Fix typos in CreateTMADescriptorVariable #138877
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/138877
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit a5e235e with merge base 72ea7ba ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command |
| kwargs["block_dim0"] if "block_dim0" in kwargs else args[4], | ||
| ] | ||
| element_size = kwargs["ptr"] if "ptr" in kwargs else args[-1] | ||
| element_size = kwargs["element_size"] if "element_size" in kwargs else args[-1] |
There was a problem hiding this comment.
| element_size = kwargs["element_size"] if "element_size" in kwargs else args[-1] | |
| element_size = kwargs.get("element_size", args[-1]) |
There was a problem hiding this comment.
And for all the other kwargs accesses with the similar type, should avoid accidental typos of the key
There was a problem hiding this comment.
It may be that args is empty and all arguments are passed via kwargs. Would kwargs.get("element_size", args[-1]) not attempt to evaluate args[-1] first? Example:
>>> args = []
>>> kwargs = {"arg": 1}
>>> kwargs["arg"] if "arg" in kwargs else args[-1]
1
>>> kwargs.get("arg", args[-1])
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
IndexError: list index out of range
|
@pytorchbot merge -f |
|
❌ 🤖 pytorchbot command failed: Try |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
This adds host-side Triton TMA support to AOTInductor. Notes: - Two helper functions, `init1DTMADescriptor` and `init2DTMADescriptor` are added to the C++ wrapper codegen on GPU, conditioned on the model having user-defined Triton kernels with host-side TMA (CUDA-specific). - C++ wrapper codegen on GPU emits TMA descriptor initialization via the aforementioned helper functions. - Special handling added for the TMA descriptors (in the Python wrapper codegen) during the compile-time autotuning, as the underlying tensor can't be passed directly to the user-defined Triton kernel. TMA descriptors are generated in-between the source tensor's buffer and the kernel call, like in the full Python wrapper codegen. - This PR concludes the host-side Triton TMA support in PT2. Pull Request resolved: #138878 Approved by: https://github.com/desertfire, https://github.com/chenyang78 ghstack dependencies: #138759, #138877
This adds host-side Triton TMA support to AOTInductor. Notes: - Two helper functions, `init1DTMADescriptor` and `init2DTMADescriptor` are added to the C++ wrapper codegen on GPU, conditioned on the model having user-defined Triton kernels with host-side TMA (CUDA-specific). - C++ wrapper codegen on GPU emits TMA descriptor initialization via the aforementioned helper functions. - Special handling added for the TMA descriptors (in the Python wrapper codegen) during the compile-time autotuning, as the underlying tensor can't be passed directly to the user-defined Triton kernel. TMA descriptors are generated in-between the source tensor's buffer and the kernel call, like in the full Python wrapper codegen. - This PR concludes the host-side Triton TMA support in PT2. Pull Request resolved: pytorch#138878 Approved by: https://github.com/desertfire, https://github.com/chenyang78 ghstack dependencies: pytorch#138759, pytorch#138877
This adds host-side Triton TMA support to AOTInductor. Notes: - Two helper functions, `init1DTMADescriptor` and `init2DTMADescriptor` are added to the C++ wrapper codegen on GPU, conditioned on the model having user-defined Triton kernels with host-side TMA (CUDA-specific). - C++ wrapper codegen on GPU emits TMA descriptor initialization via the aforementioned helper functions. - Special handling added for the TMA descriptors (in the Python wrapper codegen) during the compile-time autotuning, as the underlying tensor can't be passed directly to the user-defined Triton kernel. TMA descriptors are generated in-between the source tensor's buffer and the kernel call, like in the full Python wrapper codegen. - This PR concludes the host-side Triton TMA support in PT2. Pull Request resolved: pytorch#138878 Approved by: https://github.com/desertfire, https://github.com/chenyang78 ghstack dependencies: pytorch#138759, pytorch#138877
Stack from ghstack (oldest at bottom):
This fixes some leftover typos in
CreateTMADescriptorVariable.call_function (and close).
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @chauhang @amjames @rec