- 07 Jul, 2017 40 commits
-
-
Xiubo Li authored
When feeding the tcmu's cmd ring, we need to flush the dcache page for the cmd entry to make sure these kernel stores are visible to user space mappings of that page. For the none PAD cmd entry, this will be flushed at the end of the tcmu_queue_cmd_ring(). Signed-off-by:
Xiubo Li <lixiubo@cmss.chinamobile.com> Reviewed-by:
Mike Christie <mchristi@redhat.com> Signed-off-by:
Nicholas Bellinger <nab@linux-iscsi.org>
-
Mike Christie authored
If the uio device is open and closed multiple times, the kref count will be off due to tcmu_release getting called multiple times for each close. This patch integrates Wenji Tang's patch to add a kref_get on open that now matches the kref_put done on tcmu_release and adds a kref_put in tcmu_destroy_device to match the kref_get done in succesful tcmu_configure_device calls. Signed-off-by:
-
Mike Christie authored
destroy_device is only called if we have successfully run configure_device, so drop the duplicate tcmu_dev_configured check. Signed-off-by:
-
Mike Christie authored
g_device_list is no longer needed because we now use the idr code for lookups and seaches. Signed-off-by:
Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by:
-
Mike Christie authored
This converts the xcopy code to use the idr helper. The next patch will drop the g_device_list and make g_device_mutex local to the target_core_device.c file. Signed-off-by:
-
Mike Christie authored
This adds a wrapper around idr_for_each so the xcopy code can loop over the devices in the next patch. Signed-off-by:
-
Mike Christie authored
This makes the device add, del reconfig operations sync. It fixes the issue where for add and reconfig, we do not know if userspace successfully completely the operation, so we leave invalid kernel structs or report incorrect status for the config/reconfig operations. Signed-off-by:
-
Mike Christie authored
This adds a helper to find a se_device by dev_index. It will be used in the next patches so tcmu's netlink interface can execute commands on specific devices. Signed-off-by:
-
Mike Christie authored
In the next patches we will add tcmu netlink support that allows userspace to send commands to target_core_user. To execute operations on a se_device/tcmu_dev we need to be able to look up a dev by any old id. This patch replaces the se_device->dev_index with a idr created id. The next patches will also remove the g_device_list and replace it with the idr. Signed-off-by:
-
Mike Christie authored
With this patch free_device is now used to free what is allocated in the alloc_device callback and destroy_device tears down the resources that are setup in the configure_device callback. This patch will be needed in the next patch where tcmu needs to be able to look up the device in the destroy callback. Signed-off-by:
-
Mike Christie authored
1. TCMU_ATTR_TYPE is too generic when it describes only the reconfiguration type, so rename to TCMU_ATTR_RECONFIG_TYPE. 2. Only return the reconfig type when it is a TCMU_CMD_RECONFIG_DEVICE command. 3. CONFIG_* type is not needed. We can pass the value along with an ATTR to userspace, so it does not need to read sysfs/configfs. 4. Fix leak in tcmu_dev_path_store and rename to dev_config to reflect it is more than just a path that can be changed. 6. Don't update kernel struct value if netlink sending fails. Signed-off-by:
"Bryant G. Ly" <bryantly@linux.vnet.ibm.com> Signed-off-by:
-
Nicholas Bellinger authored
This patch drops two incorrect usages of tcm_qla2xxx_free_cmd() during TMR ABORT within tcm_qla2xxx_handle_data_work() and tcm_qla2xxx_aborted_task(), which where attempting to dispatch into workqueue context to do tcm_qla2xxx_complete_free() and subsequently invoke transport_generic_free_cmd(). This is incorrect because during TMR ABORT target-core will drop the outstanding se_cmd->cmd_kref references once it has quiesced the se_cmd via transport_wait_for_tasks(), and in the case of qla2xxx it should not attempt to do it's own transport_generic_free_cmd() once the abort has occured. As reported by Pascal, this was originally manifesting as a BUG_ON(cmd->cmd_in_wq) in qlt_free_cmd() during TMR ABORT, with a LIO backend that had sufficently high enough WRITE latency to trigger a host side TMR ABORT_TASK. In addition, for the case in tcm_qla2xxx_write_pending_status() and tcm_qla2xxx_handle_data_work() that waits for outstanding FCP WRITE data transfer to complete before preceeding with a TMR ABORT, avoid se_cmd->t_transport_stop_comp that is already used by transport_wait_for_tasks() and use a qla2xxx internal struct completion instead. Reported-by:
Pascal de Bruijn <p.debruijn@unilogic.nl> Tested-by:
Himanshu Madhani <himanshu.madhani@cavium.com> Cc: Quinn Tran <quinn.tran@cavium.com> Cc: <stable@vger.kernel.org> # 3.10+ Signed-off-by:
-
Nicholas Bellinger authored
This patch fixes a NULL pointer dereference in isert_login_recv_done() of isert_conn->cm_id due to isert_cma_handler() -> isert_connect_error() resetting isert_conn->cm_id = NULL during a failed login attempt. As per Sagi, we will always see the completion of all recv wrs posted on the qp (given that we assigned a ->done handler), this is a FLUSH error completion, we just don't get to verify that because we deref NULL before. The issue here, was the assumption that dereferencing the connection cm_id is always safe, which is not true since: commit 4a579da2 Author: Sagi Grimberg <sagig@mellanox.com> Date: Sun Mar 29 15:52:04 2015 +0300 iser-target: Fix possible deadlock in RDMA_CM connection error As I see it, we have a direct reference to the isert_device from isert_conn which is the one-liner fix that we actually need like we do in isert_rdma_read_done() and isert_rdma_write_done(). Reported-by:Andrea Righi <righi.andrea@gmail.com> Tested-by:
Sagi Grimberg <sagi@grimberg.me> Cc: <stable@vger.kernel.org> # 3.10+ Signed-off-by:
-
Colin Ian King authored
The array tcmu_attrib_attrs does not need to be in global scope, so make it static. Cleans up sparse warning: "symbol 'tcmu_attrib_attrs' was not declared. Should it be static?" Signed-off-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Xiubo Li authored
Because the unmap code just after the schdule() returned may take a long time and if the kthread_stop() is fired just when in this routine, the module removal maybe stuck too. Signed-off-by:
-
Jiang Yi authored
This patch addresses a COMPARE_AND_WRITE se_device->caw_sem leak, that would be triggered during normal se_cmd shutdown or abort via __transport_wait_for_tasks(). This would occur because target_complete_cmd() would catch this early and do complete_all(&cmd->t_transport_stop_comp), but since target_complete_ok_work() or target_complete_failure_work() are never called to invoke se_cmd->transport_complete_callback(), the COMPARE_AND_WRITE specific callbacks never release caw_sem. To address this special case, go ahead and release caw_sem directly from target_complete_cmd(). (Remove '&& success' from check, to release caw_sem regardless of scsi_status - nab) Signed-off-by:
Jiang Yi <jiangyilism@gmail.com> Cc: <stable@vger.kernel.org> # 3.14+ Signed-off-by:
-
Bryant G. Ly authored
This patch adds more info about the attribute being changed, so that usersapce can easily figure out what is happening. Signed-off-by:
-
Bryant G. Ly authored
This allows for userspace to change the device path after it has been created. Thus giving the user the ability to change the path. The use case for this is to allow for virtual optical to have media change. Signed-off-by:
-
Bryant G. Ly authored
Allow tcmu backstores to be able to set the device size after it has been configured via set attribute. Part of support in userspace to support certain backstores changing device size. Signed-off-by:
-
Bryant G. Ly authored
This gives tcmu the ability to handle events that can cause reconfiguration, such as resize, path changes, write_cache, etc... Signed-off-by:
-
Bryant G. Ly authored
This will enable the toggling of write_cache in tcmu through targetcli-fb Signed-off-by:
-
Bryant G. Ly authored
ibmvscsis always returned 0 for the tpg/tag, since it did not parse the value passed in by the user. When functions like ALUA members exports the value, it will be incorrect because targetcli/rtslib starts the tpg numbering at 1. Signed-off-by:
-
Bart Van Assche authored
If an iSCSI command is rejected before iscsit_process_scsi_cmd() is called, .reject_reason is set but iscsit_process_scsi_cmd() is not called. This means that the "if (cmd->reject_reason) ..." code in this function can be removed without changing the behavior of this function. Signed-off-by:
Bart Van Assche <bart.vanassche@sandisk.com> Cc: Hannes Reinecke <hare@suse.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Andy Grover <agrover@redhat.com> Cc: David Disseldorp <ddiss@suse.de> Signed-off-by:
-
Bart Van Assche authored
Since .se_tfo is only set if a command has been submitted to the LIO core, check .se_tfo instead of .iscsi_opcode. Since __iscsit_free_cmd() only affects SCSI commands but not TMFs, calling that function for TMFs does not change behavior. This patch does not change the behavior of iscsit_free_cmd(). Signed-off-by:
-
Bart Van Assche authored
Initialize .data_direction to DMA_NONE in iscsit_allocate_cmd() such that the second argument of __iscsit_free_cmd() can be left out. Note: this patch causes the first part of __iscsit_free_cmd() no longer to be skipped for TMFs. That's fine since no data segments are associated with TMFs. Signed-off-by:
Hannes Reinecke <hare@suse.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Andy Grover <agrover@redhat.com> Cc: David Disseldorp <ddiss@suse.de> Signed-off-by:
-
Bart Van Assche authored
Target drivers must guarantee that struct se_cmd and struct se_tmr_req exist as long as target_tmr_work() is in progress. This is why the tcm_loop driver today passes 1 as second argument to transport_generic_free_cmd() from inside the TMF code. Instead of making the TMF code wait, make the TMF code obtain two references (SCF_ACK_KREF) and drop one reference from inside the .check_stop_free() callback. Signed-off-by:
-
Bart Van Assche authored
Use target_submit_tmr() instead of open-coding this function. The only functional change is that TMFs are now added to sess_cmd_list, something the current code does not do. This behavior change is a bug fix because it makes LUN RESETs wait for other TMFs that are in progress for the same LUN. Signed-off-by:
-
Bart Van Assche authored
This patch simplifies the implementation of the tcm_loop driver but does not change its behavior. Signed-off-by:
-
Bart Van Assche authored
This patch simplifies the tcm_loop implementation but does not change any functionality. Signed-off-by:
-
Bart Van Assche authored
Introduce target_show_cmd() and use it where appropriate. If transport_wait_for_tasks() takes too long, make it show the state of the command it is waiting for. (Add missing brackets around multi-line conditions - nab) Signed-off-by:
-
Nicholas Bellinger authored
With commit 25cdda95 in place to address the initial login PDU asynchronous socket close OOPs, go ahead and kill off the left-over iscsi_target_do_cleanup() and ->login_cleanup_work. Reported-by:
-
Bart Van Assche authored
Target drivers must guarantee that struct se_cmd and struct se_tmr_req exist as long as target_tmr_work() is in progress. Since the last access by the LIO core is a call to .check_stop_free() and since the Xen scsiback .check_stop_free() drops a reference to the TMF, it is already guaranteed that the struct se_cmd that corresponds to the TMF exists as long as target_tmr_work() is in progress. Hence change the second argument of transport_generic_free_cmd() from 1 into 0. Signed-off-by:
Juergen Gross <jgross@suse.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.com> Cc: David Disseldorp <ddiss@suse.de> Cc: xen-devel@lists.xenproject.org Signed-off-by:
-
Bart Van Assche authored
This patch simplifies the implementation of the scsiback driver but does not change its behavior. Signed-off-by:
-
Bart Van Assche authored
scsiback_release_cmd() must not dereference se_cmd->se_tmr_req because that memory is freed by target_free_cmd_mem() before scsiback_release_cmd() is called. Fix this use-after-free by inlining struct scsiback_tmr into struct vscsibk_pend. Signed-off-by:
-
Bart Van Assche authored
Do not only report the state of the I/O context before srpt_abort_cmd() was called but also the new state assigned by srpt_abort_cmd() Signed-off-by:
-
Bart Van Assche authored
Move the code for parsing an XCOPY command from the context of the iSCSI receiver thread to the context of the XCOPY workqueue. Keep the simple XCOPY checks in the context of the iSCSI receiver thread. Move the code for allocating and freeing struct xcopy_op from the code that parses an XCOPY command to its caller. This patch fixes the following deadlock: ====================================================== [ INFO: possible circular locking dependency detected ] 4.10.0-rc7-dbg+ #1 Not tainted ------------------------------------------------------- rmdir/13321 is trying to acquire lock: (&sess->cmdsn_mutex){+.+.+.}, at: [<ffffffffa02cb47d>] iscsit_free_all_ooo_cmdsns+0x2d/0xb0 [iscsi_target_mod] but task is already holding lock: (&sb->s_type->i_mutex_key#14){++++++}, at: [<ffffffff811c6e20>] vfs_rmdir+0x50/0x140 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (&sb->s_type->i_mutex_key#14){++++++}: lock_acquire+0x71/0x90 down_write+0x3f/0x70 configfs_depend_item+0x3a/0xb0 [configfs] target_depend_item+0x13/0x20 [target_core_mod] target_xcopy_locate_se_dev_e4+0xdd/0x1a0 [target_core_mod] target_do_xcopy+0x34b/0x970 [target_core_mod] __target_execute_cmd+0x22/0xa0 [target_core_mod] target_execute_cmd+0x233/0x2c0 [target_core_mod] iscsit_execute_cmd+0x208/0x270 [iscsi_target_mod] iscsit_sequence_cmd+0x10b/0x190 [iscsi_target_mod] iscsit_get_rx_pdu+0x37d/0xcd0 [iscsi_target_mod] iscsi_target_rx_thread+0x6e/0xa0 [iscsi_target_mod] kthread+0x102/0x140 ret_from_fork+0x31/0x40 -> #0 (&sess->cmdsn_mutex){+.+.+.}: __lock_acquire+0x10e6/0x1260 lock_acquire+0x71/0x90 mutex_lock_nested+0x5f/0x670 iscsit_free_all_ooo_cmdsns+0x2d/0xb0 [iscsi_target_mod] iscsit_close_session+0xac/0x200 [iscsi_target_mod] lio_tpg_close_session+0x9f/0xb0 [iscsi_target_mod] target_shutdown_sessions+0xc3/0xd0 [target_core_mod] core_tpg_del_initiator_node_acl+0x91/0x140 [target_core_mod] target_fabric_nacl_base_release+0x20/0x30 [target_core_mod] config_item_release+0x5a/0xc0 [configfs] config_item_put+0x1d/0x1f [configfs] configfs_rmdir+0x1a6/0x300 [configfs] vfs_rmdir+0xb7/0x140 do_rmdir+0x1f4/0x200 SyS_rmdir+0x11/0x20 entry_SYSCALL_64_fastpath+0x23/0xc6 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&sb->s_type->i_mutex_key#14); lock(&sess->cmdsn_mutex); lock(&sb->s_type->i_mutex_key#14); lock(&sess->cmdsn_mutex); *** DEADLOCK *** 3 locks held by rmdir/13321: #0: (sb_writers#10){.+.+.+}, at: [<ffffffff811e1aff>] mnt_want_write+0x1f/0x50 #1: (&default_group_class[depth - 1]#2/1){+.+.+.}, at: [<ffffffff811cc8ce>] do_rmdir+0x15e/0x200 #2: (&sb->s_type->i_mutex_key#14){++++++}, at: [<ffffffff811c6e20>] vfs_rmdir+0x50/0x140 stack backtrace: CPU: 2 PID: 13321 Comm: rmdir Not tainted 4.10.0-rc7-dbg+ #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.0.0-prebuilt.qemu-project.org 04/01/2014 Call Trace: dump_stack+0x86/0xc3 print_circular_bug+0x1c7/0x220 __lock_acquire+0x10e6/0x1260 lock_acquire+0x71/0x90 mutex_lock_nested+0x5f/0x670 iscsit_free_all_ooo_cmdsns+0x2d/0xb0 [iscsi_target_mod] iscsit_close_session+0xac/0x200 [iscsi_target_mod] lio_tpg_close_session+0x9f/0xb0 [iscsi_target_mod] target_shutdown_sessions+0xc3/0xd0 [target_core_mod] core_tpg_del_initiator_node_acl+0x91/0x140 [target_core_mod] target_fabric_nacl_base_release+0x20/0x30 [target_core_mod] config_item_release+0x5a/0xc0 [configfs] config_item_put+0x1d/0x1f [configfs] configfs_rmdir+0x1a6/0x300 [configfs] vfs_rmdir+0xb7/0x140 do_rmdir+0x1f4/0x200 SyS_rmdir+0x11/0x20 entry_SYSCALL_64_fastpath+0x23/0xc6 Signed-off-by: -
Bart Van Assche authored
Introduce the function get_unaligned_be24(). Use {get,put}_unaligned_be*() where appropriate. This patch does not change any functionality. Signed-off-by:Christoph Hellwig <hch@lst.de> Reviewed-by:
-
Bart Van Assche authored
Avoid that aborting a command before it has been submitted onto a workqueue triggers the following warning: INFO: trying to register non-static key. the code is fine but needs lockdep annotation. turning off the locking correctness validator. CPU: 3 PID: 46 Comm: kworker/u8:1 Not tainted 4.12.0-rc2-dbg+ #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.0.0-prebuilt.qemu-project.org 04/01/2014 Workqueue: tmr-iblock target_tmr_work [target_core_mod] Call Trace: dump_stack+0x86/0xcf register_lock_class+0xe8/0x570 __lock_acquire+0xa1/0x11d0 lock_acquire+0x59/0x80 flush_work+0x42/0x2b0 __cancel_work_timer+0x10c/0x180 cancel_work_sync+0xb/0x10 core_tmr_lun_reset+0x352/0x740 [target_core_mod] target_tmr_work+0xd6/0x130 [target_core_mod] process_one_work+0x1ca/0x3f0 worker_thread+0x49/0x3b0 kthread+0x109/0x140 ret_from_fork+0x31/0x40 Signed-off-by:
-
Bart Van Assche authored
The last user of se_device.dev_list was removed through commit 0fd97ccf ("target: kill struct se_subsystem_dev"). Hence also remove se_device.dev_list. Signed-off-by:
-
Bart Van Assche authored
Now that a symbolic value has been introduced for WRITE_VERIFY_16, use it. This patch does not change any functionality. References: commit c2d26f18 ("target: Add WRITE_VERIFY_16") Signed-off-by:
-
