Overall LGTM with a few trivial nits

If you have a preferred fix please commandeer this revision and update the new patch.

Oh I was wondering if there was an upstream!

change should also go here: https://github.com/davidchisnall/dtc

Thank you for the investigation and patch. I commented on a few minor items, I will update them before committing to svn (i.e., no need to reupload).

Overall seems reasonable

committed to FreeBSD, will upstream soon

Ok, please review my proposed ObsoleteFiles.inc/OptionalObsoleteFiles.inc changes in D23124

update ObsoleteFiles.inc and OptionalObsoleteFiles.inc

Tracking PR: PR228919

rS357664

The secure/ and stand/ ones have an upstream or at least other copies around and I'll exclude them initially, but will commit the rest shortly.

Tree-based cache added in rS357450 instead

please upload with context and presumably the new entry should be sorted in alpha oder

I agree with @grehan, vmm_is_svm is preferable rather than changing all cases to vmm_is_amd || vmm_is_hygon

for future updates please include context - see https://wiki.freebsd.org/Phabricator for info
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

In D23531#516575, @emaste wrote:

pe.c: In function 'create_pe':
pe.c:177:3: error: 'strncpy' specified bound 8 equals destination size [-Werror=stringop-truncation]

strncpy(psh.sh_name, name, sizeof(psh.sh_name));
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

And note there is still one error after this change:

cc -O2 -Wno-error -DWITH_PE=1 -I. -I/tmp/cirrus-ci-build/elfcopy -I/tmp/cirrus-ci-build/elfcopy/../common -I/tmp/cirrus-ci-build/elfcopy/../libelf -I/tmp/cirrus-ci-build/elfcopy/../libelftc -I/tmp/cirrus-ci-build/elfcopy/../libpe  -Wall -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -Wno-uninitialized -Wreturn-type -Wcast-qual -Wpointer-arith -Wwrite-strings -Wswitch -Wshadow -Werror   -c pe.c
pe.c: In function 'create_pe':
pe.c:177:3: error: 'strncpy' specified bound 8 equals destination size [-Werror=stringop-truncation]
   strncpy(psh.sh_name, name, sizeof(psh.sh_name));
   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
*** Error code 1

Using memcpy avoids the warning and makes use of the already-calculated strlen, and I'm inclined to commit this as a first step. That said, I'd like to replace all of these with a better idiom.

OK with me, with usr.bin/Makefile update

In D10335#516241, @jhb wrote:

I think it would be good to first import lib9p into the vendor/ space and then contrib/

• rebase
• remove gperf remnants
• remove cc_tools leftovers from Makefile.inc1

-       # Clean out improper TARGET=MACHINE files
-       ${_+_}cd ${.CURDIR}/gnu/usr.bin/cc/cc_tools; ${NXBTMAKE} cleandir

Can you add a link to the PR/patch submission in here?

Also can you add a proposed commit message - it will be something like:

This is a change you'd keep locally while testing, it won't be committed upstream. (You can "abandon" it in Phabricator.)

Please include context with the change - see https://wiki.freebsd.org/Phabricator

I would do it differently

In D23335#515308, @rgrimes wrote:

when dump/restore are invoked with the leading r they change behavior and speak "rmt". Yes, there is probably much documentaiton missing.

In D23335#515296, @rgrimes wrote:

You need r{dump,restore} to talk rmt protocol to a remote tape drive, you can not just pipe over ssh to a remote tape drive.

rmt, rdump, rrestore all support something other than rsh(1) as a transport mechanism, ie ssh

In D23335#515229, @rgrimes wrote:

How are the symlinks /etc/termcap, /etc/unbound, and /etc/aliases handled by pkgbase? You certainly can't go removing those.