This is what the kernel reports and what PAL seems to be doing.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25578>
This should be large enough for all engines.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25578>
We overalign it anyway, so there is no change in behavior.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25578>
No change in behavior. The previous overalignment is preserved.
It sets ib_pad_dw_mask sooner.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25578>
This doesn't fix anything because gds_needed should already be TRUE
because it's initialized at pipeline bind time, but this will be needed
for skipping GDS allocation on GFX11.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25284>
According to RadeonSI, this is required for preemption, user queues,
and we only have to wait for VS after streamout which should be more
performant.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25284>
Primitive restart is also applied to non-indexed draws on AMD GPUs. On
GFX11, DISABLE_FOR_AUTO_INDEX can be set but we will need a different
solution for older GPUs.
This fixes all line related flakes in CI (at least).
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25639>
this patch calls the init and finish functions of the vk
runtime astc decoder. initializes emulate_astc flag. sets
up the additional plane to store decoded texture.
v2: fix _tex_dataformat() and _tex_numformat() (Chia-I Wu)
use correct function for bufferToImage (Chia-I Wu)
v3: add radv_is_layout_emulated() (Chia-I Wu)
avoid repeated pattern (Chia-I Wu)
v4: not create all pipelines on_demand (Chia-I Wu)
v5: current code does not support astc hdr (Chia-I Wu)
v6: keep luts in staging buffer only (Chia-I Wu)
v7: use 2DArray for both input and output
v8: document todo to use fp16 (Chia-I Wu)
not required to move meta init anymore (Chia-I Wu)
move astc_emulation_format to vk_texcompress_astc.h (Chia-I Wu)
v9: remove LAYOUT check (Chia-I Wu)
check on iview->vk.view_format
move setting tiled flags for astc (Chia-I Wu)
remove is format emulated check in radv_is_storage_image* (Chia-I Wu)
use LAYOUT_ASTC for if check (Chia-I Wu)
no 1D support (Chia-I Wu)
calculate start end offset in 2x blk size
v10: remove old wrong code (Chia-I Wu)
v11: use existing defined local format variable (Chia-I Wu)
dst image layout is always VK_IMAGE_LAYOUT_GENERAL (Chia-I Wu)
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Acked-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24672>
It works with just one counter.
This mitigates https://gitlab.freedesktop.org/drm/amd/-/issues/2902
quite a lot when you run dEQP-VK.transform_feedback.* in parallel on
more than 16 threads with RDNA3.
For example, on my GPU the kernel reports 16 GDS OA counters which means
that if you run VKCTS with 16 threads (ie. 16 Vulkan devices are
created) it's fine. Otherwise, the kernel might report ENOMEM.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25619>
Otherwise, we have dangling BO pointers in the global BO list. Not
quite sure why this hasn't been triggered before.
Cc: mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25623>
First, we need to give the parent_instr field a unique name to be able to
replace with a helper. We have parent_instr fields for both nir_src and
nir_def, so let's rename nir_src::parent_instr in preparation for rework.
This was done with a combination of sed and manual fix-ups.
Then we use semantic patches plus manual fixups:
@@
expression s;
@@
-s->renamed_parent_instr
+nir_src_parent_instr(s)
@@
expression s;
@@
-s.renamed_parent_instr
+nir_src_parent_instr(&s)
@@
expression s;
@@
-s->parent_if
+nir_src_parent_if(s)
@@
expression s;
@@
-s.renamed_parent_if
+nir_src_parent_if(&s)
@@
expression s;
@@
-s->is_if
+nir_src_is_if(s)
@@
expression s;
@@
-s.is_if
+nir_src_is_if(&s)
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Acked-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24671>
This re-introduces "radv: fix alignment of DGC command buffers" and
"radv/amdgpu: fix alignment of command buffers" which were valid
changes.
IBs need to be aligned to the IB size requirement, not the number of
padded NOPs.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/25588>
Next part don't know whether p_end_with_regs args are loaded from
memory ops or not, need to wait it's done here.
Other memory load needs to be waited too like:
a = load_mem()
b = ...
if (...) {
wait_mem(a)
store_mem(a)
}
p_end_with_regs(b)
"a" still needs to be waited, otherwise next shader part regs may
be overwritten by unfinished memory loads.
Memory stores are waited too. When >=gfx10 and last VGT has no
parameter export, we need to wait all memeory stores done before
pos export (see ac_nir_export_position). So when merged shader
(ES+GS or VS+GS) is partially built, first stage needs to wait
all memory stores done, otherwise second stage don't know if
any memory stores pending before.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Signe-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
radeonsi may generate empty main shader or an empty exit block
for p_end_with_regs to jump to.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
PS with epilog does not need to fix_exports. And radeonsi use
p_end_with_regs so does not have jump instruction at last.
radeonsi may also have exec restore instruction, so may break
before reach to p_end_with_regs.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
radeonsi need to compact color export for ps epilog while radv does not.
radv will fill empty color slot, so won't affected by this change.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
ps needs to handle wqm:
1. main part may compute with args from prolog in wqm mode, so
prolog need to compute these args in wqm mode too.
2. prolog and main part need to end with exact exec, so next
shader part which inherit previous shader part's exec won't
do valid job for helper threads
1 need p_end_with_regs to operate in wqm mode and itself can't
be exact, otherwise some move instruction added by it won't be
in wqm mode so helper threads' compute result is not passed to
next shader part as args.
2 is done by p_end_wqm added by finish_program automatically
after p_end_with_regs.
Piglit tests can trigger the problem:
1. gl-2.1-polygon-stipple-fs
a. ps prolog call discard_if
b. ps main pass wqm exec to epilog
c. ps epilog export color for discarded pixel
2. fs-fwidth-color.shader_test
a. ps prolog need to pass args computed in wqm mode
b. set p_end_with_regs to exact will end wqm mode before
the move instructions, so helper threads's result is not
passed to next shader part
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
p_end_with_regs just partially end the program, next part need
exec mask to be set correctly. For example p_end_wqm will generate
a exec restore from WQM mode after p_end_with_regs.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
It's a HW init reg, not driver spec user sgpr. radv just
doesn't use it. Move it to amd common for aco ps prolog
usage.
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Signed-off-by: Qiang Yu <yuq825@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24973>
