Kernel 1.3.37 problems...

Joerg Reuter (jr196930@lykos.tng.oche.de)
Tue, 7 Nov 1995 17:47:51 +0000 (GMT)


Hello Linus,
hello list readers ;-)

I have some problems with kernel 1.3.37. Although it seems to be the
most stable kernel of the 1.3 series, I still have some strange
phenomenas... The following kernel oops occured on 486DX2/66 VLB-board
with an Adaptec 152x compatible SCSI controller, a Conner floppy
streamer mounted and zftape-v0.1 module loaded... And xdm running,
X library version 3.1.1, 256 colour server. (Yes, I should upgrade
to 3.1.2 lib's, but you haven't seen the last phone bill...)

Anyway, the first one is not related to zftape (module unloaded):

Oops: 0002
CPU: 0
EIP: 0010:0011d150
EFLAGS: 00010013
eax: 07200720 ebx: 00040000 ecx: 00000000 edx: 001d1fa0
esi: 00000003 edi: 00000002 ebp: ffffffff esp: 0066ff5c
ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
Process cpp (pid: 16212, process nr: 37, stackpage=0066f000)
Stack: 00000f00 0006e101 00000000 fffffff2 00053000 00000293 0012cd1f 00000003
00000000 ffffffff 006d5018 000001b6 00000000 bfffec20 007f1418 007f1f2f
00121c35 0006e100 0066ffb4 006d5018 0006e119 bfffec34 0066ffbc 0010a7d1
Call Trace: 0012cd1f 00121c35 0010a7d1
Code: 89 50 04 89 d8 c1 e8 0c 89 f1 d3 e8 8b 14 b5 74 1f 1d 00 0f

>>EIP: 11d150 <___get_free_pages+98/1f0>
Trace: 12cd1f <_getname+bf/130>
Trace: 121c35 <_sys_open+1d/60>
Trace: 10a7d1 <_system_call+59/a0>

Code: 11d150 <___get_free_pages+98/1f0> movl %edx,0x4(%eax)
Code: 11d153 <___get_free_pages+9b/1f0> movl %ebx,%eax
Code: 11d155 <___get_free_pages+9d/1f0> shrl $0xc,%eax
Code: 11d158 <___get_free_pages+a0/1f0> movl %esi,%ecx
Code: 11d15a <___get_free_pages+a2/1f0> shrl %cl,%eax
Code: 11d15c <___get_free_pages+a4/1f0> movl 0x1d1f74(,%esi,4),%edx

(a whole bunch of kernel Oops following)

The next one was within a module (zftape?):

Oops: 0000
CPU: 0
EIP: 0010:01025d1b
EFLAGS: 00010246
eax: 00007400 ebx: 000000f5 ecx: 00000000 edx: 000088da
esi: 01034000 edi: 00677000 ebp: 00007400 esp: 002f1e64
ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
Process tar (pid: 2107, process nr: 32, stackpage=002f1000)
Stack: 00002800 00000296 000000e7 00000000 01034000 00000000 00000004 70617466
6f635f65 6572706d 735f7373 006b6565 0101ff95 00000296 000000e7 00001067
00002800 00000002 002f1f30 002f1ef4 00000000 00000000 00000003 00001067
Call Trace: 01034000 0101ff95 01020152 0101e490 01028610 010218bd 0100ba02
0102812c 0012ebd3 0010a7d1
Code: 8b 0c 32 a1 78 92 02 01 39 0c 98 74 30 83 3d 58 83 02 01 ff

Trace: 1034000
Trace: 101ff95
Trace: 1020152
Trace: 101e490
Trace: 1028610
Trace: 10218bd
Trace: 100ba02
Trace: 102812c
Trace: 12ebd3 <_sys_ioctl+12b/148>
Trace: 10a7d1 <_system_call+59/a0>

Code: movl (%edx,%esi,1),%ecx
Code: movl 0x1029278,%eax
Code: cmpl %ecx,(%eax,%ebx,4)
Code: je 0000003d <_EIP+3d>
Code: cmpl $0xffffffff,0x1028358

(system recovered and no problems afterwards.)

Finally, today, after 3h of working (doing a backup on the tape),
funny coloured characters on switching the console and 30 seconds later:

Oops: 0000
CPU: 0
EIP: 0010:00123bb8
EFLAGS: 00010202
eax: 0081a000 ebx: 07200720 ecx: 00026f00 edx: 00000067
esi: 0002d000 edi: 0081a000 ebp: bffffb4c esp: 002cef7c
ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
Process tar (pid: 6139, process nr: 29, stackpage=002ce000)
Stack: 00000000 0081a000 00121aab 00000000 0002d000 00000000 0081a000 00121c4a
0081a000 00000000 0002d000 002c3c0c 0002e000 00000004 0081a000 0010a7d1
00026f00 00000000 0002d000 0002e000 00000004 bffffb4c ffffffda 0002002b
Call Trace: 00121aab 00121c4a 0010a7d1
Code: 66 83 7b 0e 00 75 39 53 e8 e3 fe ff ff 31 c0 b9 0c 00 00 00

>>EIP: 123bb8 <_get_empty_filp+30/98>
Trace: 121aab <_do_open+13/180>
Trace: 121c4a <_sys_open+32/60>
Trace: 10a7d1 <_system_call+59/a0>

Code: 123bb8 <_get_empty_filp+30/98> cmpw $0x0,0xe(%ebx)
Code: 123bbd <_get_empty_filp+35/98> jne 123bf8 <_get_empty_filp+70/98>
Code: 123bbf <_get_empty_filp+37/98> pushl %ebx
Code: 123bc0 <_get_empty_filp+38/98> call fffffef0 <gcc2_compiled.+fffffef0>
Code: 123bc5 <_get_empty_filp+3d/98> xorl %eax,%eax
Code: 123bc7 <_get_empty_filp+3f/98> movl $0xc,%ecx

(and a huge amount of Oops following)

Any ideas what's going on?

Joerg Reuter ampr-net: dl1bke@db0pra.ampr.org
AX-25 : DL1BKE @ DB0ACH.#NRW.DEU.EU
Internet: jreuter@lykos.tng.oche.de