asahi: Fix agx_set_framebuffer_state for MRT

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

asahi: Set data_valid for the correct level

By inspection.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

asahi: Implement stencil texturing

Stencil texturing is easy: S8_UINT is textured like R8_UINT (with a
little swizzle fixup), and stencil is always S8_UINT thanks to
u_transfer_helper. So we just need to do some fixups to make
u_transfer_helper's seperate_stencil work and everything will work out.

Passes dEQP-GLES31.functional.stencil_texturing.*

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

asahi: Add internal formats for RGB10A2

We need to use I16 as the interchange format here. Fixes:

   dEQP-GLES3.functional.fragment_out.basic.uint.rgb10_a2ui*

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

asahi: Implement perf_debug

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

asahi: Free the scanout resource

Fixes memory leaks with renderonly.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Handle collects in backwards isel

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Assert more invariants in RA

Was helpful for debugging.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Validate part of SSA form

To debug backend pass problems.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Split off NIR preprocessing from compiling

So we can specialize after lowering I/O.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Handle scalar texture destinations

Fixes dEQP-GLES3.functional.shaders.texture_functions.texturelod.sampler2dshadow_fragment.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Don't depend on GenXML

Separation of concerns, unused #include.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

agx: Lower packs/unpacks and bitfields

Needed for GLES3. These could be optimized.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19811>

util: Move src/gallium/include/pipe/p_format.h to src/util/format/u_formats.h

Because p_format.h shared between vulkan drivers and opengl drivers

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19629>

ci/freedreno: disable flaking Civilization V

See: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7491

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19862>

util: Rename PIPE_ALIGN_STACK to UTIL_ALIGN_STACK and moved into util/compiler.h

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

llvmpipe: fixes error: ‘enum pipe_blendfactor’ declared  in lp_test_blend.c

error  message:
In file included from ../../src/gallium/drivers/llvmpipe/lp_test_blend.c:44:
../../src/gallium/drivers/llvmpipe/lp_bld_blend.h:47:21: error: ‘enum pipe_blendfactor’ declared inside parameter list will not be visible outside of this definition or declaration [-Werror]
   47 |                enum pipe_blendfactor factor_src,

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

util: convert usage of uint to unsigned in u_format.h

Because uint comes from pipe/p_compiler.h, and that depends will be removed in future

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

util: use void * instead of byte * for util_copy_rect in u_format.h

As ubyte comes from p_compiler.h,  so do not use it,
and the code calles util_copy_rect may use args with type uint8_t*, ubyte* or unsigned char*,
so use the type void* that consistence with memcpy

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

tgsi: fixes error: ‘enum pipe_shader_type’ declared  in tgsi_info.h

error message:
../../src/gallium/auxiliary/tgsi/tgsi_info.h:92:30: error: ‘enum pipe_shader_type’ declared inside parameter list will not be visible outside of this definition or declaration [-Werror]
   92 | tgsi_get_processor_name(enum pipe_shader_type processor);

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

aco: fixes error: 'uint' was not declared in aco_instruction_selection.cpp

uint is from pipe/p_compiler.h

error message:
../../src/amd/compiler/aco_instruction_selection.cpp:11061:4: error: 'uint' was not declared in this scope; did you mean 'rint'?
11061 |    uint en_mask = 1;

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

util: Remove the usage of enum pipe_error in u_hash_table.*

Use 0 instead of PIPE_OK in u_hash_table.c

Because pipe_error and PIPE_OK is comes from pipe/p_defines.h that doesn't belong
to src/util/

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

util: Remove redundant #include "util/u_inlines.h" in u_trace.c

util/u_inlines.h is comes from src/gallium/auxiliary/util/u_inlines.h,
so when possible, do not use it under src/util folder

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19676>

intel/compiler: Remove redundant argument from brw_nir_create_passthrough_tcs

Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19831>

intel/perf: fix printf formatting of size_t for 32bit builds

Fixes: bdacd6df5adc ("intel/perf: add a non installable tool to print metrics")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7736
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19856>

ci/dzn: add flake

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19858>

docs: update Python requirement to 3.7

Reported in https://gitlab.freedesktop.org/mesa/mesa/-/issues/7716

Reviewed-by: Eric Engestrom <eric@igalia.com>
Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19791>

include: Update the OpenCL headers

This syncs all our headers with their canonical versions from Khronos.
Of particular note, CL headers repo does not have cl.hpp and cl2.hpp has
been depricated all in favor of opencl.hpp which is the new canonical
header.  Fortunately, nothing in Mesa uses any of the C++ headers so
this shouldn't break anything.

Acked-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19816>

clover: Use braces arround a union initializer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19816>

scripts: Fix khronos-update.py to use main for OpenCL headers

Also, cl.hpp no longer exists in OpenCL-Headers to fix it at the last
commit where it did exist.

Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19816>

Update 00-mesa-defaults.conf

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19619>

nir/lower_explicit_io: fix metadata preserve

This pass can insert if blocks, therefore no dominance/block_index for
you.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable
Reviewed-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19818>

nir/lower_bit_size: lower uadd_sat/iadd_sat/isub_sat to unsaturated alu

The unsaturated arithmetic won't overflow/borrow, and may be faster.

No fossil-db changes.

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/19473>

nir/lower_bit_size: optimize usub_sat lowering

The result should never be larger than uint_max. This doesn't need a
special path.

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/19473>

nir/algebraic: optimize open-coded uadd_sat/usub_sat

fossil-db (navi21):
Totals from 19 (0.01% of 135636) affected shaders:
Instrs: 40730 -> 40688 (-0.10%)
CodeSize: 217708 -> 217568 (-0.06%)
Latency: 261466 -> 261373 (-0.04%)
InvThroughput: 74944 -> 74896 (-0.06%)

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/19473>

radv: lower 8/16-bit uadd_carry/usub_borrow

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7615
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19473>

nir/lower_bit_size: lower uadd_carry

8/16-bit uadd_carry can exist in SPIR-V.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Gitlab: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7615
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19473>

crocus: Pass devinfo directly in crocus_setup_uniforms

Instead of reaching through brw_compiler.  This will make easy
future changes on brw_compiler side.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19834>

radv: advertise extendedDynamicState3ColorBlendEnable

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/19579>

radv: add support for dynamic color blend enable

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/19579>

ci/freedreno: temporarily disable godot-trive

Test if this trace impact surrounding traces causing rendering changes.

Ref: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7732

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/panfrost: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/iris: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/llvmpipe: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/virgl: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/broadcom: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/freedreno: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

ci/amd: drop glmark2 traces, useless

Signed-off-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19838>

nir/lower_shader_calls: Get rid of any brw occurences

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19749>

tu: Rewrite inline uniform implementation

Now we always push the inline uniforms, using an indirect CP_LOAD_STATE.
There is just enough space to be able to always push them if we expose
the minimum possible per-stage limits. This aligns us with Qualcomm and
removes the need to setup a uniform descriptor which will be
problematic.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18703>

nir: Fix nir_chase_binding() vecN handling

In the comments we claimed to handle vecN instructions, for the case
where an offset is trimmed from the descriptor, but we didn't ignore the
offset itself and in effect only handled identity vecN's (which copy
propagation would normally remove already!), so the handling of vecN was
useless and this relied on copy propagation cleaning things up. Fix it
to ignore everything except the components in the original source.

Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18703>

pvr: Minor cleanup around pvr_emit_vdm_index_list()

Signed-off-by: Matt Coster <matt.coster@imgtec.com>
Reviewed-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19802>

vulkan/wsi/wayland: Set num_modifier_lists = 0 if num_drm_modifiers == 0

This case was missed in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18826 ,
resulting in native Wayland apps failing with

 ../src/vulkan/wsi/wsi_common_drm.c:452: wsi_configure_native_image: Assertion `!"Failed to find a supported modifier!  This should never " "happen because LINEAR should always be available"' failed.

if the Wayland compositor advertises only the INVALID modifier.

Fixes: c315e20d6198 ("vulkan/wsi/wayland: Configure images via params passed to wsi_swapchain_init()")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19795>

pvr: Support clear colours with more than 32 bits

The support for this seems to already exist, but was blocked by a
conditional. Replace with an assert to ensure the format can be packed
into 4x32 bit words (true for all currently supported formats).

Signed-off-by: Matt Coster <matt.coster@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19799>

pvr: Fix clang-format error.

Signed-off-by: Karmjit Mahil <Karmjit.Mahil@imgtec.com>
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19798>

virgl/video: Check driver supported profiles and entrypoints

Since the support of video by the device and the driver may be
different, it is necessary to check on the driver side as well.

Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

virgl/video: Add support for H.265 encoding

Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Signed-off-by: Weishi Li <liweishi@kylinos.cn>
Signed-off-by: Liming Sun <sunliming@kylinos.cn>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

virgl/video: Add support for H.264 encoding

Implement encoding framework and support H.264 encoding.

Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Signed-off-by: Ming Xie <xieming@kylinos.cn>
Signed-off-by: Liming Sun <sunliming@kylinos.cn>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

virgl/video: Pass "max_references" to backend when creating a video codec

The "max_references" is an important value in sequence parameters,
which will be used when encoding SPS headers.

Signed-off-by: Boyuan Zhang <Boyuan.Zhang@amd.com>
Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

frontends/va/enc: save intra_idr_period in the context

The va frontend loses some information when processing parameter
buffers, but the virgl video encoding driver needs it to work.

Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

virgl/video: Move hardware-related definitions from virgl_video.h to virgl_video_hw.h

Considering the increasing number of hardware-related definitions,
it is better to put them in a separate file.

Signed-off-by: Feng Jiang <jiangfeng@kylinos.cn>
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18831>

iris: Increase the maximum grid size

I had simply copied these values from another driver when adding initial
compute support to iris.  The actual hardware limit is UINT32_MAX (see
the GPGPU_WALKER/COMPUTE_WALKER ThreadGroupID{X,Y,Z}Dimension fields).

Thanks to Karol Herbst for noticing the unnecessarily low limit.

References: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7676
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Tested-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19826>

turnip: set ycbcr feature bits properly

Set ycbcr feature bits only for ycbcr formats.  Besides, we can control
chroma locations only for planar formats and we support
VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_SEPARATE_RECONSTRUCTION_FILTER_BIT
on newer gens.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19609>

freedreno: add has_separate_chroma_filter to fd_dev_info

The blob driver does not support
VK_FORMAT_FEATURE_SAMPLED_IMAGE_YCBCR_CONVERSION_SEPARATE_RECONSTRUCTION_FILTER_BIT
before a6xx_gen3.  It still sets CHROMA_LINEAR bit according to
chromaFilter, but the bit has no effect before a6xx_gen3 (confirmed on
a618 with blob version 512.490.0).

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19609>

freedreno: Find curses properly for building fdperf

Upstream ncurses doesn't ship any legacy ncurses.pc pkg-config symlinks
(just ncursesw.pc) which Void Linux strictly follows[1] for example.

Meson (since 0.54) has a generic curses dependency which works perfectly
in this case[2], thus also bump the toplevel minimum meson version from
0.53.

[1] https://github.com/void-linux/void-packages/tree/master/srcpkgs/ncurses
[2] https://mesonbuild.com/Dependencies.html#curses

Signed-off-by: Jami Kettunen <jami.kettunen@protonmail.com>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19664>

ci: Update dozen flakes

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/7549

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19824>

svga: Fixes -Werror,-Wunused-but-set-variable for clang-15 in svga/svga_tgsi_vgpu10.c

error message:
../../src/gallium/drivers/svga/svga_tgsi_vgpu10.c:5199:13: error: variable 'indicesMask' set but not used [-Werror,-Wunused-but-set-variable]   unsigned indicesMask = 0;
            ^

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19527>

softpipe: fixes -Werror,-Wunused-but-set-variable for clang-15 in sp_tile_cache.c

error message:
../../src/gallium/drivers/softpipe/sp_tile_cache.c:346:9: error: variable 'numCleared' set but not used [-Werror,-Wunused-but-set-variable]
   uint numCleared = 0;
        ^

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19527>

radv: fixes -Werror,-Wunused-but-set-variable for clang-15 in radv_shader_args.c

error message:
../../src/amd/vulkan/radv_shader_args.c:457:16: error: variable 'arg_count' set but not used [-Werror,-Wunused-but-set-variable]
      unsigned arg_count = 0;
               ^

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19527>

intel: fixes -Werror,-Wunused-but-set-variable for clang-15

one of those error message:
../../src/intel/compiler/brw_vec4_cmod_propagation.cpp:53:8: error: variable 'ip' set but not used [-Werror,-Wunused-but-set-variable]
   int ip = block->end_ip + 1;

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19527>

intel: Fixes -Werror,-Wbitwise-instead-of-logical for clang-15 in brw_nir_lower_shader_calls.c

error message:
error: use of bitwise '|' with boolean operands [-Werror,-Wbitwise-instead-of-logical]

Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19527>

r300: don't use abs twice in transform_TRUNC on older hardware

r400:
total instructions in shared programs: 27594 -> 27523 (-0.26%)
instructions in affected programs: 3594 -> 3523 (-1.98%)
helped: 61
HURT: 0
helped stats (abs) min: 1 max: 4 x̄: 1.16 x̃: 1
helped stats (rel) min: 1.16% max: 8.70% x̄: 2.55% x̃: 1.64%
95% mean confidence interval for instructions value: -1.29 -1.04
95% mean confidence interval for instructions %-change: -3.13% -1.97%
Instructions are helped.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19807>

meson: Enable initialized-but-unused warning for MSVC

Reviewed-by: Yonggang Luo <luoyonggang@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19792>

gallium/vl: Remove unused win32 screen var

Reviewed-By: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19792>

docs: update calendar for 22.3.0-rc3

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19822>

zink/ci: Add automatic test on Adreno 618 with Turnip

Make the existing manual test automatic, sharding to 3 boards.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19071>

ci/freedreno: Test on a618

Do the same testing we currently do on a630.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19071>

ci/turnip: Shard the tests on a618 over more boards

12 kingoftown Chromebooks have been added to Collabora's LAVA lab.

Use them to run the same tests as before, as the jobs never take less
than 10 minutes.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19071>

radv/rra: Fix copying accel structs that were not built yet

In the case that radv_GetEventStatus always returns true, the loop will
never exit.

Fixes: 5749806 ("radv: Add Radeon Raytracing Analyzer trace dumping utilities")
Reviewed-by: Friedrich Vock <friedrich.vock@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19772>

vulkan: Update the XML and headers to 1.3.235

Acked-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19815>

d3d12: Video Screen - Do not crash if HEVC not supported, return no support instead

Fixes: 8f654b90 ("d3d12: Fix HEVC wrong caps detection due to bad parenthesis in condition")
Signed-off-by: Sil Vilerino <sivileri@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19813>

iris: Get VM id from iris_bufmgr's fd

iris_bufmgr_init_global_vm() was getting the VM id from the callers
fd not the iris_bufmgr's duplicated fd.

If i915 driver decides to always return unique vm ids, like a counter
that is initialized per device. So fd0 would have vm id = 0 by
default, fd1(dup of fd0) would have vm id = 1... in this scenario it
would cause a mismatch. But it is not the current i915 implementation.

Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19467>

intel/perf: add ACM/DG2 performance metrics

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/18893>

intel/perf: use the new OA format for Gfx12.5+

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/18893>

drm-uapi: import i915_drm.h changes for the new OA report type

To be replace by a proper drm-next update.

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/18893>

intel/perf: fix B/C counters accumulation in non query mode

When we're not using queries, all the counters from the
MI_REPORT_PERF_COUNT are available. This is the case when using
perfetto with the global pps datasource that capture global counter
values.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 8750f43a9077 ("intel/perf: add performance query layout using MI_SRM")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18893>

anv: tweak performance query timeout based on number of passes

This avoids device lost events when we replay a command buffer 1k
times on DG2.

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/18893>

anv/perf: fixup counter/query mapping

The intel_perf_counter_pass::pass field is actually useless and
invalid.

Once you have mapped all the counters to all the metrics, the order of
the metrics capture is dictated by intel_perf_get_n_passes().

When reading values that is the order we should follow.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 2001a80d4a81 ("anv: Implement VK_KHR_performance_query")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18893>

intel/perf: simplify pass computation loop

We don't need to go through all the metric sets as we're already built
a bitset matching per counter to figure out in which metric set a
particular counter is.

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/18893>

intel/perf: don't ralloc on perf context a temporary hash table

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/18893>

intel/perf: allocate cleared counter infos

This array of structure needs to be initialized to 0 as it contains a
bitset we don't explicitly clear.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 3144bc1d3369 ("intel/perf: move query_mask and location out of gen_perf_query_counter")
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/18893>

intel/perf: add a non installable tool to print metrics

Useful to look at the layout of the queries.

v2: Rework based on Marcin's comment

v3: Rebase

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/18893>

intel/perf: fix overflow in index types

With DG2, the number of perf groups and metrics climbs into the
thousands.  16bit fields are not sufficient for storing metrics
indices, and the build throws warnings when compiling the generated
intel_perf_metrics.c

Use a 32bit integer for these values.

Reviewed-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/18893>

intel/perf: support new variable names

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/18893>

intel/perf: support new operators for upcoming metrics

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/18893>

intel/perf: support more than 64 queries

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/18893>

intel/perf: fix variable type assumption error

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/18893>

ci: don't remove wget from the packages of debian/x86_test-vk

The 'wget' package was in the list of ephemeral packages used only during the
build. 'Wget' installation comes from the build for the 'x86_build-base', and
the inclusion on this list only affects its purge. Jobs that use this image,
like 'venus-lavapipe' in 'virglrenderer', cannot use it as other GL jobs can.
So, please don't consider this package ephemeral when it comes from the upper
layer builder.

This 'wget' package is a known dependency on a job that uses this image. So,
even if this package is installed in the base image, it is good to have it
explicitly mentioned in the packages to install.

Signed-off-by: Sergi Blanch Torne <sergi.blanch.torne@collabora.com>
Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19784>

ci: insert debian/x86_test-vk in image tags

Using the variable DEBIAN_X86_TEST_IMAGE_PATH in a Vulkan test, we realize the
reference to the Vulkan one is missing. Even more, hardcoded within mesa. With
this, a variable for Vulkan image has a reference VK on it, as well as the GL
one. But the old variable is still there, pointing to the value that can be
expected to avoid any brake on other CI elements.

Also, updated the build tag to regenerate the image.

Signed-off-by: Sergi Blanch Torne <sergi.blanch.torne@collabora.com>
Reviewed-by: David Heidelberg <david.heidelberg@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19784>

anv: get rid of ilog2_round_up

__builtin_clz(value - 1) is undefined for with value=1 (because
__builtin_clz(0) is undefined).

Because we set rt_pipeline->stack_size = 1 when a ray tracing pipeline
doesn't need any stack allocation to differentiate from a dynamic size
(rt_pipeline->stack_size = 0) we can run into this undefinied behavior
issue.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: f68d64dac015 ("anv: Add support for vkCmdSetRayTracingPipelineStackSizeKHR")
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19781>