Attachment 294563 Details for Bug 423521 – patch -- fix reference counting for NFS4 callback thread

[patch] patch -- fix reference counting for NFS4 callback thread

0001-NFS-fix-reference-counting-for-NFSv4-callback-threa.patch (text/plain), 2.26 KB, created by Jeff Layton on 2008-02-11 14:44:07 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Jeff Layton

Created: 2008-02-11 14:44:07 UTC

Size: 2.26 KB

patch

obsolete

>From c259bbd0436ad28ecacc9db4edbe4cb3fda97d65 Mon Sep 17 00:00:00 2001
>From: Jeff Layton <jlayton@redhat.com>
>Date: Mon, 11 Feb 2008 09:29:52 -0500
>Subject: [PATCH] NFS: fix reference counting for NFSv4 callback thread
>
>The reference counting for the NFSv4 callback thread stays artificially
>high. When this thread comes down, it doesn't properly tear down the
>svc_serv causing a memory leak. In my testing on an older kernel on
>x86_64, memory would leak out of the 8k kmalloc slab so we're leaking at
>least 2 pages every time the thread comes down.
>
>svc_create creates the svc_serv with a sv_nrthreads count of 1, and then
>svc_create_thread() increments that count. So, whenever the callback
>thread is started it has a sv_nrthreads count of 2. When coming down, it
>calls svc_exit_thread which decrements that count and if it hits 0, it
>tears everything down.
>
>The problem is that nfs_callback_up should be calling svc_destroy on
>the svc_serv on both success and failure. This is how lockd_up_proto
>handles the reference counting, and doing that here fixes the leak.
>
>Signed-off-by: Jeff Layton <jlayton@redhat.com>
>---
> fs/nfs/callback.c |   12 ++++++------
> 1 files changed, 6 insertions(+), 6 deletions(-)
>
>diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
>index bd185a5..5e4a66f 100644
>--- a/fs/nfs/callback.c
>+++ b/fs/nfs/callback.c
>@@ -105,7 +105,7 @@ static void nfs_callback_svc(struct svc_rqst *rqstp)
>  */
> int nfs_callback_up(void)
> {
>-	struct svc_serv *serv;
>+	struct svc_serv *serv = NULL;
> 	int ret = 0;
> 
> 	lock_kernel();
>@@ -122,24 +122,24 @@ int nfs_callback_up(void)
> 	ret = svc_create_xprt(serv, "tcp", nfs_callback_set_tcpport,
> 			      SVC_SOCK_ANONYMOUS);
> 	if (ret <= 0)
>-		goto out_destroy;
>+		goto out_err;
> 	nfs_callback_tcpport = ret;
> 	dprintk("Callback port = 0x%x\n", nfs_callback_tcpport);
> 
> 	ret = svc_create_thread(nfs_callback_svc, serv);
> 	if (ret < 0)
>-		goto out_destroy;
>+		goto out_err;
> 	nfs_callback_info.serv = serv;
> 	wait_for_completion(&nfs_callback_info.started);
> out:
>+	if (serv)
>+		svc_destroy(serv);
> 	mutex_unlock(&nfs_callback_mutex);
> 	unlock_kernel();
> 	return ret;
>-out_destroy:
>+out_err:
> 	dprintk("Couldn't create callback socket or server thread; err = %d\n",
> 		ret);
>-	svc_destroy(serv);
>-out_err:
> 	nfs_callback_info.users--;
> 	goto out;
> }
>-- 
>1.5.3.8
>

Actions: View | Diff

Attachments on bug 423521: 294150 | 294151 | 294563 | 295296 | 302874