Skip to contentSkip to conversation

Fido and dizzy collaboration?

DISCUSSION Richard Purdie 3 participants 4 messages Sep 1, 2015 – Sep 4, 2015

lore
Share & tools
lore
#0 lore raw

Hi Armin, Joshua,

I've just been looking at the patches queued for the dizzy/fido branches. In fact I ended up merging them before I realised there are some issues.

The problem is there are dizzy patches which aren't in fido, which breaks our policies. Could the two of you work to resolve that and give me fido branch to pull which resolved it please?

I didn't take the dizzy patch related to S = ${WORKDIR} since that looked more like a change appropriate to master only. The bitbake patches also need handling on the bitbake list if we really do want to backport them to 1.26 and 1.24.

Cheers,

Richard

#1 lore raw
Quoted 10 lines · Richard Purdie

Hi Armin, Joshua,

I've just been looking at the patches queued for the dizzy/fido branches. In fact I ended up merging them before I realised there are some issues.

The problem is there are dizzy patches which aren't in fido, which breaks our policies. Could the two of you work to resolve that and give
me fido branch to pull which resolved it please?

Sure, I'll start working on this today.

Regards,

Joshua

#2 lore raw
In reply to Richard Purdie

Hi Richard,

Quoted 10 lines · Richard Purdie

Hi Armin, Joshua,

I've just been looking at the patches queued for the dizzy/fido branches. In fact I ended up merging them before I realised there are some issues.

The problem is there are dizzy patches which aren't in fido, which breaks our policies. Could the two of you work to resolve that and give
me fido branch to pull which resolved it please?

I just had a look at changes merged to dizzy in the past two weeks, assuming the issues were recent, and found the following changes in dizzy which I didn't have in fido:
* Some QEMU CVE fixes, those don't apply to the newer QEMU in fido (2.2.0 vs. 2.1.0).
* A fix for CVE-2015-5477 in bind, fido has a different patch to include this - probably not worth a revert & merge though?
* Three lib/oeqa changes - I've cherry-picked them to my joshuagl/fido -next branch[1].

In future I'll keep an eye on changes tagged for dizzy too and try to help avoid this happening again.

Regards,

Joshua

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

#3 lore raw
In reply to Richard Purdie
Quoted 8 lines · Richard Purdie

Hi Armin, Joshua,

I've just been looking at the patches queued for the dizzy/fido branches. In fact I ended up merging them before I realised there are some issues.

The problem is there are dizzy patches which aren't in fido, which breaks our policies.

I thought the changes had to be in Master first. Are you saying I need to wait for them to be in fido first?

Could the two of you work to resolve that and give

Quoted 1 line

me fido branch to pull which resolved it please?

Will do.

- armin

Quoted 10 lines

I didn't take the dizzy patch related to S = ${WORKDIR} since that looked more like a change appropriate to master only. The bitbake patches also need handling on the bitbake list if we really do want to backport them to 1.26 and 1.24.

Cheers,

Richard