15 messages in org.freebsd.freebsd-scsiRe: sa(4) jamming| From | Sent On | Attachments |
|---|---|---|
| J Wunsch | Apr 27, 2001 12:18 am | |
| J Wunsch | Apr 27, 2001 2:02 am | |
| J Wunsch | Apr 27, 2001 8:53 am | |
| Justin T. Gibbs | Apr 27, 2001 9:48 am | |
| Justin T. Gibbs | Apr 27, 2001 9:59 am | |
| J Wunsch | Apr 28, 2001 12:03 pm | |
| Matthew Jacob | Apr 28, 2001 12:13 pm | |
| Kenneth D. Merry | Apr 28, 2001 10:32 pm |  .Other |
| J Wunsch | Apr 29, 2001 9:49 am | |
| Matthew Jacob | Apr 29, 2001 10:03 am | |
| Kenneth D. Merry | Apr 29, 2001 4:20 pm | |
| J Wunsch | Apr 30, 2001 11:35 am | |
| Kenneth D. Merry | Apr 30, 2001 1:57 pm | |
| J Wunsch | Apr 30, 2001 3:00 pm | |
| Kenneth D. Merry | Apr 30, 2001 4:08 pm |
| Subject: | Re: sa(4) jamming | |
|---|---|---|
| From: | J Wunsch (j...@uriah.heep.sax.de) | |
| Date: | Apr 30, 2001 11:35:42 am | |
| List: | org.freebsd.freebsd-scsi | |
As Kenneth D. Merry wrote:
Hopefully that's true. :) I think another part of it may be that not as many people are brave enough to run -current. :) (I go quite a long time before updating some of my machines.)
I usually, too. My reason for upgrading recently was the acquisition of new hardware (i finally got rid of my old P133 box as the main workstation).
What else would trigger a NO SENSE situation?
It's possible you could do it with a CDROM drive. There are a couple of audio-related sense codes that return SS_NOP:
/* R */{SST(0x00, 0x12, SS_NOP, "Audio play operation paused") },
[etc]
Hmm, i can't seem to trigger them, patched or unpatched system. After looking into the SCSI-2 documents, it seems they are not returned as sense codes but in the Audio Status field of an READ SUB-CHANNEL command.
You could always just disable the sa(4) driver patch and do it that way. That would probably be the easiest way to test it.
Well, i did this, on my test machine. Since i already did that previously (without setting action_string), i knew it fixes the ILI problem, but exposes the filemark -> EIO problem. That still happens, also with the addition of setting action_string (of course). Also, hitting the filemark yields a NO SENSE situation:
(sa0:sym0:0:1:0): READ(06). CDB: 8 0 0 80 0 0 (sa0:sym0:0:1:0): CAM Status: SCSI Status Error (sa0:sym0:0:1:0): SCSI Status: Check Condition (sa0:sym0:0:1:0): NO SENSE info:8000 asc:0,1 (sa0:sym0:0:1:0): Filemark detected (sa0:sym0:0:1:0): Retries Exhausted (sa0:sym0:0:1:0): error 5 (sa0:sym0:0:1:0): Unretryable Error
I can't seem to trigger printing the action_string ("No Recovery Action Needed") in either situation (ILI or filemark), however.
Anyway, it seems that patch is the right thing:
diff -u -r1.33 cam_periph.c --- /sys/cam/cam_periph.c 2001/03/28 09:17:49 1.33 +++ /sys/cam/cam_periph.c 2001/04/29 16:49:40 @@ -1369,6 +1369,9 @@
switch (err_action & SS_MASK) { case SS_NOP: + action_string = "No Recovery Action Needed"; + error = 0; + break; case SS_RETRY: action_string = "Retrying Command"; error = ERESTART;
-- cheers, J"org .-.-. --... ...-- -.. . DL8DTL
http://www.sax.de/~joerg/ NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)
To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message