In order to do this we have to modify push constant set up to handle
ranges. We also have to tweak the way we handle dirty bits a bit so
that we re-push whenever a descriptor set changes.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
There are several places where we look up opcodes in an array of stages.
Assert that the we don't end up going out-of-bounds.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
We want to call brw_nir_analyze_ubo_ranges immedately after
anv_nir_apply_pipeline_layout and it badly wants constants. We could
run an optimization step and let constant folding do it but that's way
more expensive than needed. It's really easy to just handle constants
in apply_pipeline_layout.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
This rewires the logic for assigning uniform locations to work in terms
of "complex alignments". The basic idea is that, as we walk the list of
instructions, we keep track of the alignment and continuity requirements
of each slot and assert that the alignments all match up. We then use
those alignments in the compaction stage to ensure that everything gets
placed at a properly aligned register. The old mechanism handled
alignments by special-casing each of the bit sizes and placing 64-bit
values first followed by 32-bit values.
The old scheme had the advantage of never leaving a hole since all the
64-bit values could be tightly packed and so could the 32-bit values.
However, the new scheme has no type size special cases so it handles not
only 32 and 64-bit types but should gracefully extend to 16 and 8-bit
types as the need arises.
Tested-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
The testing for this extension is currently very poor. The CTS tests
only test accessing UBOs and SSBOs at dynamic offsets so none of our
constant-offset paths get triggered at all. Also, there's an assertion
in our handling of nir_intrinsic_load_uniform that offset % 4 == 0 which
is never triggered indicating that nothing every gets loaded from an
offset which is not a dword. Both push constants and the constant
offset pull paths are complex enough, we really don't want to ship
without tests. We'll turn the extension back on once we have decent
tests.
VCE processing IBs starts from session and task info at first level,
other commands processed subsequently. The task info for destroy is
embedded to destroy command, resulting that feedback command is not
properly procoessed. This is causing kernel spin VM fault messages on
Polaris and Vega10 card when running ends at encode application.
The fix is also verified on VCE physical mode card.
Signed-off-by: Leo Liu <leo.liu@amd.com>
Cc: mesa-stable@lists.freedesktop.org
Acked-by: Christian König <christian.koenig@amd.com>
Power8, Power8NV, and Power9 are supported on an equal footing
with X86.
Cc: "17.2" "17.3" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Ben Crocker <bcrocker@redhat.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
[Eric: changed formatting, reworded a bit (with Ben's ack)]
Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com>
Might be useful to know the VRAM/GTT usage, the number of VRAM
CPU page faults, etc. Nothing is currently using that new
interface, but it's a first step.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
dota2 binds a ton of index buffers but the type is always 16-bit.
Note that we have to invalidate the type when switching from
indexed draws to normal draws.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
dota2 always calls vkResetCommandBuffer() before
vkBeginCommandBuffer() which is quite useless.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
RADV_CMD_BUFFER_STATUS_INVALID is not used for now, but I think
it makes sense to declare it. Could be used later with better
command buffer error handling.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Copied from RadeonSI.
This fixes all CTS
dEQP-VK.renderpass.dedicated_allocation.formats.d32_sfloat_s8_uint.clear.*
And some other ones which use the same format.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
This resolves an apparent game bug described in 85564. The game
doesn't properly handle ARB_get_program_binary with 0 supported
formats.
V2 (Timothy Arceri):
- less driver code as more has been moved into the common helpers.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85564
Signed-off-by: Timothy Arceri <tarceri@itsqueeze.com>
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com> (v1)
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
The GL_ARB_get_program_binary extension spec says:
"If ProgramBinary fails to load a binary, no error is generated, but
any information about a previous link or load of that program object
is lost."
v2:
* Re-initialize shProg->data after clear. (Jordan)
(Required after 6a72eba755)
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
V2 (Timothy Arceri):
- add extra code comment
- stop passing around void *binary and just pass
program_binary_header *hdr instead.
- move to src/mesa/main rather than src/util
V3 (Timothy Arceri):
- Move more code out of the backend and into the common
helpers.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Mesa supports either 0 or 1 formats. If 1 format is supported, it is
GL_PROGRAM_BINARY_FORMAT_MESA as defined in the
GL_MESA_program_binary_formats extension spec.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
The ARB_get_program_binary extension requires that uniform values in a
program be restored to their initial value just after linking.
This patch saves off the initial values just after linking. When the
program is restored by glProgramBinary, we can use this to copy the
initial value of uniforms into UniformDataSlots.
V2 (Timothy Arceri):
- Store UniformDataDefaults only when serializing GLSL as this
is what we want for both disk cache and ARB_get_program_binary.
This saves us having to come back later and reset the Uniforms
on program binary restores.
Signed-off-by: Timothy Arceri <tarceri@itsqueeze.com>
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com> (v1)
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
This will allow us to use the program serialization to implement
ARB_get_program_binary.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Thus was merged into the OpenGL Registry in version
667c5a253781834b40a6ae9eb19d05af4542cfe1.
Ref: https://github.com/KhronosGroup/OpenGL-Registry/pull/127
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
This addresses a long-standing back-end compiler bug that could lead
to cross-channel data corruption in loops executed non-uniformly. In
some cases live variables extending through a loop divergence point
(e.g. a non-uniform break) into a convergence point (e.g. the end of
the loop) wouldn't be considered live along all physical control flow
paths the SIMD thread could possibly have taken in between due to some
channels remaining in the loop for additional iterations.
This patch fixes the problem by extending the CFG with physical edges
that don't exist in the idealized non-vectorized program, but
represent valid control flow paths the SIMD EU may take due to the
divergence of logical threads. This makes sense because the i965 IR
is explicitly SIMD, and it's not uncommon for instructions to have an
influence on neighboring channels (e.g. a force_writemask_all header
setup), so the behavior of the SIMD thread as a whole needs to be
considered.
No changes in shader-db.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This makes the dataflow propagation logic of the copy propagation pass
more intelligent in cases where the destination of a copy is known to
be undefined for some incoming CFG edges, building upon the
definedness information provided by the last patch. Helps a few
programs, and avoids a handful shader-db regressions from the next
patch.
shader-db results on ILK:
total instructions in shared programs: 6541547 -> 6541523 (-0.00%)
instructions in affected programs: 360 -> 336 (-6.67%)
helped: 8
HURT: 0
LOST: 0
GAINED: 10
shader-db results on BDW:
total instructions in shared programs: 8174323 -> 8173882 (-0.01%)
instructions in affected programs: 7730 -> 7289 (-5.71%)
helped: 5
HURT: 2
LOST: 0
GAINED: 4
shader-db results on SKL:
total instructions in shared programs: 8185669 -> 8184598 (-0.01%)
instructions in affected programs: 10364 -> 9293 (-10.33%)
helped: 5
HURT: 2
LOST: 0
GAINED: 2
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Currently the liveness analysis pass would extend a live interval up
to the top of the program when no unconditional and complete
definition of the variable is found that dominates all of its uses.
This can lead to a serious performance problem in shaders containing
many partial writes, like scalar arithmetic, FP64 and soon FP16
operations. The number of oversize live intervals in such workloads
can cause the compilation time of the shader to explode because of the
worse than quadratic behavior of the register allocator and scheduler
when running out of registers, and it can also cause the running time
of the shader to explode due to the amount of spilling it leads to,
which is orders of magnitude slower than GRF memory.
This patch fixes it by computing the intersection of our current live
intervals with the subset of the program that can possibly be reached
from any definition of the variable. Extending the storage allocation
of the variable beyond that is pretty useless because its value is
guaranteed to be undefined at a point that cannot be reached from any
definition.
According to Jason, this improves performance of the subgroup Vulkan
CTS tests significantly (e.g. the runtime of the dvec4 broadcast test
improves by nearly 50x).
No significant change in the running time of shader-db (with 5%
statistical significance).
shader-db results on IVB:
total cycles in shared programs: 61108780 -> 60932856 (-0.29%)
cycles in affected programs: 16335482 -> 16159558 (-1.08%)
helped: 5121
HURT: 4347
total spills in shared programs: 1309 -> 1288 (-1.60%)
spills in affected programs: 249 -> 228 (-8.43%)
helped: 3
HURT: 0
total fills in shared programs: 1652 -> 1597 (-3.33%)
fills in affected programs: 262 -> 207 (-20.99%)
helped: 4
HURT: 0
LOST: 2
GAINED: 209
shader-db results on BDW:
total cycles in shared programs: 67617262 -> 67361220 (-0.38%)
cycles in affected programs: 23397142 -> 23141100 (-1.09%)
helped: 8045
HURT: 6488
total spills in shared programs: 1456 -> 1252 (-14.01%)
spills in affected programs: 465 -> 261 (-43.87%)
helped: 3
HURT: 0
total fills in shared programs: 1720 -> 1465 (-14.83%)
fills in affected programs: 471 -> 216 (-54.14%)
helped: 4
HURT: 0
LOST: 2
GAINED: 162
shader-db results on SKL:
total cycles in shared programs: 65436248 -> 65245186 (-0.29%)
cycles in affected programs: 22560936 -> 22369874 (-0.85%)
helped: 8457
HURT: 6247
total spills in shared programs: 437 -> 437 (0.00%)
spills in affected programs: 0 -> 0
helped: 0
HURT: 0
total fills in shared programs: 870 -> 854 (-1.84%)
fills in affected programs: 16 -> 0
helped: 1
HURT: 0
LOST: 0
GAINED: 107
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This should allow the post-RA scheduler to do a slightly better job at
hiding latency in presence of instructions incurring bank conflicts.
The main purpuse of this patch is not to improve performance though,
but to get conflict cycles to show up in shader-db statistics in order
to make sure that regressions in the bank conflict mitigation pass
don't go unnoticed.
Acked-by: Matt Turner <mattst88@gmail.com>
Unnecessary GRF bank conflicts increase the issue time of ternary
instructions (the overwhelmingly most common of which is MAD) by
roughly 50%, leading to reduced ALU throughput. This pass attempts to
minimize the number of bank conflicts by rearranging the layout of the
GRF space post-register allocation. It's in general not possible to
eliminate all of them without introducing extra copies, which are
typically more expensive than the bank conflict itself.
In a shader-db run on SKL this helps roughly 46k shaders:
total conflicts in shared programs: 1008981 -> 600461 (-40.49%)
conflicts in affected programs: 816222 -> 407702 (-50.05%)
helped: 46234
HURT: 72
The running time of shader-db itself on SKL seems to be increased by
roughly 2.52%±1.13% with n=20 due to the additional work done by the
compiler back-end.
On earlier generations the pass is somewhat less effective in relative
terms because the hardware incurs a bank conflict anytime the last two
sources of the instruction are duplicate (e.g. while trying to square
a value using MAD), which is impossible to avoid without introducing
copies. E.g. for a shader-db run on SNB:
total conflicts in shared programs: 944636 -> 623185 (-34.03%)
conflicts in affected programs: 853258 -> 531807 (-37.67%)
helped: 31052
HURT: 19
And on BDW:
total conflicts in shared programs: 1418393 -> 987539 (-30.38%)
conflicts in affected programs: 1179787 -> 748933 (-36.52%)
helped: 47592
HURT: 70
On SKL GT4e this improves performance of GpuTest Volplosion by 3.64%
±0.33% with n=16.
NOTE: This patch intentionally disregards some i965 coding conventions
for the sake of reviewability. This is addressed by the next
squash patch which introduces an amount of (for the most part
boring) boilerplate that might distract reviewers from the
non-trivial algorithmic details of the pass.
The following patch is squashed in:
SQUASH: intel/fs/bank_conflicts: Roll back to the nineties.
Acked-by: Matt Turner <mattst88@gmail.com>
To demonstrate this bug run meson with the options:
-Ddri-drivers= -Dglx=gallium-xlib
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
src/gallium/winsys/pl111/drm/libpl111winsys.a(pl111_drm_winsys.c.o): In function `pl111_drm_screen_create':
pl111_drm_winsys.c:(.text+0x33): undefined reference to `vc4_drm_screen_create_renderonly'
Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
This patch is ported from RadeonSI and it has two effects.
It fixes a rendering issue which affects F1 2017 and Dawn
of War 3 (Vega only) because LLVM was ending up by generating
the new v_mad_mix_{hi,lo} instructions which appear to be
buggy in some way. Not sure if Mesa is generating something
wrong or if the issue is in LLVM only. Anyway, that explains why
the DOW3 issue can't be reproduced with GL on Vega.
It also improves performance because v_cvt_pkrtz_f16 is faster,
and because I guess the rounding mode behaviour is similar between
GL and VK, we can use it. About performance, it improves Talos
by +3/4% but I don't see any other impacts.
No CTS regressions on Polaris.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
To avoid any vulkan driver to include the GL mtypes.h. Renamed as
eventually this could be used by drivers not using nir.
v2: remove compiler/spirv/spirv.h from mtypes (Alejandro)
v3: added the definition at compiler/shader_info.h (Jason Ekstrand)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
At this point dc_job->cache_item_metadata.keys always equals
NULL, so call to free() is useless
Fixes: b86ecea344 ("util/disk_cache: write cache item metadata to disk")
Signed-off-by: Vadym Shovkoplias <vadym.shovkoplias@globallogic.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
