Commits

Branch selector

 main

User selector

kerbowa

Datepicker

All time

Commit History

Commits on May 12, 2025

• [AMDGPU] Move kernarg preload logic to separate pass (#130434)

  Show description for 2c9a46c
  kerbowa

  authored
  2c9a46c

  Copy full SHA for 2c9a46c

Commits on Mar 24, 2025

• [AMDGPU] Relax lds dma waitcnt with no aliasing pair (#131842)

  Show description for e75f586
  kerbowa

  authored
  e75f586

  Copy full SHA for e75f586

Commits on Jan 11, 2025

• [AMDGPU] Add target hook to isGlobalMemoryObject (#112781)

  Show description for 657fb44
  kerbowa

  authored
  657fb44

  Copy full SHA for 657fb44

Commits on Jan 10, 2025

• [AMDGPU] Add backward compatibility layer for kernarg preloading (#119167)

  Show description for 2e5c298
  kerbowa

  authored
  2e5c298

  Copy full SHA for 2e5c298

Commits on Dec 8, 2024

• [AMDGPU] Fix hidden kernarg preload count inconsistency (#116759)

  Show description for b1d4246
  kerbowa

  authored
  b1d4246

  Copy full SHA for b1d4246

Commits on Dec 7, 2024

• [AMDGPU] Fix user SGPR alloc order in docs (#119092)

  Show description for aebd338
  kerbowa

  authored
  aebd338

  Copy full SHA for aebd338

Commits on Oct 7, 2024

• [AMDGPU] Support preloading hidden kernel arguments (#98861)

  Show description for c4d8920
  kerbowa

  authored
  c4d8920

  Copy full SHA for c4d8920

Commits on Sep 23, 2024

• [AMDGPU] Include unused preload kernarg in KD total SGPR count (#104743)

  Show description for 954ab83
  kerbowa

  authored
  954ab83

  Copy full SHA for 954ab83

Commits on Aug 23, 2024

• [AMDGPU] Fix crash in allowsMisalignedMemoryAccesses with i1 (#105794)

  kerbowa

  authored
  ceb587a

  Copy full SHA for ceb587a

Commits on Aug 19, 2024

• [AMDGPU][NFC] Fix preload-kernarg.ll test after attributor move (#98840)

  Show description for 7d5281a
  kerbowa

  authored
  7d5281a

  Copy full SHA for 7d5281a

Commits on May 20, 2024

• [AMDGPU] Fix kernarg preloading crash with some types and alignments (#91625)

  Show description for f652777
  kerbowa

  authored
  f652777

  Copy full SHA for f652777

Commits on Mar 31, 2024

• [AMDGPU] Use directive for kernarg preload header padding (#86004)

  kerbowa

  authored
  b5b34db

  Copy full SHA for b5b34db

• [AMDGPU] Extend MFMA padding option to gfx90a+ (#86768)

  Show description for 0234d90
  kerbowa

  authored
  0234d90

  Copy full SHA for 0234d90

Commits on Feb 13, 2024

• [AMDGPU] Enable kernel arg preloading with gfx90a (#81180)

  Show description for 4bcbeae
  kerbowa

  authored
  4bcbeae

  Copy full SHA for 4bcbeae

Commits on Oct 19, 2023

• [AMDGPU] Add doc updates for kernarg preloading (#67516)

  kerbowa

  authored
  d681461

  Copy full SHA for d681461

Commits on Sep 25, 2023

• [AMDGPU] Add DAG ISel support for preloaded kernel arguments

  Show description for 0455596
  kerbowa

  committed
  0455596

  Copy full SHA for 0455596

• [AMDGPU] Add IR lowering changes for preloaded kernargs

  Show description for 7b70af2
  kerbowa

  committed
  7b70af2

  Copy full SHA for 7b70af2

Commits on Sep 19, 2023

• [AMDGPU] Add ASM and MC updates for preloading kernargs

  Show description for 69447d6
  kerbowa

  committed
  69447d6

  Copy full SHA for 69447d6

• [AMDGPU] Use inreg for hint to preload kernel arguments

  Show description for 60a227c
  kerbowa

  committed
  60a227c

  Copy full SHA for 60a227c

Commits on Sep 12, 2023

• [AMDGPU] Add utilities to track number of user SGPRs. NFC.

  Show description for 343be51
  kerbowa

  committed
  343be51

  Copy full SHA for 343be51

Commits on Jun 3, 2023

• [AMDGPU] Don't flush vmcnt for loops with use/def pairs

  Show description for e501ed8
  kerbowa

  committed
  e501ed8

  Copy full SHA for e501ed8

Commits on Mar 20, 2023

• [AMDGPU] Add release note for ommited barrier waitcnt

  Show description for 09f756c
  kerbowa

  committed
  09f756c

  Copy full SHA for 09f756c

Commits on Mar 18, 2023

• [AMDGPU] Reserve extra SGPR blocks wth XNACK "any" TID Setting

  Show description for 864a2b2
  kerbowa

  committed
  864a2b2

  Copy full SHA for 864a2b2

Commits on Jan 25, 2023

• [ScheduleDAG] Fix removing edges with weak deps

  Show description for 913837e
  kerbowa

  committed
  913837e

  Copy full SHA for 913837e

Commits on Dec 10, 2022

• [AMDGPU] Update MFMASmallGemmOpt with better performing stategy

  Show description for f9c76a1
  kerbowa

  committed
  f9c76a1

  Copy full SHA for f9c76a1

Commits on Aug 19, 2022

• [AMDGPU] Add iglp_opt builtin and MFMA GEMM Opt strategy

  Show description for b0f4678
  kerbowa

  committed
  b0f4678

  Copy full SHA for b0f4678

Commits on Aug 5, 2022

• [AMDGPU] Pre-commit tests for D130797

  kerbowa

  committed
  b568cb1

  Copy full SHA for b568cb1

Commits on Aug 2, 2022

• [AMDGPU] Add CL option for max-ilp scheduler.

  Show description for 3dfa562
  kerbowa

  committed
  3dfa562

  Copy full SHA for 3dfa562

• [AMDGPU] Add llvm_unreachable to switch statement added in d7100b398.

  kerbowa

  committed
  40eec27

  Copy full SHA for 40eec27

• [AMDGPU] Add GCNMaxILPSchedStrategy

  Show description for d7100b3
  kerbowa

  committed
  d7100b3

  Copy full SHA for d7100b3

Commits on Jul 30, 2022

• [AMDGPU] Remove unused function

  kerbowa

  committed
  7898426

  Copy full SHA for 7898426

Commits on Jul 29, 2022

• [AMDGPU] Omit unnecessary waitcnt before barriers

  Show description for 2c82a12
  kerbowa

  committed
  2c82a12

  Copy full SHA for 2c82a12

Commits on Jul 28, 2022

• [AMDGPU] Add isMeta flag to SCHED_GROUP_BARRIER

  kerbowa

  committed
  0f93a45

  Copy full SHA for 0f93a45

• [AMDGPU] Add amdgcn_sched_group_barrier builtin

  Show description for f5b2168
  kerbowa

  committed
  f5b2168

  Copy full SHA for f5b2168

• [AMDGPU] Aggressively schedule to reduce RP in occupancy limited regions

  Show description for ba0d079
  kerbowa

  committed
  ba0d079

  Copy full SHA for ba0d079

Pagination

Previous

Next