Add -Wno-error=nonnull for test/cpp/api/ #99468
Conversation
On some platforms the build might fail due to the nonnull error being triggered by different compiler behavior. Fix for pytorch#99278.
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/99468
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit 367387b: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
|
|
this PR solved my issue and let pytorch to compile. |
|
Can we add this only for riscv? |
It seems the problem is not arch specific. I got similar problem on x86 cpu with gcc-12.3.0 installed from conda-forge and confirmed that this PR fixes the problem. I think it makes sense to apply this regardless of archs. |
| # the nonnull check might trigger for some build configurations, | ||
| # probably happening due to different code optimization | ||
| # (see e.g. https://rkoucha.fr/tech_corner/nonnull_gcc_attribute.html) | ||
| # this happened for a riscv build: https://github.com/pytorch/pytorch/issues/99278 |
There was a problem hiding this comment.
Might make sense to mention that it is not riscv build specific.
There was a problem hiding this comment.
@JackTemaki Are you going to continue work on it?
There was a problem hiding this comment.
Yes I can, it was just not clear for me what changes are needed. Just an altered message, referencing also the issues unrelated to riscv?
|
Looks like this PR hasn't been updated in a while so we're going to go ahead and mark this as |
|
I am also seeing the problem on X86 with gcc/gxx installed by Conda. And this patch works. Is it possible to reopen this PR and land it for all archs? Thanks. @JackTemaki @malfet |
|
solved my problem too :) |
On some platforms the build might fail due to the nonnull error being triggered by different compiler behavior.
Fixes #99278.