JDevlieghere wrote:

Thanks for the context. My initial concern, which sounds validated by your 
explanation, is that in the most degenerate case, we may hit each breakpoint 
individually and do no batching at all. In other words, the test would still 
need to pass if no batching is happening at all so it would pass without the 
original change as well.

I think this plus the GDB remote test case where we spoof the packets and 
confirm that we've actually batched all 10 would give us the coverage we're 
looking for. The latter would fail without the batching, and this one may 
uncover other issues when talking to a real stub, albeit without necessarily 
enforcing that any batching has happened. 

https://github.com/llvm/llvm-project/pull/182415
_______________________________________________
lldb-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits

Reply via email to