From: <Aiv...@un...> - 2004-02-10 12:06:59
|
>I bought a NVIDIA RIVA TNT2 Model 64 PCI Card >and set up my XF86Config-4 (see attachment). >My first card is still a 5600XT. > >I tried with and without echo "1" > ...hackvideo On the right way. >I'm running the 2.6.0-ruby debian source package, with the debian 4.3 >XServer. Nvidia driver 5336. > ># cat /proc/bus/console/*/* >isa0060/serio0/input0 >usb-0000:00:02.2-1.4/input0 > >I created links X0 and X1 to /usr/X11R6/bin/XFree86 >My startup commands are: >X1 :1 -layout sesam1 vt17 >X0 :0 -layout sesam0 vt7 > >If I'm starting X1 I got something like a freeze: on the console I get >the ouput of the start then the console freezes (but I can reeboot). >remote admin is working,too. xf86 blocks VGA output. Manualy start second xf86 sever You can only from xterm. Also gdm desktop manager can do this work. >The log of a test with X1 :0 -layout sesam1 vt17 is attached. >My configuration works with Xinerama flawlessly. > Aivils Stoss |
From: Peter S. <pet...@gm...> - 2004-02-10 15:09:50
|
Aiv...@un... wrote: > xf86 blocks VGA output. Manualy start second xf86 sever You can > only from xterm. Also gdm desktop manager can do this work. > > Nope. Again if I'm trying to start a server on the second card (gdm or xterm) the Xserver init seems to stop (see log). The started process ("X1 :2 -layout sesam1 vt17") can't be killed. The IRQs of the cards are different: CPU0 0: 75535198 XT-PIC timer 1: 12217 XT-PIC i8042 2: 0 XT-PIC cascade 3: 3044317 XT-PIC eth0 4: 547 XT-PIC eth1 5: 39153 XT-PIC ohci_hcd 8: 1 XT-PIC rtc 10: 5600 XT-PIC SiS SI7012, nvidia 11: 5939092 XT-PIC ide2, ide3, nvidia 12: 163 XT-PIC ohci_hcd, saa7146 (0) 14: 260451 XT-PIC ide0 15: 30 XT-PIC ide1 NMI: 0 ERR: 0 Has anyone any luck with getting two nvidia cards with the 5336 driver up and running? I have actually no clue where to start the debugging. Could 2.6.2-ruby solve my problem? Peter |
From: Peter S. <pet...@gm...> - 2004-02-11 09:40:54
|
Peter Schrammel wrote: > Aiv...@un... wrote: > >> xf86 blocks VGA output. Manualy start second xf86 sever You can >> only from xterm. Also gdm desktop manager can do this work. >> >> > Nope. Again if I'm trying to start a server on the second card (gdm or > xterm) the Xserver init seems to stop (see log). The started process > ("X1 :2 -layout sesam1 vt17") can't be killed. > > The IRQs of the cards are different: > CPU0 > 0: 75535198 XT-PIC timer > 1: 12217 XT-PIC i8042 > 2: 0 XT-PIC cascade > 3: 3044317 XT-PIC eth0 > 4: 547 XT-PIC eth1 > 5: 39153 XT-PIC ohci_hcd > 8: 1 XT-PIC rtc > 10: 5600 XT-PIC SiS SI7012, nvidia > 11: 5939092 XT-PIC ide2, ide3, nvidia > 12: 163 XT-PIC ohci_hcd, saa7146 (0) > 14: 260451 XT-PIC ide0 > 15: 30 XT-PIC ide1 > NMI: 0 > ERR: 0 > > Has anyone any luck with getting two nvidia cards with the 5336 driver > up and running? I have actually no clue where to start the debugging. > Could 2.6.2-ruby solve my problem? I found a hint that there could be something wrong with the nvidia driver: If I start the second X server I always get a segfault, but after that the interrupt table looks like above. The nvidia driver does not release the interrupt. If I start the second X server again it hangs. Of course. Again...damn closed source :-(( Peter |
From: Peter S. <pet...@gm...> - 2004-02-13 20:25:08
|
I think I'm giving up.... Now I'm useing the nvidia (closed source driver) for my first head, and the nvidia (open source driver) for my second head (no dri). nothing hangs!! But I have still problems: The second head doesn't wake up. The logfiles are ok. But the monitor stays dark. If I config the PCI card to the be primary head in the BIOS, it's the other way round. Could this be caused by some missing POST calls? Another thing is that the XServer with the nv driver is useing the CPU to 100%. Hm... going to buy a Matrox card... BTW: Compiled 2.6.2 kernel with ruby, no fb-support. |
From: Peter S. <pet...@gm...> - 2004-02-26 20:42:29
|
Here I go again...sorry for bothering you but perhaps it's some usefull input. I got two consoles running. But something strange happens: When gdm starts up the first X-Server everything's fine (mouse und keyboard inputs are ok) but as it starts up the second X-Server the first freezes (no mouse movement and no keyboard action). The second X-Server is fully functional. If I do a 'cat /dev/input/event0' (the keyboard of X-Server 1) and hit keys I get some output. The same with /dev/input/mouse0. It seems that the start of the second X-server locks something. BTW if I'm useing the closed source nvidia diver in X-Server 1 I get the same but with 100% CPU load. Again no clue where to go from here...thanks for your help in advance. Peter Kernel 2.6.3-ruby, XFree-4.3.0-pre (Debian) No fb. No DRI. Card 1: NVIDIA 5600XT Card 2: Matrox Mystique 2 Usb mice, 1 ps2 keyboard,1 usb-keyboard |