I did not implement:
CNL's restriction on 64-bit int + align16, because I don't think
we'll ever use this combination regardless of hardware generation.
The restriction on immediate DF -> F conversions, because there's no
reason to ever generate that, and I don't even know how DF -> F
conversions are supposed to work in Align16 since (1) the dst stride
must be 1, but (2) the dst stride would have to be 2 for src and dst
strides to be aligned.
Some restrictions require something like strides to match between src
and dest. For multi-source instructions, I'd rather encapsulate the
logic for not inserting already present errors in ERROR_IF than
open-coding it multiple places.
The type suffixes were wrong, and the 16 was missing the 0 prefix.
Fixes: 92f787ff86 ("i965: Add support for disassembling 64-bit integer immediates")
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
... without the float -> double conversion. Low power parts have
additional restrictions when it comes to operating on 64-bit types, and
the instruction used to do the conversion violates one of them:
specifically, the restriction that "Source and Destination horizontal
stride must be aligned to the same qword".
Previously we generated a float and then converted, but we can avoid the
conversion by using the same extract-the-sign-bit + or-in-1.0 algorithm
by directly operating on the high four bytes of each double-precision
component in the result.
In SIMD8 and SIMD16 this cuts one instruction from the implementation,
and more importantly that instruction is the one which violated the
regioning restriction.
Along the way I removed some comments that I did not think helped, and
some code about double comparisons which does not seem to be necessary
today.
This prevents validation failures caught by the new EU validation code
added in later patches.
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
64-bit operations on Atom parts have additional restrictions over their
big-core counterparts (validated by later patches).
Specifically, the restriction that "Source and Destination horizontal
stride must be aligned to the same qword" is violated by most shift
operations since NIR uses a 32-bit value as the shift count argument,
and this causes instructions like
shl(8) g19<1>Q g5<4,4,1>Q g23<4,4,1>UD
where src1 has a 32-bit stride, but the dest and src0 have a 64-bit
stride.
This caused ~4 pixels in the ARB_shader_ballot piglit test
fs-readInvocation-uint.shader_test to be incorrect. Unfortunately no
ARB_gpu_shader_int64 test hit this case because they operate on
uniforms, and their scalar regions are an exception to the restriction.
We work around this by effectively unpacking the shift count, so that we
can read it with a 64-bit stride in the shift instruction. Unfortunately
the unpack (a MOV with a dst stride of 2) is a partial write, and cannot
be copy-propagated or CSE'd.
Bugzilla: https://bugs.freedesktop.org/101984
A typo caused us to copy src0's reg file to src1 rather than reading
src1's as intended. This caused us to fail to compact instructions like
mov(8) g4<1>D 0D { align1 1Q };
because src1 was set to immediate rather than architecture file. Fixing
this reenables compaction (after the precompact() pass changes the data
types):
mov(8) g4<1>UD 0x00000000UD { align1 1Q compacted };
Fixes: 1cb0a7941b ("i965: Switch to using the logical register types")
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This is just legacy cruft. We don't push these values; we pass them in
as vertex attributes.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
In truth gtest is an external dependency that upstream expects you to
"vendor" into your own tree. As such, it makes sense to treat it more
like a dependency than an internal library, and collect it's
requirements together in a dependency object.
v2: - include with -isystem instead of setting compiler args (Eric)
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
These shouldn't matter for non-cubes, and we always enable them all
for cubes, so we may as well set them all the time.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
I decided to use the one-boolean-per-cube-face approach because it's
clearer which bits correspond to which cube face.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
We can start reading the URB at the first offset that contains varyings
that are actually read in the URB. We still need to make sure that we
read at least one varying to honor hardware requirements.
This helps alleviate a problem introduced with 99df02ca26 for
separate shader objects: without separate shader objects we assign
locations sequentially, however, since that commit we have changed the
method for SSO so that the VUE slot assigned depends on the number of
builtin slots plus the location assigned to the varying. This fixed
layout is intended to help SSO programs by avoiding on-the-fly recompiles
when swapping out shaders, however, it also means that if a varying uses
a large location number close to the maximum allowed by the SF/FS units
(31), then the offset introduced by the number of builtin slots can push
the location outside the range and trigger an assertion.
This problem is affecting at least the following CTS tests for
enhanced layouts:
KHR-GL45.enhanced_layouts.varying_array_components
KHR-GL45.enhanced_layouts.varying_array_locations
KHR-GL45.enhanced_layouts.varying_components
KHR-GL45.enhanced_layouts.varying_locations
which use SSO and the the location layout qualifier to select such
location numbers explicitly.
This change helps these tests because for SSO we always have to include
things such as VARYING_SLOT_CLIP_DIST{0,1} even if the fragment shader is
very unlikely to read them, so by doing this we free builtin slots from
the fixed VUE layout and we avoid the tests to crash in this scenario.
Of course, this is not a proper fix, we'd still run into problems if someone
tries to use an explicit max location and read gl_ViewportIndex, gl_LayerID or
gl_CullDistancein in the FS, but that would be a much less common bug and we
can probably wait to see if anyone actually runs into that situation in a real
world scenario before making the decision that more aggresive changes are
required to support this without reverting 99df02ca26.
v2:
- Add a debug message when we skip clip distances (Ilia)
- we also need to account for this when we compute the urb setup
for the fragment shader stage, so add a compiler util to compute
the first slot that we need to read from the URB instead of
replicating the logic in both places.
v3:
- Make the util more generic so it can account for all unused slots
at the beginning of the URB, that will make it more useful (Ken).
- Drop the debug message, it was not what Ilia was asking for.
Suggested-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Allows the instructions to be compacted. The documentation claims that
some of these only accept UD types, even though the type doesn't change
the operation performed. Just normalize the types to ensure we get
instruction compaction.
The only functional changes are for FBL and CBIT (always use UD types)
and FBH (always use the same types).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Triggering the push model when 64-bit inputs are involved is not easy due to
the constrains on the maximum number of registers that we allow for this mode,
however, for GS with 'points' primitive type and just a couple of double
varyings we can trigger this and it just doesn't work because the
implementation is not 64-bit aware at all. For now, let's make sure that we
don't attempt this model whith 64-bit inputs and we always fall back to pull
model for them.
Also, don't enable the VUE handles in the thread payload on the fly when we
find an input for which we need the pull model, this is not safe: if we need
to resort to the pull model we need to account for that when we setup the
thread payload so we compute the first non-payload register properly. If we
didn't do that correctly and we enable it on-the-fly here then we will end up
VUE handles on the first non-payload register which will probably lead to
GPU hangs. Instead, always enable the VUE handles for the pull model so we
can safely use them when needed. The GS is going to resort to pull model
almost in every situation anyway, so this shouldn't make a significant
difference and it makes things easier and safer.
v2: Always enable the VUE handles for pull model, this is easier and safer
and the GS is going to fallback to pull model almost always anyway (Ken)
v3: Only clamp the URB read length if we are over the maximum reserved for
push inputs as we were doing in the original code (Ken).
v4: No need to clamp the urb read length if invocations > 1
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This allows building and installing the Intel "anv" Vulkan driver using
meson and ninja, the driver has been tested against the CTS and has
seems to pass the same series of tests (they both segfault when the CTS
tries to run wayland wsi tests).
There are still a mess of TODO, XXX, and FIXME comments in here. Those
are mostly for meson bugs I'm trying to fix, or for additional things to
implement for other drivers/features.
I have configured all intermediate libraries and optional tools to not
build by default, meaning they will only be built if they're pulled in
as a dependency of a target that will actually be installed) this allows
us to avoid massive if chains, while ensuring that only the bits that
need to be built are.
v2: - enable anv, x11, and wayland by default
- add configure option to disable valgrind
v3: - fix typo in meson_options (Nicholas)
v4: - Remove dead code (Eric)
- Remove change to generator that was from v0 (Eric)
- replace if chain with loop (Eric)
- Fix typos (Eric)
- define HAVE_DLOPEN for both libdl and builtin dl cases (Eric)
v5: - rebase on util string buffer implementation
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net> (v4)
Meson doesn't allow setting environment variables for custom targets, so
we either need to not pass this as an environment variable or use a
shell script to wrap the invocation. The chosen solution has the
advantage of working for both autotools and meson.
v2: - put rules back in top scope (Ken)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
In the vec4 backend, SHADER_OPCODE_UNTYPED_ATOMIC's src[1] is the
surface index. We want to copy propagate so we can use an immediate
message descriptor, rather than an indirect send.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Atomic operation sources are scalar values, but we were failing to
select the .x component of the second operand. For example,
atomicCounterCompSwapARB(counter, 5u, 10u)
would generate
mov(8) vgrf4.x:D, 5D
mov(8) vgrf5.x:D, 10D
mov(8) vgrf9.x:UD, vgrf4.xyzw:D
mov(8) vgrf9.y:UD, vgrf5.xyzw:D
which wrongly selects the .y component of vgrf5, so the actual 10u value
would get dead code eliminated. The swizzle works for the other source,
but both of them ought to be .xxxx.
Fixes the compare and swap CTS tests in:
KHR-GL45.shader_atomic_counter_ops_tests.ShaderAtomicCounterOpsExchangeTestCase
Cc: "17.2 17.1 17.0 13.0" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Embarassingly, someone enabled the ARB_shader_atomic_counter_ops
extension for Gen7+ but never added the intrinsics to the switch
statement in the vec4 backend, so they just hit an unreachable()
call and died.
Fixes: 40dd45d0c6 (i965: Enable ARB_shader_atomic_counter_ops)
Cc: "17.2 17.1 17.0 13.0" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
I was implementing the same enum support in broadcom's gen_pack_header.py,
and did this same simplification there.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
In Vulkan, for 'z' (depth) component, the scale and translate values
for the viewport transformation are:
pz = maxDepth - minDepth
oz = minDepth
zf = pz × zd + oz
Being zd, the third component in vertex's normalized device coordinates.
Fixes: dEQP-VK.draw.inverted_depth_ranges.*
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Cc: mesa-stable@lists.freedesktop.org
This can occur if the shader is capturing some of the values from the
VUE header for transform feedback, but the shader hasn't written all of
them.
Reviewed-by: Juan A. Suarez Romero <jasuarez@igalia.com>
The Broadwell method of handling uncompressed views of compressed
textures was to make the texture linear and have a tiled shadow copy.
This isn't needed on Sky Lake because the HALIGN and VALIGN parameters
are specified in surface elements and required to be a multiple of 4.
This means that we can just use the X/Y Offset fields and we can avoid
the shadow copy song and dance. This also makes ASTC work because ASTC
can't be linear and so the shadow copy method doesn't work there.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
In order to get support everywhere, this gets a bit complicated. On Sky
Lake and later, everything is fine because HALIGN/VALIGN are specified
in surface elements and are required to be at least 4 so any offsetting
we may need to do falls neatly within the heavy restrictions placed on
the X/Y Offset parameter of RENDER_SURFACE_STATE. On Broadwell and
earlier, HALIGN/VALIGN are specified in pixels and are hard-coded to
align to exactly the block size of the compressed texture. This means
that, when reinterpreted as a non-compressed texture, the tile offsets
may be anything and we can't rely on X/Y Offset.
In order to work around this issue, we fall back to linear where we can
trivially offset to whatever element we so choose. However, since
linear texturing performance is terrible, we create a tiled shadow copy
of the image to use for texturing. Whenever the user does a layout
transition from anything to SHADER_READ_ONLY_OPTIMAL, we use blorp to
copy the contents of the texture from the linear copy to the tiled
shadow copy. This assumes that the client will use the image far more
for texturing than as a storage image or render target.
Even though we don't need the shadow copy on Sky Lake, we implement it
this way first to make testing easier. Due to the hardware restriction
that ASTC must not be linear, ASTC does not work yet.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This struct represents a full surface state including the addresses of
the referenced main and auxiliary surfaces (if any). This makes
relocation setup substantially simpler and allows us to move 100% of the
surface state setup logic into anv_image where it belongs. Before, we
were manually fishing data out of surface states when emitting
relocations so we knew how to offset aux address. It's best to keep all
of the surface state emit logic together. This also gets us closer, at
least cosmetically, to a world of no relocations where addresses are
placed in surface states up-front.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This gives us a single centralized place where we take an image view and
use it to fill out a surface state.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Jason Ekstrand