Skip to contentSkip to conversation

kfree in fs/nfsd/nfs4xdr.c:savemem()

PATCH In review Benny Halevy 2 participants 3 messages Dec 27, 2006 – Dec 31, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 kfree in fs/nfsd/nfs4xdr.c:savemem()

The wrong pointer is being kfree'd in savemem() when defer_free returns with an error.

Benny

fs/nfsd/nfs4xdr.c 2006-12-25 09:15:58.000000000 +0200 renamed from fs/nfsd/nfs4xdr.c 2006-12-18 19:18:45.000000000 +0200 +1 -1
diff -rp -U 11 a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c
@@ -197,23 +197,23 @@ static char *savemem(struct nfsd4_compou
        if (p == argp->tmp) {
                new = kmalloc(nbytes, GFP_KERNEL);
                if (!new) return NULL;
                p = new;
                memcpy(p, argp->tmp, nbytes);
        } else {
                if (p != argp->tmpp)
                        BUG();
                argp->tmpp = NULL;
        }
        if (defer_free(argp, kfree, p)) {
-               kfree(new);
+               kfree(p);
                return NULL;
        } else
                return (char *)p;
 }
 
 
 static int
 nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval)
 {
        u32 bmlen;
        DECODE_HEAD;

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#1 lore raw
Quoted 2 lines · Benny Halevy

The wrong pointer is being kfree'd in savemem() when defer_free returns with an error.

Makes sense to me, thanks!

That makes "new" unnecessary. Any objection to the following? (untested).

Since this seems to be a rare memory leak, I'm inclined to leave this till after 2.6.20.

--b.

fs/nfsd/nfs4xdr.c +3 -5
index dabd527..91284a0 100644
@@ -199,18 +199,16 @@ defer_free(struct nfsd4_compoundargs *argp,
 
 static char *savemem(struct nfsd4_compoundargs *argp, __be32 *p, int nbytes)
 {
-	void *new = NULL;
 	if (p == argp->tmp) {
-		new = kmalloc(nbytes, GFP_KERNEL);
-		if (!new) return NULL;
-		p = new;
+		p = kmalloc(nbytes, GFP_KERNEL);
+		if (!p) return NULL;
 		memcpy(p, argp->tmp, nbytes);
 	} else {
 		BUG_ON(p != argp->tmpp);
 		argp->tmpp = NULL;
 	}
 	if (defer_free(argp, kfree, p)) {
-		kfree(new);
+		kfree(p);
 		return NULL;
 	} else
 		return (char *)p;

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs

#2 lore raw
Quoted 9 lines · J. Bruce Fields
Quoted 2 lines · Benny Halevy

The wrong pointer is being kfree'd in savemem() when defer_free returns with an error.

Makes sense to me, thanks!

That makes "new" unnecessary. Any objection to the following? (untested).

This seems to be right on the spot...

Quoted 3 lines

Since this seems to be a rare memory leak, I'm inclined to leave this till after 2.6.20.

Agreed.

Quoted 20 lines

--b.

fs/nfsd/nfs4xdr.c +2 -4
index dabd527..91284a0 100644
@@ -199,18 +199,16 @@ defer_free(struct nfsd4_compoundargs *argp,
 
 static char *savemem(struct nfsd4_compoundargs *argp, __be32 *p, int nbytes)
 {
-	void *new = NULL;
 	if (p == argp->tmp) {
-		new = kmalloc(nbytes, GFP_KERNEL);
-		if (!new) return NULL;
-		p = new;
+		p = kmalloc(nbytes, GFP_KERNEL);
+		if (!p) return NULL;
 		memcpy(p, argp->tmp, nbytes);

fs/nfsd/nfs4xdr.c defer_free(struct nfsd4_compoundargs *argp, · line 214

Quoted 10 lines
fs/nfsd/nfs4xdr.c +1 -1
 	} else {
 		BUG_ON(p != argp->tmpp);
 		argp->tmpp = NULL;
 	}
 	if (defer_free(argp, kfree, p)) {
-		kfree(new);
+		kfree(p);
 		return NULL;
 	} else
 		return (char *)p;

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs