[Lldb-commits] [lldb] [lldb] Step over non-lldb breakpoints (PR #174348)

[Duncan McBain via lldb-commits]

DuncanMcBain wrote:

Thanks for this extended testing, David. It's actually a bit strange and I'm worried that the patch is making this slightly strange situation a bit worse.

This might be expected, but building with debug information changes where the breakpoint lands when you issue `b main`. I didn't expect it at any rate.

This behaviour implies that the StopUnixSignal event isn't generated for the trap instruction actually in-memory, and I cannot claim to know why that would (or rather would not) happen. It does look like the breakpoint considers itself hit repeatedly when you continue and the breakpoint is enabled.

Comparing with an unmodified lldb (version 16), it looks like the behaviour is the same, so it's maybe at least not a regression? I am really not sure how best to try to fix this though, as I don't know why my code isn't being called.

https://github.com/llvm/llvm-project/pull/174348