Fix build break on osx/arm64 Debug #102631
Conversation
dotnet-issue-labeler
bot
added
the
area-CodeGen-coreclr
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch |
src/coreclr/jit/regset.cpp
Outdated
| @@ -44,6 +44,34 @@ XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX | |||
| XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX | |||
| */ | |||
|
|
|||
| #ifdef SWIFT_SUPPORT | |||
| regMaskTP rsAllCalleeSavedMask; | |||
| regMaskTP rsIntCalleeSavedMask; | |||
There was a problem hiding this comment.
Do you think it would be cleaner to initialize these here instead of in (Never mind, I realized we want the constructor to reset these values between methods compiled) Also, it might be nice to add a comment explaining these aren't RegSet's constructor below?const or constexpr because rsClearRegsModified (and only this method) may modify them in Swift interop scenarios.
There was a problem hiding this comment.
Do you think it would be cleaner to initialize these here instead of in RegSet's constructor below?
sure.
constexpr
Was there a specific TP improvement we saw by marking them constexpr? I am inclined to just make them and it is just cleaner.
regMaskTP rsAllCalleeSavedMask = RBM_CALLEE_SAVED;
regMaskTP rsIntCalleeSavedMask = RBM_INT_CALLEE_SAVED;There was a problem hiding this comment.
Will check if there is any TP impact with this change.
There was a problem hiding this comment.
Was there a specific TP improvement we saw by marking them constexpr?
This was part of a bigger change, so I'm not sure what the TP impact is of changing this. I'm curious to see as well. If there is no impact, then maybe we should move them back into the RegSet class.
|
Thanks @kunalspathak! Build on osx-arm64 is succeeded. 👍 |
Fixes: #102618 and the errors reported in
#102297 (comment)