[PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Shiyang Ruan]

unshare copies data from source to destination. But if the source is
HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
result will be unexpectable.

Fixes: d984648e428b ("fsdax,xfs: port unshare to fsdax")
Signed-off-by: Shiyang Ruan <ruansy.fnst@fujitsu.com>
---
 fs/dax.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/fs/dax.c b/fs/dax.c
index 9800b93ee14d..5d2e9b10030e 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -1258,15 +1258,20 @@ static s64 dax_unshare_iter(struct iomap_iter *iter)
 	/* don't bother with blocks that are not shared to start with */
 	if (!(iomap->flags & IOMAP_F_SHARED))
 		return length;
-	/* don't bother with holes or unwritten extents */
-	if (srcmap->type == IOMAP_HOLE || srcmap->type == IOMAP_UNWRITTEN)
-		return length;
 
 	id = dax_read_lock();
 	ret = dax_iomap_direct_access(iomap, pos, length, &daddr, NULL);
 	if (ret < 0)
 		goto out_unlock;
 
+	/* zero the distance if srcmap is HOLE or UNWRITTEN */
+	if (srcmap->flags & IOMAP_F_SHARED || srcmap->type == IOMAP_UNWRITTEN) {
+		memset(daddr, 0, length);
+		dax_flush(iomap->dax_dev, daddr, length);
+		ret = length;
+		goto out_unlock;
+	}
+
 	ret = dax_iomap_direct_access(srcmap, pos, length, &saddr, NULL);
 	if (ret < 0)
 		goto out_unlock;
-- 
2.39.2

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Andrew Morton]

On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:

> unshare copies data from source to destination. But if the source is
> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
> result will be unexpectable.

Please provide much more detail on the user-visible effects of the bug.
For example, are we leaking kernel memory contents to userspace?

Thanks.

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Shiyang Ruan]

在 2023/3/23 7:03, Andrew Morton 写道:
> On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> 
>> unshare copies data from source to destination. But if the source is
>> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
>> result will be unexpectable.
> 
> Please provide much more detail on the user-visible effects of the bug.
> For example, are we leaking kernel memory contents to userspace?

This fixes fail of generic/649.


--
Thanks,
Ruan.

> 
> Thanks.
> 
> 

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Darrick J. Wong]

On Thu, Mar 23, 2023 at 02:50:38PM +0800, Shiyang Ruan wrote:
> 
> 
> 在 2023/3/23 7:03, Andrew Morton 写道:
> > On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > 
> > > unshare copies data from source to destination. But if the source is
> > > HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
> > > result will be unexpectable.
> > 
> > Please provide much more detail on the user-visible effects of the bug.
> > For example, are we leaking kernel memory contents to userspace?
> 
> This fixes fail of generic/649.

Please make a note of easy(ish) reproducers when you know about them,
e.g.

"Found by running generic/649 while mounting with -o dax=always on
pmem."

So that anyone scanning around for fix backports has an easier time
doing an A/B comparison to assess whether or not they need the fix.
Especially if there's any chance that person is you circa 2024. ;)

--D

> 
> --
> Thanks,
> Ruan.
> 
> > 
> > Thanks.
> > 
> > 

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Andrew Morton]

On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:

> 
> 
> 在 2023/3/23 7:03, Andrew Morton 写道:
> > On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > 
> >> unshare copies data from source to destination. But if the source is
> >> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
> >> result will be unexpectable.
> > 
> > Please provide much more detail on the user-visible effects of the bug.
> > For example, are we leaking kernel memory contents to userspace?
> 
> This fixes fail of generic/649.

OK, but this doesn't really help.  I'm trying to determine whether this
fix should be backported into -stable kernels and whether it should be
fast-tracked into Linus's current -rc tree.

But to determine this I (and others) need to know what effect the bug
has upon our users.

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Shiyang Ruan]

在 2023/3/24 6:11, Andrew Morton 写道:
> On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> 
>>
>>
>> 在 2023/3/23 7:03, Andrew Morton 写道:
>>> On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
>>>
>>>> unshare copies data from source to destination. But if the source is
>>>> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
>>>> result will be unexpectable.
>>>
>>> Please provide much more detail on the user-visible effects of the bug.
>>> For example, are we leaking kernel memory contents to userspace?
>>
>> This fixes fail of generic/649.
> 
> OK, but this doesn't really help.  I'm trying to determine whether this
> fix should be backported into -stable kernels and whether it should be
> fast-tracked into Linus's current -rc tree.
> 
> But to determine this I (and others) need to know what effect the bug
> has upon our users.

I didn't get any bug report form users.  I just found this by running 
xfstests.  The phenomenon of this problem is: if we funshare a reflinked 
file which contains HOLE extents, the result of the HOLE extents should 
be zero but actually not (unexpectable data).

The other patch also has no reports from users.


--
Thanks,
Ruan.

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Matthew Wilcox]

On Fri, Mar 24, 2023 at 09:50:54AM +0800, Shiyang Ruan wrote:
> 
> 
> 在 2023/3/24 6:11, Andrew Morton 写道:
> > On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > 
> > > 
> > > 
> > > 在 2023/3/23 7:03, Andrew Morton 写道:
> > > > On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > > > 
> > > > > unshare copies data from source to destination. But if the source is
> > > > > HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
> > > > > result will be unexpectable.
> > > > 
> > > > Please provide much more detail on the user-visible effects of the bug.
> > > > For example, are we leaking kernel memory contents to userspace?
> > > 
> > > This fixes fail of generic/649.
> > 
> > OK, but this doesn't really help.  I'm trying to determine whether this
> > fix should be backported into -stable kernels and whether it should be
> > fast-tracked into Linus's current -rc tree.
> > 
> > But to determine this I (and others) need to know what effect the bug
> > has upon our users.
> 
> I didn't get any bug report form users.  I just found this by running
> xfstests.  The phenomenon of this problem is: if we funshare a reflinked
> file which contains HOLE extents, the result of the HOLE extents should be
> zero but actually not (unexpectable data).

You still aren't answering the question.  If this did happen to a user,
what would they see in the file?  Random data?  Something somebody else
wrote some time ago?  A copy of /etc/passwd, perhaps?  A copy of your
credit card number?

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Shiyang Ruan]

在 2023/3/24 11:33, Matthew Wilcox 写道:
> On Fri, Mar 24, 2023 at 09:50:54AM +0800, Shiyang Ruan wrote:
>>
>>
>> 在 2023/3/24 6:11, Andrew Morton 写道:
>>> On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
>>>
>>>>
>>>>
>>>> 在 2023/3/23 7:03, Andrew Morton 写道:
>>>>> On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
>>>>>
>>>>>> unshare copies data from source to destination. But if the source is
>>>>>> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
>>>>>> result will be unexpectable.
>>>>>
>>>>> Please provide much more detail on the user-visible effects of the bug.
>>>>> For example, are we leaking kernel memory contents to userspace?
>>>>
>>>> This fixes fail of generic/649.
>>>
>>> OK, but this doesn't really help.  I'm trying to determine whether this
>>> fix should be backported into -stable kernels and whether it should be
>>> fast-tracked into Linus's current -rc tree.
>>>
>>> But to determine this I (and others) need to know what effect the bug
>>> has upon our users.
>>
>> I didn't get any bug report form users.  I just found this by running
>> xfstests.  The phenomenon of this problem is: if we funshare a reflinked
>> file which contains HOLE extents, the result of the HOLE extents should be
>> zero but actually not (unexpectable data).
> 
> You still aren't answering the question.  If this did happen to a user,
> what would they see in the file?  Random data?  Something somebody else
> wrote some time ago?  A copy of /etc/passwd, perhaps?  A copy of your
> credit card number?

Ok.  If this happenned to a user, the HOLE or UNWRITTEN part will be old 
data of the new allocated extent because it didn't be cleared.


--
Thanks,
Ruan.

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Matthew Wilcox]

On Fri, Mar 24, 2023 at 11:42:53AM +0800, Shiyang Ruan wrote:
> 
> 
> 在 2023/3/24 11:33, Matthew Wilcox 写道:
> > On Fri, Mar 24, 2023 at 09:50:54AM +0800, Shiyang Ruan wrote:
> > > 
> > > 
> > > 在 2023/3/24 6:11, Andrew Morton 写道:
> > > > On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > > > 
> > > > > 
> > > > > 
> > > > > 在 2023/3/23 7:03, Andrew Morton 写道:
> > > > > > On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
> > > > > > 
> > > > > > > unshare copies data from source to destination. But if the source is
> > > > > > > HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
> > > > > > > result will be unexpectable.
> > > > > > 
> > > > > > Please provide much more detail on the user-visible effects of the bug.
> > > > > > For example, are we leaking kernel memory contents to userspace?
> > > > > 
> > > > > This fixes fail of generic/649.
> > > > 
> > > > OK, but this doesn't really help.  I'm trying to determine whether this
> > > > fix should be backported into -stable kernels and whether it should be
> > > > fast-tracked into Linus's current -rc tree.
> > > > 
> > > > But to determine this I (and others) need to know what effect the bug
> > > > has upon our users.
> > > 
> > > I didn't get any bug report form users.  I just found this by running
> > > xfstests.  The phenomenon of this problem is: if we funshare a reflinked
> > > file which contains HOLE extents, the result of the HOLE extents should be
> > > zero but actually not (unexpectable data).
> > 
> > You still aren't answering the question.  If this did happen to a user,
> > what would they see in the file?  Random data?  Something somebody else
> > wrote some time ago?  A copy of /etc/passwd, perhaps?  A copy of your
> > credit card number?
> 
> Ok.  If this happenned to a user, the HOLE or UNWRITTEN part will be old
> data of the new allocated extent because it didn't be cleared.

ie it's the data that was in whatever file happened to use that space
last, so this is a security bug because it's a data leak, and a backport
is needed, and you should have indicated that by putting a cc: stable
tag on the patch?

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Shiyang Ruan]

在 2023/3/24 12:17, Matthew Wilcox 写道:
> On Fri, Mar 24, 2023 at 11:42:53AM +0800, Shiyang Ruan wrote:
>>
>>
>> 在 2023/3/24 11:33, Matthew Wilcox 写道:
>>> On Fri, Mar 24, 2023 at 09:50:54AM +0800, Shiyang Ruan wrote:
>>>>
>>>>
>>>> 在 2023/3/24 6:11, Andrew Morton 写道:
>>>>> On Thu, 23 Mar 2023 14:50:38 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
>>>>>
>>>>>>
>>>>>>
>>>>>> 在 2023/3/23 7:03, Andrew Morton 写道:
>>>>>>> On Wed, 22 Mar 2023 11:11:09 +0000 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:
>>>>>>>
>>>>>>>> unshare copies data from source to destination. But if the source is
>>>>>>>> HOLE or UNWRITTEN extents, we should zero the destination, otherwise the
>>>>>>>> result will be unexpectable.
>>>>>>>
>>>>>>> Please provide much more detail on the user-visible effects of the bug.
>>>>>>> For example, are we leaking kernel memory contents to userspace?
>>>>>>
>>>>>> This fixes fail of generic/649.
>>>>>
>>>>> OK, but this doesn't really help.  I'm trying to determine whether this
>>>>> fix should be backported into -stable kernels and whether it should be
>>>>> fast-tracked into Linus's current -rc tree.
>>>>>
>>>>> But to determine this I (and others) need to know what effect the bug
>>>>> has upon our users.
>>>>
>>>> I didn't get any bug report form users.  I just found this by running
>>>> xfstests.  The phenomenon of this problem is: if we funshare a reflinked
>>>> file which contains HOLE extents, the result of the HOLE extents should be
>>>> zero but actually not (unexpectable data).
>>>
>>> You still aren't answering the question.  If this did happen to a user,
>>> what would they see in the file?  Random data?  Something somebody else
>>> wrote some time ago?  A copy of /etc/passwd, perhaps?  A copy of your
>>> credit card number?
>>
>> Ok.  If this happenned to a user, the HOLE or UNWRITTEN part will be old
>> data of the new allocated extent because it didn't be cleared.
> 
> ie it's the data that was in whatever file happened to use that space
> last, so this is a security bug because it's a data leak, and a backport
> is needed, and you should have indicated that by putting a cc: stable
> tag on the patch?

Yes, cc stable is needed.  Then should I send a new patch with the tag 
added?


--
Thanks,
Ruan.

Re: [PATCH] fsdax: unshare: zero destination if srcmap is HOLE or UNWRITTEN

[Andrew Morton]

On Fri, 24 Mar 2023 12:28:07 +0800 Shiyang Ruan <ruansy.fnst@fujitsu.com> wrote:

> > ie it's the data that was in whatever file happened to use that space
> > last, so this is a security bug because it's a data leak, and a backport
> > is needed, and you should have indicated that by putting a cc: stable
> > tag on the patch?
> 
> Yes, cc stable is needed.  Then should I send a new patch with the tag 
> added?

Thanks.  I updated the changelog and added cc:stable.

: unshare copies data from source to destination.  But if the source is
: HOLE or UNWRITTEN extents, we should zero the destination, otherwise
: the HOLE or UNWRITTEN part will be user-visible old data of the new
: allocated extent because it wasn't cleared.
: 
: Found by running generic/649 while mounting with -o dax=always on pmem.