[Lldb-commits] [PATCH] D97644: Allow RegisterContext to track if behaves-like-frame-0, allow LanguageRuntime for above frame 0

Mon Mar 1 10:53:47 PST 2021

clayborg added a comment.

In D97644#2593228 <https://reviews.llvm.org/D97644#2593228>, @jasonmolenda wrote:

> Another alternative to RegisterContext::BehavesLikeZerothFrame that I've thought of is RegisterContext::GetPCForSymbolication, similar to GetPC() today.  I think everyone who is decrementing $pc is doing it for symbolication, so having this hint and then leaving it to everyone to decrement-or-not may not be a great choice.  It may be easier for higher-levels if RegisterContext can provide an Address suitable for symbolication directly.

If "behaves_like_zeroth_frame" is in the register context for the purpose of backing up the address, then I vote for RegisterContext::GetPCForSymbolication(). We can actually backup the PC by the min instruction size if we really want to get fancy. It would be really nice to have this centralized in one location so we don't have a bunch of code locations backing up the PC.

> For the Swift async unwinds, the saved pc values we have are not normal ABI style calls, and backing up the resume pc results in bad symbolication. I've been meaning to add a capability like this to RegisterContext for a while for DWARFExpression::Evaluate where we get a frame's pc value to look up a location list entry too (and do not decr the pc resulting in bogus values being displayed for noreturn calls) - I'll fix that in a separate patch.

What does this mean for ReportCrash and other tools that like to just walk the frame pointer chain? Does that still produce valid stack trace, just with some Swift async frames missing?

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D97644/new/

https://reviews.llvm.org/D97644