Additional work is needed for storage images with DCC without DCC image stores to not be broken.
Fixes black screens in Doom Eternal.
Fixes: #5345
Signed-off-by: Joshua Ashton <joshua@froggi.es>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12818>
Doesn't seem to be causing any issues right now but could with modifiers potentially.
Matches what is in RadeonSI where the comment is also shamelessly stolen from.
Signed-off-by: Joshua Ashton <joshua@froggi.es>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12811>
VK_FORMAT_R32_SFLOAT seems pretty common and it seems we can be a
little smarter when shader image 32-bit float atomics aren't enabled.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12406>
This seems to be unsupported.
COMPRESSION_EN=1 and WRITE_COMPRESS_ENABLE=1 don't update HTILE
with image stores.
Note that there is no issue because depth/stencil images will be
decompressed for image stores, and TC-compat HTILE is disabled.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12450>
The driver doesn't enable TC-compat HTILE for storage images, so this
was actually always TRUE.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12450>
Images that support comp-to-single don't have to be fast-cleared at
all, so the predicate is unnecessary.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12323>
It doesn't make sense to ask for the depth-only or stencil-only format
if there is no depth or stencil. One bit of radv_image.c did seem to
take advantage of the default case in vk_format_depth_only so throw an
`if (vk_format_has_depth(format))` around it.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12023>
It seems we can enable DCC if the possible formats differ in signedness
and are otherwise compatible. We just need a fast-clear eliminate for
certain clear colors.
Improves Trine 4 performance.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9387>
Only GFX10+ is affected because older chips don't support
comp-to-single. For them, we need to implement FCE on compute with DCC
and eventually CMASK.
Fixes the gap between concurrent vs exclusive queue with Scarlet Nexus,
also gives a boost with Doom Eternal.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12088>
Lots of the MAX2 args end up subtracting two unsigned numbers, which
blows up when the result is negative.
Fixes: 4c99d6ff54 ("radv: flush L2 for images affected by the pipe misaligned issue on GFX10+")
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12272>
Otherwise, it might not be correctly initialized for Android.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12103>
This doesn't fix anything known. Found by inspection.
Cc: 21.1 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11302>
With DCC and mipmaps, some mips can't be compressed and it makes
sense to check this here.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10667>
It seems this gives 2-3% improvements most of the time. This also
enables DCC for concurrent images.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10454>
To make them readable by shaders, only needed on GFX10+.
This also fixes corruption with Control and MSAA.
Cc: 21.1 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10610>
Otherwise, we need a FMASK_DECOMPRESS which is only supported
on the gfx queue.
Fixes rendering with Forza Horizon 4 on Polaris10.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10482>
It's already implemented and it works quite fine on GFX10+, so it
should be safe to enable it on GFX8-9.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10488>
DCC decompress/FMASK expand are supported on compute queues. Since
the driver doesn't perform fast clears with concurrent images, we
don't need to perform a FCE on compute (we can't anyways).
This fixes a performance regression with Control and
VKD3D_CONFIG=multi_queue.
One more optimization (as discussed with Bas) is to implement FCE
on compute to allow fast clears.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10323>
ACO doesn't create a waitcnt for barriers between texture samples and
image stores because texture samples are supposed to use read-only
memory. It could also schedule the barrier to above the texture sample.
We also have use a larger memory scope to avoid an ACO optimization.
Tested on GFX8 with Sachsa Willems deferred sample. With some DCC
decompressions and the compute path forced.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Cc: 21.1 <mesa-stable>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9496>
The retile map is removed and replaced by direct DCC address computations
in the retile shader using the new function ac_nir_dcc_addr_from_coord.
The RADV code is disabled.
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10003>
Only one bit is set in alignments, so store the bit offset (log2) and
change the type from uint32_t to uint8_t.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
Images are always aligned to 256B (enforced by register and descriptor
fields) and limited to 40-bit addresses. This saves some space.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10083>
