Upstream LLVM commit 75c7bca740935a0cca462e28475dd6b046a6872c removed
the option to pass a Module pointer to DataLayout's constructor.
This commit does the same, that upstream did for their code in clover.
Reference: 75c7bca740
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11863
Signed-off-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31350>
This implement PIPE_CAP_CONTEXT_PRIORITY_MASK and handle priority flags
for v10.
This effectively expose EGL_IMG_context_priority and
EGL_NV_context_priority_realtime.
Expose of what is currently supported from the panthor uAPI.
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30991>
This adds support for the new DEV_QUERY_GROUP_PRIORITIES_INFO query from
panthor to report and will be used to report appropriate priority mask
in the Gallium driver.
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30991>
Some drivers are able to optimize buffer usage when it is known that
external visibility of buffer changes only needs to be guaranteed
after well defined transition points, like eglSwapBuffers.
Add a BO flag to specify this behavior and set it with GBM surfaces.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30077>
Importing an opaque fd no longer returns the mapped pointer but
a pointer to llvmpipe_memory_allocation to support both memfds and
dma bufs. This saves the pointer with the correct type and we can
unwrap the mapped data in a correct way.
Fixes: d74ea2c117 ("llvmpipe: Implement dmabuf handling")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31022>
If the buffer is not unmapped then a reference to the DMA buffer
is still kept. The DRM device is then not properly cleaned up.
This can cause the application to keep a DRM device open which is
still DRM master without any visible open file descriptors.
Fixes: d74ea2c117 ("llvmpipe: Implement dmabuf handling")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31022>
Make the external memory object a reference counted object. Without
that deleting the memory object in OpenGL would also delete and unmap
the memfd, which causes any operation on a texture or buffer that is
bound to that memory object to access invalid memory.
Fixes: 1608a815e3 ("llvmpipe: add support for EXT_memory_object(_fd)")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31022>
During compute state save/restore, let's track all the descriptor sets.
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30798>
Instead of including always the simulator file and have the code under
`#ifdefs` depending if the simulator is available or not, let's not
include the simulator directory entirely if the simulator is not
available.
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31356>
This is only relevant for hierarchical coding using sub-layers.
Fixes: 72f52329cd ("vulkan/video: add a nal_unit lookup for hevc")
Reviewed-by: Hyunjun Ko <zzoon@igalia.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Colin Marc <hi@colinmarc.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31303>
If the pipeline is created with no
VkPipelineSampleLocationsStateCreateInfoEXT, but
VK_DYNAMIC_STATE_SAMPLE_LOCATIONS_ENABLE_EXT is set, we end up
dereferencing a NULL pointer.
Fixes future dEQP-VK.pipeline.*.extended_dynamic_state.*.sample_locations_enable_no_create_info
Fixes: 1deb83fb86 ("vulkan: Add more dynamic multisample states")
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31377>
The problem with radeonsi+ACO is that UBO loads from vec4 uniforms using
only 1 component always load all 4 components. This fixes that.
We are only interested in shrinking UBO and SSBO loads, but I added more
intrinsics because why not.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29384>
As in OpenGL, if alpha-to-one is required for Vulkan generate a
blend shader for it (since the hardware does not support alpha_to_one
natively).
Signed-off-by: Eric R. Smith <eric.smith@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31194>
When SAMPLE_ALPHA_TO_ONE is enabled we need to use a blend shader, as
the BlendDescriptor.alpha_to_one bit is deprecated (does not work)
on bifrost and valhall.
In order to generate the appopriate blend shader code, we must put the
alpha_to_one status into pan_blend_state.
Signed-off-by: Eric R. Smith <eric.smith@collabora.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31194>
The " - far:" pattern never landed upstream (yet) but was part of some
enhanced fault debugging we have in the CrOS kernel. Update crashdec to
handle both cases.
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31349>
These target features have been remove in LLVM 19, and cause
warnings like the following to be printed:
'-avx512er' is not a recognized feature for this target (ignoring feature)
'-avx512pf' is not a recognized feature for this target (ignoring feature)
Do not pass these target features on LLVM 19 and newer.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11870
Reviewed-by: Dave Airlie <airlied@redhat.com>
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31321>
This extension extends EGL_IMG_context_priority with a new
EGL_CONTEXT_PRIORITY_REALTIME_NV attribute.
Effectively, Gallium drivers would need to implement
PIPE_CONTEXT_REALTIME_PRIORITY flag in context_create and expose
PIPE_CONTEXT_PRIORITY_REALTIME on PIPE_CAP_CONTEXT_PRIORITY_MASK to
enable this extension.
Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30989>
When tests were added, there was a single pipe (float), so there wasn't
a pipe to compare in `operator==`. Add it there now and adjust
expectations accordingly.
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31335>
To prepare for factoring out a larger part of panvk_cmd_draw to a
separate prepare function, move tiler idvs flag calculation to a
separate function.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31314>
There is no need to update the index_array_size register unless the draw
is indexed and the index buffer has changed. For non-indexed draws, this
field is not read.
To achieve this, add a new function prepare_index_buffer that updates
both the size and the pointer registers in case the state is dirty.
As this decouples the index buffer size register from the drawcall, set
the full index buffer size.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31314>
MALI_ATTRIBUTE_FREQUENCY_INSTANCE allows us to use the passed vertex
attribute divisor directly without having to multiply it by vertex
count.
This decouples the vertex attribute building from drawcalls.
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31314>
