Is it expected that the noted meta-arm sha1 is on branch "master" and not "gatesgarth" ?
Not expected but we worked it out. There is a patch from Ross on this list which fixes the issue in yocto-autobuilder2. For the purposes of the release I'm thinking we should rerun the meta-arm tests and swap in the gatesgarth revision.
Copying Vineela/Michael so they know about the issue.