Since a couple of releases ago my box fails to reboot, spitting out
scsi related errors and a kernel panic instead. At first I tought
this to be the double driver registration issue, but since the
latest kernel still has this problem...
This is 2.5.46 smp on a box with 2 scsi drivers like so:
CONFIG_SCSI=y
CONFIG_BLK_DEV_SD=y
CONFIG_SCSI_AHA152X=y
CONFIG_SCSI_AIC7XXX=y
CONFIG_AIC7XXX_CMDS_PER_DEVICE=8
CONFIG_AIC7XXX_RESET_DELAY_MS=5000
The messages look like this (scsi0 is the aic7xxx):
Rebooting.
<a pause of several seconds here>
scsi0: Dumping Card State in Data-out phase, at SEQADDR 0x0
ACCUM = 0x0, SINDEX = 0x0, DINDEX = 0x0, ARG_2 = 0x0
HCNT = 0x0
SCSISEQ = 0x0, SBLKCTL = 0xc0
DFCNTRL = 0x0, DFSTATUS = 0x29
LASTPHASE = 0x0, SCSISIGI = 0x0, SXFRCTL0 = 0x0
SSTAT0 = 0x0, SSTAT1= 0x8
STACK == 0x0, 0x0, 0x0, 0x0
SCB count = 4
Kernel NEXTQSCB = 2
Card NEXTQSCB = 0
QINFIFO entries: 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 ... <a lot more of these >
Waiting Queue entries: 0:255 1:255 2:255
Disconnected Queue entries: 0:255 1:255 2:255
QOUTFIFO entries:
Sequencer Free SCB List: 0 1 2
Pending list: 3
Kernel Free SCB list: 1 0
Untagged Q(0): 3
DevQ(0:0:0): 0 waiting
qinpos = 1, SCB index = 2
Kernel panic: Loop 1