Re: core dump on kldload atapicam

On Tue, 2009-09-22 at 14:51 +0300, Jaakko Heinonen wrote:
> On 2009-09-21, Warren Block wrote:
> > >> I can confirm that .. kldload atapicam still results in a "Fatal trap 12:
> > >> page fault while in kernel mode" on 8.0-RC1
> 
> The problem was introduced in r195534. Does the following patch work for
> you?
> 
> %%%
> Index: sys/dev/ata/atapi-cam.c
> ===================================================================
> --- sys/dev/ata/atapi-cam.c	(revision 197399)
> +++ sys/dev/ata/atapi-cam.c	(working copy)
> _at__at_ -418,8 +418,10 _at__at_ atapi_action(struct cam_sim *sim, union 
>  		break;
>  	    }
>  	}
> -	cpi->maxio = softc->ata_ch->dma.max_iosize ?
> -	    softc->ata_ch->dma.max_iosize : DFLTPHYS;
> +	if (softc->ata_ch != NULL && softc->ata_ch->dma.max_iosize != 0)
> +	    cpi->maxio = softc->ata_ch->dma.max_iosize;
> +	else
> +	    cpi->maxio = DFLTPHYS;
>  	ccb->ccb_h.status = CAM_REQ_CMP;
>  	xpt_done(ccb);
>  	return;
> %%%
> 
I can confirm that:

a) problem still exists in r197402
b) your patch (hand-applied) fixes it.

I have not done any extensive testing of the device apart from mounting,
unmounting and ejecting /dev/cd0.

Thank you very much for your help.

-- 
Alexandre Kovalenko (Олександр Коваленко)