Before calling add partition or resize partition, there is no check
on whether the length is aligned with the logical block size.
If the logical block size of the disk is larger than 512 bytes,
then the partition size maybe not the multiple of the logical block size,
and when the last sector is read, bio_truncate() will adjust the bio size,
resulting in an IO error if the size of the read command is smaller than
the logical block size.If integrity data is supported, this will also
result in a null pointer dereference when calling bio_integrity_free.
Signed-off-by: Min Li <[email protected]>
---
Changes from v1:
- Add a space after /* and before */.
- Move length alignment check before the "start = p.start >> SECTOR_SHIFT"
- Move check for p.start being aligned together with this length alignment check.
---
block/ioctl.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/block/ioctl.c b/block/ioctl.c
index 3be11941fb2d..c40b382dd58f 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -33,14 +33,18 @@ static int blkpg_do_ioctl(struct block_device *bdev,
if (op == BLKPG_DEL_PARTITION)
return bdev_del_partition(disk, p.pno);
+ /* check if partition is aligned to blocksize */
+ if (p.start & (bdev_logical_block_size(bdev) - 1))
+ return -EINVAL;
+ /* check if length is aligned to blocksize */
+ if (p.length & (bdev_logical_block_size(bdev) - 1))
+ return -EINVAL;
+
start = p.start >> SECTOR_SHIFT;
length = p.length >> SECTOR_SHIFT;
switch (op) {
case BLKPG_ADD_PARTITION:
- /* check if partition is aligned to blocksize */
- if (p.start & (bdev_logical_block_size(bdev) - 1))
- return -EINVAL;
return bdev_add_partition(disk, p.pno, start, length);
case BLKPG_RESIZE_PARTITION:
return bdev_resize_partition(disk, p.pno, start, length);
--
2.34.1
On 6/27/2023 4:09 AM, Min Li wrote:
> Before calling add partition or resize partition, there is no check
> on whether the length is aligned with the logical block size.
> If the logical block size of the disk is larger than 512 bytes,
> then the partition size maybe not the multiple of the logical block size,
> and when the last sector is read, bio_truncate() will adjust the bio size,
> resulting in an IO error if the size of the read command is smaller than
> the logical block size.If integrity data is supported, this will also
> result in a null pointer dereference when calling bio_integrity_free.
>
> Signed-off-by: Min Li <[email protected]>
>
> ---
> Changes from v1:
>
> - Add a space after /* and before */.
> - Move length alignment check before the "start = p.start >> SECTOR_SHIFT"
> - Move check for p.start being aligned together with this length alignment check.
> ---
Reviewed-by: Chaitanya Kulkarni <[email protected]>
-ck
On 6/27/23 20:09, Min Li wrote:
> Before calling add partition or resize partition, there is no check
> on whether the length is aligned with the logical block size.
> If the logical block size of the disk is larger than 512 bytes,
> then the partition size maybe not the multiple of the logical block size,
> and when the last sector is read, bio_truncate() will adjust the bio size,
> resulting in an IO error if the size of the read command is smaller than
> the logical block size.If integrity data is supported, this will also
> result in a null pointer dereference when calling bio_integrity_free.
>
> Signed-off-by: Min Li <[email protected]>
See Greg's comment: this likely need a "Fixes:" tag. And I think that the tag
is: fa9156ae597c ("block: refactor blkpg_ioctl"). But please double check.
>
> ---
> Changes from v1:
>
> - Add a space after /* and before */.
> - Move length alignment check before the "start = p.start >> SECTOR_SHIFT"
> - Move check for p.start being aligned together with this length alignment check.
> ---
> block/ioctl.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/block/ioctl.c b/block/ioctl.c
> index 3be11941fb2d..c40b382dd58f 100644
> --- a/block/ioctl.c
> +++ b/block/ioctl.c
> @@ -33,14 +33,18 @@ static int blkpg_do_ioctl(struct block_device *bdev,
> if (op == BLKPG_DEL_PARTITION)
> return bdev_del_partition(disk, p.pno);
>
> + /* check if partition is aligned to blocksize */
> + if (p.start & (bdev_logical_block_size(bdev) - 1))
> + return -EINVAL;
> + /* check if length is aligned to blocksize */
> + if (p.length & (bdev_logical_block_size(bdev) - 1))
> + return -EINVAL;
long long blksz_mask = bdev_logical_block_size(bdev) - 1;
/* Check that the partition is aligned to the block size */
if ((p.start & blksz_mask) || (p.length & blksz_mask))
return -EINVAL;
would be cleaner and avoid the rather redundant comments.
> +
> start = p.start >> SECTOR_SHIFT;
> length = p.length >> SECTOR_SHIFT;
>
> switch (op) {
> case BLKPG_ADD_PARTITION:
> - /* check if partition is aligned to blocksize */
> - if (p.start & (bdev_logical_block_size(bdev) - 1))
> - return -EINVAL;
> return bdev_add_partition(disk, p.pno, start, length);
> case BLKPG_RESIZE_PARTITION:
> return bdev_resize_partition(disk, p.pno, start, length);
--
Damien Le Moal
Western Digital Research
On Tue, Jun 27, 2023 at 01:39:26PM +0900, Damien Le Moal wrote:
> See Greg's comment: this likely need a "Fixes:" tag. And I think that the tag
> is: fa9156ae597c ("block: refactor blkpg_ioctl"). But please double check.
No, the lack of checks goes back all the way to when this code was
added long before git was a thing. So I don't think a Fixes tag makes
all that much sense here.
Looks good:
Reviewed-by: Christoph Hellwig <[email protected]>
On Tue, Jun 27, 2023 at 06:59:24AM +0200, Christoph Hellwig wrote:
> On Tue, Jun 27, 2023 at 01:39:26PM +0900, Damien Le Moal wrote:
> > See Greg's comment: this likely need a "Fixes:" tag. And I think that the tag
> > is: fa9156ae597c ("block: refactor blkpg_ioctl"). But please double check.
>
> No, the lack of checks goes back all the way to when this code was
> added long before git was a thing. So I don't think a Fixes tag makes
> all that much sense here.
Ok, then how about a normal "cc: [email protected]" added to the s-o-b
area so it gets picked up for all stable trees as this is an issue that
has always been there?
thanks,
greg k-h
On Tue, Jun 27, 2023 at 08:18:32AM +0200, Greg KH wrote:
> > No, the lack of checks goes back all the way to when this code was
> > added long before git was a thing. So I don't think a Fixes tag makes
> > all that much sense here.
>
> Ok, then how about a normal "cc: [email protected]" added to the s-o-b
> area so it gets picked up for all stable trees as this is an issue that
> has always been there?
Yes, that sounds sensible.
On Tue, Jun 27, 2023 at 01:39:26PM +0900, Damien Le Moal wrote:
> > diff --git a/block/ioctl.c b/block/ioctl.c
> > index 3be11941fb2d..c40b382dd58f 100644
> > --- a/block/ioctl.c
> > +++ b/block/ioctl.c
> > @@ -33,14 +33,18 @@ static int blkpg_do_ioctl(struct block_device *bdev,
> > if (op == BLKPG_DEL_PARTITION)
> > return bdev_del_partition(disk, p.pno);
> >
> > + /* check if partition is aligned to blocksize */
> > + if (p.start & (bdev_logical_block_size(bdev) - 1))
> > + return -EINVAL;
> > + /* check if length is aligned to blocksize */
> > + if (p.length & (bdev_logical_block_size(bdev) - 1))
> > + return -EINVAL;
>
> long long blksz_mask = bdev_logical_block_size(bdev) - 1;
>
> /* Check that the partition is aligned to the block size */
> if ((p.start & blksz_mask) || (p.length & blksz_mask))
> return -EINVAL;
A Minor nit on top of your comment:
unsigned int blksz = bdev_logical_block_size(bdev);
/* Check that the partition is aligned to the block size */
if (!IS_ALIGNED(p.start, blksz) || !IS_ALIGNED(p.length, blksz))
return -EINVAL;
>
> would be cleaner and avoid the rather redundant comments.
>
On Tue, Jun 27, 2023 at 10:13:39AM +0200, Pankaj Raghav wrote:
> On Tue, Jun 27, 2023 at 01:39:26PM +0900, Damien Le Moal wrote:
> > > diff --git a/block/ioctl.c b/block/ioctl.c
> > > index 3be11941fb2d..c40b382dd58f 100644
> > > --- a/block/ioctl.c
> > > +++ b/block/ioctl.c
> > > @@ -33,14 +33,18 @@ static int blkpg_do_ioctl(struct block_device *bdev,
> > > if (op == BLKPG_DEL_PARTITION)
> > > return bdev_del_partition(disk, p.pno);
> > >
> > > + /* check if partition is aligned to blocksize */
> > > + if (p.start & (bdev_logical_block_size(bdev) - 1))
> > > + return -EINVAL;
> > > + /* check if length is aligned to blocksize */
> > > + if (p.length & (bdev_logical_block_size(bdev) - 1))
> > > + return -EINVAL;
> >
> > long long blksz_mask = bdev_logical_block_size(bdev) - 1;
> >
> > /* Check that the partition is aligned to the block size */
> > if ((p.start & blksz_mask) || (p.length & blksz_mask))
> > return -EINVAL;
>
> A Minor nit on top of your comment:
>
> unsigned int blksz = bdev_logical_block_size(bdev);
>
> /* Check that the partition is aligned to the block size */
> if (!IS_ALIGNED(p.start, blksz) || !IS_ALIGNED(p.length, blksz))
> return -EINVAL;
or you can even do ...
if (!IS_ALIGNED(p.start | p.length), blksz)
Do I win the code golf trophy?