All of lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot@busybox.net
Subject: [Buildroot] feature or pebkac
Date: Fri, 30 Aug 2013 09:58:31 +0200	[thread overview]
Message-ID: <20130830095831.12c3c743@skate> (raw)
In-Reply-To: <521F9B91.7040903@mind.be>

Dear Arnout Vandecappelle,

On Thu, 29 Aug 2013 21:05:53 +0200, Arnout Vandecappelle wrote:
> On 08/29/13 20:17, rh wrote:
> > I have found that changing the url for a custom kernel tar ball does not
> > trigger a rebuild when running make.  I have removed linux-custom dir
> > and then make fetches the new tar ball and rebuilds.
> 
>   That's right. Buildroot doesn't try to be perfect for rebuilds when the 
> configuration changes - but if it is simple to fix, patches are welcome!

As Thomas De Schampheleire later explained, the problem is that for
custom tarballs, we can't easily find a directory name that's unique
for each tarball, except by having a name that is a complete
translation of the tarball URL, such as:

output/build/linux-http-www.company.com-pub-linux-3.10-board-foo.tar.gz/

That's more or less what Thomas DS proposed, and I wasn't too happy
with it due to the horrible directory name. Another option would be to
hash the tarball location and do:

output/build/linux-${hash}/

but then it's difficult for the user to map which linux-<foo> directory
corresponds to which version he has built.

> > But I don't need all that and I now use --single-branch argument
> > to git clone followed by git archive to create a tar ball.
> 
>   That's a great idea! However, this option is only available since git 
> 1.7.10 and I think that many production environments still have older git 
> versions.

Can't we do like what we do for shallow clones? Try if --single-branch
is supported and works, and if so, all right, if not, fall back to the
normal way?

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com

  parent reply	other threads:[~2013-08-30  7:58 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20130829111708.c603e5bf976541d3abd3a43a@lavabit.com>
2013-08-29 19:05 ` [Buildroot] feature or pebkac Arnout Vandecappelle
2013-08-29 20:25   ` Thomas De Schampheleire
     [not found]     ` <20130829153937.f53bdce09f0a855807dd0a5d@lavabit.com>
2013-08-30  7:35       ` Thomas De Schampheleire
2013-08-30  7:58   ` Thomas Petazzoni [this message]
2013-08-30  8:22     ` Arnout Vandecappelle
2013-08-30  8:30       ` Thomas Petazzoni
2013-08-30 12:45       ` Peter Korsgaard

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=20130830095831.12c3c743@skate \
    --to=thomas.petazzoni@free-electrons.com \
    --cc=buildroot@busybox.net \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.