Re: 2.6.13-rt12: irqs hard off for 657 usecs
From: Danny ter Haar
Date: Sat Oct 08 2005 - 07:34:57 EST
Ingo Molnar <mingo@xxxxxxx> wrote:
>> Something appears to have disabled IRQs for 657 usecs.
>hm ... i fixed one such bug in rt.c recently, but more could be lurking.
>Could you send me your .config?
could this be the same bug that's been hitting me since 2.6.13* ?
usenetgateway (heavy used server) with scsi/gig-E ethernet which
crashes within couple of days.
Last kernel i tried was 2.6.14-rc3-git5 and that was the first kernel
which also reported:
warning: many lost ticks.
Your time source seems to be instable or some driver is hogging interupts
Falling back to HPET
printk: 4 messages suppressed.
The crash itself:
scsi0:0:0:0: Attempting to queue an ABORT message:CDB: 0x2a 0x0 0x1 0x79 0x80 0x72 0x0 0x0 0x28 0x0
scsi0: At time of recovery, card was not paused
>>>>>>>>>>>>>>>>>> Dump Card State Begins <<<<<<<<<<<<<<<<<
scsi0: Dumping Card State at program address 0x26 Mode 0x33
Card was paused
SCSISIGI[0x0]:(P_DATAOUT) SCSIPHASE[0x0] SCSIBUS[0x0]
SEQINTCTL[0x0] SEQ_FLAGS[0x0] SEQ_FLAGS2[0x0] SSTAT0[0x0]
SSTAT1[0x0] SSTAT2[0x0] SSTAT3[0x0] PERRDIAG[0x0]
LQISTAT0[0x0] LQISTAT1[0x0] LQISTAT2[0x0] LQOSTAT0[0x0]
SCB Count = 128 CMDS_PENDING = 32 LASTSCB 0x6 CURRSCB 0x1c NEXTSCB 0xff80
qinstart = 44079 qinfifonext = 44079
14 FIFO_USE[0x0] SCB_CONTROL[0x60]:(TAG_ENB|DISCENB) SCB_SCSIID[0x7]
On all previous kernel crashes either scsi or ethernet suddenly quit working.
Since scsi driver is exact same version as 2.6.12-mm1 which for me keeps working
perfectly, it sounds logic that if the IRQ system gets fedup some driver is likely
to grind the kernel to a painfull halt.
Just my feedback
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/