[Lldb-commits] [lldb] New ThreadPlanSingleThreadTimeout to resolve potential deadlock in single thread stepping (PR #90930)

[via lldb-commits]

================
@@ -58,8 +58,13 @@ class ThreadPlanStepRange : public ThreadPlan {
   // run' plan, then just single step.
   bool SetNextBranchBreakpoint();
 
+  // Whether the input stop info is caused by the next branch breakpoint.
----------------
jimingham wrote:

This routine does NOT distinguish between the cases where the next branch breakpoint is the ONLY breakpoint at this site, and where it's one of many breakpoints that share this site.  Later, you distinguish between the two cases where this is used (e.g. in ThreadPlanStepRange::NextRangeBreakpointExplainsStop.
It might be interesting to see if it makes sense to incorporate the logic that does that into IsNextBranchBreakpointStop.  If it does that would allow some code reuse.  If it does not, you should note in the comment here that it does not distinguish between the two cases, so someone doesn't make a mistake later on.

https://github.com/llvm/llvm-project/pull/90930