linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: James Bottomley <James.Bottomley@HansenPartnership.com>
To: "rusty@rustcorp.com.au" <rusty@rustcorp.com.au>,
	"masami.hiramatsu.pt@hitachi.com"
	<masami.hiramatsu.pt@hitachi.com>
Cc: linux-kernel <linux-kernel@vger.kernel.org>,
	linux-scsi <linux-scsi@vger.kernel.org>
Subject: module: fix module_refcount() return when running in a module exit routine
Date: Sun, 18 Jan 2015 08:55:46 -0800	[thread overview]
Message-ID: <1421600146.2080.8.camel@HansenPartnership.com> (raw)

From: James Bottomley <JBottomley@Parallels.com>

After e513cc1 module: Remove stop_machine from module unloading,
module_refcount() is returning (unsigned long)-1 when called from within
a routine that runs in module_exit.  This is confusing the scsi device
put code which is coded to detect a module_refcount() of zero for
running within a module exit routine and not try to do another
module_put.  The fix is to restore the original behaviour of
module_refcount() and return zero if we're running inside an exit
routine.

Fixes: e513cc1c07e2ab93a4514eec9833e031df3e30bb
Reported-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

---

diff --git a/kernel/module.c b/kernel/module.c
index 3965511..c33a113 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -774,7 +774,12 @@ static int try_stop_module(struct module *mod, int flags, int *forced)
 
 unsigned long module_refcount(struct module *mod)
 {
-	return (unsigned long)atomic_read(&mod->refcnt) - MODULE_REF_BASE;
+	unsigned long ret = atomic_read(&mod->refcnt);
+
+	if (ret == 0)
+		/* ref is already zero (probably in module exit) */
+		return 0;
+	return ret - MODULE_REF_BASE;
 }
 EXPORT_SYMBOL(module_refcount);
 



             reply	other threads:[~2015-01-18 16:55 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-01-18 16:55 James Bottomley [this message]
2015-01-18 23:37 ` module: fix module_refcount() return when running in a module exit routine Rusty Russell
2015-01-19  5:18 ` Masami Hiramatsu
2015-01-19  5:51   ` Rusty Russell
2015-01-19  8:28     ` Christoph Hellwig
2015-01-19 16:07       ` James Bottomley
2015-01-19 16:08     ` James Bottomley
2015-01-20  0:45       ` Rusty Russell
2015-01-20  2:17         ` Masami Hiramatsu
2015-01-20 17:23         ` James Bottomley
2015-01-21  5:30           ` Rusty Russell
2015-01-22 16:50           ` Christoph Hellwig
2015-01-22 17:02             ` James Bottomley
2015-01-23  2:54               ` Rusty Russell
2015-01-23 13:17                 ` Christoph Hellwig
2015-01-23 18:42                   ` James Bottomley
2015-01-23 23:35                     ` Rusty Russell
2015-01-26 17:16                     ` Christoph Hellwig
2015-01-28  9:23                     ` Bart Van Assche
2015-01-28 21:45                       ` James Bottomley
2015-01-29 12:16                         ` Bart Van Assche

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=1421600146.2080.8.camel@HansenPartnership.com \
    --to=james.bottomley@hansenpartnership.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=masami.hiramatsu.pt@hitachi.com \
    --cc=rusty@rustcorp.com.au \
    /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;
as well as URLs for NNTP newsgroup(s).