[libcxx-commits] [PATCH] D134604: [clang] Instiantiate early substituted entities with sugared template arguments

Louis Dionne via Phabricator via libcxx-commits libcxx-commits at lists.llvm.org
Tue Sep 27 15:30:51 PDT 2022


ldionne added a comment.

In D134604#3815634 <https://reviews.llvm.org/D134604#3815634>, @mizvekov wrote:

> I was aware of those discussions, and I even made a workaround to avoid this extra cost of running unrelated pipelines.
> The patch on the very bottom of this stack (D134079 <https://reviews.llvm.org/D134079>) disables all the other pipelines except the bootstrapping one.

Oh, awesome, I wasn't aware of that. Okay, this mitigates a lot of my concern then. The only remaining annoyance is the additional traffic on the libc++ review group, but that's manageable if it gets you the testing you need until we've improved the CI situation w/ Clang.

> I have two other patches in the stack which I will need to keep running the libcxx-ci as I update them: D131858 <https://reviews.llvm.org/D131858> and D127695 <https://reviews.llvm.org/D127695>. These have had non-trivial libcxx breakages, and I have even seen new problems introduced as you keep working on the ranges implementation.

Understood, in that case I think your usage of the CI is definitely justified. I mostly wanted to get the word out that this is not a pattern we want to encourage more generally, for the reasons explained, but I think it doesn't apply here.

> Thanks by the way for the awesome work on the libcxx-CI, and I hope in the near future these workarounds will not be needed anymore!

Hey, thanks to you for wanting to ensure your patches don't break libc++ :-)!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D134604/new/

https://reviews.llvm.org/D134604



More information about the libcxx-commits mailing list