2018-12-24 13:08:18

by Sahitya Tummala

[permalink] [raw]
Subject: [PATCH] f2fs: fix use-after-free issue with sbi->stat_info

iput() on sbi->node_inode can update sbi->stat_info
in the below context, if the f2fs_write_checkpoint()
has failed with error.

f2fs_balance_fs_bg+0x1ac/0x1ec
f2fs_write_node_pages+0x4c/0x260
do_writepages+0x80/0xbc
__writeback_single_inode+0xdc/0x4ac
writeback_single_inode+0x9c/0x144
write_inode_now+0xc4/0xec
iput+0x194/0x22c
f2fs_put_super+0x11c/0x1e8
generic_shutdown_super+0x70/0xf4
kill_block_super+0x2c/0x5c
kill_f2fs_super+0x44/0x50
deactivate_locked_super+0x60/0x8c
deactivate_super+0x68/0x74
cleanup_mnt+0x40/0x78

Fix this by moving f2fs_destroy_stats() further below iput().

Signed-off-by: Sahitya Tummala <[email protected]>
---
fs/f2fs/super.c | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index e184ad4e..df41a3a 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -1058,9 +1058,6 @@ static void f2fs_put_super(struct super_block *sb)
f2fs_write_checkpoint(sbi, &cpc);
}

- /* f2fs_write_checkpoint can update stat informaion */
- f2fs_destroy_stats(sbi);
-
/*
* normally superblock is clean, so we need to release this.
* In addition, EIO will skip do checkpoint, we need this as well.
@@ -1080,6 +1077,12 @@ static void f2fs_put_super(struct super_block *sb)
iput(sbi->node_inode);
iput(sbi->meta_inode);

+ /*
+ * iput() can update stat information, if f2fs_write_checkpoint()
+ * above failed with error.
+ */
+ f2fs_destroy_stats(sbi);
+
/* destroy f2fs internal modules */
f2fs_destroy_node_manager(sbi);
f2fs_destroy_segment_manager(sbi);
--
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project.



2018-12-25 03:08:03

by Chao Yu

[permalink] [raw]
Subject: Re: [PATCH] f2fs: fix use-after-free issue with sbi->stat_info

On 2018/12/24 21:06, Sahitya Tummala wrote:
> iput() on sbi->node_inode can update sbi->stat_info
> in the below context, if the f2fs_write_checkpoint()
> has failed with error.
>
> f2fs_balance_fs_bg+0x1ac/0x1ec
> f2fs_write_node_pages+0x4c/0x260
> do_writepages+0x80/0xbc
> __writeback_single_inode+0xdc/0x4ac
> writeback_single_inode+0x9c/0x144
> write_inode_now+0xc4/0xec
> iput+0x194/0x22c
> f2fs_put_super+0x11c/0x1e8
> generic_shutdown_super+0x70/0xf4
> kill_block_super+0x2c/0x5c
> kill_f2fs_super+0x44/0x50
> deactivate_locked_super+0x60/0x8c
> deactivate_super+0x68/0x74
> cleanup_mnt+0x40/0x78
>
> Fix this by moving f2fs_destroy_stats() further below iput().
>
> Signed-off-by: Sahitya Tummala <[email protected]>
> ---
> fs/f2fs/super.c | 9 ++++++---
> 1 file changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> index e184ad4e..df41a3a 100644
> --- a/fs/f2fs/super.c
> +++ b/fs/f2fs/super.c
> @@ -1058,9 +1058,6 @@ static void f2fs_put_super(struct super_block *sb)
> f2fs_write_checkpoint(sbi, &cpc);
> }
>
> - /* f2fs_write_checkpoint can update stat informaion */
> - f2fs_destroy_stats(sbi);

The code order in error path of fill_super is almost the same as the one of
put_super, could you please check that as well?

Thanks,

> -
> /*
> * normally superblock is clean, so we need to release this.
> * In addition, EIO will skip do checkpoint, we need this as well.
> @@ -1080,6 +1077,12 @@ static void f2fs_put_super(struct super_block *sb)
> iput(sbi->node_inode);
> iput(sbi->meta_inode);
>
> + /*
> + * iput() can update stat information, if f2fs_write_checkpoint()
> + * above failed with error.
> + */
> + f2fs_destroy_stats(sbi);
> +
> /* destroy f2fs internal modules */
> f2fs_destroy_node_manager(sbi);
> f2fs_destroy_segment_manager(sbi);
>


2018-12-26 05:09:03

by Sahitya Tummala

[permalink] [raw]
Subject: Re: [PATCH] f2fs: fix use-after-free issue with sbi->stat_info

On Tue, Dec 25, 2018 at 11:06:45AM +0800, Chao Yu wrote:
> On 2018/12/24 21:06, Sahitya Tummala wrote:
> > iput() on sbi->node_inode can update sbi->stat_info
> > in the below context, if the f2fs_write_checkpoint()
> > has failed with error.
> >
> > f2fs_balance_fs_bg+0x1ac/0x1ec
> > f2fs_write_node_pages+0x4c/0x260
> > do_writepages+0x80/0xbc
> > __writeback_single_inode+0xdc/0x4ac
> > writeback_single_inode+0x9c/0x144
> > write_inode_now+0xc4/0xec
> > iput+0x194/0x22c
> > f2fs_put_super+0x11c/0x1e8
> > generic_shutdown_super+0x70/0xf4
> > kill_block_super+0x2c/0x5c
> > kill_f2fs_super+0x44/0x50
> > deactivate_locked_super+0x60/0x8c
> > deactivate_super+0x68/0x74
> > cleanup_mnt+0x40/0x78
> >
> > Fix this by moving f2fs_destroy_stats() further below iput().
> >
> > Signed-off-by: Sahitya Tummala <[email protected]>
> > ---
> > fs/f2fs/super.c | 9 ++++++---
> > 1 file changed, 6 insertions(+), 3 deletions(-)
> >
> > diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> > index e184ad4e..df41a3a 100644
> > --- a/fs/f2fs/super.c
> > +++ b/fs/f2fs/super.c
> > @@ -1058,9 +1058,6 @@ static void f2fs_put_super(struct super_block *sb)
> > f2fs_write_checkpoint(sbi, &cpc);
> > }
> >
> > - /* f2fs_write_checkpoint can update stat informaion */
> > - f2fs_destroy_stats(sbi);
>
> The code order in error path of fill_super is almost the same as the one of
> put_super, could you please check that as well?
>

Thanks Chao, for pointing it out. Yes, the same issue exists in error path of
fill_super as well. I will fix it.

> Thanks,
>
> > -
> > /*
> > * normally superblock is clean, so we need to release this.
> > * In addition, EIO will skip do checkpoint, we need this as well.
> > @@ -1080,6 +1077,12 @@ static void f2fs_put_super(struct super_block *sb)
> > iput(sbi->node_inode);
> > iput(sbi->meta_inode);
> >
> > + /*
> > + * iput() can update stat information, if f2fs_write_checkpoint()
> > + * above failed with error.
> > + */
> > + f2fs_destroy_stats(sbi);
> > +
> > /* destroy f2fs internal modules */
> > f2fs_destroy_node_manager(sbi);
> > f2fs_destroy_segment_manager(sbi);
> >
>

--
--
Sent by a consultant of the Qualcomm Innovation Center, Inc.
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum.