Sorry, I slightly let the ball drop here. How do I create a patch that only applies on one major version?

Ping?

This was done independently by @gahr in rP412743.

So, I can't not approve... :)

• Update to v2.6.
• Add PyQt RUN_DEPENDS

I will now! :)

I suppose that's what I get for updating the port. :)

1. Updating D5547: Fix shebang lines in DTrace Toolkit scripts.

On the MAKE_JOBS_UNSAFE thing: I see what you mean now that I try to build with make -j12. Sorry, I'm still pretty new to porting... do I just set MAKE_JOBS_UNSAFE=yes in the Makefile?

libdistance uses pkg-config --cflags tcl as part of its build configuration, which works fine on 10 and 11, but not on 9:

Thanks, all, for the comments! It sounds like there might be some rationalization of suffix rules in the works? If so, perhaps I should hold off so that we only have to add a single set of rules rather than adding the same thing twice in bsd.lib.mk and sys.mk...

Closed by rP402448.

Ping?

Thanks: I've moved the LIBDIR stuff to the port Makefile rather than patches. It's currently not that much nicer, but I'm hoping to upstream a bunch of the other changes and reduce the number of patches required in the ports tree.

• Move LIBDIR, etc., to port Makefile when possible.

Thanks again for all of your help... hopefully this is the last newbie question!

• Oops: also install header file.

• Drop PREFIX?= from patches.

• Add USES=tcl rather than tcl BUILD_DEPENDS.
• Add URL to pkg-descr.

• Apply changes suggested by reviewers.

• Add new port: devel/myrepos.
• Minor changes suggested by portlint -A.

• Add new port: devel/libdistance.
• Apply fixes suggested by portlint -A.

Last issue fixed; port added in r397538. Thanks for your help!

Thanks very much! Just to be clear, as a src-only committer, does your comment mean that I ought to commit the change to the ports tree with an Approved by: mat line?

• Remove duplicate MASTER_SITES.
• Rename port to libpololu-avr.
• Add USES=zip, since the distfile is a zip file.

This seems good to me, except for the stdin business...

I think we need an answer to the CAP_IOCTL question before this change proceeds further.

• Add myself as port maintainer.

• Remove redundant GitHub information.

• Change maintainer to ports@FreeBSD.org.

Marking request for XXXJA in the fdopen()+O_BENEATH case as done.

• Add an XXXJA to the fdopen() with O_BENEATH case.

• Let ktrace output depend on nonrelativeerrno.

Apply some fixes from portlint -A.

I'm glad it was helpful!

Sorry for all of the recent noise: it seems that Arcanist requires extra care when doing distributed development!

• Implement O_BENEATH flag for openat(2).
• Rename nameidata::ni_strictrelative to ni_nonrelativeerrno.
• Remove ndrequire_strict_relative_lookups().
• Increase style(9) compliance.
• Remove unused namei() flag.

• Let ktrace output depend on nonrelativeerrno.

Any update on the stdin thing?

So... any progress on @rwatson's CAP_IOCTL comment?

Sorry that I've been a bit unresponsive. I see no problems here, but then again I'm neither a Ports nor a Rust expert!

Well, with all respect to kib, I disagree. We don't need to have a fully-worked-out attack vector before we add security features to our tools, and I think that the ELF format is one of those things that we should treat with some suspicion.

Can you be more specific about kib's objections?

The cargo bootstrap process is unfortunate, but there's probably not much we can do about that. I had been trying to write a tool to bootstrap Cargo from just the Rust compiler, but it's not hardly working yet.

Ok, so I believe that I have addressed all of the comments above... please let me know if I've missed anything! If I have indeed addressed everything, it might be appropriate to hit the LGTM button now. :)

• Increase style(9) compliance.
• Remove unused namei() flag.

Remove ndrequire_strict_relative_lookups().

Thanks for these comments! I should be updating (and shrinking) the diff soon... just need a little more testing first.

Additional commit:

• Rename nameidata::ni_strictrelative to ni_nonrelativeerrno.

User error: I meant to type arc diff --update D2808

Thanks for the feedback!

Hmmm, it looks like the Rust version was updated in r386722. I guess I can abandon this diff now...

Thanks for your review!

Great: if things look ok to you, would you be the right person to commit the change (I'm not a ports committer)?