Bug 676330

Summary: i915: 32bit - i915_driver_load hangs in set_mtrr
Product: [openSUSE] openSUSE 11.4 Reporter: Holger Sickenberg <holgi>
Component: InstallationAssignee: Rafael Wysocki <rjw>
Status: RESOLVED DUPLICATE QA Contact: Jiri Srain <jsrain>
Severity: Major    
Priority: P5 - None CC: aj, asn, cobexer, coolo, jeffm, jslaby
Version: RC 2   
Target Milestone: ---   
Hardware: i386   
OS: SUSE Other   
Whiteboard:
Found By: Component Test Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: hwinfo.txt
/screenlog.0

Description Holger Sickenberg 2011-03-02 16:50:00 UTC 
After installing stage 1 and reboot I get (sometimes): 

doing fast boot
Creating device nodes with udev
[   71.667379] BUG: soft lockup - CPU#0 stuck for 63s! [modprobe:168]
[   71.668002] Process modprobe (pid: 168, ti=f738a000 task=f73844b0 task.ti=f738a000)
[   71.668002] Stack:
[   71.668002] Call Trace:
[   71.668002] Code: c0 74 d6 8d 54 15 00 89 f1 89 14 24 89 d8 ba 50 71 21 c0 e8 11 73 07 00 eb bf 8d b4 26 00 00 00 00 8b 44 24 14 85 c0 74 08 f3 90 <8b> 06 85 c0 75 f8 a1 f4 8a 98 c0 e8 6e 79 1e 00 83 e8 01 89 44 
[   71.759378] BUG: soft lockup - CPU#1 stuck for 63s! [migration/1:8]
[   71.760003] Process migration/1 (pid: 8, ti=f5490000 task=f546c1b0 task.ti=f5490000)
[   71.760003] Stack:
[   71.760003] Call Trace:
[   71.760003] Code: 28 66 90 83 fe 03 74 3b f0 ff 4b 14 0f 94 c0 84 c0 89 f1 74 0f 8b 73 10 8b 43 08 83 c6 01 89 43 14 89 73 10 83 f9 04 74 3a f3 90 <8b> 73 10 39 ce 74 f2 83 fe 02 75 cc fa 8d 76 00 eb cb 8d b6 00
Comment 1 Holger Sickenberg 2011-03-02 16:57:15 UTC 
Created attachment 417111 [details]
hwinfo.txt

hwinfo of the machine
Comment 2 Jiri Slaby 2011-03-08 10:01:55 UTC 
(In reply to comment #0)
> After installing stage 1 and reboot I get (sometimes): 
> 
> doing fast boot
> Creating device nodes with udev
> [   71.667379] BUG: soft lockup - CPU#0 stuck for 63s! [modprobe:168]
> [   71.668002] Process modprobe (pid: 168, ti=f738a000 task=f73844b0
> task.ti=f738a000)

Ok, some module crashed. could you enable initcall_debug:
http://en.opensuse.org/SDB:Debugging_boot_hang#Initcall_debug
?
Comment 3 Holger Sickenberg 2011-03-09 08:59:39 UTC 
Created attachment 418255 [details]
/screenlog.0

Output with debug attached.
Comment 4 Andreas Schneider 2011-04-12 13:24:32 UTC 
If you remove the 'vga=XXX' option on the kernel command line it works, this doesn't happen.
Comment 5 Jeff Mahoney 2011-04-15 20:38:34 UTC 
[   75.313441] BUG: soft lockup - CPU#0 stuck for 61s! [migration/0:6]
[   75.316004] Modules linked in: i915(+) drm_kms_helper drm i2c_algo_bit i2c_core uhci_hcd video rtc_cmos rtc_core rtc_lib button ehci_hcd usbcore output fan processor thermal thermal_sys hwmon ata_generic ata_piix ahci libahci libata
[   75.316004] Modules linked in: i915(+) drm_kms_helper drm i2c_algo_bit i2c_core uhci_hcd video rtc_cmos rtc_core rtc_lib button ehci_hcd usbcore output fan processor thermal thermal_sys hwmon ata_generic ata_piix ahci libahci libata
[   75.316004] 
[   75.316004] Pid: 6, comm: migration/0 Not tainted 2.6.37.1-1.2-default #1                  /D915GEV                        
[   75.316004] EIP: 0060:[<c028e4f0>] EFLAGS: 00000293 CPU: 0
[   75.316004] EIP is at stop_machine_cpu_stop+0x50/0xb0
[   75.316004] EAX: 00000000 EBX: f73ddf4c ECX: 00000001 EDX: 00000001
[   75.316004] ESI: 00000001 EDI: 00000000 EBP: f73ddf4c ESP: f546bef8
[   75.316004]  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
[   75.316004] Process migration/0 (pid: 6, ti=f546a000 task=f5466130 task.ti=f546a000)
[   75.316004] Stack:
[   75.316004]  f60031b0 f60031b0 f5466130 f73ddf0c c028e401 00000002 c08e1aa0 f5466130
[   75.316004]  00000000
Comment 6 Rafael Wysocki 2011-04-15 22:58:58 UTC 
@Holger: Does it help if you boot with vga=0 ?
Comment 7 Christoph Obexer 2011-05-08 08:21:48 UTC 
is this a duplicate of https://bugzilla.novell.com/show_bug.cgi?id=688714 ?
i also had a hanging modprobe and migration process on the 11.4 -default kernel
Comment 8 Rafael Wysocki 2011-05-09 19:40:54 UTC 
Yes, it is.

*** This bug has been marked as a duplicate of bug 688714 ***
Comment 9 Holger Sickenberg 2011-05-10 08:28:06 UTC 
No, it does not help to boor with vga=0 - same problems with this option.
Comment 10 Rafael Wysocki 2011-05-10 21:58:49 UTC 
Could you please follow up in bug #688714 ?