Am Freitag, den 11.12.2015, 17:36 +0000 schrieb Russell King - ARM Linux: > On Fri, Dec 11, 2015 at 05:15:40PM +0100, Daniel Vetter wrote: > > On Fri, Dec 11, 2015 at 10:02:45AM +0000, Russell King - ARM Linux > > wrote: > > > On Fri, Dec 11, 2015 at 04:58:08PM +1000, Dave Airlie wrote: > > > > I've seen etnaviv, rockchip(?), vc4 gpu api, can I get plans > > > > for if > > > > people would like these in now, also anything I've missed on > > > > the list. > > > > > > I would definitely like to see etnaviv make it in for the next > > > merge > > > window, but that depends on it being reviewed, and I haven't seen > > > anything from DRM people yet. > > > > > > I've queued up some of the TDA998x and Armada DRM changes (3 and > > > 5 > > > patches respectively) which I'll send you shortly if they haven't > > > already been merged via some other route. > > > > I did look at etnaviv on v1, if all the things I've raised there > > have been > > addressed (and it looks like, but no time for detailed checking): > > I did keep a list of your points, and made sure that we'd addressed > them all.  We went a little further towards the end with your 'flags' > suggestion for several of the ioctls, which I think was a very good > point you raised. > > > Acked-by: Daniel Vetter <daniel.vetter at ffwll.ch> > > Thanks! > > > For detailed review it would be best to just get some of the new > > big > > submissions to cross review I think. But personally I'd be ok with > > etnaviv going in as is, trusting that you've done plenty of review > > within your group. > > We have had a certain amount of review within our group - Christian > reviewed many of my early patches, and I've reviewed Lucas' patches. > There could have been more review. > > I was rather hoping for some review of the changes since your last > comments, especially with the locking changes.  I'm fairly confident > with the locking changes (which were particularly hairy) as I've been > running them for some time now with lockdep enabled.  The > particularly > "hairy" bit was in etnaviv_gem_get_iova(). > While I didn't review all of your patches in-depth, I think we've got the locking changes pretty well covered. The review I did there wasn't just some "I think this looks okay", but me taking the time to go through all paths I could envision and validating that your locking scheme works for them.
Regards, LucasÂ
