Commit 2aaa6ebb "build/amd: add amd-use-llvm build option"
allows to build radeonsi and radv without LLVM dependency
so LLVM constraint is applicable when amd-use-llvm=true
Fixes: 82047fa8 ("amd: drop support for LLVM 15, 16, 17")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37695>
We don't need to have this restriction, we just need to fence_flush()
before trying to access the fd. This path is exercised in CL CTS
gl-interop test if cl_gl_sharing is exposed.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37696>
If we need to blit a resource to a buffer that can be exported (ie.
because it was suballocated, etc), always use the aux-context. This
can be called off of the driver thread, so it isn't thread-safe to
use the main context, even if it is not NULL.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37696>
The layout_resource_for_handle() path is only intended for images/
textures. For buffers, call fdl_layout_buffer() directly.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37696>
As long as fdl is happy to layout the imported buffer type, we shouldn't
restrict it further. The cl_gl_sharing extends import/export beyond
winsys buffers, and these restrictions were getting in the way.
Signed-off-by: Rob Clark <rob.clark@oss.qualcomm.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37696>
The bindless sampler heap was introduced in Gfx11 which ELK doesn't
support.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Alyssa Anne Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37692>
It's always set to a fixed value and not used in many places. Use the
value directly where it's needed.
Suggested-by: Lucas Fryzek <lfryzek@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37648>
With EXT_shader_object, it became possible to compile shaders
independently and then use them together later, so we cannot rely on the
lack of task shader data to decide that no task shader will be used. The
flag VK_SHADER_CREATE_NO_TASK_SHADER_BIT_EXT exists for that purpose,
but it doesn't really make any difference for us. Always assume that if
the mesh shader is reading the task payload, it's going to be used with
one, as otherwise the application is doing it wrong.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13983
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37648>
Make use of the prebuilt cuttlefish packages from the Debian
Trixie base. These packages are built and published to an apt
repository under ci-deb-repo
Signed-off-by: Ritesh Raj Sarraf <ritesh.sarraf@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37633>
Set the has_f2i32_rtne shader compiler option to indicate hardware
support for it. This enables NIR's late algebraic optimization pass to
generate more efficient code for float-to-int conversions.
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Simon Perretta <simon.perretta@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37680>
Sometimes the compute shader workgroup size requires a larger SIMD width
than the minimum in order to fit in the available threads. In that case
we'll skip the SIMD8 shader, and need to try SIMD16 regardless of how
the register pressure estimate looks.
Fixes: 3af4e63061 ("brw: Skip compilation of larger SIMDs when pressure is too high")
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Tested-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37649>
Since commit a7727f692f, MESA_TRACE_SCOPE take printf-like
arguments, so it is easier to write a meaningful label for the CPU
trace.
This commit replaces one MESA_TRACE_FUNC to provide extra info, and as
we are here, we add another MESA_TRACE_SCOPE and a MESA_TRACE_FUNC to
track the bo allocation/freeing.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37639>
I didn't bother switching either iris or elk/hasvk but one could.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37517>
to allow drivers to do more complex printing without needing a subclass
paradigm. To be used on Intel.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37517>
It remove a duplication and also it will be used in a future patch
from other file.
Reviewed-by: Caio Oliveira <caio.oliveira@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37670>
LLVM PR#146819 changed the signature of `setObjectLinkingLayerCreator`,
dropping the Triple argument. The PR was first included in the LLVM 21
series, and the new signature is gated behind a version check for that.
`LLVMOrcThreadSafeContextGetContext` was removed in LLVM commit b18e5b6,
and the ORC examples in the LLVM tree seem to just create a context
instead, which we replicate here.
With this commit, mesa successfully builds the llvmpipe gallium driver
on riscv64 with LLVM 21.1.0.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13785
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13852
Reviewed-by: David Heidelberg <david@ixit.cz>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37027>
Different threads calling gallivm share the same LPJIT (and the
underlying LLJIT) instance, which could be only bound to a single cache
object at the same time.
Pass the object cache when looking up the symbol and put it under the
protect of lookup_mutex to prevent accessing wrong cache.
This seems to fix some MissingSymbolDefinitions error when running
Plasma Shell with llvmpipe on RISC-V.
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37497>
This header contains a lot of includes, and gets everywhere. Let's make
sure we don't include it from headers, which makes this much easier to
manage.
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37432>
