Commit Graph

73 Commits

Author SHA1 Message Date
nmzik 7e884f55e0 shader: reuse dominating descriptor loads across image samples 2026-10-13 21:27:07 +00:00
nmzik d70cbd4184 shader: bound descriptor worklists by their finite bit masks 2026-10-13 21:27:07 +00:00
nmzik ed1195d1e7 shader: prove descriptor bounds from simplified image EXEC predicates 2026-10-13 21:27:07 +00:00
nmzik 25e1cc7ec9 shader: preserve descriptor bounds through guarded branch joins 2026-10-13 21:27:07 +00:00
nmzik a287a54f26 shader: prove guarded descriptor counts for interleaved image records 2026-10-13 21:27:07 +00:00
nmzik 68d7bc94e2 shader: correlate masked first-lane material rows 2026-10-13 21:27:07 +00:00
nmzik 8ee31abd9d shader: bound indirect image selectors by counted scalar rows 2026-10-13 21:27:07 +00:00
nmzik 5dc6914688 shader: lower bounded UInt formatted buffer copies to shared stores 2026-10-13 21:27:07 +00:00
nmzik 145578ac03 shader: materialize image keys from proven GPU bitmap subsets 2026-10-13 21:27:07 +00:00
nmzik 0e9c77b0a2 shader: bound material selectors by proven one-wave LDS append lists 2026-10-13 21:27:07 +00:00
nmzik 49ebbfd627 shader: prove descending image ranges and prune impossible register selectors
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-13 21:27:07 +00:00
nmzik 7688093ca3 shader: resolve GPU scalar image selectors within buffer bounds
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-13 21:27:07 +00:00
nmzik fd66f580b4 shader: resolve packed buffer selectors with shared indexed stores
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-13 21:27:07 +00:00
nmzik 1c301c9aab Revert "shader: keep scalar payload and terminal conditions on the GPU" 2026-10-09 21:10:15 +02:00
nmzik 8c81386568 shader: share finite descriptor planning with GPU-selected inline samplers 2026-10-09 15:06:39 +00:00
nmzik 973809cde0 shader: retain bounded scalar material image selectors 2026-10-09 15:06:39 +00:00
nmzik 5b68a26cce shader: keep scalar payload and terminal conditions on the GPU 2026-10-09 15:06:39 +00:00
nmzik 488c0acae9 shader: resolve writable descriptors selected by immutable branches 2026-10-09 15:06:39 +00:00
nmzik 3df456675d shader: own control flow on IR blocks with direct targets 2026-10-09 01:59:41 +02:00
nmzik 190f838ea4 shader: preserve workgroup bounds for scalar image descriptor tables 2026-10-07 21:25:53 +04:00
nmzik 5a491abbda shader: remove redundant descriptor sorting and stale null check 2026-10-06 14:53:17 +04:00
Claxten cd06601022 shader: replace CPU selector chains with direct image-table lookup 2026-10-06 14:53:17 +04:00
nmzik c23de62184 shader: resolve nested descriptor selector tables with exact scalar offsets 2026-10-06 14:53:17 +04:00
nmzik 266c12cec6 Revert "shader: keep scalar data native and preserve raw addresses"
This reverts commit cd9fc6ff2d.
2026-10-05 18:39:31 +04:00
nmzik 5c83060481 shader: resolve masked image keys in strided descriptor records 2026-10-05 14:43:21 +04:00
nmzik b7d3c572c1 shader: clear unused sampler bits after constant folding
Co-authored-by: Çağatay Güçlü <33038942+defektu@users.noreply.github.com>
2026-10-05 02:24:51 +04:00
nmzik 0ee8800905 shader: read image descriptors indexed by workgroup
Co-authored-by: Çağatay Güçlü <33038942+defektu@users.noreply.github.com>
2026-10-05 02:24:51 +04:00
nmzik 3ba0e971e9 shader: materialize exact buffer capacity through shared clean reads
Co-authored-by: shadowbeat070 <70843316+shadowbeat070@users.noreply.github.com>
2026-10-04 18:39:34 +04:00
nmzik 92ac54a493 shader: eliminate impossible relative writes with closed loop bounds
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik 8b05b6730e shader: bind dispatch spans while keeping GPU record counts native
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik cd9fc6ff2d shader: keep scalar data native and preserve raw addresses 2026-10-04 18:39:34 +04:00
nmzik 1f6be3d570 shader: preserve descriptor key SSA across native-resolved image uses
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik b183cb1011 shader: share bounded material-key scans across scalar and vector loads 2026-10-04 18:39:34 +04:00
nmzik 5f01308b20 shader: reuse aligned indirect buffer loads for scalar DWORD reads
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
Co-authored-by: tototomate123 <rushwars201966@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik e0c90acee0 shader: preserve EXEC witnesses across reconverged descriptor loops 2026-10-04 18:39:34 +04:00
nmzik 843b577880 shader: resolve bounded GPU material keys through sampled image arrays
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik a55a28380d shader: decode GPU-selected formatted buffer loads with native bounds
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik b4d32394b6 shader: keep GPU-derived scalar resource reads on the GPU
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
Co-authored-by: TheCruZ <thecruzwtf@gmail.com>
2026-10-04 18:39:34 +04:00
nmzik 429d0af62b shader: bind bounded GPU-selected buffer store destinations
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-04 18:39:34 +04:00
frans bd611bfe15 Honor explicit gather LOD with shared image mip selection
Co-authored-by: YuhiAida <21134775+YuhiAida@users.noreply.github.com>
Co-authored-by: BlackCAT <52720489+psnwd@users.noreply.github.com>
Co-authored-by: shadowbeat070 <70843316+shadowbeat070@users.noreply.github.com>
Co-authored-by: Senaxx <michelbeld@gmail.com>
Co-authored-by: Mac <me@itsmemac.com>
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
2026-10-03 14:19:33 +04:00
BlackCAT ef4c5baf3a Implement RDNA2 64-bit image atomic maximum
Co-authored-by: shadowbeat070 <70843316+shadowbeat070@users.noreply.github.com>
2026-10-03 14:19:32 +04:00
nmzik e317465d8a fix linux CI (#972)
* fix linux CI

* Fix indirect image initialization in materialization tests
2026-10-02 17:20:12 +04:00
nmzik 7804eef9b6 Lower finite GPU-selected image descriptors through shared indirect bindings 2026-10-02 15:37:17 +04:00
nmzik 337f432016 shader: execute PS5 BVH intersections and local FLAT stacks
Co-authored-by: Ali Almohaya <AliAlmohaya@gmail.com>
Co-authored-by: Mac <me@itsmemac.com>
Co-authored-by: Senaxx <michelbeld@gmail.com>
Co-authored-by: Jean-Philippe <github@meunier.jp>
2026-10-02 00:51:15 +04:00
nmzik 0791f921f6 UFC 5 in-game (#883)
* shader: release descriptor dependencies

* shader: simplify shared-exit control flow

* shader: fix scalar memory loads

* shader: fix LDS counter bounds

* renderer: skip inactive pixel shaders

* shader: respect FP32 mode in comparisons

* renderer: fix sampler border colors

* shader: resolve descriptors from control flow

* shader: resolve descriptors before resource planning

* shader: handle scalar guards for image bounds

* shader: handle reduced guards for image loops

* shader: fix wave guards and material indices

* shader: preserve scalar buffer offset alignment and carry

* shader: prove masked loop bounds across iteration transitions

* shader: preserve source and IR before resource tracking failures

* shader: track material keys selected from active lanes

* shader: remove obsolete post-dominator analysis

* shader: follow bounds guards through control-flow joins
2026-09-28 16:03:05 +04:00
nmzik 392f39e2f6 shader: support indirect BUFFER_LOAD_DWORDX3 2026-09-24 20:46:04 +04:00
nmzik 963a0817dc Bound uniformized image keys by guarded material mask
Co-authored-by: TarkusR <90155059+TarkusR@users.noreply.github.com>
2026-09-24 11:13:33 +04:00
nmzik 1617ee77e5 Share CFG reachability for descriptor bounds analysis 2026-09-24 11:13:33 +04:00
nmzik 3b89e36d1e Track bounded indirect image tables from signed shader loops
Co-authored-by: TarkusR <90155059+TarkusR@users.noreply.github.com>
2026-09-24 11:13:33 +04:00
nmzik 66f1a82365 Resolve scalar image tables during resource planning 2026-09-23 15:41:02 +04:00