On Tue, 29 Oct 2013, Daniel Vetter wrote:
Oh right, I've forgotten that between the review and writing the mail ;-) I guess we could try to bend the stable rules a bit and just submit all 6. It's a regression fix after all, and at least personally I prefer the most minimal backports to avoid diverging between upstream and stable kernel branches.
But I guess that's Dave's call to make.
How about taking the patch series into drm-next and marking patch #6 only with CC: stable@ and specifying hashes of the prerequisite 5 patches for cherry-pick?
I think that is fully compliant with stable rules (at least that's my understanding of line 41 of Documentation/stable_kernel_rules.txt) and would not be bending anything.
I presume that if this is acceptable, Dave would have to add stable@ tags for cherry-picking because hash values may change in his tree if drm-next changes relative to my base (which is state of drm-next as of yesterday PM).
-- Ilija