 0
 SO_VERSION_PATCH := 0
 ifeq ($(filter darwin%,$(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
 EXPORT_FILE             := libvpx.ver
 LIBVPX_SO_SYMLINKS      := $(addprefix $(LIBSUBDIR)/, \
                             libvpx.so libvpx.so.$(SO_VERSION_MAJOR) \
                             libvpx.so.$(SO_VERSION_MAJOR).$(SO_VERSION_MINOR))
 endif
-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)
@@ -306,19 +265,6 @@ 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 "_"$$2}' $^ >>$@
-CLEAN-OBJS += libvpx.def
-
-libvpx_dll.a: $(LIBVPX_SO)
-	@echo "    [IMPLIB] $@"
-	$(qexec)emximp -o $@ $<
-CLEAN-OBJS += libvpx_dll.a
-
 define libvpx_symlink_template
 $(1): $(2)
 	@echo "    [LN]     $(2) $$@"
@@ -334,12 +280,11 @@ $(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)' > $@
@@ -365,6 +310,9 @@ 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
 #
@@ -403,15 +351,11 @@ 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 > $@
@@ -475,25 +419,7 @@ test_libvpx.$(VCPROJ_SFX): $(LIBVPX_TEST_SRCS) vpx.$(VCPROJ_SFX) gtest.$(VCPROJ_

 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

@@ -504,54 +430,45 @@ 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

@@ -596,16 +513,15 @@ 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

@@ -623,12 +539,11 @@ 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;

--- 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/active_map_refresh_test.cc
+++ b/test/active_map_refresh_test.cc
@@ -1,127 +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;
-
-  ::libvpx_test::Y4mVideoSource video("desktop_credits.y4m", 0, 30);
-  ::libvpx_test::Y4mVideoSource video_holder("desktop_credits.y4m", 0, 30);
-  video_holder.Begin();
-  y4m_holder_ = &video_holder;
-
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-}
-
-VP9_INSTANTIATE_TEST_CASE(ActiveMapRefreshTest,
-                          ::testing::Values(::libvpx_test::kRealTime),
-                          ::testing::Range(5, 6));
-}  // namespace
--- a/test/android/Android.mk
+++ b/test/android/Android.mk
@@ -51,6 +51,4 @@ 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/blockiness_test.cc
+++ b/test/blockiness_test.cc
@@ -8,11 +8,10 @@
 *  be found in the AUTHORS file in the root of the source tree.
 */

+
+#include <string.h>
 #include <limits.h>
 #include <stdio.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"

 #include "./vpx_config.h"
 #if CONFIG_VP9_ENCODER
@@ -23,6 +22,7 @@
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"

 #include "vpx_mem/vpx_mem.h"

--- a/test/borders_test.cc
+++ b/test/borders_test.cc
@@ -80,7 +80,4 @@ TEST_P(BordersTest, TestLowBitrate) {

 VP9_INSTANTIATE_TEST_CASE(BordersTest, ::testing::Values(
    ::libvpx_test::kTwoPassGood));
-
-VP10_INSTANTIATE_TEST_CASE(BordersTest, ::testing::Values(
-    ::libvpx_test::kTwoPassGood));
 }  // 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_VP8_ENCODER || CONFIG_VP9_ENCODER || CONFIG_VP10_ENCODER
+#if CONFIG_VP8_ENCODER || CONFIG_VP9_ENCODER
 #include "vpx/vp8cx.h"
 #endif
-#if CONFIG_VP8_DECODER || CONFIG_VP9_DECODER || CONFIG_VP10_DECODER
+#if CONFIG_VP8_DECODER || CONFIG_VP9_DECODER
 #include "vpx/vp8dx.h"
 #endif

@@ -233,8 +233,6 @@ class VP9CodecFactory : public CodecFactory {
                                               int usage) const {
 #if CONFIG_VP9_ENCODER
    return vpx_codec_enc_config_default(&vpx_codec_vp9_cx_algo, cfg, usage);
-#elif CONFIG_VP10_ENCODER
-    return vpx_codec_enc_config_default(&vpx_codec_vp10_cx_algo, cfg, usage);
 #else
    return VPX_CODEC_INCAPABLE;
 #endif
@@ -253,96 +251,7 @@ const libvpx_test::VP9CodecFactory kVP9;
 #define VP9_INSTANTIATE_TEST_CASE(test, ...)
 #endif  // CONFIG_VP9

-/*
- * VP10 Codec Definitions
- */
-#if CONFIG_VP10
-class VP10Decoder : public Decoder {
- public:
-  VP10Decoder(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
-      : Decoder(cfg, flag, deadline) {}
-
- protected:
-  virtual vpx_codec_iface_t* CodecInterface() const {
-#if CONFIG_VP10_DECODER
-    return &vpx_codec_vp10_dx_algo;
-#else
-    return NULL;
-#endif
-  }
-};
-
-class VP10Encoder : public Encoder {
- public:
-  VP10Encoder(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;
-#else
-    return NULL;
-#endif
-  }
-};
-
-class VP10CodecFactory : public CodecFactory {
- public:
-  VP10CodecFactory() : 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_VP10_DECODER
-    return new VP10Decoder(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_VP10_ENCODER
-    return new VP10Encoder(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_VP10_ENCODER
-    return vpx_codec_enc_config_default(&vpx_codec_vp10_cx_algo, cfg, usage);
-#else
-    return VPX_CODEC_INCAPABLE;
-#endif
-  }
-};
-
-const libvpx_test::VP10CodecFactory kVP10;
-
-#define VP10_INSTANTIATE_TEST_CASE(test, ...)\
-  INSTANTIATE_TEST_CASE_P(VP10, test, \
-      ::testing::Combine( \
-          ::testing::Values(static_cast<const libvpx_test::CodecFactory*>( \
-               &libvpx_test::kVP10)), \
-          __VA_ARGS__))
-#else
-#define VP10_INSTANTIATE_TEST_CASE(test, ...)
-#endif  // CONFIG_VP10

 }  // namespace libvpx_test
+
 #endif  // TEST_CODEC_FACTORY_H_
--- a/test/consistency_test.cc
+++ b/test/consistency_test.cc
@@ -8,11 +8,10 @@
 *  be found in the AUTHORS file in the root of the source tree.
 */

+
+#include <string.h>
 #include <limits.h>
 #include <stdio.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"

 #include "./vpx_config.h"
 #if CONFIG_VP9_ENCODER
@@ -23,11 +22,12 @@
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.h"
-#include "vpx_dsp/ssim.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "vp9/encoder/vp9_ssim.h"
 #include "vpx_mem/vpx_mem.h"

 extern "C"
-double vpx_get_ssim_metrics(uint8_t *img1, int img1_pitch,
+double vp9_get_ssim_metrics(uint8_t *img1, int img1_pitch,
                            uint8_t *img2, int img2_pitch,
                            int width, int height,
                            Ssimv *sv2, Metrics *m,
@@ -65,7 +65,7 @@ class ConsistencyTestBase : public ::testing::Test {
    vpx_free(reference_data_[1]);
    reference_data_[1] = NULL;

-    delete[] ssim_array_;
+    delete ssim_array_;
  }

  virtual void TearDown() {
@@ -144,7 +144,7 @@ class ConsistencyVP9Test
  double CheckConsistency(int frame) {
    EXPECT_LT(frame, 2)<< "Frame to check has to be less than 2.";
    return
-        vpx_get_ssim_metrics(source_data_[frame], source_stride_,
+        vp9_get_ssim_metrics(source_data_[frame], source_stride_,
                             reference_data_[frame], reference_stride_,
                             width_, height_, ssim_array_, &metrics_, 1);
  }
--- a/test/convolve_test.cc
+++ b/test/convolve_test.cc
--- a/test/cpu_speed_test.cc
+++ b/test/cpu_speed_test.cc
@@ -140,9 +140,4 @@ VP9_INSTANTIATE_TEST_CASE(
    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood,
                      ::libvpx_test::kRealTime),
    ::testing::Range(0, 9));
-
-VP10_INSTANTIATE_TEST_CASE(
-    CpuSpeedTest,
-    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood),
-    ::testing::Range(0, 3));
 }  // namespace
--- 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 {

@@ -372,7 +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);
@@ -564,8 +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;

@@ -575,14 +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;
    }
@@ -607,27 +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;
    }
@@ -655,22 +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,
@@ -742,237 +737,9 @@ TEST_P(DatarateTestVP9Large, DenoiserOffOn) {
 }
 #endif  // CONFIG_VP9_TEMPORAL_DENOISING

-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 < 2; ++i) {
-        svc_params_.max_quantizers[i] = 63;
-        svc_params_.min_quantizers[i] = 0;
-      }
-      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;
-      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_;
-      effective_datarate_ = (bits_total_ - bits_in_last_frame_) / 1000.0
-          / (cfg_.rc_buf_initial_sz / 1000.0 + duration_);
-      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_;
-  double effective_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,
-    unsigned int total_rate) {
-  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, OnePassCbrSvc) {
-  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 = 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;
-  // TODO(wonkap/marpan): No frame drop for now, we need to implement correct
-  // frame dropping for SVC.
-  cfg_.rc_dropframe_thresh = 0;
-  ::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. Also check if test can pass at lower bitrate (~200k).
-  for (int i = 400; 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,
-        cfg_.rc_target_bitrate);
-    ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-    ASSERT_GE(cfg_.rc_target_bitrate, effective_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(GetMismatchFrames(), (unsigned int) 0);
-  }
-}
-
-// 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, OnePassCbrSvc4threads) {
-  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;
-  // TODO(wonkap/marpan): No frame drop for now, we need to implement correct
-  // frame dropping for SVC.
-  cfg_.rc_dropframe_thresh = 0;
-  ::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,
-      cfg_.rc_target_bitrate);
-  ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
-  ASSERT_GE(cfg_.rc_target_bitrate, effective_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(GetMismatchFrames(), (unsigned int) 0);
-}
-
 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));
-VP9_INSTANTIATE_TEST_CASE(DatarateOnePassCbrSvc,
-                          ::testing::Values(::libvpx_test::kRealTime),
-                          ::testing::Range(5, 8));
 }  // namespace
--- a/test/dct16x16_test.cc
+++ b/test/dct16x16_test.cc
@@ -13,18 +13,15 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./vp9_rtcd.h"
 #include "vp9/common/vp9_entropy.h"
-#include "vp9/common/vp9_scan.h"
 #include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"

 using libvpx_test::ACMRandom;

@@ -40,6 +37,30 @@ static int round(double x) {
 #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;
@@ -248,12 +269,12 @@ 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,
@@ -268,11 +289,11 @@ void iht16x16_ref(const tran_low_t *in, uint8_t *dest, int stride,

 #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,
@@ -294,28 +315,28 @@ void iht16x16_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
 }

 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_VP9_HIGHBITDEPTH
@@ -335,13 +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]);
+      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(16, uint16_t, dst16[kNumCoeffs]);
-      DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+      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_].
@@ -395,9 +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_].
@@ -416,13 +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) {
@@ -449,19 +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]);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, dst, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint8_t, ref, kNumCoeffs);
 #if CONFIG_VP9_HIGHBITDEPTH
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
+    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)
@@ -511,13 +539,13 @@ 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]);
+    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(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+    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) {
@@ -571,12 +599,12 @@ class Trans16x16TestBase {
    const int count_test_block = 10000;
    const int eob = 10;
    const int16_t *scan = vp9_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]);
+    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(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
+    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) {
@@ -798,14 +826,14 @@ using std::tr1::make_tuple;
 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)));
+        make_tuple(&vp9_fdct16x16_c, &vp9_idct16x16_256_add_c, 0, VPX_BITS_8)));
 #endif  // CONFIG_VP9_HIGHBITDEPTH

 #if CONFIG_VP9_HIGHBITDEPTH
@@ -838,16 +866,16 @@ INSTANTIATE_TEST_CASE_P(
 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_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(
@@ -865,19 +893,27 @@ INSTANTIATE_TEST_CASE_P(
 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(&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),
@@ -897,20 +933,4 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(&idct16x16_12,
                   &idct16x16_256_add_12_sse2, 3167, VPX_BITS_12)));
 #endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-
-#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    MSA, 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(&vp9_fht16x16_msa, &vp9_iht16x16_256_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vp9_fht16x16_msa, &vp9_iht16x16_256_add_msa, 1, VPX_BITS_8),
-        make_tuple(&vp9_fht16x16_msa, &vp9_iht16x16_256_add_msa, 2, VPX_BITS_8),
-        make_tuple(&vp9_fht16x16_msa, &vp9_iht16x16_256_add_msa, 3,
-                   VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/dct32x32_test.cc
+++ b/test/dct32x32_test.cc
@@ -13,18 +13,16 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./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"

 using libvpx_test::ACMRandom;

@@ -82,15 +80,15 @@ typedef std::tr1::tuple<FwdTxfmFunc, InvTxfmFunc, int, vpx_bit_depth_t>

 #if CONFIG_VP9_HIGHBITDEPTH
 void idct32x32_8(const tran_low_t *in, uint8_t *out, int stride) {
-  vpx_highbd_idct32x32_1024_add_c(in, out, stride, 8);
+  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_VP9_HIGHBITDEPTH

@@ -121,13 +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]);
+  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(16, uint16_t, dst16[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+  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) {
@@ -186,16 +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) {
@@ -214,13 +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) {
@@ -232,7 +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));

@@ -257,13 +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]);
+  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(16, uint16_t, dst16[kNumCoeffs]);
-  DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+  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) {
@@ -315,61 +315,61 @@ using std::tr1::make_tuple;
 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)));
+        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)));
+        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_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)));
+        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_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)));
+        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_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)));
 #endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

@@ -377,19 +377,9 @@ INSTANTIATE_TEST_CASE_P(
 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)));
+        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
-
-#if HAVE_MSA && !CONFIG_VP9_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)));
-#endif  // HAVE_MSA && !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"

@@ -26,9 +25,6 @@ TEST(DecodeAPI, InvalidParams) {
 #endif
 #if CONFIG_VP9_DECODER
    &vpx_codec_vp9_dx_algo,
-#endif
-#if CONFIG_VP10_DECODER
-    &vpx_codec_vp10_dx_algo,
 #endif
  };
  uint8_t buf[1] = {0};
@@ -133,13 +129,8 @@ TEST(DecodeAPI, Vp9InvalidDecode) {
  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());
-#if CONFIG_VP9_HIGHBITDEPTH
  EXPECT_EQ(VPX_CODEC_MEM_ERROR,
            vpx_codec_decode(&dec, video.cxdata(), frame_size, NULL, 0));
-#else
-  EXPECT_EQ(VPX_CODEC_UNSUP_BITSTREAM,
-            vpx_codec_decode(&dec, video.cxdata(), frame_size, NULL, 0));
-#endif
  vpx_codec_iter_t iter = NULL;
  EXPECT_EQ(NULL, vpx_codec_get_frame(&dec, &iter));

--- a/test/decode_test_driver.cc
+++ b/test/decode_test_driver.cc
@@ -7,11 +7,9 @@
 *  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"

--- a/test/encode_test_driver.cc
+++ b/test/encode_test_driver.cc
@@ -10,14 +10,13 @@

 #include <string>

-#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/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) {
@@ -42,15 +41,6 @@ void Encoder::InitEncoder(VideoSource *video) {
                               log2_tile_columns);
      ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
    } else
-#endif
-#if CONFIG_VP10_ENCODER
-    if (CodecInterface() == &vpx_codec_vp10_cx_algo) {
-      // Default to 1 tile column for VP10.
-      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();
-    } else
 #endif
    {
 #if CONFIG_VP8_ENCODER
@@ -195,7 +185,6 @@ void EncoderTest::RunLoop(VideoSource *video) {

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

    unsigned long dec_init_flags = 0;  // NOLINT
    // Use fragment decoder if encoder outputs partitions.
--- 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_VP8_ENCODER || CONFIG_VP9_ENCODER || 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_VP8_ENCODER || CONFIG_VP9_ENCODER || 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();
--- 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),
@@ -194,8 +193,6 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
     pattern_switch_ = frame_switch;
   }

-  bool svc_support_;
-
 private:
  double psnr_;
  unsigned int nframes_;
@@ -305,10 +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;
@@ -354,10 +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;
@@ -590,13 +579,8 @@ TEST_P(ErrorResilienceTestLargeCodecControls, CodecControl3TemporalLayers) {
  }
 }

-VP8_INSTANTIATE_TEST_CASE(ErrorResilienceTestLarge, ONE_PASS_TEST_MODES,
-                          ::testing::Values(true));
+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,
-                          ::testing::Values(true));
-// 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));
+VP9_INSTANTIATE_TEST_CASE(ErrorResilienceTestLarge, ONE_PASS_TEST_MODES);
 }  // namespace
--- a/test/fdct4x4_test.cc
+++ b/test/fdct4x4_test.cc
@@ -13,17 +13,15 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./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;

@@ -41,7 +39,7 @@ 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);
+  vp9_fdct4x4_c(in, out, stride);
 }

 void fht4x4_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
@@ -55,11 +53,11 @@ void fwht4x4_ref(const int16_t *in, tran_low_t *out, int stride,

 #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) {
@@ -71,20 +69,20 @@ void iht4x4_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
 }

 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_VP9_HIGHBITDEPTH
@@ -104,13 +102,13 @@ class Trans4x4TestBase {
    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]);
+      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(16, uint16_t, dst16[kNumCoeffs]);
-      DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+      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].
@@ -144,7 +142,6 @@ class Trans4x4TestBase {
        const uint32_t diff =
            bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
 #else
-        ASSERT_EQ(VPX_BITS_8, bit_depth_);
        const uint32_t diff = dst[j] - src[j];
 #endif
        const uint32_t error = diff * diff;
@@ -166,9 +163,9 @@ class Trans4x4TestBase {
  void RunCoeffCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 5000;
-    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_].
@@ -187,13 +184,15 @@ class Trans4x4TestBase {
  void RunMemCheck() {
    ACMRandom rnd(ACMRandom::DeterministicSeed());
    const int count_test_block = 5000;
-    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) {
@@ -220,13 +219,13 @@ class Trans4x4TestBase {
  void RunInvAccuracyCheck(int limit) {
    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]);
+    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(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, src16[kNumCoeffs]);
+    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) {
@@ -419,14 +418,14 @@ using std::tr1::make_tuple;
 INSTANTIATE_TEST_CASE_P(
    C, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_10, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_12, 0, VPX_BITS_12),
-        make_tuple(&vpx_fdct4x4_c, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8)));
+        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)));
+        make_tuple(&vp9_fdct4x4_c, &vp9_idct4x4_16_add_c, 0, VPX_BITS_8)));
 #endif  // CONFIG_VP9_HIGHBITDEPTH

 #if CONFIG_VP9_HIGHBITDEPTH
@@ -461,20 +460,20 @@ INSTANTIATE_TEST_CASE_P(
    ::testing::Values(
        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, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8)));
+        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(&vp9_fwht4x4_c, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8)));
+        make_tuple(&vp9_fwht4x4_c, &vp9_iwht4x4_16_add_c, 0, VPX_BITS_8)));
 #endif  // CONFIG_VP9_HIGHBITDEPTH

 #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)));
+        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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
@@ -492,23 +491,15 @@ INSTANTIATE_TEST_CASE_P(
 INSTANTIATE_TEST_CASE_P(
    MMX, Trans4x4WHT,
    ::testing::Values(
-        make_tuple(&vp9_fwht4x4_mmx, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8)));
-#endif
-
-#if CONFIG_USE_X86INC && HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && \
-    !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    SSE2, Trans4x4WHT,
-    ::testing::Values(
-        make_tuple(&vp9_fwht4x4_c, &vpx_iwht4x4_16_add_sse2, 0, VPX_BITS_8)));
+        make_tuple(&vp9_fwht4x4_mmx, &vp9_iwht4x4_16_add_c, 0, VPX_BITS_8)));
 #endif

 #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)));
+        make_tuple(&vp9_fdct4x4_sse2,
+                   &vp9_idct4x4_16_add_sse2, 0, VPX_BITS_8)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4HT,
    ::testing::Values(
@@ -522,33 +513,27 @@ INSTANTIATE_TEST_CASE_P(
 INSTANTIATE_TEST_CASE_P(
    SSE2, Trans4x4DCT,
    ::testing::Values(
-        make_tuple(&vpx_highbd_fdct4x4_c,    &idct4x4_10_sse2, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_10_sse2, 0, VPX_BITS_10),
-        make_tuple(&vpx_highbd_fdct4x4_c,    &idct4x4_12_sse2, 0, VPX_BITS_12),
-        make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_12_sse2, 0, VPX_BITS_12),
-        make_tuple(&vpx_fdct4x4_sse2,      &vpx_idct4x4_16_add_c, 0,
+        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(&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
-
-#if HAVE_MSA && !CONFIG_VP9_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)));
-INSTANTIATE_TEST_CASE_P(
-    MSA, Trans4x4HT,
-    ::testing::Values(
-        make_tuple(&vp9_fht4x4_msa, &vp9_iht4x4_16_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vp9_fht4x4_msa, &vp9_iht4x4_16_add_msa, 1, VPX_BITS_8),
-        make_tuple(&vp9_fht4x4_msa, &vp9_iht4x4_16_add_msa, 2, VPX_BITS_8),
-        make_tuple(&vp9_fht4x4_msa, &vp9_iht4x4_16_add_msa, 3, VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/fdct8x8_test.cc
+++ b/test/fdct8x8_test.cc
@@ -13,40 +13,18 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./vp9_rtcd.h"
 #include "vp9/common/vp9_entropy.h"
-#include "vp9/common/vp9_scan.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], int stride) {
  const double kInvSqrt2 = 0.707106781186547524400844362104;
  for (int k = 0; k < 8; k++) {
@@ -81,9 +59,26 @@ void reference_8x8_dct_2d(const int16_t input[kNumCoeffs],
  }
 }

+using libvpx_test::ACMRandom;
+
+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) {
-  vpx_fdct8x8_c(in, out, stride);
+  vp9_fdct8x8_c(in, out, stride);
 }

 void fht8x8_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
@@ -92,11 +87,11 @@ void fht8x8_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {

 #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) {
@@ -108,28 +103,28 @@ void iht8x8_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
 }

 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_VP9_HIGHBITDEPTH
@@ -144,8 +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;

@@ -215,13 +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]);
+    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(16, uint16_t, dst16[64]);
-    DECLARE_ALIGNED(16, uint16_t, src16[64]);
+    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) {
@@ -292,14 +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]);
+    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(16, uint16_t, dst16[64]);
-    DECLARE_ALIGNED(16, uint16_t, src16[64]);
+    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) {
@@ -381,13 +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]);
+    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(16, uint16_t, src16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, dst16[kNumCoeffs]);
+    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) {
@@ -439,9 +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];
@@ -469,12 +464,12 @@ 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]);
+    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(16, uint16_t, dst16[kNumCoeffs]);
-    DECLARE_ALIGNED(16, uint16_t, ref16[kNumCoeffs]);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, dst16, kNumCoeffs);
+    DECLARE_ALIGNED_ARRAY(16, uint16_t, ref16, kNumCoeffs);
 #endif
    const int16_t *scan = vp9_default_scan_orders[TX_8X8].scan;

@@ -658,14 +653,14 @@ using std::tr1::make_tuple;
 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)));
+        make_tuple(&vp9_fdct8x8_c, &vp9_idct8x8_64_add_c, 0, VPX_BITS_8)));
 #endif  // CONFIG_VP9_HIGHBITDEPTH

 #if CONFIG_VP9_HIGHBITDEPTH
@@ -685,6 +680,8 @@ INSTANTIATE_TEST_CASE_P(
        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(
@@ -695,10 +692,12 @@ INSTANTIATE_TEST_CASE_P(
 #endif  // CONFIG_VP9_HIGHBITDEPTH

 #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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

@@ -713,10 +712,12 @@ INSTANTIATE_TEST_CASE_P(
 #endif  // HAVE_NEON && !CONFIG_VP9_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,
@@ -731,16 +732,18 @@ INSTANTIATE_TEST_CASE_P(
 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(
@@ -764,26 +767,14 @@ INSTANTIATE_TEST_CASE_P(
                   &idct8x8_64_add_12_sse2, 6225, VPX_BITS_12)));
 #endif  // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

-#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 && \
-    !CONFIG_VP9_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_VP9_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(&vp9_fht8x8_msa, &vp9_iht8x8_64_add_msa, 0, VPX_BITS_8),
-        make_tuple(&vp9_fht8x8_msa, &vp9_iht8x8_64_add_msa, 1, VPX_BITS_8),
-        make_tuple(&vp9_fht8x8_msa, &vp9_iht8x8_64_add_msa, 2, VPX_BITS_8),
-        make_tuple(&vp9_fht8x8_msa, &vp9_iht8x8_64_add_msa, 3, VPX_BITS_8)));
-#endif  // HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 }  // namespace
--- a/test/frame_size_tests.cc
+++ b/test/frame_size_tests.cc
@@ -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/idct8x8_test.cc
+++ b/test/idct8x8_test.cc
@@ -14,7 +14,8 @@

 #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"

@@ -67,6 +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;
@@ -87,7 +125,7 @@ TEST(VP9Idct8x8Test, AccuracyCheck) {
    reference_dct_2d(input, output_r);
    for (int j = 0; j < 64; ++j)
      coeff[j] = round(output_r[j]);
-    vpx_idct8x8_64_add_c(coeff, dst, 8);
+    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
@@ -10,11 +10,10 @@

 #include "./vpx_config.h"
 #include "./vp8_rtcd.h"
-
-#include "third_party/googletest/src/include/gtest/gtest.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,
@@ -114,8 +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
@@ -0,0 +1,396 @@
+/*
+ *  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 <string.h>
+#include "test/acm_random.h"
+#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 "vp8/common/blockd.h"
+#include "vpx_mem/vpx_mem.h"
+
+namespace {
+
+using libvpx_test::ACMRandom;
+
+class IntraPredBase {
+ public:
+  virtual ~IntraPredBase() { libvpx_test::ClearSystemState(); }
+
+ 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;
+  }
+
+  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();
+    }
+  }
+
+  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_;
+};
+
+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:
+  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() {
+    pred_fn_ = GetParam();
+    SetupMacroblock(mb_, mi_, data_array_, kBlockSize, kStride, 1);
+  }
+
+  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));
+  }
+
+  IntraPredYFunc pred_fn_;
+  static uint8_t* data_array_;
+  static MACROBLOCKD * mb_;
+  static MODE_INFO *mi_;
+};
+
+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
+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
+
+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);
+
+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
+
+}  // namespace
--- a/test/invalid_file_test.cc
+++ b/test/invalid_file_test.cc
@@ -63,22 +63,9 @@ class InvalidFileTest
    EXPECT_NE(res, EOF) << "Read result data failed";

    // Check results match.
-    const DecodeParam input = GET_PARAM(1);
-    if (input.threads > 1) {
-      // The serial decode check is too strict for tile-threaded decoding as
-      // there is no guarantee on the decode order nor which specific error
-      // will take precedence. Currently a tile-level error is not forwarded so
-      // the frame will simply be marked corrupt.
-      EXPECT_TRUE(res_dec == expected_res_dec ||
-                  res_dec == VPX_CODEC_CORRUPT_FRAME)
-          << "Results don't match: frame number = " << video.frame_number()
-          << ". (" << decoder->DecodeError() << "). Expected: "
-          << expected_res_dec << " or " << VPX_CODEC_CORRUPT_FRAME;
-    } else {
-      EXPECT_EQ(expected_res_dec, res_dec)
-          << "Results don't match: frame number = " << video.frame_number()
-          << ". (" << decoder->DecodeError() << ")";
-    }
+    EXPECT_EQ(expected_res_dec, res_dec)
+        << "Results don't match: frame number = " << video.frame_number()
+        << ". (" << decoder->DecodeError() << ")";

    return !HasFailure();
  }
@@ -125,9 +112,7 @@ TEST_P(InvalidFileTest, ReturnCode) {

 const DecodeParam kVP9InvalidFileTests[] = {
  {1, "invalid-vp90-02-v2.webm"},
-#if CONFIG_VP9_HIGHBITDEPTH
  {1, "invalid-vp90-2-00-quantizer-00.webm.ivf.s5861_r01-05_b6-.v2.ivf"},
-#endif
  {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"},
@@ -158,7 +143,7 @@ TEST_P(InvalidFileInvalidPeekTest, ReturnCode) {
 }

 const DecodeParam kVP9InvalidFileInvalidPeekTests[] = {
-  {1, "invalid-vp90-01-v3.webm"},
+  {1, "invalid-vp90-01-v2.webm"},
 };

 VP9_INSTANTIATE_TEST_CASE(InvalidFileInvalidPeekTest,
--- a/test/lpf_8_test.cc
+++ b/test/lpf_8_test.cc
@@ -13,13 +13,13 @@
 #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 "./vpx_config.h"
+#include "./vp9_rtcd.h"
 #include "vp9/common/vp9_entropy.h"
 #include "vp9/common/vp9_loopfilter.h"
 #include "vpx/vpx_integer.h"
@@ -60,49 +60,49 @@ typedef std::tr1::tuple<dual_loop_op_t, dual_loop_op_t, int> dualloop8_param_t;
 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) {
-  vpx_highbd_lpf_vertical_16_sse2(s, p, blimit, limit, thresh, 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) {
-  vpx_highbd_lpf_vertical_16_c(s, p, blimit, limit, thresh, 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) {
-  vpx_highbd_lpf_vertical_16_dual_sse2(s, p, blimit, limit, thresh, 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) {
-  vpx_highbd_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh, 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) {
-  vpx_lpf_vertical_16_sse2(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_c(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_dual_sse2(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
+  vp9_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
 }
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif  // HAVE_SSE2
@@ -114,43 +114,29 @@ void wrapper_vertical_16_dual_c(uint8_t *s, int p, const uint8_t *blimit,
 void wrapper_vertical_16_neon(uint8_t *s, int p, const uint8_t *blimit,
                              const uint8_t *limit, const uint8_t *thresh,
                              int count) {
-  vpx_lpf_vertical_16_neon(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_c(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_dual_neon(s, p, blimit, limit, thresh);
+  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) {
-  vpx_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
+  vp9_lpf_vertical_16_dual_c(s, p, blimit, limit, thresh);
 }
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif  // HAVE_NEON_ASM

-#if HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
-void wrapper_vertical_16_msa(uint8_t *s, int p, const uint8_t *blimit,
-                             const uint8_t *limit, const uint8_t *thresh,
-                             int count) {
-  vpx_lpf_vertical_16_msa(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) {
-  vpx_lpf_vertical_16_c(s, p, blimit, limit, thresh);
-}
-#endif  // HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
-
 class Loop8Test6Param : public ::testing::TestWithParam<loop8_param_t> {
 public:
  virtual ~Loop8Test6Param() {}
@@ -196,11 +182,11 @@ TEST_P(Loop8Test6Param, OperationCheck) {
  const int count_test_block = number_of_iterations;
 #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]);
+  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;
@@ -281,11 +267,11 @@ TEST_P(Loop8Test6Param, ValueCheck) {
  const int count_test_block = number_of_iterations;
 #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]);
+  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;
@@ -352,11 +338,11 @@ TEST_P(Loop8Test9Param, OperationCheck) {
  const int count_test_block = number_of_iterations;
 #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]);
+  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;
@@ -454,11 +440,11 @@ TEST_P(Loop8Test9Param, ValueCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
  const int count_test_block = number_of_iterations;
 #if CONFIG_VP9_HIGHBITDEPTH
-  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]);
+  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;
@@ -534,46 +520,46 @@ using std::tr1::make_tuple;
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 8, 1),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 8, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 8, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 8, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 8, 2),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 8, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 8, 1),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 8, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 8, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 8, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 8, 2),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 8, 1),
        make_tuple(&wrapper_vertical_16_sse2,
                   &wrapper_vertical_16_c, 8, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 10, 1),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 10, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 10, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 10, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 10, 2),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 10, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 10, 1),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 10, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 10, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 10, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 10, 2),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 10, 1),
        make_tuple(&wrapper_vertical_16_sse2,
                   &wrapper_vertical_16_c, 10, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
-                   &vpx_highbd_lpf_horizontal_4_c, 12, 1),
-        make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
-                   &vpx_highbd_lpf_vertical_4_c, 12, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
-                   &vpx_highbd_lpf_horizontal_8_c, 12, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 12, 1),
-        make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
-                   &vpx_highbd_lpf_horizontal_16_c, 12, 2),
-        make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
-                   &vpx_highbd_lpf_vertical_8_c, 12, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_4_sse2,
+                   &vp9_highbd_lpf_horizontal_4_c, 12, 1),
+        make_tuple(&vp9_highbd_lpf_vertical_4_sse2,
+                   &vp9_highbd_lpf_vertical_4_c, 12, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_8_sse2,
+                   &vp9_highbd_lpf_horizontal_8_c, 12, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 12, 1),
+        make_tuple(&vp9_highbd_lpf_horizontal_16_sse2,
+                   &vp9_highbd_lpf_horizontal_16_c, 12, 2),
+        make_tuple(&vp9_highbd_lpf_vertical_8_sse2,
+                   &vp9_highbd_lpf_vertical_8_c, 12, 1),
        make_tuple(&wrapper_vertical_16_sse2,
                   &wrapper_vertical_16_c, 12, 1),
        make_tuple(&wrapper_vertical_16_dual_sse2,
@@ -586,13 +572,11 @@ INSTANTIATE_TEST_CASE_P(
 INSTANTIATE_TEST_CASE_P(
    SSE2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_8_sse2, &vpx_lpf_horizontal_8_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_sse2, &vpx_lpf_horizontal_16_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_sse2, &vpx_lpf_horizontal_16_c, 8, 2),
-        make_tuple(&vpx_lpf_vertical_8_sse2, &vpx_lpf_vertical_8_c, 8, 1),
-        make_tuple(&wrapper_vertical_16_sse2, &wrapper_vertical_16_c, 8, 1),
-        make_tuple(&wrapper_vertical_16_dual_sse2,
-                   &wrapper_vertical_16_dual_c, 8, 1)));
+        make_tuple(&vp9_lpf_horizontal_8_sse2, &vp9_lpf_horizontal_8_c, 8, 1),
+        make_tuple(&vp9_lpf_horizontal_16_sse2, &vp9_lpf_horizontal_16_c, 8, 1),
+        make_tuple(&vp9_lpf_horizontal_16_sse2, &vp9_lpf_horizontal_16_c, 8, 2),
+        make_tuple(&vp9_lpf_vertical_8_sse2, &vp9_lpf_vertical_8_c, 8, 1),
+        make_tuple(&wrapper_vertical_16_sse2, &wrapper_vertical_16_c, 8, 1)));
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif

@@ -600,8 +584,8 @@ INSTANTIATE_TEST_CASE_P(
 INSTANTIATE_TEST_CASE_P(
    AVX2, Loop8Test6Param,
    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_16_avx2, &vpx_lpf_horizontal_16_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_avx2, &vpx_lpf_horizontal_16_c, 8,
+        make_tuple(&vp9_lpf_horizontal_16_avx2, &vp9_lpf_horizontal_16_c, 8, 1),
+        make_tuple(&vp9_lpf_horizontal_16_avx2, &vp9_lpf_horizontal_16_c, 8,
                   2)));
 #endif

@@ -610,42 +594,42 @@ INSTANTIATE_TEST_CASE_P(
 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)));
+        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

@@ -659,60 +643,37 @@ 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_16_neon,
-                   &vpx_lpf_horizontal_16_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_neon,
-                   &vpx_lpf_horizontal_16_c, 8, 2),
+        make_tuple(&vp9_lpf_horizontal_16_neon,
+                   &vp9_lpf_horizontal_16_c, 8, 1),
+        make_tuple(&vp9_lpf_horizontal_16_neon,
+                   &vp9_lpf_horizontal_16_c, 8, 2),
        make_tuple(&wrapper_vertical_16_neon,
                   &wrapper_vertical_16_c, 8, 1),
        make_tuple(&wrapper_vertical_16_dual_neon,
                   &wrapper_vertical_16_dual_c, 8, 1),
+        make_tuple(&vp9_lpf_horizontal_8_neon,
+                   &vp9_lpf_horizontal_8_c, 8, 1),
+        make_tuple(&vp9_lpf_vertical_8_neon,
+                   &vp9_lpf_vertical_8_c, 8, 1),
 #endif  // HAVE_NEON_ASM
-        make_tuple(&vpx_lpf_horizontal_8_neon,
-                   &vpx_lpf_horizontal_8_c, 8, 1),
-        make_tuple(&vpx_lpf_vertical_8_neon,
-                   &vpx_lpf_vertical_8_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_4_neon,
-                   &vpx_lpf_horizontal_4_c, 8, 1),
-        make_tuple(&vpx_lpf_vertical_4_neon,
-                   &vpx_lpf_vertical_4_c, 8, 1)));
+        make_tuple(&vp9_lpf_horizontal_4_neon,
+                   &vp9_lpf_horizontal_4_c, 8, 1),
+        make_tuple(&vp9_lpf_vertical_4_neon,
+                   &vp9_lpf_vertical_4_c, 8, 1)));
 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)));
+        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_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
-INSTANTIATE_TEST_CASE_P(
-    MSA, Loop8Test6Param,
-    ::testing::Values(
-        make_tuple(&vpx_lpf_horizontal_8_msa, &vpx_lpf_horizontal_8_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_msa, &vpx_lpf_horizontal_16_c, 8, 1),
-        make_tuple(&vpx_lpf_horizontal_16_msa, &vpx_lpf_horizontal_16_c, 8, 2),
-        make_tuple(&vpx_lpf_vertical_8_msa, &vpx_lpf_vertical_8_c, 8, 1),
-        make_tuple(&wrapper_vertical_16_msa, &wrapper_vertical_16_c, 8, 1)));
-
-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_VP9_HIGHBITDEPTH)
-
 }  // 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/partial_idct_test.cc
+++ b/test/partial_idct_test.cc
@@ -13,13 +13,12 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./vp9_rtcd.h"
 #include "vp9/common/vp9_blockd.h"
 #include "vp9/common/vp9_scan.h"
 #include "vpx/vpx_integer.h"
@@ -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) {
@@ -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;
@@ -202,62 +201,62 @@ 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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
 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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE

@@ -265,79 +264,45 @@ INSTANTIATE_TEST_CASE_P(
 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 && CONFIG_USE_X86INC && ARCH_X86_64 && \
-    !CONFIG_VP9_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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-INSTANTIATE_TEST_CASE_P(
-    MSA, 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_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
-
 }  // namespace
--- a/test/pp_filter_test.cc
+++ b/test/pp_filter_test.cc
@@ -110,9 +110,4 @@ INSTANTIATE_TEST_CASE_P(SSE2, VP8PostProcessingFilterTest,
    ::testing::Values(vp8_post_proc_down_and_across_mb_row_sse2));
 #endif

-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(MSA, VP8PostProcessingFilterTest,
-    ::testing::Values(vp8_post_proc_down_and_across_mb_row_msa));
-#endif
-
 }  // namespace
--- a/test/quantize_test.cc
+++ b/test/quantize_test.cc
@@ -11,13 +11,13 @@
 #include <string.h>

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "./vp8_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_config.h"
+#include "./vp8_rtcd.h"
 #include "vp8/common/blockd.h"
 #include "vp8/common/onyx.h"
 #include "vp8/encoder/block.h"
@@ -192,12 +192,4 @@ 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/register_state_check.h
+++ b/test/register_state_check.h
@@ -30,9 +30,7 @@

 #if defined(_WIN64)

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

@@ -98,7 +96,7 @@ class RegisterStateCheck {

 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 {
@@ -113,7 +111,7 @@ class RegisterStateCheck {

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

@@ -121,7 +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
@@ -81,15 +81,6 @@ 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;
-};
-
 unsigned int ScaleForFrameNumber(unsigned int frame, unsigned int val) {
  if (frame < 10)
    return val;
@@ -129,6 +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));
@@ -196,27 +196,13 @@ 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);
    }
  }

@@ -241,7 +227,6 @@ class ResizeInternalTest : public ResizeTest {
 #endif

  double frame0_psnr_;
-  bool change_config_;
 #if WRITE_COMPRESSED_STREAM
  FILE *outfile_;
  unsigned int out_frames_;
@@ -252,7 +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;
@@ -277,143 +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 ResizeInternalRealtimeTest : public ::libvpx_test::EncoderTest,
-  public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
- protected:
-  ResizeInternalRealtimeTest() : EncoderTest(GET_PARAM(0)) {}
-  virtual ~ResizeInternalRealtimeTest() {}
-
-  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));
-  }
-
-  void DefaultConfig() {
-    cfg_.g_w = 352;
-    cfg_.g_h = 288;
-    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_;
-};
-
-// 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(ResizeInternalRealtimeTest, TestInternalResizeDown) {
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 299);
-  DefaultConfig();
-  change_bitrate_ = false;
-  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++;
-    }
-  }
-
-  // Verify that we get 1 resize down event in this test.
-  ASSERT_EQ(1, resize_count) << "Resizing should occur.";
-}
-
-// 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(ResizeInternalRealtimeTest, TestInternalResizeDownUpChangeBitRate) {
-  ::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
-                                       30, 1, 0, 299);
-  DefaultConfig();
-  change_bitrate_ = true;
-  // Disable dropped frames.
-  cfg_.rc_dropframe_thresh = 0;
-  // Starting bitrate low.
-  cfg_.rc_target_bitrate = 100;
-  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;
-    }
-  }
-
-  // Verify that we get 2 resize events in this test.
-  ASSERT_EQ(2, resize_count) << "Resizing should occur twice.";
-}
-
 vpx_img_fmt_t CspForFrameNumber(int frame) {
  if (frame < 10)
    return VPX_IMG_FMT_I420;
@@ -524,9 +371,6 @@ VP9_INSTANTIATE_TEST_CASE(ResizeTest,
                          ::testing::Values(::libvpx_test::kRealTime));
 VP9_INSTANTIATE_TEST_CASE(ResizeInternalTest,
                          ::testing::Values(::libvpx_test::kOnePassBest));
-VP9_INSTANTIATE_TEST_CASE(ResizeInternalRealtimeTest,
-                          ::testing::Values(::libvpx_test::kRealTime),
-                          ::testing::Range(5, 9));
 VP9_INSTANTIATE_TEST_CASE(ResizeCspTest,
                          ::testing::Values(::libvpx_test::kRealTime));
 }  // namespace
--- a/test/sad_test.cc
+++ b/test/sad_test.cc
--- a/test/sixtap_predict_test.cc
+++ b/test/sixtap_predict_test.cc
@@ -11,15 +11,13 @@
 #include <math.h>
 #include <stdlib.h>
 #include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_config.h"
-#include "./vp8_rtcd.h"
 #include "test/acm_random.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
 #include "test/util.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"

@@ -201,7 +199,7 @@ const SixtapPredictFunc sixtap_16x16_neon = vp8_sixtap_predict16x16_neon;
 const SixtapPredictFunc sixtap_8x8_neon = vp8_sixtap_predict8x8_neon;
 const SixtapPredictFunc sixtap_8x4_neon = vp8_sixtap_predict8x4_neon;
 INSTANTIATE_TEST_CASE_P(
-    NEON, SixtapPredictTest, ::testing::Values(
+    DISABLED_NEON, SixtapPredictTest, ::testing::Values(
        make_tuple(16, 16, sixtap_16x16_neon),
        make_tuple(8, 8, sixtap_8x8_neon),
        make_tuple(8, 4, sixtap_8x4_neon)));
@@ -240,16 +238,4 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(8, 4, sixtap_8x4_ssse3),
        make_tuple(4, 4, sixtap_4x4_ssse3)));
 #endif
-#if HAVE_MSA
-const SixtapPredictFunc sixtap_16x16_msa = vp8_sixtap_predict16x16_msa;
-const SixtapPredictFunc sixtap_8x8_msa = vp8_sixtap_predict8x8_msa;
-const SixtapPredictFunc sixtap_8x4_msa = vp8_sixtap_predict8x4_msa;
-const SixtapPredictFunc sixtap_4x4_msa = vp8_sixtap_predict4x4_msa;
-INSTANTIATE_TEST_CASE_P(
-    MSA, SixtapPredictTest, ::testing::Values(
-        make_tuple(16, 16, sixtap_16x16_msa),
-        make_tuple(8, 8, sixtap_8x8_msa),
-        make_tuple(8, 4, sixtap_8x4_msa),
-        make_tuple(4, 4, sixtap_4x4_msa)));
-#endif
 }  // namespace
--- a/test/subtract_test.cc
+++ b/test/subtract_test.cc
@@ -0,0 +1,123 @@
+/*
+ *  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/acm_random.h"
+#include "test/clear_system_state.h"
+#include "test/register_state_check.h"
+#include "./vpx_config.h"
+#include "./vp8_rtcd.h"
+#include "vp8/common/blockd.h"
+#include "vp8/encoder/block.h"
+#include "vpx_mem/vpx_mem.h"
+
+typedef void (*SubtractBlockFunc)(BLOCK *be, BLOCKD *bd, int pitch);
+
+namespace {
+
+class SubtractBlockTest : public ::testing::TestWithParam<SubtractBlockFunc> {
+ public:
+  virtual void TearDown() {
+    libvpx_test::ClearSystemState();
+  }
+};
+
+using libvpx_test::ACMRandom;
+
+TEST_P(SubtractBlockTest, SimpleSubtract) {
+  ACMRandom rnd(ACMRandom::DeterministicSeed());
+  BLOCK be;
+  BLOCKD bd;
+  // in libvpx, this stride is always 16
+  const int kDiffPredStride = 16;
+  const int kSrcStride[] = {32, 16, 8, 4, 0};
+  const int kBlockWidth = 4;
+  const int kBlockHeight = 4;
+
+  // Allocate... align to 16 for mmx/sse tests
+  uint8_t *source = reinterpret_cast<uint8_t*>(
+      vpx_memalign(16, kBlockHeight * kSrcStride[0] * sizeof(*source)));
+  be.src_diff = reinterpret_cast<int16_t*>(
+      vpx_memalign(16, kBlockHeight * kDiffPredStride * sizeof(*be.src_diff)));
+  bd.predictor = reinterpret_cast<unsigned char*>(
+      vpx_memalign(16, kBlockHeight * kDiffPredStride * sizeof(*bd.predictor)));
+
+  for (int i = 0; kSrcStride[i] > 0; ++i) {
+    // start at block0
+    be.src = 0;
+    be.base_src = &source;
+    be.src_stride = kSrcStride[i];
+
+    // set difference
+    int16_t *src_diff = be.src_diff;
+    for (int r = 0; r < kBlockHeight; ++r) {
+      for (int c = 0; c < kBlockWidth; ++c) {
+        src_diff[c] = static_cast<int16_t>(0xa5a5u);
+      }
+      src_diff += kDiffPredStride;
+    }
+
+    // set destination
+    uint8_t *base_src = *be.base_src;
+    for (int r = 0; r < kBlockHeight; ++r) {
+      for (int c = 0; c < kBlockWidth; ++c) {
+        base_src[c] = rnd.Rand8();
+      }
+      base_src += be.src_stride;
+    }
+
+    // set predictor
+    uint8_t *predictor = bd.predictor;
+    for (int r = 0; r < kBlockHeight; ++r) {
+      for (int c = 0; c < kBlockWidth; ++c) {
+        predictor[c] = rnd.Rand8();
+      }
+      predictor += kDiffPredStride;
+    }
+
+    ASM_REGISTER_STATE_CHECK(GetParam()(&be, &bd, kDiffPredStride));
+
+    base_src = *be.base_src;
+    src_diff = be.src_diff;
+    predictor = bd.predictor;
+    for (int r = 0; r < kBlockHeight; ++r) {
+      for (int c = 0; c < kBlockWidth; ++c) {
+        EXPECT_EQ(base_src[c], (src_diff[c] + predictor[c])) << "r = " << r
+                                                             << ", c = " << c;
+      }
+      src_diff += kDiffPredStride;
+      predictor += kDiffPredStride;
+      base_src += be.src_stride;
+    }
+  }
+  vpx_free(be.src_diff);
+  vpx_free(source);
+  vpx_free(bd.predictor);
+}
+
+INSTANTIATE_TEST_CASE_P(C, SubtractBlockTest,
+                        ::testing::Values(vp8_subtract_b_c));
+
+#if HAVE_NEON
+INSTANTIATE_TEST_CASE_P(NEON, SubtractBlockTest,
+                        ::testing::Values(vp8_subtract_b_neon));
+#endif
+
+#if HAVE_MMX
+INSTANTIATE_TEST_CASE_P(MMX, SubtractBlockTest,
+                        ::testing::Values(vp8_subtract_b_mmx));
+#endif
+
+#if HAVE_SSE2
+INSTANTIATE_TEST_CASE_P(SSE2, SubtractBlockTest,
+                        ::testing::Values(vp8_subtract_b_sse2));
+#endif
+
+}  // namespace
--- a/test/superframe_test.cc
+++ b/test/superframe_test.cc
@@ -16,13 +16,8 @@

 namespace {

-const int kTestMode = 0;
-const int kSuperframeSyntax = 1;
-
-typedef std::tr1::tuple<libvpx_test::TestMode,int> SuperframeTestParam;
-
 class SuperframeTest : public ::libvpx_test::EncoderTest,
-    public ::libvpx_test::CodecTestWithParam<SuperframeTestParam> {
+    public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
 protected:
  SuperframeTest() : EncoderTest(GET_PARAM(0)), modified_buf_(NULL),
      last_sf_pts_(0) {}
@@ -30,13 +25,9 @@ class SuperframeTest : public ::libvpx_test::EncoderTest,

  virtual void SetUp() {
    InitializeConfig();
-    const SuperframeTestParam input = GET_PARAM(1);
-    const libvpx_test::TestMode mode = std::tr1::get<kTestMode>(input);
-    const int syntax = std::tr1::get<kSuperframeSyntax>(input);
-    SetMode(mode);
+    SetMode(GET_PARAM(1));
    sf_count_ = 0;
    sf_count_max_ = INT_MAX;
-    is_vp10_style_superframe_ = syntax;
  }

  virtual void TearDown() {
@@ -59,8 +50,7 @@ class SuperframeTest : public ::libvpx_test::EncoderTest,
    const uint8_t marker = buffer[pkt->data.frame.sz - 1];
    const int frames = (marker & 0x7) + 1;
    const int mag = ((marker >> 3) & 3) + 1;
-    const unsigned int index_sz =
-        2 + mag * (frames - is_vp10_style_superframe_);
+    const unsigned int index_sz = 2 + mag  * frames;
    if ((marker & 0xe0) == 0xc0 &&
        pkt->data.frame.sz >= index_sz &&
        buffer[pkt->data.frame.sz - index_sz] == marker) {
@@ -85,7 +75,6 @@ class SuperframeTest : public ::libvpx_test::EncoderTest,
    return pkt;
  }

-  int is_vp10_style_superframe_;
  int sf_count_;
  int sf_count_max_;
  vpx_codec_cx_pkt_t modified_pkt_;
@@ -103,11 +92,6 @@ TEST_P(SuperframeTest, TestSuperframeIndexIsOptional) {
  EXPECT_EQ(sf_count_, 1);
 }

-VP9_INSTANTIATE_TEST_CASE(SuperframeTest, ::testing::Combine(
-    ::testing::Values(::libvpx_test::kTwoPassGood),
-    ::testing::Values(0)));
-
-VP10_INSTANTIATE_TEST_CASE(SuperframeTest, ::testing::Combine(
-    ::testing::Values(::libvpx_test::kTwoPassGood),
-    ::testing::Values(CONFIG_MISC_FIXES)));
+VP9_INSTANTIATE_TEST_CASE(SuperframeTest, ::testing::Values(
+    ::libvpx_test::kTwoPassGood));
 }  // namespace
--- a/test/svc_test.cc
+++ b/test/svc_test.cc
@@ -453,7 +453,6 @@ TEST_F(SvcTest, OnePassEncodeOneFrame) {

 TEST_F(SvcTest, OnePassEncodeThreeFrames) {
  codec_enc_.g_pass = VPX_RC_ONE_PASS;
-  codec_enc_.g_lag_in_frames = 0;
  vpx_fixed_buf outputs[3];
  memset(&outputs[0], 0, sizeof(outputs));
  Pass2EncodeNFrames(NULL, 3, 2, &outputs[0]);
--- a/test/test-data.mk
+++ b/test/test-data.mk
@@ -12,13 +12,11 @@ LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_12_420.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_12_422.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_12_444.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_12_440.yuv
-LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_8_420_a10-1.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_8_420.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_8_422.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_8_444.y4m
 LIBVPX_TEST_DATA-$(CONFIG_ENCODERS) += park_joy_90p_8_440.yuv

-LIBVPX_TEST_DATA-$(CONFIG_VP9_ENCODER) += desktop_credits.y4m
 LIBVPX_TEST_DATA-$(CONFIG_VP9_ENCODER) += niklas_1280_720_30.y4m
 LIBVPX_TEST_DATA-$(CONFIG_VP9_ENCODER) += rush_hour_444.y4m
 LIBVPX_TEST_DATA-$(CONFIG_VP9_ENCODER) += screendata.y4m
@@ -688,8 +686,8 @@ LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += vp93-2-20-12bit-yuv444.webm.md5
 endif  # CONFIG_VP9_HIGHBITDEPTH

 # Invalid files for testing libvpx error checking.
-LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-01-v3.webm
-LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-01-v3.webm.res
+LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-01-v2.webm
+LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-01-v2.webm.res
 LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-02-v2.webm
 LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-02-v2.webm.res
 LIBVPX_TEST_DATA-$(CONFIG_VP9_DECODER) += invalid-vp90-03-v3.webm
--- a/test/test-data.sha1
+++ b/test/test-data.sha1
--- a/test/test.mk
+++ b/test/test.mk
@@ -36,7 +36,6 @@ LIBVPX_TEST_SRCS-$(CONFIG_VP9_DECODER) += external_frame_buffer_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_DECODER) += invalid_file_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_DECODER) += user_priv_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_DECODER) += vp9_frame_parallel_test.cc
-LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += active_map_refresh_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += active_map_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += borders_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += cpu_speed_test.cc
@@ -67,7 +66,6 @@ LIBVPX_TEST_SRCS-$(CONFIG_DECODERS)    += ../tools_common.h
 LIBVPX_TEST_SRCS-$(CONFIG_DECODERS)    += ../webmdec.cc
 LIBVPX_TEST_SRCS-$(CONFIG_DECODERS)    += ../webmdec.h
 LIBVPX_TEST_SRCS-$(CONFIG_DECODERS)    += webm_video_source.h
-LIBVPX_TEST_SRCS-$(CONFIG_VP9_DECODER) += vp9_skip_loopfilter_test.cc
 endif

 LIBVPX_TEST_SRCS-$(CONFIG_DECODERS)    += decode_api_test.cc
@@ -92,7 +90,6 @@ endif
 ## shared library builds don't make these functions accessible.
 ##
 ifeq ($(CONFIG_SHARED),)
-LIBVPX_TEST_SRCS-$(CONFIG_VP9)         += lpf_8_test.cc

 ## VP8
 ifneq ($(CONFIG_VP8_ENCODER)$(CONFIG_VP8_DECODER),)
@@ -106,11 +103,13 @@ endif
 LIBVPX_TEST_SRCS-$(CONFIG_POSTPROC)    += pp_filter_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP8_DECODER) += vp8_decrypt_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP8_ENCODER) += set_roi.cc
+LIBVPX_TEST_SRCS-$(CONFIG_VP8_ENCODER) += subtract_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP8_ENCODER) += variance_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP8_ENCODER) += vp8_fdct4x4_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP8_ENCODER) += quantize_test.cc

 LIBVPX_TEST_SRCS-yes                   += idct_test.cc
+LIBVPX_TEST_SRCS-yes                   += intrapred_test.cc
 LIBVPX_TEST_SRCS-yes                   += sixtap_predict_test.cc
 LIBVPX_TEST_SRCS-yes                   += vpx_scale_test.cc

@@ -143,6 +142,7 @@ LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += fdct4x4_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += fdct8x8_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += variance_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += vp9_subtract_test.cc
+LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += lpf_8_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += vp9_avg_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += vp9_error_block_test.cc
 LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += vp9_quantize_test.cc
@@ -158,19 +158,11 @@ endif
 ifeq ($(CONFIG_VP9_ENCODER)$(CONFIG_VP9_TEMPORAL_DENOISING),yesyes)
 LIBVPX_TEST_SRCS-$(HAVE_SSE2) += vp9_denoiser_sse2_test.cc
 endif
-LIBVPX_TEST_SRCS-$(CONFIG_VP9_ENCODER) += vp9_arf_freq_test.cc

 endif # VP9

 LIBVPX_TEST_SRCS-$(CONFIG_ENCODERS)    += sad_test.cc

-TEST_INTRA_PRED_SPEED_SRCS-$(CONFIG_VP9) := test_intra_pred_speed.cc
-TEST_INTRA_PRED_SPEED_SRCS-$(CONFIG_VP9) += ../md5_utils.h ../md5_utils.c
-
-## VP10
-LIBVPX_TEST_SRCS-$(CONFIG_VP10_ENCODER) += vp10_dct_test.cc
-LIBVPX_TEST_SRCS-$(CONFIG_VP10) += vp10_inv_txfm_test.cc
-
 endif # CONFIG_SHARED

 include $(SRC_PATH_BARE)/test/test-data.mk
--- a/test/test_intra_pred_speed.cc
+++ b/test/test_intra_pred_speed.cc
@@ -1,384 +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.
- */
-//  Test and time VPX intra-predictor functions
-
-#include <stdio.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vpx_dsp_rtcd.h"
-#include "test/acm_random.h"
-#include "test/clear_system_state.h"
-#include "test/md5_helper.h"
-#include "vpx/vpx_integer.h"
-#include "vpx_ports/mem.h"
-#include "vpx_ports/vpx_timer.h"
-
-// -----------------------------------------------------------------------------
-
-namespace {
-
-typedef void (*VpxPredFunc)(uint8_t *dst, ptrdiff_t y_stride,
-                            const uint8_t *above, const uint8_t *left);
-
-const int kNumVp9IntraPredFuncs = 13;
-const char *kVp9IntraPredNames[kNumVp9IntraPredFuncs] = {
-  "DC_PRED", "DC_LEFT_PRED", "DC_TOP_PRED", "DC_128_PRED", "V_PRED", "H_PRED",
-  "D45_PRED", "D135_PRED", "D117_PRED", "D153_PRED", "D207_PRED", "D63_PRED",
-  "TM_PRED"
-};
-
-void TestIntraPred(const char name[], VpxPredFunc const *pred_funcs,
-                   const char *const pred_func_names[], int num_funcs,
-                   const char *const signatures[], int block_size,
-                   int num_pixels_per_test) {
-  libvpx_test::ACMRandom rnd(libvpx_test::ACMRandom::DeterministicSeed());
-  const int kBPS = 32;
-  const int kTotalPixels = 32 * kBPS;
-  DECLARE_ALIGNED(16, uint8_t, src[kTotalPixels]);
-  DECLARE_ALIGNED(16, uint8_t, ref_src[kTotalPixels]);
-  DECLARE_ALIGNED(16, uint8_t, left[kBPS]);
-  DECLARE_ALIGNED(16, uint8_t, above_mem[2 * kBPS + 16]);
-  uint8_t *const above = above_mem + 16;
-  for (int i = 0; i < kTotalPixels; ++i) ref_src[i] = rnd.Rand8();
-  for (int i = 0; i < kBPS; ++i) left[i] = rnd.Rand8();
-  for (int i = -1; i < kBPS; ++i) above[i] = rnd.Rand8();
-  const int kNumTests = static_cast<int>(2.e10 / num_pixels_per_test);
-
-  // some code assumes the top row has been extended:
-  // d45/d63 C-code, for instance, but not the assembly.
-  // TODO(jzern): this style of extension isn't strictly necessary.
-  ASSERT_LE(block_size, kBPS);
-  memset(above + block_size, above[block_size - 1], 2 * kBPS - block_size);
-
-  for (int k = 0; k < num_funcs; ++k) {
-    if (pred_funcs[k] == NULL) continue;
-    memcpy(src, ref_src, sizeof(src));
-    vpx_usec_timer timer;
-    vpx_usec_timer_start(&timer);
-    for (int num_tests = 0; num_tests < kNumTests; ++num_tests) {
-      pred_funcs[k](src, kBPS, above, left);
-    }
-    libvpx_test::ClearSystemState();
-    vpx_usec_timer_mark(&timer);
-    const int elapsed_time =
-        static_cast<int>(vpx_usec_timer_elapsed(&timer) / 1000);
-    libvpx_test::MD5 md5;
-    md5.Add(src, sizeof(src));
-    printf("Mode %s[%12s]: %5d ms     MD5: %s\n", name, pred_func_names[k],
-           elapsed_time, md5.Get());
-    EXPECT_STREQ(signatures[k], md5.Get());
-  }
-}
-
-void TestIntraPred4(VpxPredFunc const *pred_funcs) {
-  static const int kNumVp9IntraFuncs = 13;
-  static const char *const kSignatures[kNumVp9IntraFuncs] = {
-    "4334156168b34ab599d9b5b30f522fe9",
-    "bc4649d5ba47c7ff178d92e475960fb0",
-    "8d316e5933326dcac24e1064794b5d12",
-    "a27270fed024eafd762c95de85f4da51",
-    "c33dff000d4256c2b8f3bf9e9bab14d2",
-    "44d8cddc2ad8f79b8ed3306051722b4f",
-    "eb54839b2bad6699d8946f01ec041cd0",
-    "ecb0d56ae5f677ea45127ce9d5c058e4",
-    "0b7936841f6813da818275944895b574",
-    "9117972ef64f91a58ff73e1731c81db2",
-    "c56d5e8c729e46825f46dd5d3b5d508a",
-    "c0889e2039bcf7bcb5d2f33cdca69adc",
-    "309a618577b27c648f9c5ee45252bc8f",
-  };
-  TestIntraPred("Intra4", pred_funcs, kVp9IntraPredNames, kNumVp9IntraFuncs,
-                kSignatures, 4, 4 * 4 * kNumVp9IntraFuncs);
-}
-
-void TestIntraPred8(VpxPredFunc const *pred_funcs) {
-  static const int kNumVp9IntraFuncs = 13;
-  static const char *const kSignatures[kNumVp9IntraFuncs] = {
-    "7694ddeeefed887faf9d339d18850928",
-    "7d726b1213591b99f736be6dec65065b",
-    "19c5711281357a485591aaf9c96c0a67",
-    "ba6b66877a089e71cd938e3b8c40caac",
-    "802440c93317e0f8ba93fab02ef74265",
-    "9e09a47a15deb0b9d8372824f9805080",
-    "b7c2d8c662268c0c427da412d7b0311d",
-    "78339c1c60bb1d67d248ab8c4da08b7f",
-    "5c97d70f7d47de1882a6cd86c165c8a9",
-    "8182bf60688b42205acd95e59e967157",
-    "08323400005a297f16d7e57e7fe1eaac",
-    "95f7bfc262329a5849eda66d8f7c68ce",
-    "815b75c8e0d91cc1ae766dc5d3e445a3",
-  };
-  TestIntraPred("Intra8", pred_funcs, kVp9IntraPredNames, kNumVp9IntraFuncs,
-                kSignatures, 8, 8 * 8 * kNumVp9IntraFuncs);
-}
-
-void TestIntraPred16(VpxPredFunc const *pred_funcs) {
-  static const int kNumVp9IntraFuncs = 13;
-  static const char *const kSignatures[kNumVp9IntraFuncs] = {
-    "b40dbb555d5d16a043dc361e6694fe53",
-    "fb08118cee3b6405d64c1fd68be878c6",
-    "6c190f341475c837cc38c2e566b64875",
-    "db5c34ccbe2c7f595d9b08b0dc2c698c",
-    "a62cbfd153a1f0b9fed13e62b8408a7a",
-    "143df5b4c89335e281103f610f5052e4",
-    "d87feb124107cdf2cfb147655aa0bb3c",
-    "7841fae7d4d47b519322e6a03eeed9dc",
-    "f6ebed3f71cbcf8d6d0516ce87e11093",
-    "3cc480297dbfeed01a1c2d78dd03d0c5",
-    "b9f69fa6532b372c545397dcb78ef311",
-    "a8fe1c70432f09d0c20c67bdb6432c4d",
-    "b8a41aa968ec108af447af4217cba91b",
-  };
-  TestIntraPred("Intra16", pred_funcs, kVp9IntraPredNames, kNumVp9IntraFuncs,
-                kSignatures, 16, 16 * 16 * kNumVp9IntraFuncs);
-}
-
-void TestIntraPred32(VpxPredFunc const *pred_funcs) {
-  static const int kNumVp9IntraFuncs = 13;
-  static const char *const kSignatures[kNumVp9IntraFuncs] = {
-    "558541656d84f9ae7896db655826febe",
-    "b3587a1f9a01495fa38c8cd3c8e2a1bf",
-    "4c6501e64f25aacc55a2a16c7e8f0255",
-    "b3b01379ba08916ef6b1b35f7d9ad51c",
-    "0f1eb38b6cbddb3d496199ef9f329071",
-    "911c06efb9ed1c3b4c104b232b55812f",
-    "9225beb0ddfa7a1d24eaa1be430a6654",
-    "0a6d584a44f8db9aa7ade2e2fdb9fc9e",
-    "b01c9076525216925f3456f034fb6eee",
-    "d267e20ad9e5cd2915d1a47254d3d149",
-    "ed012a4a5da71f36c2393023184a0e59",
-    "f162b51ed618d28b936974cff4391da5",
-    "9e1370c6d42e08d357d9612c93a71cfc",
-  };
-  TestIntraPred("Intra32", pred_funcs, kVp9IntraPredNames, kNumVp9IntraFuncs,
-                kSignatures, 32, 32 * 32 * kNumVp9IntraFuncs);
-}
-
-}  // namespace
-
-// Defines a test case for |arch| (e.g., C, SSE2, ...) passing the predictors
-// to |test_func|. The test name is 'arch.test_func', e.g., C.TestIntraPred4.
-#define INTRA_PRED_TEST(arch, test_func, dc, dc_left, dc_top, dc_128, v, h, \
-                        d45, d135, d117, d153, d207, d63, tm)               \
-  TEST(arch, test_func) {                                                   \
-    static const VpxPredFunc vpx_intra_pred[] = {                           \
-        dc,   dc_left, dc_top, dc_128, v,   h, d45,                         \
-        d135, d117,    d153,   d207,   d63, tm};                            \
-    test_func(vpx_intra_pred);                                              \
-  }
-
-// -----------------------------------------------------------------------------
-// 4x4
-
-INTRA_PRED_TEST(C, TestIntraPred4, vpx_dc_predictor_4x4_c,
-                vpx_dc_left_predictor_4x4_c, vpx_dc_top_predictor_4x4_c,
-                vpx_dc_128_predictor_4x4_c, vpx_v_predictor_4x4_c,
-                vpx_h_predictor_4x4_c, vpx_d45_predictor_4x4_c,
-                vpx_d135_predictor_4x4_c, vpx_d117_predictor_4x4_c,
-                vpx_d153_predictor_4x4_c, vpx_d207_predictor_4x4_c,
-                vpx_d63_predictor_4x4_c, vpx_tm_predictor_4x4_c)
-
-#if HAVE_SSE && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSE, TestIntraPred4, vpx_dc_predictor_4x4_sse,
-                vpx_dc_left_predictor_4x4_sse, vpx_dc_top_predictor_4x4_sse,
-                vpx_dc_128_predictor_4x4_sse, vpx_v_predictor_4x4_sse, NULL,
-                NULL, NULL, NULL, NULL, NULL, NULL, vpx_tm_predictor_4x4_sse)
-#endif  // HAVE_SSE && CONFIG_USE_X86INC
-
-#if HAVE_SSSE3 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSSE3, TestIntraPred4, NULL, NULL, NULL, NULL, NULL,
-                vpx_h_predictor_4x4_ssse3, vpx_d45_predictor_4x4_ssse3, NULL,
-                NULL, vpx_d153_predictor_4x4_ssse3,
-                vpx_d207_predictor_4x4_ssse3, vpx_d63_predictor_4x4_ssse3, NULL)
-#endif  // HAVE_SSSE3 && CONFIG_USE_X86INC
-
-#if HAVE_DSPR2
-INTRA_PRED_TEST(DSPR2, TestIntraPred4, vpx_dc_predictor_4x4_dspr2, NULL, NULL,
-                NULL, NULL, vpx_h_predictor_4x4_dspr2, NULL, NULL, NULL, NULL,
-                NULL, NULL, vpx_tm_predictor_4x4_dspr2)
-#endif  // HAVE_DSPR2
-
-#if HAVE_NEON
-INTRA_PRED_TEST(NEON, TestIntraPred4, vpx_dc_predictor_4x4_neon,
-                vpx_dc_left_predictor_4x4_neon, vpx_dc_top_predictor_4x4_neon,
-                vpx_dc_128_predictor_4x4_neon, vpx_v_predictor_4x4_neon,
-                vpx_h_predictor_4x4_neon, vpx_d45_predictor_4x4_neon,
-                vpx_d135_predictor_4x4_neon, NULL, NULL, NULL, NULL,
-                vpx_tm_predictor_4x4_neon)
-#endif  // HAVE_NEON
-
-#if HAVE_MSA
-INTRA_PRED_TEST(MSA, TestIntraPred4, vpx_dc_predictor_4x4_msa,
-                vpx_dc_left_predictor_4x4_msa, vpx_dc_top_predictor_4x4_msa,
-                vpx_dc_128_predictor_4x4_msa, vpx_v_predictor_4x4_msa,
-                vpx_h_predictor_4x4_msa, NULL, NULL, NULL, NULL, NULL,
-                NULL, vpx_tm_predictor_4x4_msa)
-#endif  // HAVE_MSA
-
-// -----------------------------------------------------------------------------
-// 8x8
-
-INTRA_PRED_TEST(C, TestIntraPred8, vpx_dc_predictor_8x8_c,
-                vpx_dc_left_predictor_8x8_c, vpx_dc_top_predictor_8x8_c,
-                vpx_dc_128_predictor_8x8_c, vpx_v_predictor_8x8_c,
-                vpx_h_predictor_8x8_c, vpx_d45_predictor_8x8_c,
-                vpx_d135_predictor_8x8_c, vpx_d117_predictor_8x8_c,
-                vpx_d153_predictor_8x8_c, vpx_d207_predictor_8x8_c,
-                vpx_d63_predictor_8x8_c, vpx_tm_predictor_8x8_c)
-
-#if HAVE_SSE && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSE, TestIntraPred8, vpx_dc_predictor_8x8_sse,
-                vpx_dc_left_predictor_8x8_sse, vpx_dc_top_predictor_8x8_sse,
-                vpx_dc_128_predictor_8x8_sse, vpx_v_predictor_8x8_sse, NULL,
-                NULL, NULL, NULL, NULL, NULL, NULL, NULL)
-#endif  // HAVE_SSE && CONFIG_USE_X86INC
-
-#if HAVE_SSE2 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSE2, TestIntraPred8, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-                NULL, NULL, NULL, NULL, NULL, vpx_tm_predictor_8x8_sse2)
-#endif  // HAVE_SSE2 && CONFIG_USE_X86INC
-
-#if HAVE_SSSE3 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSSE3, TestIntraPred8, NULL, NULL, NULL, NULL, NULL,
-                vpx_h_predictor_8x8_ssse3, vpx_d45_predictor_8x8_ssse3, NULL,
-                NULL, vpx_d153_predictor_8x8_ssse3,
-                vpx_d207_predictor_8x8_ssse3, vpx_d63_predictor_8x8_ssse3, NULL)
-#endif  // HAVE_SSSE3 && CONFIG_USE_X86INC
-
-#if HAVE_DSPR2
-INTRA_PRED_TEST(DSPR2, TestIntraPred8, vpx_dc_predictor_8x8_dspr2, NULL, NULL,
-                NULL, NULL, vpx_h_predictor_8x8_dspr2, NULL, NULL, NULL, NULL,
-                NULL, NULL, vpx_tm_predictor_8x8_c)
-#endif  // HAVE_DSPR2
-
-#if HAVE_NEON
-INTRA_PRED_TEST(NEON, TestIntraPred8, vpx_dc_predictor_8x8_neon,
-                vpx_dc_left_predictor_8x8_neon, vpx_dc_top_predictor_8x8_neon,
-                vpx_dc_128_predictor_8x8_neon, vpx_v_predictor_8x8_neon,
-                vpx_h_predictor_8x8_neon, vpx_d45_predictor_8x8_neon, NULL,
-                NULL, NULL, NULL, NULL, vpx_tm_predictor_8x8_neon)
-
-#endif  // HAVE_NEON
-
-#if HAVE_MSA
-INTRA_PRED_TEST(MSA, TestIntraPred8, vpx_dc_predictor_8x8_msa,
-                vpx_dc_left_predictor_8x8_msa, vpx_dc_top_predictor_8x8_msa,
-                vpx_dc_128_predictor_8x8_msa, vpx_v_predictor_8x8_msa,
-                vpx_h_predictor_8x8_msa, NULL, NULL, NULL, NULL, NULL,
-                NULL, vpx_tm_predictor_8x8_msa)
-#endif  // HAVE_MSA
-
-// -----------------------------------------------------------------------------
-// 16x16
-
-INTRA_PRED_TEST(C, TestIntraPred16, vpx_dc_predictor_16x16_c,
-                vpx_dc_left_predictor_16x16_c, vpx_dc_top_predictor_16x16_c,
-                vpx_dc_128_predictor_16x16_c, vpx_v_predictor_16x16_c,
-                vpx_h_predictor_16x16_c, vpx_d45_predictor_16x16_c,
-                vpx_d135_predictor_16x16_c, vpx_d117_predictor_16x16_c,
-                vpx_d153_predictor_16x16_c, vpx_d207_predictor_16x16_c,
-                vpx_d63_predictor_16x16_c, vpx_tm_predictor_16x16_c)
-
-#if HAVE_SSE2 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSE2, TestIntraPred16, vpx_dc_predictor_16x16_sse2,
-                vpx_dc_left_predictor_16x16_sse2,
-                vpx_dc_top_predictor_16x16_sse2,
-                vpx_dc_128_predictor_16x16_sse2, vpx_v_predictor_16x16_sse2,
-                NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-                vpx_tm_predictor_16x16_sse2)
-#endif  // HAVE_SSE2 && CONFIG_USE_X86INC
-
-#if HAVE_SSSE3 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSSE3, TestIntraPred16, NULL, NULL, NULL, NULL, NULL,
-                vpx_h_predictor_16x16_ssse3, vpx_d45_predictor_16x16_ssse3,
-                NULL, NULL, vpx_d153_predictor_16x16_ssse3,
-                vpx_d207_predictor_16x16_ssse3, vpx_d63_predictor_16x16_ssse3,
-                NULL)
-#endif  // HAVE_SSSE3 && CONFIG_USE_X86INC
-
-#if HAVE_DSPR2
-INTRA_PRED_TEST(DSPR2, TestIntraPred16, vpx_dc_predictor_16x16_dspr2, NULL,
-                NULL, NULL, NULL, vpx_h_predictor_16x16_dspr2, NULL, NULL, NULL,
-                NULL, NULL, NULL, NULL)
-#endif  // HAVE_DSPR2
-
-#if HAVE_NEON
-INTRA_PRED_TEST(NEON, TestIntraPred16, vpx_dc_predictor_16x16_neon,
-                vpx_dc_left_predictor_16x16_neon,
-                vpx_dc_top_predictor_16x16_neon,
-                vpx_dc_128_predictor_16x16_neon, vpx_v_predictor_16x16_neon,
-                vpx_h_predictor_16x16_neon, vpx_d45_predictor_16x16_neon, NULL,
-                NULL, NULL, NULL, NULL, vpx_tm_predictor_16x16_neon)
-#endif  // HAVE_NEON
-
-#if HAVE_MSA
-INTRA_PRED_TEST(MSA, TestIntraPred16, vpx_dc_predictor_16x16_msa,
-                vpx_dc_left_predictor_16x16_msa, vpx_dc_top_predictor_16x16_msa,
-                vpx_dc_128_predictor_16x16_msa, vpx_v_predictor_16x16_msa,
-                vpx_h_predictor_16x16_msa, NULL, NULL, NULL, NULL, NULL,
-                NULL, vpx_tm_predictor_16x16_msa)
-#endif  // HAVE_MSA
-
-// -----------------------------------------------------------------------------
-// 32x32
-
-INTRA_PRED_TEST(C, TestIntraPred32, vpx_dc_predictor_32x32_c,
-                vpx_dc_left_predictor_32x32_c, vpx_dc_top_predictor_32x32_c,
-                vpx_dc_128_predictor_32x32_c, vpx_v_predictor_32x32_c,
-                vpx_h_predictor_32x32_c, vpx_d45_predictor_32x32_c,
-                vpx_d135_predictor_32x32_c, vpx_d117_predictor_32x32_c,
-                vpx_d153_predictor_32x32_c, vpx_d207_predictor_32x32_c,
-                vpx_d63_predictor_32x32_c, vpx_tm_predictor_32x32_c)
-
-#if HAVE_SSE2 && CONFIG_USE_X86INC
-#if ARCH_X86_64
-INTRA_PRED_TEST(SSE2, TestIntraPred32, vpx_dc_predictor_32x32_sse2,
-                vpx_dc_left_predictor_32x32_sse2,
-                vpx_dc_top_predictor_32x32_sse2,
-                vpx_dc_128_predictor_32x32_sse2, vpx_v_predictor_32x32_sse2,
-                NULL, NULL, NULL, NULL, NULL, NULL, NULL,
-                vpx_tm_predictor_32x32_sse2)
-#else
-INTRA_PRED_TEST(SSE2, TestIntraPred32, vpx_dc_predictor_32x32_sse2,
-                vpx_dc_left_predictor_32x32_sse2,
-                vpx_dc_top_predictor_32x32_sse2,
-                vpx_dc_128_predictor_32x32_sse2, vpx_v_predictor_32x32_sse2,
-                NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL)
-#endif  // ARCH_X86_64
-#endif  // HAVE_SSE2 && CONFIG_USE_X86INC
-
-#if HAVE_SSSE3 && CONFIG_USE_X86INC
-INTRA_PRED_TEST(SSSE3, TestIntraPred32, NULL, NULL, NULL, NULL, NULL,
-                vpx_h_predictor_32x32_ssse3, vpx_d45_predictor_32x32_ssse3,
-                NULL, NULL, vpx_d153_predictor_32x32_ssse3,
-                vpx_d207_predictor_32x32_ssse3, vpx_d63_predictor_32x32_ssse3,
-                NULL)
-#endif  // HAVE_SSSE3 && CONFIG_USE_X86INC
-
-#if HAVE_NEON
-INTRA_PRED_TEST(NEON, TestIntraPred32, vpx_dc_predictor_32x32_neon,
-                vpx_dc_left_predictor_32x32_neon,
-                vpx_dc_top_predictor_32x32_neon,
-                vpx_dc_128_predictor_32x32_neon, vpx_v_predictor_32x32_neon,
-                vpx_h_predictor_32x32_neon, NULL, NULL, NULL, NULL, NULL, NULL,
-                vpx_tm_predictor_32x32_neon)
-#endif  // HAVE_NEON
-
-#if HAVE_MSA
-INTRA_PRED_TEST(MSA, TestIntraPred32, vpx_dc_predictor_32x32_msa,
-                vpx_dc_left_predictor_32x32_msa, vpx_dc_top_predictor_32x32_msa,
-                vpx_dc_128_predictor_32x32_msa, vpx_v_predictor_32x32_msa,
-                vpx_h_predictor_32x32_msa, NULL, NULL, NULL, NULL, NULL,
-                NULL, vpx_tm_predictor_32x32_msa)
-#endif  // HAVE_MSA
-
-#include "test/test_libvpx.cc"
--- a/test/test_libvpx.cc
+++ b/test/test_libvpx.cc
@@ -8,9 +8,6 @@
 *  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"
 #if ARCH_X86 || ARCH_X86_64
 #include "vpx_ports/x86.h"
@@ -22,11 +19,10 @@ extern void vp8_rtcd();
 #if CONFIG_VP9
 extern void vp9_rtcd();
 #endif  // CONFIG_VP9
-extern void vpx_dsp_rtcd();
 extern void vpx_scale_rtcd();
 }
+#include "third_party/googletest/src/include/gtest/gtest.h"

-#if ARCH_X86 || ARCH_X86_64
 static void append_negative_gtest_filter(const char *str) {
  std::string filter = ::testing::FLAGS_gtest_filter;
  // Negative patterns begin with one '-' followed by a ':' separated list.
@@ -34,7 +30,6 @@ static void append_negative_gtest_filter(const char *str) {
  filter += str;
  ::testing::FLAGS_gtest_filter = filter;
 }
-#endif  // ARCH_X86 || ARCH_X86_64

 int main(int argc, char **argv) {
  ::testing::InitGoogleTest(&argc, argv);
@@ -42,22 +37,22 @@ int main(int argc, char **argv) {
 #if ARCH_X86 || ARCH_X86_64
  const int simd_caps = x86_simd_caps();
  if (!(simd_caps & HAS_MMX))
-    append_negative_gtest_filter(":MMX.*:MMX/*");
+    append_negative_gtest_filter(":MMX/*");
  if (!(simd_caps & HAS_SSE))
-    append_negative_gtest_filter(":SSE.*:SSE/*");
+    append_negative_gtest_filter(":SSE/*");
  if (!(simd_caps & HAS_SSE2))
-    append_negative_gtest_filter(":SSE2.*:SSE2/*");
+    append_negative_gtest_filter(":SSE2/*");
  if (!(simd_caps & HAS_SSE3))
-    append_negative_gtest_filter(":SSE3.*:SSE3/*");
+    append_negative_gtest_filter(":SSE3/*");
  if (!(simd_caps & HAS_SSSE3))
-    append_negative_gtest_filter(":SSSE3.*:SSSE3/*");
+    append_negative_gtest_filter(":SSSE3/*");
  if (!(simd_caps & HAS_SSE4_1))
-    append_negative_gtest_filter(":SSE4_1.*:SSE4_1/*");
+    append_negative_gtest_filter(":SSE4_1/*");
  if (!(simd_caps & HAS_AVX))
-    append_negative_gtest_filter(":AVX.*:AVX/*");
+    append_negative_gtest_filter(":AVX/*");
  if (!(simd_caps & HAS_AVX2))
-    append_negative_gtest_filter(":AVX2.*:AVX2/*");
-#endif  // ARCH_X86 || ARCH_X86_64
+    append_negative_gtest_filter(":AVX2/*");
+#endif

 #if !CONFIG_SHARED
 // Shared library builds don't support whitebox tests
@@ -69,7 +64,6 @@ int main(int argc, char **argv) {
 #if CONFIG_VP9
  vp9_rtcd();
 #endif  // CONFIG_VP9
-  vpx_dsp_rtcd();
  vpx_scale_rtcd();
 #endif  // !CONFIG_SHARED

--- a/test/test_vector_test.cc
+++ b/test/test_vector_test.cc
@@ -135,7 +135,6 @@ TEST_P(TestVectorTest, MD5Match) {

 // Test VP8 decode in serial mode with single thread.
 // NOTE: VP8 only support serial mode.
-#if CONFIG_VP8_DECODER
 VP8_INSTANTIATE_TEST_CASE(
    TestVectorTest,
    ::testing::Combine(
@@ -144,10 +143,8 @@ VP8_INSTANTIATE_TEST_CASE(
        ::testing::ValuesIn(libvpx_test::kVP8TestVectors,
                            libvpx_test::kVP8TestVectors +
                                libvpx_test::kNumVP8TestVectors)));
-#endif  // CONFIG_VP8_DECODER

 // Test VP9 decode in serial mode with single thread.
-#if CONFIG_VP9_DECODER
 VP9_INSTANTIATE_TEST_CASE(
    TestVectorTest,
    ::testing::Combine(
@@ -157,6 +154,8 @@ VP9_INSTANTIATE_TEST_CASE(
                            libvpx_test::kVP9TestVectors +
                                libvpx_test::kNumVP9TestVectors)));

+
+#if CONFIG_VP9_DECODER
 // Test VP9 decode in frame parallel mode with different number of threads.
 INSTANTIATE_TEST_CASE_P(
    VP9MultiThreadedFrameParallel, TestVectorTest,
--- a/test/test_vectors.cc
+++ b/test/test_vectors.cc
@@ -165,10 +165,7 @@ const char *const kVP9TestVectors[] = {
  "vp90-2-11-size-351x287.webm", "vp90-2-11-size-351x288.webm",
  "vp90-2-11-size-352x287.webm", "vp90-2-12-droppable_1.ivf",
  "vp90-2-12-droppable_2.ivf", "vp90-2-12-droppable_3.ivf",
-#if !CONFIG_SIZE_LIMIT || \
-    (DECODE_WIDTH_LIMIT >= 20400 && DECODE_HEIGHT_LIMIT >= 120)
  "vp90-2-13-largescaling.webm",
-#endif
  "vp90-2-14-resize-fp-tiles-1-16.webm",
  "vp90-2-14-resize-fp-tiles-1-2-4-8-16.webm",
  "vp90-2-14-resize-fp-tiles-1-2.webm", "vp90-2-14-resize-fp-tiles-1-4.webm",
--- a/test/tile_independence_test.cc
+++ b/test/tile_independence_test.cc
@@ -104,5 +104,4 @@ TEST_P(TileIndependenceTest, MD5Match) {

 VP9_INSTANTIATE_TEST_CASE(TileIndependenceTest, ::testing::Range(0, 2, 1));

-VP10_INSTANTIATE_TEST_CASE(TileIndependenceTest, ::testing::Range(0, 2, 1));
 }  // namespace
--- a/test/tools_common.sh
+++ b/test/tools_common.sh
@@ -408,9 +408,6 @@ YUV_RAW_INPUT="${LIBVPX_TEST_DATA_PATH}/hantro_collage_w352h288.yuv"
 YUV_RAW_INPUT_WIDTH=352
 YUV_RAW_INPUT_HEIGHT=288

-Y4M_NOSQ_PAR_INPUT="${LIBVPX_TEST_DATA_PATH}/park_joy_90p_8_420_a10-1.y4m"
-Y4M_720P_INPUT="${LIBVPX_TEST_DATA_PATH}/niklas_1280_720_30.y4m"
-
 # Setup a trap function to clean up after tests complete.
 trap cleanup EXIT

@@ -432,7 +429,6 @@ vlog "$(basename "${0%.*}") test configuration:
  VPX_TEST_VERBOSE_OUTPUT=${VPX_TEST_VERBOSE_OUTPUT}
  YUV_RAW_INPUT=${YUV_RAW_INPUT}
  YUV_RAW_INPUT_WIDTH=${YUV_RAW_INPUT_WIDTH}
-  YUV_RAW_INPUT_HEIGHT=${YUV_RAW_INPUT_HEIGHT}
-  Y4M_NOSQ_PAR_INPUT=${Y4M_NOSQ_PAR_INPUT}"
+  YUV_RAW_INPUT_HEIGHT=${YUV_RAW_INPUT_HEIGHT}"

 fi  # End $VPX_TEST_TOOLS_COMMON_SH pseudo include guard.
--- a/test/util.h
+++ b/test/util.h
@@ -19,7 +19,8 @@
 // Macros
 #define GET_PARAM(k) std::tr1::get< k >(GetParam())

-inline double compute_psnr(const vpx_image_t *img1, const vpx_image_t *img2) {
+static double compute_psnr(const vpx_image_t *img1,
+                           const vpx_image_t *img2) {
  assert((img1->fmt == img2->fmt) &&
         (img1->d_w == img2->d_w) &&
         (img1->d_h == img2->d_h));
--- a/test/variance_test.cc
+++ b/test/variance_test.cc
--- a/test/video_source.h
+++ b/test/video_source.h
@@ -11,9 +11,6 @@
 #define TEST_VIDEO_SOURCE_H_

 #if defined(_WIN32)
-#undef NOMINMAX
-#define NOMINMAX
-#define WIN32_LEAN_AND_MEAN
 #include <windows.h>
 #endif
 #include <cstdio>
@@ -51,7 +48,7 @@ static std::string GetDataPath() {
 #undef TO_STRING
 #undef STRINGIFY

-inline FILE *OpenTestDataFile(const std::string& file_name) {
+static FILE *OpenTestDataFile(const std::string& file_name) {
  const std::string path_to_source = GetDataPath() + "/" + file_name;
  return fopen(path_to_source.c_str(), "rb");
 }
--- a/test/vp10_dct_test.cc
+++ b/test/vp10_dct_test.cc
@@ -1,111 +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 <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_ports/msvc.h"
-
-#undef CONFIG_COEFFICIENT_RANGE_CHECKING
-#define CONFIG_COEFFICIENT_RANGE_CHECKING 1
-#include "vp10/encoder/dct.c"
-
-using libvpx_test::ACMRandom;
-
-namespace {
-void reference_dct_1d(const double *in, double *out, int size) {
-  const double PI = 3.141592653589793238462643383279502884;
-  const double kInvSqrt2 = 0.707106781186547524400844362104;
-  for (int k = 0; k < size; ++k) {
-    out[k] = 0;
-    for (int n = 0; n < size; ++n) {
-      out[k] += in[n] * cos(PI * (2 * n + 1) * k / (2 * size));
-    }
-    if (k == 0)
-      out[k] = out[k] * kInvSqrt2;
-  }
-}
-
-typedef void (*FdctFuncRef)(const double *in, double *out, int size);
-typedef void (*IdctFuncRef)(const double *in, double *out, int size);
-typedef void (*FdctFunc)(const tran_low_t *in, tran_low_t *out);
-typedef void (*IdctFunc)(const tran_low_t *in, tran_low_t *out);
-
-class TransTestBase {
- public:
-  virtual ~TransTestBase() {}
-
- protected:
-  void RunFwdAccuracyCheck() {
-    tran_low_t *input  = new tran_low_t[txfm_size_];
-    tran_low_t *output = new tran_low_t[txfm_size_];
-    double *ref_input  = new double[txfm_size_];
-    double *ref_output = new double[txfm_size_];
-
-    ACMRandom rnd(ACMRandom::DeterministicSeed());
-    const int count_test_block = 5000;
-    for (int ti =  0; ti < count_test_block; ++ti) {
-      for (int ni = 0; ni < txfm_size_; ++ni) {
-        input[ni] = rnd.Rand8() - rnd.Rand8();
-        ref_input[ni] = static_cast<double>(input[ni]);
-      }
-
-      fwd_txfm_(input, output);
-      fwd_txfm_ref_(ref_input, ref_output, txfm_size_);
-
-      for (int ni = 0; ni < txfm_size_; ++ni) {
-        EXPECT_LE(
-            abs(output[ni] - static_cast<tran_low_t>(round(ref_output[ni]))),
-            max_error_);
-      }
-    }
-
-    delete[] input;
-    delete[] output;
-    delete[] ref_input;
-    delete[] ref_output;
-  }
-
-  double max_error_;
-  int txfm_size_;
-  FdctFunc fwd_txfm_;
-  FdctFuncRef fwd_txfm_ref_;
-};
-
-typedef std::tr1::tuple<FdctFunc, FdctFuncRef, int, int> FdctParam;
-class Vp10FwdTxfm
-    : public TransTestBase,
-      public ::testing::TestWithParam<FdctParam> {
- public:
-  virtual void SetUp() {
-    fwd_txfm_ = GET_PARAM(0);
-    fwd_txfm_ref_ = GET_PARAM(1);
-    txfm_size_ = GET_PARAM(2);
-    max_error_ = GET_PARAM(3);
-  }
-  virtual void TearDown() {}
-};
-
-TEST_P(Vp10FwdTxfm, RunFwdAccuracyCheck) {
-  RunFwdAccuracyCheck();
-}
-
-INSTANTIATE_TEST_CASE_P(
-    C, Vp10FwdTxfm,
-    ::testing::Values(
-        FdctParam(&fdct4, &reference_dct_1d, 4, 1),
-        FdctParam(&fdct8, &reference_dct_1d, 8, 1),
-        FdctParam(&fdct16, &reference_dct_1d, 16, 2)));
-}  // namespace
--- a/test/vp10_inv_txfm_test.cc
+++ b/test/vp10_inv_txfm_test.cc
@@ -1,321 +0,0 @@
-/*
- *  Copyright (c) 2013 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 "./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 "vpx/vpx_integer.h"
-#include "vp10/common/vp10_inv_txfm.h"
-
-using libvpx_test::ACMRandom;
-
-namespace {
-const double PI = 3.141592653589793238462643383279502884;
-const double kInvSqrt2 = 0.707106781186547524400844362104;
-
-void reference_idct_1d(const double *in, double *out, int size) {
-  for (int n = 0; n < size; ++n) {
-    out[n] = 0;
-    for (int k = 0; k < size; ++k) {
-      if (k == 0)
-        out[n] += kInvSqrt2 * in[k] * cos(PI * (2 * n + 1) * k / (2 * size));
-      else
-        out[n] += in[k] * cos(PI * (2 * n + 1) * k / (2 * size));
-    }
-  }
-}
-
-typedef void (*IdctFuncRef)(const double *in, double *out, int size);
-typedef void (*IdctFunc)(const tran_low_t *in, tran_low_t *out);
-
-class TransTestBase {
- public:
-  virtual ~TransTestBase() {}
-
- protected:
-  void RunInvAccuracyCheck() {
-    tran_low_t *input  = new tran_low_t[txfm_size_];
-    tran_low_t *output = new tran_low_t[txfm_size_];
-    double *ref_input  = new double[txfm_size_];
-    double *ref_output = new double[txfm_size_];
-
-    ACMRandom rnd(ACMRandom::DeterministicSeed());
-    const int count_test_block = 5000;
-    for (int ti =  0; ti < count_test_block; ++ti) {
-      for (int ni = 0; ni < txfm_size_; ++ni) {
-        input[ni] = rnd.Rand8() - rnd.Rand8();
-        ref_input[ni] = static_cast<double>(input[ni]);
-      }
-
-      fwd_txfm_(input, output);
-      fwd_txfm_ref_(ref_input, ref_output, txfm_size_);
-
-      for (int ni = 0; ni < txfm_size_; ++ni) {
-        EXPECT_LE(
-            abs(output[ni] - static_cast<tran_low_t>(round(ref_output[ni]))),
-            max_error_);
-      }
-    }
-
-    delete[] input;
-    delete[] output;
-    delete[] ref_input;
-    delete[] ref_output;
-  }
-
-  double max_error_;
-  int txfm_size_;
-  IdctFunc fwd_txfm_;
-  IdctFuncRef fwd_txfm_ref_;
-};
-
-typedef std::tr1::tuple<IdctFunc, IdctFuncRef, int, int> IdctParam;
-class Vp10InvTxfm
-    : public TransTestBase,
-      public ::testing::TestWithParam<IdctParam> {
- public:
-  virtual void SetUp() {
-    fwd_txfm_ = GET_PARAM(0);
-    fwd_txfm_ref_ = GET_PARAM(1);
-    txfm_size_ = GET_PARAM(2);
-    max_error_ = GET_PARAM(3);
-  }
-  virtual void TearDown() {}
-};
-
-TEST_P(Vp10InvTxfm, RunInvAccuracyCheck) {
-  RunInvAccuracyCheck();
-}
-
-INSTANTIATE_TEST_CASE_P(
-    C, Vp10InvTxfm,
-    ::testing::Values(
-        IdctParam(&vp10_idct4_c, &reference_idct_1d, 4, 1),
-        IdctParam(&vp10_idct8_c, &reference_idct_1d, 8, 2),
-        IdctParam(&vp10_idct16_c, &reference_idct_1d, 16, 4),
-        IdctParam(&vp10_idct32_c, &reference_idct_1d, 32, 6))
-);
-
-typedef void (*FwdTxfmFunc)(const int16_t *in, tran_low_t *out, int stride);
-typedef void (*InvTxfmFunc)(const tran_low_t *in, uint8_t *out, int stride);
-typedef std::tr1::tuple<FwdTxfmFunc,
-                        InvTxfmFunc,
-                        InvTxfmFunc,
-                        TX_SIZE, int> PartialInvTxfmParam;
-const int kMaxNumCoeffs = 1024;
-class Vp10PartialIDctTest
-    : public ::testing::TestWithParam<PartialInvTxfmParam> {
- public:
-  virtual ~Vp10PartialIDctTest() {}
-  virtual void SetUp() {
-    ftxfm_ = GET_PARAM(0);
-    full_itxfm_ = GET_PARAM(1);
-    partial_itxfm_ = GET_PARAM(2);
-    tx_size_  = GET_PARAM(3);
-    last_nonzero_ = GET_PARAM(4);
-  }
-
-  virtual void TearDown() { libvpx_test::ClearSystemState(); }
-
- protected:
-  int last_nonzero_;
-  TX_SIZE tx_size_;
-  FwdTxfmFunc ftxfm_;
-  InvTxfmFunc full_itxfm_;
-  InvTxfmFunc partial_itxfm_;
-};
-
-TEST_P(Vp10PartialIDctTest, RunQuantCheck) {
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  int size;
-  switch (tx_size_) {
-    case TX_4X4:
-      size = 4;
-      break;
-    case TX_8X8:
-      size = 8;
-      break;
-    case TX_16X16:
-      size = 16;
-      break;
-    case TX_32X32:
-      size = 32;
-      break;
-    default:
-      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]);
-
-  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]);
-
-  int max_error = 0;
-  for (int i = 0; i < count_test_block; ++i) {
-    // clear out destination buffer
-    memset(dst1, 0, sizeof(*dst1) * block_size);
-    memset(dst2, 0, sizeof(*dst2) * block_size);
-    memset(test_coef_block1, 0, sizeof(*test_coef_block1) * block_size);
-    memset(test_coef_block2, 0, sizeof(*test_coef_block2) * block_size);
-
-    ACMRandom rnd(ACMRandom::DeterministicSeed());
-
-    for (int i = 0; i < count_test_block; ++i) {
-      // Initialize a test block with input range [-255, 255].
-      if (i == 0) {
-        for (int j = 0; j < block_size; ++j)
-          input_extreme_block[j] = 255;
-      } else if (i == 1) {
-        for (int j = 0; j < block_size; ++j)
-          input_extreme_block[j] = -255;
-      } else {
-        for (int j = 0; j < block_size; ++j) {
-          input_extreme_block[j] = rnd.Rand8() % 2 ? 255 : -255;
-        }
-      }
-
-      ftxfm_(input_extreme_block, output_ref_block, size);
-
-      // 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]]
-                         = (output_ref_block[j] / 1828) * 1828;
-    }
-
-    ASM_REGISTER_STATE_CHECK(full_itxfm_(test_coef_block1, dst1, size));
-    ASM_REGISTER_STATE_CHECK(partial_itxfm_(test_coef_block1, dst2, size));
-
-    for (int j = 0; j < block_size; ++j) {
-      const int diff = dst1[j] - dst2[j];
-      const int error = diff * diff;
-      if (max_error < error)
-        max_error = error;
-    }
-  }
-
-  EXPECT_EQ(0, max_error)
-      << "Error: partial inverse transform produces different results";
-}
-
-TEST_P(Vp10PartialIDctTest, ResultsMatch) {
-  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  int size;
-  switch (tx_size_) {
-    case TX_4X4:
-      size = 4;
-      break;
-    case TX_8X8:
-      size = 8;
-      break;
-    case TX_16X16:
-      size = 16;
-      break;
-    case TX_32X32:
-      size = 32;
-      break;
-    default:
-      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]);
-  const int count_test_block = 1000;
-  const int max_coeff = 32766 / 4;
-  const int block_size = size * size;
-  int max_error = 0;
-  for (int i = 0; i < count_test_block; ++i) {
-    // clear out destination buffer
-    memset(dst1, 0, sizeof(*dst1) * block_size);
-    memset(dst2, 0, sizeof(*dst2) * block_size);
-    memset(test_coef_block1, 0, sizeof(*test_coef_block1) * block_size);
-    memset(test_coef_block2, 0, sizeof(*test_coef_block2) * block_size);
-    int max_energy_leftover = max_coeff * max_coeff;
-    for (int j = 0; j < last_nonzero_; ++j) {
-      int16_t coef = static_cast<int16_t>(sqrt(1.0 * max_energy_leftover) *
-                                          (rnd.Rand16() - 32768) / 65536);
-      max_energy_leftover -= coef * coef;
-      if (max_energy_leftover < 0) {
-        max_energy_leftover = 0;
-        coef = 0;
-      }
-      test_coef_block1[vp10_default_scan_orders[tx_size_].scan[j]] = coef;
-    }
-
-    memcpy(test_coef_block2, test_coef_block1,
-           sizeof(*test_coef_block2) * block_size);
-
-    ASM_REGISTER_STATE_CHECK(full_itxfm_(test_coef_block1, dst1, size));
-    ASM_REGISTER_STATE_CHECK(partial_itxfm_(test_coef_block2, dst2, size));
-
-    for (int j = 0; j < block_size; ++j) {
-      const int diff = dst1[j] - dst2[j];
-      const int error = diff * diff;
-      if (max_error < error)
-        max_error = error;
-    }
-  }
-
-  EXPECT_EQ(0, max_error)
-      << "Error: partial inverse transform produces different results";
-}
-using std::tr1::make_tuple;
-
-INSTANTIATE_TEST_CASE_P(
-    C, Vp10PartialIDctTest,
-    ::testing::Values(
-        make_tuple(&vpx_fdct32x32_c,
-                   &vp10_idct32x32_1024_add_c,
-                   &vp10_idct32x32_34_add_c,
-                   TX_32X32, 34),
-        make_tuple(&vpx_fdct32x32_c,
-                   &vp10_idct32x32_1024_add_c,
-                   &vp10_idct32x32_1_add_c,
-                   TX_32X32, 1),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vp10_idct16x16_256_add_c,
-                   &vp10_idct16x16_10_add_c,
-                   TX_16X16, 10),
-        make_tuple(&vpx_fdct16x16_c,
-                   &vp10_idct16x16_256_add_c,
-                   &vp10_idct16x16_1_add_c,
-                   TX_16X16, 1),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vp10_idct8x8_64_add_c,
-                   &vp10_idct8x8_12_add_c,
-                   TX_8X8, 12),
-        make_tuple(&vpx_fdct8x8_c,
-                   &vp10_idct8x8_64_add_c,
-                   &vp10_idct8x8_1_add_c,
-                   TX_8X8, 1),
-        make_tuple(&vpx_fdct4x4_c,
-                   &vp10_idct4x4_16_add_c,
-                   &vp10_idct4x4_1_add_c,
-                   TX_4X4, 1)));
-}  // namespace
--- a/test/vp8_boolcoder_test.cc
+++ b/test/vp8_boolcoder_test.cc
@@ -16,13 +16,13 @@
 #include <string.h>
 #include <sys/types.h>

-#include "third_party/googletest/src/include/gtest/gtest.h"
-
 #include "test/acm_random.h"
-#include "vp8/decoder/dboolhuff.h"
-#include "vp8/encoder/boolhuff.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
 #include "vpx/vpx_integer.h"

+#include "vp8/encoder/boolhuff.h"
+#include "vp8/decoder/dboolhuff.h"
+
 namespace {
 const int num_tests = 10;

--- a/test/vp8_denoiser_sse2_test.cc
+++ b/test/vp8_denoiser_sse2_test.cc
@@ -52,13 +52,13 @@ TEST_P(VP8DenoiserTest, BitexactCheck) {
  // mc_avg_block is the denoised reference block,
  // avg_block_c is the denoised result from C code,
  // avg_block_sse2 is the denoised result from SSE2 code.
-  DECLARE_ALIGNED(16, uint8_t, sig_block_c[kNumPixels]);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, sig_block_c, kNumPixels);
  // Since in VP8 denoiser, the source signal will be changed,
  // we need another copy of the source signal as the input of sse2 code.
-  DECLARE_ALIGNED(16, uint8_t, sig_block_sse2[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, mc_avg_block[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, avg_block_c[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, avg_block_sse2[kNumPixels]);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, sig_block_sse2, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, mc_avg_block, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, avg_block_c, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, avg_block_sse2, kNumPixels);

  for (int i = 0; i < count_test_block; ++i) {
    // Generate random motion magnitude, 20% of which exceed the threshold.
--- a/test/vp8_fdct4x4_test.cc
+++ b/test/vp8_fdct4x4_test.cc
@@ -15,10 +15,10 @@
 #include <string.h>
 #include <sys/types.h>

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

 namespace {
--- a/test/vp9_arf_freq_test.cc
+++ b/test/vp9_arf_freq_test.cc
@@ -1,252 +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 "vp9/encoder/vp9_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_VP9_HIGHBITDEPTH
-  // Add list of profile 2/3 test videos here ...
-#endif  // CONFIG_VP9_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 ArfFreqTest
-    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith3Params<TestVideoParam, \
-                                                 TestEncodeParam, int> {
- protected:
-  ArfFreqTest()
-      : EncoderTest(GET_PARAM(0)),
-        test_video_param_(GET_PARAM(1)),
-        test_encode_param_(GET_PARAM(2)),
-        min_arf_requested_(GET_PARAM(3)) {
-  }
-
-  virtual ~ArfFreqTest() {}
-
-  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 vp9_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(ArfFreqTest, 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;
-
-  libvpx_test::VideoSource *video;
-  if (is_extension_y4m(test_video_param_.filename)) {
-    video = new libvpx_test::Y4mVideoSource(test_video_param_.filename,
-                                            0, kFrames);
-  } else {
-    video = 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));
-  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);
-  }
-  delete(video);
-}
-
-VP9_INSTANTIATE_TEST_CASE(
-    ArfFreqTest,
-    ::testing::ValuesIn(kTestVectors),
-    ::testing::ValuesIn(kEncodeVectors),
-    ::testing::ValuesIn(kMinArfVectors));
-
-#if CONFIG_VP9_HIGHBITDEPTH
-# if CONFIG_VP10_ENCODER
-// TODO(angiebird): 25-29 fail in high bitdepth mode.
-INSTANTIATE_TEST_CASE_P(
-    DISABLED_VP10, ArfFreqTest,
-    ::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(
-    ArfFreqTest,
-    ::testing::ValuesIn(kTestVectors),
-    ::testing::ValuesIn(kEncodeVectors),
-    ::testing::ValuesIn(kMinArfVectors));
-#endif  // CONFIG_VP9_HIGHBITDEPTH
-}  // namespace
--- a/test/vp9_avg_test.cc
+++ b/test/vp9_avg_test.cc
@@ -8,22 +8,22 @@
 *  be found in the AUTHORS file in the root of the source tree.
 */

+
+#include <string.h>
 #include <limits.h>
 #include <stdio.h>
-#include <string.h>
-
-#include "third_party/googletest/src/include/gtest/gtest.h"

 #include "./vpx_config.h"
 #if CONFIG_VP9_ENCODER
 #include "./vp9_rtcd.h"
 #endif
+#include "vpx_mem/vpx_mem.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"
+#include "third_party/googletest/src/include/gtest/gtest.h"

 using libvpx_test::ACMRandom;

@@ -121,79 +121,6 @@ class AverageTest
  }
 };

-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_;
-};
-

 uint8_t* AverageTestBase::source_data_ = NULL;

@@ -216,36 +143,6 @@ TEST_P(AverageTest, Random) {
  }
 }

-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();
-}
-
 using std::tr1::make_tuple;

 INSTANTIATE_TEST_CASE_P(
@@ -254,6 +151,7 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(16, 16, 1, 8, &vp9_avg_8x8_c),
        make_tuple(16, 16, 1, 4, &vp9_avg_4x4_c)));

+
 #if HAVE_SSE2
 INSTANTIATE_TEST_CASE_P(
    SSE2, AverageTest,
@@ -265,17 +163,6 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(16, 16, 5, 4, &vp9_avg_4x4_sse2),
        make_tuple(32, 32, 15, 4, &vp9_avg_4x4_sse2)));

-INSTANTIATE_TEST_CASE_P(
-    SSE2, IntProRowTest, ::testing::Values(
-        make_tuple(16, &vp9_int_pro_row_sse2, &vp9_int_pro_row_c),
-        make_tuple(32, &vp9_int_pro_row_sse2, &vp9_int_pro_row_c),
-        make_tuple(64, &vp9_int_pro_row_sse2, &vp9_int_pro_row_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    SSE2, IntProColTest, ::testing::Values(
-        make_tuple(16, &vp9_int_pro_col_sse2, &vp9_int_pro_col_c),
-        make_tuple(32, &vp9_int_pro_col_sse2, &vp9_int_pro_col_c),
-        make_tuple(64, &vp9_int_pro_col_sse2, &vp9_int_pro_col_c)));
 #endif

 #if HAVE_NEON
@@ -286,29 +173,6 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(16, 16, 5, 8, &vp9_avg_8x8_neon),
        make_tuple(32, 32, 15, 8, &vp9_avg_8x8_neon)));

-INSTANTIATE_TEST_CASE_P(
-    NEON, IntProRowTest, ::testing::Values(
-        make_tuple(16, &vp9_int_pro_row_neon, &vp9_int_pro_row_c),
-        make_tuple(32, &vp9_int_pro_row_neon, &vp9_int_pro_row_c),
-        make_tuple(64, &vp9_int_pro_row_neon, &vp9_int_pro_row_c)));
-
-INSTANTIATE_TEST_CASE_P(
-    NEON, IntProColTest, ::testing::Values(
-        make_tuple(16, &vp9_int_pro_col_neon, &vp9_int_pro_col_c),
-        make_tuple(32, &vp9_int_pro_col_neon, &vp9_int_pro_col_c),
-        make_tuple(64, &vp9_int_pro_col_neon, &vp9_int_pro_col_c)));
-#endif
-
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(
-    MSA, AverageTest,
-    ::testing::Values(
-        make_tuple(16, 16, 0, 8, &vp9_avg_8x8_msa),
-        make_tuple(16, 16, 5, 8, &vp9_avg_8x8_msa),
-        make_tuple(32, 32, 15, 8, &vp9_avg_8x8_msa),
-        make_tuple(16, 16, 0, 4, &vp9_avg_4x4_msa),
-        make_tuple(16, 16, 5, 4, &vp9_avg_4x4_msa),
-        make_tuple(32, 32, 15, 4, &vp9_avg_4x4_msa)));
 #endif

 }  // namespace
--- a/test/vp9_boolcoder_test.cc
+++ b/test/vp9_boolcoder_test.cc
@@ -14,10 +14,11 @@

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

+#include "vp9/decoder/vp9_reader.h"
+#include "vp9/encoder/vp9_writer.h"
+
 #include "test/acm_random.h"
 #include "vpx/vpx_integer.h"
-#include "vpx_dsp/bitreader.h"
-#include "vpx_dsp/bitwriter.h"

 using libvpx_test::ACMRandom;

@@ -49,9 +50,9 @@ TEST(VP9, TestBitIO) {
        const int random_seed = 6432;
        const int kBufferSize = 10000;
        ACMRandom bit_rnd(random_seed);
-        vpx_writer bw;
+        vp9_writer bw;
        uint8_t bw_buffer[kBufferSize];
-        vpx_start_encode(&bw, bw_buffer);
+        vp9_start_encode(&bw, bw_buffer);

        int bit = (bit_method == 0) ? 0 : (bit_method == 1) ? 1 : 0;
        for (int i = 0; i < kBitsToTest; ++i) {
@@ -60,16 +61,16 @@ TEST(VP9, TestBitIO) {
          } else if (bit_method == 3) {
            bit = bit_rnd(2);
          }
-          vpx_write(&bw, bit, static_cast<int>(probas[i]));
+          vp9_write(&bw, bit, static_cast<int>(probas[i]));
        }

-        vpx_stop_encode(&bw);
+        vp9_stop_encode(&bw);

        // First bit should be zero
        GTEST_ASSERT_EQ(bw_buffer[0] & 0x80, 0);

-        vpx_reader br;
-        vpx_reader_init(&br, bw_buffer, kBufferSize, NULL, NULL);
+        vp9_reader br;
+        vp9_reader_init(&br, bw_buffer, kBufferSize, NULL, NULL);
        bit_rnd.Reset(random_seed);
        for (int i = 0; i < kBitsToTest; ++i) {
          if (bit_method == 2) {
@@ -77,7 +78,7 @@ TEST(VP9, TestBitIO) {
          } else if (bit_method == 3) {
            bit = bit_rnd(2);
          }
-          GTEST_ASSERT_EQ(vpx_read(&br, probas[i]), bit)
+          GTEST_ASSERT_EQ(vp9_read(&br, probas[i]), bit)
              << "pos: " << i << " / " << kBitsToTest
              << " bit_method: " << bit_method
              << " method: " << method;
--- a/test/vp9_denoiser_sse2_test.cc
+++ b/test/vp9_denoiser_sse2_test.cc
@@ -52,10 +52,10 @@ TEST_P(VP9DenoiserTest, BitexactCheck) {
  // mc_avg_block is the denoised reference block,
  // avg_block_c is the denoised result from C code,
  // avg_block_sse2 is the denoised result from SSE2 code.
-  DECLARE_ALIGNED(16, uint8_t, sig_block[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, mc_avg_block[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, avg_block_c[kNumPixels]);
-  DECLARE_ALIGNED(16, uint8_t, avg_block_sse2[kNumPixels]);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, sig_block, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, mc_avg_block, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, avg_block_c, kNumPixels);
+  DECLARE_ALIGNED_ARRAY(16, uint8_t, avg_block_sse2, kNumPixels);

  for (int i = 0; i < count_test_block; ++i) {
    // Generate random motion magnitude, 20% of which exceed the threshold.
--- a/test/vp9_encoder_parms_get_to_decoder.cc
+++ b/test/vp9_encoder_parms_get_to_decoder.cc
@@ -8,16 +8,43 @@
 *  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 "vp9/vp9_dx_iface.h"
+#include "test/yuv_video_source.h"
+#include "test/util.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"
+#include "vp9/decoder/vp9_decoder.h"
+
+typedef vpx_codec_stream_info_t vp9_stream_info_t;
+struct vpx_codec_alg_priv {
+  vpx_codec_priv_t        base;
+  vpx_codec_dec_cfg_t     cfg;
+  vp9_stream_info_t       si;
+  struct VP9Decoder      *pbi;
+  int                     postproc_cfg_set;
+  vp8_postproc_cfg_t      postproc_cfg;
+  vpx_decrypt_cb          decrypt_cb;
+  void                   *decrypt_state;
+  vpx_image_t             img;
+  int                     img_avail;
+  int                     flushed;
+  int                     invert_tile_order;
+  int                     frame_parallel_decode;
+
+  // External frame buffer info to save for VP9 common.
+  void *ext_priv;  // Private data associated with the external frame buffers.
+  vpx_get_frame_buffer_cb_fn_t get_ext_fb_cb;
+  vpx_release_frame_buffer_cb_fn_t release_ext_fb_cb;
+};
+
+static vpx_codec_alg_priv_t *get_alg_priv(vpx_codec_ctx_t *ctx) {
+  return (vpx_codec_alg_priv_t *)ctx->priv;
+}

 namespace {

+const unsigned int kFramerate = 50;
 const int kCpuUsed = 2;

 struct EncodePerfTestVideo {
@@ -38,29 +65,37 @@ struct EncodeParameters {
  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, VPX_CR_FULL_RANGE, VPX_CS_BT_709},
-  {0, 0, 1, 0, 0, VPX_CR_FULL_RANGE, VPX_CS_BT_2020},
-  {0, 2, 0, 0, 1, VPX_CR_STUDIO_RANGE, VPX_CS_UNKNOWN, { 640, 480 }},
-  // TODO(JBB): Test profiles (requires more work).
+    {0, 0, 0, 1, 0, VPX_CS_BT_601},
+    {0, 0, 0, 0, 0, VPX_CS_BT_709},
+    {0, 0, 1, 0, 0, VPX_CS_BT_2020},
+    {0, 2, 0, 0, 1, VPX_CS_UNKNOWN},
+    // TODO(JBB): Test profiles (requires more work).
 };

-class VpxEncoderParmsGetToDecoder
+int is_extension_y4m(const char *filename) {
+  const char *dot = strrchr(filename, '.');
+  if (!dot || dot == filename)
+    return 0;
+  else
+    return !strcmp(dot, ".y4m");
+}
+
+class Vp9EncoderParmsGetToDecoder
    : public ::libvpx_test::EncoderTest,
-      public ::libvpx_test::CodecTestWith2Params<EncodeParameters,
+      public ::libvpx_test::CodecTestWith2Params<EncodeParameters, \
                                                 EncodePerfTestVideo> {
 protected:
-  VpxEncoderParmsGetToDecoder()
-      : EncoderTest(GET_PARAM(0)), encode_parms(GET_PARAM(1)) {}
+  Vp9EncoderParmsGetToDecoder()
+      : EncoderTest(GET_PARAM(0)),
+        encode_parms(GET_PARAM(1)) {
+  }

-  virtual ~VpxEncoderParmsGetToDecoder() {}
+  virtual ~Vp9EncoderParmsGetToDecoder() {}

  virtual void SetUp() {
    InitializeConfig();
@@ -76,7 +111,6 @@ class VpxEncoderParmsGetToDecoder
                                  ::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);
@@ -87,44 +121,37 @@ class VpxEncoderParmsGetToDecoder
      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,
+                                  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);
-    VP9_COMMON *const common = &worker_data->pbi->common;
+    vpx_codec_ctx_t* vp9_decoder = decoder->GetDecoder();
+    vpx_codec_alg_priv_t* priv =
+        (vpx_codec_alg_priv_t*) get_alg_priv(vp9_decoder);
+
+    VP9Decoder* pbi = priv->pbi;
+    VP9_COMMON* common = &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(common->base_qindex, 0);
+      EXPECT_EQ(common->y_dc_delta_q, 0);
+      EXPECT_EQ(common->uv_dc_delta_q, 0);
+      EXPECT_EQ(common->uv_ac_delta_q, 0);
+      EXPECT_EQ(common->tx_mode, ONLY_4X4);
    }
-    EXPECT_EQ(encode_parms.error_resilient, common->error_resilient_mode);
+    EXPECT_EQ(common->error_resilient_mode, encode_parms.error_resilient);
    if (encode_parms.error_resilient) {
-      EXPECT_EQ(1, common->frame_parallel_decoding_mode);
-      EXPECT_EQ(0, common->use_prev_frame_mvs);
+      EXPECT_EQ(common->frame_parallel_decoding_mode, 1);
+      EXPECT_EQ(common->use_prev_frame_mvs, 0);
    } else {
-      EXPECT_EQ(encode_parms.frame_parallel,
-                common->frame_parallel_decoding_mode);
+      EXPECT_EQ(common->frame_parallel_decoding_mode,
+                encode_parms.frame_parallel);
    }
-    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(common->color_space, encode_parms.cs);
+    EXPECT_EQ(common->log2_tile_cols, encode_parms.tile_cols);
+    EXPECT_EQ(common->log2_tile_rows, encode_parms.tile_rows);

    EXPECT_EQ(VPX_CODEC_OK, res_dec) << decoder->DecodeError();
    return VPX_CODEC_OK == res_dec;
@@ -136,18 +163,31 @@ class VpxEncoderParmsGetToDecoder
  EncodeParameters encode_parms;
 };

-TEST_P(VpxEncoderParmsGetToDecoder, BitstreamParms) {
+// TODO(hkuang): This test conflicts with frame parallel decode. So disable it
+// for now until fix.
+TEST_P(Vp9EncoderParmsGetToDecoder, DISABLED_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);
+  libvpx_test::VideoSource *video;
+  if (is_extension_y4m(test_video_.name)) {
+    video = new libvpx_test::Y4mVideoSource(test_video_.name,
+                                            0, test_video_.frames);
+  } else {
+    video = new libvpx_test::YUVVideoSource(test_video_.name,
+                                            VPX_IMG_FMT_I420,
+                                            test_video_.width,
+                                            test_video_.height,
+                                            kFramerate, 1, 0,
+                                            test_video_.frames);
+  }

  ASSERT_NO_FATAL_FAILURE(RunLoop(video));
-  delete video;
+  delete(video);
 }

-VP9_INSTANTIATE_TEST_CASE(VpxEncoderParmsGetToDecoder,
-                          ::testing::ValuesIn(kVP9EncodeParameterSet),
-                          ::testing::ValuesIn(kVP9EncodePerfTestVectors));
+VP9_INSTANTIATE_TEST_CASE(
+    Vp9EncoderParmsGetToDecoder,
+    ::testing::ValuesIn(kVP9EncodeParameterSet),
+    ::testing::ValuesIn(kVP9EncodePerfTestVectors));
+
 }  // namespace
--- a/test/vp9_end_to_end_test.cc
+++ b/test/vp9_end_to_end_test.cc
@@ -8,13 +8,12 @@
 *  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 "test/util.h"
+#include "third_party/googletest/src/include/gtest/gtest.h"

 namespace {

@@ -187,23 +186,4 @@ VP9_INSTANTIATE_TEST_CASE(
    ::testing::ValuesIn(kTestVectors),
    ::testing::ValuesIn(kCpuUsedVectors));

-#if CONFIG_VP9_HIGHBITDEPTH
-# if CONFIG_VP10_ENCODER
-// TODO(angiebird): many fail in high bitdepth mode.
-INSTANTIATE_TEST_CASE_P(
-    DISABLED_VP10, EndToEndTestLarge,
-    ::testing::Combine(
-        ::testing::Values(static_cast<const libvpx_test::CodecFactory *>(
-            &libvpx_test::kVP10)),
-        ::testing::ValuesIn(kEncodingModeVectors),
-        ::testing::ValuesIn(kTestVectors),
-        ::testing::ValuesIn(kCpuUsedVectors)));
-# endif  // CONFIG_VP10_ENCODER
-#else
-VP10_INSTANTIATE_TEST_CASE(
-    EndToEndTestLarge,
-    ::testing::ValuesIn(kEncodingModeVectors),
-    ::testing::ValuesIn(kTestVectors),
-    ::testing::ValuesIn(kCpuUsedVectors));
-#endif  // CONFIG_VP9_HIGHBITDEPTH
 }  // namespace
--- a/test/vp9_error_block_test.cc
+++ b/test/vp9_error_block_test.cc
@@ -14,14 +14,13 @@

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

-#include "./vpx_config.h"
-#include "./vp9_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_config.h"
+#include "./vp9_rtcd.h"
 #include "vp9/common/vp9_entropy.h"
-#include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"

 using libvpx_test::ACMRandom;
@@ -58,8 +57,8 @@ class ErrorBlockTest

 TEST_P(ErrorBlockTest, OperationCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff[4096]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff[4096]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff,   4096);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff, 4096);
  int err_count_total = 0;
  int first_failure = -1;
  intptr_t block_size;
@@ -67,22 +66,12 @@ TEST_P(ErrorBlockTest, OperationCheck) {
  int64_t ret;
  int64_t ref_ssz;
  int64_t ref_ret;
-  const int msb = bit_depth_ + 8 - 1;
  for (int i = 0; i < kNumIterations; ++i) {
    int err_count = 0;
    block_size = 16 << (i % 9);  // All block sizes from 4x4, 8x4 ..64x64
    for (int j = 0; j < block_size; j++) {
-      // coeff and dqcoeff will always have at least the same sign, and this
-      // can be used for optimization, so generate test input precisely.
-      if (rnd(2)) {
-        // Positive number
-        coeff[j]   = rnd(1 << msb);
-        dqcoeff[j] = rnd(1 << msb);
-      } else {
-        // Negative number
-        coeff[j]   = -rnd(1 << msb);
-        dqcoeff[j] = -rnd(1 << msb);
-      }
+      coeff[j]   = rnd(2 << 20) - (1 << 20);
+      dqcoeff[j] = rnd(2 << 20) - (1 << 20);
    }
    ref_ret = ref_error_block_op_(coeff, dqcoeff, block_size, &ref_ssz,
                                  bit_depth_);
@@ -95,14 +84,14 @@ TEST_P(ErrorBlockTest, OperationCheck) {
    err_count_total += err_count;
  }
  EXPECT_EQ(0, err_count_total)
-      << "Error: Error Block Test, C output doesn't match optimized output. "
+      << "Error: Error Block Test, C output doesn't match SSE2 output. "
      << "First failed at test case " << first_failure;
 }

 TEST_P(ErrorBlockTest, ExtremeValues) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff[4096]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff[4096]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff,   4096);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff, 4096);
  int err_count_total = 0;
  int first_failure = -1;
  intptr_t block_size;
@@ -110,36 +99,23 @@ TEST_P(ErrorBlockTest, ExtremeValues) {
  int64_t ret;
  int64_t ref_ssz;
  int64_t ref_ret;
-  const int msb = bit_depth_ + 8 - 1;
-  int max_val = ((1 << msb) - 1);
+  int max_val = ((1 << 20) - 1);
  for (int i = 0; i < kNumIterations; ++i) {
    int err_count = 0;
-    int k = (i / 9) % 9;
+    int k = (i / 9) % 5;

    // Change the maximum coeff value, to test different bit boundaries
-    if ( k == 8 && (i % 9) == 0 ) {
+    if ( k == 4 && (i % 9) == 0 ) {
      max_val >>= 1;
    }
    block_size = 16 << (i % 9);  // All block sizes from 4x4, 8x4 ..64x64
    for (int j = 0; j < block_size; j++) {
-      if (k < 4) {
-        // Test at positive maximum values
-        coeff[j]   = k % 2 ? max_val : 0;
-        dqcoeff[j] = (k >> 1) % 2 ? max_val : 0;
-      } else if (k < 8) {
-        // Test at negative maximum values
-        coeff[j]   = k % 2 ? -max_val : 0;
-        dqcoeff[j] = (k >> 1) % 2 ? -max_val : 0;
+      if (k < 4) {  // Test at maximum values
+        coeff[j]   = k % 2 ? max_val : -max_val;
+        dqcoeff[j] = (k >> 1) % 2 ? max_val : -max_val;
      } else {
-        if (rnd(2)) {
-          // Positive number
-          coeff[j]   = rnd(1 << 14);
-          dqcoeff[j] = rnd(1 << 14);
-        } else {
-          // Negative number
-          coeff[j]   = -rnd(1 << 14);
-          dqcoeff[j] = -rnd(1 << 14);
-        }
+        coeff[j]   = rnd(2 << 14) - (1 << 14);
+        dqcoeff[j] = rnd(2 << 14) - (1 << 14);
      }
    }
    ref_ret = ref_error_block_op_(coeff, dqcoeff, block_size, &ref_ssz,
@@ -153,30 +129,13 @@ TEST_P(ErrorBlockTest, ExtremeValues) {
    err_count_total += err_count;
  }
  EXPECT_EQ(0, err_count_total)
-      << "Error: Error Block Test, C output doesn't match optimized output. "
+      << "Error: Error Block Test, C output doesn't match SSE2 output. "
      << "First failed at test case " << first_failure;
 }

 using std::tr1::make_tuple;

-#if CONFIG_USE_X86INC
-int64_t wrap_vp9_highbd_block_error_8bit_c(const tran_low_t *coeff,
-                                           const tran_low_t *dqcoeff,
-                                           intptr_t block_size,
-                                           int64_t *ssz, int bps) {
-  assert(bps == 8);
-  return vp9_highbd_block_error_8bit_c(coeff, dqcoeff, block_size, ssz);
-}
-
 #if HAVE_SSE2
-int64_t wrap_vp9_highbd_block_error_8bit_sse2(const tran_low_t *coeff,
-                                              const tran_low_t *dqcoeff,
-                                              intptr_t block_size,
-                                              int64_t *ssz, int bps) {
-  assert(bps == 8);
-  return vp9_highbd_block_error_8bit_sse2(coeff, dqcoeff, block_size, ssz);
-}
-
 INSTANTIATE_TEST_CASE_P(
    SSE2, ErrorBlockTest,
    ::testing::Values(
@@ -185,27 +144,7 @@ INSTANTIATE_TEST_CASE_P(
        make_tuple(&vp9_highbd_block_error_sse2,
                   &vp9_highbd_block_error_c, VPX_BITS_12),
        make_tuple(&vp9_highbd_block_error_sse2,
-                   &vp9_highbd_block_error_c, VPX_BITS_8),
-        make_tuple(&wrap_vp9_highbd_block_error_8bit_sse2,
-                   &wrap_vp9_highbd_block_error_8bit_c, VPX_BITS_8)));
+                   &vp9_highbd_block_error_c, VPX_BITS_8)));
 #endif  // HAVE_SSE2
-
-#if HAVE_AVX
-int64_t wrap_vp9_highbd_block_error_8bit_avx(const tran_low_t *coeff,
-                                              const tran_low_t *dqcoeff,
-                                              intptr_t block_size,
-                                              int64_t *ssz, int bps) {
-  assert(bps == 8);
-  return vp9_highbd_block_error_8bit_avx(coeff, dqcoeff, block_size, ssz);
-}
-
-INSTANTIATE_TEST_CASE_P(
-    AVX, ErrorBlockTest,
-    ::testing::Values(
-        make_tuple(&wrap_vp9_highbd_block_error_8bit_avx,
-                   &wrap_vp9_highbd_block_error_8bit_c, VPX_BITS_8)));
-#endif  // HAVE_AVX
-
-#endif  // CONFIG_USE_X86INC
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 }  // namespace
--- a/test/vp9_ethread_test.cc
+++ b/test/vp9_ethread_test.cc
@@ -18,11 +18,11 @@
 #include "test/y4m_video_source.h"

 namespace {
-class VPxEncoderThreadTest
+class VP9EncoderThreadTest
    : public ::libvpx_test::EncoderTest,
      public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
 protected:
-  VPxEncoderThreadTest()
+  VP9EncoderThreadTest()
      : EncoderTest(GET_PARAM(0)),
        encoder_initialized_(false),
        tiles_(2),
@@ -36,7 +36,7 @@ class VPxEncoderThreadTest

    md5_.clear();
  }
-  virtual ~VPxEncoderThreadTest() {
+  virtual ~VP9EncoderThreadTest() {
    delete decoder_;
  }

@@ -105,7 +105,7 @@ class VPxEncoderThreadTest
  std::vector<std::string> md5_;
 };

-TEST_P(VPxEncoderThreadTest, EncoderResultTest) {
+TEST_P(VP9EncoderThreadTest, EncoderResultTest) {
  std::vector<std::string> single_thr_md5, multi_thr_md5;

  ::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 15, 20);
@@ -130,13 +130,8 @@ TEST_P(VPxEncoderThreadTest, EncoderResultTest) {
 }

 VP9_INSTANTIATE_TEST_CASE(
-    VPxEncoderThreadTest,
+    VP9EncoderThreadTest,
    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood,
                      ::libvpx_test::kRealTime),
    ::testing::Range(1, 9));
-
-VP10_INSTANTIATE_TEST_CASE(
-    VPxEncoderThreadTest,
-    ::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood),
-    ::testing::Range(1, 3));
 }  // namespace
--- a/test/vp9_intrapred_test.cc
+++ b/test/vp9_intrapred_test.cc
@@ -10,17 +10,17 @@

 #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 "third_party/googletest/src/include/gtest/gtest.h"
+
+#include "./vpx_config.h"
+#include "./vp9_rtcd.h"
 #include "vp9/common/vp9_blockd.h"
 #include "vp9/common/vp9_pred_common.h"
 #include "vpx_mem/vpx_mem.h"
+#include "test/util.h"

 namespace {

@@ -120,10 +120,10 @@ class VP9IntraPredTest

 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]);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, left_col, 2*32);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, above_data, 2*32+32);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, dst, 3 * 32 * 32);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_dst, 3 * 32 * 32);
  RunTest(left_col, above_data, dst, ref_dst);
 }

@@ -131,171 +131,168 @@ using std::tr1::make_tuple;

 #if HAVE_SSE2
 #if CONFIG_VP9_HIGHBITDEPTH
-#if CONFIG_USE_X86INC
 #if ARCH_X86_64
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_8, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_32x32_sse2,
-                                       &vpx_highbd_dc_predictor_32x32_c, 32, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_16x16_sse2,
-                                       &vpx_highbd_tm_predictor_16x16_c, 16, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_32x32_sse2,
-                                       &vpx_highbd_tm_predictor_32x32_c, 32, 8),
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 8),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16, 8),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 8),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16, 8),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 8)));
+                            make_tuple(&vp9_highbd_dc_predictor_32x32_sse2,
+                                       &vp9_highbd_dc_predictor_32x32_c, 32, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_16x16_sse2,
+                                       &vp9_highbd_tm_predictor_16x16_c, 16, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_32x32_sse2,
+                                       &vp9_highbd_tm_predictor_32x32_c, 32, 8),
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 8),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16, 8),
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 8),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16, 8),
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 8)));
 #else
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_8, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 8),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16, 8),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 8),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16, 8),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 8),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 8)));
-#endif  // !ARCH_X86_64
-
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 8),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16, 8),
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 8),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16, 8),
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 8),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 8)));
+#endif
 #if ARCH_X86_64
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_10, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_32x32_sse2,
-                                       &vpx_highbd_dc_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_dc_predictor_32x32_sse2,
+                                       &vp9_highbd_dc_predictor_32x32_c, 32,
                                       10),
-                            make_tuple(&vpx_highbd_tm_predictor_16x16_sse2,
-                                       &vpx_highbd_tm_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_tm_predictor_16x16_sse2,
+                                       &vp9_highbd_tm_predictor_16x16_c, 16,
                                       10),
-                            make_tuple(&vpx_highbd_tm_predictor_32x32_sse2,
-                                       &vpx_highbd_tm_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_tm_predictor_32x32_sse2,
+                                       &vp9_highbd_tm_predictor_32x32_c, 32,
                                       10),
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 10),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 10),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16,
                                       10),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 10),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 10),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16,
                                       10),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32,
                                       10),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 10)));
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 10)));
 #else
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_10, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 10),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 10),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16,
                                       10),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 10),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16, 10),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32, 10),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 10),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 10)));
-#endif  // !ARCH_X86_64
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 10),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16, 10),
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32, 10),
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 10),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 10)));
+#endif

 #if ARCH_X86_64
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_12, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_32x32_sse2,
-                                       &vpx_highbd_dc_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_dc_predictor_32x32_sse2,
+                                       &vp9_highbd_dc_predictor_32x32_c, 32,
                                       12),
-                            make_tuple(&vpx_highbd_tm_predictor_16x16_sse2,
-                                       &vpx_highbd_tm_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_tm_predictor_16x16_sse2,
+                                       &vp9_highbd_tm_predictor_16x16_c, 16,
                                       12),
-                            make_tuple(&vpx_highbd_tm_predictor_32x32_sse2,
-                                       &vpx_highbd_tm_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_tm_predictor_32x32_sse2,
+                                       &vp9_highbd_tm_predictor_32x32_c, 32,
                                       12),
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 12),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 12),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16,
                                       12),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 12),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 12),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16,
                                       12),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32,
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32,
                                       12),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 12)));
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 12)));
 #else
 INSTANTIATE_TEST_CASE_P(SSE2_TO_C_12, VP9IntraPredTest,
                        ::testing::Values(
-                            make_tuple(&vpx_highbd_dc_predictor_4x4_sse,
-                                       &vpx_highbd_dc_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_dc_predictor_8x8_sse2,
-                                       &vpx_highbd_dc_predictor_8x8_c, 8, 12),
-                            make_tuple(&vpx_highbd_dc_predictor_16x16_sse2,
-                                       &vpx_highbd_dc_predictor_16x16_c, 16,
+                            make_tuple(&vp9_highbd_dc_predictor_4x4_sse,
+                                       &vp9_highbd_dc_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_dc_predictor_8x8_sse2,
+                                       &vp9_highbd_dc_predictor_8x8_c, 8, 12),
+                            make_tuple(&vp9_highbd_dc_predictor_16x16_sse2,
+                                       &vp9_highbd_dc_predictor_16x16_c, 16,
                                       12),
-                            make_tuple(&vpx_highbd_v_predictor_4x4_sse,
-                                       &vpx_highbd_v_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_v_predictor_8x8_sse2,
-                                       &vpx_highbd_v_predictor_8x8_c, 8, 12),
-                            make_tuple(&vpx_highbd_v_predictor_16x16_sse2,
-                                       &vpx_highbd_v_predictor_16x16_c, 16, 12),
-                            make_tuple(&vpx_highbd_v_predictor_32x32_sse2,
-                                       &vpx_highbd_v_predictor_32x32_c, 32, 12),
-                            make_tuple(&vpx_highbd_tm_predictor_4x4_sse,
-                                       &vpx_highbd_tm_predictor_4x4_c, 4, 12),
-                            make_tuple(&vpx_highbd_tm_predictor_8x8_sse2,
-                                       &vpx_highbd_tm_predictor_8x8_c, 8, 12)));
-#endif  // !ARCH_X86_64
-#endif  // CONFIG_USE_X86INC
+                            make_tuple(&vp9_highbd_v_predictor_4x4_sse,
+                                       &vp9_highbd_v_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_v_predictor_8x8_sse2,
+                                       &vp9_highbd_v_predictor_8x8_c, 8, 12),
+                            make_tuple(&vp9_highbd_v_predictor_16x16_sse2,
+                                       &vp9_highbd_v_predictor_16x16_c, 16, 12),
+                            make_tuple(&vp9_highbd_v_predictor_32x32_sse2,
+                                       &vp9_highbd_v_predictor_32x32_c, 32, 12),
+                            make_tuple(&vp9_highbd_tm_predictor_4x4_sse,
+                                       &vp9_highbd_tm_predictor_4x4_c, 4, 12),
+                            make_tuple(&vp9_highbd_tm_predictor_8x8_sse2,
+                                       &vp9_highbd_tm_predictor_8x8_c, 8, 12)));
+#endif
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 #endif  // HAVE_SSE2
 }  // namespace
--- a/test/vp9_lossless_test.cc
+++ b/test/vp9_lossless_test.cc
@@ -7,10 +7,8 @@
 *  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 "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"
@@ -127,8 +125,4 @@ VP9_INSTANTIATE_TEST_CASE(LosslessTest,
                          ::testing::Values(::libvpx_test::kRealTime,
                                            ::libvpx_test::kOnePassGood,
                                            ::libvpx_test::kTwoPassGood));
-
-VP10_INSTANTIATE_TEST_CASE(LosslessTest,
-                           ::testing::Values(::libvpx_test::kOnePassGood,
-                                             ::libvpx_test::kTwoPassGood));
 }  // namespace
--- a/test/vp9_quantize_test.cc
+++ b/test/vp9_quantize_test.cc
@@ -14,15 +14,13 @@

 #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_config.h"
+#include "./vp9_rtcd.h"
 #include "vp9/common/vp9_entropy.h"
-#include "vp9/common/vp9_scan.h"
-#include "vpx/vpx_codec.h"
 #include "vpx/vpx_integer.h"

 using libvpx_test::ACMRandom;
@@ -82,18 +80,18 @@ class VP9Quantize32Test : public ::testing::TestWithParam<QuantizeParam> {

 TEST_P(VP9QuantizeTest, OperationCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff_ptr[256]);
-  DECLARE_ALIGNED(16, int16_t, zbin_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, round_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_shift_ptr[2]);
-  DECLARE_ALIGNED(16, tran_low_t, qcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_qcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_dqcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, int16_t, dequant_ptr[2]);
-  DECLARE_ALIGNED(16, uint16_t, eob_ptr[1]);
-  DECLARE_ALIGNED(16, uint16_t, ref_eob_ptr[1]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, zbin_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, round_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_shift_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, qcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_qcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_dqcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, dequant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, eob_ptr, 1);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_eob_ptr, 1);
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < number_of_iterations; ++i) {
@@ -141,18 +139,18 @@ TEST_P(VP9QuantizeTest, OperationCheck) {

 TEST_P(VP9Quantize32Test, OperationCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff_ptr[1024]);
-  DECLARE_ALIGNED(16, int16_t, zbin_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, round_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_shift_ptr[2]);
-  DECLARE_ALIGNED(16, tran_low_t, qcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_qcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_dqcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, int16_t, dequant_ptr[2]);
-  DECLARE_ALIGNED(16, uint16_t, eob_ptr[1]);
-  DECLARE_ALIGNED(16, uint16_t, ref_eob_ptr[1]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, zbin_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, round_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_shift_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, qcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_qcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_dqcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, dequant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, eob_ptr, 1);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_eob_ptr, 1);
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < number_of_iterations; ++i) {
@@ -200,18 +198,18 @@ TEST_P(VP9Quantize32Test, OperationCheck) {

 TEST_P(VP9QuantizeTest, EOBCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff_ptr[256]);
-  DECLARE_ALIGNED(16, int16_t, zbin_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, round_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_shift_ptr[2]);
-  DECLARE_ALIGNED(16, tran_low_t, qcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_qcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_dqcoeff_ptr[256]);
-  DECLARE_ALIGNED(16, int16_t, dequant_ptr[2]);
-  DECLARE_ALIGNED(16, uint16_t, eob_ptr[1]);
-  DECLARE_ALIGNED(16, uint16_t, ref_eob_ptr[1]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, zbin_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, round_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_shift_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, qcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_qcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_dqcoeff_ptr, 256);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, dequant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, eob_ptr, 1);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_eob_ptr, 1);
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < number_of_iterations; ++i) {
@@ -264,18 +262,18 @@ TEST_P(VP9QuantizeTest, EOBCheck) {

 TEST_P(VP9Quantize32Test, EOBCheck) {
  ACMRandom rnd(ACMRandom::DeterministicSeed());
-  DECLARE_ALIGNED(16, tran_low_t, coeff_ptr[1024]);
-  DECLARE_ALIGNED(16, int16_t, zbin_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, round_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_ptr[2]);
-  DECLARE_ALIGNED(16, int16_t, quant_shift_ptr[2]);
-  DECLARE_ALIGNED(16, tran_low_t, qcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, dqcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_qcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, tran_low_t, ref_dqcoeff_ptr[1024]);
-  DECLARE_ALIGNED(16, int16_t, dequant_ptr[2]);
-  DECLARE_ALIGNED(16, uint16_t, eob_ptr[1]);
-  DECLARE_ALIGNED(16, uint16_t, ref_eob_ptr[1]);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, coeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, zbin_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, round_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, quant_shift_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, qcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, dqcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_qcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, tran_low_t, ref_dqcoeff_ptr, 1024);
+  DECLARE_ALIGNED_ARRAY(16, int16_t, dequant_ptr, 2);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, eob_ptr, 1);
+  DECLARE_ALIGNED_ARRAY(16, uint16_t, ref_eob_ptr, 1);
  int err_count_total = 0;
  int first_failure = -1;
  for (int i = 0; i < number_of_iterations; ++i) {
@@ -331,21 +329,21 @@ using std::tr1::make_tuple;
 INSTANTIATE_TEST_CASE_P(
    SSE2, VP9QuantizeTest,
    ::testing::Values(
-        make_tuple(&vpx_highbd_quantize_b_sse2,
-                   &vpx_highbd_quantize_b_c, VPX_BITS_8),
-        make_tuple(&vpx_highbd_quantize_b_sse2,
-                   &vpx_highbd_quantize_b_c, VPX_BITS_10),
-        make_tuple(&vpx_highbd_quantize_b_sse2,
-                   &vpx_highbd_quantize_b_c, VPX_BITS_12)));
+        make_tuple(&vp9_highbd_quantize_b_sse2,
+                   &vp9_highbd_quantize_b_c, VPX_BITS_8),
+        make_tuple(&vp9_highbd_quantize_b_sse2,
+                   &vp9_highbd_quantize_b_c, VPX_BITS_10),
+        make_tuple(&vp9_highbd_quantize_b_sse2,
+                   &vp9_highbd_quantize_b_c, VPX_BITS_12)));
 INSTANTIATE_TEST_CASE_P(
    SSE2, VP9Quantize32Test,
    ::testing::Values(
-        make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
-                   &vpx_highbd_quantize_b_32x32_c, VPX_BITS_8),
-        make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
-                   &vpx_highbd_quantize_b_32x32_c, VPX_BITS_10),
-        make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
-                   &vpx_highbd_quantize_b_32x32_c, VPX_BITS_12)));
+        make_tuple(&vp9_highbd_quantize_b_32x32_sse2,
+                   &vp9_highbd_quantize_b_32x32_c, VPX_BITS_8),
+        make_tuple(&vp9_highbd_quantize_b_32x32_sse2,
+                   &vp9_highbd_quantize_b_32x32_c, VPX_BITS_10),
+        make_tuple(&vp9_highbd_quantize_b_32x32_sse2,
+                   &vp9_highbd_quantize_b_32x32_c, VPX_BITS_12)));
 #endif  // HAVE_SSE2
 #endif  // CONFIG_VP9_HIGHBITDEPTH
 }  // namespace
--- a/test/vp9_skip_loopfilter_test.cc
+++ b/test/vp9_skip_loopfilter_test.cc
@@ -1,180 +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 <string>
-
-#include "test/codec_factory.h"
-#include "test/decode_test_driver.h"
-#include "test/md5_helper.h"
-#include "test/util.h"
-#include "test/webm_video_source.h"
-
-namespace {
-
-const char kVp9TestFile[] = "vp90-2-08-tile_1x8_frame_parallel.webm";
-const char kVp9Md5File[] = "vp90-2-08-tile_1x8_frame_parallel.webm.md5";
-
-// Class for testing shutting off the loop filter.
-class SkipLoopFilterTest {
- public:
-  SkipLoopFilterTest()
-      : video_(NULL),
-        decoder_(NULL),
-        md5_file_(NULL) {}
-
-  ~SkipLoopFilterTest() {
-    if (md5_file_ != NULL)
-      fclose(md5_file_);
-    delete decoder_;
-    delete video_;
-  }
-
-  // If |threads| > 0 then set the decoder with that number of threads.
-  void Init(int num_threads) {
-    expected_md5_[0] = '\0';
-    junk_[0] = '\0';
-    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();
-    if (num_threads > 0)
-      cfg.threads = num_threads;
-    decoder_ = new libvpx_test::VP9Decoder(cfg, 0);
-    ASSERT_TRUE(decoder_ != NULL);
-
-    OpenMd5File(kVp9Md5File);
-  }
-
-  // Set the VP9 skipLoopFilter control value.
-  void SetSkipLoopFilter(int value, vpx_codec_err_t expected_value) {
-    decoder_->Control(VP9_SET_SKIP_LOOP_FILTER, value, expected_value);
-  }
-
-  vpx_codec_err_t DecodeOneFrame() {
-    const vpx_codec_err_t res =
-        decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
-    if (res == VPX_CODEC_OK) {
-      ReadMd5();
-      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;
-      ReadMd5();
-    }
-    return VPX_CODEC_OK;
-  }
-
-  // Checks if MD5 matches or doesn't.
-  void CheckMd5(bool matches) {
-    libvpx_test::DxDataIterator dec_iter = decoder_->GetDxData();
-    const vpx_image_t *img = dec_iter.Next();
-    CheckMd5Vpx(*img, matches);
-  }
-
- private:
-  // 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;
-  }
-
-  // Reads the next line of the MD5 file.
-  void ReadMd5() {
-    ASSERT_TRUE(md5_file_ != NULL);
-    const int res = fscanf(md5_file_, "%s  %s", expected_md5_, junk_);
-    ASSERT_NE(EOF, res) << "Read md5 data failed";
-    expected_md5_[32] = '\0';
-  }
-
-  // Checks if the last read MD5 matches |img| or doesn't.
-  void CheckMd5Vpx(const vpx_image_t &img, bool matches) {
-    ::libvpx_test::MD5 md5_res;
-    md5_res.Add(&img);
-    const char *const actual_md5 = md5_res.Get();
-
-    // Check MD5.
-    if (matches)
-      ASSERT_STREQ(expected_md5_, actual_md5) << "MD5 checksums don't match";
-    else
-      ASSERT_STRNE(expected_md5_, actual_md5) << "MD5 checksums match";
-  }
-
-  libvpx_test::WebMVideoSource *video_;
-  libvpx_test::VP9Decoder *decoder_;
-  FILE *md5_file_;
-  char expected_md5_[33];
-  char junk_[128];
-};
-
-TEST(SkipLoopFilterTest, ShutOffLoopFilter) {
-  const int non_zero_value = 1;
-  const int num_threads = 0;
-  SkipLoopFilterTest skip_loop_filter;
-  skip_loop_filter.Init(num_threads);
-  skip_loop_filter.SetSkipLoopFilter(non_zero_value, VPX_CODEC_OK);
-  ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeRemainingFrames());
-  skip_loop_filter.CheckMd5(false);
-}
-
-TEST(SkipLoopFilterTest, ShutOffLoopFilterSingleThread) {
-  const int non_zero_value = 1;
-  const int num_threads = 1;
-  SkipLoopFilterTest skip_loop_filter;
-  skip_loop_filter.Init(num_threads);
-  skip_loop_filter.SetSkipLoopFilter(non_zero_value, VPX_CODEC_OK);
-  ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeRemainingFrames());
-  skip_loop_filter.CheckMd5(false);
-}
-
-TEST(SkipLoopFilterTest, ShutOffLoopFilter8Threads) {
-  const int non_zero_value = 1;
-  const int num_threads = 8;
-  SkipLoopFilterTest skip_loop_filter;
-  skip_loop_filter.Init(num_threads);
-  skip_loop_filter.SetSkipLoopFilter(non_zero_value, VPX_CODEC_OK);
-  ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeRemainingFrames());
-  skip_loop_filter.CheckMd5(false);
-}
-
-TEST(SkipLoopFilterTest, WithLoopFilter) {
-  const int non_zero_value = 1;
-  const int num_threads = 0;
-  SkipLoopFilterTest skip_loop_filter;
-  skip_loop_filter.Init(num_threads);
-  skip_loop_filter.SetSkipLoopFilter(non_zero_value, VPX_CODEC_OK);
-  skip_loop_filter.SetSkipLoopFilter(0, VPX_CODEC_OK);
-  ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeRemainingFrames());
-  skip_loop_filter.CheckMd5(true);
-}
-
-TEST(SkipLoopFilterTest, ToggleLoopFilter) {
-  const int num_threads = 0;
-  SkipLoopFilterTest skip_loop_filter;
-  skip_loop_filter.Init(num_threads);
-
-  for (int i = 0; i < 10; ++i) {
-    skip_loop_filter.SetSkipLoopFilter(i % 2, VPX_CODEC_OK);
-    ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeOneFrame());
-  }
-  ASSERT_EQ(VPX_CODEC_OK, skip_loop_filter.DecodeRemainingFrames());
-  skip_loop_filter.CheckMd5(false);
-}
-
-}  // namespace
--- a/test/vp9_subtract_test.cc
+++ b/test/vp9_subtract_test.cc
@@ -9,13 +9,11 @@
 */

 #include "third_party/googletest/src/include/gtest/gtest.h"
-
-#include "./vp9_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 "./vpx_config.h"
+#include "./vp9_rtcd.h"
 #include "vp9/common/vp9_blockd.h"
 #include "vpx_mem/vpx_mem.h"

@@ -91,19 +89,15 @@ TEST_P(VP9SubtractBlockTest, SimpleSubtract) {
 }

 INSTANTIATE_TEST_CASE_P(C, VP9SubtractBlockTest,
-                        ::testing::Values(vpx_subtract_block_c));
+                        ::testing::Values(vp9_subtract_block_c));

 #if HAVE_SSE2 && CONFIG_USE_X86INC
 INSTANTIATE_TEST_CASE_P(SSE2, VP9SubtractBlockTest,
-                        ::testing::Values(vpx_subtract_block_sse2));
+                        ::testing::Values(vp9_subtract_block_sse2));
 #endif
 #if HAVE_NEON
 INSTANTIATE_TEST_CASE_P(NEON, VP9SubtractBlockTest,
-                        ::testing::Values(vpx_subtract_block_neon));
-#endif
-#if HAVE_MSA
-INSTANTIATE_TEST_CASE_P(MSA, VP9SubtractBlockTest,
-                        ::testing::Values(vpx_subtract_block_msa));
+                        ::testing::Values(vp9_subtract_block_neon));
 #endif

 }  // namespace vp9
--- a/test/vp9_thread_test.cc
+++ b/test/vp9_thread_test.cc
@@ -18,33 +18,33 @@
 #if CONFIG_WEBM_IO
 #include "test/webm_video_source.h"
 #endif
-#include "vpx_util/vpx_thread.h"
+#include "vp9/common/vp9_thread.h"

 namespace {

 using std::string;

-class VPxWorkerThreadTest : public ::testing::TestWithParam<bool> {
+class VP9WorkerThreadTest : public ::testing::TestWithParam<bool> {
 protected:
-  virtual ~VPxWorkerThreadTest() {}
+  virtual ~VP9WorkerThreadTest() {}
  virtual void SetUp() {
-    vpx_get_worker_interface()->init(&worker_);
+    vp9_get_worker_interface()->init(&worker_);
  }

  virtual void TearDown() {
-    vpx_get_worker_interface()->end(&worker_);
+    vp9_get_worker_interface()->end(&worker_);
  }

-  void Run(VPxWorker* worker) {
+  void Run(VP9Worker* worker) {
    const bool synchronous = GetParam();
    if (synchronous) {
-      vpx_get_worker_interface()->execute(worker);
+      vp9_get_worker_interface()->execute(worker);
    } else {
-      vpx_get_worker_interface()->launch(worker);
+      vp9_get_worker_interface()->launch(worker);
    }
  }

-  VPxWorker worker_;
+  VP9Worker worker_;
 };

 int ThreadHook(void* data, void* return_value) {
@@ -53,12 +53,12 @@ int ThreadHook(void* data, void* return_value) {
  return *reinterpret_cast<int*>(return_value);
 }

-TEST_P(VPxWorkerThreadTest, HookSuccess) {
+TEST_P(VP9WorkerThreadTest, HookSuccess) {
  // should be a no-op.
-  EXPECT_NE(vpx_get_worker_interface()->sync(&worker_), 0);
+  EXPECT_NE(vp9_get_worker_interface()->sync(&worker_), 0);

  for (int i = 0; i < 2; ++i) {
-    EXPECT_NE(vpx_get_worker_interface()->reset(&worker_), 0);
+    EXPECT_NE(vp9_get_worker_interface()->reset(&worker_), 0);

    int hook_data = 0;
    int return_value = 1;  // return successfully from the hook
@@ -67,17 +67,17 @@ TEST_P(VPxWorkerThreadTest, HookSuccess) {
    worker_.data2 = &return_value;

    Run(&worker_);
-    EXPECT_NE(vpx_get_worker_interface()->sync(&worker_), 0);
+    EXPECT_NE(vp9_get_worker_interface()->sync(&worker_), 0);
    EXPECT_FALSE(worker_.had_error);
    EXPECT_EQ(5, hook_data);

    // should be a no-op.
-    EXPECT_NE(vpx_get_worker_interface()->sync(&worker_), 0);
+    EXPECT_NE(vp9_get_worker_interface()->sync(&worker_), 0);
  }
 }

-TEST_P(VPxWorkerThreadTest, HookFailure) {
-  EXPECT_NE(vpx_get_worker_interface()->reset(&worker_), 0);
+TEST_P(VP9WorkerThreadTest, HookFailure) {
+  EXPECT_NE(vp9_get_worker_interface()->reset(&worker_), 0);

  int hook_data = 0;
  int return_value = 0;  // return failure from the hook
@@ -86,29 +86,29 @@ TEST_P(VPxWorkerThreadTest, HookFailure) {
  worker_.data2 = &return_value;

  Run(&worker_);
-  EXPECT_FALSE(vpx_get_worker_interface()->sync(&worker_));
+  EXPECT_FALSE(vp9_get_worker_interface()->sync(&worker_));
  EXPECT_EQ(1, worker_.had_error);

  // Ensure _reset() clears the error and _launch() can be called again.
  return_value = 1;
-  EXPECT_NE(vpx_get_worker_interface()->reset(&worker_), 0);
+  EXPECT_NE(vp9_get_worker_interface()->reset(&worker_), 0);
  EXPECT_FALSE(worker_.had_error);
-  vpx_get_worker_interface()->launch(&worker_);
-  EXPECT_NE(vpx_get_worker_interface()->sync(&worker_), 0);
+  vp9_get_worker_interface()->launch(&worker_);
+  EXPECT_NE(vp9_get_worker_interface()->sync(&worker_), 0);
  EXPECT_FALSE(worker_.had_error);
 }

-TEST_P(VPxWorkerThreadTest, EndWithoutSync) {
+TEST_P(VP9WorkerThreadTest, EndWithoutSync) {
  // Create a large number of threads to increase the chances of detecting a
  // race. Doing more work in the hook is no guarantee as any race would occur
  // post hook execution in the main thread loop driver.
  static const int kNumWorkers = 64;
-  VPxWorker workers[kNumWorkers];
+  VP9Worker workers[kNumWorkers];
  int hook_data[kNumWorkers];
  int return_value[kNumWorkers];

  for (int n = 0; n < kNumWorkers; ++n) {
-    vpx_get_worker_interface()->init(&workers[n]);
+    vp9_get_worker_interface()->init(&workers[n]);
    return_value[n] = 1;  // return successfully from the hook
    workers[n].hook = ThreadHook;
    workers[n].data1 = &hook_data[n];
@@ -117,7 +117,7 @@ TEST_P(VPxWorkerThreadTest, EndWithoutSync) {

  for (int i = 0; i < 2; ++i) {
    for (int n = 0; n < kNumWorkers; ++n) {
-      EXPECT_NE(vpx_get_worker_interface()->reset(&workers[n]), 0);
+      EXPECT_NE(vp9_get_worker_interface()->reset(&workers[n]), 0);
      hook_data[n] = 0;
    }

@@ -126,16 +126,16 @@ TEST_P(VPxWorkerThreadTest, EndWithoutSync) {
    }

    for (int n = kNumWorkers - 1; n >= 0; --n) {
-      vpx_get_worker_interface()->end(&workers[n]);
+      vp9_get_worker_interface()->end(&workers[n]);
    }
  }
 }

-TEST(VPxWorkerThreadTest, TestInterfaceAPI) {
-  EXPECT_EQ(0, vpx_set_worker_interface(NULL));
-  EXPECT_TRUE(vpx_get_worker_interface() != NULL);
+TEST(VP9WorkerThreadTest, TestInterfaceAPI) {
+  EXPECT_EQ(0, vp9_set_worker_interface(NULL));
+  EXPECT_TRUE(vp9_get_worker_interface() != NULL);
  for (int i = 0; i < 6; ++i) {
-    VPxWorkerInterface winterface = *vpx_get_worker_interface();
+    VP9WorkerInterface winterface = *vp9_get_worker_interface();
    switch (i) {
      default:
      case 0: winterface.init = NULL; break;
@@ -145,7 +145,7 @@ TEST(VPxWorkerThreadTest, TestInterfaceAPI) {
      case 4: winterface.execute = NULL; break;
      case 5: winterface.end = NULL; break;
    }
-    EXPECT_EQ(0, vpx_set_worker_interface(&winterface));
+    EXPECT_EQ(0, vp9_set_worker_interface(&winterface));
  }
 }

@@ -190,7 +190,7 @@ string DecodeFile(const string& filename, int num_threads) {
 void DecodeFiles(const FileList files[]) {
  for (const FileList *iter = files; iter->name != NULL; ++iter) {
    SCOPED_TRACE(iter->name);
-    for (int t = 1; t <= 8; ++t) {
+    for (int t = 2; t <= 8; ++t) {
      EXPECT_EQ(iter->expected_md5, DecodeFile(iter->name, t))
          << "threads = " << t;
    }
@@ -202,21 +202,21 @@ void DecodeFiles(const FileList files[]) {
 // hang.
 namespace impl {

-void Init(VPxWorker *const worker) { memset(worker, 0, sizeof(*worker)); }
-int Reset(VPxWorker *const /*worker*/) { return 1; }
-int Sync(VPxWorker *const worker) { return !worker->had_error; }
+void Init(VP9Worker *const worker) { memset(worker, 0, sizeof(*worker)); }
+int Reset(VP9Worker *const /*worker*/) { return 1; }
+int Sync(VP9Worker *const worker) { return !worker->had_error; }

-void Execute(VPxWorker *const worker) {
+void Execute(VP9Worker *const worker) {
  worker->had_error |= !worker->hook(worker->data1, worker->data2);
 }

-void Launch(VPxWorker *const worker) { Execute(worker); }
-void End(VPxWorker *const /*worker*/) {}
+void Launch(VP9Worker *const worker) { Execute(worker); }
+void End(VP9Worker *const /*worker*/) {}

 }  // namespace impl

-TEST(VPxWorkerThreadTest, TestSerialInterface) {
-  static const VPxWorkerInterface serial_interface = {
+TEST(VP9WorkerThreadTest, TestSerialInterface) {
+  static const VP9WorkerInterface serial_interface = {
    impl::Init, impl::Reset, impl::Sync, impl::Launch, impl::Execute, impl::End
  };
  // TODO(jzern): Avoid using a file that will use the row-based thread
@@ -225,23 +225,23 @@ TEST(VPxWorkerThreadTest, TestSerialInterface) {
  // progress in the row above before proceeding.
  static const char expected_md5[] = "b35a1b707b28e82be025d960aba039bc";
  static const char filename[] = "vp90-2-03-size-226x226.webm";
-  VPxWorkerInterface default_interface = *vpx_get_worker_interface();
+  VP9WorkerInterface default_interface = *vp9_get_worker_interface();

-  EXPECT_NE(vpx_set_worker_interface(&serial_interface), 0);
+  EXPECT_NE(vp9_set_worker_interface(&serial_interface), 0);
  EXPECT_EQ(expected_md5, DecodeFile(filename, 2));

  // Reset the interface.
-  EXPECT_NE(vpx_set_worker_interface(&default_interface), 0);
+  EXPECT_NE(vp9_set_worker_interface(&default_interface), 0);
  EXPECT_EQ(expected_md5, DecodeFile(filename, 2));
 }

-TEST(VP9DecodeMultiThreadedTest, NoTilesNonFrameParallel) {
+TEST(VP9DecodeMultiThreadedTest, Decode) {
  // no tiles or frame parallel; this exercises loop filter threading.
  EXPECT_EQ("b35a1b707b28e82be025d960aba039bc",
            DecodeFile("vp90-2-03-size-226x226.webm", 2));
 }

-TEST(VP9DecodeMultiThreadedTest, FrameParallel) {
+TEST(VP9DecodeMultiThreadedTest, Decode2) {
  static const FileList files[] = {
    { "vp90-2-08-tile_1x2_frame_parallel.webm",
      "68ede6abd66bae0a2edf2eb9232241b6" },
@@ -255,7 +255,8 @@ TEST(VP9DecodeMultiThreadedTest, FrameParallel) {
  DecodeFiles(files);
 }

-TEST(VP9DecodeMultiThreadedTest, FrameParallelResize) {
+// Test tile quantity changes within one file.
+TEST(VP9DecodeMultiThreadedTest, Decode3) {
  static const FileList files[] = {
    { "vp90-2-14-resize-fp-tiles-1-16.webm",
      "0cd5e632c326297e975f38949c31ea94" },
@@ -306,21 +307,8 @@ TEST(VP9DecodeMultiThreadedTest, FrameParallelResize) {

  DecodeFiles(files);
 }
-
-TEST(VP9DecodeMultiThreadedTest, NonFrameParallel) {
-  static const FileList files[] = {
-    { "vp90-2-08-tile_1x2.webm", "570b4a5d5a70d58b5359671668328a16" },
-    { "vp90-2-08-tile_1x4.webm", "988d86049e884c66909d2d163a09841a" },
-    { "vp90-2-08-tile_1x8.webm", "0941902a52e9092cb010905eab16364c" },
-    { "vp90-2-08-tile-4x1.webm", "06505aade6647c583c8e00a2f582266f" },
-    { "vp90-2-08-tile-4x4.webm", "85c2299892460d76e2c600502d52bfe2" },
-    { NULL, NULL }
-  };
-
-  DecodeFiles(files);
-}
 #endif  // CONFIG_WEBM_IO

-INSTANTIATE_TEST_CASE_P(Synchronous, VPxWorkerThreadTest, ::testing::Bool());
+INSTANTIATE_TEST_CASE_P(Synchronous, VP9WorkerThreadTest, ::testing::Bool());

 }  // namespace
--- a/test/vpx_scale_test.cc
+++ b/test/vpx_scale_test.cc
@@ -10,10 +10,11 @@

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

-#include "./vpx_config.h"
-#include "./vpx_scale_rtcd.h"
 #include "test/clear_system_state.h"
 #include "test/register_state_check.h"
+
+#include "./vpx_config.h"
+#include "./vpx_scale_rtcd.h"
 #include "vpx_mem/vpx_mem.h"
 #include "vpx_scale/yv12config.h"

--- a/Show More
+++ b/Show More