[PATCH] D152973: Reland "[gold] Add preliminary FatLTO support to the Gold plugin""

Mon Jul 24 15:55:38 PDT 2023

paulkirth added a comment.

In D152973#4529866 <https://reviews.llvm.org/D152973#4529866>, @nikic wrote:

> In D152973#4529730 <https://reviews.llvm.org/D152973#4529730>, @paulkirth wrote:
>
>> In D152973#4529666 <https://reviews.llvm.org/D152973#4529666>, @MaskRay wrote:
>>
>>> If `-ffat-lto-objects` doesn't need to go to the LLVM 17.0.0 release, we can do a soft transition:
>>
>> Given that Rust pins the LLVM revision to a stable release I think it may be good to get it into Clang 17. Doubly so, since it seems to me that they already have a version of FatLTO, and would benefit from using the FatLTO pipeline.
>
> I don't think we are going to use the FatLTO pipeline as currently implemented in Rust, due to its bad compile-time characteristics. We need a FatLTO implementation that does not require compiling the module twice.

That's unfortunate, and is something I hope we can address moving forward. I vaguely recall your suggestion that after D148010 <https://reviews.llvm.org/D148010> we could probably do better w.r.t. optimizing the pipeline, is that still correct? Also, out of curiosity, how does `rustc` avoid the edge cases brought up in the initial patches, like missing certain transforms for ThinLTO + SamplePGO?

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D152973/new/

https://reviews.llvm.org/D152973