From: Kurt Garloff <garloff@suse.de>
To: Christoph Hellwig <hch@infradead.org>
Cc: Linux SCSI list <linux-scsi@vger.kernel.org>,
Patrick Mansfield <patmans@us.ibm.com>,
James Bottomley <James.Bottomley@steeleye.com>,
Andrew Morton <akpm@osdl.org>
Subject: Re: PATCH 4/5: scsi-scan-dont-att-pq-notcon
Date: Wed, 21 Apr 2004 18:16:05 +0200 [thread overview]
Message-ID: <20040421161605.GA643@tpkurt.garloff.de> (raw)
In-Reply-To: <20040421163336.A7382@infradead.org>
[-- Attachment #1: Type: text/plain, Size: 3159 bytes --]
On Wed, Apr 21, 2004 at 04:33:36PM +0100, Christoph Hellwig wrote:
> (also needs a respin of the sd.c part of the previous patch, btw - and while
> you're at it can you kill the date in the scsi_scan.c comment? Everyone can
> see when it was applied in bk annotate or the web interface for it)
OTOH, there's so few to be changed that I can resend right away ;-)
garloff@suse.de
Bugfix
Don't mark devices with PQ = 1 (or 3) offline, as they can't be accessed
at all then any more, not even by sg.
Don't attach other upper level drivers if PQ != 0.
SPC3, 7.4.2.
diff -uNrp linux-2.6.5.replun/drivers/scsi/scsi_scan.c linux-2.6.5.nooffline/drivers/scsi/scsi_scan.c
--- linux-2.6.5.replun/drivers/scsi/scsi_scan.c 2004-04-21 14:26:31.000000000 +0200
+++ linux-2.6.5.nooffline/drivers/scsi/scsi_scan.c 2004-04-21 14:28:13.000000000 +0200
@@ -542,16 +542,10 @@ static int scsi_add_lun(struct scsi_devi
* 011 the same. Stay compatible with previous code, and create a
* Scsi_Device for a PQ of 1
*
- * XXX Save the PQ field let the upper layers figure out if they
- * want to attach or not to this device, do not set online FALSE;
- * otherwise, offline devices still get an sd allocated, and they
- * use up an sd slot.
- */
- if (((inq_result[0] >> 5) & 7) == 1) {
- SCSI_LOG_SCAN_BUS(3, printk(KERN_INFO "scsi scan: peripheral"
- " qualifier of 1, device offlined\n"));
- sdev->online = FALSE;
- }
+ * Don't set the device offline here; rather let the upper
+ * level drivers eval the PQ to decide whether they should
+ * attach. So remove ((inq_result[0] >> 5) & 7) == 1 check.
+ */
sdev->removable = (0x80 & inq_result[1]) >> 7;
sdev->lockable = sdev->removable;
diff -uNrp linux-2.6.5.replun/include/scsi/scsi.h linux-2.6.5.nooffline/include/scsi/scsi.h
--- linux-2.6.5.replun/include/scsi/scsi.h 2004-04-20 21:51:40.000000000 +0200
+++ linux-2.6.5.nooffline/include/scsi/scsi.h 2004-04-21 14:28:13.000000000 +0200
@@ -365,6 +365,13 @@ struct scsi_lun {
#define SCSI_2 3
#define SCSI_3 4
+/*
+ * INQ PERIPHERAL QUALIFIERS
+ */
+#define SCSI_INQ_PQ_CON 0x00
+#define SCSI_INQ_PQ_NOT_CON 0x01
+#define SCSI_INQ_PQ_NOT_CAP 0x03
+
/*
* Here are some scsi specific ioctl commands which are sometimes useful.
diff -uNrp linux-2.6.5.attach/drivers/scsi/scsi_sysfs.c linux-2.6.5.noattach/drivers/scsi/scsi_sysfs.c
--- linux-2.6.5.attach/drivers/scsi/scsi_sysfs.c 2004-04-04 05:36:26.000000000 +0200
+++ linux-2.6.5.noattach/drivers/scsi/scsi_sysfs.c 2004-04-21 17:22:22.000000000 +0200
@@ -155,7 +155,9 @@ struct class sdev_class = {
/* all probing is done in the individual ->probe routines */
static int scsi_bus_match(struct device *dev, struct device_driver *gendrv)
{
- return 1;
+ struct scsi_device *sdp = to_scsi_device(dev);
+ int pq = (sdp->inquiry[0] >> 5) & 7;
+ return (pq == SCSI_INQ_PQ_CON)? 1: 0;
}
struct bus_type scsi_bus_type = {
Regards,
--
Kurt Garloff <garloff@suse.de> Cologne, DE
SUSE LINUX AG, Nuernberg, DE SUSE Labs (Head)
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
next prev parent reply other threads:[~2004-04-21 16:16 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-04-18 18:57 Patches for SCSI scanning Kurt Garloff
2004-04-18 23:16 ` James Bottomley
2004-04-20 11:54 ` Kurt Garloff
2004-04-20 12:04 ` Christoph Hellwig
2004-04-20 13:02 ` Kurt Garloff
2004-04-20 14:38 ` James Bottomley
2004-04-20 16:03 ` Kurt Garloff
2004-04-20 16:08 ` James Bottomley
2004-04-21 13:48 ` Kurt Garloff
2004-04-21 15:36 ` James Bottomley
2004-04-21 13:45 ` Kurt Garloff
2004-04-21 14:10 ` PATCH 1/5: scsi-scan-deprecate-forcelun Kurt Garloff
2004-04-21 14:12 ` PATCH 2/5: scsi-scan-blist_replun Kurt Garloff
2004-04-21 15:14 ` Christoph Hellwig
2004-04-21 15:30 ` Kurt Garloff
2004-04-21 16:03 ` Kurt Garloff
2004-04-21 14:13 ` PATCH 3/5: scsi-scan-no-offl-pq-notcon Kurt Garloff
2004-04-21 14:14 ` PATCH 4/5: scsi-scan-dont-att-pq-notcon Kurt Garloff
2004-04-21 15:02 ` Christoph Hellwig
2004-04-21 15:24 ` Kurt Garloff
2004-04-21 15:33 ` Mike Anderson
2004-04-21 15:33 ` Christoph Hellwig
2004-04-21 16:08 ` Kurt Garloff
2004-04-21 16:18 ` James Bottomley
2004-04-21 16:55 ` Patrick Mansfield
2004-04-21 22:51 ` Kurt Garloff
2004-04-22 20:39 ` Patrick Mansfield
2004-04-22 20:45 ` James Bottomley
2004-04-21 16:58 ` Kurt Garloff
2004-04-21 16:16 ` Kurt Garloff [this message]
2004-04-21 15:40 ` James Bottomley
2004-04-21 14:14 ` PATCH 5/5: scsi-scan-inq-timeout Kurt Garloff
2004-04-21 20:24 ` Patrick Mansfield
2004-04-21 22:48 ` Kurt Garloff
2004-04-21 23:49 ` Patrick Mansfield
2004-04-20 16:26 ` Patches for SCSI scanning Patrick Mansfield
2004-04-20 16:42 ` Kurt Garloff
2004-04-20 17:44 ` Patrick Mansfield
2004-04-21 13:52 ` Kurt Garloff
2004-04-20 10:24 ` Fabien Salvi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20040421161605.GA643@tpkurt.garloff.de \
--to=garloff@suse.de \
--cc=James.Bottomley@steeleye.com \
--cc=akpm@osdl.org \
--cc=hch@infradead.org \
--cc=linux-scsi@vger.kernel.org \
--cc=patmans@us.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox