public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: "H . J . Lu" <hjl@lucon.org>
To: Kristian Hogsberg <hogsberg@users.sourceforge.net>
Cc: Andrew Morton <akpm@zip.com.au>,
	hogsberg@users.sourceforge.net, jamesg@filanet.com,
	linux-1394devel@lists.sourceforge.net,
	lkml <linux-kernel@vger.kernel.org>
Subject: Re: sbp2.c on SMP
Date: Fri, 16 Nov 2001 13:25:20 -0800	[thread overview]
Message-ID: <20011116132520.A6204@lucon.org> (raw)
In-Reply-To: <3BEF27D1.7793AE8E@zip.com.au> <3BEF27D1.7793AE8E@zip.com.au> <20011113191721.A9276@lucon.org> <3BF21B79.5F188A0D@zip.com.au> <20011115193234.A22081@lucon.org> <m3snbeofnw.fsf@dk20037170.bang-olufsen.dk>
In-Reply-To: <m3snbeofnw.fsf@dk20037170.bang-olufsen.dk>; from hogsberg@users.sf.net on Fri, Nov 16, 2001 at 05:15:47PM +0100

On Fri, Nov 16, 2001 at 05:15:47PM +0100, Kristian Hogsberg wrote:
> This is true, but only because the struct list_head is the first
> element in struct node_entry.  If it wasn't, lh would have been -16 or
> so, as Andrew says.
> 
> In any case, it's the wrong fix, because the error is elsewhere:
> neither the host_info list or the node list should contain NULL
> entries.  This is just curing the symptoms.  HJ, could you provide
> some details on the crash?  Do you have the sbp2 module loaded when
> you insmod/rmmod ohci1394, and if so, does it crash without sbp2
> loaded?
> 

Found it. You have to use list_for_each_safe when you remove things.
Here is a patch.


H.J.
----
--- linux-2.4.9-12.2mod/drivers/ieee1394/nodemgr.c.rmmod	Tue Nov 13 19:15:44 2001
+++ linux-2.4.9-12.2mod/drivers/ieee1394/nodemgr.c	Fri Nov 16 13:14:22 2001
@@ -558,7 +558,7 @@ done_reset_host:
 
 static void nodemgr_remove_host(struct hpsb_host *host)
 {
-	struct list_head *lh;
+	struct list_head *lh, *spare;
 	struct host_info *hi = NULL;
 	struct node_entry *ne;
 	unsigned long flags;
@@ -568,7 +568,7 @@ static void nodemgr_remove_host(struct h
 
 	/* First remove all node entries for this host */
 	write_lock_irqsave(&node_lock, flags);
-	list_for_each(lh, &node_list) {
+	list_for_each_safe(lh, spare, &node_list) {
 		ne = list_entry(lh, struct node_entry, list);
 
 		/* Only checking this host */
--- linux-2.4.9-12.2mod/drivers/ieee1394/sbp2.c.rmmod	Thu Nov 15 17:08:49 2001
+++ linux-2.4.9-12.2mod/drivers/ieee1394/sbp2.c	Fri Nov 16 13:14:37 2001
@@ -754,13 +754,13 @@ static int sbp2util_create_command_orb_p
 static void sbp2util_remove_command_orb_pool(struct scsi_id_instance_data *scsi_id,
 					     struct sbp2scsi_host_info *hi)
 {
-	struct list_head *lh;
+	struct list_head *lh, *spare;
 	struct sbp2_command_info *command;
 	unsigned long flags;
         
 	sbp2_spin_lock(&scsi_id->sbp2_command_orb_lock, flags);
 	if (!list_empty(&scsi_id->sbp2_command_orb_completed)) {
-		list_for_each(lh, &scsi_id->sbp2_command_orb_completed) {
+		list_for_each_safe(lh, spare, &scsi_id->sbp2_command_orb_completed) {
 			command = list_entry(lh, struct sbp2_command_info, list);
 
 			/* Release our generic DMA's */
--- linux-2.4.9-12.2mod/drivers/ieee1394/video1394.c.rmmod	Sun Nov 11 21:06:48 2001
+++ linux-2.4.9-12.2mod/drivers/ieee1394/video1394.c	Fri Nov 16 13:15:28 2001
@@ -1592,7 +1592,7 @@ static void video1394_remove_host (struc
 {
 	struct ti_ohci *ohci;
 	unsigned long flags;
-	struct list_head *lh;
+	struct list_head *lh, *spare;
 
 	/* We only work with the OHCI-1394 driver */
 	if (strcmp(host->template->name, OHCI1394_DRIVER_NAME))
@@ -1603,7 +1603,7 @@ static void video1394_remove_host (struc
         spin_lock_irqsave(&video1394_cards_lock, flags);
 	if (!list_empty(&video1394_cards)) {
 		struct video_card *p;
-		list_for_each(lh, &video1394_cards) {
+		list_for_each_safe(lh, spare, &video1394_cards) {
 			p = list_entry(lh, struct video_card, list);
 			if (p ->ohci == ohci) {
 				remove_card(p);

  parent reply	other threads:[~2001-11-16 21:25 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-11-12  1:37 sbp2.c on SMP Andrew Morton
2001-11-12  4:54 ` H . J . Lu
2001-11-12  5:14   ` Andrew Morton
2001-11-12  5:28     ` H . J . Lu
2001-11-12  8:50 ` Alan Cox
2001-11-12 17:34 ` Jens Axboe
2001-11-14  3:17 ` H . J . Lu
2001-11-14  7:21   ` Andrew Morton
2001-11-16  3:32     ` H . J . Lu
2001-11-16 16:15       ` Kristian Hogsberg
2001-11-16 16:30         ` H . J . Lu
2001-11-16 21:25         ` H . J . Lu [this message]
2001-11-16 22:40           ` Kristian Hogsberg
2001-11-26 15:43           ` Oops 2.4.15-pre1aa1 Sven Heinicke
  -- strict thread matches above, loose matches on Subject: below --
2001-11-12  4:49 sbp2.c on SMP Douglas Gilbert

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=20011116132520.A6204@lucon.org \
    --to=hjl@lucon.org \
    --cc=akpm@zip.com.au \
    --cc=hogsberg@users.sourceforge.net \
    --cc=jamesg@filanet.com \
    --cc=linux-1394devel@lists.sourceforge.net \
    --cc=linux-kernel@vger.kernel.org \
    /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