[PATCH] D99179: [RFC] [Coroutines] Enable printing coroutine frame in debugger if program is compiled with -g

[Chuanqi Xu via Phabricator via llvm-commits]

ChuanqiXu added a comment.

In D99179#2688049 <https://reviews.llvm.org/D99179#2688049>, @dblaikie wrote:

> Something @probinson mentioned in the llvm-dev thread is that any debug info this creates should be sure that it doesn't produce duplicate unique entities - if it has to create any support types. Does the code correctly only produce one of them, even if there are multiple coroutines being lowered?

Thanks for reminding! I don't know why I don't get the responding from @probinson even in the spam list. And I find it here: https://groups.google.com/g/llvm-dev/c/hfZ6z_MJFOw.
It looks like it wouldn't create redundant base types. But it would still create one coroutine frame type for each coroutine. It makes sense because the layout of coroutine is different with each other (Although it could be the same, as the updated test case tells).
And I didn't do type cache in the module level actually. So I guess it is handled by the infrastructure of the DI system.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D99179/new/

https://reviews.llvm.org/D99179