Also moved everything in a struct and then return the struct from
the helper function, so it is clear in the caller what part of the
pipeline gets modified.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
We don't need the pipeline state struct anymore.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This gives about 2% performance improvement on dota2 for me.
This is mostly a mechanical copy and replacement, but at bind time
we still do:
1) Some stuff that is only based on num_samples changes.
2) Some command buffer state setting.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This function can get access for a 64-bit dvec4, which means we
have to load 8 components.
This fixes:
R600_DEBUG=nir ./bin/shader_runner generated_tests/spec/arb_gpu_shader_fp64/execution/built-in-functions/fs-abs-dvec4.shader_test -auto
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Allows nir drivers to either use a single or dual locations for
vs double inputs.
i965 uses dual locations for both OpenGL and Vulkan drivers, for
now gallium OpenGL drivers only use a single location.
The following patch will also make use of this option when
calling nir_shader_gather_info().
Reviewed-by: Karol Herbst <kherbst@redhat.com>
This ports a fix from amdvlk, to fix the sizing for mip levels
when block compressed images are viewed using uncompressed views.
My original fix didn't power the clamping, but it looks like
the clamping is required to stop the sizing going too large.
Fixes:
dEQP-VK.image.texel_view_compatible.graphic.extended*bc*
Doesn't crash DOW3 anymore.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Signed-off-by: Dave Airlie <airlied@redhat.com>
It did not signal syncobjs in the fence, and also signalled too early
if there was work on the queue already, as we have to wait till that
work is done.
Fixes: d27aaae4d2 "radv: Add external fence support."
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This fixes the scenario where the input is a struct. With this
the Unreal engines Elemental demo now works on radeonsi.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Changes since v1:
* Rebased on top of e68150de26 and
82adf53308.
Signed-off-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
shader-db doesn't show any regression and 32-bit pointers with byval
are declared as VGPRs for some reason.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
UBOs are constants buffers.
Cc: "18.0" <mesa-stable@lists.freedesktop.org>
Fixes: 41c36c45 ("amd/common: use ac_build_buffer_load() for emitting UBO loads")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Alex Smith <asmith@feralinteractive.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This attribute is similar to the definition of restrict in
C99 and it might help LLVM.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This allows to reduce the number of dwords that are loaded
with buffer_load_format_xyzw. For example, when the only used
channel is 1, the driver will emit buffer_load_format_x instead.
Shader stats for DOW3 (with some local hacky scripts for SPIRV):
143 shaders in 143 tests
Totals:
SGPRS: 5344 -> 5352 (0.15 %)
VGPRS: 3476 -> 3452 (-0.69 %)
Spilled SGPRs: 30 -> 29 (-3.33 %)
Spilled VGPRs: 0 -> 0 (0.00 %)
Private memory VGPRs: 0 -> 0 (0.00 %)
Scratch size: 0 -> 0 (0.00 %) dwords per thread
Code Size: 269860 -> 269808 (-0.02 %) bytes
LDS: 0 -> 0 (0.00 %) blocks
Max Waves: 1267 -> 1272 (0.39 %)
Wait states: 0 -> 0 (0.00 %)
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
For both versions of llvm.amdgcn.buffer.load.{format}.*.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
The GPU hangs when the driver forces a PS_PARTIAL_FLUSH after
a dispatch call (and vice versa for graphics). Something has
changed in the kernel driver because it used to work.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
It's better to do that in ac_shader_info.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
According to LLVM, only pre-GFX9 targets do not flush denorms
for fmin/fmax.
All dEQP-VK.glsl.builtin.precision.* still pass.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This seems to be broken, at least the cts tests fail.
This fixes:
dEQP-VK.renderpass.suballocation.multisample.d32_sfloat_s8_uint.samples_4
dEQP-VK.renderpass.suballocation.multisample.d32_sfloat_s8_uint.samples_8
2 samples seems to pass fine, amdvlk doesn't appear to enable TC for
possibly some other reasons here.
This is most likely a hack.
v1.1: add a bit of explaination text. (Samuel)
Fixes: ad3d98da9 (radv: enable tc compatible htile for d32s8 also.)
Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Previous code is correct but as the first case statement uses
a break, keep it consistent.
CID: 1428579
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This was just found while reading for other stuff,
src/core/hw/gfxip/gfx6/gfx6DepthStencilView.cpp.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
We need to enable the pos float location 2 mode anytime we have
persample not just when forced by the frag shader.
This fixes:
dEQP-VK.pipeline.multisample.min_sample_shading*
Fixes: 58c97a079 (radv: enable location at sample when persample is forced.)
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is ported from radeonsi and fixes:
dEQP-VK.pipeline.multisample_shader_builtin.sample_mask.bit_*
v2: don't call this path for radeonsi, it does it in the epilog.
use the radeonsi code path.
v3: handle NULL pCreateInfo->pMultisampleState properly (Samuel)
v3.1: set ps_iter_samples default to 1 (Bas)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: bdcbe7c76 (radv: add sample mask input support)
Signed-off-by: Dave Airlie <airlied@redhat.com>
radeonsi has a workaround for this, but it uses a R16A16 format,
which vulkan doesn't have, we could probably come up with a work
around but for now just avoid hw resolves.
Fixes:
dEQP-VK.renderpass.suballocation.multisample.r16g16_*norm*
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: 2a04f5481d (radv/meta: select resolve paths)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From reading AMDVLK it currently never uses hw resolve paths.
This patch takes from radeonsi which doesn't use hw resolve
for integer formats, and does the same for radv.
This fixes:
dEQP-VK.renderpass.suballocation.multisample*uint tests.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: 2a04f5481d (radv/meta: select resolve paths)
Signed-off-by: Dave Airlie <airlied@redhat.com>
Some of the hw resolve passes need the SPI color format setup
correctly.
This fixes lots of 16-bit and 32-bit format tests in
dEQP-VK.renderpass.suballocation.multisample*
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: f4e499ec79 "radv: add initial non-conformant radv vulkan driver"
Signed-off-by: Dave Airlie <airlied@redhat.com>
This did the wrong thing if we had e.g. an array for which only some
of the attributes use the instance index. Tripped up some new CTS
tests.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Fixes the following piglit tests:
arb_shader_image_load_store/layer/image3d/layered binding test
arb_shader_image_load_store/max-size/image3d max size test/2048x8x8x1
arb_shader_image_load_store/max-size/image3d max size test/8x2048x8x1
arb_shader_image_load_store/max-size/image3d max size test/8x8x2048x1
arb_shader_image_load_store/semantics/imageload/vertex shader/rgba32f/image3d test
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Fixes the following piglit test on radeonsi:
./bin/arb_enhanced_layouts-gs-stream-location-aliasing
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
addrlib asserts when that happens, and supporting it is not
required so lets not allow this for now.
It also assert on fmask, but we don't have the number of samples here.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
This gets memcpy'd and written driectly, and due to alignment, this
resulted in uninitialized gaps. This makes those gaps go away.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
The inidividual init parts don't clean up their own stuff on failure.
CC: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Dave Airlie <airlied@redhat.com>
This adds the meson.build, meson_options.txt, and a few scripts that are
used exclusively by the meson build.
v2: - Remove accidentally included changes needed to test make dist with
LLVM > 3.9
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Acked-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
The user SGPR location can change between pipelines, so we need to
emit it again to the pottentially changed SGPR index.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
When a channel was not set we also did not increase the LDS address,
while that obviously should happen.
The output loading code was inadvertently fixed which resulted in a
mismatch causing the SaschaWillems tessellation demo to result
in corrupt rendering.
Fixes: 7898eb9a60 "ac: rework load_tcs_{inputs,outputs}"
Reviewed-by: Dave Airlie <airlied@redhat.com>
The bindings also have an index field.
Fixes: 49d035122e "radv: Add single pipeline cache key."
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104677
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Passes
dEQP-VK.api.smoke.*
dEQP-VK.wsi.android.*
with android-cts-7.1_r12 .
Unlike the initial anv implementation this does
use syncobjs instead of waiting on the CPU.
This is missing meson build coverage for now.
One possible todo is that linux 4.15 now has a
sycall that allows us to export amdgpu fence to
a sync_file, which allows us not to force all
fences and semaphores to use syncobjs. However,
I had trouble with my kernel crashing regularly
with NULL pointers, and I'm not sure how beneficial
it is in the first place given that intel uses
syncobjs for all fences if available.
Reviewed-by: Dave Airlie <airlied@redhat.com>
The view index user sgpr wasn't being accounted for properly,
this refactors out the code to decide if it's required and then
uses that info to account for it.
Fixes: 180c1b924e (ac/nir: Add shader support for multiviews.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Looks like the decompress does not handle invalid encodings well,
which happens with random memory. Of course apps should not use it
with random memory, but they are allowed to ....
Fixes: 44fcf58744 "radv: Disable DCC for GENERAL layout and compute transfer dest."
Reviewed-by: Dave Airlie <airlied@redhat.com>
This shares more code and calls the new shared load_tess_varyings()
abi so that the radeonsi nir path now supports tcs output loads.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
The code to load outputs is essentially the same as load inputs
so we make the interface more generic to maximise code sharing.
We will make use of the new support in the following patch.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
This is not hooked up to any messages yet, but useful for e.g.
renderdoc if you add some messages during development.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This is an optimisation that is recommended by Matt Arsenault,
and used by RadeonSI, but it's not compatible with Vulkan.
Note that AC_FLOAT_MODE_UNSAFE_FP_MATH includes the no signed
zeros flag in LLVM.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
When that debug option is not used, we use the default float mode
because the no signed zeros optimisation is not Vulkan compatible.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This also replaces llvm.AMDGPU.kilp by llvm.AMDGPU.kill with
LLVM < 6. Similar to RadeonSI codepath.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This can't work for two reasons:
- TESSINNER/TESSOUTER are shader input values, so never translated
to the intrinsic ops
- the shader info pass scans the current stage but we want to know
in TCS, if TES reads the tess factors.
This fixes 6 regressions related to
deqp-vk/tessellation/shader_input_output/tess_level_{inner,outer}_XXX_tes
This reverts commit 5ba1a61648.
InstanceID is in VGPR2, not 1.
One more failure that CTS didn't catch up...
Reported-by: Alex Smith <asmith@feralinteractive.com>
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This shouldn't be scanned in the pipeline.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Fixes a number of int64 piglit tests, for example:
generated_tests/spec/arb_gpu_shader_int64/execution/built-in-functions/fs-sign-i64vec2.shader_test
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
V2: just zero-extend the 32-bit value.
Fixes a number of int64 piglet tests, for example:
generated_tests/spec/arb_gpu_shader_int64/execution/conversion/frag-conversion-explicit-bool-int64_t.shader_test
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes a bunch of arb_gpu_shader_fp64 piglit tests for example:
generated_tests/spec/arb_gpu_shader_fp64/execution/built-in-functions/fs-mix-double-double-double.shader_test
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
If the number of instances hasn't changed and we've already
emitted it, don't emit it again.
If the vertex shader is the same and the first_instance, vertex_offset
haven't changed don't emit them again.
This increases the fps in GL_vs_VK -t 1 -m -api vk from around 40
to around 60 here, it may not impact anything else.
Dieter also reported smoketest going from 1060->1200 fps.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Running dota2 since the below commit crashes with an llvm assert.
Trim the vector like the other user. This possible could also be
avoided by not padding inside the load vec3->vec4.
Fixes: 41c36c4549 (amd/common: use ac_build_buffer_load() for emitting UBO loads)
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This creates two new internal dependencies, idep_nir_headers and
idep_nir. The former encapsulates the generation of nir_opcodes.h and
nir_builder_opcodes.h and adding src/compiler/nir as an include path.
This ensures that any target that needs nir headers will have the
includes and that the generated headers will be generated before the
target is build. The second, idep_nir, includes the first and
additionally links to libnir.
This is intended to make it easier to avoid race conditions in the build
when using nir, since the number of consumers for libnir and it's
headers are quite high.
Acked-by: Eric Engestrom <eric.engestrom@imgtec.com>
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
These were added after adderlib was mesonified, but it still good to use
them instead of open coding them.
Acked-by: Eric Engestrom <eric.engestrom@imgtec.com>
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Currently the meosn build has a mix of two styles:
arg : [foo, ...
bar],
and
arg : [
foo, ...,
bar,
]
For consistency let's pick one. I've picked the later style, which I
think is more readable, and is more common in the mesa code base.
v2: - fix commit message
Acked-by: Eric Engestrom <eric.engestrom@imgtec.com>
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Without this we end up with the llvm error message:
"Both operands to a binary operator are not of the same type!"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Without this we end up with the llvm error message:
"Both operands to a binary operator are not of the same type!"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Without this we end up with the llvm error message:
"Both operands to a binary operator are not of the same type!"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Per spec:
"Additionally, exporting a fence payload to a handle with copy transference has the same side effects
on the source fence’s payload as executing a fence reset operation. If the fence was using a
temporarily imported payload, the fence’s prior permanent payload will be restored."
And similar for semaphores:
"Additionally, exporting a semaphore payload to a handle with copy transference has the same side
effects on the source semaphore’s payload as executing a semaphore wait operation. If the
semaphore was using a temporarily imported payload, the semaphore’s prior permanent payload
will be restored."
Fixes: 42bc25a79c "radv: Advertise sync fd import and export."
Reviewed-by: Dave Airlie <airlied@redhat.com>
Fixes the follow test for radeonsi nir:
tests/spec/arb_tessellation_shader/execution/quads.shader_test
Also stops 8 other tests from crashing, they now just fail e.g.
tcs-output-array-float-index-rd-after-barrier.shader_test
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
If they were promoted from inputs/outputs, they could have a
non-zero value left over, which messed with our store handling.
Fixes: 06f05040eb "radv: Link shaders."
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
v2: move those helpers to the header and use static inline
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> (v1)
Tested with a modified deferred demo and no regressions in a 1.0.2
mustpass run.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
The EXT values are really large, e.g.
VK_DYNAMIC_STATE_DISCARD_RECTANGLE_EXT = 1000099000, so 1 << value
is not going to fit into a 32-bit mask.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
It makes more sense to rely on nir_intrinsic_load_push_constant
instead of the pipeline layout.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
nir_to_llvm_context will always be NULL for radeonsi so we need
work around this.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Fixes the following piglit tests in radeonsi:
vs-tcs-tes-tessinner-tessouter-inputs-quads.shader_test
vs-tcs-tes-tessinner-tessouter-inputs-tris.shader_test
vs-tes-tessinner-tessouter-inputs-quads.shader_test
vs-tes-tessinner-tessouter-inputs-tris.shader_test
v2: make use of si_shader_io_get_unique_index_patch()
via the helper in the previous patch rather than
shader_io_get_unique_index()
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com> (v1)
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
VGPR1 = InstanceID / StepRate0; // StepRate0 can be set to 1
Ported from RadeonSI.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
For Vega10 and Raven that need a special workaround for the
scissor bug.
This seems to give a minor boost for Talos and Dota 2, at least.
To reduce the cost of memcmp, the driver checks if it's
really useful to do the comparison.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
VGPR1 is only needed for topology that needs 3 offsets like
triangles or quads.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Similar to RadeonSI.
This fixes:
dEQP-VK.image.texel_view_compatible.graphic.basic.attachment_read.bc*r16g16b16a16_sfloat
dEQP-VK.image.extended_usage_bit.attachment_write.r16_sfloat
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
When allocate_user_sgprs() was called, ctx->stage was actually
unset and 0 is for the vertex shader. This doesn't change
anything for now because of the spill support thing.
Though, the number of user SGPRs has to be fixed for merged
shaders on GFX9. It was broken before anyway.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This might decrease VGPR spilling, because we no longer
have to use v4i32 for 2D fetches when level == 0. We now
use v2i32 for those cases.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Copied from radeonsi.
Putting in the correct metadata flush commands for eventually not
flushing L2 on CB/DB switch.
Does not remove the need for V_028A90_CACHE_FLUSH_AND_INV_TS_EVENT
at the moment.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
These are just shaders reads, so we need to invalidate L1.
Fixes: 6dbb0eaccc "radv: handle subpass cache flushes"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
This can still be improved, but let's start with this.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
It makes more sense to move all scan stuff in the same place.
Also, we don't really need to duplicate the uses_primid field
for each stages.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
sync_files are in linux since 4.7, while the amdgpu fence_to_handle
ioctl is only in 4.15.
In particular we don't need it for sync_file in radv, because
everything happens via syncobjs, which got support earlier than
fence_to_handle.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
When rasterization is disabled we can have that few.
Fixes: 76603aa90b "radv: Drop the default viewport when 0 viewports are given."
Reviewed-by: Dave Airlie <airlied@redhat.com>
Seems like users are actually hitting 0xFFFFFFFF actually making
things broken for them, and the mad max regression is fixed, so
lets put this in once more.
v2: Use 0xf for depth-only htile. (Dave)
Fixes: af2844116f "radv: Revert HTILE reset word to 0xFFFFFFFF."
Reviewed-by: Dave Airlie <airlied@redhat.com>
Overall it does not really help or hurt. The deferred demo gets 1%
improvement and some games a 3% decrease, so I don't think this
should be enabled by default.
But with the code upstream it is easier to experiment with it.
v2: Remove initializing the registers from si_emit_config.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Those are implemented as texture sampling, so we need to make the
texture TC-compatible too.
Fixes: 34d23e82ca "radv: set some dcc parameters depending on if texture will be sampled"
Reviewed-by: Fredrik Höglund <fredrik@kde.org>
Before this DCC was in practice disabled for most games. This
enables practical DCC use. Expect a 5-10% perf increase on a
bunch of games on vega @ 4k.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
If both source and destination are DCC compressed, and their formats
are not compatible, we need to decompress one of them to make
sure we can do reinterpretation (which needs src format == dst format)
.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Apps can use this for render feedback loops, where things are
defined if they render each pixel only once. However, DCC fails
here, as the level of coherence is a block not a pixel, so disable it.
This is also going to help implementing other stuff.
Even if we optimize this later to only happen if there actually is
a loop (if possible at all ...), then the machinery is still useful
to exclude images accessible by the SDMA queue when that is implemented.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
It should already be valid there + the RB will update it during
rendering.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
For fast clear eliminate and decompressions, we always use the most compressed
format.
For clears, the code already creates a renderpass on demand with the exact same
layout as specified.
Otherwise we start distinguishing between GENERAL and TRANSFER_DST_OPTIMAL.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
We do an in place copy where we read compressed and write decompressed.
By doing this in sizes that cover entire DCC blocks and waiting for all
reads in the block before starting to write we avoid corruption.
In the end we clear the DCC metadata to 0xffffffff.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
We don't get a layout when binding to a descriptor set, but can
assume that the LAYOUT is GENERAL.
For DCC stores with the DCC bits set will result in a hang, so
better be safe than sorry.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
This reverts commit 5951578043.
The mentioned commit causes a hang in DoW3 on Vega.
Fixes: 5951578043 "radv/gfx9: fix block compression texture views."
Acked-by: Dave Airlie <airlied@redhat.com>
This should fix:
dEQP-VK.pipeline.sampler.view_type.*.format.b4g4r4a4_unorm_pack16.address_modes.all_mode_clamp_to_border_opaque_black
and a few others in that area.
Fixes: b11c4a5546 (radv: add texture descriptor/fmask/cmask support for GFX9)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
amdvlk is probably more subtle than this but it never uses
the inv cb/db variants, we fail some CTS tests without this.
Fixes:
dEQP-VK.renderpass.dedicated_allocation.formats.d32_sfloat_s8_uint.input*.
Fixes: c2fbeb7ca0 (radv: add GFX9 cache flushing support.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> (for now :-)
Signed-off-by: Dave Airlie <airlied@redhat.com>
This ports a fix from amdvlk, to fix the sizing for mip levels
when block compressed images are viewed using uncompressed views.
Fixes:
dEQP-VK.image.texel_view_compatible.graphic.extended*bc*
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes some of the broken:
dEQP-VK.synchronization.op.multi_queue.*64x64x8* tests.
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes some of the broken:
dEQP-VK.synchronization.op.multi_queue.*64x64x8* tests.
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes some of the broken:
dEQP-VK.synchronization.op.multi_queue.*64x64x8* tests.
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
It's legal to a pipeline stat query on a compute queue,
but we'd emit the wrong packet here. This should fix it to emit
the correct packet.
Noticed while inspecting the mpv hang.
Fixes: ad61eac250 (radv: factor out eop event writing code. (v2))
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The event emission wasn't sending the correct packet for gfx8 compute
queues, which explains why it works on vega fine.
This fixes the mpv vulkan hang.
Fixes: ad61eac250 (radv: factor out eop event writing code. (v2))
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
These seem mildly unstable on vega, crashing CTS in various fun ways,
and looks like leaking memory.
Disable for now, but leave the option to enable them.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Framebuffer is from 0,0, not (dst.x, dst.y).
Fixes: 69136f4e63 "radv/meta: add resolve pass using fragment/vertex shaders"
Reviewed-by: Dave Airlie <airlied@redhat.com>
The position start at (dst.x, dst.y), so if we want the source to
start at (src.x, src.y), we have to offset by (src.x-dst.x,src.y-dst.y).
Haven't tested that this fixed anything yet, but found by inspection.
Fixes: 69136f4e63 "radv/meta: add resolve pass using fragment/vertex shaders"
Reviewed-by: Dave Airlie <airlied@redhat.com>
the samples_identical instruction returns 0 if they are differet, so
we have to do the extra work if the result is 0, not if it is != 0.
Fixes: f4e499ec79 "radv: add initial non-conformant radv vulkan driver"
Reviewed-by: Dave Airlie <airlied@redhat.com>
For copies the texture unit needs to know the depth format so
it can read the htile data properly.
This fixes:
dEQP-VK.renderpass.suballocation.formats.d32_sfloat_s8_uint.load.clear
Fixes: ad3d98da9f (radv: enable tc compatible htile for d32s8 also.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This needs to correspond to the bit depth of the Z plane.
noticed in passing reading amdvlk.
Fixes: fc6c77e162 (radv: fix TC-compat HTILE with VK_FORMAT_D32_SFLOAT_S8_UINT on Vega)
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
... to set_vs_specific_input_locs().
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
The idea is to clean up the add arguments logic.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
... to declare_global_input_sgprs().
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
... to declare_vs_specific_inputs_sgprs().
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
It's a no-op anyway but it looked strange to me, remove it.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This is ported from amdvlk which sets the independent 64b blocks
only for image which will sample dcc.
I'm not sure how to port this to radeonsi.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
These are just taken from amdvlk, we probably knew these already,
but may as well port them now.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
DCC was disabled when the image format is !!supported, which is one ! too many.
Ironically the commit that introduced it was supposed to lead to more DCC use ...
Fixes: 969537d935 "radv: Add support for more DCC compression with VK_KHR_image_format_list."
Reviewed-by: Dave Airlie <airlied@redhat.com>
This fixes the layout issue for the blit path as well.
This fixes:
dEQP-VK.api.copy_and_blit.core.blit_image.all_formats.depth_stencil.d32_sfloat_s8_uint_d32_sfloat_s8_uint*
v2: use compatible render passes.
v2.1: use enum
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Cc: "17.2 17.3" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
If we are doing a general->general transfer with HIZ enabled,
we want to hit the tile surface disable bits in radv_emit_fb_ds_state,
however we never get the current layout to know we are in general
and meta hardcoded the transfer layout which is always tile enabled.
This fixes:
dEQP-VK.api.copy_and_blit.core.image_to_image.all_formats.depth_stencil.d32_sfloat_s8_uint_d32_sfloat_s8_uint.optimal_general
dEQP-VK.api.copy_and_blit.core.image_to_image.all_formats.depth_stencil.d32_sfloat_s8_uint_d32_sfloat_s8_uint.general_general
v2: refactor some shared helpers for blit patches
v3: we only need multiple render passes as they should be compatible.
v3.1: use enum (Bas)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Cc: "17.2 17.3" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This just refactors the gfx9 blit2d pipeline creation
to be less lines of code.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This add support for a 3D image reading path to the blit 2d paths,
like I did for the clear paths.
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Alex Smith <asmith@feralinteractive.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
On GFX9 we must access 3D textures with 3D samplers AFAICS.
This fixes:
dEQP-VK.api.image_clearing.core.clear_color_image.3d.single_layer
on GFX9 for me.
v1.1: fix tex->sampler_dim to dim
v2: send layer in from outside
v3: don't regress on pre-gfx9
Fixes: e38685cc62 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Tested-by: Alex Smith <asmith@feralinteractive.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
looking at traces I noticed we'd set slice_max too large sometimes.
This should fix it.
v2: fix missing - 1
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes vmfaults seen on vega with:
dEQP-VK.pipeline.multisample_interpolation.sample_interpolate_at_single_sample_.128_128_1.samples_1
These were caused by the don't allocate cmask but it was just accidental.
The actual problem was the shader was trying to get the sample positions from
a buffer, but the buffer was never getting configured to contain them, as the
previous shader never needed them.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Fixes: 1171b304f3 (radv: overhaul fragment shader sample positions.)
Signed-off-by: Dave Airlie <airlied@redhat.com>
Found by inspection.
Cc: 17.3 <mesa-stable@lists.freedesktop.org>
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
This is not because the vertex stage needs some push constants
that other stages need them too. This should reduce the number
of loaded SGPRs in some situations.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
pipeline->layout can't be NULL now.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
pipeline->layout can't be NULL now.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
They are dummy objects but the spec requires layout to not be
NULL, this just makes sure we are creating valid pipeline layout
objects. This will allow us to remove some useless checks.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
It uses slightly more memory (though still bounded by the number
of mapped ranges), but gives less quadratic behavior.
Cuts 4 minutes from the runtime of the CTS *.sparse.* tests.
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
This wasn't calculating the correct value, this along with
a nir patch fixes a regression in:
dEQP-VK.tessellation.shader_input_output.barrier
Fixes: 043d14db30 (ac/nir: don't write tcs outputs to LDS that aren't read back.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This reverts commit 2294d35b24.
We can't do this without adjusting the input SGPRs/VGPRs logic.
For now, just revert it. I will send a proper solution later.
It fixes a rendering issue in F1 2017 that CTS didn't catch up.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Alex Smith <asmith@feralinteractive.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
anv merges the tess info correctly, but radv wasn't doing this.
This fixes hangs in
dEQP-VK.tessellation.winding.default_domain.hlsl_triangles_ccw
Fixes: 60fc0544e0 (radv/pipeline: handle tessellation shader compilation)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>