Btrfs: don't leak reloc root nodes on error
diff mbox

Message ID 1472844343-19086-1-git-send-email-jbacik@fb.com
State Accepted
Headers show

Commit Message

Josef Bacik Sept. 2, 2016, 7:25 p.m. UTC
We don't track the reloc roots in any sort of normal way, so the only way the
root/commit_root nodes get free'd is if the relocation finishes successfully and
the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
Thanks,

Signed-off-by: Josef Bacik <jbacik@fb.com>
---
 fs/btrfs/relocation.c | 4 ++++
 1 file changed, 4 insertions(+)

Comments

Liu Bo Sept. 3, 2016, 1:08 a.m. UTC | #1
On Fri, Sep 02, 2016 at 03:25:43PM -0400, Josef Bacik wrote:
> We don't track the reloc roots in any sort of normal way, so the only way the
> root/commit_root nodes get free'd is if the relocation finishes successfully and
> the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
> Thanks,

Looks good.

> 
> Signed-off-by: Josef Bacik <jbacik@fb.com>
> ---
>  fs/btrfs/relocation.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> index 7fc6ea7..62dfc2c 100644
> --- a/fs/btrfs/relocation.c
> +++ b/fs/btrfs/relocation.c
> @@ -2349,6 +2349,10 @@ void free_reloc_roots(struct list_head *list)
>  	while (!list_empty(list)) {
>  		reloc_root = list_entry(list->next, struct btrfs_root,
>  					root_list);
> +		free_extent_buffer(reloc_root->node);
> +		free_extent_buffer(reloc_root->commit_root);
> +		reloc_root->node = NULL;
> +		reloc_root->commit_root = NULL;

What about reloc_root itself?

Thanks,

-liubo

>  		__del_reloc_root(reloc_root);
>  	}
>  }
> -- 
> 2.7.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Sterba Sept. 5, 2016, 3:20 p.m. UTC | #2
On Fri, Sep 02, 2016 at 06:08:35PM -0700, Liu Bo wrote:
> On Fri, Sep 02, 2016 at 03:25:43PM -0400, Josef Bacik wrote:
> > We don't track the reloc roots in any sort of normal way, so the only way the
> > root/commit_root nodes get free'd is if the relocation finishes successfully and
> > the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
> > Thanks,
> 
> Looks good.
> 
> > 
> > Signed-off-by: Josef Bacik <jbacik@fb.com>
> > ---
> >  fs/btrfs/relocation.c | 4 ++++
> >  1 file changed, 4 insertions(+)
> > 
> > diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> > index 7fc6ea7..62dfc2c 100644
> > --- a/fs/btrfs/relocation.c
> > +++ b/fs/btrfs/relocation.c
> > @@ -2349,6 +2349,10 @@ void free_reloc_roots(struct list_head *list)
> >  	while (!list_empty(list)) {
> >  		reloc_root = list_entry(list->next, struct btrfs_root,
> >  					root_list);
> > +		free_extent_buffer(reloc_root->node);
> > +		free_extent_buffer(reloc_root->commit_root);
> > +		reloc_root->node = NULL;
> > +		reloc_root->commit_root = NULL;
> 
> What about reloc_root itself?

> >  		__del_reloc_root(reloc_root);

It's deleted at the end of __del_reloc_root
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Liu Bo Sept. 6, 2016, 7:22 p.m. UTC | #3
On Mon, Sep 05, 2016 at 05:20:53PM +0200, David Sterba wrote:
> On Fri, Sep 02, 2016 at 06:08:35PM -0700, Liu Bo wrote:
> > On Fri, Sep 02, 2016 at 03:25:43PM -0400, Josef Bacik wrote:
> > > We don't track the reloc roots in any sort of normal way, so the only way the
> > > root/commit_root nodes get free'd is if the relocation finishes successfully and
> > > the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
> > > Thanks,
> > 
> > Looks good.
> > 
> > > 
> > > Signed-off-by: Josef Bacik <jbacik@fb.com>
> > > ---
> > >  fs/btrfs/relocation.c | 4 ++++
> > >  1 file changed, 4 insertions(+)
> > > 
> > > diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> > > index 7fc6ea7..62dfc2c 100644
> > > --- a/fs/btrfs/relocation.c
> > > +++ b/fs/btrfs/relocation.c
> > > @@ -2349,6 +2349,10 @@ void free_reloc_roots(struct list_head *list)
> > >  	while (!list_empty(list)) {
> > >  		reloc_root = list_entry(list->next, struct btrfs_root,
> > >  					root_list);
> > > +		free_extent_buffer(reloc_root->node);
> > > +		free_extent_buffer(reloc_root->commit_root);
> > > +		reloc_root->node = NULL;
> > > +		reloc_root->commit_root = NULL;
> > 
> > What about reloc_root itself?
> 
> > >  		__del_reloc_root(reloc_root);
> 
> It's deleted at the end of __del_reloc_root

__del_reloc_root() is dealing with the associated rb_node, not
reloc_root.

I posted a patch to free reloc_root while free'ing its corresponding fs
root, but it'd be good if we can free it here.

Thanks,

-liubo
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Sterba Sept. 21, 2016, 2 p.m. UTC | #4
On Tue, Sep 06, 2016 at 12:22:01PM -0700, Liu Bo wrote:
> On Mon, Sep 05, 2016 at 05:20:53PM +0200, David Sterba wrote:
> > On Fri, Sep 02, 2016 at 06:08:35PM -0700, Liu Bo wrote:
> > > On Fri, Sep 02, 2016 at 03:25:43PM -0400, Josef Bacik wrote:
> > > > We don't track the reloc roots in any sort of normal way, so the only way the
> > > > root/commit_root nodes get free'd is if the relocation finishes successfully and
> > > > the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
> > > > Thanks,
> > > 
> > > Looks good.
> > > 
> > > > 
> > > > Signed-off-by: Josef Bacik <jbacik@fb.com>
> > > > ---
> > > >  fs/btrfs/relocation.c | 4 ++++
> > > >  1 file changed, 4 insertions(+)
> > > > 
> > > > diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> > > > index 7fc6ea7..62dfc2c 100644
> > > > --- a/fs/btrfs/relocation.c
> > > > +++ b/fs/btrfs/relocation.c
> > > > @@ -2349,6 +2349,10 @@ void free_reloc_roots(struct list_head *list)
> > > >  	while (!list_empty(list)) {
> > > >  		reloc_root = list_entry(list->next, struct btrfs_root,
> > > >  					root_list);
> > > > +		free_extent_buffer(reloc_root->node);
> > > > +		free_extent_buffer(reloc_root->commit_root);
> > > > +		reloc_root->node = NULL;
> > > > +		reloc_root->commit_root = NULL;
> > > 
> > > What about reloc_root itself?
> > 
> > > >  		__del_reloc_root(reloc_root);
> > 
> > It's deleted at the end of __del_reloc_root
> 
> __del_reloc_root() is dealing with the associated rb_node, not
> reloc_root.
> 
> I posted a patch to free reloc_root while free'ing its corresponding fs
> root, but it'd be good if we can free it here.

Can you please point me to the patch?
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Liu Bo Sept. 21, 2016, 10:45 p.m. UTC | #5
On Wed, Sep 21, 2016 at 04:00:05PM +0200, David Sterba wrote:
> On Tue, Sep 06, 2016 at 12:22:01PM -0700, Liu Bo wrote:
> > On Mon, Sep 05, 2016 at 05:20:53PM +0200, David Sterba wrote:
> > > On Fri, Sep 02, 2016 at 06:08:35PM -0700, Liu Bo wrote:
> > > > On Fri, Sep 02, 2016 at 03:25:43PM -0400, Josef Bacik wrote:
> > > > > We don't track the reloc roots in any sort of normal way, so the only way the
> > > > > root/commit_root nodes get free'd is if the relocation finishes successfully and
> > > > > the reloc root is deleted.  Fix this by free'ing them in free_reloc_roots.
> > > > > Thanks,
> > > > 
> > > > Looks good.
> > > > 
> > > > > 
> > > > > Signed-off-by: Josef Bacik <jbacik@fb.com>
> > > > > ---
> > > > >  fs/btrfs/relocation.c | 4 ++++
> > > > >  1 file changed, 4 insertions(+)
> > > > > 
> > > > > diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
> > > > > index 7fc6ea7..62dfc2c 100644
> > > > > --- a/fs/btrfs/relocation.c
> > > > > +++ b/fs/btrfs/relocation.c
> > > > > @@ -2349,6 +2349,10 @@ void free_reloc_roots(struct list_head *list)
> > > > >  	while (!list_empty(list)) {
> > > > >  		reloc_root = list_entry(list->next, struct btrfs_root,
> > > > >  					root_list);
> > > > > +		free_extent_buffer(reloc_root->node);
> > > > > +		free_extent_buffer(reloc_root->commit_root);
> > > > > +		reloc_root->node = NULL;
> > > > > +		reloc_root->commit_root = NULL;
> > > > 
> > > > What about reloc_root itself?
> > > 
> > > > >  		__del_reloc_root(reloc_root);
> > > 
> > > It's deleted at the end of __del_reloc_root
> > 
> > __del_reloc_root() is dealing with the associated rb_node, not
> > reloc_root.
> > 
> > I posted a patch to free reloc_root while free'ing its corresponding fs
> > root, but it'd be good if we can free it here.
> 
> Can you please point me to the patch?

Btrfs: fix memory leak of reloc_root
https://patchwork.kernel.org/patch/9238383/

Thanks,

-liubo
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch
diff mbox

diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 7fc6ea7..62dfc2c 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2349,6 +2349,10 @@  void free_reloc_roots(struct list_head *list)
 	while (!list_empty(list)) {
 		reloc_root = list_entry(list->next, struct btrfs_root,
 					root_list);
+		free_extent_buffer(reloc_root->node);
+		free_extent_buffer(reloc_root->commit_root);
+		reloc_root->node = NULL;
+		reloc_root->commit_root = NULL;
 		__del_reloc_root(reloc_root);
 	}
 }