C++ Interface Test

Workflow runs for C++ Interface Test

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#34:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 202613m 17s
  More actions for workflow run 34

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#33:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 202638m 21s
  More actions for workflow run 33

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#32:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 202641m 51s
  More actions for workflow run 32

• Add gfx950 Q256 fused MXFP4 prefill pipeline

  skipped:
  C++ Interface Test#31:Pull request#5988synchronizebyvstakhov-amd
  feature/fmoe-q256-prefill

  Oct 23, 20261s
  More actions for workflow run 31

• Add gfx950 Q256 fused MXFP4 prefill pipeline

  skipped:
  C++ Interface Test#30:Pull request#5988synchronizebyvstakhov-amd
  feature/fmoe-q256-prefill

  Oct 23, 20262s
  More actions for workflow run 30

• [CI] Upgrade shared PyTorch images to ROCm 10.1 and PyTorch 2.14

  skipped:
  C++ Interface Test#29:Pull request#6882synchronizebyrkayaith
  rkayaith:ci-rocm101-main

  Oct 23, 20261s
  More actions for workflow run 29

• Add gfx950 Q256 fused MXFP4 prefill pipeline

  skipped:
  C++ Interface Test#28:Pull request#5988synchronizebyvstakhov-amd
  feature/fmoe-q256-prefill

  Oct 23, 20261s
  More actions for workflow run 28

• [CI] Upgrade shared PyTorch images to ROCm 10.1 and PyTorch 2.14

  skipped:
  C++ Interface Test#27:Pull request#6882openedbyrkayaith
  rkayaith:ci-rocm101-main

  Oct 23, 20262s
  More actions for workflow run 27

• [HIP] [FlyDSL] [JIT] Feat/topk per row gfx950

  completed successfully:
  C++ Interface Test#26:Pull request#5686synchronizebyMHYangAMD
  AFDEAPAC:feat/topk-per-row-gfx950-avo

  Oct 23, 202650m 48s
  More actions for workflow run 26

• [Triton/Gluon] [HIP] Add hadamard_rotate to fused_rms_gated_mxfp4_quant

  completed successfully:
  C++ Interface Test#25:Pull request#6866synchronizebymjkvaak-amd
  mjkvaak-amd:triton/fused-rms-gated-mxfp4-hadamard32

  Oct 23, 202638m 25s
  More actions for workflow run 25

• [HIP] Fuse a 32-point Hadamard rotation into the MXFP4 per-1x32 quant and AR + RMSNorm epilogue

  completed successfully:
  C++ Interface Test#24:Pull request#6734synchronizebymjkvaak-amd
  mjkvaak-amd:mxfp4-hadamard32-quant

  Oct 23, 202635m 10s
  More actions for workflow run 24

• [HIP] [FlyDSL] [JIT] [gfx1250] Add MXFP4 a4w4 GEMM with tuned DSR1 configs

  completed successfully:
  C++ Interface Test#23:Pull request#6621synchronizebyaoli26
  gfx1250/flydsl-mxfp4-gemm

  Oct 23, 202618m 33s
  More actions for workflow run 23

• [HIP] [Bugfix] topk_gating kernels: trap when vec_size does not divide num_experts

  completed successfully:
  C++ Interface Test#22:Pull request#6323synchronizebyyzhou103
  truong-v:fix/hip-topk-gating-vec-tail

  Oct 23, 20266m 42s
  More actions for workflow run 22

• [HIP] [FlyDSL] [JIT] [gfx1250] Add MXFP4 a4w4 GEMM with tuned DSR1 configs

  completed successfully:
  C++ Interface Test#21:Pull request#6621synchronizebyaoli26
  gfx1250/flydsl-mxfp4-gemm

  Oct 23, 20266m 0s
  More actions for workflow run 21

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#20:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 20265m 13s
  More actions for workflow run 20

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#19:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 20265m 31s
  More actions for workflow run 19

• fused_norm_rope_att_rope_cast kernel

  skipped:
  C++ Interface Test#18:Pull request#6821synchronizebyyzhou103
  yzhou103:perf/dsv41-pr-fusion-20261022

  Oct 23, 20261s
  More actions for workflow run 18

• MHA v4 sparsified online (sol) attention for gfx950

  skipped:
  C++ Interface Test#17:Pull request#6822synchronizebynsakkine
  nsakkine:mha_v4-sol-attn

  Oct 23, 20261s
  More actions for workflow run 17

• [HIP] [FlyDSL] [JIT] [gfx1250] Add MXFP4 a4w4 GEMM with tuned DSR1 configs

  failed:
  C++ Interface Test#16:Pull request#6621synchronizebyaoli26
  gfx1250/flydsl-mxfp4-gemm

  Oct 23, 20265m 2s
  More actions for workflow run 16

• [Triton/Gluon] [HIP] Add hadamard_rotate to fused_rms_gated_mxfp4_quant

  completed successfully:
  C++ Interface Test#15:Pull request#6866ready_for_reviewbymjkvaak-amd
  mjkvaak-amd:triton/fused-rms-gated-mxfp4-hadamard32

  Oct 23, 20267m 28s
  More actions for workflow run 15

• [Triton/Gluon] [HIP] Add hadamard_rotate to fused_rms_gated_mxfp4_quant

  skipped:
  C++ Interface Test#14:Pull request#6866openedbymjkvaak-amd
  mjkvaak-amd:triton/fused-rms-gated-mxfp4-hadamard32

  Oct 23, 202611s
  More actions for workflow run 14

• [HIP] Fuse a 32-point Hadamard rotation into the MXFP4 per-1x32 quant and AR + RMSNorm epilogue

  completed successfully:
  C++ Interface Test#13:Pull request#6734ready_for_reviewbymjkvaak-amd
  mjkvaak-amd:mxfp4-hadamard32-quant

  Oct 23, 20264m 38s
  More actions for workflow run 13

• [HIP] Fuse a 32-point Hadamard rotation into the MXFP4 per-1x32 quant and AR + RMSNorm epilogue

  skipped:
  C++ Interface Test#12:Pull request#6734synchronizebymjkvaak-amd
  mjkvaak-amd:mxfp4-hadamard32-quant

  Oct 23, 20261s
  More actions for workflow run 12

• [ASM] [HIP] Mha v4: LSE on sorted-sparse, kernel race fixes, gfx950 refresh

  failed:
  C++ Interface Test#11:Pull request#6841synchronizebyjcaraban
  mha_v4_lse_sparse_etc

  Oct 23, 202612m 33s
  More actions for workflow run 11

• fused_norm_rope_att_rope_cast kernel

  skipped:
  C++ Interface Test#10:Pull request#6821synchronizebyyzhou103
  yzhou103:perf/dsv41-pr-fusion-20261022

  Oct 23, 20262s
  More actions for workflow run 10