Rob Clark
7a7ac8cd40
freedreno/ir3: Fix reg size validation
...
8b types also live in half-regs
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
8a6934dfe8
freedreno/ir3: Fix load/store_global_ir3 type
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
81eefe0090
freedreno/ir3: 8bit fixes
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
f7b2d613c5
freedreno/ir3: 16b bools
...
A create_immed_for_instr() type thing could be useful to make the immed
type match other src(s) for instruction..
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
9a2562a545
freedreno/ir3: Deal with zero-source instructions
...
Needed by the next patch, which starts treating bools as 16bit exposing
a bug that was previously accidentially hidden for instructions like
ELECT_MACRO. Needed for next patch.
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
b6e11225a2
freedreno: Fix set_global_binding
...
The gallium interface is a bit awkward, but pointer sizes are actually
64b despite what the API suggests.
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
0a35ba5c43
freedreno/ir3: Move lower_idiv_options
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
e544a9db16
freedreno/ir3: Add support for load_kernel_input
...
Used for function arguments to compute kernels (ie. OpenCL).
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
e10c76d277
freedreno/ir3: implement load_work_dim intrinsic
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
3bd265a393
freedreno/ir3: vec8+vec16 support
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
f5bbf77be8
freedreno: implement set_compute_state()
...
This interface should really go away, but for now just implement it
to directly update constant state (ie. what clover *should* be doing
instead)
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
1e9f27f37f
freedreno/ir3: Handle MESA_SHADER_KERNEL
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
83a1bca952
freedreno: Skip built-in shaders for clover
...
Avoids assert:
../src/compiler/glsl_types.cpp:1134: static const glsl_type *glsl_type::get_array_instance(const glsl_type *, unsigned int, unsigned int): Assertion `glsl_type_users > 0' failed.
caused by us trying to compile built-in shaders (ie. clear, gmem<->mem,
etc) before clover has initialized glsl_types. But we don't need these
shaders for compute-only contexts.
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
dff8a0c4cb
isaspec: inherite parent's bitset gpu gen requirements
...
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
d77b9fb518
isaspec: Fix gpu_id for default_options
...
We forgot to set this. It starts to matter in the next patch, otherwise
pre-pass to detect branch targets (needed for backwards jumps/branches)
will not work.
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
c4e7541b9d
freedreno/ir3: use stg.a/ldg.a only if offset is reg or doesn't fit
...
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
064c806d23
freedreno/ir3: Add load/store_global lowering
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
d85eb9268a
freedreno/ir3: set proper dst size for {store,load}_{global,shared}_ir3
...
We want to pass 64b variables.
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
1ef43a0be7
freedreno/ir3: disallow immediate addr/offset for ldg/ldg.a
...
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
f45b7c58c4
freedreno/ir3: Lower 64b phis
...
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
bee9212efb
ir3/freedreno: add 64b undef lowering
...
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Rob Clark
2d65e6f56d
freedreno/ir3: 64b intrinsic lowering
...
Both for OpenCL and VK_KHR_buffer_device_address
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Danylo Piliaiev
1eee1fda11
nir/lower_amul: do not lower 64bit amul to imul24
...
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com >
Signed-off-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13300 >
2021-10-21 18:59:57 +00:00
Sagar Ghuge
b83c9b21a6
intel/compiler: Set correct cache policy for A64 byte scattered read
...
This doesn't impact any performance since the previous typo value
matches the current cache control value.
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com >
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13458 >
2021-10-21 17:32:23 +00:00
Marek Olšák
272af39be1
amd/addrlib: cosmetic addrlib update
...
Acked-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/13459 >
2021-10-21 16:26:06 +00:00
Marek Olšák
69a1b02b68
amd/addrlib: change how the license is formatted to match internal tree
...
It's the same MIT license.
Acked-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/13459 >
2021-10-21 16:26:06 +00:00
Sajeesh Sidharthan
cf0bc4fb55
frontends/va/av1: handle multiple slice params
...
Multiple slice params in a single vaRenderPicture function call
is not handled. This patch will fix overwriting slice params
when multiple slice params received in one buffer.
Change-Id: I880df5bc35dfbd64382a178074482548882ee4af
Signed-off-by: Sajeesh Sidharthan <sajeesh.sidharthan@amd.com >
Reviewed-by: Leo Liu <leo.liu@amd.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13463 >
2021-10-21 16:12:30 +00:00
Samuel Pitoiset
996e81fb70
aco: fix loading 64-bit inputs with fragment shaders
...
Fixes a bunch of 64-bit IO tests with piglit and Zink.
Cc: 21.3 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13454 >
2021-10-21 12:50:55 +02:00
Iago Toral Quiroga
75bd37dc6a
broadcom/compiler: disallow tsy barrier in thrsw delay slots
...
A TSY barrier becomes effective at the point of the next thread switch,
so if we have one coming after a previous thread switch we need to
be careful not to emit it in its delay slots, or we would be effectively
moving the barrier earlier than intended.
Fixes simulator assert crash in:
dEQP-VK.graphicsfuzz.two-for-loops-with-barrier-function
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13468 >
2021-10-21 12:40:00 +02:00
Emma Anholt
9202e8cbaf
turnip: Make copy_format() and tu6_plane_format() return pipe_format
...
[ Connor: Keep the argument to copy_format() a VkFormat, fold in
plane_format() conversion. ]
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Emma Anholt
68f8bbb37e
util: Move freedreno's snorm-to-unorm to util/, adding remaining cases.
...
I want it in turnip too.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Emma Anholt
cbdc8e09bf
turnip: Switch format_to_ifmt() to take a pipe_format.
...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Emma Anholt
e4e8db0132
turnip: Switch tu6_format_color() to a pipe_format.
...
To handle Y8 specially, we want a PIPE_FORMAT instead of VK_FORMAT. There
are some redundant vk-to-pipe conversions, but they're going to go away
shortly.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Emma Anholt
3b68fc0c6a
turnip: Switch tu6_format_texture() to a pipe_format.
...
To handle Y8 specially, we want a PIPE_FORMAT instead of VK_FORMAT.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Connor Abbott
cfabdbd7d3
tu/clear_blit: Move around copy_format()/tu6_plane_format()
...
We want these functions to take a Vulkan format and return a
pipe_format, but tu6_plane_format() was getting redundantly called on
the result of copy_format() and copy_format() was also getting called
twice with image to image copies. Pull these functions further up the
call chain so that they're only called once.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13379 >
2021-10-21 08:46:31 +00:00
Iago Toral Quiroga
acb83e1b13
v3dv: enable Vulkan 1.1
...
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13465 >
2021-10-21 10:12:38 +02:00
Emma Anholt
bd81a23620
ci/piglit-runner: Fix funny indentation of the piglit-runner command.
...
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
440f207a1f
ci/deqp-runner: Move more non-suite logic under the non-suite 'if'.
...
Changing these variables won't do anything for you otherwise.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >.
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
92748e40ef
ci/deqp-runner: Don't start GPU hang detection for making junit results.
...
It's just CPU-side post-processing, not running tests.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
61ca900b69
ci/deqp-runner: Drop LD_LIBRARY_PATH=/usr/local for libkms workaround.
...
deqp hasn't been linking against that in quite some time.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
899174c210
ci/deqp-runner: Move remaining asan runs to --env LD_PRELOAD=
...
This should improve their reliability and speed a little by getting
deqp-runner off of asan. This removes the last jobs setting
TEST_LD_PRELOAD, so remove passing that variable around from other
scripts.
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
37c690ad1a
ci/deqp-runner: Drop silly CSV env vars.
...
One was unused, the other was used once.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
b978688df6
ci/deqp-runner: Use new deqp-runner's built-in renderer/version checks.
...
This is prettier in the log files, less shell code, and for non-suite mode
adds checking that the driver has the right git sha1. Also, no need for
suites to have a DEQP_VER to say which dEQP we should run for the renderer
check.
The version checks can help us make sure that GL version exposed doesn't
accidentally regress, and the ".*git" checks that we're using a git
version of Mesa rather than something that snuck in through distro
packages.
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
9ddfd297e0
ci/deqp-runner: Simplify the --jobs argument setup.
...
We can use the general "how parallel should we go on this runner?" env var
and save a bunch of massaging env var names. Fixes how PIGLIT_PARALLEL
looked like it was useful but actually wasn't passed through to HW
runners.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Emma Anholt
59f3a8e6b4
ci/deqp-runner: Drop SUMMARY_LIMIT env var.
...
Nobody uses it any more, and you could just put it in DEQP_OPTIONS.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com >
Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13372 >
2021-10-21 07:34:19 +00:00
Vinson Lee
670fd8123b
radv: Fix memory leak on error path.
...
Fix defect reported by Coverity Scan.
Resource leak (RESOURCE_LEAK)
leaked_storage: Variable prolog going out of scope leaks the storage it points to
Fixes: 80841196b2 ("radv: implement dynamic vertex input state using vertex shader prologs")
Suggested-by: Rhys Perry <pendingchaos02@gmail.com >
Signed-off-by: Vinson Lee <vlee@freedesktop.org >
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13402 >
2021-10-21 07:14:40 +00:00
Samuel Pitoiset
b797ecac7a
ac/rgp: remove useless code related to GFX6-7
...
RGP only supports GFX8+. RADV doesn't allow SQTT on < GFX8 and
RadeonSI only allows it on GFX9+.
Signed-off-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/13451 >
2021-10-21 06:44:50 +00:00
Samuel Pitoiset
8304392c35
radv: add an assertion to prevent GPU hangs when VRS isn't supported
...
Just hit this case with a buggy CTS test.
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/13361 >
2021-10-21 08:07:54 +02:00
Caio Marcelo de Oliveira Filho
9a32a7fdfe
util: Move test sources to tests/ directory
...
Acked-by: Matt Turner <mattst88@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13425 >
2021-10-20 21:40:31 -07:00
Caio Marcelo de Oliveira Filho
abf2af64ac
util: Convert sparse array multithread test to use gtest
...
Reviewed-by: Dylan Baker <dylan@pnwbakers.com >
Acked-by: Matt Turner <mattst88@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13425 >
2021-10-20 21:40:31 -07:00