Messages in this thread | | | Date | Fri, 19 Mar 2004 16:46:55 -0300 (BRT) | From | Marcelo Tosatti <> | Subject | Re: 2.4.25 oops in do_softirq |
| |
On Fri, 19 Mar 2004, Matt Bernstein wrote:
> Hi, > > Our big student fileserver has crashed two nights in a row (while the > dumps were running). I didn't have time to capture the oops the first > time; this time I've copied it by hand. The modules having changed are > credit to my initrd. The machine is a dual-Athlon running Debian woody. > > We've recently changed the gigabit NIC from a 33MHz, 64-bit fibre acenic > under 2.4.23 (stable for at least weeks) to a 66MHz, 64-bit copper ns83820 > under 2.4.25 while changing our switches. It had run fine for a week. > > Here's some possibly irrelevant information: I can provide plenty more if > need be! (I've tried to run a memtest86+ but had to abort at 9am, as our > students have a deadline..) > > Cheers, > > Matt > > Linux version 2.4.25 (mb@absinthe-pc) (gcc version 2.96 20000731 (Red Hat Linux 7.3 2.96-118)) #1 SMP Wed Mar 10 11:22:38 GMT 2004 > > 00:09.0 Ethernet controller: National Semiconductor Corporation DP83820 10/100/1000 Ethernet Controller > Subsystem: Netgear: Unknown device 622a > Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR- FastB2B- > Status: Cap- 66Mhz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- > Latency: 64 (2750ns min, 13000ns max), cache line size 10 > Interrupt: pin A routed to IRQ 17 > Region 0: I/O ports at 1000 [size=256] > Region 1: Memory at e8000000 (32-bit, non-prefetchable) [size=4K] > Expansion ROM at <unassigned> [disabled] [size=64K] > > Unable to handle kernel NULL pointer dereference at virtual address > 00000000 printing eip: > c011e5c0 > *pde = 00000000 > Oops: 0002 > CPU: 0 > EIP: 0010:[<c011e5c0>] > Using defaults from ksymoops -t elf32-i386 -a i386 > EFLAGS: 00010213 > eax: 00000000 ebx: 00000002 ecx: 00000000 edx: c024d688 ********
Looks like one bit got flipped, causing the crash.
Most probably bad hardware. Try memtest86 for longer periods (the weekend is here :))
> 00000000 <_EIP>: > Code; c011e5c0 <do_softirq+70/e0> <===== > 0: f7 c3 01 00 00 00 test $0x1,%ebx <=====
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |