Re: [PATCH 0/6] Installed-header hygiene (final, hopefully)

[Carlos O'Donell <carlos at redhat dot com>]

On 09/22/2016 09:30 AM, Zack Weinberg wrote:
> I have made all the changes requested on previous iterations of the
> patch series.  I have also reorganized it to indicate which groups of
> patches I propose to squash together for landing: 1[abcd], 2[abc], and
> 3[abc].
> 
> I believe that the only change not yet approved is the test script in
> patch 6, which Carlos wanted Joseph to look at in detail.

Just to reiterate, I think the test script is good as-is, but it would
be better if such a framework got review from a second senior developer.

This is the time in the review where you might reach out to someone else
to get specific review of the test scripts, that should be sufficient
consensus for me that it's good to go in e.g. one more developer review.

Regarding test machinery like this the top-three people that come to mind
are: Joseph Myers, Florian Weimer, and myself. You might reach out to Florian
and has him specifically for help reviewing the test script. All three of us
have recently worked on test machinery in one way or another.

> Each complete squash group builds and passes the test suite on
> x86_64-pc-linux-gnu, and after stripping debug information there are
> no changes to any files outside of /usr/include (except for nscd,
> which embeds the build time, blech).  Builds used gcc 6.2.0 (current
> default for Debian unstable) and kernel headers from Linux 4.7.4
> (ditto), and were configured with --prefix=/usr --enable-obsolete-rpc.

Awesome.

-- 
Cheers,
Carlos.