Re: after upgrading fedora rawhide this morning, no graphical desktop

On Thu, 13 Mar 2014, Kevin Martin wrote: > On 03/13/2014 07:57 AM, Robert P. J. Day wrote: > > > > recently, i upgraded my ASUS G74S laptop to fedora rawhide and it > > was running nicely. then this morning, i did another "yum update", > > which appeared to update well over 200 packages (including a slightly > > newer kernel), after which, when i booted, i had no graphical desktop > > anymore, just the little blue and white fedora logo. > > > > i can still switch to VC2 and log in at the command line (where i am > > now), so i can certainly check log files, but i don't see anything > > immediately amiss. > > > > i rebooted both to the earlier rawhide kernel, and even back to the > > latest fedora 20 official kernel -- same result, the fedora logo in > > the middle of the screen on VC1, but the ability to log in on another > > virtual console. > > > > has anyone else run into this? i have an nvidia graphics card, and > > am running the nouveau driver. i'll keep poking around the log files, > > and if you have any suggestions, i'm all ears. > > > > rday > > > Hmm, sounds similar to what I'm experiencing. When you go into VC2 > what does "lsmod | grep nouveau" show? I've found that I've been > having to manually "modprobe nouveau modeset=1" since doing my > update about 4 days ago. I'm not sure why nouveau won't load and I > find that if I don't set the modeset=1 when I do the manual modprobe > that I still can't get X. here's mine: nouveau 1087913 2 mxm_wmi 12865 1 nouveau i2c_algo_bit 13257 1 nouveau drm_kms_helper 50287 1 nouveau ttm 80479 1 nouveau drm 283937 4 ttm,drm_kms_helper,nouveau i2c_core 38476 6 drm,i2c_i801,drm_kms_helper,i2c_algo_bit,nouveau,videodev wmi 18804 3 mxm_wmi,nouveau,asus_wmi video 19206 2 nouveau,asus_wmi i could, i guess, manually unload all modules dependent on nouveau, or just reboot and add "nouveau.modeset=1" at the kernel command line, right? i also checked /sys/module/nouveau/parameters/modeset, which currently has a value of -1, which i don't understand. still, if it's a module issue, why would rebooting under an earlier, working kernel not have fixed this?