Use the RT descriptors pointer as dummy VA.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Natalie Vock <natalie.vock@gmx.de>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41780>
To avoid weird crashes caused by trying to load unused attributes.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Natalie Vock <natalie.vock@gmx.de>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41780>
can_bind_const_buffer_as_vertex is a direct copy of
screen->caps.can_bind_const_buffer_as_vertex.
Read the cap from the screen directly.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
allow_st_finalize_nir_twice is a direct copy of
screen->caps.call_finalize_nir_in_linker.
Read the cap from the screen directly at each call site.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
prefer_real_buffer_in_constbuf0 is a direct copy of
screen->caps.prefer_real_buffer_in_constbuf0.
Read the cap from the screen directly.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
lower_two_sided_color is the negation of screen->caps.two_sided_color.
Replace all call sites with !screen->caps.two_sided_color directly.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
has_indirect_partial_stride is a direct copy of
screen->caps.multi_draw_indirect_partial_stride.
Read the cap from the screen directly.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
has_multi_draw_indirect is a direct copy of
screen->caps.multi_draw_indirect.
Read the cap from the screen directly at each call site.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
has_shareable_shaders is a direct copy of screen->caps.shareable_shaders.
Read the cap from the screen directly at each call site.
Fix st_format_test which was missing st_context.screen initialization,
exposed by this change.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
This cap is a direct alias of screen->caps.astc_void_extents_need_denorm_flush.
Replace all st_context usages with the screen cap directly.
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41184>
More common and this implicitly enables this for Path Of Exile and X4
Foundations. Though, zero VRAM allocs is already the default in AMDGPU,
so that doesn't change anything in practice (except for very old
kernels).
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41735>
libdrm dups the fd internally, so local_fd and get_fd() are different
fd number but they point to the same file descriptor. Close it right
after the amdgpu device is initialized to avoid keeping two fds open
for the same thing.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41753>
Update the with_gfx_compute flag so decode-only builds do not
unnecessarily compile gfx/compute code.
Signed-off-by: Thong Thai <thong.thai@amd.com>
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41493>
Removes gfx, tgsi and driver_trace code from auxiliary for
non-gfx/compute builds, to reduce the number of files that gets
compiled.
Signed-off-by: Thong Thai <thong.thai@amd.com>
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41493>
This allows rusticl to make use of the native fma instructions giving us
better OpenCL performance.
e.g. ProjectPhysX_OpenCL-Benchmark on my GA102:
FP32 0.610 -> 11.474 TFLOPs/s
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41720>
Vulkan dispatch commands should multiply group count by local size,
but pre-compiled dispatches should not. For example, the predicate
indirect shader has a local size of 32 and a grid size equal to the
max draw count, which was resulting in a dispatch total grid size of
(max_draw_count * 32), overrunning the buffer.
Reviewed-by: Aitor Camacho <aitor@lunarg.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41740>
Ported from HoneyKrisp upload pools, with some adaptations such as
providing the MTLBuffer and offset for use with certain commands.
Reviewed-by: Aitor Camacho <aitor@lunarg.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41740>
The register values will depend on new fields in PS_STATE and it doesn't
seem like dynamic state belongs in radv_emit_fragment_shader_state.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41689>
to reduce the number of initialized PS VGPRs, increasing the PS wave launch
rate.
The pass will have more RADV-specific stuff.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41689>
These are SGPR inputs, so they are uniform in subgroups but may
have different values in different subgroups.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41584>
These intrinsics are generally divergent between different
subgroups, but they can be uniform when all their sources
are also uniform.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41584>
AMD SMEM instructions are always uniform within a subgroup,
but they may be divergent across subgroups, ie. each subgroup
may have a different value from the same SMEM instruction.
This needs to be considered for divergence across subgroups
as well as for vertex divergence, because vertices of the
same primitive may be split between different waves.
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/41584>
