[Openchrome-users] K8M890 testing needed!!!

[bob davis]

> 
> Please check out the latest changes from svn.  Thomas has committed a
> bunch of fixes for VT switching to trunk, that I have merged over to the
> VT3336 branch.
> 
> If you are feeling very adventurous you should try and back out the
> attached patch (patch -R ) and see if VT switching works without it.


Hi Jon

I updated to 257 and have same results.

(!!) (development build, at svn revision 257M)

1. VT switching still is messing up. I had to look up on the web to 
realize that VT switching is using <ctrl><alt><f8> to switch back to 
console mode. To be clear its only the console mode(<alt><f1-f6>) that 
is graphically messed up. When I switch back to gui(<alt><f7>) 
everything is fine. So that would mean that the function VIALeaveVT is 
the one screwing up?

2. xv still locking. It was better when I was getting BadAlloc msgs. 
Only because it didnt crash my system and gave me more info.

I also removed the patch and there was no difference in action. I had to 
remove the patch by hand because it was for a different version. I 
basically commented out VIASaveScreen() in the function VIAEnterVT. Also 
commented out the block beginning with if (pVia->NoAccel).

I noticed gabriel mansi modified his kernel agpgart module. Is this 
something I would have to do? I can see that the agpgart module is 
loaded but I also see the errors from drm in dmesg "reclaim buffers 
locked deadlock".

Here is xorg.0.log when I <ctrl><alt><f8>

(II) VIA(0): VIALeaveVT
(II) VIA(0): ViaCursorStore
(II) VIA(0): VIARestore
(II) VIA(0): ViaLCDPower: On.
(II) VIA(0): ViaDisablePrimaryFIFO
(II) Open ACPI successful (/var/run/acpid.socket)
(II) AIGLX: Resuming AIGLX clients after VT switch
(II) VIA(0): VIAEnterVT
(II) VIA(0): VIASave
(II) VIA(0): Primary
(II) VIA(0): Unlocking...
(II) VIA(0): Crtc...
(II) VIA(0): Trying VBE Mode 1280x800 (0xc1b8) Refresh 59.99:
(II) VIA(0): ViaVbeSetRefresh
(II) VIA(0): Active Device: 2
(II) VIA(0): Refresh Rate Index: 0
(II) VIA(0): VBESetVBEMode failed...but worked OK without customized 
refresh and dotclock.
(II) VIA(0): ViaCursorRestore
(II) VIA(0): VIALoadPalette

thanks
bob