Re: Error message: md: si_iosize_max not set

[Vincent Stemen <vs1@xxxxxxx>]

On 2007-09-04, Matthew Dillon <dillon@apollo.backplane.com> wrote:
>:Hi.
>:
>:We are getting an error printed may times with each ramdisk access on
>:a custom ramdisk rootfs we are creating on Dragonfly 1.10.1.  
>:The error is
>:
>:    md: si_iosize_max not set!
>:...
>:
>:Is this a bug, or is there something we are supposed to be doing
>:different starting with 1.10?
>:
>:Regards.
>:Vincent
>
>     It's a bug.  Try this patch.
>
> 					-Matt
> 					Matthew Dillon 
> 					<dillon@backplane.com>
>
> Index: md.c
>===================================================================
> RCS file: /cvs/src/sys/dev/disk/md/md.c,v
> retrieving revision 1.17
> diff -u -p -r1.17 md.c
> --- md.c	31 Jul 2007 20:04:48 -0000	1.17
> +++ md.c	4 Sep 2007 02:18:45 -0000
> @@ -373,6 +373,8 @@ 		DEVSTAT_TYPE_DIRECT | DEVSTAT_TYPE_IF_
>  		DEVSTAT_PRIORITY_OTHER);
>  	sc->dev = disk_create(sc->unit, &sc->disk, &md_ops);
>  	sc->dev->si_drv1 = sc;
> +	sc->dev->si_iosize_max = DFLTPHYS;
> +
>  	return (sc);
>  }
>  

Thank you for the very fast response Matt.

I thought it was appropriate to move this thread to the bugs list.

I patched and recompiled the kernel, but probably will not get a chance to test
it out until tomorrow.  I'll let you know.

In the mean time, I thought I would point out an apparent bug in cvs
on Dragonfly that your patch exposed.

Since it is a different topic, I will start a new thread following this posting
to give you the details.