6733 commits

Author	SHA1	Message	Date
Lionel Landwerlin	94f8d0072d	nir: add pass to propagate image format to intrinsics ... Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Alyssa Anne Rosenzweig <alyssa.rosenzweig@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36773>	2025-10-07 08:54:26 +00:00
Lionel Landwerlin	0922a0dd50	nir/lower_tex: remove unused options ... Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Alyssa Anne Rosenzweig <alyssa.rosenzweig@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37692>	2025-10-03 20:19:03 +00:00
Lionel Landwerlin	97dde5bc10	nir/lower_tex: add an callback to lower txd ops ... Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Alyssa Anne Rosenzweig <alyssa.rosenzweig@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37692>	2025-10-03 20:19:02 +00:00
Daniel Schürmann	0e3bc3d8c0	nir/opt_offsets: call allow_offset_wrap() for try_fold_shared2() ... This prevents applying wrapping offsets on GFX6. Fixes: e1a692f74b ('nir/opt_offsets: allow for unsigned wraps when folding load/store_shared2_amd offsets') Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37667>	2025-10-03 07:54:12 +00:00
Kenneth Graunke	25cb6dfbf7	nir: Add load_simd_width_intel to divergence analysis ... For some reason we missed adding this. This prevents some asserts from triggering when I call divergence analysis at certain points in an upcoming patch. Cc: mesa-stable Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36750>	2025-09-30 19:44:02 +00:00
sjfricke	05ea82a766	nir: Fix gnu-empty-initializer warning ... Found with clang 14 Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37615>	2025-09-30 19:09:31 +00:00
Ella Stanforth	082e6369f9	nir: add v3d specific intrinsic normalised to float conversion ... Reviewed-by: Iago Toral Quiroga <itoral@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35820>	2025-09-30 12:48:42 +00:00
Simon Perretta	a1acd6f8d1	pvr, pco: add primitive support for VK_KHR_robustness2.nullDescriptor ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>	2025-09-30 12:15:54 +00:00
Simon Perretta	2a7ebf2ae0	nir/lower_alpha: extend to support dynamic a2c ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>	2025-09-30 12:15:53 +00:00
Simon Perretta	6dc5e1e109	pco: fully support Vulkan 1.2 image atomics ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>	2025-09-30 12:15:48 +00:00
Aitor Camacho	06dbd4c33c	nir: Set cursor in lower_sampler_lod_bias ... Reviewed-by: Alyssa Anne Rosenzweig <alyssa.rosenzweig@intel.com> Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37494>	2025-09-29 23:24:52 +00:00
Simon Perretta	b111b8a844	pvr, pco: implement prerequisites for sampleRateShading ... - Implement load_interpolated_input and friends. - Optimize load_barycentric_* cases that can be simplified. - Initial support for non-standard sample locations. Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Frank Binns <frank.binns@imgtec.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37540>	2025-09-27 23:45:54 +01:00
Simon Perretta	83aecc8f3f	mesa/st, nir: commonize unlower_io_to_vars pass ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37540>	2025-09-27 23:45:54 +01:00
Aleksi Sapon	8949473023	nir: Fix nir.h MSVC compilation for C++ source files ... This kind of C initializer is not accepted by MSVC in C++ mode. Fixed: 75292ae7 ("nir: Fix gnu-empty-initializer warning ") Reviewed-by: Karol Herbst <kherbst@redhat.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37604>	2025-09-26 18:25:22 +00:00
Georg Lehmann	46a4569c22	nir/opt_undef: prefer 0 over NaN for pack_half_2x16_rtz_split ... Using NaN doesn't usually allow any extra optimizations, and 0 is an inline constant on AMD hw where this opcode is used with undef for fragment shader exports. Foz-DB GFX1201: Totals from 889 (1.11% of 80287) affected shaders: Instrs: 1676365 -> 1676348 (-0.00%) CodeSize: 8827040 -> 8821760 (-0.06%) Latency: 13346728 -> 13346699 (-0.00%) InvThroughput: 1799283 -> 1799262 (-0.00%) Copies: 108125 -> 108102 (-0.02%) VALU: 974875 -> 974852 (-0.00%) Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37552>	2025-09-26 15:11:26 +00:00
Aleksi Sapon	75292ae7e4	nir: Fix gnu-empty-initializer warning ... This also causes a build error on older MSVC. Fixes: 75381670 ("nir,rusticl: NIR_PASS/nir_pass! validation fixes and improvements") Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37569>	2025-09-25 18:14:22 +00:00
Daniel Schürmann	e1a692f74b	nir/opt_offsets: allow for unsigned wraps when folding load/store_shared2_amd offsets ... Totals from 131 (0.16% of 79839) affected shaders: (Navi48) Instrs: 217026 -> 216541 (-0.22%); split: -0.24%, +0.01% CodeSize: 1150136 -> 1146772 (-0.29%); split: -0.31%, +0.02% Latency: 4225732 -> 4225549 (-0.00%); split: -0.01%, +0.00% InvThroughput: 840231 -> 839823 (-0.05%); split: -0.05%, +0.00% VClause: 3815 -> 3816 (+0.03%) Copies: 15414 -> 15358 (-0.36%); split: -0.38%, +0.02% PreSGPRs: 6322 -> 6323 (+0.02%) PreVGPRs: 6064 -> 6062 (-0.03%) VALU: 117317 -> 116873 (-0.38%); split: -0.40%, +0.02% SALU: 25384 -> 25331 (-0.21%); split: -0.22%, +0.02% Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37453>	2025-09-24 14:28:24 +00:00
Rhys Perry	7538167096	nir: add NIR_DEBUG=progress_validation ... Fails if a shader was changed but the pass didn't report progress. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Marek Olšák <maraeo@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35069>	2025-09-24 08:20:28 +00:00
Rhys Perry	706ba80057	nir: fix NIR_DEBUG=extended_validation ... This broke after divergence became metadata because the divergence analysis pass does not support all instructions. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Marek Olšák <maraeo@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35069>	2025-09-24 08:20:28 +00:00
Rhys Perry	026e2527bf	nir/opt_if: rewrite progress reporting and metadata invalidation ... This would unconditionally invalid all metadata except nir_metadata_control_flow and then invalidate that if opt_if_safe_cf_list and opt_if_regs_cf_list made no progress. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Marek Olšák <maraeo@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35069>	2025-09-24 08:20:27 +00:00
Rhys Perry	da23b17c8b	nir/opt_if: fix progress reporting with multiple function impls ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Marek Olšák <maraeo@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35069>	2025-09-24 08:20:27 +00:00
Rhys Perry	12ee2b0fd4	nir: fix progress reporting in nir_io_add_const_offset_to_base ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Marek Olšák <maraeo@gmail.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/35069>	2025-09-24 08:20:27 +00:00
Pierre-Eric Pelloux-Prayer	81f3a5a035	nir/opcodes: remove invalid comment ... "\b" is interpreted by python which results in an invalid char being written to the C file. Reviewed-by: Marek Olšák <marek.olsak@amd.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37331>	2025-09-23 09:09:55 +02:00
Pierre-Eric Pelloux-Prayer	cc4b50b023	nir/opcodes: use u_overflow to fix incorrect checks ... Operands of an addition will be promoted to int making the a+b<a kind of checks ineffective. Use u_overflow.h helpers to perform the check correctly. The commit would be simpler if it used __typeof__ like so: util_add_check_overflow(__typeof__(src0), src0, src1) But typeof only became a standard in C23 so this commit instead extends nir_opcodes a bit to allow opcodes that need the dest_type to get it. Reviewed-by: Marek Olšák <marek.olsak@amd.com> Reviewed-by: Dylan Baker <dylan.c.baker@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37331>	2025-09-23 09:09:55 +02:00
Simon Perretta	7b7fb811ab	pvr, pco: switch to clc load/store sr and idfwdf shaders ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37439>	2025-09-22 14:52:05 +01:00
Simon Perretta	c2127bf4f7	pvr, pco: switch to usc generated clear attachment shaders ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37439>	2025-09-22 14:52:04 +01:00
Simon Perretta	6dd0a5ee2d	pvr, pco: switch to clc query shaders ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37439>	2025-09-22 14:52:04 +01:00
Simon Perretta	3fd3d7ee69	pvr, pco: switch to clc vertex passthrough shaders ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37439>	2025-09-22 14:52:04 +01:00
Simon Perretta	96e4026273	nir: print loop unroll info if present ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Reviewed-by: Georg Lehmann <dadschoorse@gmail.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37439>	2025-09-22 14:52:03 +01:00
Timothy Arceri	870ce22754	nir: fix uniform cloning helper again ... UBOs in different stages can have the same instance name for different UBOs so here we make sure to also check that types match before deciding we have a match. Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13254 Fixes: b47b8d16d9 ("nir: expose reusable linking helpers for cloning uniform loads") Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37428>	2025-09-18 22:28:54 +00:00
Alyssa Rosenzweig	c0c70cb392	nir/lower_flatshade: clean up ... while in the area. Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com> Reviewed-by: Dylan Baker <dylan.c.baker@intel.com> Reviewed-by: Iago Toral Quiroga <itoral@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37448>	2025-09-18 13:37:03 +00:00
Mary Guillemard	d9b11cd7a2	nir/print: Fix load_converted_output_pan and load_readonly_output_pan ... We were not printing IO infos properly for those intrinsics. Signed-off-by: Mary Guillemard <mary.guillemard@collabora.com> Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37452>	2025-09-18 10:41:50 +00:00
Gert Wollny	3b3c3ccf56	nir+r600: add option to avoid contracting fabs into ffma ... On r600 ternary operations can't use the fabs source modifier, so converting "fadd(fabs(fmul(a, b), c)" to "ffma(fabs(a), fabs(b), c)" adds one more instruction in the backend, hence avoid this. Signed-off-by: Gert Wollny <gert.wollny@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37440>	2025-09-17 21:03:58 +00:00
Christian Gmeiner	a7d2570296	nir/opt_algebraic: optimize f2i32(fround_even(x)) to f2i32_rtne(x) ... Add late optimization to fuse f2i32 and fround_even operations into a single f2i32_rtne instruction when the intermediate fround_even result is only used once. This eliminates redundant rounding since f2i32_rtne performs round-to-nearest-even conversion directly. Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com> Tested-by: Simon Perretta <simon.perretta@imgtec.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37426>	2025-09-17 20:31:59 +00:00
Karmjit Mahil	9c6183604f	nir, ir3: Add lower_fmulz_with_abs_min backend option ... This commits adds the `lower_fmulz_with_abs_min` which lowers `fmulz` -> `min(abs(a), abs(b)) == 0.0 ? 0.0 : a * b` `ffmaz` -> `min(abs(a), abs(b)) == 0.0 ? c : ffma(a, b, c) This is useful for ISAs which have `abs` for free on `min` such as ir3. Adreno A750 Benchmark of 10 runs of 5 DX9 single frame trimmed captures looped 2048 times using u_trace measuring `start_render_pass` to `end_render_pass` results: sysmem: -1.91156%, -2.21791%, -2.02533%, -2.21666%, -2.33272%, -2.67349%, -1.75278%, -2.05923%, -2.26892%, -2.10506% Avg: ~ -2.16% ST.S: ~ 0.25% gmem: -3.61496%, -3.66682%, -3.80901%, -3.51198%, -3.72950%, -3.71413%, -3.64467%, -3.67092%, -3.90640%, -3.83888% Avg: ~ -3.71% ST.S: ~ 0.12% Signed-off-by: Karmjit Mahil <karmjit.mahil@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31479>	2025-09-17 15:02:50 +00:00
Karmjit Mahil	8d19ffef0a	nir: Add more matches for fmulz ... In some cases after other passes, `(a == 0.0 ? 0 : b)` can be turned into `(a != 0.0 ? b : 0)`, so let's match those cases too. Also matching `min(abs(a), abs(b)) == 0.0 ? 0.0 : a * b`. Signed-off-by: Karmjit Mahil <karmjit.mahil@igalia.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/31479>	2025-09-17 15:02:50 +00:00
Rhys Perry	810d4ff299	nir/lower_atomics: set ACCESS_COHERENT ... The lack of ACCESS_COHERENT was probably fine, the worst it would do would add a pointless iteration after loading an out-of-date value. Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Daniel Schürmann <daniel@schuermann.dev> Reviewed-by: Karol Herbst <kherbst@redhat.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37417>	2025-09-17 11:29:21 +00:00
Rhys Perry	e819160705	nir/lower_atomics_to_ssbo: set ACCESS_COHERENT for loads ... Signed-off-by: Rhys Perry <pendingchaos02@gmail.com> Reviewed-by: Daniel Schürmann <daniel@schuermann.dev> Reviewed-by: Karol Herbst <kherbst@redhat.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37417>	2025-09-17 11:29:21 +00:00
Ella Stanforth	753af683b5	pco: Switch to common alpha_to_coverage intrinsic ... Signed-off-by: Ella Stanforth <ella@igalia.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:43 +00:00
Ella Stanforth	5ab606ccda	nir: assert when we do not have a sample count when not using intrinsic ... Signed-off-by: Ella Stanforth <ella@igalia.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:42 +00:00
Simon Perretta	6edb72d28b	pco: replace {un,}packing alu ops with intrinsics ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:19 +00:00
Simon Perretta	8104ef4e01	pco: support 1010102 snorm, [us]scaled formats ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:19 +00:00
Simon Perretta	c3325b22d8	pco: image atomics support ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:12 +00:00
Simon Perretta	672541d036	nir, asahi: commonize interleave_agx ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:12 +00:00
Simon Perretta	a67120cda3	pvr, pco: full support for tile buffer eot handling ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:12 +00:00
Simon Perretta	59dc07e02c	pco: improve image write using pck.prog ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:12 +00:00
Simon Perretta	5fa1bb9194	pvr, pco: alpha to coverage support ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:11 +00:00
Simon Perretta	22c67a3c4e	pvr, pco: add dummy stores for tilebuffer-only loadops ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:11 +00:00
Simon Perretta	297a0c269a	pvr, pco: tile buffer support ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:11 +00:00
Simon Perretta	de4dd8e9ea	pvr, pco: fragment shader metadata boilerplate code ... Signed-off-by: Simon Perretta <simon.perretta@imgtec.com> Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36412>	2025-09-16 18:26:11 +00:00