Re: sparc64 problems w/ Fujitsu M2954ESP drive

[francisco <frisco_(_at_)_blackant_(_dot_)_net>]

I switched the 4 gig Fujitsu with the 4 gig Seagate that didn't give me
problems on boot up.  This seemed to work, but after a few hours i got the
following messages:

sd1(esp0:1:0): Check Condition on opcode 0x28
    SENSE KEY: Illegal Request
     ASC/ASCQ: Illegal Field in CDB
     FRU CODE: 0x1
         SKSV: Error in CDB, Offset 6, bit 15

(esp0:1:0): selection failed; 0 left in FIFO [intr 18, stat 87, step 3]

sd1(esp0:1:0): esp0: timed out [ecb 0x1e0a1c0 (flags 0x1, dleft 1000, stat
0)],
<state 2, nexus 0x1e0a1c0, phase(l 7, c 7, p 3), resid 0, msg(q 20,o 0) >
sd1(esp0:1:0): esp0: timed out [ecb 0x1e0a1c0 (flags 0x41, dleft 1000,
stat 0)],
 <state 2, nexus 0x1e0a1c0, phase(l 7, c 7, p 3), resid 0, msg(q 20,o 0) >
AGAIN
esp0: invalid state: 6esp0: resetting SCSI bus
esp0: SCSI bus reset
esp0: invalid state: 6esp0: resetting SCSI bus
esp0: SCSI bus reset


These last 2 messages were continually repeated on console.
I tried to halt the machine but the command just hung.  Upon hard reboot
the partitions on sd1 needed serious repair.  In single user mode, i can
generate the following error:
sd1(esp0:1:0): Check Condition on opcode 0x28
    SENSE KEY: Illegal Request
     ASC/ASCQ: Invalid Command Operation Code
     FRU CODE: 0x2
         SKSV: Error in CDB, Offset 0, bit 15

by working with large files on an fs on sd1, but the system doesn't hang.

looking through the archives, it seems my problem is similar to
http://marc.theaimsgroup.com/?l=openbsd-sparc&m=100852314013046&w=2
(though i dont get kernel panic) but i dont see any followup/resolution to
that mail.

is this a problem of bad hardware or with esp driver?

-f
http://www.blackant.net/