Bug 135439

Summary: Xorg display freezes when switching from text console to X
Product: [openSUSE] SUSE LINUX 10.0 Reporter: Helge Deller <helge.deller>
Component: X.OrgAssignee: Stefan Dirsch <sndirsch>
Status: RESOLVED WONTFIX QA Contact: Stefan Dirsch <sndirsch>
Severity: Normal    
Priority: P5 - None    
Version: Final   
Target Milestone: ---   
Hardware: HP   
OS: SuSE Linux 10.0   
Whiteboard:
Found By: Customer Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: Xorg.0.log file
New xorg.conf file
Xorg.0.log file
X11 straced
xorg config file (remove/add "dri")
new log file (with dri enabled)
strace of X

Description Helge Deller 2005-11-24 08:01:56 UTC
Using SUSE 10.0 (final) on an HP Pavillon ze4423EA laptop with ATI IGP320 chipset, Xorg freezes when switching from a text console (e.g. cntrl-alt-F1) back to the X11 screen (cntrl-alt-F7).

How to reproduce:
a) start up normal X and login
b) switch to text console (cntrl-alt-F1)
c) text console shows up correctly
d) now switch back (cntrl-alt-F7) -> X shows up, but is not completely restored/repainted, Mouse pointer still works, but else X stopps painting/doing anything.
 
"top" shows X continously using 99% CPU time.

"strace" on the X-process shows:
ioctl(6, 0x6444, 0)                     = -1 EBUSY (Device or resource busy)
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn()                             = ? (mask now [])
ioctl(6, 0x6444, 0)                     = -1 EBUSY (Device or resource busy)
ioctl(6, 0x6444, 0)                     = -1 EBUSY (Device or resource busy)
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn()                             = ? (mask now [])
ioctl(6, 0x6444, 0)                     = -1 EBUSY (Device or resource busy)
--- SIGALRM (Alarm clock) @ 0 (0) ---
sigreturn()                             = ? (mask now [])
ioctl(6, 0x6444 <unfinished ...>
Process 4636 detached

justin:~ # rpm -qa | grep -i Xorg
xorg-x11-Xvnc-6.8.2-100
xorg-x11-Mesa-6.8.2-100
xorg-x11-6.8.2-100.2
xorg-x11-fonts-scalable-6.8.2-100
xorg-x11-devel-6.8.2-100
xorg-x11-server-6.8.2-100.2
xorg-x11-fonts-75dpi-6.8.2-100
xorg-x11-server-glx-6.8.2-100
xorg-x11-Mesa-devel-6.8.2-100
xorg-x11-libs-6.8.2-100.2
xorg-x11-Xnest-6.8.2-100

some lspci output:
00:00.0 Host bridge: ATI Technologies Inc AGP Bridge [IGP 320M] (rev 13)
        Flags: bus master, 66MHz, medium devsel, latency 32
        Memory at d4000000 (32-bit, prefetchable) [size=64M]
        Memory at d0400000 (32-bit, prefetchable) [size=4K]
        I/O ports at 8090 [disabled] [size=4]
        Capabilities: [a0] AGP version 2.0

00:01.0 PCI bridge: ATI Technologies Inc PCI Bridge [IGP 320M] (rev 01) (prog-if 00 [Normal decode])
        Flags: bus master, 66MHz, medium devsel, latency 99
        Bus: primary=00, secondary=01, subordinate=01, sec-latency=68
        I/O behind bridge: 00009000-00009fff
        Memory behind bridge: d0100000-d01fffff
        Prefetchable memory behind bridge: e0000000-efffffff

01:05.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility U1 (prog-if 00 [VGA])
        Subsystem: Hewlett-Packard Company Pavilion ze4400 builtin Video
        Flags: bus master, stepping, fast Back2Back, 66MHz, medium devsel, latency 66, IRQ 10
        Memory at e0000000 (32-bit, prefetchable) [size=256M]
        I/O ports at 9000 [size=256]
        Memory at d0100000 (32-bit, non-prefetchable) [size=64K]
        [virtual] Expansion ROM at d0120000 [disabled] [size=128K]
        Capabilities: [58] AGP version 2.0
        Capabilities: [50] Power Management version 2

The same bug existed in SUSE 9.3 as well, SUSE 9.2 was OK.
Comment 1 Helge Deller 2005-11-24 08:04:42 UTC
Created attachment 58548 [details]
Xorg.0.log file
Comment 2 Stefan Dirsch 2005-11-24 08:43:03 UTC
Could you also attach /etc/X11/xorg.conf? Thanks.
Comment 3 Helge Deller 2005-11-24 19:25:30 UTC
Created attachment 58643 [details]
New xorg.conf file

this file was created with sax2. Then I only changed the default size "640x480" to "1024x768", since this was not possible to set with sax2 (the dialogs were out of the view!)
Comment 4 Helge Deller 2005-11-24 19:27:33 UTC
Created attachment 58644 [details]
Xorg.0.log file
Comment 5 Helge Deller 2005-11-24 19:28:38 UTC
Created attachment 58645 [details]
X11 straced

strace of started X on the command line with "X", then switching to text console...
Comment 6 Helge Deller 2005-11-24 19:30:22 UTC
files attached now.
Comment 7 Stefan Dirsch 2005-11-24 21:43:17 UTC
According to the first logfile you did enable DRI (3D support), but this doesn't match the config file you attached!

It might help to disable the kernel framebffer (vga=normal in /boot/grub/menu.lst) or use the kernel framebuffer resolution you want to use for X11 (vga=0x317).
Comment 8 Helge Deller 2005-11-27 14:09:38 UTC
You were absolutely right.
I did some further testing, and with the new attached basic xorg.conf file I can reproduce the problem.
When I add the loading of the "dri" module, the system hangs as I described it above when switching back to X. If I remove the "dri" X module, everything works without problems.
So it seems to be some bug in "dri", right ?
Comment 9 Helge Deller 2005-11-27 14:12:04 UTC
Created attachment 58788 [details]
xorg config file (remove/add "dri")
Comment 10 Helge Deller 2005-11-27 14:16:02 UTC
Created attachment 58789 [details]
new log file (with dri enabled)
Comment 11 Helge Deller 2005-11-27 14:18:28 UTC
Created attachment 58790 [details]
strace of X

This is a "strace" log file.
X was started on the text console, then switched with Cntrl-Alt-F1 to VT1, then switched back with Alt-F7 to X -> X11 hangs forever and does not repaint all
Comment 12 Stefan Dirsch 2005-11-27 15:04:55 UTC
> It might help to disable the kernel framebffer (vga=normal in
> /boot/grub/menu.lst) or use the kernel framebuffer resolution you want to
> use for X11 (vga=0x317).

Could you give it a try?
Comment 13 Helge Deller 2005-11-28 19:55:38 UTC
I tried vga=normal -> same problem.
I tried setting X11 to 16bpp (as the framebuffer vga=0x317) -> same problem.
Comment 14 Stefan Dirsch 2005-11-29 14:59:12 UTC
Still nothing obvious I can see. :-(
Comment 15 Stefan Dirsch 2006-01-27 11:50:37 UTC
We don't have the hardware for investigating this. Need to close this as WONTFIX. :-(