--lto-newpm-passes has been in the tree in a while. I used it for a bit and I haven't faced any problems, so I decided to add the last missing bit (--lto-aa-pipeline).
The question (probably for Chandler) is: what's the best way to test this?
Details
Details
Diff Detail
Diff Detail
- Repository
- rL LLVM
Event Timeline
Comment Actions
Yes, this is roughly how I expected the API to be used. Are you running into problems with it?
You might want to name it with a newpm prefix much like the other flag, but I'll leave that mostly up to the LLD folks -- I suspect the local conventions are the most important there.
Comment Actions
No, not at all.
You might want to name it with a newpm prefix much like the other flag, but I'll leave that mostly up to the LLD folks -- I suspect the local conventions are the most important there.
Sure, thanks for your review.