Message ID | 20230508162219.1731964-1-mlombard@redhat.com (mailing list archive) |
---|---|
Headers | show |
Series | Fix possible hangs and race conditions during login | expand |
On 5/8/23 11:22 AM, Maurizio Lombardi wrote: > PATCH 1 fixes a possible hang when during a login operation the initiator > stops sending data to the target and the login_work doesn't get scheduled. > > PATCH 3 fixes a bug in iscsi_target_locate_portal(), in the error > code path the login semaphore may end up being released even if > no one called down() against it, this may allow > multiple threads to access the semaphore-protected code. > > v3: fix the comments' format > > v2: the login timeout of 15 seconds now is relative to the entire > login operation and not per-PDU. > Reviewed-by: Mike Christie <michael.christie@oracle.com>
On Mon, 08 May 2023 18:22:16 +0200, Maurizio Lombardi wrote: > PATCH 1 fixes a possible hang when during a login operation the initiator > stops sending data to the target and the login_work doesn't get scheduled. > > PATCH 3 fixes a bug in iscsi_target_locate_portal(), in the error > code path the login semaphore may end up being released even if > no one called down() against it, this may allow > multiple threads to access the semaphore-protected code. > > [...] Applied to 6.4/scsi-fixes, thanks! [1/3] target: iscsi: fix hang in the iSCSI login code https://git.kernel.org/mkp/scsi/c/13247018d68f [2/3] target: iscsi: remove unused transport_timer https://git.kernel.org/mkp/scsi/c/98a8c2bf938a [3/3] target: iscsi: prevent login threads from racing between each other https://git.kernel.org/mkp/scsi/c/2a737d3b8c79