There were fixes in 3.17 that were supposed to help this, but apparently they didn't help enough. See https://bugs.freedesktop.org/show_bug.cgi?id=72180 -- basically some sort of card setup failure on our part is causing our ctxsw to die, but nvidia's appears to be more resilient to the screwups. On Fri, Dec 5, 2014 at 1:16 PM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote:> Well, I downgraded back to 2 monitors and it turns out that this problem has reappeared when using only 2 monitors as well. There must have been a regression of some sort somewhere along the line, possibly in one of the latest kernels? What can I provide to help this issue get fixed? > > -Rob > >> On Dec 5, 2014, at 8:07 AM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >> >> Hi, >> >> Using 3 monitors on the GeForce GTX 650 Ti card (2 DVI and 1 VGA) works great out of the box on Fedora 20. However, every couple of days the graphics driver seems to crash. The same thing used to happen when 2 monitors were in use, but that has pretty much stabilized as of late. Does anyone have a stable setup with this card using 3 monitors? Will 3 monitors be supported? >> >> Thanks, >> Rob >> >> $ uname -r >> 3.17.4-200.fc20.x86_64 >> >> $ lspci | grep Ti >> 01:00.0 VGA compatible controller: NVIDIA Corporation GK106 [GeForce GTX 650 Ti] (rev a1) >> >> $ dmesg >> [141447.170973] traps: polkitd[9707] general protection ip:7f383508c022 sp:7fffdd1ac110 error:0 in libmozjs-17.0.so[7f3834f4d000+3b3000] >> [141869.521246] nouveau E[ PFIFO][0000:01:00.0] SCHED_ERROR [ CTXSW_TIMEOUT ] >> [141869.521258] nouveau E[ PFIFO][0000:01:00.0] PGRAPH engine fault on channel 4, recovering... >> [142044.384110] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >> [142059.381836] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >> [142059.381933] nouveau E[ PFIFO][0000:01:00.0] read fault at 0x0000023000 [PTE] from PBDMA0/HOST_CPU on channel 0x007f966000 [unknown] >> [142074.437553] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >> [142089.435278] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >> [142114.092541] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >> [142129.090266] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >> [142145.229819] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142160.227546] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142175.225272] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142190.222999] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142205.220724] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142220.218450] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> [142235.216176] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] > > _______________________________________________ > Nouveau mailing list > Nouveau at lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/nouveau
FYI, I’ve downgraded to kernel 3.16.7-200.fc20.x86_64 and have not had any lockups for several days. So I guess this means I’ll stay at 3.16.7-200.fc20.x86_64 for the time being. Whatever fixes happened in 3.17 seemed to have made things worse, at least for me. -Rob> On Dec 5, 2014, at 6:32 PM, Ilia Mirkin <imirkin at alum.mit.edu> wrote: > > There were fixes in 3.17 that were supposed to help this, but > apparently they didn't help enough. See > https://bugs.freedesktop.org/show_bug.cgi?id=72180 -- basically some > sort of card setup failure on our part is causing our ctxsw to die, > but nvidia's appears to be more resilient to the screwups. > > On Fri, Dec 5, 2014 at 1:16 PM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >> Well, I downgraded back to 2 monitors and it turns out that this problem has reappeared when using only 2 monitors as well. There must have been a regression of some sort somewhere along the line, possibly in one of the latest kernels? What can I provide to help this issue get fixed? >> >> -Rob >> >>> On Dec 5, 2014, at 8:07 AM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >>> >>> Hi, >>> >>> Using 3 monitors on the GeForce GTX 650 Ti card (2 DVI and 1 VGA) works great out of the box on Fedora 20. However, every couple of days the graphics driver seems to crash. The same thing used to happen when 2 monitors were in use, but that has pretty much stabilized as of late. Does anyone have a stable setup with this card using 3 monitors? Will 3 monitors be supported? >>> >>> Thanks, >>> Rob >>> >>> $ uname -r >>> 3.17.4-200.fc20.x86_64 >>> >>> $ lspci | grep Ti >>> 01:00.0 VGA compatible controller: NVIDIA Corporation GK106 [GeForce GTX 650 Ti] (rev a1) >>> >>> $ dmesg >>> [141447.170973] traps: polkitd[9707] general protection ip:7f383508c022 sp:7fffdd1ac110 error:0 in libmozjs-17.0.so[7f3834f4d000+3b3000] >>> [141869.521246] nouveau E[ PFIFO][0000:01:00.0] SCHED_ERROR [ CTXSW_TIMEOUT ] >>> [141869.521258] nouveau E[ PFIFO][0000:01:00.0] PGRAPH engine fault on channel 4, recovering... >>> [142044.384110] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >>> [142059.381836] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >>> [142059.381933] nouveau E[ PFIFO][0000:01:00.0] read fault at 0x0000023000 [PTE] from PBDMA0/HOST_CPU on channel 0x007f966000 [unknown] >>> [142074.437553] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >>> [142089.435278] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >>> [142114.092541] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >>> [142129.090266] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >>> [142145.229819] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142160.227546] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142175.225272] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142190.222999] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142205.220724] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142220.218450] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> [142235.216176] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >> >> _______________________________________________ >> Nouveau mailing list >> Nouveau at lists.freedesktop.org >> http://lists.freedesktop.org/mailman/listinfo/nouveau
I’ve had lockups on 3.16.7-200.fc20.x86_64 as well, although they seem to occur somewhat less frequently. Is there work being done on this issue that may fix the bug in a newer kernel? (I would like to avoid proprietary drivers.) Thanks, Rob> On Dec 10, 2014, at 9:00 AM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: > > FYI, I’ve downgraded to kernel 3.16.7-200.fc20.x86_64 and have not had any lockups for several days. So I guess this means I’ll stay at 3.16.7-200.fc20.x86_64 for the time being. Whatever fixes happened in 3.17 seemed to have made things worse, at least for me. > > -Rob > >> On Dec 5, 2014, at 6:32 PM, Ilia Mirkin <imirkin at alum.mit.edu> wrote: >> >> There were fixes in 3.17 that were supposed to help this, but >> apparently they didn't help enough. See >> https://bugs.freedesktop.org/show_bug.cgi?id=72180 -- basically some >> sort of card setup failure on our part is causing our ctxsw to die, >> but nvidia's appears to be more resilient to the screwups. >> >> On Fri, Dec 5, 2014 at 1:16 PM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >>> Well, I downgraded back to 2 monitors and it turns out that this problem has reappeared when using only 2 monitors as well. There must have been a regression of some sort somewhere along the line, possibly in one of the latest kernels? What can I provide to help this issue get fixed? >>> >>> -Rob >>> >>>> On Dec 5, 2014, at 8:07 AM, Rob Jansen <rob.g.jansen at nrl.navy.mil> wrote: >>>> >>>> Hi, >>>> >>>> Using 3 monitors on the GeForce GTX 650 Ti card (2 DVI and 1 VGA) works great out of the box on Fedora 20. However, every couple of days the graphics driver seems to crash. The same thing used to happen when 2 monitors were in use, but that has pretty much stabilized as of late. Does anyone have a stable setup with this card using 3 monitors? Will 3 monitors be supported? >>>> >>>> Thanks, >>>> Rob >>>> >>>> $ uname -r >>>> 3.17.4-200.fc20.x86_64 >>>> >>>> $ lspci | grep Ti >>>> 01:00.0 VGA compatible controller: NVIDIA Corporation GK106 [GeForce GTX 650 Ti] (rev a1) >>>> >>>> $ dmesg >>>> [141447.170973] traps: polkitd[9707] general protection ip:7f383508c022 sp:7fffdd1ac110 error:0 in libmozjs-17.0.so[7f3834f4d000+3b3000] >>>> [141869.521246] nouveau E[ PFIFO][0000:01:00.0] SCHED_ERROR [ CTXSW_TIMEOUT ] >>>> [141869.521258] nouveau E[ PFIFO][0000:01:00.0] PGRAPH engine fault on channel 4, recovering... >>>> [142044.384110] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >>>> [142059.381836] nouveau E[Xorg[1058]] failed to idle channel 0xcccc0000 [Xorg[1058]] >>>> [142059.381933] nouveau E[ PFIFO][0000:01:00.0] read fault at 0x0000023000 [PTE] from PBDMA0/HOST_CPU on channel 0x007f966000 [unknown] >>>> [142074.437553] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >>>> [142089.435278] nouveau E[chrome[2451]] failed to idle channel 0xcccc0000 [chrome[2451]] >>>> [142114.092541] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >>>> [142129.090266] nouveau E[gnome-shell[1916]] failed to idle channel 0xcccc0000 [gnome-shell[1916]] >>>> [142145.229819] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142160.227546] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142175.225272] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142190.222999] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142205.220724] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142220.218450] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>>> [142235.216176] nouveau E[gnome-session-c[16010]] failed to idle channel 0xcccc0000 [gnome-session-c[16010]] >>> >>> _______________________________________________ >>> Nouveau mailing list >>> Nouveau at lists.freedesktop.org >>> http://lists.freedesktop.org/mailman/listinfo/nouveau >