Looks ok

(email followup to copyright holders sent Jun 18)

See linux_map_osrel

Minor tidyup

Also support V2. It's identical to V3 except that in Linux it produces a warning if used.

Update copyright and minor cleanup

I like the idea, but I think the version comparison ought to be simplified and made more general. What do you think about parsing the version into a __FreeBSD_version style single integer in the sysctl handler and simplifying any comparisons of the value?

I did this in my tree this morning after our IRC discussion but went out for the afternoon. Committed after IRC discussion with @allanjude

In D15827#334689, @rgrimes wrote:

I actually worry less about man page DEPRECATION, and more about having the binary spit something out so anyone using it for who knows what gets a heads up that it is/may be going away soon.

perhaps this for deprecation notice:

Index: usr.sbin/asf/asf.8
===================================================================
--- usr.sbin/asf/asf.8  (revision 335216)
+++ usr.sbin/asf/asf.8  (working copy)
@@ -37,6 +37,15 @@
 .Op Fl o Ar outfile
 .Op Fl X Ar suffix
 .Op Ar modules-path Op Ar outfile
+.Sh DEPRECATION NOTICE
+The
+.Nm
+utility is not present in
+.Fx 12.0
+and later.
+The
+.Xr kgdb 1
+debugger handles kernel module state internally.
 .Sh DESCRIPTION
 By default,
 .Nm

Objdump removal tracking pr in https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=229046

Committed to rendered man page in rS335219.

In D15237#334538, @andrew wrote:

How will this work if we decide to add linux32 on arm64? I expect we will need to implement the legacy syscalls there.

LGTM

It might be worth moving the clarification you provided in response to my query into a comment above linprocfs_dominfree.

rS329166

Change seems sensible to me.

No underscore in the commit metadata :)

Thanks for this change - looks good to me, but one small question first. Also for future uploads can you include full context as on https://wiki.freebsd.org/Phabricator - e.g. one of:

git diff -U999999 other-branch > change.diff
git show -U999999 <commit-hash> > change.diff
svn diff --diff-cmd=diff -x -U999999 > change.diff

This ends up calling into ffs_sbget which initializes it like so:

Is setting it in the environment expected to work? What about make SRCCONF=/foo/bar/baz ...?

Ok, can someone else who knows more about this take that as an action item?

In D15211#329988, @sbruno wrote:

In D15211#329987, @imp wrote:

No, we've always just used efiFOO for the efi programs.

Oh, so you're saying that "uefisign" should be a symlink to efisign? I don't know the background here.

Successfully tested the updated version in QEMU in both EFI and CSM modes.

This is not quite the right way to phrase it. Discussion is ongoing on mailing lists now, let's let that continue before revisiting this.

Another report https://forums.freebsd.org/threads/so-freebsd11-amd64-memstick-is-uefi-only.50660/

Overall looks good to me. Potential worries would be stack usage and 4k block size issues.

Related PRs:

• PR227422
• PR227954