[4214] in linux-scsi channel archive

home help back first fref pref prev next nref lref last post

Re: ncr53c875 loop

daemon@ATHENA.MIT.EDU (Gerard Roudier)
Wed Jun 24 15:38:02 1998

Date: 	Wed, 24 Jun 1998 20:33:50 +0200 (MET DST)
From: Gerard Roudier <groudier@club-internet.fr>
To: Keith Owens <kaos@ocs.com.au>
cc: linux-scsi@vger.rutgers.edu
In-Reply-To: <19980623235448.1772.qmail@mail.ocs.com.au>


On Wed, 24 Jun 1998, Keith Owens wrote:

> I have an intermittent problem where ncr53c875 will loop trying to
> reload the script while detecting ID 2, Micropolis 3243NT.  Happens on
> 2.0.34 and 2.1.106, both SMP, no overclocking.  Is this hardware or
> software?  You don't have to tell me that the 3243 is a pile of crud,
> this is the third replacement.

The drive is disconnected the bus when the chip was not expecting this 
to happen. Devices may use this to signal a severe error condition 
to the initiator. Donnot have idea about what the drive is complaining
for, however the driver should not reset the bus.
I would suggest you to try this patch:

--- linux/drivers/scsi/ncr53c8xx.c.ORIG	Sat May 16 15:32:49 1998
+++ linux/drivers/scsi/ncr53c8xx.c	Wed Jun 24 20:19:26 1998
@@ -7138,8 +7138,9 @@
 		if (INB (nc_scr1) != 0xff) {
 			OUTB (nc_scr1, HS_UNEXPECTED);
 			OUTL (nc_dsp, NCB_SCRIPT_PHYS (np, cleanup));
-		};
-		ncr_start_reset(np, driver_setup.settle_delay);
+		}
+		else
+			ncr_start_reset(np, driver_setup.settle_delay);
 		return;
 	};
 
And you also could you give a try to latest 3.0b driver version
which is available at ftp://ftp.tux.org/pub/roudier.
This version is normaly quite reliable and is expecting to be prooved so 
in order to be proposed for incorporation into both stable and current 
kernels.

[...]
> ncr53c875-0:2: ERROR (0:4) (8-0-0) (0/5) @ (script 15c:87030000).
> ncr53c875-0: script cmd = 72580000
> ncr53c875-0: regdump: da 00 80 05 47 00 02 0e 02 08 00 00 80 00 0e 02.
> ncr53c875-0: have to clear fifos.
> ncr53c875-0: unexpected disconnect
> ncr53c875-0: restart (scsi reset).
> ncr53c875-0: copying script fragments into the on-board RAM ...

Regards,
   Gerard.


-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu

home help back first fref pref prev next nref lref last post