In D44162#1007615, @jlduran_gmail.com wrote:

It is worth noting that building with clang 18 currently fails (https://api.cirrus-ci.com/v1/task/5600331233492992/logs/build_world.log), I imagine this will be ironed out as we go along.

No objections from me :)

@devnexen_gmail.com I think you implemented https://github.com/llvm/llvm-project/commit/691b12a2dcc12fa43517d23f2a9b6039616eebc8, but the problem is that it specifcally did _not_ call elf_aux_info since that is intercepted, and might lead to segfaults.

Hmm https://github.com/llvm/llvm-project/commit/691b12a2dcc12fa43517d23f2a9b6039616eebc8 moved this function out of the file, but reverted back to calling elf_aux_info()! That was really a mistake, I'll mention that upstream.

In D44030#1005530, @arichardson wrote:

In D44030#1004970, @kevans wrote:

In D44030#1004924, @arichardson wrote:

Have you tried running the lsan test suite with this change?

I'm working on it, but just trying to get the llvm test suite running has been... rough. I chose to redirect from your question to the tangentially related and still valuable: "Have you tried running the asan test suite with this change?" since we can know how much of the asan test suite has passed already.

After fumbling around and finally figuring out that I needed to configure llvm/ with -DLLVM_ENABLE_PROJECTS="clang;compiler-rt" because some of the test prereqs are in llvm or clang, most of the tests that actually run fail because they're trying to do something like:

22:56 <@kevans91_>  70252 103012 Asan-aarch64-inline CALL  execve(0,0x81278f78,0x81278f98)
22:56 <@kevans91_>  70252 103012 Asan-aarch64-inline RET   execve -1 errno 9 Bad file descriptor

... and this is the realization that I just burned a not-insignificant chunk of time because this hasn't been upstreamed yet, three months later: https://cgit.freebsd.org/src/commit/?id=4c9a0adad18263ec8725d9bfc5f560c6ad1da8bd

If you or @dim upload that change I can give it an approval and we can merge it.

In D44030#1004924, @arichardson wrote:

Have you tried running the lsan test suite with this change?

Yeah, it's fine to test this locally but it should really be submitted upstream. If this goes through changes upstream, due to comments, it's a bit of a burden to maintain downstream.

Rename option to WITH_LLVM_FULL_DEBUGINFO, off by default.

Maybe I should start with putting LLVM_LIMIT_DEBUGINFO in __DEFAULT_YES_OPTIONS for now? Then it would not influence most people's builds at all, but it is an easy option for those that do like full debuginfo?

Hmm... I did a test run on universe15a with MK_LLVM_LIMIT_DEBUGINFO=yes and MK_LLVM_LIMIT_DEBUG_INFO=no:

In D35327#998729, @vishwin wrote:

This is probably needed for the latest versions of www/webkit2-gtk, which use C++20 and these constructors/destructors amongst other things.

LGTM

Do we still have MIPS in 13? If so I'd like to leave it in for ease of maintenance. Do we already disable LLVM_TARGET_MIPS by default in src.opts.mk (or wherever the default is set)?