jryans wrote:

Hmm, I am unsure which implementation approach is best here... Mainly, I don't 
think I have enough historical context to understand why some people (as 
discussed in https://github.com/llvm/llvm-project/pull/211804) seem to 
occasionally desire a mode of "disable function body optimisation, but allow 
IPO". Over in that PR, @jhuber6 
[says](https://github.com/llvm/llvm-project/pull/211804#issuecomment-5233353639):

> We still want optnone functions to participate in IPO optimizations

...but I wonder when that's actually desired...?

Perhaps it is simpler for this PR to proceed as-is (adding `noipa` attribute) 
for now, and then a separate PR / RFC could try to untangle the history of 
"optnone, but with IPO", and then make any changes needed after that discussion?

https://github.com/llvm/llvm-project/pull/218782
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to