This work is sponsored by, and copyright, Google.
Previously all subpartitions except the eob=1 (DC) case ran with
the same runtime:
Cortex A7 A8 A9 A53
vp9_inv_dct_dct_16x16_sub16_add_neon: 3188.1 2435.4 2499.0 1969.0
vp9_inv_dct_dct_32x32_sub32_add_neon: 18531.7 16582.3 14207.6 12000.3
By skipping individual 4x16 or 4x32 pixel slices in the first pass,
we reduce the runtime of these functions like this:
vp9_inv_dct_dct_16x16_sub1_add_neon: 274.6 189.5 211.7 235.8
vp9_inv_dct_dct_16x16_sub2_add_neon: 2064.0 1534.8 1719.4 1248.7
vp9_inv_dct_dct_16x16_sub4_add_neon: 2135.0 1477.2 1736.3 1249.5
vp9_inv_dct_dct_16x16_sub8_add_neon: 2446.7 1828.7 1993.6 1494.7
vp9_inv_dct_dct_16x16_sub12_add_neon: 2832.4 2118.3 2266.5 1735.1
vp9_inv_dct_dct_16x16_sub16_add_neon: 3211.7 2475.3 2523.5 1983.1
vp9_inv_dct_dct_32x32_sub1_add_neon: 756.2 456.7 862.0 553.9
vp9_inv_dct_dct_32x32_sub2_add_neon: 10682.2 8190.4 8539.2 6762.5
vp9_inv_dct_dct_32x32_sub4_add_neon: 10813.5 8014.9 8518.3 6762.8
vp9_inv_dct_dct_32x32_sub8_add_neon: 11859.6 9313.0 9347.4 7514.5
vp9_inv_dct_dct_32x32_sub12_add_neon: 12946.6 10752.4 10192.2 8280.2
vp9_inv_dct_dct_32x32_sub16_add_neon: 14074.6 11946.5 11001.4 9008.6
vp9_inv_dct_dct_32x32_sub20_add_neon: 15269.9 13662.7 11816.1 9762.6
vp9_inv_dct_dct_32x32_sub24_add_neon: 16327.9 14940.1 12626.7 10516.0
vp9_inv_dct_dct_32x32_sub28_add_neon: 17462.7 15776.1 13446.2 11264.7
vp9_inv_dct_dct_32x32_sub32_add_neon: 18575.5 17157.0 14249.3 12015.1
I.e. in general a very minor overhead for the full subpartition case due
to the additional loads and cmps, but a significant speedup for the cases
when we only need to process a small part of the actual input data.
In common VP9 content in a few inspected clips, 70-90% of the non-dc-only
16x16 and 32x32 IDCTs only have nonzero coefficients in the upper left
8x8 or 16x16 subpartitions respectively.
This is cherrypicked from libav commit
9c8bc74c2b40537b0997f646c87c008042d788c2.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
* commit '80fbb7becae530167373fe5178966b7d7604306e':
checkasm: vp8.mc: initialize the full src buffer after ec32574209f
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
* commit '8c816c0c9b12fdefd9046415e97df299880bc9b8':
checkasm/arm: align the clobber check data properly for ldrd
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
* commit 'ec32574209f36467ef0d22c21a7e811ba98c15b6':
checkasm: vp8: mc: test unequal width/height for partitions
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
* commit 'e48746deec48e9ff195841bc3266b4e153a878cd':
checkasm: h264dsp: Move the x and y variables into the randomize_buffer macro
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>
This fixes valgrind warnings about conditional jumps based on
uninitialized data (even though the uninitialized data only ever
was compared with a direct copy of the same uninitialized data).
Signed-off-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The stack used by checkasm_checked_call_vfp was a multiple of 4 when the
checked function is called. AAPCS requires a double word (8 byte)
aligned stack public interfaces. Since both calls are public interfaces
the stack is misaligned when the checked is called.
Might fix the SIGBUS error in the armv7-linux-clang-3.7 fate config.
This avoids listing the same feature multiple times in the
test output. Previously the output contained something like this:
SSE2:
- hevc_mc.qpel [OK]
- hevc_mc.epel [OK]
- hevc_mc.unweighted_pred [OK]
- hevc_mc.qpel [OK]
- hevc_mc.epel [OK]
- hevc_mc.unweighted_pred [OK]
Signed-off-by: Martin Storsjö <martin@martin.st>
This avoids the risk of accidentally clobbering such variables outside
of the macro if the same variables are used there.
Signed-off-by: Martin Storsjö <martin@martin.st>
This fixes valgrind warnings about conditional jumps based on
uninitialized data (even though the uninitialized data only ever
was compared with a direct copy of the same uninitialized data).
Signed-off-by: Martin Storsjö <martin@martin.st>
The functions may not clean up properly after using MMX
registers. For the normal testing calls, the checkasm_checked_call
functions will do the cleanup (and check that functions that
should clean up do it as well), but when benchmarking functions
that don't clean up, we don't currently properly clean up at all.
This causes issues if a benchmarked function is followed by testing
of a function that is supposed to not clobber the MMX/FPU state but
doesn't touch it at all.
Signed-off-by: Martin Storsjö <martin@martin.st>
Restore alphabetical order in lists, break overly long lines, do some
prettyprinting, add some explanatory section comments, group parts
together that belong together logically.
Also bench a smaller buffer. This drastically reduces --bench runtime
and reports smaller, more readable numbers.
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>
Some debuggers/profilers use this metadata to determine which function a
given instruction is in; without it they get can confused by local labels
(if you haven't stripped those). On the other hand, some tools are still
confused even with this metadata. e.g. this fixes `gdb`, but not `perf`.
Currently only implemented for ELF.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Some debuggers/profilers use this metadata to determine which function a
given instruction is in; without it they get can confused by local labels
(if you haven't stripped those). On the other hand, some tools are still
confused even with this metadata. e.g. this fixes `gdb`, but not `perf`.
Currently only implemented for ELF.
Use two separate functions, depending on whether VFP/NEON is available.
This is set to require armv5te - it uses blx, which is only available
since armv5t, but we don't have a separate configure item for that.
(It also uses ldrd, which requires armv5te, but this could be avoided
if necessary.)
Signed-off-by: Martin Storsjö <martin@martin.st>
