Date: Mon, 27 Aug 2001 13:01:31 -0600 From: "Kenneth D. Merry" <ken@kdm.org> To: Gerrit Kuehn <gerrit@pmp.uni-hannover.de> Cc: freebsd-scsi@FreeBSD.ORG Subject: Re: problems with camcontrol rescan Message-ID: <20010827130130.A20496@panzer.kdm.org> In-Reply-To: <01082713450808.22981@mako.pmp.uni-hannover.de>; from gerrit@pmp.uni-hannover.de on Mon, Aug 27, 2001 at 01:41:22PM %2B0200 References: <0108241008070E.15703@mako.pmp.uni-hannover.de> <20010824140054.A5088@panzer.kdm.org> <01082713450808.22981@mako.pmp.uni-hannover.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Aug 27, 2001 at 13:41:22 +0200, Gerrit Kuehn wrote:
> Am Fri, 24 Aug 2001 schrieb Kenneth D. Merry
>
> > It sounds like what is likely happening is that the controller isn't
> > returning some of the CCBs that are being sent down to scan the bus. Since
> > the CCBs don't get returned, the rescan hangs forever. This is just
> > speculation, though; I haven't looked at the driver.
>
> BTW: Joerg Wunsch, who advised me to state my problem in this list here, already
> did have a quick look at the source of the driver. This is what he sent me:
>
> case XPT_SCAN_BUS:
> {
> #ifdef DC395x_trm_DEBUG1
> printf(" XPT_SCAN_BUS \n");
> #endif
> pccb->ccb_h.status = CAM_REQ_INVALID;
> xpt_done(pccb);
> }
> break;
>
> ...
>
> case XPT_SCAN_LUN:
> {
> #ifdef DC395x_trm_DEBUG1
> printf(" XPT_SCAN_LUN \n");
> #endif
> pccb->ccb_h.status = CAM_REQ_INVALID;
> xpt_done(pccb);
> }
> break;
>
>
> >From this fragment of code I really don't understand why there isn't at least
> an error message.
Well, the HBA driver doesn't have to deal with XPT_SCAN_BUS and
XPT_SCAN_LUN CCBs. Those CCB types are handled by the transport layer
and then translated into SCSI commands which are then sent to the HBA
driver.
HBA drivers should send back an error if you send down a scan bus or
scan lun ccb, but there's probably no need to print something.
> And what is more: what happens when FreeBSD is scanning the devices on
> boot-time? That procedure obviously is working fine...
It's pretty much the same code path at boot time. That's why it's somewhat
puzzling that it's hanging up. Rescanning seems to work with other HBAs,
though.
Ken
--
Kenneth Merry
ken@kdm.org
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010827130130.A20496>