Is this behavior desirable? It seems very ambiguous to me...

Oof... that's a bit of an increase. Wonder if there's a way to improve things a bit more perf wise (I'd file a bug to track the investigation).

Abandoning as @mm is planning on cutting a new libarchive release soon which will include this fix (and others).

Ship it!

sys/modules/asmc/Makefile: condense SRCS

In D41705#1244931, @ngie wrote:

I scrolled back several releases to see what our prior "Open Issues" sections look like.

In D54512#1245481, @ngie wrote:

BTW, isn't the proposed change what Errata Notes sort of do...?

BTW, isn't the proposed change what Errata Notes sort of do...?

Sorry for the delayed review on this. LGTM!

This is an interesting project. If we put this in tree, FreeBSD should really adopt mfsbsd from @mm and do a full vendor import for the component.

(here are some suggested changes to help get the process going..)

In D54427#1244543, @ngie wrote:

NAK on the "blind" atf_check removal: it makes potentially valid issues that should be tracked in test case runs into silent errors. You have to dig the prior test case output to try and piece together what went wrong after this change, making failures far more annoying or impossible to triage.

NAK on the "blind" atf_check removal: it makes potentially valid issues that should be tracked in test case runs into silent errors. You have to dig the prior test case output to try and piece together what went wrong after this change, making failures far more annoying or impossible to triage.

I'll take up the patch and fix up the whitespace bug since it's a trivial change which adds helpful documentation.

In D54062#1243684, @jrtc27 wrote:

.gitattributes is a footgun and can give confusing behaviour. tools/build/checkstyle9.pl should enforce this for GitHub PRs already, and clang-format can (will?) normalise this too. There are significant downsides for this with little benefit that I can see.

In D54016#1243704, @jlduran wrote:

Thank you!
I wanted as few diffs with NetBSD as possible, but using nitems() is a good suggestion; I'll probably submit it to GNATS.

In D54374#1242935, @cperciva wrote:

Hmm, I thought the way this worked was that env_setup set defaults and then the config file was sourced to override those values.

Please provide an example (spec) of what you're implementing in the Testplan section.

@markj: this change avoids a large amount of grief when building or editing sources on hosts that are incorrectly configured, and/or on Windows hosts. I'm not sure what exactly Linnea did to trigger this, but I assume it had to do with hosting or editing the sources on a Windows box.

I still think it's a good idea to push ${CWARNFLAGS.${COMPILER_TYPE}.${.IMPSRC:T}} into bsd.sys.mk, but this is ok too.

Incorporate feedback from @bz

In D53591#1223710, @bz wrote:

You do not need the .if if you do it as suggested here:
https://lists.freebsd.org/archives/freebsd-hackers/2025-August/004872.html

Fix typo spotted by @bz.

Incorporate corrections/suggestions from @delphij

In D53591#1223568, @delphij wrote:

I think what bz@ really wanted is something like:

.if ${COMPILER_TYPE} == "clang"
CWARNFLAGS.drv.c += -Wno-override-init -Wno-initializer-overrides
.endif

but your change should be sufficient to fix gcc build.