public inbox for isar-users@googlegroups.com
 help / color / mirror / Atom feed
From: Henning Schild <henning.schild@siemens.com>
To: Christopher Larson <kergoth@gmail.com>
Cc: Alexander Smirnov <asmirnov@ilbers.de>, <isar-users@googlegroups.com>
Subject: Re: [PATCH] dpkg.bbclass: add lockfile for do_build
Date: Thu, 11 Jan 2018 18:20:10 +0100	[thread overview]
Message-ID: <20180111182010.6d503748@mmd1pvb1c.ad001.siemens.net> (raw)
In-Reply-To: <CABcZAN=5wr=EbQCFPzipwMgiA2QGGb0bTG9x82dB7-H3a5AE_A@mail.gmail.com>

Am Thu, 11 Jan 2018 08:20:54 -0700
schrieb Christopher Larson <kergoth@gmail.com>:

> On Thu, Jan 11, 2018 at 5:14 AM, Henning Schild
> <henning.schild@siemens.com> wrote:
> 
> > Am Thu, 11 Jan 2018 13:31:45 +0300
> > schrieb Alexander Smirnov <asmirnov@ilbers.de>:
> >  
> > > Hello Christopher,
> > >
> > > thank you for this catch!
> > >
> > > On 01/10/2018 07:04 PM, Christopher Larson wrote:  
> > > > All recipes using this class will run build.sh in the same
> > > > chroot, and build.sh runs apt-get to install deps, so the build
> > > > can fail due to dpkg being unable to acquire its lock. Add a
> > > > bitbake lockfile to ensure that only one do_build task for
> > > > recipes inheriting this class runs at a time.  
> > >
> > > Serialization of do_build tasks leads to whole Isar serialization,
> > > what is significant issue if you have lots of packages to build.
> > > So I think the "build" and "deps installation" should be split in
> > > separate tasks and only the second one should be protected by
> > > locks.  
> >
> > The fact that Isar uses multiconfig to build many targets from the
> > same tree is not intuitive, we discussed this before. So i am not
> > surprised new contributors get that "wrong". Alex i suggest you
> > split the task and adopt Christophers changes.
> >
> > Christopher, have a look at scripts/ci_build.sh and you will see
> > that Isar is capable of building multiple images at once in the
> > same tree. If you are building 10 images you will have 10
> > buildchroots that have to be serialized seperately. A feature that
> > probably nobody but the CI-bot will ever use ;).  
> 
> 
> Fair enough, thanks, I’ll re-test with 1) a lockfile in the
> buildchroot or 2) having build.sh wait on a lock itself, if possible.

I just read your signature, and now i guess you can handle the
complexity of multiconf without Alex helping you out. Watch out for
changes to that code in current next.

Henning

      reply	other threads:[~2018-01-11 17:20 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-10 16:04 Christopher Larson
2018-01-11 10:31 ` Alexander Smirnov
2018-01-11 12:14   ` Henning Schild
2018-01-11 15:20     ` Christopher Larson
2018-01-11 17:20       ` Henning Schild [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180111182010.6d503748@mmd1pvb1c.ad001.siemens.net \
    --to=henning.schild@siemens.com \
    --cc=asmirnov@ilbers.de \
    --cc=isar-users@googlegroups.com \
    --cc=kergoth@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox