Anton Khirnov
31b5b3badc
lavu/opt: deprecate av_opt_ptr()
...
It has no more internal callers, and we do not want to support direct
pointer access via AVOptions, as that constrains AVOption API
extensions.
2024-10-16 16:46:29 +02:00
Lynne
931d45d4d6
vulkan: do not create imageviews with video encode/decode usage
...
This function is only used for filtering and generic compute.
The issue is that a view inherits the usage flags from the image
by default, and the spec says the view format must be compatible
with the usage. VkImageViewUsageCreateInfo allows us to filter out
the indeded uses of the imageview.
Pffff.
2024-10-16 12:48:16 +02:00
Lynne
4b128de44a
vulkan: enable selecting a compatible representation of format
...
When using **integer** images inside shaders, it turns out
that conversion doesn't automatically happen, but we need to
explicitly use the imageviews to get the image exposed as
a suitable representation for the shader.
Finally enables bitexact image representations.
2024-10-16 12:48:15 +02:00
Lynne
76e8afa8a6
hwcontext_vulkan: always enable MUTABLE creation flag
...
We need it even for something as simple as bitexact opening
of images.
2024-10-16 12:48:13 +02:00
Lynne
98f6d43146
vulkan: keep track of mapped memory in the buffer structure
...
Can/could be useful to know which buffers are mapped.
2024-10-15 17:45:20 +02:00
Lynne
3afdfe830f
vulkan: add support for AV_PIX_FMT_RGB96
2024-10-15 17:45:19 +02:00
Lynne
903f7f053a
vulkan: add support for AV_PIX_FMT_RGBA128
2024-10-15 17:45:18 +02:00
Lynne
2336e68565
lavu: bump minor and add APIChanges entries for new 32bpc pixfmts
2024-10-15 17:45:18 +02:00
Lynne
4f387a1c4e
lavu/pixfmt: add AV_PIX_FMT_RGB96
2024-10-15 17:45:17 +02:00
Lynne
281bba1d26
lavu/pixfmt: add AV_PIX_FMT_RGBA128
...
This format is useful for doing certain lossless transforms on images,
RCT in particular, which require you to escalate the size from 16 to
32 bits to avoid overflows.
APIchanges will be done alongside when comitting.
2024-10-15 17:45:12 +02:00
Martin Schitter
c993a91bea
avutil: add RGBF16 pix_fmt
...
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
2024-10-14 20:46:24 +02:00
James Almer
fd8b0dcfed
avutil/hwcontext_vulkan: add proper maps for XV3{0,6}
...
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-14 15:12:49 -03:00
James Almer
d41e5b0fc7
avutil/hwcontext_videotoolbox: add support for AYUV pixel format
...
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-12 11:21:14 -03:00
Lynne
0e19b05375
vulkan: fix layout qualifiers for 32-bit RGBA float storage reps
...
r8
r16/r16f
r32f
Sure, GLSL. Makes sense.
2024-10-12 00:38:49 +02:00
Lynne
8c47cf9eea
hwcontext_vulkan: enable shaderBufferInt64Atomics if supported
2024-10-11 23:54:23 +02:00
Lynne
168e8ae9bf
hwcontext_vulkan: add support for AV_PIX_FMT_RGBF32
2024-10-11 23:54:23 +02:00
Lynne
d55a351741
hwcontext_vulkan: add support for AV_PIX_FMT_RGBAF32
2024-10-11 23:54:22 +02:00
Lynne
1addd0fdca
vulkan: use correct signed image type for storage images
...
Using signed or unsigned integer formats/layouts requires that
"uimage" or "iimage" are used.
2024-10-11 23:54:22 +02:00
Lynne
c982376076
vulkan: extend ff_vk_shader_rep_fmt to be useful for bitexactness
...
The original either reported 8 or 16-bit conversion from the
original, rather than being able to return the actual original.
This makes it usable in a situation where preserving exactness
is required.
2024-10-11 23:54:18 +02:00
James Almer
1cead90292
avutil/pixdesc: use a bigger variable type when writing bitstream formats
...
Fixes fate-imgutils and fate-pixelutils under gcc-usan after
29ea34728f
2024-10-10 16:03:22 -03:00
Lynne
80a0567a2f
vulkan: fix ImageView flexible array struct allocation
...
Same as the previous commit, the compiler may insert padding.
Thanks to Marvin Scholz for pointing this out.
2024-10-09 02:39:34 +02:00
Marvin Scholz
f5e2914a89
libavutil/vulkan: fix flexible array struct allocation
...
The flexible array member struct can have padding added by
the compiler which was not taken into account properly, which
could lead to a heap buffer overflow.
2024-10-09 02:36:12 +02:00
Lynne
4ad4f8101a
hwcontext_vulkan: set hwctx->device_features
...
This was forgotten during the recent device feature refactor.
2024-10-09 01:01:32 +02:00
Lynne
31def0c3ce
hwcontext_vulkan: remove redundant hostQueryReset setting
...
Its set just a few lines earlier.
2024-10-09 01:01:32 +02:00
Lynne
b8d1b4358b
vulkan: add a #define when printf debugging is enabled
2024-10-09 01:01:31 +02:00
James Almer
29ea34728f
avutil/pixfmt: add V30X pixel format
...
This maps to the 444YpCbCr10 pixel format as defined by Apple.
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-08 19:45:10 -03:00
James Almer
1f9424e0b1
avutil/pixfmt: add VYU444 pixel format
...
This maps to the 444YpCbCr8 pixel format as defined by Apple, which is ordered
Cr Y' Cb.
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-08 14:10:07 -03:00
James Almer
dc3ef53819
avutil/pixfmt: add UYVA pixel format
...
This maps to the 4444YpCbCrA8 pixel format as defined by Apple, which is ordered
Cb Y' Cr A.
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-08 14:10:07 -03:00
James Almer
263108a504
avutil/pixfmt: add AYUV pixel format
...
This maps to the 4444AYpCbCr8 pixel format as defined by Apple, which is ordered
A Y’ Cb Cr.
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-08 14:10:07 -03:00
James Almer
2e79690a70
avutil/pixfmt: extend the VUY{A,X} description
...
Signed-off-by: James Almer <jamrial@gmail.com>
2024-10-08 14:10:07 -03:00
Lynne
6871fd3170
vulkan: always enable GL_EXT_control_flow_attributes
2024-10-06 07:13:04 +02:00
Lynne
8c250da8f3
vulkan: always enable GL_EXT_shader_explicit_arithmetic_types
...
Implicit types were a mistake.
lowp/mediump/highp was a sin.
2024-10-06 07:12:59 +02:00
Lynne
9391d4ccc5
lavu/vulkan_spirv: use correct inclusion guards
...
Forgot to change them after moving the file over from libavfilter.
2024-10-06 06:04:36 +02:00
IndecisiveTurtle
d80676117a
libavutil/vulkan: Prevent crash on shaders with no descriptors
...
Needed to prevent crashes on vc2 vulkan encoder patch
2024-10-06 01:32:50 +02:00
Lynne
892f64ad9b
x86/tx_float: remove HAVE_AVX2_EXTERNAL checks
...
It'll always be enabled.
Thanks, nasm.
2024-10-06 01:32:49 +02:00
Lynne
b17a240c8d
Revert "x86/tx_float: set all operands for shufps"
...
This reverts commit 74f5fb6db8
2024-10-06 01:32:49 +02:00
Lynne
24c5a58e55
Revert "x86/tx_float: add missing check for AVX2"
...
This reverts commit f4097e4c1f
2024-10-06 01:32:48 +02:00
Lynne
bf643f989b
Revert "x86/tx_float: add missing preprocessor wrapper for AVX2 functions"
...
This reverts commit 750f378bec
2024-10-06 01:32:48 +02:00
Lynne
b890482d05
Revert "x86/tx_float: change a condition in a preprocessor check"
...
This reverts commit 0d8f43c74d
2024-10-06 01:32:47 +02:00
Lynne
0a1f8107bb
vulkan: move SPIR-V compilation code to libavutil
...
The code is not currently used by libavutil, its just where our
common Vulkan code is.
Since SPIR-V compilation will be needed by lavc, move it, rather
than having lavc including lavfi.
2024-10-06 01:21:44 +02:00
Lynne
ac092c6707
hwcontext_vulkan: guard all uses of new spec defines and fix stray bracket
...
This fixes compilation with less recent Vulkan headers.
2024-10-04 10:41:03 +02:00
Lynne
a304cbeb8d
vulkan: add profiling debug setting
...
This simply keeps all shader optimizations, but allows debug
data to be generated.
2024-10-04 10:10:46 +02:00
Lynne
832947be02
vulkan: don't enable GL_EXT_buffer_reference by default
...
Only nlmeans_vulkan uses it.
2024-10-04 10:10:45 +02:00
Lynne
356d1cc8ff
vulkan: parse instance list and add the DEBUG_UTILS extension
...
Required to let users know whether debugging is active.
2024-10-04 10:10:44 +02:00
Lynne
e3676d96cb
hwcontext_vulkan: move device feature struct setup to a new function
2024-10-04 10:10:43 +02:00
Lynne
535e5eb7f3
hwcontext_vulkan: enable VK_KHR_shader_relaxed_extended_instruction
2024-10-04 10:10:43 +02:00
Lynne
0d5bfd0b21
vulkan_functions: change extension type to a typedef uint64_t
...
We were getting a bit too close for comfort to the 32-bit limit
on enums.
2024-10-04 10:10:42 +02:00
Lynne
d80f9f55c8
vulkan: always enable GL_EXT_scalar_block_layout
...
This makes std430 (which we use everywhere already) fully match C
layout.
Extension was made mandatory in 1.2.
2024-10-04 10:10:42 +02:00
Lynne
37d5cb84e8
vulkan: check if current buffer has finished execution before picking another
...
This saves resources, as dependencies are freed/reclaimed with a lower latency,
and provies a speedup.
2024-10-04 10:10:42 +02:00
Lynne
877c5a9692
vulkan: use shader objects if supported
...
Shader objects finally allow completely independent shaders.
2024-10-04 10:10:41 +02:00
