Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1455664 > unrolled thread

Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to overlay inode

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to  overlay inode Eryu Guan <eguan@redhat.com> - 2016-08-03 09:40 +0200
    Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to  overlay inode Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-08-03 10:00 +0200
      Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to  overlay inode Eryu Guan <eguan@redhat.com> - 2016-08-03 10:20 +0200
        Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to  overlay inode Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2016-08-03 13:20 +0200

#1455664 — Re: [PATCH 4.4 133/146] ovl: Copy up underlying inodes ->i_mode to overlay inode

On Mon, Jul 25, 2016 at 01:56:28PM -0700, Greg Kroah-Hartman wrote:
> 4.4-stable review patch.  If anyone has any objections, please let me know.
> 
> ------------------
> 
> From: Vivek Goyal <vgoyal@redhat.com>
> 
> commit 07a2daab49c549a37b5b744cbebb6e3f445f12bc upstream.
> 
> Right now when a new overlay inode is created, we initialize overlay
> inode's ->i_mode from underlying inode ->i_mode but we retain only
> file type bits (S_IFMT) and discard permission bits.
> 
> This patch changes it and retains permission bits too. This should allow
> overlay to do permission checks on overlay inode itself in task context.
> 
> [SzM] It also fixes clearing suid/sgid bits on write.

This patch introduced a hang when writing to suid file, fstests
generic/355 could reproduce the hang easily, it only failed the test
without this patch and didn't hang the kernel.

Should we skip it for now and wait for a further fix?

(The 4.6-stable tree faces the same question)

Thanks,
Eryu

> 
> Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
> Reported-by: Eryu Guan <eguan@redhat.com>
> Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
> Fixes: 4bacc9c9234c ("overlayfs: Make f_path always point to the overlay and f_inode to the underlay")
> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> 
> ---
>  fs/overlayfs/inode.c     |    3 +--
>  fs/overlayfs/overlayfs.h |    1 +
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> --- a/fs/overlayfs/inode.c
> +++ b/fs/overlayfs/inode.c
> @@ -412,12 +412,11 @@ struct inode *ovl_new_inode(struct super
>  	if (!inode)
>  		return NULL;
>  
> -	mode &= S_IFMT;
> -
>  	inode->i_ino = get_next_ino();
>  	inode->i_mode = mode;
>  	inode->i_flags |= S_NOATIME | S_NOCMTIME;
>  
> +	mode &= S_IFMT;
>  	switch (mode) {
>  	case S_IFDIR:
>  		inode->i_private = oe;
> --- a/fs/overlayfs/overlayfs.h
> +++ b/fs/overlayfs/overlayfs.h
> @@ -181,6 +181,7 @@ static inline void ovl_copyattr(struct i
>  {
>  	to->i_uid = from->i_uid;
>  	to->i_gid = from->i_gid;
> +	to->i_mode = from->i_mode;
>  }
>  
>  /* dir.c */
> 
> 

[toc] | [next] | [standalone]

#1455670

On Wed, Aug 03, 2016 at 03:36:44PM +0800, Eryu Guan wrote:
> On Mon, Jul 25, 2016 at 01:56:28PM -0700, Greg Kroah-Hartman wrote:
> > 4.4-stable review patch.  If anyone has any objections, please let me know.
> > 
> > ------------------
> > 
> > From: Vivek Goyal <vgoyal@redhat.com>
> > 
> > commit 07a2daab49c549a37b5b744cbebb6e3f445f12bc upstream.
> > 
> > Right now when a new overlay inode is created, we initialize overlay
> > inode's ->i_mode from underlying inode ->i_mode but we retain only
> > file type bits (S_IFMT) and discard permission bits.
> > 
> > This patch changes it and retains permission bits too. This should allow
> > overlay to do permission checks on overlay inode itself in task context.
> > 
> > [SzM] It also fixes clearing suid/sgid bits on write.
> 
> This patch introduced a hang when writing to suid file, fstests
> generic/355 could reproduce the hang easily, it only failed the test
> without this patch and didn't hang the kernel.
> 
> Should we skip it for now and wait for a further fix?

Does Linus's tree have the same problem?

> (The 4.6-stable tree faces the same question)

Are we just missing a patch to be applied here?

And this is already in the released stable kernels...

thanks,

greg k-h

[toc] | [prev] | [next] | [standalone]

#1455687

On Wed, Aug 03, 2016 at 09:45:06AM +0200, Greg Kroah-Hartman wrote:
> On Wed, Aug 03, 2016 at 03:36:44PM +0800, Eryu Guan wrote:
> > On Mon, Jul 25, 2016 at 01:56:28PM -0700, Greg Kroah-Hartman wrote:
> > > 4.4-stable review patch.  If anyone has any objections, please let me know.
> > > 
> > > ------------------
> > > 
> > > From: Vivek Goyal <vgoyal@redhat.com>
> > > 
> > > commit 07a2daab49c549a37b5b744cbebb6e3f445f12bc upstream.
> > > 
> > > Right now when a new overlay inode is created, we initialize overlay
> > > inode's ->i_mode from underlying inode ->i_mode but we retain only
> > > file type bits (S_IFMT) and discard permission bits.
> > > 
> > > This patch changes it and retains permission bits too. This should allow
> > > overlay to do permission checks on overlay inode itself in task context.
> > > 
> > > [SzM] It also fixes clearing suid/sgid bits on write.
> > 
> > This patch introduced a hang when writing to suid file, fstests
> > generic/355 could reproduce the hang easily, it only failed the test
> > without this patch and didn't hang the kernel.
> > 
> > Should we skip it for now and wait for a further fix?
> 
> Does Linus's tree have the same problem?

Yes, 4.7 kernel hangs as well.

> 
> > (The 4.6-stable tree faces the same question)
> 
> Are we just missing a patch to be applied here?

AFAICT, no patch is available to fix it yet. But I've reported the bug
to Miklos (in RH bugzilla).

> 
> And this is already in the released stable kernels...

This patch is not in 4.7-rc7, it first appears in 4.7 kernel, seems like
the soak time is not long enough to let testings finish :) I finished my
4.7 kernel testing and identified the culprit this Monday.

Thanks,
Eryu

[toc] | [prev] | [next] | [standalone]

#1455749

On Wed, Aug 03, 2016 at 04:04:55PM +0800, Eryu Guan wrote:
> On Wed, Aug 03, 2016 at 09:45:06AM +0200, Greg Kroah-Hartman wrote:
> > On Wed, Aug 03, 2016 at 03:36:44PM +0800, Eryu Guan wrote:
> > > On Mon, Jul 25, 2016 at 01:56:28PM -0700, Greg Kroah-Hartman wrote:
> > > > 4.4-stable review patch.  If anyone has any objections, please let me know.
> > > > 
> > > > ------------------
> > > > 
> > > > From: Vivek Goyal <vgoyal@redhat.com>
> > > > 
> > > > commit 07a2daab49c549a37b5b744cbebb6e3f445f12bc upstream.
> > > > 
> > > > Right now when a new overlay inode is created, we initialize overlay
> > > > inode's ->i_mode from underlying inode ->i_mode but we retain only
> > > > file type bits (S_IFMT) and discard permission bits.
> > > > 
> > > > This patch changes it and retains permission bits too. This should allow
> > > > overlay to do permission checks on overlay inode itself in task context.
> > > > 
> > > > [SzM] It also fixes clearing suid/sgid bits on write.
> > > 
> > > This patch introduced a hang when writing to suid file, fstests
> > > generic/355 could reproduce the hang easily, it only failed the test
> > > without this patch and didn't hang the kernel.
> > > 
> > > Should we skip it for now and wait for a further fix?
> > 
> > Does Linus's tree have the same problem?
> 
> Yes, 4.7 kernel hangs as well.
> 
> > 
> > > (The 4.6-stable tree faces the same question)
> > 
> > Are we just missing a patch to be applied here?
> 
> AFAICT, no patch is available to fix it yet. But I've reported the bug
> to Miklos (in RH bugzilla).

Ok, thanks, just mark whatever fix you all come up with, with a stable
marking, and I will pick up the fix when it gets merged to Linus's tree.

thanks,

greg k-h

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web