Messages in this thread | | | Date | Thu, 6 Mar 1997 08:08:25 +0100 (MET) | From | "Michael L. Galbraith" <> | Subject | network oops/proc fs destroyed in 2.1.28 |
| |
Shortly after boot of 2.1.28, the following oops occurs. It doesn't matter if I use gated or routed, the same thing happens. This is stock 2.1.28 w. Ted/Linus fix + Dave Miller fix.. conflicts w. Ted/Linus manually resolved.
Oops happens even when I don't touch the machine after ~30 secs.
Network: ne2000 clone w. one alias and hisax with ippp0 and ippp1 ifconfiged down from the get-go.
Gcc 2.7.2.2,libc-5.4.23 + almost every net option known to linux.
More info avail upon request if this is not enough.
-Mike --------------------------------the meat------------------------------------
Unable to handle kernel paging request at virtual address c58304d4 current->tss.cr3 = 049ae000, Dr3 = 049ae000 *pde = 003e2063 *pte = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<c013f54f>] EFLAGS: 00010246 eax: 00000000 ebx: c58304d4 ecx: 00000000 edx: 00000000 esi: c49b3f5c edi: c49b3f5c ebp: c49b3f3c esp: c49b3f1c ds: 0018 es: 0018 ss: 0018 Process routed (pid: 143, process nr: 20, stackpage=c49b3000) Stack: bffff78c c4a9ec0c bffff78c 00008912 000003a0 bffff814 00000400 bffff7b4 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 c013ff6e bffff78c c015c62f 00008912 bffff78c 00008912 c4a9ec0c bffff78c Call Trace: [<c013ff6e>] [<c015c62f>] [<c013a621>] [<c012e93b>] [<c010a4c8>] Code: 8b 13 89 d6 89 ef fc ac aa 84 c0 75 fa 66 8b 43 46 66 89 44 Using `/boot/2.1.28/System.map' to map addresses to symbols.
>>EIP: c013f54f <dev_ifconf+7b/148> Trace: c013ff6e <dev_ioctl+1aa/1fc> Trace: c015c62f <inet_ioctl+3c7/43c> Trace: c013a621 <sock_ioctl+21/28> Trace: c012e93b <sys_ioctl+14b/160> Trace: c010a4c8 <system_call+38/40>
Code: c013f54f <dev_ifconf+7b/148>
After the first oops, `cat /proc/net/dev` produces... every time.
Unable to handle kernel paging request at virtual address c5830508 current->tss.cr3 = 0445d000, Dr3 = 0445d000 *pde = 003e2063 *pte = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<c013f626>] EFLAGS: 00010282 eax: c42ef15b ebx: c58304d4 ecx: c01b6876 edx: 0000015b esi: c42ef15b edi: c42ef000 ebp: 00000000 esp: c446cf34 ds: 0018 es: 0018 ss: 0018 Process cat (pid: 242, process nr: 15, stackpage=c446c000) Stack: 0000015b c58304d4 c013f6f1 c42ef15b c58304d4 c4a22000 c01d1508 c42ef000 00001000 00000000 c0171edd c42ef000 c446cf98 00000000 00000c00 00000000 c4a22000 ffffffea 00001000 c431d8b8 04259067 00000c00 c01d1508 00000000 Call Trace: [<c58304d4>] [<c013f6f1>] [<c58304d4>] [<c0171edd>] [<c0124fcb>] [<c010a4c8>] Code: 8b 43 34 85 c0 74 0b 53 ff d0 83 c4 04 89 c2 eb 03 90 31 d2 Using `/boot/2.1.28/System.map' to map addresses to symbols.
>>EIP: c013f626 <sprintf_stats+a/94> Trace: c58304d4 Trace: c013f6f1 <dev_get_info+41/90> Trace: c58304d4 Trace: c0171edd <proc_readnet+9d/130> Trace: c0124fcb <sys_read+9b/c0> Trace: c010a4c8 <system_call+38/40>
Code: c013f626 <sprintf_stats+a/94>
If an ISDN connection is active at the time of the oops, the link stays up.. forever.. no control what so ever. All utils which access /proc are toast.
| |