Previously, the microsoft arm assembler errored out, saying that
bilinear_taps_coeff was an undefined symbol.
Change-Id: Ib938f0b454c41ccbd801e70a7c9acc0fa04e3c55
Allows building the library with the gcc -pedantic option, for improved
portabilty. In particular, this commit removes usage of C99/C++ style
single-line comments and dynamic struct initializers. This is a
continuation of the work done in commit 97b766a46, which removed most
of these warnings for decode only builds.
Change-Id: Id453d9c1d9f44cc0381b10c3869fabb0184d5966
Besides imposing a performance penalty at startup in most
configurations, these relocations break the dynamic linker for
native Fennec, since it does not support them at all.
Change-Id: Id5dc768609354ebb4379966eb61a7313e6fd18de
This is a proof of concept RTCD implementation to replace the current
system of nested includes, prototypes, INVOKE macros, etc. Currently
only the decoder specific functions are implemented in the new system.
Additional functions will be added in subsequent commits.
Overview:
RTCD "functions" are implemented as either a global function pointer
or a macro (when only one eligible specialization available).
Functions which have RTCD specializations are listed using a simple
DSL identifying the function's base name, its prototype, and the
architecture extensions that specializations are available for.
Advantages over the old system:
- No INVOKE macros. A call to an RTCD function looks like an ordinary
function call.
- No need to pass vtables around.
- If there is only one eligible function to call, the function is
called directly, rather than indirecting through a function pointer.
- Supports the notion of "required" extensions, so in combination with
the above, on x86_64 if the best function available is sse2 or lower
it will be called directly, since all x86_64 platforms implement
sse2.
- Elides all references to functions which will never be called, which
could reduce binary size. For example if sse2 is required and there
are both mmx and sse2 implementations of a certain function, the
code will have no link time references to the mmx code.
- Significantly easier to add a new function, just one file to edit.
Disadvantages:
- Requires global writable data (though this is not a new requirement)
- 1 new generated source file.
Change-Id: Iae6edab65315f79c168485c96872641c5aa09d55
A processor with ARMv7 instructions does not
necessarily have NEON dsp extensions. This CL
has the added side effect of allowing the ability
to enable/disable the dsp extensions cleanly.
Change-Id: Ie1e879b8fe131885bc3d4138a0acc9ffe73a36df
This patch removes the local copies of the dequantize
constants and implements John's idea as described
in "Make a local copy of the dequantized data" commit.
Change-Id: Ic6b7d681f00bf63263f71ff1e39ab2f80729e8b2
These functions are now used by the encoder.
This is WIP with the goal of creating a common idct/add for
the encoder and decoder. A boost of 1.8% was seen for
the HD rt test clip used.
[Tero] Added needed changes to ARM side.
Change-Id: Ibbb8000be09034203d7adffc457d3c3f8b06a5bf
to the dqcoeff or qcoeff buffer. The encoder would
populate the dc coeffs of the y blocks as a separate
stage (recon_dcblock) and the decoder would use a special
version of the idct. This change eliminates the extra copy
and reduces the code footprint.
[Tero] Added needed changes to armv6 and NEON assembly.
Change-Id: I83202ffdbaf83f6e5dd69f4ba2519fcf0b13b3ba
Added ARM optimized intra 4x4 prediction
- 2x faster on Profiler compared to C-code compiled with -O3
- Function interface changed a little to improve BLOCKD structure
access
Change-Id: I9bc2b723155943fe0cf03dd9ca5f1760f7a81f54
Instead of using the predict buffer, the decoder now writes
the predictor into the recon buffer. For blocks with eob=0,
unnecessary idcts can be eliminated. This gave a performance
boost of ~1.8% for the HD clips used.
Tero: Added needed changes to ARM side and scheduled some
assembly code to prevent interlocks.
Patch Set 6: Merged (I1bcdca7a95aacc3a181b9faa6b10e3a71ee24df3)
into this commit because of similarities in the idct
functions.
Patch Set 7: EC bug fix.
Change-Id: Ie31d90b5d3522e1108163f2ac491e455e3f955e6
The current code stores pointers to coefficient tables and loads them to
access the tables contents. As these pointers are stored in the code
sections, it means we end up with text relocations. eu-findtextrel will
thus complain about code not compiled with -fpic/-fPIC.
Since the pointers are stored in the code sections, we can actually cheat
and let the assembler generate relative addressing when accessing the
coefficient tables, and just load their location with adr.
Change-Id: Ib74ae2d3f2bab80b29991355f2dbe6955f38f6ae
rvct 4.1 was complaining about vstmia.16, store multiple expects 64 data type.
optimized the implementation.
Change-Id: I0701052cabd685c375637bbc3796ff6d88f5972c
the decision to run the regular or simple loopfilter is made outside the
function and managed with pointers
stop tracking the option in two places. use filter_type exclusively
Change-Id: I39d7b5d1352885efc632c0a94aaf56b72cc2fe15
The vp8_build_intra_predictors_mby and vp8_build_intra_predictors_mby_s
functions had global function pointers rather than using the RTCD
framework. This can show up as a potential data race with tools such as
helgrind. See https://bugzilla.mozilla.org/show_bug.cgi?id=640935
for an example.
Change-Id: I29c407f828ac2bddfc039f852f138de5de888534
make reference version of bilinear_filters short.
use reference versions of bilinear_filters and sub_pel_filters when
possible.
recognize that Width was being passed into
filter_block2d_bil_first_pass multiple times. ARM version had already
fixed this. propegate to C.
change references to src_pixels_per_line to src_pitch and standardize on
src/dst (instead of input/output).
recognize that first_pass is only run in the verticle and second_pass
only horizontal. ARM version had already fixed this. propegate to C
Change-Id: I292d376d239a9a7ca37ec2bf03cc0720606983e2
common/arm/vpx_asm_offsets moves up a level. prepare for muxing with
encoder/arm/vpx_vp8_enc_asm_offsets
Change-Id: I89a04a5235447e66571995c9d9b4b6edcb038e24
Adds following targets to configure script to support RVCT compilation
without operating system support (for Profiler or bare metal images).
- armv5te-none-rvct
- armv6-none-rvct
- armv7-none-rvct
To strip OS specific parts from the code "os_support"-config was added
to script and CONFIG_OS_SUPPORT flag is used in the code to exclude OS
specific parts such as OS specific includes and function calls for
timers and threads etc. This was done to enable RVCT compilation for
profiling purposes or running the image on bare metal target with
Lauterbach.
Removed separate AREA directives for READONLY data in armv6 and neon
assembly files to fix the RVCT compilation. Otherwise
"ldr <reg>, =label" syntax would have been needed to prevent linker
errors. This syntax is not supported by older gnu assemblers.
Change-Id: I14f4c68529e8c27397502fbc3010a54e505ddb43
This eliminates a large set of warnings exposed by the Mozilla build
system (Use of C++ comments in ISO C90 source, commas at the end of
enum lists, a couple incomplete initializers, and signed/unsigned
comparisons).
It also eliminates many (but not all) of the warnings expose by newer
GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite
without checking the return values).
There are a few spurious warnings left on my system:
../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used
uninitialized in this function
gcc seems to be unable to figure out that the value shortcut doesn't
change between the two if blocks that test it here.
../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned
expression >= 0 is always true
../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned
expression >= 0 is always true
This is true, so far as it goes, but it's comparing against an enum, and the C
standard does not mandate that enums be unsigned, so the checks can't be
removed.
Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
ARM NEON has a platform specific version of vp8_recon16x16mb, though
it's just a stub to extract the various parameters from the
MACROBLOCKD struct and pass them to vp8_recon16x16mb_neon(). Using
that function's prototype directly will be a better long term solution,
but it's quite an invasive change.
Change-Id: I04273149e2ade34749e2d09e7edb0c396e1dd620
Some of the ARM functions differed from their generic counterparts
only by unrolling their loops. Since this change may be useful
on other platforms, or might even supercede the looped version
in the generic case, move it back to the generic file.
This code is left under #if ARCH_ARM for now, but it may be worth
considering a different (possibly new) conditional for these. If
it turns out that this should be runtime selectable, these
functions will have to move to the RTCD infrastructure. Don't want
to take that step at this time without more profile data.
Change-Id: I4612fdbc606fbebba4971a690fb743ad184ff15f
These functions were true duplicates of functions present in the
generic code. This fixes some of the link errors when building
with --enable-shared --enable-pic.
Change-Id: Idff26599d510d954e439207883607ad6b74df20c
there were four versions for the regular and
macroblock loopfilters:
horizontal [y|uv]
vertical [y|uv]
this moves all the common code into 2 functions:
vp8_loop_filter_neon
vp8_mbloop_filter_neon
this provides no gain in performance. there's a bit
of jitter, but it trends down ~0.25-0.5%. however,
this is a huge gain maintenance. also, there is the
potential to drop some stack usage in the macroblock
loopfilter.
Change-Id: I91506f07d2f449631ff67ad6f1b3f3be63b81a92
The primary goal is to allow a binary to be built which supports
NEON, but can fall back to non-NEON routines, since some Android
devices do not have NEON, even if they are otherwise ARMv7 (e.g.,
Tegra).
The configure-generated flags HAVE_ARMV7, etc., are used to decide
which versions of each function to build, and when
CONFIG_RUNTIME_CPU_DETECT is enabled, the correct version is chosen
at run time.
In order for this to work, the CFLAGS must be set to something
appropriate (e.g., without -mfpu=neon for ARMv7, and with
appropriate -march and -mcpu for even earlier configurations), or
the native C code will not be able to run.
The ASFLAGS must remain set for the most advanced instruction set
required at build time, since the ARM assembler will refuse to emit
them otherwise.
I have not attempted to make any changes to configure to do this
automatically.
Doing so will probably require the addition of new configure options.
Many of the hooks for RTCD on ARM were already there, but a lot of
the code had bit-rotted, and a good deal of the ARM-specific code
is not integrated into the RTCD structs at all.
I did not try to resolve the latter, merely to add the minimal amount
of protection around them to allow RTCD to work.
Those functions that were called based on an ifdef at the calling
site were expanded to check the RTCD flags at that site, but they
should be added to an RTCD struct somewhere in the future.
The functions invoked with global function pointers still are, but
these should be moved into an RTCD struct for thread safety (I
believe every platform currently supported has atomic pointer
stores, but this is not guaranteed).
The encoder's boolhuff functions did not even have _c and armv7
suffixes, and the correct version was resolved at link time.
The token packing functions did have appropriate suffixes, but the
version was selected with a define, with no associated RTCD struct.
However, for both of these, the only armv7 instruction they actually
used was rbit, and this was completely superfluous, so I reworked
them to avoid it.
The only non-ARMv4 instruction remaining in them is clz, which is
ARMv5 (not even ARMv5TE is required).
Considering that there are no ARM-specific configs which are not at
least ARMv5TE, I did not try to detect these at runtime, and simply
enable them for ARMv5 and above.
Finally, the NEON register saving code was completely non-reentrant,
since it saved the registers to a global, static variable.
I moved the storage for this onto the stack.
A single binary built with this code was tested on an ARM11 (ARMv6)
and a Cortex A8 (ARMv7 w/NEON), for both the encoder and decoder,
and produced identical output, while using the correct accelerated
functions on each.
I did not test on any earlier processors.
Change-Id: I45cbd63a614f4554c3b325c45d46c0806f009eaa
The existing code applied a 6-tap filter with 0's on either end.
We're already paying the branch penalty to avoid computing the two
extra columns needed as input to this filter.
We might as well save time computing the filter as well.
This reduces the inner loop from 21 instructions to 16, the number
of loads per iteration from 4 to 1, and the number of multiplies
from 7 to 4.
The gain in overall decoding performance, however, is small (less
than 1%).
This change also means we now valgrind clean on ARMv6, which is
its real purpose.
The errors reported here were valgrind's fault (it does not detect
that 0 times an uninitialized value is initialized), but Julian
Seward says it would slow down valgrind considerably to make such
checks.
Speeding up libvpx rather, even by a small amount, seems a much
better idea if only to enable proper valgrind checking of the
rest of the codec.
Change-Id: Ifb376ea195e086b60f61daf1097d8910c4d8ff16
This function was accessing values below the stack pointer, which
can be corrupted by signal delivery at any time.
Change-Id: I92945b30817562eb0340f289e74c108da72aeaca
previous implementation compared each set of values to limit and then
&'d them together, requiring a compare and & for each value.
this does the accumulation first, requiring only one compare
Change-Id: Ia5e3a1a50e47699c88470b8c41964f92a0dc1323
Changes 'The VP8 project' to 'The WebM project', for consistency
with other webmproject.org repositories.
Fixes issue #97.
Change-Id: I37c13ed5fbdb9d334ceef71c6350e9febed9bbba
Remove the dependency on postproc.c for the encoder in general, the only
unchecked need for it is when CONFIG_PSNR is enabled. All other cases
are already wrapped in CONFIG_POSTPROC. In the CONFIG_PSNR case the file
will still be included.
Additionally, when VP8_SET_POSTPROC is used with the encoder when post
processing has been disabled an error will be returned.
This addresses issue #153.
Change-Id: Ia6dfe20167f7077734a6058cbd1d794550346089
move some things around, reorder some instructions
constant 0 is used several times. load it once per call in horiz,
once per loop in vert.
separate saturating instructions to avoid stalls.
just use one usub8 call to set GE flags, rather than uqsub8 followed by
usub8 w/ 0
document some stalls for further consideration
Change-Id: Ic3877e0ddbe314bb8a17fd5db73501a7d64570ec
test cases were causing a crash because the count was being read
incorrectly. after fixing that, noticed that the output was not
matching. fixed that.
Change-Id: Idb0edb887736bd566a3cf6d4aa1a03ea8d20eb27
asm_offsets contains some definitions which are no longer used. this
was one of them. v6 build works now
Change-Id: If370cfa8acd145de4fead2d9a11b048fccc090df
Jeff Muizelaar posted some changes to the idct/reconstruction c code.
This is the equivalent update for the arm assembly.
This shows a good boost on v6, and a minor boost on neon.
Here are some numbers for highway in qcif, 2641 frames:
HEAD neon: ~161 fps
new neon: ~162 fps
HEAD v6: ~102 fps
new v6: ~106 fps
The following functions have been updated for armv6 and neon:
vp8_dc_only_idct_add
vp8_dequant_idct_add
vp8_dequant_dc_idct_add
Conflicts:
vp8/decoder/arm/armv6/dequantdcidct_v6.asm
vp8/decoder/arm/armv6/dequantidct_v6.asm
Resolved by removing these files. When I rewrote the functions, I also
moved the files to dequant_dc_idct_v6.asm/dequant_idct_v6.asm
Change-Id: Ie3300df824d52474eca1a5134cf22d8b7809a5d4
This moves the prediction step before the idct and combines the idct and
reconstruction steps into a single step. Combining them seems to give an
overall decoder performance improvement of about 1%.
Change-Id: I90d8b167ec70d79c7ba2ee484106a78b3d16e318
Restructured and rewrote SSE2 loopfilter functions. Combined u and
v into one function to take advantage of SSE2 128-bit registers.
Tests on test clips showed a 4% decoder performance improvement on
Linux desktop.
Change-Id: Iccc6669f09e17f2224da715f7547d6f93b0a4987
When the license headers were updated, they accidentally contained
trailing whitespace, so unfortunately we have to touch all the files
again.
Change-Id: I236c05fade06589e417179c0444cb39b09e4200d
Change bitreading functions to use a larger window which is refilled less
often.
This makes it cheap enough to do bounds checking each time the window is
refilled, which avoids the need to copy the input into a large circular
buffer.
This uses less memory and speeds up the total decode time by 1.6% on an ARM11,
2.8% on a Cortex A8, and 2.2% on x86-32, but less than 1% on x86-64.
Inlining vp8dx_bool_decoder_fill() has a big penalty on x86-32, as does moving
the refill loop to the front of vp8dx_decode_bool().
However, having the refill loop between computation of the split values and
the branch in vp8_decode_mb_tokens() is a big win on ARM (presumably due to
memory latency and code size: refilling after normalization duplicates the
code in the DECODE_AND_BRANCH_IF_ZERO and DECODE_AND_LOOP_IF_ZERO cases.
Unfortunately, refilling at the end of vp8dx_bool_decoder_fill() and at the
beginning of each decode step in vp8_decode_mb_tokens() means the latter
requires an extra refill at the end.
Platform-specific versions could avoid the problem, but would require most of
detokenize.c to be duplicated.
Change-Id: I16c782a63376f2a15b78f8086d899b987204c1c7
