public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] seqfile: fix uninitialized memory allocation in mounts_open()
       [not found] <20070815085637.GA19293@mail.ustc.edu.cn>
@ 2007-08-15  8:56 ` Fengguang Wu
  2007-08-15  9:04   ` Alexey Dobriyan
       [not found] ` <20070815090116.GA19893@mail.ustc.edu.cn>
       [not found] ` <20070815090549.GB19893@mail.ustc.edu.cn>
  2 siblings, 1 reply; 5+ messages in thread
From: Fengguang Wu @ 2007-08-15  8:56 UTC (permalink / raw)
  To: Andrew Morton; +Cc: linux-kernel, Al Viro

proc_mounts.m.buf is not explicitly zeroed at allocation time, which
may later confuse the seqfile code and triggle an kfree(m->buf).

Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn>
---

--- linux.orig/fs/proc/base.c
+++ linux/fs/proc/base.c
@@ -380,7 +380,7 @@ static int mounts_open(struct inode *ino
 
 	if (ns) {
 		ret = -ENOMEM;
-		p = kmalloc(sizeof(struct proc_mounts), GFP_KERNEL);
+		p = kzalloc(sizeof(struct proc_mounts), GFP_KERNEL);
 		if (p) {
 			file->private_data = &p->m;
 			ret = seq_open(file, &mounts_op);


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] seqfile: fix uninitialized memory allocation in mounts_open()
       [not found] ` <20070815090116.GA19893@mail.ustc.edu.cn>
@ 2007-08-15  9:01   ` Fengguang Wu
  0 siblings, 0 replies; 5+ messages in thread
From: Fengguang Wu @ 2007-08-15  9:01 UTC (permalink / raw)
  To: Andrew Morton; +Cc: linux-kernel, Al Viro

On Wed, Aug 15, 2007 at 04:56:37PM +0800, Fengguang Wu wrote:
> proc_mounts.m.buf is not explicitly zeroed at allocation time, which
> may later confuse the seqfile code and triggle an kfree(m->buf).

It's weird it did not show up in real world.  Could I be wrong?

> --- linux.orig/fs/proc/base.c
> +++ linux/fs/proc/base.c
> @@ -380,7 +380,7 @@ static int mounts_open(struct inode *ino
>  
>  	if (ns) {
>  		ret = -ENOMEM;
> -		p = kmalloc(sizeof(struct proc_mounts), GFP_KERNEL);
> +		p = kzalloc(sizeof(struct proc_mounts), GFP_KERNEL);
>  		if (p) {
>  			file->private_data = &p->m;
>  			ret = seq_open(file, &mounts_op);
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] seqfile: fix uninitialized memory allocation in mounts_open()
  2007-08-15  8:56 ` [PATCH] seqfile: fix uninitialized memory allocation in mounts_open() Fengguang Wu
@ 2007-08-15  9:04   ` Alexey Dobriyan
       [not found]     ` <20070815090855.GB19293@mail.ustc.edu.cn>
  0 siblings, 1 reply; 5+ messages in thread
From: Alexey Dobriyan @ 2007-08-15  9:04 UTC (permalink / raw)
  To: Fengguang Wu; +Cc: Andrew Morton, linux-kernel, Al Viro

On 8/15/07, Fengguang Wu <wfg@mail.ustc.edu.cn> wrote:
> proc_mounts.m.buf is not explicitly zeroed at allocation time, which
> may later confuse the seqfile code and triggle an kfree(m->buf).

It's cleared in seq_open().

> --- linux.orig/fs/proc/base.c
> +++ linux/fs/proc/base.c
> @@ -380,7 +380,7 @@ static int mounts_open(struct inode *ino
>
>  	if (ns) {
>  		ret = -ENOMEM;
> -		p = kmalloc(sizeof(struct proc_mounts), GFP_KERNEL);
> +		p = kzalloc(sizeof(struct proc_mounts), GFP_KERNEL);
>  		if (p) {
>  			file->private_data = &p->m;
>  			ret = seq_open(file, &mounts_op);

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [PATCH] seqfile: merge duplite code to seq_open_private()
       [not found] ` <20070815090549.GB19893@mail.ustc.edu.cn>
@ 2007-08-15  9:05   ` Fengguang Wu
  0 siblings, 0 replies; 5+ messages in thread
From: Fengguang Wu @ 2007-08-15  9:05 UTC (permalink / raw)
  To: Andrew Morton; +Cc: linux-kernel, Al Viro

There are dozens of calls to seq_open() that need to set m->private.
Introduce seq_open_private() to remove code duplications.

Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn>
---
Andrew: more conversions will be made if this patch is ok.

 fs/nfs/client.c          |   24 ++----------------------
 fs/seq_file.c            |   10 ++++++++++
 include/linux/seq_file.h |    1 +
 3 files changed, 13 insertions(+), 22 deletions(-)

--- linux-2.6.23-rc2-mm2.orig/include/linux/seq_file.h
+++ linux-2.6.23-rc2-mm2/include/linux/seq_file.h
@@ -33,6 +33,7 @@ struct seq_operations {
 };
 
 int seq_open(struct file *, const struct seq_operations *);
+int seq_open_private(struct file *, struct seq_operations *, void *);
 ssize_t seq_read(struct file *, char __user *, size_t, loff_t *);
 loff_t seq_lseek(struct file *, loff_t, int);
 int seq_release(struct inode *, struct file *);
--- linux-2.6.23-rc2-mm2.orig/fs/seq_file.c
+++ linux-2.6.23-rc2-mm2/fs/seq_file.c
@@ -53,6 +53,16 @@ int seq_open(struct file *file, const st
 }
 EXPORT_SYMBOL(seq_open);
 
+int seq_open_private(struct file *file, struct seq_operations *op, void *data)
+{
+	int ret = seq_open(file, op);
+
+	if (!ret)
+		((struct seq_file *)file->private_data)->private = data;
+
+	return ret;
+}
+EXPORT_SYMBOL(seq_open_private);
 /**
  *	seq_read -	->read() method for sequential files.
  *	@file: the file to read from
--- linux-2.6.23-rc2-mm2.orig/fs/nfs/client.c
+++ linux-2.6.23-rc2-mm2/fs/nfs/client.c
@@ -1188,17 +1188,7 @@ static const struct file_operations nfs_
  */
 static int nfs_server_list_open(struct inode *inode, struct file *file)
 {
-	struct seq_file *m;
-	int ret;
-
-	ret = seq_open(file, &nfs_server_list_ops);
-	if (ret < 0)
-		return ret;
-
-	m = file->private_data;
-	m->private = PDE(inode)->data;
-
-	return 0;
+	return seq_open_private(file, &nfs_server_list_ops, PDE(inode)->data);
 }
 
 /*
@@ -1258,17 +1248,7 @@ static int nfs_server_list_show(struct s
  */
 static int nfs_volume_list_open(struct inode *inode, struct file *file)
 {
-	struct seq_file *m;
-	int ret;
-
-	ret = seq_open(file, &nfs_volume_list_ops);
-	if (ret < 0)
-		return ret;
-
-	m = file->private_data;
-	m->private = PDE(inode)->data;
-
-	return 0;
+	return seq_open_private(file, &nfs_volume_list_ops, PDE(inode)->data);
 }
 
 /*


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH] seqfile: fix uninitialized memory allocation in mounts_open()
       [not found]     ` <20070815090855.GB19293@mail.ustc.edu.cn>
@ 2007-08-15  9:08       ` Fengguang Wu
  0 siblings, 0 replies; 5+ messages in thread
From: Fengguang Wu @ 2007-08-15  9:08 UTC (permalink / raw)
  To: Alexey Dobriyan; +Cc: Andrew Morton, linux-kernel, Al Viro

On Wed, Aug 15, 2007 at 01:04:05PM +0400, Alexey Dobriyan wrote:
> On 8/15/07, Fengguang Wu <wfg@mail.ustc.edu.cn> wrote:
> > proc_mounts.m.buf is not explicitly zeroed at allocation time, which
> > may later confuse the seqfile code and triggle an kfree(m->buf).
> 
> It's cleared in seq_open().

So it is :)

> > --- linux.orig/fs/proc/base.c
> > +++ linux/fs/proc/base.c
> > @@ -380,7 +380,7 @@ static int mounts_open(struct inode *ino
> >
> >  	if (ns) {
> >  		ret = -ENOMEM;
> > -		p = kmalloc(sizeof(struct proc_mounts), GFP_KERNEL);
> > +		p = kzalloc(sizeof(struct proc_mounts), GFP_KERNEL);
> >  		if (p) {
> >  			file->private_data = &p->m;
> >  			ret = seq_open(file, &mounts_op);


^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2007-08-15  9:09 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <20070815085637.GA19293@mail.ustc.edu.cn>
2007-08-15  8:56 ` [PATCH] seqfile: fix uninitialized memory allocation in mounts_open() Fengguang Wu
2007-08-15  9:04   ` Alexey Dobriyan
     [not found]     ` <20070815090855.GB19293@mail.ustc.edu.cn>
2007-08-15  9:08       ` Fengguang Wu
     [not found] ` <20070815090116.GA19893@mail.ustc.edu.cn>
2007-08-15  9:01   ` Fengguang Wu
     [not found] ` <20070815090549.GB19893@mail.ustc.edu.cn>
2007-08-15  9:05   ` [PATCH] seqfile: merge duplite code to seq_open_private() Fengguang Wu

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox