review.tizen.org Git - platform/upstream/mesa.git/log

asahi: Skip draws if the scissor culls everything

We can't pack the scissor descriptor for these, and there would be no rendering
anyway, so detect this condition and skip the draw.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Mark scissor dirty if rast->scissor changes

Although there is a scissor enable bit in the hardware rasterizer state, we
cannot rely on it alone as we also "scissor" to the viewport.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Track scissor states

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Dirty track viewport descriptor

Mitigates the extra CPU cost from packing in the previous commit, and
avoids the redundant memcpy.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Defer viewport pack

Nontrivial interaction between viewport state and scissor state, so
defer until draw time to deal with it.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Add scissor enable bit

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Add "set scissor" command

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

asahi: Fix scissor descriptor definition

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11084>

docs: Even more gratutious nitpicks

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11085>

agx: Zero r0l before first use of control flow

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add break/continue support

Following Dougall's notes closely.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Implement loops in the simplest way

Again, optimizations are possible, but for now go for conformance.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Optimize out empty else blocks

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Implement emit_if the simplest way

Lots of optimizations are possible from here.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Fix up branch offsets at pack time

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Model jump instructions

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Pack control flow instructions

Nearly ALU.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add push_exec alias

Use the same canonical form as Metal, so the disassembler can alias it
back.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Model pop_exec

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Model control flow instructions

Thankfully the encoding is regular, if a bit complicated.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add inner loop nesting count field

Needed for proper handling of break/continue with nested if-else.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add branch target to IR

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add invert_cond (ccn) to IR

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Add nest field to IR

Needed to model control flow instructions.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Track block offsets

For fixing branch offsets.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Track current_block

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Implement boolean mov

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Enable 1-bit load_const

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Pack ld_var Dx

In the expected place.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

agx: Drop cmdline version back to ES3.0

The patch needed for this to work was dropped. Fixes the standalone
compiler.

Fixes: 972409dacb9 ("asahi: Stub command-line compiler for AGX G13B")
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

asahi: Fix meson.build definition to depend on agx_pack.h

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11023>

doc: Gratuituous promotion of Wayland

This is really just a dummy commit to get the docs regenerated so
we can stop 404ing.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11083>

gv100/ir: fix quadop/pop lowering

Fixes: texture_cube_map_array.sampling

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11061>

gm107/ir: emit barrier sources for quadon/pop

We drop them later on, but it's actually how that needs to be done on
Volta.

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11061>

gv100/ir: add support for barrier thread state files for OP_CVT

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11061>

nv50/ir: add barrier and thread_state files

Those are required on Turing+ to handle cross thread synchronization and
other goodies as the c/r stack is gone.

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11061>

nv50/ir/ra: fixes upcoming barrier file

Replaces some FILE_ADDRESS with LAST_REGISTER_FILE and makes RA not choke
on instructions using TS values.

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11061>

CI: Disable rk3399-gru-kevin jobs for now

We lost a few machines from the rota due to an infrastructure issue and
we don't have enough for good parallelism.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11082>

vulkan/util: generate vk_dispatch_table that combines all dispatch tables

Zink uses this, as it doesn't need to differentiate all the entrypoints.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11045>

radeon/vcn: enable parsing support for st_rps_bits

Set st_rps_bits in hevc message buffer and set corresponding flag to indicate
that st_rps_bits will be used for parsing the short_term_ref_pic_set structure.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10889>

frontends/vdpau: disable UseStRpsBits for vdpau hevc

vdpau interface doesn't provide st_rps_bits, it uses NumDeltaPocsOfRefRpsIdx
instead. So disabling the flag to indicate st_rps_bits will not be used.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10889>

frontends/va: get st_rps_bits from VA pic param hevc

Get st_rps_bits from VAPictureParameterBufferHEVC, and set the flag that
indicates st_rps_bits will be used for parsing the short_term_ref_pic_set
structure

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10889>

vl: add st_rps_bits for HEVC decode

st_rps_bits is used for accelorater to skip parsing the short_term_ref_pic_set
structure, which is needed for HEVC decode.

Signed-off-by: Boyuan Zhang <boyuan.zhang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10889>

travis: Download XQuartz from GitHub.

Bintray was sunsetted on May 1st, 2021.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11026>

util/prim_restart: revert part of bad fix

When drawing using util_translate_prim_restart_ib, zink explicitly
ignores pipe_draw_start_count_bias::start, because
util_translate_prim_restart_ib used to create a new index-buffer without
padding at the start.

This makes a lot of sense, because creating a padded index buffer is
just wasteful.

So let's walk back on the choice of starting to pad the output buffer.

Fixes: 1272c2e0524 ("util/prim_restart: fix util_translate_prim_restart_ib")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4851
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11059>

zink: add support for string-markers

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11004>

zink: untangle have_EXT_debug_utils and ZINK_DEBUG_VALIDATION

EXT_debug_utils is useful for more than just validation, so let's
untangle these a bit.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11004>

zink: implement half-float packing

This cap isn't really optional for drivers that implement GL SPIR-V
support, so let's just implement it.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4846
Reviewed-by: Hoe Hao Cheng <haochengho12907@gmail.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11057>

zink: mark 2d-arrays as cube-compatible

OpenGL allows reinterpreting 2D array textures as cubemaps,
so we need to always set the cube-compatible flag for them.

This fixes a piglit test on Lavapipe.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10913>

zink: only mark resources as cube-compatible if supported

We need to check if the number of samples are valid for the image
first if we've going to set the cube-compatible bit.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10913>

lavapipe: do not interpret cube-compatible as cubemap

Now that we handle the cubemap-array views properly, we no longer need
to do this. It's allowed in Gallium to create cubemap views of 2d
arrays, so this should work fine.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10913>

lavapipe: handle cube-array image-views

It seems this was missed in the past.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10913>

venus: fix mismatched bo mmap_size for export and multiple imports

Spec requires apps to use the size returned from
vkGetAndroidHardwareBufferPropertiesANDROID for AHB import, which
includes dedicated image/buffer import and non-dedicated buffer import.

Spec requires venus to use the size from image and buffer memory
requirement for dma_buf fd import if it's dedicated. If not dedicated,
the actual payload size should be used.

For AHB export allocation of VkImage, the app provided size will be 0,
and internally we must use the size from image memory requirement.

For AHB export allocation of VkBuffer, the app provided size comes from
buffer memory requirements and it can be smaller than the actual AHB
size due to page alignment. Internally that's the size we must use.

For AHB import, app must use the size from AHB prop query. Internally,
we have to override that with the size from memory requirement if the
import operation is dedicated to a VkImage or a VkBuffer. If not
dedicated, the actual payload size should be used.

The not working scenario is:
1. App creates an AHB backed VkBuffer, and the exported AHB size is
larger than the buffer memory requirement (very common).
2. App imports the AHB without a dedicated VkBuffer. Then the entire
AHB payload will be imported and the mmap_size might increase.

Test: dEQP-VK.api.external.memory.android_hardware_buffer.*

Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11055>

venus: refactor for property query of dma_buf fd

With a TODO to route host storage size instead of that from the guest.

Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11055>

venus: fix misaligned bo_flags between import and query

For importing an dma_buf fd, export info is not required. Leaving the
bo_flags missing VIRTGPU_BLOB_FLAG_USE_SHAREABLE bit as well as the
VIRTGPU_BLOB_FLAG_USE_CROSS_DEVICE bit. Upon querying fd properties,
DMA_BUF handle type will be specified which generates a new bo_flags
not matching the prior one.

This patch aligns the above 2 bits for the dma_buf import path.

Test: vkGetAndroidHardwareBufferPropertiesANDROID should succeed with
a valid AHB with AHARDWAREBUFFER_FORMAT_BLOB format.

Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11055>

venus: rename dmabuf to dma_buf when it represents a type

Signed-off-by: Yiwei Zhang <zzyiwei@chromium.org>
Reviewed-by: Chia-I Wu <olvaffe@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11055>

svga: fix texture rectangle sampling when no sampler view declaration is defined

It is valid to not have a sampler view declaration for the corresponding
sampler in a TGSI shader, and hence we should not rely on the sampler view
declaration to determine if we need to adjust the unnormalized coordinates
for texture rectangle sampling.

This patch is to prep for tgsi shaders that are translated from nir which
in many cases do not issue sampler view declarations.

Fixes: 584b1070372a0e ("st/mesa: Drop the TGSI paths for drawpixels and use nir-to-tgsi")

Reviewed-by: Neha Bhende <bhenden@vmware.com
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11011>

svga: Use shader_key info to declare resources if TGSI shader is missing it

Sometimes, TGSI shader doesn't have SVIEW declaration if it is not
utilize in shader. In such cases, declare those resources with the
help of information stored in shader key.

Fixes: 584b1070372a0e ("st/mesa: Drop the TGSI paths for drawpixels and use nir-to-tgsi")

Tested with piglit, gleretrace

Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11011>

svga: Add target and sampler_return_type info into shader key

Fixes: 584b1070372a0e ("st/mesa: Drop the TGSI paths for drawpixels and use nir-to-tgsi")

Reviewed-by: Charmaine Lee <charmainel@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11011>

anv_android: Add missing type

Fixes: 2797c75426cf ("anv: fix AHB leak upon exportable allocation")
Signed-off-by: Roman Stratiienko <r.stratiienko@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11068>

venus: ignore pInheritanceInfo when we should

pInheritanceInfo should be ignored when the command buffer is primary.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4850
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11067>

venus: remember cmd buffer level and queue family

We need them for wsi queue ownership transfer and others.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11067>

venus: add struct vn_command_buffer_builder

We are going to remember more states.

Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Reviewed-by: Yiwei Zhang <zzyiwei@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11067>

lavapipe: add more format mappings for vertex buffer formats

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10957>

mesa: don't save/restore VAO NumUpdates and IsDynamic to fix update tracking

NumUpdates is used to indicate whether a VAO is static or dynamic, but if
we restored an older value, it could incorrectly indicate that it's not
dynamic.

This fixes a small performance issue with torcs.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

mesa: don't call FLUSH_VERTICES in glPopClientAttrib

These states don't affect draws, so any enqueued Begin/End vertices don't
need flushing.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

mesa: optimize unreferencing VBOs in glPopClientAttrib

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

mesa: optimize glPush/PopClientAttrib for GL_CLIENT_VERTEX_ARRAY_BIT

This improves performance in torcs by 6%.

The idea is to skip saving and restoring vertex attribs and bindings that
have never been changed.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

mesa: don't call _mesa_set_draw_vao in glPushClientAttrib

Pushing states doesn't affect draws.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

mesa: move _mesa_copy_vertex_attrib/buffer functions to their only use

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10994>

aux/indices: employ Delete The Code methodology

the util functions will assert for unknown prim types, so the switches
are redundant now

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10965>

aux/indices: break out index count conversion into separate function

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10965>

aux/indices: break out index size conversion to separate function

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10965>

aux/indices: break out primitive type conversion to separate function

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10965>

llvmpipe: ci updates

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11049>

gallivm: fix oob imageLoad with formats that have <4 components

it's expected that these loads will have full alpha

fixes spec@arb_shader_image_load_store@invalid

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11049>

aux/trace: avoid deadlock in screen::flush_frontbuffer hook

this hook usually requires mapping a resource, which will trigger a
trace dump internally, which requires locking the output mutex, so
finish the trace dump before calling it

Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11052>

aco: Eliminate SALU comparison when SCC can be used instead.

For example:

s0, scc = s_and_u32 ...
scc = s_cmp_eq_u32 s0, 0
p_cbranch_sccz

is turned into:

s0, scc = s_and_u32 ...
p_cbranch_sccnz

Fossil DB results on Sienna Cichlid:

Totals from 85267 (56.91% of 149839) affected shaders:
CodeSize: 202539256 -> 202237268 (-0.15%)
Instrs: 38964493 -> 38888996 (-0.19%)
Latency: 750062328 -> 749913450 (-0.02%); split: -0.02%, +0.00%
InvThroughput: 167408952 -> 167405157 (-0.00%)

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

aco: Use s_cbranch_vccz/nz in post-RA optimization.

A simple post-RA optimization which takes advantage of the
s_cbranch_vccz and s_cbranch_vccnz instructions.

It works on the following pattern:

vcc = v_cmp ...
scc = s_and vcc, exec
p_cbranch scc

The result looks like this:

vcc = v_cmp ...
p_cbranch vcc

Fossil DB results on Sienna Cichlid:

Totals from 4814 (3.21% of 149839) affected shaders:
CodeSize: 15371176 -> 15345964 (-0.16%)
Instrs: 3028557 -> 3022254 (-0.21%)
Latency: 21872753 -> 21823476 (-0.23%); split: -0.23%, +0.00%
InvThroughput: 4470282 -> 4468691 (-0.04%); split: -0.04%, +0.00%

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

aco: Introduce a new, post-RA optimizer.

This commit adds the skeleton of a new ACO post-RA optimizer,
which is intended to be a simple pass called after RA, and
is meant to do code changes which can only be done
after RA.

It is currently empty, the actual optimizations will be added
in their own commits. It only has a DCE pass, which deletes
some dead code generated by the spiller.

Fossil DB results on Sienna Cichlid:

Totals from 375 (0.25% of 149839) affected shaders:
CodeSize: 2933056 -> 2907192 (-0.88%)
Instrs: 534154 -> 530706 (-0.65%)
Latency: 12088064 -> 12084907 (-0.03%); split: -0.03%, +0.00%
InvThroughput: 4433454 -> 4432421 (-0.02%); split: -0.02%, +0.00%
Copies: 81649 -> 78203 (-4.22%)

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

aco: New writeout overloads for the test framework.

These will be used by future tests.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

aco: Add Operand(Temp, PhysReg) constructor.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

aco: Don't DCE instructions that write non-temps, eg. exec.

No Fossil DB changes.
This commit makes DCE usable after RA.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7779>

Revert "ci: Configure DUTs for max performance"

This reverts commit eef5409df4c3f1192687eaa6157875e4650c1206.

I suspect this has caused a lot of the CI instability today -- some flakes
were already added, but the a630-traces job is still flaking. Revert
until a fix makes it stable.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11024>

radv: fix generating hang reports if mutable descriptors are used

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/11031>

v3dv: remove unused v3dv_zs_buffer_from_vk_format

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11050>

mesa: fix error set for glCompressedTexSubImage calls

Desktop and ES expect a different error code here when dealing with
generic compressed format tokens. This fixes failures with upcoming
new tests for compressed texture related API calls.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11009>

mesa/st: Don't assert !unify_interfaces in the passthrough edge flags case

Technically, this is correct, and I think it's the right long-term
solution. However, Iris has code that undoes the damage caused by
nir_lower_passthrough_edgeflags (see iris_fix_edge_flags in
iris_program.c), so it should be safe to do the lowering here.

I'm not marking this as closing mesa#4838 because I think we should move
the Iris code up to here, make a different version of the NIR pass, or
something different... to properly fix this problem. In the mean time,
this gets a bunch of tests to stop crashing. :)

Fixes: a76ec17f12c ("mesa/st: Fix iris regression with clip distances.")
Acked-by: Emma Anholt <emma@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11013>

turnip: Make sure that SNORM blits don't clamp ambiguous -1.0 values.

The CTS expects that some paths transfer SNORM data exactly, but the HW
will clamp 0x80 values to 0x81 in the process. We can treat snorm as
unorm, though, and get working compression without the clamping happening.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10735>

turnip: Reorganize copy_format()'s switch statement.

Now that we need FALLTHROUGH macros we weren't saving much by falling
through, and things were weirdly ordered anyway with depth intermixed with
color formats and the default case tucked in the middle of the switch
statement. Replace with pretty obvious ordering of normal color, planar
color, depth, then default.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10735>

nv50/ir: don't optimize shl(mul_hi, a) to mul_hi

Fixes shader_storage_buffer_object.advanced-unsizedArrayLength* CTS tests

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11021>

nv50/ir: when constant folding shl(mul, a) we need to copy muls type

Also clear subop while at it as we might convert a shl.wrap to mul.

Signed-off-by: Karol Herbst <kherbst@redhat.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11021>

panfrost: Performance configuration

Configure a panfrost performance structure with tables of categories and
counters for the current product id. An array for storing counter values
read from the GPU is also managed by this structure. A generic read
function can be used to retrieve the value of a counter from the conter
values array.

v2: Generate tables instead of calling register functions.
v3: Simplify counter read function and `pan_gen_perf.py` write method.
v4: Accumulate counter values from all cores.
v5: Wrap `STATIC_ASSERT`s within unused functions.

Signed-off-by: Antonio Caggiano <antonio.caggiano@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10844>

panfrost: Counter definitions

Add Mali events XML files generated with Panfrost HWC helper.
https://gitlab.freedesktop.org/fahien/panfrost-hwc-helper/

v2: Restore license headers.
v3: Fix shader core and memory system counter offsets.

Signed-off-by: Antonio Caggiano <antonio.caggiano@collabora.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10844>

st/mesa: fix clearing of 1D array textures

st_gl_texture_dims_to_pipe_dims isn't usable in this case since its
API doesn't match the usage so fix up the dimensions manually.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3646

Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10523>

ci/freedreno: Skip Portal 2 trace on a630, due to flakiness

Sometimes the reticle is missing. Skip it for now to keep the number of
pipeline failures due to flakes low.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11033>

radv: fix missing default state for DB_DFSM_CONTROL

Fixes: 69ae02151d7 ("radv: remove DFSM")
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/11028>

zink: improve unsupported feature warning message

this should print the device name and the missing features in order to
be less opaque

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10959>

zink: use actual const for const offset

When we emit constants, we don't know what type they'll be used as, so
we just emit them as uint, and then bitcast them to whatever we need.

But this isn't a good idea for ConstOffset, which needs to actually be a
const value, and not a const value bitcasted. So we sadly have to
open-code the const emitting here to avoid the problem.

Fixes: e963d35efe1 ("zink: use ConstOffset for nir_tex_src_offset")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4831
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11032>