kern/112119: system hangs when starts k3b on RELENG_6

Scott Long scottl at samsco.org
Fri Apr 27 20:28:35 UTC 2007


Thomas Quinot wrote:
> * Scott Long, 2007-04-27 :
> 
>> Oh hell, I know exactly what the problem is!  The opcode for a
>> TEST_UNIT_READY is 0x00.  This is probably the command that is
>> generating the CHECK_CONDITION.  The test for saved_cmd is entirely
>> bogus.
> 
> Hmmmm. Looks like a very plausible culprit. Good catch Scott!
> (I felt there had to be something wrong when I wrote that test,
> incidentally, precisely because of TEST_UNIT_READY).
> 
> Nikolay, Ganbold, (and others), here's another patch against 1.42.2.3,
> please let me know if it works for you.
> 
> Thomas.
> 
> Index: atapi-cam.c
> ===================================================================
> RCS file: /space/mirror/ncvs/src/sys/dev/ata/atapi-cam.c,v
> retrieving revision 1.50
> diff -u -r1.50 atapi-cam.c
> --- atapi-cam.c	14 Mar 2007 01:59:00 -0000	1.50
> +++ atapi-cam.c	27 Apr 2007 19:26:09 -0000
> @@ -729,7 +743,7 @@
>  		 * issued a REQUEST SENSE automatically and that operation
>  		 * returned without error.
>  		 */
> -		if (request->u.atapi.saved_cmd != 0 && request->error == 0) {
> +		if (request->u.atapi.sense.key != 0 && request->error == 0) {
>  		    bcopy (&request->u.atapi.sense, &csio->sense_data, sizeof(struct atapi_sense));
>  		    csio->ccb_h.status |= CAM_AUTOSNS_VALID;
>  		}

Right, just make sure that this isn't filled with garbage before you
send the command.  If you're using ata_alloc_request(), and you're not
recycling requests internally, then you should be fine.

Scott



More information about the freebsd-stable mailing list