This will effectively enable the optimization in anv.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This allows us to drop legacy userclip plane handling in both the vec4
and FS backends, and simplifies a few interfaces.
v2 (Jason Ekstrand):
- Move brw_nir_lower_legacy_clipping to brw_nir_uniforms.cpp because
it's i965-specific.
- Handle adding the params in brw_nir_lower_legacy_clipping
- Call brw_nir_lower_legacy_clipping from brw_codegen_vs_prog
Co-authored-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
The rules for gl_SubgroupSize in Vulkan require that it be a constant
that can be queried through the API. However, all GL requires is that
it's a uniform. Instead of always claiming that the subgroup size in
the shader is 32 in GL like we have to do for Vulkan, claim 8 for
geometry stages, the maximum for fragment shaders, and the actual size
for compute.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Instead of lowering the subgroup size so early, wait until we have more
information. In particular, we're going to want different subgroup
sizes from different stages depending on the API. We also defer
lowering of subgroup masks because the ge/gt masks require the subgroup
size to generate a subgroup mask.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Make sure that a <group> tag within another <group> tag work just fine.
v2: rename 'halfbyte' to 'byte' to match the size (Lionel).
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Now we can decode a <group> tag inside another <group> tag, and properly
print its indices and content.
v2: Use push/pop stack to fields, groups and iters (Lionel).
v3: Add assert(iter->level < DECODE_MAX_ARRAY_DEPTH) (Lionel).
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
We currently use the group->next pointer to iterate through the <group>
tags. This change them to be a type of field, so we can descend into
them while iterating, and then go back to the original position. Will be
useful when we want to decode <group>'s inside <group>'s, and when there
are more <field>'s after a <group> tag.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
A gen_group (group in most of the code) can be of several types:
- instruction
- struct
- register
- group (?!?)
The <group> tag actually represents an array of elements. So at least
in our code, lets call it an array to avoid confusion with gen_group.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Refactor the code from gen_spec_load_from_path() into a separate
function, that can be used with a xml file that doesn't fit the genX.xml
filename format.
Will be used soon for implementing unit tests for gen_decoder.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
When using gen_spec_load_from path, only abort decoding if the read
length is 0. Previously, we were aborting if finding an EOF, even if
something was read from the file.
Also only kill the decoded file if no commands or structs were found,
and print a message in such case.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This doesn't fix any bug at the moment because the next statement is
'true' which happens to be APIMODE_D3D, but if that changes it could.
The fixes tags is as far I could go but the error predates it (2016 is
probably far enough).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 8db6f2e6eb ("anv/pipeline: Roll genX_pipeline_util.h into genX_pipeline.c")
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Normally, we haven't worried too much about stack sizes as Linux tends
to be fairly friendly towards large stacks. However, when running DXVK
apps under wine, we're suddenly subject to Windows' more stringent stack
limitations and can run out of space more easily. In particular, some
of the shaders in Elite Dangerous: Horizons have quite a few registers
and the arrays in split_virtual_grfs are large enough to blow a 1 MiB
stack leading to crashes during shader compilation.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108662
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Cc: mesa-stable@lists.freedesktop.org
anv_format is supposed to have a pointer back to the associated
VkFormat, we were missed this for depth/stencil formats.
This doesn't fix anything afaict, but will be needed for future
changes.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 465de47bad ("anv: associate vulkan formats with aspects")
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Spec says :
"timestampComputeAndGraphics specifies support for timestamps on all
graphics and compute queues. If this limit is set to VK_TRUE, all
queues that advertise the VK_QUEUE_GRAPHICS_BIT or
VK_QUEUE_COMPUTE_BIT in the VkQueueFamilyProperties::queueFlags
support VkQueueFamilyProperties::timestampValidBits of at least 36."
On gen7+ this should be true (we only have 32bits of timestamp on
gen6 and below).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 802f00219a ("anv/device: Update features and limits")
Reported-by: Timothy Strelchun <timothy.strelchun@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
In many cases, the compiler can just copy-prop the strided MOV whereas
the conversion is a bit trickier. This cuts 5% of the instructions off
of one particular Vulkan CTS test which does lots of load_ssbo.
Reviewed-by: Matt Turner <mattst88@gmail.com>
This fixes some validation errors generated by certain D->W conversions
but is likely not a full solution. Calculating an actual register
stride is a far more complex problem in general and should probably be
handled by the brw_fs_generator.
Reviewed-by: Matt Turner <mattst88@gmail.com>
When running on ICL the
dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13 needs more than 1M for
the shader, so bump it.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Now that the 64-bit lowering passes do a complete lowering in one go, we
don't need to loop anymore. We do, however, have to ensure that int64
lowering happens after double lowering because double lowering can
produce int64 ops.
Reviewed-by: Eric Anholt <eric@anholt.net>
In 6ce8592836 we started looking at the dynamic stencil state and
disabling stencil writes when the stencil mask is zero. Unfortunately,
we never updated the PMA fix code accordingly so 3DSTATE_WM_DEPTH_STENCIL
and the PMA fix were getting out-of-sync causing hangs.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109203
Fixes: 6ce8592836 "anv: Disable stencil writes when both write..."
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Specifically needed for nativewindow for some VK_EXT_external_memory_android_hardware_buffers
functions, where we call into some AHardwareBuffer functions.
The legacy Android ext did not have us call into any Android function
at all and hence it was not noticed.
Fixes: 755c633b8d "anv: Fix vulkan build in meson."
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Chad Versace <chadversary@chromium.org>
anv_render_pass_compile() turns an unused attachment into a NULL
depth_stencil_attachment pointer so check that pointer before
accessing it.
Found with updates to existing CTS tests.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 208be8eafa ("anv: Make subpass::depth_stencil_attachment a pointer")
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
When using softpin, the first allocation was not calculating the
padding and offset correctly for the case the first allocation needed
to grow. We were missing initialize the state.end right after
expanding the pool for the first time.
This is not a problem for non-softpin since there we don't use
leftover padding so the ends would re-arrange incrementally.
This fixes running dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13 in
SKL -- the test uses a shader larger than the initial size for the
instruction pool.
Fixes: dfc9ab2ccd "anv/allocator: Add padding information."
Reviewed-by: Rafael Antognolli <rafael.antognolli@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
For bindless SSBO access, we have to do 64-bit address calculations. On
ICL and above, we don't have 64-bit integer support so we have to lower
the address calculations to 32-bit arithmetic. If we don't run the
optimization loop before lowering, we won't fold any of the address
chain calculations before lowering 64-bit arithmetic and they aren't
really foldable afterwards. This cuts the size of the generated code in
the compute shader in dEQP-VK.ssbo.phys.layout.random.16bit.scalar.13 by
around 30%.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
c8665005: ("intel/compiler: Don't always require precise lowering of flrp")
forgot to remove some comments that didn't apply any more after the
change.
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrnd.net>
With separate stencil usage, we can't just grab the usage from the image
directly and have to consider the per-aspect usage instead.
Fixes: 1be38f9178 "anv:Use VK_EXT_separate_stencil_usage to avoid..."
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
A couple patches later in this series use the flag to avoid a few
thousand shader-db regresions on all vec4 platforms.
I'm not particularly enamored with the name of this flag. However, I
suspect the Intel vec4 backend is the only backend that will benefit
from it. Specifically, the cases where this helps are all cases where
we want to prevent nir_opt_algebraic from rearranging instructions to
create 3-source instructions, such as ffma and flrp, with additional
immediate value or uniform sources.
The earlier commit "intel/vec4: Try to emit a single load for multiple
3-src instruction operands" solves most of the problems caused by
additional immediate values, but the restrictions on register strides
that cause problems for uniforms and shader inputs persist.
Reviewed-by: Matt Turner <mattst88@gmail.com>
If a 3-source instruction uses immediate values 1.0 and -1.0, just load
1.0 into a register. Use the negation source modifier to get -1.0.
This has trivial impact now, but it prevents a few thousand regressions
on vec4 platforms with "nir/algebraic: Recognize open-coded flrp(-1, 1,
a) and flrp(1, -1, a)"
All Gen6 and Gen7 platforms had similar results. (Haswell shown)
total instructions in shared programs: 13487412 -> 13487406 (<.01%)
instructions in affected programs: 541 -> 535 (-1.11%)
helped: 6
HURT: 0
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.36% max: 2.08% x̄: 1.65% x̃: 1.80%
95% mean confidence interval for instructions value: -1.00 -1.00
95% mean confidence interval for instructions %-change: -2.33% -0.97%
Instructions are helped.
total cycles in shared programs: 376402564 -> 376402500 (<.01%)
cycles in affected programs: 10348 -> 10284 (-0.62%)
helped: 10
HURT: 1
helped stats (abs) min: 2 max: 26 x̄: 7.00 x̃: 2
helped stats (rel) min: 0.13% max: 2.05% x̄: 0.89% x̃: 0.79%
HURT stats (abs) min: 6 max: 6 x̄: 6.00 x̃: 6
HURT stats (rel) min: 0.29% max: 0.29% x̄: 0.29% x̃: 0.29%
95% mean confidence interval for cycles value: -11.72 0.08
95% mean confidence interval for cycles %-change: -1.20% -0.36%
Inconclusive result (value mean confidence interval includes 0).
No shader-db changes on any other Intel platform.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Gen11 SLM is not on L3 anymore, so now the hardware has two separate
fences. Add a way to control which fence types to use.
At this time, we don't have enough information in NIR to control the
visibility of the memory being fenced, so for now be conservative and
assume that fences will need a stall. With more information later
we'll be able to reduce those.
Fixes Vulkan CTS tests in ICL:
dEQP-VK.memory_model.message_passing.core11.u32.coherent.fence_fence.atomicwrite.device.payload_nonlocal.workgroup.guard_local.buffer.comp
dEQP-VK.memory_model.message_passing.core11.u32.coherent.fence_fence.atomicwrite.device.payload_local.buffer.guard_nonlocal.workgroup.comp
dEQP-VK.memory_model.message_passing.core11.u32.coherent.fence_fence.atomicwrite.device.payload_local.image.guard_nonlocal.workgroup.comp
dEQP-VK.memory_model.message_passing.core11.u32.coherent.fence_fence.atomicwrite.workgroup.payload_local.buffer.guard_nonlocal.workgroup.comp
dEQP-VK.memory_model.message_passing.core11.u32.coherent.fence_fence.atomicwrite.workgroup.payload_local.image.guard_nonlocal.workgroup.comp
The whole set of supported tests in dEQP-VK.memory_model.* group
should be passing in ICL now.
v2: Pass BTI around instead of having an enum. (Jason)
Emit two SHADER_OPCODE_MEMORY_FENCE instead of one that gets
transformed into two. (Jason)
List tests fixed. (Lionel)
v3: For clarity, split the decision of which fences to emit from the
emission code. (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Right now, all keys have two things in common: a program string ID and a
sampler_prog_key_data. I'd like to add another thing or two and need a
place to put it. This commit adds a new brw_base_prog_key struct which
contains those two common bits.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
I'm not 100% sure how this ever worked because gem_create usually shoots
you if the BO size isn't page-aligned.
Reviewed-by: Chad Versace <chadversary@chromium.org>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This is the MOCS setting used for the A64 stateless messages which we
sometimes use for SSBO operations.
Fixes: 48ed2a7bb0 "anv: Implement VK_EXT_buffer_device_address"
Fixes: 79fb0d27f3 "anv: Implement SSBOs bindings with GPU addr..."
Reviewed-by: Chad Versace <chadversary@chromium.org>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Add libmesa_nir to a common LOCAL_STATIC_LIBRARIES defined by
ANV_STATIC_LIBRARIES so that its include path can be imported
automatically. Then ANV_INCLUDES is unnecessary and could be
eliminated.
Signed-off-by: Chih-Wei Huang <cwhuang@linux.org.tw>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Emil Velikov <emil.velikov@collabora.com>
The dummy library libmesa_anv_entrypoints is totally unnecessary.
The four VULKAN_GENERATED_FILES could be generated and built in
libmesa_vulkan_common directly. The libraries using the generated
headers should get it via the exported include path.
Signed-off-by: Chih-Wei Huang <cwhuang@linux.org.tw>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Emil Velikov <emil.velikov@collabora.com>
The libmesa_anv_entrypoints and libmesa_genxml are dummy libraries.
There is no reason to put them into LOCAL_WHOLE_STATIC_LIBRARIES.
Move libmesa_vulkan_util to the vulkan HAL which really needs it.
Signed-off-by: Chih-Wei Huang <cwhuang@linux.org.tw>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Emil Velikov <emil.velikov@collabora.com>
Daniel Schürmann