[PATCH] D101980: [RFC] [Coroutines] Put the contents of byval argument to the frame

[Xun Li via Phabricator via llvm-commits]

lxfind added a comment.

Sorry for being late in reviewing this patch. I have been out of work for a few days.

The description of the problem is not accurate.
Even for byval parameters, their content will still be copied into the local stack for coroutines correctly. This is handled by the front-end.
So parameter copying is not the problem here.
The problem is exactly what @rjmccall said about the assumptions optimizations typically make about parameters.
The bug linked in this patch, is an example of such issue: for a byval parameter A, a memcpy from A to B, followed by another memcpy from B to C, will be optimized into a memcpy from A to C. And if the memcpy of B to C happens after a coroutine suspension, this leads to memory corruption because A would have died by that point.
D101510 <https://reviews.llvm.org/D101510> is one attempt I tried to fix this, but it's too hacky.
D100415 <https://reviews.llvm.org/D100415> is more long-term approach, but it's much harder than I thought and I don't know how long it will take.
So the idea for this patch is to see if we can find a way to fix this problem within the coroutine passes only.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D101980/new/

https://reviews.llvm.org/D101980