[PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[David Howells]

    
If a DIO read or an unbuffered read request extends beyond the EOF, the
server will return a short read and a status code indicating that EOF was
hit, which gets translated to -ENODATA.  Note that the client does not cap
the request at i_size, but asks for the amount requested in case there's a
race on the server with a third party.

Now, on the client side, the request will get split into multiple
subrequests if rsize is smaller than the full request size.  A subrequest
that starts before or at the EOF and returns short data up to the EOF will
be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
indicating to netfslib that we can't read more.

If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
not be flagged, its error will be set to -ENODATA and it will be abandoned.
This will cause the request as a whole to fail with -ENODATA.

Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
the EOF marker.

This can be reproduced by mounting with "cache=none,sign,vers=1.0" and
doing a read of a file that's significantly bigger than the size of the
file (e.g. attempting to read 64KiB from a 16KiB file).

Fixes: a68c74865f51 ("cifs: Fix SMB1 readv/writev callback in the same way as SMB2/3")
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Steve French <sfrench@samba.org>
cc: Paulo Alcantara <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: linux-cifs@vger.kernel.org
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
---
 fs/smb/client/cifssmb.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/smb/client/cifssmb.c b/fs/smb/client/cifssmb.c
index 645831708e1b..1871d2c1a8e0 100644
--- a/fs/smb/client/cifssmb.c
+++ b/fs/smb/client/cifssmb.c
@@ -1395,7 +1395,7 @@ cifs_readv_callback(struct mid_q_entry *mid)
 	} else {
 		size_t trans = rdata->subreq.transferred + rdata->got_bytes;
 		if (trans < rdata->subreq.len &&
-		    rdata->subreq.start + trans == ictx->remote_i_size) {
+		    rdata->subreq.start + trans >= ictx->remote_i_size) {
 			rdata->result = 0;
 			__set_bit(NETFS_SREQ_HIT_EOF, &rdata->subreq.flags);
 		} else if (rdata->got_bytes > 0) {

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[Paulo Alcantara]

David Howells <dhowells@redhat.com> writes:

>     
> If a DIO read or an unbuffered read request extends beyond the EOF, the
> server will return a short read and a status code indicating that EOF was
> hit, which gets translated to -ENODATA.  Note that the client does not cap
> the request at i_size, but asks for the amount requested in case there's a
> race on the server with a third party.
>
> Now, on the client side, the request will get split into multiple
> subrequests if rsize is smaller than the full request size.  A subrequest
> that starts before or at the EOF and returns short data up to the EOF will
> be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
> indicating to netfslib that we can't read more.
>
> If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
> not be flagged, its error will be set to -ENODATA and it will be abandoned.
> This will cause the request as a whole to fail with -ENODATA.
>
> Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
> the EOF marker.
>
> This can be reproduced by mounting with "cache=none,sign,vers=1.0" and
> doing a read of a file that's significantly bigger than the size of the
> file (e.g. attempting to read 64KiB from a 16KiB file).
>
> Fixes: a68c74865f51 ("cifs: Fix SMB1 readv/writev callback in the same way as SMB2/3")
> Signed-off-by: David Howells <dhowells@redhat.com>
> cc: Steve French <sfrench@samba.org>
> cc: Paulo Alcantara <pc@manguebit.org>
> cc: Shyam Prasad N <sprasad@microsoft.com>
> cc: linux-cifs@vger.kernel.org
> cc: netfs@lists.linux.dev
> cc: linux-fsdevel@vger.kernel.org

Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>

Dave, looks like we're missing a similar fix for smb2_readv_callback()
as well.

Can you handle it?

Thanks.

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[Steve French]

On Wed, Dec 3, 2025 at 12:03 PM Paulo Alcantara <pc@manguebit.org> wrote:
>
> David Howells <dhowells@redhat.com> writes:
>
> >
> > If a DIO read or an unbuffered read request extends beyond the EOF, the
> > server will return a short read and a status code indicating that EOF was
> > hit, which gets translated to -ENODATA.  Note that the client does not cap
> > the request at i_size, but asks for the amount requested in case there's a
> > race on the server with a third party.
> >
> > Now, on the client side, the request will get split into multiple
> > subrequests if rsize is smaller than the full request size.  A subrequest
> > that starts before or at the EOF and returns short data up to the EOF will
> > be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
> > indicating to netfslib that we can't read more.
> >
> > If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
> > not be flagged, its error will be set to -ENODATA and it will be abandoned.
> > This will cause the request as a whole to fail with -ENODATA.
> >
> > Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
> > the EOF marker.
> >
> > This can be reproduced by mounting with "cache=none,sign,vers=1.0" and
> > doing a read of a file that's significantly bigger than the size of the
> > file (e.g. attempting to read 64KiB from a 16KiB file).
> >
> > Fixes: a68c74865f51 ("cifs: Fix SMB1 readv/writev callback in the same way as SMB2/3")
> > Signed-off-by: David Howells <dhowells@redhat.com>
> > cc: Steve French <sfrench@samba.org>
> > cc: Paulo Alcantara <pc@manguebit.org>
> > cc: Shyam Prasad N <sprasad@microsoft.com>
> > cc: linux-cifs@vger.kernel.org
> > cc: netfs@lists.linux.dev
> > cc: linux-fsdevel@vger.kernel.org
>
> Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
>
> Dave, looks like we're missing a similar fix for smb2_readv_callback()
> as well.
>
> Can you handle it?

Any luck reproducing it for smb2/smb3/smb3.1.1?

-- 
Thanks,

Steve

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[Steve French]

Paulo,
Added your reviewed by to David's patches but wanted to doublecheck
that I didn't apply it to too many of them since I couldn't find one
of your notes

Does this look ok for your RB on all 14 of these - or just the SMB1 one one?

a6fd899da60f (HEAD -> for-next, origin/for-next, origin/HEAD) cifs:
Remove dead function prototypes
1b7270c879f5 smb: server: defer the initial recv completion logic to
smb_direct_negotiate_recv_work()
9d095775a0cb smb: server: initialize recv_io->cqe.done = recv_done just once
667246dbce2d smb: smbdirect: introduce smbdirect_socket.connect.{lock,work}
2b4e375e4006 cifs: Do some preparation prior to organising the
function declarations
c3bdaf3afd87 cifs: Add a tracepoint to log EIO errors
cb416ff96b83 cifs: Don't need state locking in smb2_get_mid_entry()
a64fa1835237 cifs: Remove the server pointer from smb_message
960cd2e1e28a cifs: Fix specification of function pointers
2fdd780130d1 cifs: Replace SendReceiveBlockingLock() with
SendReceive() plus flags
bb8172e800b3 cifs: Clean up some places where an extra kvec[] was
required for rfc1002
41daa3d4a238 cifs: Make smb1's SendReceive() wrap cifs_send_recv()
3ed72b50d276 cifs: Remove the RFC1002 header from smb_hdr
271b1138e8b4 cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

On Wed, Dec 3, 2025 at 12:03 PM Paulo Alcantara <pc@manguebit.org> wrote:
>
> David Howells <dhowells@redhat.com> writes:
>
> >
> > If a DIO read or an unbuffered read request extends beyond the EOF, the
> > server will return a short read and a status code indicating that EOF was
> > hit, which gets translated to -ENODATA.  Note that the client does not cap
> > the request at i_size, but asks for the amount requested in case there's a
> > race on the server with a third party.
> >
> > Now, on the client side, the request will get split into multiple
> > subrequests if rsize is smaller than the full request size.  A subrequest
> > that starts before or at the EOF and returns short data up to the EOF will
> > be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
> > indicating to netfslib that we can't read more.
> >
> > If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
> > not be flagged, its error will be set to -ENODATA and it will be abandoned.
> > This will cause the request as a whole to fail with -ENODATA.
> >
> > Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
> > the EOF marker.
> >
> > This can be reproduced by mounting with "cache=none,sign,vers=1.0" and
> > doing a read of a file that's significantly bigger than the size of the
> > file (e.g. attempting to read 64KiB from a 16KiB file).
> >
> > Fixes: a68c74865f51 ("cifs: Fix SMB1 readv/writev callback in the same way as SMB2/3")
> > Signed-off-by: David Howells <dhowells@redhat.com>
> > cc: Steve French <sfrench@samba.org>
> > cc: Paulo Alcantara <pc@manguebit.org>
> > cc: Shyam Prasad N <sprasad@microsoft.com>
> > cc: linux-cifs@vger.kernel.org
> > cc: netfs@lists.linux.dev
> > cc: linux-fsdevel@vger.kernel.org
>
> Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>
>
> Dave, looks like we're missing a similar fix for smb2_readv_callback()
> as well.
>
> Can you handle it?
>
> Thanks.
>


--
Thanks,

Steve

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[Paulo Alcantara]

Steve French <smfrench@gmail.com> writes:

> Paulo,
> Added your reviewed by to David's patches but wanted to doublecheck
> that I didn't apply it to too many of them since I couldn't find one
> of your notes
>
> Does this look ok for your RB on all 14 of these - or just the SMB1 one one?

I'd reviewed and discussed the other 14 patches with Dave already, so
please keep my R-b on all of them.

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB1

[David Howells]

Paulo Alcantara <pc@manguebit.org> wrote:

> Dave, looks like we're missing a similar fix for smb2_readv_callback()
> as well.

I couldn't reproduce the problem with smb2/3, but it's probably worth fixing
by analogy.

David

[PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB2

[David Howells]

If a DIO read or an unbuffered read request extends beyond the EOF, the
server will return a short read and a status code indicating that EOF was
hit, which gets translated to -ENODATA.  Note that the client does not cap
the request at i_size, but asks for the amount requested in case there's a
race on the server with a third party.

Now, on the client side, the request will get split into multiple
subrequests if rsize is smaller than the full request size.  A subrequest
that starts before or at the EOF and returns short data up to the EOF will
be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
indicating to netfslib that we can't read more.

If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
not be flagged, its error will be set to -ENODATA and it will be abandoned.
This will cause the request as a whole to fail with -ENODATA.

Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
the EOF marker.

Fixes: 1da29f2c39b6 ("netfs, cifs: Fix handling of short DIO read")
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Steve French <sfrench@samba.org>
cc: Paulo Alcantara <pc@manguebit.org>
cc: Shyam Prasad N <sprasad@microsoft.com>
cc: linux-cifs@vger.kernel.org
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
---
 fs/smb/client/smb2pdu.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/smb/client/smb2pdu.c b/fs/smb/client/smb2pdu.c
index d21606fa3f1f..f142f62c639d 100644
--- a/fs/smb/client/smb2pdu.c
+++ b/fs/smb/client/smb2pdu.c
@@ -4656,7 +4656,7 @@ smb2_readv_callback(struct mid_q_entry *mid)
 	} else {
 		size_t trans = rdata->subreq.transferred + rdata->got_bytes;
 		if (trans < rdata->subreq.len &&
-		    rdata->subreq.start + trans == ictx->remote_i_size) {
+		    rdata->subreq.start + trans >= ictx->remote_i_size) {
 			__set_bit(NETFS_SREQ_HIT_EOF, &rdata->subreq.flags);
 			rdata->result = 0;
 		}

Re: [PATCH] cifs: Fix handling of a beyond-EOF DIO/unbuffered read over SMB2

[Paulo Alcantara]

David Howells <dhowells@redhat.com> writes:

> If a DIO read or an unbuffered read request extends beyond the EOF, the
> server will return a short read and a status code indicating that EOF was
> hit, which gets translated to -ENODATA.  Note that the client does not cap
> the request at i_size, but asks for the amount requested in case there's a
> race on the server with a third party.
>
> Now, on the client side, the request will get split into multiple
> subrequests if rsize is smaller than the full request size.  A subrequest
> that starts before or at the EOF and returns short data up to the EOF will
> be correctly handled, with the NETFS_SREQ_HIT_EOF flag being set,
> indicating to netfslib that we can't read more.
>
> If a subrequest, however, starts after the EOF and not at it, HIT_EOF will
> not be flagged, its error will be set to -ENODATA and it will be abandoned.
> This will cause the request as a whole to fail with -ENODATA.
>
> Fix this by setting NETFS_SREQ_HIT_EOF on any subrequest that lies beyond
> the EOF marker.
>
> Fixes: 1da29f2c39b6 ("netfs, cifs: Fix handling of short DIO read")
> Signed-off-by: David Howells <dhowells@redhat.com>
> cc: Steve French <sfrench@samba.org>
> cc: Paulo Alcantara <pc@manguebit.org>
> cc: Shyam Prasad N <sprasad@microsoft.com>
> cc: linux-cifs@vger.kernel.org
> cc: netfs@lists.linux.dev
> cc: linux-fsdevel@vger.kernel.org

Reviewed-by: Paulo Alcantara (Red Hat) <pc@manguebit.org>