Linux PARISC architecture development
 help / color / mirror / Atom feed
* [parisc-linux] sstfb crash
@ 2003-04-25 12:08 Arjen Hommersom
  2003-04-25 16:57 ` Joel Soete
  2003-04-25 22:52 ` Grant Grundler
  0 siblings, 2 replies; 6+ messages in thread
From: Arjen Hommersom @ 2003-04-25 12:08 UTC (permalink / raw)
  To: parisc-linux

Hi,

I just tried to test my Voodoo2 card on a C240 machine. I compiled 2.4.20-pa32 
and set CONFIG_FB_VOODOO1=m. The compiler was 3.0.4. Unfortunately it crashes 
when i try to load it. The following messages appear just before it locks up 
hard:

# modprobe sstfb
sstfb : Voodoo2 with ICS ICS5342 dac
sstfb: framebuffer at 0xf3400000, mapped at 0xf3400000, size 2 Mb

Stack dump:
 1ec64900:  0004ff0e 10159e6c 1ec64880 002acc20
 1ec648f0:  5758595a 0003242c 0003556c 00000000
 1ec648e0:  10650800 00000000 1035e1df 1ebe5a14
 1ec648d0:  6269745f 1ebe5a14 6269745f 1ec64710
 1ec648c0:  6c6f636b 73747576 00000000 1ec647a0
 1ec648b0:  000001ff 00032adc 1f185340 0000000f

Kernel addresses on the stack:
[<10159e6c>]  [<0003242c>]  [<0003556c>]  [<00032adc>]
[<000355cc>]  [<0003556c>]  [<00034fee>]  [<00034940>]
[<00034fb4>]  [<0003556c>]  [<000355a4>]  [<101f1ac8>]
[<000355a4>]  [<101f1b90>]  [<000355a4>]  [<000345c8>]
[<1011cff0>]  [<101049fc>]  [<00032060>]  [<10109f90>]
[<0002b800>]  [<10109084>]  [<10108cf4>]  [<1013eaac>]
[<1013d610>]  [<1014db28>]  [<1011f77c>]

Kernel Fault: Code=15 regs=1ec64900 (Addr=eae3e729)

     YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
PSW: 00000000000001001111111100001110
r00-03  00000000 102ad810 101f5f68 1ebe5ab4
r04-07  00000000 00000000 00000000 1ebe5800
r08-11  ffffffff 1ec64810 6269745f 1ebaa000
r12-15  1fdabcc0 1ec64550 000435a8 00042da8
r16-19  00042da8 00042da8 000521c8 00450192
r20-23  001f5dce 2bb0e9ab eae3e71d 745f6c6f
r24-27  00000000 00000000 745f6c6f 10292010
r28-31  0000000f f3000000 1ec64900 101f20d4
sr0-3   00000000 0000034d 00000000 0000034d 
sr4-7   00000000 00000000 00000000 00000000

IASQ: 00000000 00000000 IADQ: 10135394 10135398
 IIR: 0ed8109a    ISR: 00000000  IOR: eae3e729
 CPU:        0   CR30: 1ec64000 CR31: 10348000
 ORIG_R28: 102acdc0

The System.map can be found at http://213.46.122.180/System.map.bz2

Any ideas?

Arjen Hommersom

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [parisc-linux] sstfb crash
  2003-04-25 16:57 ` Joel Soete
@ 2003-04-25 16:13   ` Arjen Hommersom
  2003-04-25 18:19     ` Joel Soete
  0 siblings, 1 reply; 6+ messages in thread
From: Arjen Hommersom @ 2003-04-25 16:13 UTC (permalink / raw)
  To: parisc-linux

[-- Attachment #1: Type: text/plain, Size: 529 bytes --]

On Friday 25 April 2003 18:57, Joel Soete wrote:
> >I just tried to test my Voodoo2 card on a C240 machine. I compiled
> > 2.4.20-pa32 and set CONFIG_FB_VOODOO1=m. The compiler was 3.0.4.
> > Unfortunately it crashes when i try to load it. The following messages
> > appear just before it locks up hard: [dump]
>
> It would be also of good help if you try to run the dump_analyser.sh (T.
> Varene) on this 'dump' with your kernel.

Thanks for your reply. I have attached the output of the dump_analyser.

regards,
Arjen Hommersom

[-- Attachment #2: dumpoutput.txt --]
[-- Type: text/plain, Size: 2507 bytes --]

IAOQ = 10135394
Func: kfree, Off: 34, Addr: 0x10135394
10135390:	00 01 0d 60 	ssm 1,r0
10135394:	35 d3 00 30 	ldo 18(r14),r19
10135398:	4a 63 00 c8 	ldw 64(r19),r3
1013539c:	36 73 00 c8 	ldo 64(r19),r19

GR0 = 00000000

GR1 = 102ad810
Func: vm_table, Off: c, Addr: 0x102ad810
102ad810:	92 e0 b0 64 	cmpiclr,>> 32,r23,r0
102ad814:	34 17 00 64 	ldi 32,r23
102ad818:	8a e6 80 48 	cmpb,>>= r6,r23,102ad844 <tcp_retransmit_timer+0xe4>
102ad81c:	34 05 00 02 	ldi 1,r5

GR2 = 101f5f68
Func: fb_alloc_cmap, Off: 3c, Addr: 0x101f5f68
101f5f60:	eb f8 1d 9c 	b,l 10227e34 <sr_audio_ioctl+0x2b4>,r31
101f5f64:	0b 83 0a 03 	add,l r3,ret0,r3
101f5f68:	08 1d 02 5a 	copy ret1,r26
101f5f6c:	08 ba 0a 1a 	add,l r26,r5,r26

GR3 = 1ebe5ab4

GR4 = 00000000

GR5 = 00000000

GR6 = 00000000

GR7 = 1ebe5800

GR8 = ffffffff

GR9 = 1ec64810

GR10 = 6269745f

GR11 = 1ebaa000

GR12 = 1fdabcc0

GR13 = 1ec64550

GR14 = 000435a8

GR15 = 00042da8

GR16 = 00042da8

GR17 = 00042da8

GR18 = 000521c8

GR19 = 00450192

GR20 = 001f5dce

GR21 = 2bb0e9ab

GR22 = eae3e71d

GR23 = 745f6c6f

GR24 = 00000000

GR25 = 00000000

GR26 = 745f6c6f

GR27 = 10292010
Func: $global$, Off: 0, Addr: 0x10292010
10292010:	6f cd 01 00 	stw,ma r13,80(sp)

GR28 = 0000000f

GR29 = f3000000

GR30 = 1ec64900

GR31 = 101f20d4
Func: pci_write_config_dword, Off: 40, Addr: 0x101f20d4
101f20d0:	0a 98 02 14 	and r24,r20,r20
101f20d4:	34 13 00 80 	ldi 40,r19
101f20d8:	82 93 20 70 	cmpb,= r19,r20,101f2118 <set_geometry_intr+0x70>
101f20dc:	34 1c 00 00 	ldi 0,ret0

Kernel symbols on the stack:
[<10159e6c>]: Func: seq_read, Off: 318, Addr: 0x10159e6c
[<0003242c>]: 
[<0003556c>]: 
[<00032adc>]: 
[<000355cc>]: 
[<0003556c>]: 
[<00034fee>]: 
[<00034940>]: 
[<00034fb4>]: 
[<0003556c>]: 
[<000355a4>]: 
[<101f1ac8>]: Func: pci_announce_device, Off: 60, Addr: 0x101f1ac8
[<000355a4>]: 
[<101f1b90>]: Func: pci_register_driver, Off: 98, Addr: 0x101f1b90
[<000355a4>]: 
[<000345c8>]: 
[<1011cff0>]: Func: sys_init_module, Off: 640, Addr: 0x1011cff0
[<101049fc>]: Func: handle_interruption, Off: 150, Addr: 0x101049fc
[<00032060>]: 
[<10109f90>]: Func: syscall_exit, Off: 0, Addr: 0x10109f90
[<0002b800>]: 
[<10109084>]: Func: intr_check_sig, Off: 0, Addr: 0x10109084
[<10108cf4>]: Func: _switch_to_ret, Off: 0, Addr: 0x10108cf4
[<1013eaac>]: Func: chrdev_open, Off: 54, Addr: 0x1013eaac
[<1013d610>]: Func: dentry_open, Off: 1e8, Addr: 0x1013d610
[<1014db28>]: Func: locate_fd, Off: 88, Addr: 0x1014db28
[<1011f77c>]: Func: it_real_fn, Off: 0, Addr: 0x1011f77c

Done.

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [parisc-linux] sstfb crash
  2003-04-25 12:08 [parisc-linux] sstfb crash Arjen Hommersom
@ 2003-04-25 16:57 ` Joel Soete
  2003-04-25 16:13   ` Arjen Hommersom
  2003-04-25 22:52 ` Grant Grundler
  1 sibling, 1 reply; 6+ messages in thread
From: Joel Soete @ 2003-04-25 16:57 UTC (permalink / raw)
  To: Arjen Hommersom; +Cc: parisc-linux

Arjen Hommersom wrote:

>Hi,
>
>I just tried to test my Voodoo2 card on a C240 machine. I compiled 2.4.20-pa32 
>and set CONFIG_FB_VOODOO1=m. The compiler was 3.0.4. Unfortunately it crashes 
>when i try to load it. The following messages appear just before it locks up 
>hard:
>
># modprobe sstfb
>sstfb : Voodoo2 with ICS ICS5342 dac
>sstfb: framebuffer at 0xf3400000, mapped at 0xf3400000, size 2 Mb
>
>Stack dump:
> 1ec64900:  0004ff0e 10159e6c 1ec64880 002acc20
> 1ec648f0:  5758595a 0003242c 0003556c 00000000
> 1ec648e0:  10650800 00000000 1035e1df 1ebe5a14
> 1ec648d0:  6269745f 1ebe5a14 6269745f 1ec64710
> 1ec648c0:  6c6f636b 73747576 00000000 1ec647a0
> 1ec648b0:  000001ff 00032adc 1f185340 0000000f
>
>Kernel addresses on the stack:
>[<10159e6c>]  [<0003242c>]  [<0003556c>]  [<00032adc>]
>[<000355cc>]  [<0003556c>]  [<00034fee>]  [<00034940>]
>[<00034fb4>]  [<0003556c>]  [<000355a4>]  [<101f1ac8>]
>[<000355a4>]  [<101f1b90>]  [<000355a4>]  [<000345c8>]
>[<1011cff0>]  [<101049fc>]  [<00032060>]  [<10109f90>]
>[<0002b800>]  [<10109084>]  [<10108cf4>]  [<1013eaac>]
>[<1013d610>]  [<1014db28>]  [<1011f77c>]
>
>Kernel Fault: Code=15 regs=1ec64900 (Addr=eae3e729)
>
>     YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
>PSW: 00000000000001001111111100001110
>r00-03  00000000 102ad810 101f5f68 1ebe5ab4
>r04-07  00000000 00000000 00000000 1ebe5800
>r08-11  ffffffff 1ec64810 6269745f 1ebaa000
>r12-15  1fdabcc0 1ec64550 000435a8 00042da8
>r16-19  00042da8 00042da8 000521c8 00450192
>r20-23  001f5dce 2bb0e9ab eae3e71d 745f6c6f
>r24-27  00000000 00000000 745f6c6f 10292010
>r28-31  0000000f f3000000 1ec64900 101f20d4
>sr0-3   00000000 0000034d 00000000 0000034d 
>sr4-7   00000000 00000000 00000000 00000000
>
>IASQ: 00000000 00000000 IADQ: 10135394 10135398
> IIR: 0ed8109a    ISR: 00000000  IOR: eae3e729
> CPU:        0   CR30: 1ec64000 CR31: 10348000
> ORIG_R28: 102acdc0
>
>The System.map can be found at http://213.46.122.180/System.map.bz2
>
>Any ideas?
>
>  
>
It would be also of good help if you try to run the dump_analyser.sh (T. 
Varene) on this 'dump' with your kernel.
You can find it at 
<http://cvs.parisc-linux.org/build-tools/dump_analyser.sh>

hth,
Joel

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [parisc-linux] sstfb crash
  2003-04-25 16:13   ` Arjen Hommersom
@ 2003-04-25 18:19     ` Joel Soete
  2003-04-26 18:14       ` Arjen Hommersom
  0 siblings, 1 reply; 6+ messages in thread
From: Joel Soete @ 2003-04-25 18:19 UTC (permalink / raw)
  To: Arjen Hommersom; +Cc: parisc-linux

Arjen Hommersom wrote:

>On Friday 25 April 2003 18:57, Joel Soete wrote:
>  
>
>>>I just tried to test my Voodoo2 card on a C240 machine. I compiled
>>>2.4.20-pa32 and set CONFIG_FB_VOODOO1=m. The compiler was 3.0.4.
>>>Unfortunately it crashes when i try to load it. The following messages
>>>appear just before it locks up hard: [dump]
>>>      
>>>
>>It would be also of good help if you try to run the dump_analyser.sh (T.
>>Varene) on this 'dump' with your kernel.
>>    
>>
>
>Thanks for your reply. I have attached the output of the dump_analyser.
>
>regards,
>Arjen Hommersom
>
>------------------------------------------------------------------------
>
Interesting, isn't it :)

>IAOQ = 10135394
>Func: kfree, Off: 34, Addr: 0x10135394
>
here is the kernel function which made panic kernel. (why? we need 
expert help)

>10135390:	00 01 0d 60 	ssm 1,r0
>10135394:	35 d3 00 30 	ldo 18(r14),r19
>10135398:	4a 63 00 c8 	ldw 64(r19),r3
>1013539c:	36 73 00 c8 	ldo 64(r19),r19
>
>GR0 = 00000000
>
>GR1 = 102ad810
>Func: vm_table, Off: c, Addr: 0x102ad810
>102ad810:	92 e0 b0 64 	cmpiclr,>> 32,r23,r0
>102ad814:	34 17 00 64 	ldi 32,r23
>102ad818:	8a e6 80 48 	cmpb,>>= r6,r23,102ad844 <tcp_retransmit_timer+0xe4>
>102ad81c:	34 05 00 02 	ldi 1,r5
>
>GR2 = 101f5f68
>Func: fb_alloc_cmap, Off: 3c, Addr: 0x101f5f68
>101f5f60:	eb f8 1d 9c 	b,l 10227e34 <sr_audio_ioctl+0x2b4>,r31
>101f5f64:	0b 83 0a 03 	add,l r3,ret0,r3
>101f5f68:	08 1d 02 5a 	copy ret1,r26
>101f5f6c:	08 ba 0a 1a 	add,l r26,r5,r26
>
>GR3 = 1ebe5ab4
>
>GR4 = 00000000
>
>GR5 = 00000000
>
>GR6 = 00000000
>
>GR7 = 1ebe5800
>
>GR8 = ffffffff
>
>GR9 = 1ec64810
>
>GR10 = 6269745f
>
>GR11 = 1ebaa000
>
>GR12 = 1fdabcc0
>
>GR13 = 1ec64550
>
>GR14 = 000435a8
>
>GR15 = 00042da8
>
>GR16 = 00042da8
>
>GR17 = 00042da8
>
>GR18 = 000521c8
>
>GR19 = 00450192
>
>GR20 = 001f5dce
>
>GR21 = 2bb0e9ab
>
>GR22 = eae3e71d
>
>GR23 = 745f6c6f
>
>GR24 = 00000000
>
>GR25 = 00000000
>
>GR26 = 745f6c6f
>
>GR27 = 10292010
>Func: $global$, Off: 0, Addr: 0x10292010
>10292010:	6f cd 01 00 	stw,ma r13,80(sp)
>
>GR28 = 0000000f
>
>GR29 = f3000000
>
>GR30 = 1ec64900
>
>GR31 = 101f20d4
>Func: pci_write_config_dword, Off: 40, Addr: 0x101f20d4
>
iirc this is the calling function

>101f20d0:	0a 98 02 14 	and r24,r20,r20
>101f20d4:	34 13 00 80 	ldi 40,r19
>101f20d8:	82 93 20 70 	cmpb,= r19,r20,101f2118 <set_geometry_intr+0x70>
>101f20dc:	34 1c 00 00 	ldi 0,ret0
>
>Kernel symbols on the stack:
>[<10159e6c>]: Func: seq_read, Off: 318, Addr: 0x10159e6c
>[<0003242c>]: 
>[<0003556c>]: 
>[<00032adc>]: 
>[<000355cc>]: 
>[<0003556c>]: 
>[<00034fee>]: 
>[<00034940>]: 
>[<00034fb4>]: 
>[<0003556c>]: 
>[<000355a4>]: 
>[<101f1ac8>]: Func: pci_announce_device, Off: 60, Addr: 0x101f1ac8
>[<000355a4>]: 
>[<101f1b90>]: Func: pci_register_driver, Off: 98, Addr: 0x101f1b90
>[<000355a4>]: 
>[<000345c8>]: 
>[<1011cff0>]: Func: sys_init_module, Off: 640, Addr: 0x1011cff0
>[<101049fc>]: Func: handle_interruption, Off: 150, Addr: 0x101049fc
>[<00032060>]: 
>[<10109f90>]: Func: syscall_exit, Off: 0, Addr: 0x10109f90
>[<0002b800>]: 
>[<10109084>]: Func: intr_check_sig, Off: 0, Addr: 0x10109084
>[<10108cf4>]: Func: _switch_to_ret, Off: 0, Addr: 0x10108cf4
>[<1013eaac>]: Func: chrdev_open, Off: 54, Addr: 0x1013eaac
>[<1013d610>]: Func: dentry_open, Off: 1e8, Addr: 0x1013d610
>[<1014db28>]: Func: locate_fd, Off: 88, Addr: 0x1014db28
>[<1011f77c>]: Func: it_real_fn, Off: 0, Addr: 0x1011f77c
>
>Done.
>  
>
I will be interested by expert analyse.

Sorry to not be helpfull,
    Joel

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [parisc-linux] sstfb crash
  2003-04-25 12:08 [parisc-linux] sstfb crash Arjen Hommersom
  2003-04-25 16:57 ` Joel Soete
@ 2003-04-25 22:52 ` Grant Grundler
  1 sibling, 0 replies; 6+ messages in thread
From: Grant Grundler @ 2003-04-25 22:52 UTC (permalink / raw)
  To: Arjen Hommersom; +Cc: parisc-linux

On Fri, Apr 25, 2003 at 02:08:38PM +0200, Arjen Hommersom wrote:
> Kernel Fault: Code=15 regs=1ec64900 (Addr=eae3e729)

data page fault - trying to access an invalid address.
 
>      YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
> PSW: 00000000000001001111111100001110
> r00-03  00000000 102ad810 101f5f68 1ebe5ab4
> r04-07  00000000 00000000 00000000 1ebe5800
> r08-11  ffffffff 1ec64810 6269745f 1ebaa000
> r12-15  1fdabcc0 1ec64550 000435a8 00042da8
> r16-19  00042da8 00042da8 000521c8 00450192
> r20-23  001f5dce 2bb0e9ab eae3e71d 745f6c6f
> r24-27  00000000 00000000 745f6c6f 10292010
> r28-31  0000000f f3000000 1ec64900 101f20d4
> sr0-3   00000000 0000034d 00000000 0000034d 
> sr4-7   00000000 00000000 00000000 00000000
> 
> IASQ: 00000000 00000000 IADQ: 10135394 10135398

"IAOQ" (not IADQ) points at the offending instruction.

grant

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [parisc-linux] sstfb crash
  2003-04-25 18:19     ` Joel Soete
@ 2003-04-26 18:14       ` Arjen Hommersom
  0 siblings, 0 replies; 6+ messages in thread
From: Arjen Hommersom @ 2003-04-26 18:14 UTC (permalink / raw)
  To: parisc-linux

On Friday 25 April 2003 20:19, Joel Soete wrote:

> >IAOQ = 10135394
> >Func: kfree, Off: 34, Addr: 0x10135394
>
> here is the kernel function which made panic kernel. (why? we need
> expert help)
>

It seems that the crash occurs at this code piece (line 902 of sstfb.c):

        if (old_bpp != var->bits_per_pixel) {
            if ((err = fb_alloc_cmap(&display->cmap, 0, 0)))
                return err;
            sstfb_install_cmap(con, info);
        }

old_bpp is some strange value (seems random, nothing realistic) and is not 
equal to var->bits_per_pixel (16 in my case). The fb_alloc_cmap is called to 
free the old colormap and that seems to be impossible. Commenting out that 
call and setting the monitor type to 1024x768 makes it work most of the 
times. 

Any suggestions for a correct fix are appreciated.

Arjen Hommersom

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2003-04-26 19:17 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-04-25 12:08 [parisc-linux] sstfb crash Arjen Hommersom
2003-04-25 16:57 ` Joel Soete
2003-04-25 16:13   ` Arjen Hommersom
2003-04-25 18:19     ` Joel Soete
2003-04-26 18:14       ` Arjen Hommersom
2003-04-25 22:52 ` Grant Grundler

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox