DuncanMcBain wrote: Thank you Jim and Jason for the reviews. I'd like to try to make sure I understand both of your comments better before I go charging in the wrong direction! Also sorry for the mild noise of push events, I was pushing some junk when I was trying to test things on another machine.
As far as the functions I've changed around in Platform.cpp, I have tried to keep the interface more or less the same, while also exposing the bit that I need (i.e. the actual instruction bytes). I can try to change the behaviour of Platform::GetSoftwareBreakpointTrapOpCode but I thought it best to tread a bit more lightly on that code because I don't know the wider effects of altering it, but I think I can tell approximately what you're suggesting. If you'd like, I can try to change GetSoftwareBreakpointTrapOpCode to read the program bytes at the bp_site location, but maybe that makes more sense as a separate PR? Otherwise, Jason, I think you're correct in your analysis. I will probably try to rework how the IsValid works on RISC-V/ARM to do a check against more byte sequences, as that's probably the cheapest way and will catch all cases. I will update this with more changes next week and make sure the branch is more up-to-date. Thank you very much for your time and have a great weekend. https://github.com/llvm/llvm-project/pull/174348 _______________________________________________ lldb-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
