rendering onto a linear-tiled image is unbelievably slow if any sort of
blending is enabled, so instead always render to optimal tiling and then
copy to linear for scanout
this doubles performance for now and can be deleted in its entirety along
with the rest of the related hacks once real wsi support is implemented
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10180>
`qpu.raddr_b` is an unsigned int, so it is always positive, even after
casting to signed int.
Fixes CID#1438117 "Operands don't affect result
(CONSTANT_EXPRESSION_RESULT)":
"result_independent_of_operands: (int)inst->qpu.raddr_b >= -16 is
always true regardless of the values of its operands. This occurs as
the logical first operand of "&&".
v2:
- Use signed pointers (Iago)
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10131>
Wrap the si_cp_wait_mem call to emit RGP_SQTT_MARKER_IDENTIFIER_BARRIER_START and
RGP_SQTT_MARKER_IDENTIFIER_BARRIER_END events.
Only for gfx9+ for now.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10105>
The devices have been running KernelCI jobs for a good while without
glitches, let's reenable Mesa jobs on them now.
This reverts commit c6fbbbbf70.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10173>
The devices have been running KernelCI jobs for a good while without
glitches, let's reenable Mesa jobs on them now.
This reverts commit 40647fcc3d.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10173>
Driver.Draw is now unused for Gallium drivers - except for st_cb_feedback which
sets its own Draw function.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10066>
__builtin_popcount is not available on all compilers.
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10159>
Some logic was tuned to buffers / atomics / images being supported in
frag stages in order to expose any support at all. Fix some of these
assumptions.
Signed-off-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/10161>
These are a bit special. Among other things, removing them will cause us
to potentially remove the load itself, defeating the purpose of the
locking. Also it's unclear whether it's legal to access the shared
memory directly when it's locked like this.
This only comes up on nv50, since on nvc0+, shared memory can't be
loaded from random ops.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Acked-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
It doesn't matter since it's 0, but all the offsets are in bytes whereas
the method expects words. So adjust by 2.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
We were previously dumping $r127 in there. This has a bad effect on
nv50, so make sure we allocate an actual register for it, even if
there's nothing using the result.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Acked-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
There does not appear to be an instruction form for this. Prevent an
immediate from being loaded into place.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
The idea is that buffers will be bound to the appropriate indices. That
means that we can just rename them to global.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Acked-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
The upstream logic will not end up using an address, so we have to force
it here. The other backends don't care either.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
This is probably not the most efficient way to go for all geometries,
but the assumption is that kernels tend to be x/y-heavy rather than
z-heavy. Iterates over each z slice and passes in the current value via
user param. (And bump all user params by a dword.)
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
This makes compute mostly work. For now we're laying out images/buffers
in a fixed offset from each other in the globals "array", but this
should be done dynamically. We're also missing passing image info to
shaders, as well as adding image formats to a shader key.
Heavily inspired by nvc0 variants of these.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Acked-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
With shader images / buffers, we can get more complex barrier requests.
This mirrors the logic in nvc0.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
Heavily inpsired by the nvc0 code. Note that these only exist for the
compute stage, so there is no shared-based indexing.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
This contains the logic updates necessary to perform necessary resource
tracking and emit update / flush commands for the relevant stages.
Inspired by some changes from Pierre Moreau.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Pierre Moreau <dev@pmoreau.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
For legacy reasons, we were using the PIPE order, instead of the
hardware order. Reorder the stages to match the order of the
BIND_TIC/etc methods, and adjust internal usage to match.
Signed-off-by: Pierre Moreau <dev@pmoreau.org>
[imirkin: fixed numbering, removed TODO comments]
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9299>
These were noticed by trial and error after fixing the fp16 render
setup.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9916>
Format 0x26 is invalid, formats are in a 4 bit field so they repeat
in increments of 16.
Frame reg flags needs to set 0x01 to actually enable fp16.
The clear color setup is also a bit different for fp16, need to use
the 16 bit values in the first two clear color registers.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9916>
Sources outside of src/util path should include "util/string_buffer.h"
Fixes the following building error in Android:
external/mesa/src/compiler/glsl/ast_type.cpp:25:10: fatal error: 'string_buffer.h' file not found
^~~~~~~~~~~~~~~~~
1 error generated.
Fixes: eeec9d56ad ("compiler/glsl: clean up output")
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10156>
Bring the scripts for parsing, dumping state and diffing of Gallium
trace files to modern day by updating them to Python 3.
Add option '-p' to some tools for outputting only plaintext
instead of ANSI / colorized format.
Also fix state parsing of some dumps by adding 'clear_render_target'
and 'get_disk_shader_cache' to ignored calls list.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4321
Signed-off-by: Matti Hamalainen <ccr@tnsp.org>
Acked-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9830>
We're currently using uint arrays instead of uvec4 arrays for UBOs and
SSBOs, which requires scalarBlockLayout. Fixing this is a lot of work,
so let's just document it for now.
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10127>
When lowering precision on integers from GLSL ES, we can end up with
16 bit integer loop counters. So let's tolerate this as well.
This was probably not caught earlier because most NIR drivers disable
GLSL-level loop-unrolling, and no non-NIR driver sets LowerPrecisionInt16
to true. This was discovered while trying to wire up int16 support for
Zink, which doesn't currently disable GLSL loop-unrolling.
Reviewed-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10125>
