Hi, On 06-07-16 19:44, Adam Jackson wrote:
I've merged some fixes back from master to the 1.18 branch, and plan to release 1.18.4 in the next few days. Give a shout if there's anything wrong on the branch, or if there's any other changes that should be picked up.
I'm not sure the "modesetting: Load on GPU-s with 0 outputs" patch is actually useful (it cannot harm, but not sure it is useful, see the mail titled "modesetting driver does not support being an offload source, should not advertise cap ?"). I got fooled by xrandr --listproviders not listing the nvidia gpu on a laptop were all outputs are permanently connected the intel gpu, but DRI_PRIME=1 likely would have worked without this change (*), as the provider stuff is not necessary for DRI_PRIME-1 with DRI3, and it seems that the modesetting actually cannot do render offloading with DRI2 (cannot work as an offload source). Depending on the outcome of the discussion surrounding this we may be better of reverting the "modesetting: Load on GPU-s with 0 outputs" patch on both master and the 1.18 branch. OTOH I think that the 24bpp packed fixes are useful, given how long we've been carrying those in RHEL, I think they would make a good addition to the 1.18 branch. Regards, Hans *) I've not checked this yet _______________________________________________ [email protected]: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: https://lists.x.org/mailman/listinfo/xorg-devel
