eventfs: Check for NULL ef in eventfs_set_attr()

[v6.6][PATCH] eventfs: Check for NULL ef in eventfs_set_attr()

Posted by Steven Rostedt 2 years, 1 month ago

From: "Steven Rostedt (Google)" <rostedt@goodmis.org>

The top level events directory dentry does not have a d_fsdata set to a
eventfs_file pointer. This dentry is still passed to eventfs_set_attr().
It can not assume that the d_fsdata is set. Check for that.

Link: https://lore.kernel.org/all/20231112104158.6638-1-milian.wolff@kdab.com/

Fixes: 9aaee3eebc91 ("eventfs: Save ownership and mode")
Reported-by: Milian Wolff <milian.wolff@kdab.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---

Note: This only affects 6.6 as the code in 6.7 here was rewritten.
   I tested 6.7 and it does not have this bug.

 fs/tracefs/event_inode.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/tracefs/event_inode.c b/fs/tracefs/event_inode.c
index 5fcfb634fec2..efbdc47c74dc 100644
--- a/fs/tracefs/event_inode.c
+++ b/fs/tracefs/event_inode.c
@@ -113,14 +113,14 @@ static int eventfs_set_attr(struct mnt_idmap *idmap, struct dentry *dentry,
 
 	mutex_lock(&eventfs_mutex);
 	ef = dentry->d_fsdata;
-	if (ef->is_freed) {
+	if (ef && ef->is_freed) {
 		/* Do not allow changes if the event is about to be removed. */
 		mutex_unlock(&eventfs_mutex);
 		return -ENODEV;
 	}
 
 	ret = simple_setattr(idmap, dentry, iattr);
-	if (!ret)
+	if (!ret && ef)
 		update_attr(ef, iattr);
 	mutex_unlock(&eventfs_mutex);
 	return ret;
-- 
2.42.0

Re: [v6.6][PATCH] eventfs: Check for NULL ef in eventfs_set_attr()

Posted by Milian Wolff 2 years, 1 month ago

On Sonntag, 12. November 2023 18:18:17 CET Steven Rostedt wrote:
> From: "Steven Rostedt (Google)" <rostedt@goodmis.org>
> 
> The top level events directory dentry does not have a d_fsdata set to a
> eventfs_file pointer. This dentry is still passed to eventfs_set_attr().
> It can not assume that the d_fsdata is set. Check for that.
> 
> Link:
> https://lore.kernel.org/all/20231112104158.6638-1-milian.wolff@kdab.com/

Thanks you Steven,

this works like a charm.

Tested-by: Milian Wolff <milian.wolff@kdab.com>

> Fixes: 9aaee3eebc91 ("eventfs: Save ownership and mode")
> Reported-by: Milian Wolff <milian.wolff@kdab.com>
> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
> ---
> 
> Note: This only affects 6.6 as the code in 6.7 here was rewritten.
>    I tested 6.7 and it does not have this bug.
> 
>  fs/tracefs/event_inode.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/tracefs/event_inode.c b/fs/tracefs/event_inode.c
> index 5fcfb634fec2..efbdc47c74dc 100644
> --- a/fs/tracefs/event_inode.c
> +++ b/fs/tracefs/event_inode.c
> @@ -113,14 +113,14 @@ static int eventfs_set_attr(struct mnt_idmap *idmap,
> struct dentry *dentry,
> 
>  	mutex_lock(&eventfs_mutex);
>  	ef = dentry->d_fsdata;
> -	if (ef->is_freed) {
> +	if (ef && ef->is_freed) {
>  		/* Do not allow changes if the event is about to be 
removed. */
>  		mutex_unlock(&eventfs_mutex);
>  		return -ENODEV;
>  	}
> 
>  	ret = simple_setattr(idmap, dentry, iattr);
> -	if (!ret)
> +	if (!ret && ef)
>  		update_attr(ef, iattr);
>  	mutex_unlock(&eventfs_mutex);
>  	return ret;


-- 
Milian Wolff | milian.wolff@kdab.com | Senior Software Engineer
KDAB (Deutschland) GmbH, a KDAB Group company
Tel: +49-30-521325470
KDAB - The Qt, C++ and OpenGL Experts

Re: [v6.6][PATCH] eventfs: Check for NULL ef in eventfs_set_attr()

Posted by Steven Rostedt 2 years, 1 month ago

On Sun, 12 Nov 2023 12:18:17 -0500
Steven Rostedt <rostedt@goodmis.org> wrote:

> From: "Steven Rostedt (Google)" <rostedt@goodmis.org>
> 
> The top level events directory dentry does not have a d_fsdata set to a
> eventfs_file pointer. This dentry is still passed to eventfs_set_attr().
> It can not assume that the d_fsdata is set. Check for that.
> 
> Link: https://lore.kernel.org/all/20231112104158.6638-1-milian.wolff@kdab.com/
> 
> Fixes: 9aaee3eebc91 ("eventfs: Save ownership and mode")
> Reported-by: Milian Wolff <milian.wolff@kdab.com>
> Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>

I guess I should have included the "Cc: stable" tag. This is a bit
different patch since it's not going to land in upstream, as the bug
doesn't exist there.

I can resend if needed.

-- Steve