Add V3D_TILE_ALLOC_INITIAL_BLOCK_SIZE = 128 and
V3D_TILE_ALLOC_OVERFLOW_BLOCK_SIZE = 64 to v3d_limits.h.
Corresponding _ENUM macros provide the 2-bit hardware encoding for the
TILE_BINNING_MODE_CFG packets.
The previous implicit 64B initial blocks were too small: a single draw
call emits ~88 bytes of per-tile BCL state, immediately overflowing
into continuation blocks. 128B initial blocks avoid the first
continuation allocation for simple single-draw passes.
Add v3d_tile_alloc_sizes() to v3d_util with the full tile alloc BO and
TSDA sizing logic. This uses the 128B initial blocks and tile_alloc
becomes proportional to the number of draws and size of the initial
blocks allocation with the cap of the previous fixed allocation. So
jobs with 0 or 1 drawcalls (blits/fills) reduce their headroom
dramatically.
The draw-proportional formula replaces a flat 512 KB continuation pool:
headroom = MIN2((tiles_size * draw_count) / 2, 512 KB)
Benchmarked on RPi5 (V3D 7.1) against GfxBench GL tests and
apitrace replays at 1080p. Tile-alloc memory reduction versus the
flat 512 KB headroom (taking into account 256kb kernel alloc per OOM):
GfxBench (5 benchmarks): -45% to -70% reduction, OOM at or below baseline
Apitrace (19 traces): -4% to -77% reduction on 20/24 traces
No FPS regressions observed on any workload.
Reviewed-by: Maíra Canal <mcanal@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40554>
For Valhall, use SHADDX instruction for 64-bit integer addition instead
of lowering it to 32-bit operations. The instruction sequence for doing
it in 32-bit costs 3 cycles but SHADDX only takes 2 cycles to perform.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40841>
va_mark_last currently expects 64-bit registers to always be split in
two, this commit changes it to check first if a 64-bit register is split
or not.
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40841>
According to the GL_EXT_multisampled_render_to_texture specification,
copy operations should be allowed when the extension is supported.
Previously, glCopyTexImage* would unconditionally fail with
GL_INVALID_OPERATION when copying from any multisampled framebuffer
(samples > 0), even when using render-to-texture attachments.
Fixes: d7b9da2673 ("mesa/main: fix artifacts with GL_EXT_multisampled_render_to_texture")
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Signed-off-by: Wujian Sun <wujian.sun_1@nxp.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40863>
When round_to_nearest_even is enabled with NEAREST filtering, texture
coordinates near texel boundaries (e.g. 0.9999999404) can be incorrectly
rounded up to the next texel instead of being floor()'d.
According to OpenCL spec section 8.2, for CLK_FILTER_NEAREST:
i = address_mode((int)floor(u))
Backport-to: *
Signed-off-by: Eric Guo <eric.guo@nxp.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40912>
Allocate to hold 256 wait fences. Since there is only one queue per
per ip per process, the idea is that there won't be app or windowing system
that would have large number of job dependencies / wait fences.
If there is an app that has wait fences greater than 256, there won't
be corruption issues since kernel will wait for the extra fences.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40698>
AMD APUs are hitting this case where they have very small discrete VRAM,
but a lot of staging memory, which can be used additionally.
Fixes: 7487ac2046 ("rusticl/device: support query_memory_info to retrieve available memory")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30123>
Is a single triangle is selected, it can be the case that the next iteration
can't merge any pair with the triangle. In that case, the HW node with a
single triangle will not have the highest hw_node_index, triggering an
assert.
Fixes: c18a7d0 ("radv: Emit compressed primitive nodes on GFX12")
Reviewed-by: Natalie Vock <natalie.vock@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/39655>
Required for SM6.6 in vkd3d-proton and used in a number of UE5 titles.
From descriptor side R64 images are R32G32_UINT, and to get storage_descriptor
we have to move early-return if format doesn't support rendering after
storage_descriptor setup.
Passes vkd3d-proton test:
test_shader_sm66_64bit_atomics
CTS tests:
dEQP-VK.image.atomic_operations.*.r64*
Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/39932>
Initial data is loaded via bindless_image_load, atomic swap via
bindless_image_atomic_swap.
Signed-off-by: Job Noorman <jnoorman@igalia.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/39932>
The timeout in
dEQP-VK.wsi.wayland.incremental_present.scale_none.fifo.identity.opaque.incremental_present
is already fixed in vulkan-cts-1.4.5.1.
dEQP-VK.wsi.xcb.maintenance1.present_{fences,modes}.* are known flakes,
see commit 9ec387efb1 ("panvk: advertise wsi maintenance extensions").
Signed-off-by: Janne Grunau <j@jannau.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40861>
When cross-compiling with MinGW, d3d12_resource_state.cpp fails to
compile with:
d3d12_resource_state.cpp:161:83: error: call to non-'constexpr'
function 'D3D12_RESOURCE_STATES operator|(D3D12_RESOURCE_STATES,
D3D12_RESOURCE_STATES)'
161 | D3D12_RESOURCE_STATE_ALL_SHADER_RESOURCE |
| D3D12_RESOURCE_STATE_COPY_SOURCE | D3D12_RESOURCE_STATE_COPY_DEST;
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from /usr/share/mingw-w64/include/minwindef.h:163,
from /usr/share/mingw-w64/include/windef.h:9,
from /usr/share/mingw-w64/include/windows.h:69,
from /usr/share/mingw-w64/include/rpc.h:16,
from /usr/share/mingw-w64/include/unknwn.h:7,
from ../subprojects/DirectX-Headers-1.0/include/wsl/winadapter.h:6,
from ../src/gallium/drivers/d3d12/d3d12_common.h:29,
from ../src/gallium/drivers/d3d12/d3d12_bufmgr.h:31,
from ../src/gallium/drivers/d3d12/d3d12_resource_state.cpp:24:
../subprojects/DirectX-Headers-1.0/include/directx/d3d12.h:3540:1:
note: 'D3D12_RESOURCE_STATES operator|(D3D12_RESOURCE_STATES,
D3D12_RESOURCE_STATES)' declared here
3540 | DEFINE_ENUM_FLAG_OPERATORS( D3D12_RESOURCE_STATES )
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
The DEFINE_ENUM_FLAG_OPERATORS macro in the MinGW winnt.h header
defines operator| for D3D12_RESOURCE_STATES as inline but not
constexpr. (The DirectX-Headers WSL stubs do define it as constexpr,
but when building with MinGW, windows.h is pulled in via winadapter.h
and its non-constexpr definition wins.) Calling a non-constexpr
function to initialize a constexpr variable is ill-formed in C++.
Fix by changing static constexpr to static const, which avoids the
constexpr context while still giving the variable static storage
duration.
Fixes: fe48cd7c5a ("d3d12: Allow state promotion for non-simultaneous access textures")
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40880>
Goldfish needs a non-stub QEMU pipe stream and access
to the goldfish_sync device.
All of this logic may be deleted once qemu over
virtio-gpu is fully enabled.
TEST=run meson2hermetic + m vulkan.ranchu
Reviewed-by: Bo Hu <bohu@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40851>
The issues with this are proving to be difficult to solve. Turn it off
for now until we have a proper fix.
Fixes: c24963d8da ("nvk: Enable zcull for VK_ATTACHMENT_LOAD_OP_LOAD")
Acked-by: Mary Guillemard <mary@mary.zone>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40894>
Fixes the following building error happening with clang:
FAILED: src/amd/vulkan/libvulkan_radeon.so.p/radv_pipeline_rt.c.o
...
../src/amd/vulkan/radv_pipeline_rt.c:1050:44: error: use of GNU empty initializer extension [-Werror,-Wgnu-empty-initializer]
struct radv_shader_stage prolog_stage = {};
^
1 error generated.
Fixes: afe51940 ("radv: Rewrite the RT prolog in NIR")
Reviewed-by: Natalie Vock <natalie.vock@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40893>
In all other cases, we immediately set atomic cap after acquiring a drm
fd:
- wsi_init
- AcquireXlib
- AcquireDrm
wsi_GetDrmDisplayEXT is the only function left where the atomic cap is
set by wsi_display_get_connector. This commit moves the set atomic cap
into wsi_GetDrmDisplayEXT.
Signed-off-by: Yuxuan Shui <yshui@codeweavers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/39166>
Usually connector property IDs are acquired in
wsi_display_get_connector, which is called by wsi_get_connectors, and in
turn by vkGetPhysicalDeviceDisplayProperties2KHR and
vkGetPhysicalDeviceDisplayPlanePropertiesKHR. Except if the drm fd is
not available when these functions are called. Which will be the case if
vkAcquireXlibDisplayEXT is not called first.
So it goes like this. First, the display is created in
vkGetRandROutputDisplayEXT. Then it's used in
vkGetPhysicalDeviceDisplayPlanePropertiesKHR, but since the drm fd is
not available at this point, connector property IDs are not initialized.
Later, this display is used in vkAcquireXlibDisplayEXT, which also
doesn't touch the property IDs. Finally in drm_atomic_commit, the
atomic commit fails with EINVAL, specifically because of the
uninitialized ID of the "CRTC_ID" property. Since it's one of the
properties drm_atomic_commit tries to set.
This commit makes sure that find_connector_properties is called in
vkAcquireXlibDisplayEXT to initialize the property IDs.
Fixes: 513ffea1d3 ("wsi/display: use atomic mode setting")
Signed-off-by: Yuxuan Shui <yshui@codeweavers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/39166>
Without this, printf messages that were sent before a crash will be
lost, which makes shader printf pretty useless for debugging crashes.
Signed-off-by: Olivia Lee <olivia.lee@collabora.com>
Reviewed-by: Lars-Ivar Hesselberg Simonsen <lars-ivar.simonsen@arm.com>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/40816>
