Openembedded Core Discussions
 help / color / mirror / Atom feed
From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: Patches and discussions about the oe-core layer
	<openembedded-core@lists.openembedded.org>
Subject: Re: Using meta-toolchain output as prebuilt toolchains
Date: Tue, 29 Mar 2011 16:32:28 +0100	[thread overview]
Message-ID: <1301412748.24596.58.camel@rex> (raw)
In-Reply-To: <4D91C35B.50708@mlbassoc.com>

On Tue, 2011-03-29 at 05:32 -0600, Gary Thomas wrote:
> On 03/28/2011 12:58 PM, Tom Rini wrote:
> > On 03/27/2011 04:44 AM, Koen Kooi wrote:
> >> What is the preferred way of 'importing' prebuilt toolchains into OE-core? In oe.dev I can use this: http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/meta/external-toolchain-angstrom.bb to 'import' an angstrom toolchain built with 'bitbake meta-toolchain', but I'm not sure how to do that in OE-core.
> >>
> >> My actual use case is actually 2 use-cases:
> >>
> >> 1) Hand people a prebuilt angstrom toolchain and migrate them to OE while keeping the toolchain
> >> 2) "Get started in 10 minutes" type of thing, I suspect an sstate mirror might be better.
> >>
> >> So any suggestions for #2 as well?
> >
> > sstate is the best way to catch #2 and I know in general there's "look
> > at this URI for the file you want" bits.  I haven't yet started kicking
> > the tires on sstate as hard as I have for packaged-staging however.
> 
> This is the approach I've been working on to pass on to my customers.  The
> idea that I can give them a packaged bundle which includes prebuilt tools
> is far more attractive than a simple poky (or oe-core plus layers) tree
> which involves many hours of building to produce a simple image.
> 
> Hopefully, the sstate method will become more stable over time.  The problem
> I've had with it recently is that there are so many variables which are used
> to decide if the sstate data is valid (see Richard's description on how this
> works) that if the smallest thing changes, the whole lot becomes useless :-(
> so it's really only valid for a [fairly] static repository.

It depends what changes your customers expect to make and whether with
those changes you consider the prebuilt data still to be valid. I agree
that the defaults do tend towards detecting even small changes. I think
over time we'll get a better feel for the things that should/shouldn't
be included in there. It is almost totally customisable and what we're
missing now is some good "policy" type data.

Cheers,

Richard





  reply	other threads:[~2011-03-29 15:34 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-27 11:44 Using meta-toolchain output as prebuilt toolchains Koen Kooi
2011-03-27 19:34 ` Khem Raj
2011-03-27 20:26   ` Koen Kooi
2011-03-28 18:58 ` Tom Rini
2011-03-29 11:32   ` Gary Thomas
2011-03-29 15:32     ` Richard Purdie [this message]
2011-03-28 19:25 ` Richard Purdie
2011-03-31 19:36   ` Denys Dmytriyenko

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=1301412748.24596.58.camel@rex \
    --to=richard.purdie@linuxfoundation.org \
    --cc=openembedded-core@lists.openembedded.org \
    /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