Re: State of bitbake world, Failed tasks 2014-04-28

[Martin Jansa <martin.jansa@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1476 bytes --]

On Mon, Apr 28, 2014 at 04:50:29AM +0200, Martin Jansa wrote:
> Be aware that qemux86-64 log is there, but it wasn't complete so it
> isn't included in this report.
> 
> It isn't complete, because with this number of failed recipes rm_work
> wasn't able to keep tmpdir small enough for my 55G tmpfs, biggest chunk
> was eaten by chromium and mongodb:
> 593M    tmp-eglibc/work/qemux86_64-oe-linux/linux-yocto-tiny-kexecboot
> 6.8G    tmp-eglibc/work/core2-64-oe-linux/mongodb
> 20G     tmp-eglibc/work/core2-64-oe-linux/chromium
> 
> Khem sent couple of patches for B!=S issues, bit thanks to him!
> 
> I have local patch to set B=S/autotools-brokensep for remaining failing
> recipes, but I'll wait a bit more for people to send fixes.

Be aware that failures in this report are just small part of whole B!=S
issues. After setting B=S in them more issues were found in recipes
which were "hidden" before because of failing dependency in first run.

Now I'm in 3rd round of adding B=S and with ~180 recipes having this it
still fails in couple newly revealed ones - I'll send the list when it's
complete (and list of failing recipes is the same as last report before
B!=S or better)

It sounds terrible, but changing the default to B!=S was good thing, we
can easily work-around it by B=S/autotools-brokensep and we'll get the
advantages of it at least in recipes which aren't broken.

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]