fido status

I just wanted to let people know what is going on with Fido.

Joshua has had a branch in testing for a while. The autobuilder gave mixed feedback for it. Partly, there was a problem since we'd introduced
a new autobuilder (Fedora 22) and this introduced some problems with subversion-native and gcc5. Partly there were also networking issues on
the autobuilders related to runqemu.

Thanks to some help from Michael, we've tracked down and sorted the networking issue (we hope!). I therefore took the decision to merge Joshua's fido-next branch myself, now, with the addition of the subversion fixes to the fido branch.

I've triggered another build which is underway at the moment and I'm confident the result should be reasonable. What I am expecting is failures for oe-selftest and for nightly-aarm64 and nightly-mips64. There are fixes just merged into master which resolve issues in all three areas, I'm hoping we can queue up a fido-next with those in and get fido building green quickly. Those issues are less of an issue given
we know what the issues are and they are testing issues, not 'real' bugs.

Quoted 10 lines · Richard Purdie

I just wanted to let people know what is going on with Fido.

Joshua has had a branch in testing for a while. The autobuilder gave mixed feedback for it. Partly, there was a problem since we'd introduced
a new autobuilder (Fedora 22) and this introduced some problems with subversion-native and gcc5. Partly there were also networking issues on
the autobuilders related to runqemu.

Quoted 4 lines

Thanks to some help from Michael, we've tracked down and sorted the networking issue (we hope!). I therefore took the decision to merge Joshua's fido-next branch myself, now, with the addition of the subversion fixes to the fido branch.

Thanks for merging Richard, the number of patches was starting to get unwieldy so I'm pleased to see this happen.

Quoted 10 lines

I've triggered another build which is underway at the moment and I'm confident the result should be reasonable. What I am expecting is failures for oe-selftest and for nightly-aarm64 and nightly-mips64. There are fixes just merged into master which resolve issues in all three areas, I'm hoping we can queue up a fido-next with those in and get fido building green quickly. Those issues are less of an issue given
we know what the issues are and they are testing issues, not 'real' bugs.

Your prediction was very accurate. I've merged a bunch of fixes for the issues you highlighted (and a couple of others) into my fido-next branch[1]. I shall endeavour to get this tested on the autobuilder and submitted for merge over the next couple of days.

Thanks again,

Joshua

1. http://cgit.openembedded.org/openembedded-core
-contrib/log/?h=joshuagl/fido-next
--
_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.openembedded.org/mailman/listinfo/openembedded-core

Quoted 42 lines · Joshua Lock

Quoted 13 lines · Richard Purdie

I just wanted to let people know what is going on with Fido.

Joshua has had a branch in testing for a while. The autobuilder gave
mixed feedback for it. Partly, there was a problem since we'd introduced
a new autobuilder (Fedora 22) and this introduced some problems with
subversion-native and gcc5. Partly there were also networking issues
on
the autobuilders related to runqemu.

Quoted 6 lines

Thanks to some help from Michael, we've tracked down and sorted the
networking issue (we hope!). I therefore took the decision to merge
Joshua's fido-next branch myself, now, with the addition of the subversion fixes to the fido branch.

Thanks for merging Richard, the number of patches was starting to get
unwieldy so I'm pleased to see this happen.

Quoted 16 lines

I've triggered another build which is underway at the moment and I'm
confident the result should be reasonable. What I am expecting is failures for oe-selftest and for nightly-aarm64 and nightly -mips64.
There are fixes just merged into master which resolve issues in all
three areas, I'm hoping we can queue up a fido-next with those in and
get fido building green quickly. Those issues are less of an issue
given
we know what the issues are and they are testing issues, not 'real'
bugs.

Quoted 6 lines

Your prediction was very accurate. I've merged a bunch of fixes for the
issues you highlighted (and a couple of others) into my fido-next branch[1]. I shall endeavour to get this tested on the autobuilder and
submitted for merge over the next couple of days.

One minor comment on the patches already merged. This change:

http://git.openembedded.org/openembedded
-core/commit/?h=fido&id=f46c2ac0fd05f0e10258d8f03bfb2ebd0c2bcbe3

was reported to cause problems for multilib [1]. It needs this follow

(quoted diff) +0 -2

-on fix:
 
  http://git.openembedded.org/openembedded-core/commit/?h=master
-next&id=78a0e916882a747c267808c08ab8bc615198b5a8

[1] http://lists.openembedded.org/pipermail/openembedded-core/2015
-May/105351.html

Quoted 9 lines · Andre McCurdy

One minor comment on the patches already merged. This change:

http://git.openembedded.org/openembedded
-core/commit/?h=fido&id=f46c2ac0fd05f0e10258d8f03bfb2ebd0c2bcbe3

was reported to cause problems for multilib [1]. It needs this follow -on fix:

Quoted 5 lines

http://git.openembedded.org/openembedded-core/commit/?h=master
-next&id=78a0e916882a747c267808c08ab8bc615198b5a8

[1] http://lists.openembedded.org/pipermail/openembedded-core/2015
-May/105351.html

Thanks for raising this Andre - I've included the above change in my fido-next branch[1].

Regards,

Joshua

1. http://cgit.openembedded.org/openembedded-core
-contrib/log/?h=joshuagl/fido-next