kernel.git/drivers/target/target_core_internal.h, branch linux-5.16.y

scsi: target: Fix ordered tag handling

2021-10-19T02:38:35Z

This patch fixes the following bugs: 1. If there are multiple ordered cmds queued and multiple simple cmds completing, target_restart_delayed_cmds() could be called on different CPUs and each instance could start a ordered cmd. They could then run in different orders than they were queued. 2. target_restart_delayed_cmds() and target_handle_task_attr() can race where: 1. target_handle_task_attr() has passed the simple_cmds == 0 check. 2. transport_complete_task_attr() then decrements simple_cmds to 0. 3. transport_complete_task_attr() runs target_restart_delayed_cmds() and it does not see any cmds on the delayed_cmd_list. 4. target_handle_task_attr() adds the cmd to the delayed_cmd_list. The cmd will then end up timing out. 3. If we are sent > 1 ordered cmds and simple_cmds == 0, we can execute them out of order, because target_handle_task_attr() will hit that simple_cmds check first and return false for all ordered cmds sent. 4. We run target_restart_delayed_cmds() after every cmd completion, so if there is more than 1 simple cmd running, we start executing ordered cmds after that first cmd instead of waiting for all of them to complete. 5. Ordered cmds are not supposed to start until HEAD OF QUEUE and all older cmds have completed, and not just simple. 6. It's not a bug but it doesn't make sense to take the delayed_cmd_lock for every cmd completion when ordered cmds are almost never used. Just replacing that lock with an atomic increases IOPs by up to 10% when completions are spread over multiple CPUs and there are multiple sessions/ mqs/thread accessing the same device. This patch moves the queued delayed handling to a per device work to serialze the cmd executions for each device and adds a new counter to track HEAD_OF_QUEUE and SIMPLE cmds. We can then check the new counter to determine when to run the work on the completion path. Link: https://lore.kernel.org/r/20210930020422.92578-3-michael.christie@oracle.com Signed-off-by: Mike Christie Signed-off-by: Martin K. Petersen

scsi: target: core: Make completion affinity configurable

2021-03-04T22:37:03Z

It may not always be best to complete the IO on same CPU as it was submitted on. This commit allows userspace to configure it. This has been useful for vhost-scsi where we have a single thread for submissions and completions. If we force the completion on the submission CPU we may be adding conflicts with what the user has setup in the lower levels with settings like the block layer rq_affinity or the driver's IRQ or softirq (the network's rps_cpus value) settings. We may also want to set it up where the vhost thread runs on CPU N and does its submissions/completions there, and then have LIO do its completion booking on CPU M, but can't configure the lower levels due to issues like using dm-multipath with lots of paths (the path selector can throw commands all over the system because it's only taking into account latency/throughput at its level). The new setting is in: /sys/kernel/config/target/$fabric/$target/param/cmd_completion_affinity Writing: -1 -> Gives the current default behavior of completing on the submission CPU. -2 -> Completes the cmd on the CPU the lower layers sent it to us from. > 0 -> Completes on the CPU userspace has specified. Link: https://lore.kernel.org/r/20210227170006.5077-26-michael.christie@oracle.com Reviewed-by: Himanshu Madhani Signed-off-by: Mike Christie Signed-off-by: Martin K. Petersen

scsi: target: core: Add workqueue based cmd submission

2021-03-04T22:37:02Z

loop and vhost/scsi do their target cmd submission from driver workqueues. This allows them to avoid an issue where the backend may block waiting for resources like tags/requests, mem/locks, etc and that ends up blocking their entire submission path and for the case of vhost-scsi both the submission and completion path. This patch adds a helper drivers can use to submit from a LIO workqueue. This code will then be extended in the next patches to fix the plugging of backend devices. We are only converting vhost/loop initially, but the workqueue based submission will work for other drivers and have similar benefits where the main target loops will not end up blocking one some backend resource. Link: https://lore.kernel.org/r/20210227170006.5077-17-michael.christie@oracle.com Tested-by: Laurence Oberman Reviewed-by: Bodo Stroesser Signed-off-by: Mike Christie Signed-off-by: Martin K. Petersen

scsi: target: Fix xcopy sess release leak

2020-07-08T04:14:34Z

transport_init_session can allocate memory via percpu_ref_init, and target_xcopy_release_pt never frees it. This adds a transport_uninit_session function to handle cleanup of resources allocated in the init function. Link: https://lore.kernel.org/r/1593654203-12442-3-git-send-email-michael.christie@oracle.com Signed-off-by: Mike Christie Signed-off-by: Martin K. Petersen

scsi: target/core: Make ABORT and LUN RESET handling synchronous

2018-12-08T02:22:55Z

Instead of invoking target driver callback functions from the context that handles an abort or LUN RESET task management function, only set the abort flag from that context and perform the actual abort handling from the context of the regular command processing flow. This approach has the advantage that the task management code becomes much easier to read and to verify since the number of potential race conditions against the command processing flow is strongly reduced. This patch has been tested by running the following two shell commands concurrently for about ten minutes for both the iSCSI and the SRP target drivers ($dev is an initiator device node connected with storage provided by the target driver under test): * fio with data verification enabled on a filesystem mounted on top of $dev. * while true; do sg_reset -d $dev; echo -n .; sleep .1; done Cc: Nicholas Bellinger Cc: Mike Christie Cc: Christoph Hellwig Cc: David Disseldorp Cc: Hannes Reinecke Signed-off-by: Bart Van Assche Signed-off-by: Martin K. Petersen

scsi: target: Fold core_tmr_handle_tas_abort() into transport_cmd_finish_abort()

2018-07-02T20:44:31Z

For the two calls to transport_cmd_finish_abort() outside core_tmr_handle_tas_abort() it is guaranteed that CMD_T_TAS is not set. Use this property to fold core_tmr_handle_tas_abort() into transport_cmd_finish_abort(). This patch does not change any functionality. Signed-off-by: Bart Van Assche Reviewed-by: Mike Christie Cc: Hannes Reinecke Cc: Christoph Hellwig Signed-off-by: Martin K. Petersen

scsi: target: prefer dbroot of /etc/target over /var/target

2018-04-19T04:47:02Z

The target database root directory, dbroot, has defaulted to /var/target for a while, but its main client, targetcli-fb, has been moving it to /etc/target for quite some time. With the plethora of target drivers now appearing, it has become more difficult to initialize this attribute before use by any child drivers. If the directory /etc/target exists, use that as the DB root. Otherwise, fall back to using /var/target. The ability to override this dbroot attribute still exists via sysfs. Signed-off-by: Lee Duncan Reviewed-by: Christoph Hellwig Signed-off-by: Martin K. Petersen

target core: add device action configfs files

2018-01-17T02:05:04Z

This patch adds a new group of files that are to be used to have the kernel module execution some action. The next patch will have target_core_user use the group/files to be able to block a device and to reset its memory buffer used to pass commands between user/kernel space. This type of file is different from the existing device attributes in that they may be write only and when written to they result in the kernel module executing some function. These need to be separate from the normal device attributes which get/set device values so userspace can continue to loop over all the attribs and get/set them during initialization. Signed-off-by: Mike Christie Signed-off-by: Nicholas Bellinger

target-core: don't use "const char*" for a buffer that is written to

2018-01-12T23:07:09Z

iscsi_parse_pr_out_transport_id launders the const away via a call to strstr(), and then modifies the buffer (writing a nul byte) through the return value. It's cleaner to be honest and simply declare the parameter as "char*", fixing up the call chain, and allowing us to drop the cast in the return statement. Amusingly, the two current callers found it necessary to cast a non-const pointer to a const. Signed-off-by: Rasmus Villemoes Signed-off-by: Nicholas Bellinger

Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending

2017-11-25T05:19:20Z

Pull SCSI target updates from Nicholas Bellinger: "This series is predominantly bug-fixes, with a few small improvements that have been outstanding over the last release cycle. As usual, the associated bug-fixes have CC' tags for stable. Also, things have been particularly quiet wrt new developments the last months, with most folks continuing to focus on stability atop 4.x stable kernels for their respective production configurations. Also at this point, the stable trees have been synced up with mainline. This will continue to be a priority, as production users tend to run exclusively atop stable kernels, a few releases behind mainline. The highlights include: - Fix PR PREEMPT_AND_ABORT null pointer dereference regression in v4.11+ (tangwenji) - Fix OOPs during removing TCMU device (Xiubo Li + Zhang Zhuoyu) - Add netlink command reply supported option for each device (Kenjiro Nakayama) - cxgbit: Abort the TCP connection in case of data out timeout (Varun Prakash) - Fix PR/ALUA file path truncation (David Disseldorp) - Fix double se_cmd completion during ->cmd_time_out (Mike Christie) - Fix QUEUE_FULL + SCSI task attribute handling in 4.1+ (Bryant Ly + nab) - Fix quiese during transport_write_pending_qf endless loop (nab) - Avoid early CMD_T_PRE_EXECUTE failures during ABORT_TASK in 3.14+ (Don White + nab)" * 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending: (35 commits) tcmu: Add a missing unlock on an error path tcmu: Fix some memory corruption iscsi-target: Fix non-immediate TMR reference leak iscsi-target: Make TASK_REASSIGN use proper se_cmd->cmd_kref target: Avoid early CMD_T_PRE_EXECUTE failures during ABORT_TASK target: Fix quiese during transport_write_pending_qf endless loop target: Fix caw_sem leak in transport_generic_request_failure target: Fix QUEUE_FULL + SCSI task attribute handling iSCSI-target: Use common error handling code in iscsi_decode_text_input() target/iscsi: Detect conn_cmd_list corruption early target/iscsi: Fix a race condition in iscsit_add_reject_from_cmd() target/iscsi: Modify iscsit_do_crypto_hash_buf() prototype target/iscsi: Fix endianness in an error message target/iscsi: Use min() in iscsit_dump_data_payload() instead of open-coding it target/iscsi: Define OFFLOAD_BUF_SIZE once target: Inline transport_put_cmd() target: Suppress gcc 7 fallthrough warnings target: Move a declaration of a global variable into a header file tcmu: fix double se_cmd completion target: return SAM_STAT_TASK_SET_FULL for TCM_OUT_OF_RESOURCES ...