In D27011#602724, @markj wrote:

I find the naming of both the internal flags and the socket options kind of confusing.

I have no objection to the change. I'm not really convinced it needs to be part of the pmap code, instead of a self-contained module in the SMMU driver, but I can see why it's convenient.

In D27015#602732, @markj wrote:

This diff doesn't appear to match the description.

In D27019#602758, @mjg wrote:

They had kmem_malloc and that got patched to malloc(...., M_EXEC) so I did not deviate from the general direction. That said, I'm indifferent to this one as long as the flag is gone from malloc itself.

In D26992#602651, @arrowd wrote:

Looks good modulo indentation issue.

Does committing this leave science/mpb in a broken state? If yes, I'd prefer you to prepare two reviews (this one and mpb) and once both of them are approved commit them in one go.

(Tinderboxed.)

They had kmem_malloc and that got patched to malloc(...., M_EXEC) so I did not deviate from the general direction. That said, I'm indifferent to this one as long as the flag is gone from malloc itself.

I think this is fine. I wouldn't even bother with malloc_exec(), you could just modify the few existing users to call kmem_malloc(M_EXEC).

This diff doesn't appear to match the description.

The reason we handle CLONE_FS by setting RFFDG is because a process' cwd (and chroot dir and umask) is stored in the fd table. I'm not sure how we could implement CLONE_FS properly, it'd require some modification to the core kernel.

Can you describe a setup that demonstrates the problem this fixes?
(Ideally in the form of a test case, but if you can describe a setup I can probably write the test case.)

I find the naming of both the internal flags and the socket options kind of confusing. Rather than having two flags to do almost the same thing, you could have UNP_WANTCRED and a second flag, maybe UNP_WANTCRED_ALWAYS, that determines whether WANTCRED gets cleared after the initial read. The new socket option would set both flags.

Thank you SOOOOOOOOOOO MUCH!!!!

I've written and switched to sysutils/quickjail to solve my problem here... simple utility that creates truly non-persistent jails because it doesn't need to observe the result and potentially run other scripts and whatnot.

Huh, that's interesting -- I got it in the manpage, but not the usage... how peculiar.

• Restore dependency of audio/lame
• Place new dependencies in RUN_DEPENDS to the end so that they're more visible in the diff.

• Remove superfluous ${ECHO_CMD} from post-install target.

Add actual content here.

In D26975#602614, @cyprien_cypou.net wrote:

It is; at some point we might need to move this into architecture-specific include. Unless Linux folks managed to fix it for newer architectures, in which case we won’t.

It comes from sys/compat/linux/linux_errno.inc, added there by the linked review.

Forgot to put the updated pkg-plist in the review, sorry!

Thanks Mateusz!

Thanks!

@zeising Is this review approved for an exp-run or not?

In D27014#602641, @scottl wrote:

I was about to upload a different patch that uses the environment. I'd rather take it in that direction since there's already a strong precedence for acquiring LOCALBASE from the environment. Need to finish the man page then I'll upload it. Please hold off on this.

Fix unpleasant indentation for OPTIONS_DEFINE.