Bug 400808

Could you please attach your /etc/X11/xorg.conf, /var/log/Xorg.0.log and /var/log/SaX.log seperately in plain/text format? Thanks.

I did cp /etc/X11/xorg.conf.install xorg.conf, to get X running. SaX still segfaults in the same way. Previous xorg.conf was very empty: less than two screens, and mostly empty sections. X failed because there were no devices configured.

Do you still want Xorg.0.log?

Created attachment 222469 [details]
xorg.conf

Created attachment 222470 [details]
Xorg.0.log

Created attachment 222471 [details]
SaX.log

info attached

>(II) PCI: 00:04:0: chip 1002,515e card 1002,515e rev 02 class 03,00,00 hdr 00
>(II) PCI: 04:00:0: chip 10de,01d1 card 1043,821e rev a1 class 03,00,00 hdr 00

So you have two graphics cards. 

1) ATI Technologies Inc:ES1000 --> radeon
2) nVidia Corporation:GeForce 7300 LE --> nv

This could be part of the problem. evdev driver is used for your keyboard. This could be the next problem. Could you do the following, please?

- switch to runlevel 3
- run 'sax2 -a -r'
- attach again /var/log/SaX.log

This will hopefully contain more information than the previous one like config file and log files for the native driver.

I'm afraid dual card configurations in most cases no longer work at all. You can force to configure only one card with

a) sax2 -r -c 0
b) sax2 -r -c 1

See also the output of 'sax2 -p'. You should chose the graphics card, which is
connected to your monitor. ;-) Probably the NVIDIA card.

Created attachment 223010 [details]
SaX.log from comment #7

Hmm, I was running SaX2, not sax2, but i guess that are synonyms?

sax2 -p tells me ATI is the chip: 0.

sax2 -r -c 0 seems to ... at least do something... I got black screen on both ATI and nVidia outputs.

(Config is like this: nVidia onboard, but as I really hate binary modules, ATI is installed as an add-on card. I'd like to use ATI.)

Unfortunately SaX.log is useless as before. :-( 

> (Config is like this: nVidia onboard, but as I really hate binary modules,
> ATI is installed as an add-on card. I'd like to use ATI.)

This sounds weird. ATI ES1000 is an onboard graphics chip. And GeForce 7300 a typical example for a discrete graphics card. Could you please check this, i.e.
also try 'sax2 -r -c 1'? Thanks!

sax2 -r -c 1 seems to configure the nVidia card. (I may be confused about what's onboard and what is not.)... I can see the picture etc.

Yes, I'm sorry, I was confused. sax2 -r -c 1 configures the nvidia card okay.

Ok. I'll try to reproduce the issue to make sure the support for ATI ES1000 still works. I have a ATI ES1000 PCI card (test sample) and a GeForce 7300 PCIe card. I will configure the NVIDIA PCIe card as primary card. So this should be comparable setup.

(alternatively, you can get remote access to affected machine.)

I can reproduce this issue.

# sax2 -p
Chip: 0  is -> NVidia GeForce 7300 GS          01:00:0 0x10de 0x01df AGP nv
Chip: 1  is -> ATI ES1000 515E                 06:00:0 0x1002 0x515e AGP radeon

# sax2 -r -c 0
< configuring NVIDIA card (primary card) works fine >

# sax2 -r -c 1
< configuring ATI (secondary card) hangs>
(**) RADEON(1): RADEONRestoreMemMapRegisters() : 
(**) RADEON(1):   MC_FB_LOCATION   : 0x881d803c
(**) RADEON(1):   MC_AGP_LOCATION  : 0xffffffc0
(**) RADEON(1):   Map Changed ! Applying ...
(**) RADEON(1): FIFO timed out: 12 entries, stat=0x4668500c
(EE) RADEON(1): FIFO timed out, resetting engine...
(**) RADEON(1): DC flush timeout: cd8ab5ce
(**) RADEON(1): EngineRestore (32/32)
(**) RADEON(1): FIFO timed out: 12 entries, stat=0x4668500c
(EE) RADEON(1): FIFO timed out, resetting engine...
(**) RADEON(1): DC flush timeout: cd8ab5ce
(**) RADEON(1): EngineRestore (32/32)
(**) RADEON(1): FIFO timed out: 12 entries, stat=0x4668500c
(EE) RADEON(1): FIFO timed out, resetting engine...
(**) RADEON(1): DC flush timeout: cd8ab5ce
(**) RADEON(1): EngineRestore (32/32)
(**) RADEON(1): FIFO timed out: 12 entries, stat=0x4668500c
(EE) RADEON(1): FIFO timed out, resetting engine...

# sax2 -r 
< dual card configuration with NVIDIA (primary) and ATI (secondary) hangs
  as well >

BIOS Setup: ATI (primary) / NVIDIA (secondary)

# sax2 -p
Chip: 0  is -> NVidia GeForce 7300 GS          01:00:0 0x10de 0x01df AGP nv
Chip: 1  is -> ATI ES1000 515E                 06:00:0 0x1002 0x515e AGP radeon

# sax2 -r -c 1 
< configuring ATI card (primary) works fine >

# sax2 -r -c 0
< configuring NVIDIA (secondary card) hangs>
SaX: startup

xc: sorry could not start configuration server
xc: for details refer to the log file:

    /var/log/SaX.log

xc: abort...
        (II) Setting vga for screen 0.
        (II) Loading sub module "int10"
        (II) LoadModule: "int10"
        (II) Loading /usr/lib/xorg/modules//libint10.so
        (II) Module int10: vendor="X.Org Foundation"
        compiled for 1.4.0.90, module version = 1.0.0
        ABI class: X.Org Video Driver, version 2.0
        (II) NV(0): Initializing int10
        (II) Attempted to read BIOS 128KB from /sys/bus/pci/devices/0000:01:00.0/rom: got 63KB

# sax2 -r -a
SaX: initializing please wait...
SaX: your current configuration will not be read in

SaX: no X-Server is running
SaX: will start own server if needed
*** buffer overflow detected ***: /usr/sbin/sysp terminated
======= Backtrace: =========
/lib/libc.so.6(__fortify_fail+0x48)[0xb7b794c8]
/lib/libc.so.6[0xb7b77500]
/lib/libc.so.6[0xb7b7679d]
/usr/sbin/sysp[0x8081038]
/usr/sbin/sysp[0x80813e7]
/usr/sbin/sysp[0x806e416]
/usr/sbin/sysp[0x8064610]
/usr/sbin/sysp[0x80506fc]
/usr/sbin/sysp[0x805117a]
/lib/libc.so.6(__libc_start_main+0xe5)[0xb7aac5f5]
/usr/sbin/sysp(__gxx_personality_v0+0x121)[0x804ba11]
======= Memory map: ========
08048000-08096000 r-xp 00000000 08:05 443878     /usr/sbin/sysp
08096000-08097000 r--p 0004d000 08:05 443878     /usr/sbin/sysp

So this is comparable to your case.

So basically the conclusion is, that nv/radeon driver don't work together.

The only way to fix this I can see, is to always only configure the primary card, i.e. remove Multicard support from SaX2. I try to do this for openSUSE 11.1/SLE11. See also Bug #381644. Therefore I set this to enhancement now with high priority.

Yes, you can reproduce it correctly.

I don't know what sysp is, or why is it corrupting its own memory, but it does not seem to be part of X, right?

sysp is part of SaX2. Still the radeon driver doesn't work, if it's not the primary card. So there's no use in debugging this I think.

*** This bug has been marked as a duplicate of bug 381644 ***