[clang] [llvm] [clang][modules] Prevent deadlock in module cache (PR #182722)

Aaron Ballman via cfe-commits cfe-commits at lists.llvm.org
Wed Jun 24 04:15:23 PDT 2026


AaronBallman wrote:

> > Just noticed this. Please do NOT do this, it is against our community policies per this accepted RFC: https://discourse.llvm.org/t/rfc-specify-a-community-policy-for-use-of-private-links/72208
> 
> The final community policy is that private links can't explain the context for the change but it's perfectly fine to supplement private links for e.g. downstream automation.
> 
> see:
> 
> > It is also acceptable to add other metadata to the commit message to automate processes, including for downstream consumers. This metadata can include links to resources that are not available to the entire community. However, such links and/or metadata should not be used in place of making the commit message self-explanatory. Note that such non-public links should not be included in the submitted code.
> 
> in: https://llvm.org/docs/DeveloperPolicy.html

I believe this use complies with community policies but at the same time, this is not something the Clang community typically sees anyone ask for and we actually had to spend a fair amount of effort excising incorrectly added radr links in the past. Is there a new push to use radr links more consistently or was this a one-off request? (If it's a push to do this more consistently then the community could use a reminder about the policy so we don't end up in a bad place again by accident.)

https://github.com/llvm/llvm-project/pull/182722


More information about the cfe-commits mailing list