Thanks for doing the update.

There is the problem that if any of the created *_COMPDIR are not used, it will leave an orphaned directory behind:

Please submit on https://bugs.freebsd.org to find someone to commit it.

Is the profiling library really useful for something (it looks like an accidental artifact of using bsd.lib.mk)? If not, I think just replacing %%PROFILE%% with @comment in pkg-plist and removing the condition would be a better solution.

exp-run was fine.

• devel/racer: Update to 2.1.46 (unbreak it)

• Update to released version of 1.53.0 (fixes regression in clippy)

• Unbreak powerpc64le? (sync with https://github.com/rust-lang/rust/pull/83572)

@mat Do you still plan to fix this?

In D30592#687089, @mat wrote:

I don't like it, FLAVORS are just changing the way something is built, it is not going to change what the software is.

If you really think your port absolutely needs to change what it is supposed to be, there are a handful of ports that already change their COMMENT according to FLAVORS, look at devel/binutils or devel/git.

@bdragon lang/rust-bootstrap@powerpc failed to build. I tried to fix it in 0ade1fd7a6cd7eef17d19cf118bf1ee15ec1c653 but I'm wondering how you managed to build it? Are we missing a patch or something else?

• Mk/Uses/cargo.mk: Fix with empty CARGO_CRATES
• lang/rust: Update to 1.52.1 (diff: https://github.com/rust-lang/rust/compare/1.52.0...1.52.1 i.e. no need to redo QA but was ok on ref13-amd64 anyway)

• Use the right base for arc diff :-/

• Apply final powerpc fixes from bdragon

• Add LibreSSL 3.3.x patch from PR 255652

In D28520#676669, @mikael wrote:

In D28520#675239, @tobik wrote:

@mikael, wondering why you abandoned? I am curious if adding build-stage=2 was already enough.

build-stage=2 made no improvement. I abandoned this because it adds too much complexity to the port.

• Update distinfo with new powerpc bootstraps