From: Chris Leech <cleech@redhat.com>
To: Showrya M N <showrya@chelsio.com>
Cc: lduncan@suse.com, michael.christie@oracle.com,
martin.petersen@oracle.com, linux-scsi@vger.kernel.org,
bharat@chelsio.com
Subject: Re: [PATCH for-rc] scsi: libiscsi: initialize iscsi_conn->dd_data only if memory is allocated
Date: Tue, 15 Jul 2025 10:49:21 -0700 [thread overview]
Message-ID: <aHaUoRFbpnjkQFYD@my-developer-toolbox-latest> (raw)
In-Reply-To: <20250627112329.19763-1-showrya@chelsio.com>
On Fri, Jun 27, 2025 at 04:53:29PM +0530, Showrya M N wrote:
> In case of an ib_fast_reg_mr allocation failure during iSER setup,
> the machine hits a panic because iscsi_conn->dd_data is initialized
> unconditionally, even when no memory is allocated (dd_size == 0).
> This leads invalid pointer dereference during connection teardown.
>
> Fix by setting iscsi_conn->dd_data only if memory is actually allocated.
iser is allocating the iser_conn along with the endpoint, and
dynamically updating iscsi_conn->dd_data to track that allocation.
That's different from all the other libiscsi drivers which allocate a
fixed sized driver area with every iscsi_conn.
I don't see any problem with conditionally setting iscsi_conn->dd_data,
and it makes sense to not have an invalid pointer for a 0 size request.
iscsi_iser_conn_stop already has the checks for a NULL dd_data for this
case of a bind failure, so that looks OK.
Signed-off-by: Chris Leech <cleech@redhat.com>
> Panic trace:
> ------------
> iser: iser_create_fastreg_desc: Failed to allocate ib_fast_reg_mr err=-12
> iser: iser_alloc_rx_descriptors: failed allocating rx descriptors / data buffers
> BUG: unable to handle page fault for address: fffffffffffffff8
> RIP: 0010:swake_up_locked.part.5+0xa/0x40
> Call Trace:
> complete+0x31/0x40
> iscsi_iser_conn_stop+0x88/0xb0 [ib_iser]
> iscsi_stop_conn+0x66/0xc0 [scsi_transport_iscsi]
> iscsi_if_stop_conn+0x14a/0x150 [scsi_transport_iscsi]
> iscsi_if_rx+0x1135/0x1834 [scsi_transport_iscsi]
> ? netlink_lookup+0x12f/0x1b0
> ? netlink_deliver_tap+0x2c/0x200
> netlink_unicast+0x1ab/0x280
> netlink_sendmsg+0x257/0x4f0
> ? _copy_from_user+0x29/0x60
> sock_sendmsg+0x5f/0x70
>
> Signed-off-by: Showrya M N <showrya@chelsio.com>
> Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com>
> ---
> drivers/scsi/libiscsi.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
> index 392d57e054db..c9f410c50978 100644
> --- a/drivers/scsi/libiscsi.c
> +++ b/drivers/scsi/libiscsi.c
> @@ -3185,7 +3185,8 @@ iscsi_conn_setup(struct iscsi_cls_session *cls_session, int dd_size,
> return NULL;
> conn = cls_conn->dd_data;
>
> - conn->dd_data = cls_conn->dd_data + sizeof(*conn);
> + if (dd_size)
> + conn->dd_data = cls_conn->dd_data + sizeof(*conn);
> conn->session = session;
> conn->cls_conn = cls_conn;
> conn->c_stage = ISCSI_CONN_INITIAL_STAGE;
> --
> 2.39.3
>
next prev parent reply other threads:[~2025-07-15 17:49 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-06-27 11:23 [PATCH for-rc] scsi: libiscsi: initialize iscsi_conn->dd_data only if memory is allocated Showrya M N
2025-07-15 1:31 ` Martin K. Petersen
2025-07-15 17:49 ` Chris Leech [this message]
2025-07-22 3:48 ` Martin K. Petersen
2025-07-25 3:00 ` Martin K. Petersen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=aHaUoRFbpnjkQFYD@my-developer-toolbox-latest \
--to=cleech@redhat.com \
--cc=bharat@chelsio.com \
--cc=lduncan@suse.com \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=michael.christie@oracle.com \
--cc=showrya@chelsio.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox