Subject: Re: 'wdc0:0:0: lost interrupt' on tibook since PMAP changes
To: Markus W Kilbinger <kilbi@rad.rwth-aachen.de>
From: Manuel Bouyer <bouyer@antioche.lip6.fr>
List: port-macppc
Date: 06/19/2001 18:33:53
--YiEDa0DAkWCtVeE4
Content-Type: text/plain; charset=us-ascii

On Tue, Jun 19, 2001 at 12:54:14AM +0200, Markus W Kilbinger wrote:
> >>>>> "Manuel" == Manuel Bouyer <bouyer@antioche.lip6.fr> writes:
> 
>     >> /*      $NetBSD: wdc_obio.c,v 1.13 2001/06/15 10:35:26 bouyer Exp $  */
>     >> 
>     >> the most actual I can find and which shows the timeouts. So it
>     >> seems to be an another problem.
> 
>     Manuel> Can you try to revert this to 1.12 ?
> 
> Yep: It solved the problem!

Not really a good new; this means that the changes I made to fix some problems
don't work on all hardware.

Can you post the dmesg; then apply the attached patch and tell me what
values have "conf" and "new conf" ?

The patch is agains wdc_obio.c 1.13

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--

--YiEDa0DAkWCtVeE4
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename=diff

--- wdc_obio.c.orig	Tue Jun 19 18:28:05 2001
+++ wdc_obio.c	Tue Jun 19 18:32:59 2001
@@ -252,6 +252,7 @@
 	int cycle_tick, act_tick, inact_tick, half_tick;
 
 
+	printf("conf = 0x%x\n", bus_space_read_4(chp->cmd_iot, chp->cmd_ioh, CONFIG_REG);
 	for (drive = 0; drive < 2; drive++) {
 		drvp = &chp->ch_drive[drive];
 		if ((drvp->drive_flags & DRIVE) == 0)
@@ -296,10 +297,7 @@
 		    (half_tick << 21) | (inact_tick << 16) | (act_tick << 11);
 	}
 	bus_space_write_4(chp->cmd_iot, chp->cmd_ioh, CONFIG_REG, conf);
-#if 0
-	printf("conf = 0x%x, cyc = %d (%d ns), act = %d (%d ns), inact = %d\n",
-	    conf, cycle_tick, min_cycle, act_tick, min_active, inact_tick);
-#endif
+	printf("new conf = 0x%x\n", conf);
 	wdc_print_modes(chp);
 }
 

--YiEDa0DAkWCtVeE4--