[PATCH] D38776: [codeview] Implement FPO data assembler directives

David Majnemer via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed Oct 11 10:09:22 PDT 2017


majnemer added a comment.

In https://reviews.llvm.org/D38776#894832, @rnk wrote:

> In https://reviews.llvm.org/D38776#894228, @majnemer wrote:
>
> > I'd also be interested in seeing a stack frame with spills and stack realignment.
>
>
> I'll do that, but it's not very interesting because we basically use a normal EBP-based prologue before AND-ing the stack pointer and optionally copying it to ESI. As far as the unwinder is concerned, it's pretty boring.


I guess I remembered the MSVC prologue sequence.

> 
> 
> In https://reviews.llvm.org/D38776#894218, @majnemer wrote:
> 
>> It'd also be nice to have a test showing how it interacts with stack cookies.
> 
> 
> I think our implementation of these happens in LLVM IR, so it isn't very interesting, but maybe worth adding.

True.

> 
> 
> In https://reviews.llvm.org/D38776#894167, @majnemer wrote:
> 
>> Should we have a shrink-wrapping test?
> 
> 
> This is definitely worth testing, but it should mostly work since it basically splits the prologue block after this stuff is inserted and the label ranges should just work out. Let's find out.

Awesome.


https://reviews.llvm.org/D38776





More information about the llvm-commits mailing list