Lynne [Sun, 27 Oct 2019 14:44:00 +0000 (14:44 +0000)]
lavfi: add Vulkan filtering framework
This commit adds a Vulkan filtering infrastructure for libavfilter.
It attempts to abstract as much as possible of the Vulkan API from filters.
The way the hwcontext and the framework are designed permits for parallel,
non-CPU-blocking filtering throughout, with the exception of up/downloading
and mapping.
Philip Langdale [Thu, 24 Oct 2019 01:11:37 +0000 (18:11 -0700)]
lavfi/vf_hwupload: Add support for HW -> HW transfers
As we find ourselves wanting a way to transfer frames between
HW devices (or more realistically, between APIs on the same device),
it's desirable to have a way to describe the relationship. While
we could imagine introducing a `hwtransfer` filter, there is
almost no difference from `hwupload`. The main new feature we need
is a way to specify the target device. Having a single device
for the filter chain is obviously insufficient if we're dealing
with two devices.
So let's add a way to specify the upload target device, and if none
is specified, continue with the existing behaviour.
We must also correctly preserve the sw_format on such a transfer.
Lynne [Wed, 28 Aug 2019 20:58:10 +0000 (21:58 +0100)]
lavu: add Vulkan hwcontext code
This commit adds the necessary code to initialize and use a Vulkan device
within the hwcontext libavutil framework.
Currently direct mapping to VAAPI and DRM frames is functional, and
transfers to CUDA and native frames are supported.
Lets hope the future Vulkan video decode extension fits well within this
framework.
Philip Langdale [Thu, 24 Oct 2019 01:01:52 +0000 (18:01 -0700)]
lavu/hwcontext: Add support for HW -> HW transfers
We are beginning to consider scenarios where a given HW Context
may be able to transfer frames to another HW Context without
passing via system memory - this would usually be when two
contexts represent different APIs on the same device (eg: Vulkan
and CUDA).
This is modelled as a transfer, as we have today, but where both
the src and the dst are hardware frames with hw contexts. We need
to be careful to ensure the contexts are compatible - particularly,
we cannot do transfers where one of the frames has been mapped via
a derived frames context - we can only do transfers for frames that
were directly allocated by the specified context.
Additionally, as we have two hardware contexts, the transfer function
could be implemented by either (or indeed both). To handle this
uncertainty, we explicitly look for ENOSYS as an indicator to try
the transfer in the other direction before giving up.
Jai Luthra [Fri, 24 Jan 2020 10:33:33 +0000 (16:03 +0530)]
mlpenc: improve lpc filtering
* fix a possible memory leak (apply_filter returned before freeing)
* use apply_filters in process_major_frame
* revert back to checking bounds with 24 bitdepth, as huff offset takes
care of it
avcodec/apedec: Fix integer overflows in predictor_decode_mono_3950()
Fixes: signed integer overflow: -2147407150 + -1871606 cannot be represented in type 'int' Fixes: 18702/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_APE_fuzzer-5679095417667584 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Marton Balint [Sun, 2 Feb 2020 21:58:08 +0000 (22:58 +0100)]
avformat/dashenc: use AV_OPT_TYPE_DICT for http_opts
This changes the separator character from comma to colon, but since this option
was only added recently I think it should be done for consistency with other
similar options.
This commit reuses the configuration options for VP8 that enables
temporal scalability for VP9. It also adds a way to enable three
preset temporal structures (refer to the documentation for more
detail) that can be used in offline encoding.
Fixes: signed integer overflow: 2147482897 + 2048 cannot be represented in type 'int' Fixes: 19240/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RALF_fuzzer-5743240326414336 Fixes: 19869/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RALF_fuzzer-5150136636538880 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/wmavoice: Fix rounding and integer anomalies in calc_input_response()
Fixes: out of array access Fixes: inf is outside the range of representable values of type 'int' Fixes: signed integer overflow: -9223372036854775808 - 1 cannot be represented in type 'long' Fixes: 19316/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_WMAVOICE_fuzzer-5677369365102592 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This limit is roughly based on the bitreader limit, its likely a much tighter limit
could be used
Fixes: left shift of 1965039647 by 1 places cannot be represented in type 'int' Fixes: 19545/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_WMAVOICE_fuzzer-5695391899320320 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/pcm: Fix invalid shift in pcm_decode_frame for LXF
Fixes: left shift of 32 by 28 places cannot be represented in type 'int' Fixes: 19472/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_PCM_LXF_fuzzer-5704364320096256 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: left shift of 79 by 28 places cannot be represented in type 'int' Fixes: 20202/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HAP_fuzzer-5719004081815552 Fixes: 20219/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HAP_fuzzer-5641738677125120 Fixes: 20389/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_HAP_fuzzer-5680721517871104 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Marton Balint [Sun, 26 Jan 2020 21:13:50 +0000 (22:13 +0100)]
avformat/udp: cancel pending IO on win32 manually
recvfrom() is not a cancellation point in pthreads-win32, see
https://sourceware.org/pthreads-win32/manual/pthread_cancel.html
In order to be able to cancel the reader thread on Win32 properly we first
shutdown the socket then call CancelIoEx to abort pending IO. Subsequent
recvfrom() calls will fail with WSAESHUTDOWN causing the thread to exit.
avcodec/mlpdsp: Fix a invalid shift in ff_mlp_rematrix_channel()
Fixes: left shift of negative value -2 Fixes: 20305/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_TRUEHD_fuzzer-5677196618498048 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Jai Luthra <me@jailuthra.in> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Before this commit s->avctx == NULL was used to infer that an encoder is
being initialzed. Code readability has been improved by directly using
!av_codec_is_decoder() instead.
dashenc: check pts to prevent division by zero error
this usecase will cause a division by zero trap:
1. dashenc has received one frame
2. os->max_pts and os->start_pts have same value
3. delta between max_pts and start_pts is 0
4. av_rescale_q(0, x, y) returns 0
5. this value is used as denominator in division
6. Bang! -> segfault
this fix checks that max_pts > start_pts.
the fix has been tested and works.
Signed-off-by: Alfred E. Heggestad <alfred.heggestad@gmail.com> Reviewed-by: Jeyapal, Karthick <kjeyapal@akamai.com> Signed-off-by: James Almer <jamrial@gmail.com>
Fixes: signed integer overflow: 2003010644 * 2 cannot be represented in type 'int' Fixes: 19593/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RALF_fuzzer-5660628006207488 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: bypassing of checks and assertion failure Fixes: asan_1003879.mp4 Found-by: Clusterfuzz + asan Reported-by: Thomas Guilbert <tguilbert@google.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: left shift of 3329 by 20 places cannot be represented in type 'int' Fixes: signed integer overflow: -199378355 + -1948950833 cannot be represented in type 'int' Fixes: 19837/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_WMALOSSLESS_fuzzer-5752565837070336 Fixes: 19839/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_WMALOSSLESS_fuzzer-5767483265122304 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/apedec: Fix undefined integer overflow in decode_array_0000()
Fixes: signed integer overflow: -2143289344 - 6246400 cannot be represented in type 'int' Fixes: 19239/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_APE_fuzzer-5173755680915456 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: out of array access Fixes: 19750/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RAWVIDEO_fuzzer-5074834119983104 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/iff: Over-allocate ham_palbuf for HAM6 IFF-PBM
IFF-PBM-HAM6 can read out of array without this overallocation Fixes: Out of array read Fixes: 19752/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IFF_ILBM_fuzzer-5675331403120640 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/x86/diracdsp: Fix incorrect src addressing in dequant_subband_32()
Fixes: Segfault (not reproducable with asm, which made this hard to debug) Fixes: decoding errors Fixes: 19854/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_DIRAC_fuzzer-5729372837511168 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avformat/aviobuf: Remove AVIOInternal and one level of indirection
In the Libav commit cae448cf, the opaque of every AVIOContext opened
by ffio_fdopen() (which is used internally by avio_open() and avio_open2())
changed: It was a simple pointer to an URLContext before, but now it was
a structure (namely AVIOInternal) containing a pointer to an URLContext
as its only member. The next commits (namely 8c0ceafb and ec4c4839) added
members to AVIOInternal to allow white-/blacklisting of protocols.
But these two commits were never merged into FFmpeg (they were only
merged as no-ops in 510046c2 and 063b26d3), because FFmpeg chose
a different way to implement this (in 93629735); and so our AVIOInternal
still has exactly one member.
This of course means that it is unnecessary to use AVIOInternal as
opaque as it is just adding a level of indirection (not only pointer
dereference, but also wrapper functions). Therefore this commit
removes AVIOInternal entirely and essentially reverts cae448cf.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Reviewed-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Steven Liu [Thu, 30 Jan 2020 09:03:48 +0000 (17:03 +0800)]
avformat/avio: add avio_protocol_get_class
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc> Suggested-by: Hendrik Leppkes <h.leppkes@gmail.com> Suggested-by: Nicolas George <george@nsup.org> Signed-off-by: Steven Liu <lq@chinaffmpeg.org>
tools/target_dec_fuzzer: Add threshold for IFF_ILBM
Fixes: Timeout (32 -> 1sec) Fixes: 20138/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IFF_ILBM_fuzzer-5634665251864576 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Peter Ross <pross@xvid.org> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Besides the obvious advantages this also fixes a potential memleak:
If only one of the arrays had been successfully allocated, the other
would leak. This also fixes Coverity issues #1440386 and #1440387.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Reviewed-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Write a few numbers directly via AV_WB32 instead of using an AVIOContext
(that is initialized only for this very purpose) to write these numbers
at known offsets into a fixed buffer.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> Signed-off-by: James Almer <jamrial@gmail.com>
projects / ffmpeg / log