Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 20 Feb 2020 13:46:33 -0800 (PST)
From:      "Rodney W. Grimes" <freebsd@gndrsh.dnsmgr.net>
To:        Ed Maste <emaste@freebsd.org>
Cc:        Pedro Giffuni <pfg@freebsd.org>, Li-Wen Hsu <lwhsu@freebsd.org>, Xin LI <d@delphij.net>, Yuri Pankov <yuripv@yuripv.me>, src-committers <src-committers@freebsd.org>, svn-src-all <svn-src-all@freebsd.org>, svn-src-head <svn-src-head@freebsd.org>
Subject:   Re: svn commit: r358153 - head/usr.sbin/services_mkdb
Message-ID:  <202002202146.01KLkXRM076042@gndrsh.dnsmgr.net>
In-Reply-To: <CAPyFy2C=tUtZaipR5ns2SWNnST-Q0XYD9%2BCF4eqA=jgwDyRqYw@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
[ Charset UTF-8 unsupported, converting... ]
> On Thu, 20 Feb 2020 at 09:54, Pedro Giffuni <pfg@freebsd.org> wrote:
> >
> > On 2020-02-20 09:46, Ed Maste wrote:
> > > On Thu, 20 Feb 2020 at 04:56, Li-Wen Hsu <lwhsu@freebsd.org> wrote:
> > >> Please note this is not fixed in ci.freebsd.org yet. The reason is it
> > >> directly goes to src/release and perform `make packagesystem` to
> > >> generate kernel.txz and base.txz.  If I read release(7) correctly,
> > >> this should be performed under a fresh installed environment.  I am
> > >> checking how to fix this.
> > > Should we revert the change until a build fix can be committed too?
> >
> > I am OK with reverting the change, and I am rather surprised such a
> > simply change could have broken the build.
> 
> Indeed, this has uncovered a broken dependency in our build process
> (i.e., that the build host's services_mkdb is being used).
> 
> IMO we should:
> 
> 1. Revert the change for now (pfg will you take care of it?)
> 2. MFC the MAX_PROTO increase as soon as possible
> 3. Try to fix the use of host's services_mkdb
> 4. Recommit after either #3 is done, or FreeBSD 12.2 and 11.4 are released

Agree fully with items 1 to 3, as to item 4 I think needs some work:
4. Recommit services to head after #3 is done
5. MFC Fixes #2 and #3 after a short bake in ^head (1 or 2 weeks)
6. MFC of services file to stable/12&11 after #2 and #3 are shipped in 12.2 and 11.4 respectfully.

This should ensure no disruptions to anyone on stable/ happens,
though I could probably be convinced that once #3 is fixed the risk
here is low.

-- 
Rod Grimes                                                 rgrimes@freebsd.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?202002202146.01KLkXRM076042>