bd1976bris wrote:

Thanks @c-rhodes,

> the PR title is misleading

Can I adjust the PR title? I didn’t want to do so in case it interfered with 
the automated backport process.

> it clearly does a lot more than that. I also see it touches LTO and 
> lib/Support so it looks like there could be risk outside of DTLTO > but I'm 
> not sure without looking at the code it changes more clearly and seeing where 
> else it's used.

The changes to the support library are only exercised by DTLTO. There were some 
minor changes to other support utilities; those changes were mechanical, and 
the paths are well covered by unit tests and broader test coverage. That said, 
if the preference is to be especially cautious, I understand the hesitation.

> @teresajohnson wdyt about backporting this as maintainer?

To add some context on why I proposed the backport: this resolves the last 
significant usability gap in DTLTO. With this applied, DTLTO in LLVM 22 should 
handle any link scenario that ThinLTO already supports, without requiring 
special build system adjustments or environment tweaks.

https://github.com/llvm/llvm-project/pull/181059
_______________________________________________
llvm-branch-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits

Reply via email to