-
Notifications
You must be signed in to change notification settings - Fork 21.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Reland] fix missing-prototypes warnings in torch_cpu (Part 4) #101949
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks!
Do you know why this didn't fail in our CI? Do we just never build with tensorexpr in CI?
@albanD We build with tensorexpr but the test coverage isn't 100%. May be we can add the failed test inside MeTa to the OSS version. |
@pytorchmergebot 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 |
@izaitsevfb is that something you can look into on your end? |
@albanD, correct me if I'm wrong, but the linker issue seems to be stemming from the way we build things like android and xplat at Meta. It could be theoretically possible to find all PT symbols that we use and pin them somehow in OSS, but that's quite a project. For example, it seems that now we're having issue with another symbol, We'll try to brainstorm the solution with @osalpekar. @cyyever, if this PR gets reverted again, please ping me next time before merge, I'll import it manually to see if it breaks anything. |
#101976)" This reverts commit 4db2dad. Reverted #101976 on behalf of https://github.com/osalpekar due to reverting to allow #101949 to be cleanly reverted ([comment](#101976 (comment)))
@pytorchbot revert -m "As noted in @izaitsevfb's comment, we are still seeing linker errors, this time due to |
@pytorchbot successfully started a revert job. Check the current status here. |
@cyyever your PR has been successfully reverted. |
#101949)" This reverts commit 4f2c007. Reverted #101949 on behalf of https://github.com/osalpekar due to As noted in @izaitsevfb's comment, we are still seeing linker errors, this time due to `nnc_prepacked_linear_clamp_run` being made a static function. ([comment](#101949 (comment)))
@izaitsevfb It seems that the changes in external_functions.cpp are the only reason this PR got reverted again. Then a simple solution is to discard them in that file. |
@cyyever, trying that now, will keep you posted. |
@cyyever, the signals looks good after making @osalpekar, please see D46093855. |
@izaitsevfb I have create a new PR #102228 for reland. |
This PR relands the changes introduced in PR #100849. The old PR turnd nnc_aten_embedding into a static function, however, it is actually used in torch/csrc/jit/tensorexpr/operators/misc.cpp.
cc @jgong5 @mingfeima @XiaobingSuper @sanchitintel @ashokei @jingxu10 @EikanWang @izaitsevfb @albanD