This does not improve performance with current hardware due to the poor
performance of segmented accesses. Performance should be slightly better
with expensive or near-future hardware that I don't have, however it is
still limited by two other factors:
- There are only 4 elements.
- The final stores are necessarily indexed and hit multiple cache lines,
thus as slow as scalar.
SpacemiT X60:
dct_unquantize_h263_inter_c: 417.8 ( 1.00x)
dct_unquantize_h263_inter_rvv_i32: 66.0 ( 6.33x)
dct_unquantize_h263_intra_c: 140.2 ( 1.00x)
dct_unquantize_h263_intra_rvv_i32: 67.7 ( 2.07x)
Note that the C benchmarks are not stable, depending heavily on the
number of coefficients picked by the RNG. The R-V V benchmarks are
however very stable and generally better than C's.
The VK spec forbids using clear commands on YUV images,
so we need to allocate separate per-plane images.
This removes the need for a separate reset shader.
It's a name that communicates its functionality in a better way.
Since the function was introduced very recently, we can safely rename it.
Signed-off-by: James Almer <jamrial@gmail.com>
Normally, this function tries to make sure all threads are saturated with
work to do before returning any frames; and will continue requesting packets
until that is the case.
However, this significantly slows down initial decoding latency when only
requesting a single frame (to e.g. configure the filter graph), and also
wastes a lot of unnecessary memory in the event that the user does not intend
to decode more frames until later.
By introducing a new `flags` paramater and a new flag
`AV_CODEC_RECEIVE_FRAME_FLAG_SYNCHRONOUS` to go along with it, we can allow
users to temporarily bypass this logic.
h->context_initialized is zero after flush, which triggers call to
ff_get_format unconditionally. ff_get_format can be heavy with
ff_hwaccel_uninit and hwaccel_init. For example, it takes 20 ms on
macOS with videotoolbox. ff_get_format should not be called if
nothing changed. ff_get_format is guarantee to be called at the
first time and when video information changed with
(must_reinit || needs_reinit).
Fix#20760.
The width 16 epel functions never use four taps in any direction*,
so don't build said functions. Saves 4352B of .text and 89B of
.text.unlikely here.
*: mx and my in vp8_mc_luma() are always even.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
For the epel functions, there can be no overflow as long as the sum
contains only one of the two large central coefficients; for bilinear
functions, there can be no overflow whatsoever.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
By changing the permutations used in the epel8_h{4,6} case
we can simply reuse the coefficient tables from the vertical epel
filters.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Doubling the register width allows to use only one pshufb and pmaddubsw.
Old benchmarks:
vp8_put_epel4_h4_c: 82.8 ( 1.00x)
vp8_put_epel4_h4_ssse3: 13.9 ( 5.96x)
New benchmarks:
vp8_put_epel4_h4_c: 82.7 ( 1.00x)
vp8_put_epel4_h4_ssse3: 11.7 ( 7.08x)
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Switching to xmm registers allows to process two rows in parallel,
leading to speedups. It is also ABI compliant (no more missing emms).
Old benchmarks:
vp8_put_epel4_v4_c: 96.8 ( 1.00x)
vp8_put_epel4_v4_ssse3: 28.2 ( 3.43x)
New benchmarks:
vp8_put_epel4_v4_c: 95.1 ( 1.00x)
vp8_put_epel4_v4_ssse3: 22.8 ( 4.17x)
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Switching to xmm registers allows to process two rows in parallel,
leading to speedups. It is also ABI compliant (no more missing emms).
Old benchmarks:
vp8_put_epel4_v6_c: 132.8 ( 1.00x)
vp8_put_epel4_v6_ssse3: 34.3 ( 3.87x)
New benchmarks:
vp8_put_epel4_v6_c: 131.5 ( 1.00x)
vp8_put_epel4_v6_ssse3: 27.1 ( 4.86x)
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
There is a register available. No change in benchmarks here.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Use GPRs on x64 and xmm registers else (using GPRs reduces codesize).
This avoids clobbering the floating point state and therefore no longer
breaks the ABI.
No change in benchmarks here.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
SSSE3 is already quite old (introduced 2006 for Intel, 2011 for AMD),
so that the overwhelming majority of our users (particularly those
that actually update their FFmpeg) will be using the SSSE3 versions.
This commit therefore removes the MMX(EXT) functions overridden
by them (which don't abide by the ABI) to get closer to a removal
of emms_c.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
A heap-use-after-free vulnerability was identified in
`libavcodec/aac/aacdec.c`. When `che_configure` frees a
`ChannelElement` (`ac->che[type][id]`), it failed to clear all
references to it in `ac->tag_che_map`. `ac->tag_che_map` caches
pointers to `ChannelElement`s and can contain cross-type mappings (e.g.,
a `TYPE_SCE` tag mapping to a `TYPE_LFE` element).
In a USAC stream reconfiguration scenario, an LFE element was freed, but
a stale pointer remained in `ac->tag_che_map`. Subsequent calls to
`ff_aac_get_che` returned this dangling pointer, leading to a crash in
`decode_usac_core_coder`.
This commit fixes the issue by iterating over the entire
`ac->tag_che_map` in `che_configure` and clearing any entries that point
to the `ChannelElement` about to be freed, ensuring no dangling pointers
remain.
Fixes: https://issues.oss-fuzz.com/issues/440220467
Intra refresh is a technique that gradually refreshes the video by encoding rows or regions as intra macroblocks/CTUs spread over multiple frames, rather than using periodic I-frames.
This provides better error resilience for video streaming while maintaining more consistent bitrate.
Disable Intra Refresh (This is the default)
ffmpeg -init_hw_device d3d12va -hwaccel d3d12va -hwaccel_output_format d3d12 \
-i input.mp4 \
-c:v h264_d3d12va \
-intra_refresh_mode none \
-intra_refresh_duration 30 \
-g 60 \
output.h264
Enable Intra Refresh
ffmpeg -init_hw_device d3d12va -hwaccel d3d12va -hwaccel_output_format d3d12 \
-i input.mp4 \
-c:v h264_d3d12va \
-intra_refresh_mode row_based \
-intra_refresh_duration 30 \
-g 60 \
output.h264
Parameters
- `-intra_refresh_mode`: Set to `row_based` to enable row-based intra refresh, or `NONE` to disable
- `-intra_refresh_duration`: Number of frames over which to spread the intra refresh (default: 0 = use GOP size)
- `-g`: GOP size (should typically be larger than intra refresh duration)
Fixes a heap-buffer-overflow in `decode_frame` where `header_len` read
from the bitstream was not validated against the remaining bytes in the
input buffer (`gb`). This allowed `gb_hdr` to be initialized with a size
exceeding the actual packet data, leading to an out-of-bounds read.
The fix adds a check to ensure `bytestream2_get_bytes_left(&gb)` is
greater than or equal to `header_len - 2` before initializing `gb_hdr`.
Fixes: https://issues.oss-fuzz.com/issues/439711053
This was accidentally removed in
357fc5243c.
This fixes test failures when built with Clang and MSVC;
surprisingly, the checkasm test did seem to pass when built with
GCC. Clang and MSVC also warn about the use of the uninitialized
variable, while GCC didn't.
Possible since 222127418b.
Reviewed-by: Kacper Michajłow <kasper93@gmail.com>
Reviewed-by: Lynne <dev@lynne.ee>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Currently several inline ASM blocks used a value as
an input and rax as clobber register. The input value
was just moved into the register which then served as loop
counter. This is wasteful, as one can just use the value's
register directly as loop counter.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Unquantizing involves calculating
(block[j] * qscale * quant_matrix[j]) / 16
where / rounds towards zero. Arithmetic right shifts
naturally round towards -inf, so the earlier code
calculated the absolute value first, then used a right-shift
and then negated the result if necessary.
This commit uses a different procedure: It biases the product
for negative values of block[j] by 0xf. The combination of
this and the arithmetic right shift is the same as rounding
towards zero.
Furthermore, a write-only store to mm7 has been removed.
Benchmarks:
dct_unquantize_mpeg2_intra_c: 214.3 ( 1.00x)
dct_unquantize_mpeg2_intra_mmx (old): 43.0 ( 4.98x)
dct_unquantize_mpeg2_intra_mmx (new): 28.4 ( 7.56x)
(The bitexact flag and the test for correctness have beem removed
from checkasm for the benchmarks.)
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The H.263 unquantize functions modified an input parameter.
(And they did so since this code was added in
7f3f5ec87b. I am surprised
that this didn't cause issues, particularly with the intra function.)
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
This is necessary so that the mpegvideo_unquantize checkasm test
does not pull mpegvideo.o and then all of libavcodec into checkasm.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
These functions currently operate on the assumption that the number
of coefficients to process is always of the form 16k+m with m<=4 or >8.
Yet this is not true when the IDCT permutation is of type FF_IDCT_PERM_LIBMPEG2
(i.e. when FF_IDCT_INT is in use).
Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Forgotten in 70a7df049c.
Using the wrong scantable matters for codecs for which both scantables
can differ, namely the MPEG-4 decoder and the WMV1/2 codecs.
For WMV1 it can lead to wrong output in case the IDCT permutation
is FF_IDCT_PERM_PARTTRANS, because in this case the entries of
of the intra scantable's raster end are not always <= the corresponding
entries of the inter scantable's raster end when the former is
initialized via ff_wmv1_scantable[1] and the latter via ff_wmv1_scantable[0].
FF_IDCT_PERM_PARTTRANS is used iff the Neon IDCT is used (for both arm
and aarch64).* Said IDCT is not used during FATE, so that this issue
went unnoticed.
WMV2 uses the same scantables, but uses a custom IDCT
which always uses FF_IDCT_PERM_NONE for which the inter_scantable,
so that the output is always correct for it.
The scantable for MPEG-4 can change mid-stream (for the decoder),
but since c41818dc5d only the intra
scantable is updated, so that both scantables can get out of sync.
In such a case the unquantize intra functions could unquantize
an incorrect number of coefficients.
Using raster_end of the wrong scantable can also lead to an
unnecessarily large amount of coefficients unquantized.
*: FF_IDCT_PERM_SIMPLE and FF_IDCT_PERM_TRANSPOSE would also not work,
but they are not used at all by arm and aarch64.
Reviewed-by: Martin Storsjö <martin@martin.st>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
To do so, simply add these init files to X86ASM-OBJS instead of OBJS
in the Makefile. The former is already used for the actual assembly
files, but using them for the C init files just works, because the build
system uses file extensions to derive whether it is a C or a NASM file.
This avoids compiling unused function stubs and also reduces our
reliance on DCE: We don't add %if checks to the asm files except
for AVX, AVX2, FMA3, FMA4, XOP and AVX512, so all the MMX-SSE4
functions will be available. It also allows to remove HAVE_X86ASM checks
in these init files.
Reviewed-by: Kacper Michajłow <kasper93@gmail.com>
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
