Michael137 wrote: > Should the MSVC change move to another PR?
Yes please > We can't run the API tests with MSVC, because its command line options are > different. Would a shell test suffice? I think as a temporary measure shell-tests seem reasonable. Getting API tests (and subsequently all the formatter tests) to run against MSVC would be great, but before doing that, a good first step is to ask on Discourse about the level of support/infrastructure we're willing to dedicate to testing MSVC. AFAIU the general practice is "if you're using MSVC, use the debugger that goes with it". I'd be interested to hear about the use-cases where a user compiled with MSVC but doesn't want to use the associated debugger. Are we talking about programs that were partially compiled with clang-cl and partially with MSVC? Is that a common workflow? I just don't work on Windows enough to be able to answer these questions personally. > I couldn't find a method on CompilerType to look up unscoped enum values in a > class/struct. Because that workaround is MSVC specific, I used the name > present in the PDB. Is this fine or should there be a method on CompilerType > for this? Lookup into anonymous scopes is a bit of a can of worms. There are a couple of issues with it and would be nice to fix eventually. But it's a bit more complicated than it seems because we have to decide whether we want to break existing SBAPI users relying on the old lookup behaviour (which doesn't account for anonymous scopes consistently). I think an MSVC-specific workaround seems fine for now. https://github.com/llvm/llvm-project/pull/172360 _______________________________________________ lldb-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
