aco: make IDSet sparse

Improves compilation time of huge shaders.

A ray tracing pipeline of Hellblade: Senua's Sacrifice compiles in about
half the time, with this patch.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Gitlab: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8179
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21022>

asahi: Fix scissor culling check when out of bounds for FB/viewport

Fixes a bunch of the `assert(maxx > minx && maxy > miny)` failures in
dEQP-EGL and probably others.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21678>

asahi: Do not overread user index buffers

We need to align the extent to 4 for the GPU, but we can't copy more
than the original size out of the user buffer. Always alloc the exact
size we need.

This does mean the GPU gets an IB extent that could include some other
stuff later in the pool, if not aligned. This is probably safe? Given
the base alignment, it should never cross a page boundary and fault or
anything like that.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21686>

iris: Add iris_bo_set_caching()

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21688>

iris: Move iris_bo_madvise() to i915/iris_bufmgr.c

Start to split functions that are not in hot paths to specific i915
files.

Also making it static as iris_bo_madvise() is only called from
iris_bufmgr.c and adding a enum iris_madvice to be used among all
backends.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21688>

aco/tests: add tests for v_fma_f32 with 2 fp16 literals

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21633>

blorp: add dependency on idep_intel_dev

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 3c9a8f7a6d ("intel/dev: generate helpers to identify platform workarounds")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8421
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21679>

crocus/meson: add back dependency on libintel_dev

Fixes: a0fa31bcdd8 ("intel/dev: create a helper dependency for libintel_dev")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21679>

anv: add gfx9 generated draw support

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove unused item_count parameter

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: use a list of generated shaders

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: move common shader code into header

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: pack more data into generated draws input

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: use 64bit int support in generation shaders

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: rename generated draws for Gfx11

We'll need different shaders for Gfx9

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: use a single generation shader for indirect draws

The indirect draw count shader can be used as a more generic case of
the indirect draw one. We'll never enter the last condition of the
shader (writing the MI_BATCH_BUFFER_START) with non count variants of
draws.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove MI_NOOPs at the end of the generation batch

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: make sure mi_memcpy lands before push constant loads

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: e2dc32d755 ("anv: move functions around to plan for generated draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: fix indirect draws VF cache tracking of index buffer

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: e2dc32d755 ("anv: move functions around to plan for generated draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: fix 3DSTATE_PS emission in generation shaders

We have to use the helper and also were missing the vector mask
programming.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove commented code

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove pre hasvk split assert

With softpin we should not always expect a BO in addresses.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: limit push constant dirtyness with generation shaders

We only use the fragment shader push constants.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: correctly program 3DSTATE_SF in generation shaders

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove BTI related flush in generation shaders

Earlier versions of the generation shaders were using the binding
table. We since switch to A64 messages. So the flush can be removed.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: remove copied code from generation shader

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: fix generated forward jump with more than 67M draws

The issue here is that for draw indirect count variants, we want to
jump after the last generated draw call to the next location where
commands are. But if we have more than 67M draws (8k * 8k chunks), we
only know the location once we've generated each of the 8k * 8k
chunks.

This change adds a CPU side pointer in the push constant struct so
that we can create a single linked list of chunks to edit and go
through to write the correct jump address after all the generated
space has been allocated.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: c950fe97a0 ("anv: implement generated (indexed) indirect draws")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: correctly reset generation address on command buffer reset

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

anv: fix incorrect parameter

cmd_buffer_update_dirty_vbs_for_gfx8_vb_flush takes a value RANDOM/SEQUENTIAL. Not a boolean.
Fortunately this worked okay because true == RANDOM

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

util/glsl2spirv: add support for include directive

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20497>

panfrost/ci: add EGL tests

Signed-off-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21504>

anv: pull Wa_14016118574 out of some loop not changing state

The WA is meant to be here to apply some state that is not propagated
properly inside the HW. But if you have a loop like :

for ( ... ) {
  emit(3DPRIMITIVE, some param);
}

You're not really changing any state, just push more draws into the
pipeline.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: f2645229c2 ("anv: implement Wa_14016118574")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21660>

anv: fixup condition for Wa_14016118574

We don't want the WA to kick-in if it's not point/line topology.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: f2645229c2 ("anv: implement Wa_14016118574")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21660>

radv/ci: adjust timeouts for Vega10 and Renoir

With latest CTS it takes much more time.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21390>

ci: uprev CTS to 1.3.5.0

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Acked-by: Eric Engestrom <eric@igalia.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21390>

radv: stop allocationg the attr ring BO for compute queues on GFX11

Only needed for graphics. This saves ~8Mib of 32-bit VRAM per compute
queue.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21632>

iris: Add initial skeleton of kmd backend

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21369>

iris: Use DRM_IOCTL_I915_GEM_CREATE_EXT in all supported kernels

As we start to refactor the iris code base to support Xe KMD here I'm
dropping DRM_IOCTL_I915_GEM_CREATE usage as much as possible and
unifying all graphics memory allocation calls to
DRM_IOCTL_I915_GEM_CREATE_EXT.

The kernel version that implemented DRM_I915_QUERY_MEMORY_REGIONS uAPI
also implemented DRM_IOCTL_I915_GEM_CREATE_EXT so we can use that
to safely call DRM_IOCTL_I915_GEM_CREATE_EXT.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21369>

iris/bufmgr: Add i915_gem_set_domain()

Avoids code duplication.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21369>

iris: Convert drm_i915_gem_memory_class_instance to intel_memory_class_instance

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21369>

intel/dev: Query and compute hardware topology for Xe

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel/dev: Implement Xe functions to handle hwconfig

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel/dev: Implement Xe functions to fill intel_device_info

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel/dev: Add INTEL_KMD_TYPE_XE

As mentioned in the previous patch, if intel-xe-kmd is disabled
it will fail to detected in run time but it will still compile all
Xe files.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel: Add Meson parameter to enable Xe KMD support

The plan is to compile all the Xe files but in run time it will fail
to detect the KMD loaded and it will fall back to software
rendering(if build).

Compiling Xe files makes sure newer commits don't break Xe even if
developers don't have Xe enabled in their build folder.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel: Pull in xe_drm.h

This is the uapi of the new Xe kernel driver.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21368>

intel/dev: Print required workarounds with intel_dev_info

With the addition of workarounds, the output from this tool is more
verbose than some users will want.  Provide optional parameters for
enabling hwconfig and workaround details.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21639>

radv: Expose VK_EXT_swapchain_maintenance1.

Passes dEQP-VK.wsi.*.maintenance1.*.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/win32: Implement VK_EXT_swapchain_maintenance1.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Jesse Natalie <jenatalie@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/display: Implement EXT_swapchain_maintenance1.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/wayland: Implement EXT_swapchain_maintenance1.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/common: Add function to modify present mode.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/common: Add comment about DEFERRED_ALLOCATION_BIT_EXT.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/common: Implement swapchain present fence.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/x11: Implement EXT_swapchain_maintenance1.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

wsi/common: Add common implementation of vkReleaseSwapchainImagesEXT.

Signed-off-by: Hans-Kristian Arntzen <post@arntzen-software.no>
Reviewed-by: Joshua Ashton <joshua@froggi.es>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20235>

lavapipe/ci: add a new flake

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

amd: lower multi-component subdword SSBO loads in NIR

because the hw and LLVM only support subdword single-component SSBO loads,
and ac_nir_to_llvm splits multi-component loads because of that, which is
inefficient.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

amd: lower subdword UBO loads in NIR

This fixes broken subdword UBO loads with LLVM.

It's only needed for LLVM, but it's done for both LLVM and ACO because
the pass can be fully validated only with ACO and the Vulkan CTS right now.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

ac/llvm: implement nir_op_unpack_32_4x8

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

aco: implement nir_op_unpack_32_4x8

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

ac/nir: add ac_nir_lower_subdword_loads to lower 8/16-bit loads to 32 bits

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

nir: skip nir_op_unpack_32_4x8 in nir_lower_alu_width

The pass can't handle it just like the other unpack opcodes and generates
invalid NIR.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

nir: return progress from nir_lower_io_to_scalar

oversight?

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19399>

intel/nir: Limit unaligned loads to vec4

This probably doesn't affect Vulkan or GL because they can't have
anything bigger than a vec4 anyway unless it's a u64vec4 and those have
to be at least 8B aligned.  This may affect CL apps if they use
__attribute__((packed)) on something with big vectors, depending on how
LLVM decides to translate that.

Fixes: f8aa83f0c86e ("intel/nir: Use nir_lower_mem_access_bit_sizes()")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Handle wider unaligned loads in lower_mem_access_bit_size

Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Make chunk_align_offset const in lower_mem_load()

This should make things more clear than changing the value from earlier
in the loop.  Also, rename chunk_offset to load_offset so they match.

Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Rename nir_mem_access_size_align::align_mul to align

It's a simple alignment so calling it align_mul is a bit misleading.

Suggested-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Rename align to whole_align in lower_mem_load

Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Add a combined alignment helper

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@colllabora.com>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Add UBO support to nir_lower_mem_access_bit_sizes

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Add mode filtering to lower_mem_access_bit_sizes

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

nir: Check against combined alignment in nir_lower_mem_access_bit_sizes

Checking against align_mul is insufficient if align_offset > 0.  We need
to check against the combined alignment instead.

Fixes: 2e2d7803c7f1 ("nir: Add a load/store bit size lowering pass")
Reviewed-by: M Henning <drawoc@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21524>

mesa/framebuffer: fix gl_framebuffer.resolve refcnt imbalance

Indeed, "resolve" is not freed at the gl_framebuffer destroy
stage.

For instance, this issue is triggered and detected with
"piglit/bin/fbo-depthstencil clear default_fb -samples=2 -auto"
while setting GALLIUM_REFCNT_LOG=refcnt.log.

Fixes: f5bde99cbdd2 ("gallium: plumb resolve attachments through from frontends -> pipe_framebuffer_state")
Signed-off-by: Patrick Lerda <patrick9876@free.fr>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21663>

radeonsi: assume shader is never NULL in si_emit_shader_*

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: simplify encoding VGPRS and SGPRS

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: check the pm4.reg_va_low_idx assertion unconditionally

This is not a hot path. We can always do this.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: add si_pm4_set_reg_va to simplify setting reg_va_low_idx for RGP

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: reindent code in si_state_binning.c

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: don't merge SET_* packets that have a different index in si_pm4_state

Oops.

Fixes: c8e2c6faf64 ("radeonsi: use SET_SH_REG_INDEX with index=3 for registers containing CU_EN")

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: reorganize si_emit_framebuffer_state for better readability

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi: remove Smart Access Memory because CPU access has large overhead

Related: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8176

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

radeonsi/ci: update flakes and gfx8-polaris11 results

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

amd/rtld: allow 64K LDS for all shader stages except for gfx6

Gfx6 can only use 32K LDS per workgroup.

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/21641>

amd: massively simplify how info->spi_cu_en is applied

Instead of having ac_set_reg_cu_en that sets the register, replace it with
ac_apply_cu_en that only returns the modified register value,
which allows a large simplification in both drivers because a lot of code
becomes duplicated after it's switched to ac_apply_cu_en.

RADV also didn't apply it to a few registers. Fixed.

This removes 82 lines of code in total.

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/21641>

ac/nir: don't use load_esgs_vertex_stride_amd on gfx6-8

An improvement for 9f1e6d8f70a8fa2c174e0070c4331f5f178e6f1.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

amd: query cache sizes from the kernel

Also rename l1_cache_size -> tcp_cache_size. L1 means shader array cache.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

amd,radeonsi: change enabled_rb_mask to 64 bits

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

amd: update amdgpu_drm.h

From kernel commit 817714d9665e.

Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21641>

asahi: Pull device name from device struct

This isn't filled in yet, but will be once the UAPI init function is
added.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Add agx_bo_mmap() calls to transfer path

We have the prototype for this already.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Add result buffer to context/batches

The result buffer is where the kernel places statistics and fault
information after the GPU executes a command. Dummy structure pending
UAPI.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Add agx_debug_fault() helper

We expect to forward GPU fault information to userspace. Since Mesa can
get that information, we can look up the fault address to log what was
the containing or nearest BO. Add a helper for that, so it can be called
from the driver.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Add APIs for DMA-BUF sync file import/export

These are generic ioctls, so it is safe to add them now.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Implement Linux driver scaffolding, sans UAPI

With macOS support out of the way, we can start implementing a lot of
the Linux driver interface and bookkeeping without actually adding the
UAPI proper. Let's do that to reduce the size of the UAPI patchset.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Align device submission API with upcoming UAPI

Nothing implemented, but this lets us get the batch tracking bits in,
including explicit sync/DMA-BUF integration which uses generic ioctls.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Add nocluster,sync,stats debug flags

These are only useful with the upcoming Linux UAPI, but there's no harm
in getting the debug scaffolding in now.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

asahi: Drop macOS backend

This might be useful in the future, but it is best reimplemented in
terms of the upcoming Linux UAPI instead of having parallel codepaths.
Let's drop it.

Signed-off-by: Asahi Lina <lina@asahilina.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21662>

venus: ensure invariance of buffer memory requirement size

Need to align the size for the initial cache miss.

Fixes: ef255444c19 ("venus: switch to lazy VkBuffer cache")

Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21644>

mesa: initialize VertexProgram._VaryingInputs before the first use

Noticed by code inspection.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21438>