On Thu, Dec 29, 2016 at 14:52:08 +0000, J. Roeleveld wrote: > On December 29, 2016 3:24:27 PM GMT+01:00, Alan Mackenzie <a...@muc.de> wrote: > >Hi, Gentoo.
> >I'm having trouble with my kernel 4.4.26. It will do everything > >expected of it except for running X Windows. > >I've checked I've got the kernel configuration set up as described in > >the gentoo wiki documents, unless I've overlooked something. > >What happens is that I run "startx", and a little while later, the X > >server reports that no screens are found. > >More precisely, the following appears in Xorg.0.log: > > KAVERI, KAVERI, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, > > HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII > >[ 47.952] (II) modesetting: Driver for Modesetting Kernel Drivers: > >kms > >[ 47.952] (--) using VT number 7 > >[ 47.963] (II) [KMS] Kernel modesetting enabled. > >[ 47.963] (WW) Falling back to old probe method for modesetting > >[ 47.963] (EE) No devices detected. > >[ 47.963] (EE) > >Fatal server error: > >[ 47.963] (EE) no screens found(EE) > >[ 47.963] (EE) > >Please consult the The X.Org Foundation support > > at http://wiki.x.org > > for help. > >[ 47.963] (EE) Please also check the log file at > >"/var/log/Xorg.0.log" for additional information. > >[ 47.963] (EE) > >Note the "No devices detected". > >The corresponding piece of the working Xorg.0.log in my older kernel > >is: > > KAVERI, KAVERI, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, > > HAWAII, HAWAII, HAWAII, HAWAII, HAWAII, HAWAII > >[ 188.705] (II) modesetting: Driver for Modesetting Kernel Drivers: > >kms > >[ 188.705] (--) using VT number 7 > >[ 188.713] (II) [KMS] Kernel modesetting enabled. > >[ 188.713] (WW) Falling back to old probe method for modesetting > >[ 188.713] (WW) VGA arbiter: cannot open kernel arbiter, no > >multi-card support > >[ 188.713] (II) RADEON(0): Creating default Display subsection in > >Screen section > > "Default Screen Section" for depth/fbbpp 24/32 > >[ 188.713] (==) RADEON(0): Depth 24, (--) framebuffer bpp 32 > >[ 188.713] (II) RADEON(0): Pixel depth = 24 bits stored in 4 bytes > >(32 bpp pixmaps) > >[ 188.713] (==) RADEON(0): Default visual is TrueColor > >[ 188.713] (==) RADEON(0): RGB weight 888 > >[ 188.713] (II) RADEON(0): Using 8 bits per RGB (8 bit DAC) > >[ 188.713] (--) RADEON(0): Chipset: "ATI Radeon HD 4550" (ChipID = > >0x9540) > >Needless to say, the kernel version is the sole difference between the > >non-working 4.4.26 and the working 4.0.5. > >Would somebody please help me track down this problem. Thanks! > I ended up skipping 4.4 kernels as I had problems as well. > Not with X, but hibernate wasn't working. > 4.6 and 4.8 kernels do seem to work better. > You could try a more recent one as well. I found the problem whilst debugging the same symptoms in kernel 4.9.6. I'd disabled "Enable PCI quirk workarounds" in General Setup. After reenabling it, my X functions again. Don't ask me what brought me to suspect this setting. I mean, my PC is "relatively recent" (2009). PCI, even PCIe, is as old as the hills. Surely they got the quirks out of the chipsets by 2009? Apparently not! > -- > Joost -- Alan Mackenzie (Nuremberg, Germany).
