The Vulkan API provides a mechanism for applications to cache their own
shaders and manage on-disk pipeline caching themselves. Generally, this
is what I would recommend to application developers and I've resisted
implementing driver-side transparent caching in the Vulkan driver for a
long time. However, not all applications do this and, for some
use-cases, it's just not practical.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Before, we were only hashing the shader if we had a shader cache to
cache things in. This means that if we ever get it wrong, we could end
up trying to cache a shader with an undefined hash. Since not having a
shader cache is an extremely uncommon case, let's optimize for code
clarity and obvious correctness over avoiding a hash operation.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
If a client is dumb enough to not specify a pipeline cache, give it a
default. We have to create one anyway for blorp so we may as well let
the client cache shaders in it.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Previously, we just hashed the entire descriptor set layout verbatim.
This meant that a bunch of extra stuff such as pointers and reference
counts made its way into the cache. It also meant that we weren't
properly hashing in the Y'CbCr conversion information information from
bound immutable samplers.
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
According to RenderDoc, this shaves 99.6% of the run time off of the
ambient occlusion pass in Skyrim Special Edition when running under DXVK
and shaves 92% off the runtime for a reasonably representative frame.
When running the actual game, Skyrim goes from being a slide-show to a
very stable and playable framerate on my SKL GT4e machine.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Every time we emit a new state base address we will need to re-emit our
binding tables, since they might have been emitted with a different base
state adress.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
CC: <mesa-stable@lists.freedesktop.org>
If we have to re-emit push constant data, we need to re-emit all
of it.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
CC: <mesa-stable@lists.freedesktop.org>
Running VK-CTS in batch execution mode was raising the
VK_ERROR_INITIALIZATION_FAILED error in multiple tests. But when the
same failing tests were run isolated they always passed.
createDevice and destroyDevice were called before and after every
tests. Because the binding_table_pool was never closed, we reached the
maximum number of open file descriptors (ulimit -n) and when that
happened every call to createDevice implied a
VK_ERROR_INITIALIZATION_FAILED error.
Fixes: c7db0ed4e9
("anv: Use a separate pool for binding tables when soft pinning")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
There's no reason for us to emit it a pile of times and then have a
whole pass to clean it up. Just emit it once like we really want.
Reviewed-by: Matt Turner <mattst88@gmail.com>
This generalizes the unlit centroid workaround so it's less code and now
supports SIMD32.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
v2 (Jason Ekstrand):
- Disallow gl_SampleId in SIMD32 on gen7
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
And handle 32-wide payload register reads in fetch_payload_reg().
v2 (Jason Ekstrand);
- Fix some whitespace and brace placement
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
While we're here, we change to using horiz_offset() instead of abusing
half().
v2 (Jason Ekstrand):
- Use horiz_offset() instead of half()
Reviewed-by: Matt Turner <mattst88@gmail.com>
The original code manually handled splitting the MOVs to 8-wide to
handle various regioning restrictions. Now that we have a SIMD width
splitting pass that handles these things, we can just emit everything at
the full width and let the SIMD splitting pass handle it. We also now
have a useful "subscript" helper which is designed exactly for the case
where you want to take a W type and read it as a vector of Bs so we may
as well use that too.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
On g4x through Sandy Bridge, src1 (the coordinates) of the PLN
instruction is required to be an even register number. When it's odd
(which can happen with SIMD32), we have to emit a LINE+MAC combination
instead. Unfortunately, we can't just fall through to the gen4 case
because the input registers are still set up for PLN which lays out the
four src1 registers differently in SIMD16 than LINE.
v2 (Jason Ekstrand):
- Take advantage of both accumulators and emit LINE LINE MAC MAC
(Based on a patch from Francisco Jerez)
- Unify the gen4 and gen4x-6 cases using a loop
v3 (Jason Ekstrand):
- Don't unify gen4 with gen4x-6 as this turns out to be more fragile
than first thought without reworking the gen4 barycentric coordinate
layout.
Reviewed-by: Matt Turner <mattst88@gmail.com>
When we don't have PLN (gen4 and gen11+), we implement LINTERP as either
LINE+MAC or a pair of MADs. In both cases, the accumulator is written
by the first of the two instructions and read by the second. Even
though the accumulator value isn't actually ever used from a logical
instruction perspective, it is trashed so we need to make the scheduler
aware. Otherwise, the scheduler could end up re-ordering instructions
and putting a LINTERP between another an instruction which writes the
accumulator and another which tries to use that result.
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Matt Turner <mattst88@gmail.com>
This reworks INTERPOLATE_AT_PER_SLOT_OFFSET to work more like an ALU
operation and less like a send. This is less code over-all and, as a
side-effect, it now properly handles execution groups and lowering so
SIMD32 support just falls out.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
We want consistent behavior in the meaning of the flag_subreg field
between SNB and IVB+.
v2 (Jason Ekstrand):
- Add some extra commentary
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
This prevents a crash in some arb_enhanced_layouts tests that would be
caused by the next commit.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Current discard handling requires dedicating the second flag register to
discard. However, control-flow in SIMD32 requires both flag registers
so it's incompatible with the current discard handling. Just don't
support SIMD32+discard for now.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
The hardware's control flow logic is 16-wide so we're out of luck
here. We could, in theory, support SIMD32 if we know the control-flow
is uniform but we don't have that information at this point.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Commit 0d905597f fixed an issue with the placement of the zip and unzip
instructions. However, as a side-effect, it reversed the order in which
we were emitting the split instructions so that they went from high
group to low instead of low to high. This is fine for most things like
texture instructions and the like but certain render target writes
really want to be emitted low to high. This commit just switches the
order back around to be low to high.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Fixes: 0d905597f "intel/fs: Be more explicit about our placement of [un]zip"
The pixel shader dispatch table is kind-of a confusing mess. This adds
some helpers for dealing with it and for easily extracting the correct
data from wm_prog_data.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Doing instruction header setup in the generator is awful for a number
of reasons. For one, we can't schedule the header setup at all. For
another, it means lots of implied writes which the instruction scheduler
and other passes can't properly read about. The second isn't a huge
problem for FB writes since they always happen at the end. We made a
similar change to sampler handling in ff4726077d.
Reviewed-by: Matt Turner <mattst88@gmail.com>
