Re: What breaks aic7xxx in post 2.6.12-rc2 ?

[James Bottomley]

On Wed, 2005-05-18 at 22:07 -0500, K.R. Foley wrote: 
> This also solves my problem that I reported in this thread
> http://marc.theaimsgroup.com/?l=linux-scsi&m=111422854418964&w=2

OK, I think the patch below is the extract of this.  Could you see if it
works by simply patching vanilla 2.6.12-rc4 with no other SCSI patches
(if it does, I'll push it for 2.6.12 final).

James

--- a/drivers/scsi/aic7xxx/aic7xxx_osm.c
+++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c
@@ -2300,6 +2300,30 @@ ahc_linux_alloc_target(struct ahc_softc 
 	targ->target = target;
 	targ->ahc = ahc;
 	ahc->platform_data->targets[target_offset] = targ;
+
+	/* FIXME: This code is just for a bug workaround in 2.6.12
+	 * The correct fix will be in 2.6.13 */
+	{
+		struct ahc_devinfo devinfo;
+		struct ahc_initiator_tinfo *tinfo;
+		struct ahc_tmode_tstate *tstate;
+		char my_channel = channel + 'A';
+		unsigned int our_id = ahc->our_id;
+		
+		if (channel)
+			our_id = ahc->our_id_b;
+
+		tinfo = ahc_fetch_transinfo(ahc, my_channel, ahc->our_id,
+					    targ->target, &tstate);
+		ahc_compile_devinfo(&devinfo, our_id, targ->target,
+				    CAM_LUN_WILDCARD, my_channel,
+				    ROLE_INITIATOR);
+		ahc_set_syncrate(ahc, &devinfo, NULL, 0, 0, 0,
+				 AHC_TRANS_GOAL, /*paused*/FALSE);
+		ahc_set_width(ahc, &devinfo, MSG_EXT_WDTR_BUS_8_BIT,
+			      AHC_TRANS_GOAL, /*paused*/FALSE);
+	}
+
 	return (targ);
 }
 


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/