Skip to content

Fix AxisAnalysis to avoid out of bound access.#3832

Merged
etiotto merged 12 commits intomainfrom
etiotto.tdesc_to_block_pointer.6
Apr 4, 2025
Merged

Fix AxisAnalysis to avoid out of bound access.#3832
etiotto merged 12 commits intomainfrom
etiotto.tdesc_to_block_pointer.6

Conversation

@etiotto
Copy link
Copy Markdown
Contributor

@etiotto etiotto commented Apr 3, 2025

Resolves issue #3830.

etiotto added 10 commits April 1, 2025 19:04
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
@etiotto etiotto self-assigned this Apr 3, 2025
@etiotto etiotto requested review from a team and alexbaden April 3, 2025 17:24
@etiotto etiotto marked this pull request as draft April 3, 2025 17:25
@etiotto
Copy link
Copy Markdown
Contributor Author

etiotto commented Apr 3, 2025

Rebase once PR #3816 lands

Comment thread third_party/intel/lib/Analysis/AxisInfo.cpp
@etiotto etiotto marked this pull request as ready for review April 4, 2025 14:35
Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
@etiotto etiotto requested a review from whitneywhtsang April 4, 2025 16:30
Comment thread third_party/intel/lib/Analysis/AxisInfo.cpp
@etiotto etiotto enabled auto-merge (squash) April 4, 2025 18:17
@etiotto etiotto merged commit 1d6ae76 into main Apr 4, 2025
6 checks passed
@etiotto etiotto deleted the etiotto.tdesc_to_block_pointer.6 branch April 4, 2025 19:31
@etiotto etiotto linked an issue Apr 4, 2025 that may be closed by this pull request
@etiotto etiotto restored the etiotto.tdesc_to_block_pointer.6 branch April 4, 2025 20:37
etiotto added a commit that referenced this pull request Apr 7, 2025
Enable the tutorial `09-persistent-matmul` and run it with both tensor
of ptrs (`matmul_kernel_persistent`) and with tensor descriptors
(`matmul_kernel_descriptor_persistent`). The latter kernel is internally
translated into blocked ptrs.


Note: rebase after PRs #3817 and #3832 lands.

---------

Signed-off-by: Tiotto, Ettore <ettore.tiotto@intel.com>
@etiotto etiotto deleted the etiotto.tdesc_to_block_pointer.6 branch April 25, 2025 13:51
david-hls pushed a commit to david-hls/intel-xpu-backend-for-triton that referenced this pull request Jun 18, 2025
This is a follow up PR of intel#3832
`wave` has been replaced with `warp` for the consistency between GPUs.
Unfortunately there are still remaining use of `wave` in the code as
below, although I've tried to minimize it.

## Referencing AMD features (HIP API or AMDGPU)
third_party/amd/backend/include/hip/:
third_party/amd/backend/include/roctracer/:
third_party/amd/backend/include/has/*:
- Cannot completely replace waves because the definition comes from
outside e.g., __AMDGCN_WAVEFRONT_SIZE, hsa_wavefront_info_t
- Mixing up `warp` and `wave` together in the same place could be even
worse.

## Using amdgpu compiler option
third_party/amd/backend/compiler.py:
python/tutorials/03-matrix-multiplication.py:
python/tutorials/06-fused-attention.py:
- `waves_per_eu` which is supposed to mapped to a CLANG attribute
`amdgpu-waves-per-eu`
- It is AMD only option and makes better sense to keep
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[convert-triton-intel-gpu-to-llvm]: Triggers assertion in Axis analysis

2 participants