Jessica Clarke

We want to stay as close to upstream as possible

1. Upstream not downstream 2. You still have diffs in exception.S 3. Please squash commits rather than adding reverts to PRs, otherwise it gets really messy

> The value of errno is zero at program startup, but is never set to zero by any library function. C99 7.5p3

On reflection, I do wonder whether we really want any of this or should just ditch it; `procstat vm` isn't printing any of it (which is probably wise, it'd be...

Why is this a CheriBSD issue

It will only archive if you build the branch as a branch though; PRs have env.CHANGE_ID set so the condition is still false. Up to you whether you add the...

> freebsd64 compat isn't implemented and is going to be complicated. I believe it will require threading a set of ifdefs through the whole management code including building the nvpair...

Hmm. This is in heavy violation of style(9), is a rather ugly concept and putting it in reloc.c is a bit arbitrary. If anything the assertions should be next to...

My guess is: https://github.com/CTSRD-CHERI/cheribsd/blob/d870550ebe149e1002fe5eed2705c7dc76b73c11/sys/fs/nfsclient/nfs_clvfsops.c#L1298 That's definitely nonsense, though works when you don't have bounds. nfs_mount_parse_from just called `strdup(nam)` to give the result, so it copies a bunch of junk past...

For example: > Timeout running cheritest: timeout running /bin/cheriabitest -a -x > /tmp/cheriabitest.xml which doesn't show up in the Tests tab since that results in no JUnit file at all.