oontvoo wrote: > I find it confusing to use the same entry type for > `Target::SetExecutableModule` and `Process::SetExitStatus`. I think it just > doesn't match the lldb object model. I suspect you somehow want to bracket > the existence of a process between a "start" and "exit" events, but there's > no guarantee that this will do that. > > SetExecutableModule can be called both before a process is started and after > that (multiple times, due to execve). And if you run the same target > repeatedly you may get multiple SetExitStatus calls without ever changing the > executable module. > > I think we need to slow down and figure out what exactly you want to collect, > and then find a place to do that.
At a high level, we want to know: What targets were debugged + related startup/exit statistics? - How long did it take to load the target? ( the first time it was load vs subsequent times ...) - Did the target exit successfully? (if not, what was the issue) How long did the "exit" take? So in that sense we care slightly "less" about a "process" but more about a "target". (We'd still like to keep track of the process as a way to "count' how many times a target (or many targets) was run in a whole debugging session). Does that make sense? https://github.com/llvm/llvm-project/pull/127834 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
