A vertex-path program with LDS, GDS or memory writes refused any s_barrier. NGG passthrough
programs start with the s_barrier that orders GS_ALLOC_REQ, before any memory access, and some
write buffers later. A barrier that runs before every load, store, atomic, image or LDS access of
the program has nothing to order, so it emits no instruction, like the barrier of a program
without memory access. Barriers after a memory access still throw.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Resolves the shader debt file bullet by bullet: main's EmitSetAttribute entry, and this branch's PERSP_SAMPLE/LINEAR_SAMPLE and pixel-position entries.
The RDNA2 ISA (8.1.7) ignores the two low address bits of dword or wider
non-formatted accesses in alignment mode DWORD. MXGP 26 (PPSA30839) sets
bits with buffer_atomic_or at odd byte offsets, which recorded an
unaligned BDA fault and dropped the atomic. A 32-bit buffer atomic now
acts on the dword that holds its address; 64-bit atomics off an 8-byte
boundary still fault.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
10_11_11_UNORM (format 30) images are read and written through an
R32_UINT view, so the hardware sampler cannot filter them. image_sample
of a 2D or 2D array view now fetches the one or four texels at LOD 0,
decodes them and blends them in the shader, with bilinear weights on
the 8-bit grid, for point or bilinear filtering with wrap or
clamp-to-edge addressing.
A shader gets the converted sample mode only after a converted
descriptor reaches one of its sampled images: that capture throws
ConvertedSampleRequired and the driver prepares the program again
with GuestContext::convertedSamples.
The R10_G11_B11 float format is sampled natively since #7271 and keeps
its refusal on the converted path.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A vertex or tessellation evaluation program whose PA_CL_VS_OUT_CNTL
enables eight guest clip distances (CLIP_DIST_ENA 0xff, both distance
vectors) failed to compile with "guest clip distances plus the all-zero
clip position distance exceed 8 clip distances", at registration and at
the draw. A program whose guest clip and cull distances together take
eight (four of each, say) compiled, but its draw failed in the pipeline
validator with "more than eight clip and cull distances in one shader".
The all-zero position distance is now added only while the guest clip and
cull distances leave one of the eight slots free. Otherwise the program
keeps its guest distances, and a primitive with a vertex at clip position
(0,0,0,0) is not removed; the EmitSetAttribute debt entry says so.
agc_position_invariance compiles one program with seven and eight clip
distances, and with three and four clip distances beside four cull
distances: the clip distance arrays are 8, 8, 4 and 4, so the all-zero
position distance is in the first and third only.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Compute LDS beyond the device's shared-memory limit lives in a storage
buffer, and its 64-bit atomics took a lock in that buffer. AMD's Windows
driver compiles the lock's acquire and release to s_barrier, which inside
the divergent lock loop pairs with the program's own barriers, so the
device-memory case of agc_driver_lds_lock_budget read half its sum.
With shaderBufferInt64Atomics the device-memory LDS now uses a 64-bit view
of the buffer instead: 64-bit atomics are compare-exchange loops on it and
32-bit atomics stay native, so no lock is taken. Without the feature the
lock is kept. The per-workgroup slice is rounded up to an even dword count
so that the 64-bit view stays aligned.
Since #4048 enabled robustBufferAccess2, AMD's Windows driver draws
nothing in agc_driver_layered_volume_draw and agc_driver_layered_volume_vertex,
and the selected loads make the high half of agc_driver_subvector_loop's
wave64 skip its loop. Until the cause is found, that driver does not get
the feature, and its loads keep their selections.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Main's #7325 adds verifyChainedTableOffset to the image table tests
where this branch adds its selected-record tests. Keep both, with
main's call after this branch's.
Main's EmitSetAttribute debt entry now says the generated rect-list and
barycentric tessellation stages forward the vertex shader's clip array.
It sits next to this branch's PERSP_SAMPLE and pixel position entries,
which main left as they were; keep main's entry and this branch's two.
Main's #7101 reads a key buffer's words in one call where this branch
names the reason a narrowing is skipped. Keep main's bulk read and
count both of its unmapped-page skips as unmapped.
The binding plan refused every buffer atomic through a V# whose base is
not a multiple of 4. On RDNA2 such an atomic works when base + offset is
on a DWORD boundary and stops the dispatch otherwise, which the runtime
buffer atomic already does, so the refusal is dropped.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A v_mul_f32 in a FLOAT_MODE that flushes f32 denormals emits the
nearest-even tininess check, the NaN result selection and then a
result flush. The check already turns every product that rounds below
2^-126 into a zero of its sign, so the flush can't change what it
keeps. DenormalFlushEliminator now proves that: a select arm guarded
by the tiny-product check of the same denormal-free factors, an arm a
select takes only when its own bits are NaN, and an and with a mask
that has no significand bits are denormal-free.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
main d7262bd6d6 adds the single-sample scan entry after the custom MSAA
sample locations entry in the AGC debt file. Keep it after this
branch's version of that entry.
v_pk_add/mul/fma/fmac/min/max_f16 threw in f16 denormal modes 0-2. On
RDNA2 (gfx1031) bit 6 clear flushes each denormal input lane to a zero
of its sign, and bit 7 clear flushes an add result below 2^-14 and a
mul or fma result whose exact value rounded to 11 significant bits
with an unbounded exponent is below 2^-14, as the scalar forms do;
min/max flush only their inputs. Packed f16 still throws for f16
rounding other than nearest even.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Adrià