From e517d3133f62c27b211f305a6dbd6f6ccac0db1b Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 26 Jul 2005 10:18:45 -0400 Subject: [SCSI] add missing scan mutex to scsi_scan_target() This patch (as543) adds a private entry point to scsi_scan_target, for use when the caller already owns the scan_mutex, and updates the kerneldoc for that routine (which was badly out-of-date). It converts scsi_scan_channel to use the new entry point. Lastly, it modifies scsi_get_host_dev to make it acquire the scan_mutex, necessary since the routine adds a new scsi_device even if it doesn't do any actual scanning. Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_scan.c | 67 +++++++++++++++++++++++++++++------------------- 1 file changed, 40 insertions(+), 27 deletions(-) diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index 19c9a23..76577fa 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -1276,27 +1276,8 @@ void scsi_rescan_device(struct device *dev) } EXPORT_SYMBOL(scsi_rescan_device); -/** - * scsi_scan_target - scan a target id, possibly including all LUNs on the - * target. - * @sdevsca: Scsi_Device handle for scanning - * @shost: host to scan - * @channel: channel to scan - * @id: target id to scan - * - * Description: - * Scan the target id on @shost, @channel, and @id. Scan at least LUN - * 0, and possibly all LUNs on the target id. - * - * Use the pre-allocated @sdevscan as a handle for the scanning. This - * function sets sdevscan->host, sdevscan->id and sdevscan->lun; the - * scanning functions modify sdevscan->lun. - * - * First try a REPORT LUN scan, if that does not scan the target, do a - * sequential scan of LUNs on the target id. - **/ -void scsi_scan_target(struct device *parent, unsigned int channel, - unsigned int id, unsigned int lun, int rescan) +static void __scsi_scan_target(struct device *parent, unsigned int channel, + unsigned int id, unsigned int lun, int rescan) { struct Scsi_Host *shost = dev_to_shost(parent); int bflags = 0; @@ -1310,9 +1291,7 @@ void scsi_scan_target(struct device *parent, unsigned int channel, */ return; - starget = scsi_alloc_target(parent, channel, id); - if (!starget) return; @@ -1358,6 +1337,33 @@ void scsi_scan_target(struct device *parent, unsigned int channel, put_device(&starget->dev); } + +/** + * scsi_scan_target - scan a target id, possibly including all LUNs on the + * target. + * @parent: host to scan + * @channel: channel to scan + * @id: target id to scan + * @lun: Specific LUN to scan or SCAN_WILD_CARD + * @rescan: passed to LUN scanning routines + * + * Description: + * Scan the target id on @parent, @channel, and @id. Scan at least LUN 0, + * and possibly all LUNs on the target id. + * + * First try a REPORT LUN scan, if that does not scan the target, do a + * sequential scan of LUNs on the target id. + **/ +void scsi_scan_target(struct device *parent, unsigned int channel, + unsigned int id, unsigned int lun, int rescan) +{ + struct Scsi_Host *shost = dev_to_shost(parent); + + down(&shost->scan_mutex); + if (scsi_host_scan_allowed(shost)) + __scsi_scan_target(parent, channel, id, lun, rescan); + up(&shost->scan_mutex); +} EXPORT_SYMBOL(scsi_scan_target); static void scsi_scan_channel(struct Scsi_Host *shost, unsigned int channel, @@ -1383,10 +1389,12 @@ static void scsi_scan_channel(struct Scsi_Host *shost, unsigned int channel, order_id = shost->max_id - id - 1; else order_id = id; - scsi_scan_target(&shost->shost_gendev, channel, order_id, lun, rescan); + __scsi_scan_target(&shost->shost_gendev, channel, + order_id, lun, rescan); } else - scsi_scan_target(&shost->shost_gendev, channel, id, lun, rescan); + __scsi_scan_target(&shost->shost_gendev, channel, + id, lun, rescan); } int scsi_scan_host_selected(struct Scsi_Host *shost, unsigned int channel, @@ -1484,12 +1492,15 @@ void scsi_forget_host(struct Scsi_Host *shost) */ struct scsi_device *scsi_get_host_dev(struct Scsi_Host *shost) { - struct scsi_device *sdev; + struct scsi_device *sdev = NULL; struct scsi_target *starget; + down(&shost->scan_mutex); + if (!scsi_host_scan_allowed(shost)) + goto out; starget = scsi_alloc_target(&shost->shost_gendev, 0, shost->this_id); if (!starget) - return NULL; + goto out; sdev = scsi_alloc_sdev(starget, 0, NULL); if (sdev) { @@ -1497,6 +1508,8 @@ struct scsi_device *scsi_get_host_dev(struct Scsi_Host *shost) sdev->borken = 0; } put_device(&starget->dev); + out: + up(&shost->scan_mutex); return sdev; } EXPORT_SYMBOL(scsi_get_host_dev); -- cgit v1.1 From 903f4fed858a7b56b260cbd55d174fe54d188fb7 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 26 Jul 2005 10:20:53 -0400 Subject: [SCSI] fix callers of scsi_remove_device() who already hold the scan muted This patch (as544) adds a private entry point to scsi_remove_device, for use when callers already own the scan_mutex. The appropriate callers are modified to use the new entry point. Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_priv.h | 1 + drivers/scsi/scsi_sysfs.c | 28 +++++++++++++++------------- 2 files changed, 16 insertions(+), 13 deletions(-) diff --git a/drivers/scsi/scsi_priv.h b/drivers/scsi/scsi_priv.h index ee6de17..d05f778 100644 --- a/drivers/scsi/scsi_priv.h +++ b/drivers/scsi/scsi_priv.h @@ -124,6 +124,7 @@ extern void scsi_sysfs_unregister(void); extern void scsi_sysfs_device_initialize(struct scsi_device *); extern int scsi_sysfs_target_initialize(struct scsi_device *); extern struct scsi_transport_template blank_transport_template; +extern void __scsi_remove_device(struct scsi_device *); extern struct bus_type scsi_bus_type; diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index dae59d1..b8052d5 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -653,7 +653,7 @@ int scsi_sysfs_add_sdev(struct scsi_device *sdev) error = attr_add(&sdev->sdev_gendev, sdev->host->hostt->sdev_attrs[i]); if (error) { - scsi_remove_device(sdev); + __scsi_remove_device(sdev); goto out; } } @@ -667,7 +667,7 @@ int scsi_sysfs_add_sdev(struct scsi_device *sdev) scsi_sysfs_sdev_attrs[i]); error = device_create_file(&sdev->sdev_gendev, attr); if (error) { - scsi_remove_device(sdev); + __scsi_remove_device(sdev); goto out; } } @@ -687,17 +687,10 @@ int scsi_sysfs_add_sdev(struct scsi_device *sdev) return error; } -/** - * scsi_remove_device - unregister a device from the scsi bus - * @sdev: scsi_device to unregister - **/ -void scsi_remove_device(struct scsi_device *sdev) +void __scsi_remove_device(struct scsi_device *sdev) { - struct Scsi_Host *shost = sdev->host; - - down(&shost->scan_mutex); if (scsi_device_set_state(sdev, SDEV_CANCEL) != 0) - goto out; + return; class_device_unregister(&sdev->sdev_classdev); device_del(&sdev->sdev_gendev); @@ -706,8 +699,17 @@ void scsi_remove_device(struct scsi_device *sdev) sdev->host->hostt->slave_destroy(sdev); transport_unregister_device(&sdev->sdev_gendev); put_device(&sdev->sdev_gendev); -out: - up(&shost->scan_mutex); +} + +/** + * scsi_remove_device - unregister a device from the scsi bus + * @sdev: scsi_device to unregister + **/ +void scsi_remove_device(struct scsi_device *sdev) +{ + down(&sdev->host->scan_mutex); + __scsi_remove_device(sdev); + up(&sdev->host->scan_mutex); } EXPORT_SYMBOL(scsi_remove_device); -- cgit v1.1 From 286f3e13a1dc7f32407629fbd7aabc8ea78c62b5 Mon Sep 17 00:00:00 2001 From: Neil Brown Date: Fri, 2 Sep 2005 13:13:54 +1000 Subject: [SCSI] fix possible deadlock in scsi_lib.c If a filesystem, while writing out data, decides that it is good to issue a cache flush on a SCSI drive (or other 'sd' device), it will call blkdev_issue_flush which calls ->issue_flush_fn which is scsi_issue_flush_fn. This calls sd_issue_flush which calls sd_sync_cache, which calls scsi_execute_request. This will (as sshdr != NULL) call kmalloc(SCSI_SENSE_BUFFERSIZE, GFP_KERNEL) If memory is tight, the presence of GFP_KERNEL may cause write requests to be sent to some filesystem to free up memory, however if that filesystem is waiting for the issue_flush_fn to complete, you could get a deadlock. I wonder if it might be more appropriate to use GFP_NOIO as in the following patch. I wonder if it might be even more appropriate to cope better with a kmalloc failure, especially as in this use, sd_sync_cache only will use the sense information to print out a more informative error message. Signed-off-by: Neil Brown Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 77f2d44..2ad60f1 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -339,7 +339,7 @@ int scsi_execute_req(struct scsi_device *sdev, const unsigned char *cmd, int result; if (sshdr) { - sense = kmalloc(SCSI_SENSE_BUFFERSIZE, GFP_KERNEL); + sense = kmalloc(SCSI_SENSE_BUFFERSIZE, GFP_NOIO); if (!sense) return DRIVER_ERROR << 24; memset(sense, 0, SCSI_SENSE_BUFFERSIZE); -- cgit v1.1 From e91442b635be776ea205fba233bdd5bc74b62bc3 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Fri, 9 Sep 2005 10:44:16 -0500 Subject: [SCSI] SCSI core: fix leakage of scsi_cmnd's From: Alan Stern This patch (as559b) adds a new routine, scsi_unprep_request, which gets called every place a request is requeued. (That includes scsi_queue_insert as well as scsi_requeue_command.) It also changes scsi_kill_requests to make it call __scsi_done with result equal to DID_NO_CONNECT << 16. (I'm not sure if it's necessary to call scsi_init_cmd_errh here; maybe you can check on that.) Finally, the patch changes the return value from scsi_end_request, to avoid returning a stale pointer in the case where the request was requeued. Fortunately the return value is used in only place, and the change actually simplified it. Signed-off-by: Alan Stern Rejections fixed up and Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 111 +++++++++++++++++++++++++++++++----------------- 1 file changed, 71 insertions(+), 40 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 2ad60f1..003f8cf 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -97,6 +97,30 @@ int scsi_insert_special_req(struct scsi_request *sreq, int at_head) } static void scsi_run_queue(struct request_queue *q); +static void scsi_release_buffers(struct scsi_cmnd *cmd); + +/* + * Function: scsi_unprep_request() + * + * Purpose: Remove all preparation done for a request, including its + * associated scsi_cmnd, so that it can be requeued. + * + * Arguments: req - request to unprepare + * + * Lock status: Assumed that no locks are held upon entry. + * + * Returns: Nothing. + */ +static void scsi_unprep_request(struct request *req) +{ + struct scsi_cmnd *cmd = req->special; + + req->flags &= ~REQ_DONTPREP; + req->special = (req->flags & REQ_SPECIAL) ? cmd->sc_request : NULL; + + scsi_release_buffers(cmd); + scsi_put_command(cmd); +} /* * Function: scsi_queue_insert() @@ -116,12 +140,14 @@ static void scsi_run_queue(struct request_queue *q); * commands. * Notes: This could be called either from an interrupt context or a * normal process context. + * Notes: Upon return, cmd is a stale pointer. */ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason) { struct Scsi_Host *host = cmd->device->host; struct scsi_device *device = cmd->device; struct request_queue *q = device->request_queue; + struct request *req = cmd->request; unsigned long flags; SCSI_LOG_MLQUEUE(1, @@ -162,8 +188,9 @@ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason) * function. The SCSI request function detects the blocked condition * and plugs the queue appropriately. */ + scsi_unprep_request(req); spin_lock_irqsave(q->queue_lock, flags); - blk_requeue_request(q, cmd->request); + blk_requeue_request(q, req); spin_unlock_irqrestore(q->queue_lock, flags); scsi_run_queue(q); @@ -552,15 +579,16 @@ static void scsi_run_queue(struct request_queue *q) * I/O errors in the middle of the request, in which case * we need to request the blocks that come after the bad * sector. + * Notes: Upon return, cmd is a stale pointer. */ static void scsi_requeue_command(struct request_queue *q, struct scsi_cmnd *cmd) { + struct request *req = cmd->request; unsigned long flags; - cmd->request->flags &= ~REQ_DONTPREP; - + scsi_unprep_request(req); spin_lock_irqsave(q->queue_lock, flags); - blk_requeue_request(q, cmd->request); + blk_requeue_request(q, req); spin_unlock_irqrestore(q->queue_lock, flags); scsi_run_queue(q); @@ -595,13 +623,14 @@ void scsi_run_host_queues(struct Scsi_Host *shost) * * Lock status: Assumed that lock is not held upon entry. * - * Returns: cmd if requeue done or required, NULL otherwise + * Returns: cmd if requeue required, NULL otherwise. * * Notes: This is called for block device requests in order to * mark some number of sectors as complete. * * We are guaranteeing that the request queue will be goosed * at some point during this call. + * Notes: If cmd was requeued, upon return it will be a stale pointer. */ static struct scsi_cmnd *scsi_end_request(struct scsi_cmnd *cmd, int uptodate, int bytes, int requeue) @@ -624,14 +653,15 @@ static struct scsi_cmnd *scsi_end_request(struct scsi_cmnd *cmd, int uptodate, if (!uptodate && blk_noretry_request(req)) end_that_request_chunk(req, 0, leftover); else { - if (requeue) + if (requeue) { /* * Bleah. Leftovers again. Stick the * leftovers in the front of the * queue, and goose the queue again. */ scsi_requeue_command(q, cmd); - + cmd = NULL; + } return cmd; } } @@ -857,15 +887,13 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, * requeueing right here - we will requeue down below * when we handle the bad sectors. */ - cmd = scsi_end_request(cmd, 1, good_bytes, result == 0); /* - * If the command completed without error, then either finish off the - * rest of the command, or start a new one. + * If the command completed without error, then either + * finish off the rest of the command, or start a new one. */ - if (result == 0 || cmd == NULL ) { + if (scsi_end_request(cmd, 1, good_bytes, result == 0) == NULL) return; - } } /* * Now, if we were good little boys and girls, Santa left us a request @@ -880,7 +908,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, * and quietly refuse further access. */ cmd->device->changed = 1; - cmd = scsi_end_request(cmd, 0, + scsi_end_request(cmd, 0, this_count, 1); return; } else { @@ -914,7 +942,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, scsi_requeue_command(q, cmd); result = 0; } else { - cmd = scsi_end_request(cmd, 0, this_count, 1); + scsi_end_request(cmd, 0, this_count, 1); return; } break; @@ -931,7 +959,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, dev_printk(KERN_INFO, &cmd->device->sdev_gendev, "Device not ready.\n"); - cmd = scsi_end_request(cmd, 0, this_count, 1); + scsi_end_request(cmd, 0, this_count, 1); return; case VOLUME_OVERFLOW: if (!(req->flags & REQ_QUIET)) { @@ -941,7 +969,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, __scsi_print_command(cmd->data_cmnd); scsi_print_sense("", cmd); } - cmd = scsi_end_request(cmd, 0, block_bytes, 1); + scsi_end_request(cmd, 0, block_bytes, 1); return; default: break; @@ -972,7 +1000,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes, block_bytes = req->hard_cur_sectors << 9; if (!block_bytes) block_bytes = req->data_len; - cmd = scsi_end_request(cmd, 0, block_bytes, 1); + scsi_end_request(cmd, 0, block_bytes, 1); } } EXPORT_SYMBOL(scsi_io_completion); @@ -1336,19 +1364,24 @@ static inline int scsi_host_queue_ready(struct request_queue *q, } /* - * Kill requests for a dead device + * Kill a request for a dead device */ -static void scsi_kill_requests(request_queue_t *q) +static void scsi_kill_request(struct request *req, request_queue_t *q) { - struct request *req; + struct scsi_cmnd *cmd = req->special; - while ((req = elv_next_request(q)) != NULL) { - blkdev_dequeue_request(req); - req->flags |= REQ_QUIET; - while (end_that_request_first(req, 0, req->nr_sectors)) - ; - end_that_request_last(req); + spin_unlock(q->queue_lock); + if (unlikely(cmd == NULL)) { + printk(KERN_CRIT "impossible request in %s.\n", + __FUNCTION__); + BUG(); } + + scsi_init_cmd_errh(cmd); + cmd->result = DID_NO_CONNECT << 16; + atomic_inc(&cmd->device->iorequest_cnt); + __scsi_done(cmd); + spin_lock(q->queue_lock); } /* @@ -1371,7 +1404,8 @@ static void scsi_request_fn(struct request_queue *q) if (!sdev) { printk("scsi: killing requests for dead queue\n"); - scsi_kill_requests(q); + while ((req = elv_next_request(q)) != NULL) + scsi_kill_request(req, q); return; } @@ -1399,10 +1433,7 @@ static void scsi_request_fn(struct request_queue *q) printk(KERN_ERR "scsi%d (%d:%d): rejecting I/O to offline device\n", sdev->host->host_no, sdev->id, sdev->lun); blkdev_dequeue_request(req); - req->flags |= REQ_QUIET; - while (end_that_request_first(req, 0, req->nr_sectors)) - ; - end_that_request_last(req); + scsi_kill_request(req, q); continue; } @@ -1415,6 +1446,14 @@ static void scsi_request_fn(struct request_queue *q) sdev->device_busy++; spin_unlock(q->queue_lock); + cmd = req->special; + if (unlikely(cmd == NULL)) { + printk(KERN_CRIT "impossible request in %s.\n" + "please mail a stack trace to " + "linux-scsi@vger.kernel.org", + __FUNCTION__); + BUG(); + } spin_lock(shost->host_lock); if (!scsi_host_queue_ready(q, shost, sdev)) @@ -1433,15 +1472,6 @@ static void scsi_request_fn(struct request_queue *q) */ spin_unlock_irq(shost->host_lock); - cmd = req->special; - if (unlikely(cmd == NULL)) { - printk(KERN_CRIT "impossible request in %s.\n" - "please mail a stack trace to " - "linux-scsi@vger.kernel.org", - __FUNCTION__); - BUG(); - } - /* * Finally, initialize any error handling parameters, and set up * the timers for timeouts. @@ -1477,6 +1507,7 @@ static void scsi_request_fn(struct request_queue *q) * cases (host limits or settings) should run the queue at some * later time. */ + scsi_unprep_request(req); spin_lock_irq(q->queue_lock); blk_requeue_request(q, req); sdev->device_busy--; -- cgit v1.1 From 788ce43aa1ad7158f894b6bb3df8ba2f63794c20 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Fri, 9 Sep 2005 13:40:23 -0500 Subject: [SCSI] SCSI core: fix leakage of scsi_cmnd's Actually, just one problem and one cosmetic fix: 1) We need to dequeue for the loop and kill case (it seems easiest simply to dequeue in the scsi_kill_request() routine) 2) There's no real need to drop the queue lock. __scsi_done() is lock agnostic, so since there's no requirement, let's just leave it in to avoid any locking issues. Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 003f8cf..d8d9848 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1370,7 +1370,8 @@ static void scsi_kill_request(struct request *req, request_queue_t *q) { struct scsi_cmnd *cmd = req->special; - spin_unlock(q->queue_lock); + blkdev_dequeue_request(req); + if (unlikely(cmd == NULL)) { printk(KERN_CRIT "impossible request in %s.\n", __FUNCTION__); @@ -1381,7 +1382,6 @@ static void scsi_kill_request(struct request *req, request_queue_t *q) cmd->result = DID_NO_CONNECT << 16; atomic_inc(&cmd->device->iorequest_cnt); __scsi_done(cmd); - spin_lock(q->queue_lock); } /* @@ -1432,7 +1432,6 @@ static void scsi_request_fn(struct request_queue *q) if (unlikely(!scsi_device_online(sdev))) { printk(KERN_ERR "scsi%d (%d:%d): rejecting I/O to offline device\n", sdev->host->host_no, sdev->id, sdev->lun); - blkdev_dequeue_request(req); scsi_kill_request(req, q); continue; } -- cgit v1.1 From 1c8e71d72026ed4c6ba0fdfd7eebd865f4fd1415 Mon Sep 17 00:00:00 2001 From: Douglas Gilbert Date: Fri, 9 Sep 2005 17:18:57 +1000 Subject: [SCSI] sg: do not set VM_IO flag on mmap-ed pages Further to the problem discussed in this post: http://marc.theaimsgroup.com/?l=linux-scsi&m=112540053711489&w=2 It seems that the sg driver does not need to set the VM_IO flag on pages that it memory maps to the user space since they are not from the IO space. Ahmed Teirelbar wants the facility and has tested this patch as I have without adverse effects. The oops protection is still important. Some users really did try and use dio transfers from the sg driver to memory mapped IO space (on a video capture card if my memory serves) during the lk 2.4 series. I'm not sure how successful it was but that will now be politely refused in lk 2.6.13+ . Changelog: - set the page flags for sg's reserved buffer mmap-ed to the user space to VM_RESERVED (rather than VM_RESERVED | VM_IO ) Signed-off-by: Douglas Gilbert Signed-off-by: James Bottomley --- drivers/scsi/sg.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index b1b69d7..9ea4765 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -61,7 +61,7 @@ static int sg_version_num = 30533; /* 2 digits for each component */ #ifdef CONFIG_SCSI_PROC_FS #include -static char *sg_version_date = "20050901"; +static char *sg_version_date = "20050908"; static int sg_proc_init(void); static void sg_proc_cleanup(void); @@ -1299,7 +1299,7 @@ sg_mmap(struct file *filp, struct vm_area_struct *vma) sg_rb_correct4mmap(rsv_schp, 1); /* do only once per fd lifetime */ sfp->mmap_called = 1; } - vma->vm_flags |= (VM_RESERVED | VM_IO); + vma->vm_flags |= VM_RESERVED; vma->vm_private_data = sfp; vma->vm_ops = &sg_mmap_vm_ops; return 0; -- cgit v1.1 From 942fc2fb73f2cac53484ebaf1c4f9af7aefaca83 Mon Sep 17 00:00:00 2001 From: Douglas Gilbert Date: Fri, 9 Sep 2005 20:07:32 +1000 Subject: [SCSI] permit READ DEFECT DATA in block/scsi_ioctl The soon to be released smartmontools 5.34 uses the READ DEFECT DATA command on SCSI disks. A disk that has defect list entries (or worse, an increasing number of them) is at risk. Currently the first invocation of smartctl causes this: scsi: unknown opcode 0x37 message to appear the console and in the log. The READ DEFECT DATA SCSI command does not change the state of a disk. Its opcode (0x37) is valid for SBC devices (e.g. disks) and SMC-2 devices (media changers) where it is called INITIALIZE STATUS ELEMENT WITH RANGE and again doesn't change the external state of the device. Changelog: - mark SCSI opcode 0x37 (READ DEFECT DATA) as safe_for_read Signed-off-by: Douglas Gilbert Acked-by: Jens Axboe Signed-off-by: James Bottomley --- drivers/block/scsi_ioctl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/scsi_ioctl.c b/drivers/block/scsi_ioctl.c index abb2df2..856c227 100644 --- a/drivers/block/scsi_ioctl.c +++ b/drivers/block/scsi_ioctl.c @@ -123,6 +123,7 @@ static int verify_command(struct file *file, unsigned char *cmd) safe_for_read(READ_12), safe_for_read(READ_16), safe_for_read(READ_BUFFER), + safe_for_read(READ_DEFECT_DATA), safe_for_read(READ_LONG), safe_for_read(INQUIRY), safe_for_read(MODE_SENSE), -- cgit v1.1 From c7ebbbce366c02e5657ac6b6059933fe0353b175 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 9 Sep 2005 16:22:50 +0200 Subject: [SCSI] SAS transport class The SAS transport class contains common code to deal with SAS HBAs, an aproximated representation of SAS topologies in the driver model, and various sysfs attributes to expose these topologies and managment interfaces to userspace. In addition to the basic SCSI core objects this transport class introduces two additional intermediate objects: The SAS PHY as represented by struct sas_phy defines an "outgoing" PHY on a SAS HBA or Expander, and the SAS remote PHY represented by struct sas_rphy defines an "incoming" PHY on a SAS Expander or end device. Note that this is purely a software concept, the underlying hardware for a PHY and a remote PHY is the exactly the same. There is no concept of a SAS port in this code, users can see what PHYs form a wide port based on the port_identifier attribute, which is the same for all PHYs in a port. This submission doesn't handle hot-plug addition or removal of SAS devices and thus doesn't do scanning in a workqueue yet, that will be added in phase2 after this submission. In a third phase I will add additional managment infrastructure. I think this submission is ready for 2.6.14, but additional comments are of course very welcome. I'd like to thanks James Smart a lot for his very useful input on the design. Signed-off-by: Christoph Hellwig Signed-off-by: James Bottomley --- drivers/scsi/Kconfig | 7 + drivers/scsi/Makefile | 1 + drivers/scsi/scsi_transport_sas.c | 819 ++++++++++++++++++++++++++++++++++++++ include/scsi/scsi_transport_sas.h | 100 +++++ 4 files changed, 927 insertions(+) create mode 100644 drivers/scsi/scsi_transport_sas.c create mode 100644 include/scsi/scsi_transport_sas.h diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 2d21265..20019b8 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -235,6 +235,13 @@ config SCSI_ISCSI_ATTRS each attached iSCSI device to sysfs, say Y. Otherwise, say N. +config SCSI_SAS_ATTRS + tristate "SAS Transport Attributes" + depends on SCSI + help + If you wish to export transport-specific information about + each attached SAS device to sysfs, say Y. + endmenu menu "SCSI low-level drivers" diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index 4b4fd94..1e4edbd 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -31,6 +31,7 @@ obj-$(CONFIG_RAID_ATTRS) += raid_class.o obj-$(CONFIG_SCSI_SPI_ATTRS) += scsi_transport_spi.o obj-$(CONFIG_SCSI_FC_ATTRS) += scsi_transport_fc.o obj-$(CONFIG_SCSI_ISCSI_ATTRS) += scsi_transport_iscsi.o +obj-$(CONFIG_SCSI_SAS_ATTRS) += scsi_transport_sas.o obj-$(CONFIG_SCSI_AMIGA7XX) += amiga7xx.o 53c7xx.o obj-$(CONFIG_A3000_SCSI) += a3000.o wd33c93.o diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c new file mode 100644 index 0000000..ac4a53a --- /dev/null +++ b/drivers/scsi/scsi_transport_sas.c @@ -0,0 +1,819 @@ +/* + * Copyright (C) 2005 Dell Inc. + * Released under GPL v2. + * + * Serial Attached SCSI (SAS) transport class. + * + * The SAS transport class contains common code to deal with SAS HBAs, + * an aproximated representation of SAS topologies in the driver model, + * and various sysfs attributes to expose these topologies and managment + * interfaces to userspace. + * + * In addition to the basic SCSI core objects this transport class + * introduces two additional intermediate objects: The SAS PHY + * as represented by struct sas_phy defines an "outgoing" PHY on + * a SAS HBA or Expander, and the SAS remote PHY represented by + * struct sas_rphy defines an "incoming" PHY on a SAS Expander or + * end device. Note that this is purely a software concept, the + * underlying hardware for a PHY and a remote PHY is the exactly + * the same. + * + * There is no concept of a SAS port in this code, users can see + * what PHYs form a wide port based on the port_identifier attribute, + * which is the same for all PHYs in a port. + */ + +#include +#include +#include + +#include +#include +#include +#include + + +#define SAS_HOST_ATTRS 0 +#define SAS_PORT_ATTRS 11 +#define SAS_RPORT_ATTRS 5 + +struct sas_internal { + struct scsi_transport_template t; + struct sas_function_template *f; + + struct class_device_attribute private_host_attrs[SAS_HOST_ATTRS]; + struct class_device_attribute private_phy_attrs[SAS_PORT_ATTRS]; + struct class_device_attribute private_rphy_attrs[SAS_RPORT_ATTRS]; + + struct transport_container phy_attr_cont; + struct transport_container rphy_attr_cont; + + /* + * The array of null terminated pointers to attributes + * needed by scsi_sysfs.c + */ + struct class_device_attribute *host_attrs[SAS_HOST_ATTRS + 1]; + struct class_device_attribute *phy_attrs[SAS_PORT_ATTRS + 1]; + struct class_device_attribute *rphy_attrs[SAS_RPORT_ATTRS + 1]; +}; +#define to_sas_internal(tmpl) container_of(tmpl, struct sas_internal, t) + +struct sas_host_attrs { + struct list_head rphy_list; + spinlock_t lock; + u32 next_target_id; +}; +#define to_sas_host_attrs(host) ((struct sas_host_attrs *)(host)->shost_data) + + +/* + * Hack to allow attributes of the same name in different objects. + */ +#define SAS_CLASS_DEVICE_ATTR(_prefix,_name,_mode,_show,_store) \ + struct class_device_attribute class_device_attr_##_prefix##_##_name = \ + __ATTR(_name,_mode,_show,_store) + + +/* + * Pretty printing helpers + */ + +#define sas_bitfield_name_match(title, table) \ +static ssize_t \ +get_sas_##title##_names(u32 table_key, char *buf) \ +{ \ + char *prefix = ""; \ + ssize_t len = 0; \ + int i; \ + \ + for (i = 0; i < sizeof(table)/sizeof(table[0]); i++) { \ + if (table[i].value & table_key) { \ + len += sprintf(buf + len, "%s%s", \ + prefix, table[i].name); \ + prefix = ", "; \ + } \ + } \ + len += sprintf(buf + len, "\n"); \ + return len; \ +} + +#define sas_bitfield_name_search(title, table) \ +static ssize_t \ +get_sas_##title##_names(u32 table_key, char *buf) \ +{ \ + ssize_t len = 0; \ + int i; \ + \ + for (i = 0; i < sizeof(table)/sizeof(table[0]); i++) { \ + if (table[i].value == table_key) { \ + len += sprintf(buf + len, "%s", \ + table[i].name); \ + break; \ + } \ + } \ + len += sprintf(buf + len, "\n"); \ + return len; \ +} + +static struct { + u32 value; + char *name; +} sas_device_type_names[] = { + { SAS_PHY_UNUSED, "unused" }, + { SAS_END_DEVICE, "end device" }, + { SAS_EDGE_EXPANDER_DEVICE, "edge expander" }, + { SAS_FANOUT_EXPANDER_DEVICE, "fanout expander" }, +}; +sas_bitfield_name_search(device_type, sas_device_type_names) + + +static struct { + u32 value; + char *name; +} sas_protocol_names[] = { + { SAS_PROTOCOL_SATA, "sata" }, + { SAS_PROTOCOL_SMP, "smp" }, + { SAS_PROTOCOL_STP, "stp" }, + { SAS_PROTOCOL_SSP, "ssp" }, +}; +sas_bitfield_name_match(protocol, sas_protocol_names) + +static struct { + u32 value; + char *name; +} sas_linkspeed_names[] = { + { SAS_LINK_RATE_UNKNOWN, "Unknown" }, + { SAS_PHY_DISABLED, "Phy disabled" }, + { SAS_LINK_RATE_FAILED, "Link Rate failed" }, + { SAS_SATA_SPINUP_HOLD, "Spin-up hold" }, + { SAS_LINK_RATE_1_5_GBPS, "1.5 Gbit" }, + { SAS_LINK_RATE_3_0_GBPS, "3.0 Gbit" }, +}; +sas_bitfield_name_search(linkspeed, sas_linkspeed_names) + + +/* + * SAS host attributes + */ + +static int sas_host_setup(struct device *dev) +{ + struct Scsi_Host *shost = dev_to_shost(dev); + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + + INIT_LIST_HEAD(&sas_host->rphy_list); + spin_lock_init(&sas_host->lock); + sas_host->next_target_id = 0; + return 0; +} + +static DECLARE_TRANSPORT_CLASS(sas_host_class, + "sas_host", sas_host_setup, NULL, NULL); + +static int sas_host_match(struct attribute_container *cont, + struct device *dev) +{ + struct Scsi_Host *shost; + struct sas_internal *i; + + if (!scsi_is_host_device(dev)) + return 0; + shost = dev_to_shost(dev); + + if (!shost->transportt) + return 0; + if (shost->transportt->host_attrs.ac.class != + &sas_host_class.class) + return 0; + + i = to_sas_internal(shost->transportt); + return &i->t.host_attrs.ac == cont; +} + +static int do_sas_phy_delete(struct device *dev, void *data) +{ + if (scsi_is_sas_phy(dev)) + sas_phy_delete(dev_to_phy(dev)); + return 0; +} + +/** + * sas_remove_host -- tear down a Scsi_Host's SAS data structures + * @shost: Scsi Host that is torn down + * + * Removes all SAS PHYs and remote PHYs for a given Scsi_Host. + * Must be called just before scsi_remove_host for SAS HBAs. + */ +void sas_remove_host(struct Scsi_Host *shost) +{ + device_for_each_child(&shost->shost_gendev, NULL, do_sas_phy_delete); +} +EXPORT_SYMBOL(sas_remove_host); + + +/* + * SAS Port attributes + */ + +#define sas_phy_show_simple(field, name, format_string, cast) \ +static ssize_t \ +show_sas_phy_##name(struct class_device *cdev, char *buf) \ +{ \ + struct sas_phy *phy = transport_class_to_phy(cdev); \ + \ + return snprintf(buf, 20, format_string, cast phy->field); \ +} + +#define sas_phy_simple_attr(field, name, format_string, type) \ + sas_phy_show_simple(field, name, format_string, (type)) \ +static CLASS_DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL) + +#define sas_phy_show_protocol(field, name) \ +static ssize_t \ +show_sas_phy_##name(struct class_device *cdev, char *buf) \ +{ \ + struct sas_phy *phy = transport_class_to_phy(cdev); \ + \ + if (!phy->field) \ + return snprintf(buf, 20, "none\n"); \ + return get_sas_protocol_names(phy->field, buf); \ +} + +#define sas_phy_protocol_attr(field, name) \ + sas_phy_show_protocol(field, name) \ +static CLASS_DEVICE_ATTR(name, S_IRUGO, show_sas_phy_##name, NULL) + +#define sas_phy_show_linkspeed(field) \ +static ssize_t \ +show_sas_phy_##field(struct class_device *cdev, char *buf) \ +{ \ + struct sas_phy *phy = transport_class_to_phy(cdev); \ + \ + return get_sas_linkspeed_names(phy->field, buf); \ +} + +#define sas_phy_linkspeed_attr(field) \ + sas_phy_show_linkspeed(field) \ +static CLASS_DEVICE_ATTR(field, S_IRUGO, show_sas_phy_##field, NULL) + +static ssize_t +show_sas_device_type(struct class_device *cdev, char *buf) +{ + struct sas_phy *phy = transport_class_to_phy(cdev); + + if (!phy->identify.device_type) + return snprintf(buf, 20, "none\n"); + return get_sas_device_type_names(phy->identify.device_type, buf); +} + +static CLASS_DEVICE_ATTR(device_type, S_IRUGO, show_sas_device_type, NULL); + +sas_phy_protocol_attr(identify.initiator_port_protocols, + initiator_port_protocols); +sas_phy_protocol_attr(identify.target_port_protocols, + target_port_protocols); +sas_phy_simple_attr(identify.sas_address, sas_address, "0x%016llx\n", + unsigned long long); +sas_phy_simple_attr(identify.phy_identifier, phy_identifier, "%d\n", u8); +sas_phy_simple_attr(port_identifier, port_identifier, "%d\n", u8); +sas_phy_linkspeed_attr(negotiated_linkrate); +sas_phy_linkspeed_attr(minimum_linkrate_hw); +sas_phy_linkspeed_attr(minimum_linkrate); +sas_phy_linkspeed_attr(maximum_linkrate_hw); +sas_phy_linkspeed_attr(maximum_linkrate); + + +static DECLARE_TRANSPORT_CLASS(sas_phy_class, + "sas_phy", NULL, NULL, NULL); + +static int sas_phy_match(struct attribute_container *cont, struct device *dev) +{ + struct Scsi_Host *shost; + struct sas_internal *i; + + if (!scsi_is_sas_phy(dev)) + return 0; + shost = dev_to_shost(dev->parent); + + if (!shost->transportt) + return 0; + if (shost->transportt->host_attrs.ac.class != + &sas_host_class.class) + return 0; + + i = to_sas_internal(shost->transportt); + return &i->phy_attr_cont.ac == cont; +} + +static void sas_phy_release(struct device *dev) +{ + struct sas_phy *phy = dev_to_phy(dev); + + put_device(dev->parent); + kfree(phy); +} + +/** + * sas_phy_alloc -- allocates and initialize a SAS PHY structure + * @parent: Parent device + * @number: Port number + * + * Allocates an SAS PHY structure. It will be added in the device tree + * below the device specified by @parent, which has to be either a Scsi_Host + * or sas_rphy. + * + * Returns: + * SAS PHY allocated or %NULL if the allocation failed. + */ +struct sas_phy *sas_phy_alloc(struct device *parent, int number) +{ + struct Scsi_Host *shost = dev_to_shost(parent); + struct sas_phy *phy; + + phy = kmalloc(sizeof(*phy), GFP_KERNEL); + if (!phy) + return NULL; + memset(phy, 0, sizeof(*phy)); + + get_device(parent); + + phy->number = number; + + device_initialize(&phy->dev); + phy->dev.parent = get_device(parent); + phy->dev.release = sas_phy_release; + sprintf(phy->dev.bus_id, "phy-%d:%d", shost->host_no, number); + + transport_setup_device(&phy->dev); + + return phy; +} +EXPORT_SYMBOL(sas_phy_alloc); + +/** + * sas_phy_add -- add a SAS PHY to the device hierachy + * @phy: The PHY to be added + * + * Publishes a SAS PHY to the rest of the system. + */ +int sas_phy_add(struct sas_phy *phy) +{ + int error; + + error = device_add(&phy->dev); + if (!error) { + transport_add_device(&phy->dev); + transport_configure_device(&phy->dev); + } + + return error; +} +EXPORT_SYMBOL(sas_phy_add); + +/** + * sas_phy_free -- free a SAS PHY + * @phy: SAS PHY to free + * + * Frees the specified SAS PHY. + * + * Note: + * This function must only be called on a PHY that has not + * sucessfully been added using sas_phy_add(). + */ +void sas_phy_free(struct sas_phy *phy) +{ + transport_destroy_device(&phy->dev); + put_device(phy->dev.parent); + put_device(phy->dev.parent); + put_device(phy->dev.parent); + kfree(phy); +} +EXPORT_SYMBOL(sas_phy_free); + +/** + * sas_phy_delete -- remove SAS PHY + * @phy: SAS PHY to remove + * + * Removes the specified SAS PHY. If the SAS PHY has an + * associated remote PHY it is removed before. + */ +void +sas_phy_delete(struct sas_phy *phy) +{ + struct device *dev = &phy->dev; + + if (phy->rphy) + sas_rphy_delete(phy->rphy); + + transport_remove_device(dev); + device_del(dev); + transport_destroy_device(dev); + put_device(dev->parent); +} +EXPORT_SYMBOL(sas_phy_delete); + +/** + * scsi_is_sas_phy -- check if a struct device represents a SAS PHY + * @dev: device to check + * + * Returns: + * %1 if the device represents a SAS PHY, %0 else + */ +int scsi_is_sas_phy(const struct device *dev) +{ + return dev->release == sas_phy_release; +} +EXPORT_SYMBOL(scsi_is_sas_phy); + +/* + * SAS remote PHY attributes. + */ + +#define sas_rphy_show_simple(field, name, format_string, cast) \ +static ssize_t \ +show_sas_rphy_##name(struct class_device *cdev, char *buf) \ +{ \ + struct sas_rphy *rphy = transport_class_to_rphy(cdev); \ + \ + return snprintf(buf, 20, format_string, cast rphy->field); \ +} + +#define sas_rphy_simple_attr(field, name, format_string, type) \ + sas_rphy_show_simple(field, name, format_string, (type)) \ +static SAS_CLASS_DEVICE_ATTR(rphy, name, S_IRUGO, \ + show_sas_rphy_##name, NULL) + +#define sas_rphy_show_protocol(field, name) \ +static ssize_t \ +show_sas_rphy_##name(struct class_device *cdev, char *buf) \ +{ \ + struct sas_rphy *rphy = transport_class_to_rphy(cdev); \ + \ + if (!rphy->field) \ + return snprintf(buf, 20, "none\n"); \ + return get_sas_protocol_names(rphy->field, buf); \ +} + +#define sas_rphy_protocol_attr(field, name) \ + sas_rphy_show_protocol(field, name) \ +static SAS_CLASS_DEVICE_ATTR(rphy, name, S_IRUGO, \ + show_sas_rphy_##name, NULL) + +static ssize_t +show_sas_rphy_device_type(struct class_device *cdev, char *buf) +{ + struct sas_rphy *rphy = transport_class_to_rphy(cdev); + + if (!rphy->identify.device_type) + return snprintf(buf, 20, "none\n"); + return get_sas_device_type_names( + rphy->identify.device_type, buf); +} + +static SAS_CLASS_DEVICE_ATTR(rphy, device_type, S_IRUGO, + show_sas_rphy_device_type, NULL); + +sas_rphy_protocol_attr(identify.initiator_port_protocols, + initiator_port_protocols); +sas_rphy_protocol_attr(identify.target_port_protocols, target_port_protocols); +sas_rphy_simple_attr(identify.sas_address, sas_address, "0x%016llx\n", + unsigned long long); +sas_rphy_simple_attr(identify.phy_identifier, phy_identifier, "%d\n", u8); + +static DECLARE_TRANSPORT_CLASS(sas_rphy_class, + "sas_rphy", NULL, NULL, NULL); + +static int sas_rphy_match(struct attribute_container *cont, struct device *dev) +{ + struct Scsi_Host *shost; + struct sas_internal *i; + + if (!scsi_is_sas_rphy(dev)) + return 0; + shost = dev_to_shost(dev->parent->parent); + + if (!shost->transportt) + return 0; + if (shost->transportt->host_attrs.ac.class != + &sas_host_class.class) + return 0; + + i = to_sas_internal(shost->transportt); + return &i->rphy_attr_cont.ac == cont; +} + +static void sas_rphy_release(struct device *dev) +{ + struct sas_rphy *rphy = dev_to_rphy(dev); + + put_device(dev->parent); + kfree(rphy); +} + +/** + * sas_rphy_alloc -- allocates and initialize a SAS remote PHY structure + * @parent: SAS PHY this remote PHY is conneted to + * + * Allocates an SAS remote PHY structure, connected to @parent. + * + * Returns: + * SAS PHY allocated or %NULL if the allocation failed. + */ +struct sas_rphy *sas_rphy_alloc(struct sas_phy *parent) +{ + struct Scsi_Host *shost = dev_to_shost(&parent->dev); + struct sas_rphy *rphy; + + rphy = kmalloc(sizeof(*rphy), GFP_KERNEL); + if (!rphy) { + put_device(&parent->dev); + return NULL; + } + memset(rphy, 0, sizeof(*rphy)); + + device_initialize(&rphy->dev); + rphy->dev.parent = get_device(&parent->dev); + rphy->dev.release = sas_rphy_release; + sprintf(rphy->dev.bus_id, "rphy-%d:%d", + shost->host_no, parent->number); + transport_setup_device(&rphy->dev); + + return rphy; +} +EXPORT_SYMBOL(sas_rphy_alloc); + +/** + * sas_rphy_add -- add a SAS remote PHY to the device hierachy + * @rphy: The remote PHY to be added + * + * Publishes a SAS remote PHY to the rest of the system. + */ +int sas_rphy_add(struct sas_rphy *rphy) +{ + struct sas_phy *parent = dev_to_phy(rphy->dev.parent); + struct Scsi_Host *shost = dev_to_shost(parent->dev.parent); + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + struct sas_identify *identify = &rphy->identify; + int error; + + if (parent->rphy) + return -ENXIO; + parent->rphy = rphy; + + error = device_add(&rphy->dev); + if (error) + return error; + transport_add_device(&rphy->dev); + transport_configure_device(&rphy->dev); + + spin_lock(&sas_host->lock); + list_add_tail(&rphy->list, &sas_host->rphy_list); + if (identify->device_type == SAS_END_DEVICE && + (identify->target_port_protocols & + (SAS_PROTOCOL_SSP|SAS_PROTOCOL_STP|SAS_PROTOCOL_SATA))) + rphy->scsi_target_id = sas_host->next_target_id++; + else + rphy->scsi_target_id = -1; + spin_unlock(&sas_host->lock); + + if (rphy->scsi_target_id != -1) { + scsi_scan_target(&rphy->dev, parent->number, + rphy->scsi_target_id, ~0, 0); + } + + return 0; +} +EXPORT_SYMBOL(sas_rphy_add); + +/** + * sas_rphy_free -- free a SAS remote PHY + * @rphy SAS remote PHY to free + * + * Frees the specified SAS remote PHY. + * + * Note: + * This function must only be called on a remote + * PHY that has not sucessfully been added using + * sas_rphy_add(). + */ +void sas_rphy_free(struct sas_rphy *rphy) +{ + struct Scsi_Host *shost = dev_to_shost(rphy->dev.parent->parent); + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + + spin_lock(&sas_host->lock); + list_del(&rphy->list); + spin_unlock(&sas_host->lock); + + transport_destroy_device(&rphy->dev); + put_device(rphy->dev.parent); + put_device(rphy->dev.parent); + put_device(rphy->dev.parent); + kfree(rphy); +} +EXPORT_SYMBOL(sas_rphy_free); + +/** + * sas_rphy_delete -- remove SAS remote PHY + * @rphy: SAS remote PHY to remove + * + * Removes the specified SAS remote PHY. + */ +void +sas_rphy_delete(struct sas_rphy *rphy) +{ + struct device *dev = &rphy->dev; + struct sas_phy *parent = dev_to_phy(dev->parent); + struct Scsi_Host *shost = dev_to_shost(parent->dev.parent); + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + + transport_destroy_device(&rphy->dev); + + scsi_remove_target(&rphy->dev); + + spin_lock(&sas_host->lock); + list_del(&rphy->list); + spin_unlock(&sas_host->lock); + + transport_remove_device(dev); + device_del(dev); + transport_destroy_device(dev); + put_device(&parent->dev); +} +EXPORT_SYMBOL(sas_rphy_delete); + +/** + * scsi_is_sas_rphy -- check if a struct device represents a SAS remote PHY + * @dev: device to check + * + * Returns: + * %1 if the device represents a SAS remote PHY, %0 else + */ +int scsi_is_sas_rphy(const struct device *dev) +{ + return dev->release == sas_rphy_release; +} +EXPORT_SYMBOL(scsi_is_sas_rphy); + + +/* + * SCSI scan helper + */ + +static struct device *sas_target_parent(struct Scsi_Host *shost, + int channel, uint id) +{ + struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); + struct sas_rphy *rphy; + struct device *dev = NULL; + + spin_lock(&sas_host->lock); + list_for_each_entry(rphy, &sas_host->rphy_list, list) { + struct sas_phy *parent = dev_to_phy(rphy->dev.parent); + if (parent->number == channel && + rphy->scsi_target_id == id) + dev = &rphy->dev; + } + spin_unlock(&sas_host->lock); + + return dev; +} + + +/* + * Setup / Teardown code + */ + +#define SETUP_RPORT_ATTRIBUTE(field) \ + i->private_rphy_attrs[count] = class_device_attr_##field; \ + i->private_rphy_attrs[count].attr.mode = S_IRUGO; \ + i->private_rphy_attrs[count].store = NULL; \ + i->rphy_attrs[count] = &i->private_rphy_attrs[count]; \ + count++ + +#define SETUP_PORT_ATTRIBUTE(field) \ + i->private_phy_attrs[count] = class_device_attr_##field; \ + i->private_phy_attrs[count].attr.mode = S_IRUGO; \ + i->private_phy_attrs[count].store = NULL; \ + i->phy_attrs[count] = &i->private_phy_attrs[count]; \ + count++ + + +/** + * sas_attach_transport -- instantiate SAS transport template + * @ft: SAS transport class function template + */ +struct scsi_transport_template * +sas_attach_transport(struct sas_function_template *ft) +{ + struct sas_internal *i; + int count; + + i = kmalloc(sizeof(struct sas_internal), GFP_KERNEL); + if (!i) + return NULL; + memset(i, 0, sizeof(struct sas_internal)); + + i->t.target_parent = sas_target_parent; + + i->t.host_attrs.ac.attrs = &i->host_attrs[0]; + i->t.host_attrs.ac.class = &sas_host_class.class; + i->t.host_attrs.ac.match = sas_host_match; + transport_container_register(&i->t.host_attrs); + i->t.host_size = sizeof(struct sas_host_attrs); + + i->phy_attr_cont.ac.class = &sas_phy_class.class; + i->phy_attr_cont.ac.attrs = &i->phy_attrs[0]; + i->phy_attr_cont.ac.match = sas_phy_match; + transport_container_register(&i->phy_attr_cont); + + i->rphy_attr_cont.ac.class = &sas_rphy_class.class; + i->rphy_attr_cont.ac.attrs = &i->rphy_attrs[0]; + i->rphy_attr_cont.ac.match = sas_rphy_match; + transport_container_register(&i->rphy_attr_cont); + + i->f = ft; + + count = 0; + i->host_attrs[count] = NULL; + + count = 0; + SETUP_PORT_ATTRIBUTE(initiator_port_protocols); + SETUP_PORT_ATTRIBUTE(target_port_protocols); + SETUP_PORT_ATTRIBUTE(device_type); + SETUP_PORT_ATTRIBUTE(sas_address); + SETUP_PORT_ATTRIBUTE(phy_identifier); + SETUP_PORT_ATTRIBUTE(port_identifier); + SETUP_PORT_ATTRIBUTE(negotiated_linkrate); + SETUP_PORT_ATTRIBUTE(minimum_linkrate_hw); + SETUP_PORT_ATTRIBUTE(minimum_linkrate); + SETUP_PORT_ATTRIBUTE(maximum_linkrate_hw); + SETUP_PORT_ATTRIBUTE(maximum_linkrate); + i->phy_attrs[count] = NULL; + + count = 0; + SETUP_RPORT_ATTRIBUTE(rphy_initiator_port_protocols); + SETUP_RPORT_ATTRIBUTE(rphy_target_port_protocols); + SETUP_RPORT_ATTRIBUTE(rphy_device_type); + SETUP_RPORT_ATTRIBUTE(rphy_sas_address); + SETUP_RPORT_ATTRIBUTE(rphy_phy_identifier); + i->rphy_attrs[count] = NULL; + + return &i->t; +} +EXPORT_SYMBOL(sas_attach_transport); + +/** + * sas_release_transport -- release SAS transport template instance + * @t: transport template instance + */ +void sas_release_transport(struct scsi_transport_template *t) +{ + struct sas_internal *i = to_sas_internal(t); + + transport_container_unregister(&i->t.host_attrs); + transport_container_unregister(&i->phy_attr_cont); + transport_container_unregister(&i->rphy_attr_cont); + + kfree(i); +} +EXPORT_SYMBOL(sas_release_transport); + +static __init int sas_transport_init(void) +{ + int error; + + error = transport_class_register(&sas_host_class); + if (error) + goto out; + error = transport_class_register(&sas_phy_class); + if (error) + goto out_unregister_transport; + error = transport_class_register(&sas_rphy_class); + if (error) + goto out_unregister_phy; + + return 0; + + out_unregister_phy: + transport_class_unregister(&sas_phy_class); + out_unregister_transport: + transport_class_unregister(&sas_host_class); + out: + return error; + +} + +static void __exit sas_transport_exit(void) +{ + transport_class_unregister(&sas_host_class); + transport_class_unregister(&sas_phy_class); + transport_class_unregister(&sas_rphy_class); +} + +MODULE_AUTHOR("Christoph Hellwig"); +MODULE_DESCRIPTION("SAS Transphy Attributes"); +MODULE_LICENSE("GPL"); + +module_init(sas_transport_init); +module_exit(sas_transport_exit); diff --git a/include/scsi/scsi_transport_sas.h b/include/scsi/scsi_transport_sas.h new file mode 100644 index 0000000..bc4aeb6 --- /dev/null +++ b/include/scsi/scsi_transport_sas.h @@ -0,0 +1,100 @@ +#ifndef SCSI_TRANSPORT_SAS_H +#define SCSI_TRANSPORT_SAS_H + +#include +#include + +struct scsi_transport_template; +struct sas_rphy; + + +enum sas_device_type { + SAS_PHY_UNUSED, + SAS_END_DEVICE, + SAS_EDGE_EXPANDER_DEVICE, + SAS_FANOUT_EXPANDER_DEVICE, +}; + +enum sas_protocol { + SAS_PROTOCOL_SATA = 0x01, + SAS_PROTOCOL_SMP = 0x02, + SAS_PROTOCOL_STP = 0x04, + SAS_PROTOCOL_SSP = 0x08, +}; + +enum sas_linkrate { + SAS_LINK_RATE_UNKNOWN, + SAS_PHY_DISABLED, + SAS_LINK_RATE_FAILED, + SAS_SATA_SPINUP_HOLD, + SAS_SATA_PORT_SELECTOR, + SAS_LINK_RATE_1_5_GBPS, + SAS_LINK_RATE_3_0_GBPS, + SAS_LINK_VIRTUAL, +}; + +struct sas_identify { + enum sas_device_type device_type; + enum sas_protocol initiator_port_protocols; + enum sas_protocol target_port_protocols; + u64 sas_address; + u8 phy_identifier; +}; + +/* The functions by which the transport class and the driver communicate */ +struct sas_function_template { +}; + +struct sas_phy { + struct device dev; + int number; + struct sas_identify identify; + enum sas_linkrate negotiated_linkrate; + enum sas_linkrate minimum_linkrate_hw; + enum sas_linkrate minimum_linkrate; + enum sas_linkrate maximum_linkrate_hw; + enum sas_linkrate maximum_linkrate; + u8 port_identifier; + struct sas_rphy *rphy; +}; + +#define dev_to_phy(d) \ + container_of((d), struct sas_phy, dev) +#define transport_class_to_phy(cdev) \ + dev_to_phy((cdev)->dev) +#define phy_to_shost(phy) \ + dev_to_shost((phy)->dev.parent) + +struct sas_rphy { + struct device dev; + struct sas_identify identify; + struct list_head list; + u32 scsi_target_id; +}; + +#define dev_to_rphy(d) \ + container_of((d), struct sas_rphy, dev) +#define transport_class_to_rphy(cdev) \ + dev_to_rphy((cdev)->dev) +#define rphy_to_shost(rphy) \ + dev_to_shost((rphy)->dev.parent) + +extern void sas_remove_host(struct Scsi_Host *); + +extern struct sas_phy *sas_phy_alloc(struct device *, int); +extern void sas_phy_free(struct sas_phy *); +extern int sas_phy_add(struct sas_phy *); +extern void sas_phy_delete(struct sas_phy *); +extern int scsi_is_sas_phy(const struct device *); + +extern struct sas_rphy *sas_rphy_alloc(struct sas_phy *); +void sas_rphy_free(struct sas_rphy *); +extern int sas_rphy_add(struct sas_rphy *); +extern void sas_rphy_delete(struct sas_rphy *); +extern int scsi_is_sas_rphy(const struct device *); + +extern struct scsi_transport_template * +sas_attach_transport(struct sas_function_template *); +extern void sas_release_transport(struct scsi_transport_template *); + +#endif /* SCSI_TRANSPORT_SAS_H */ -- cgit v1.1 From d327d082325a0d4afb3748ef8b59e734e57cfe4c Mon Sep 17 00:00:00 2001 From: adam radford Date: Fri, 9 Sep 2005 15:55:13 -0700 Subject: [SCSI] 3ware 9000: handle use_sg != 0 for emulated commands The attached patch updates the driver for the 3ware 9000 series to do the following: - Correctly handle single sgl's with use_sg = 1. This is needed with the latest scsi-block-2.6 merge otherwise the 3w-9xxx driver will not work. I tested the patch James sent a few weeks back to fix this, and it had a bug where the request_buffer was accessed in twa_scsiop_execute_scsi_complete() when it was invalid. This is a corrected variation of that patch. Signed-off-by: Adam Radford Signed-off-by: James Bottomley --- drivers/scsi/3w-9xxx.c | 30 ++++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c index bc6e462..a6ac616 100644 --- a/drivers/scsi/3w-9xxx.c +++ b/drivers/scsi/3w-9xxx.c @@ -59,6 +59,7 @@ Fix 'handled=1' ISR usage, remove bogus IRQ check. Remove un-needed eh_abort handler. Add support for embedded firmware error strings. + 2.26.02.003 - Correctly handle single sgl's with use_sg=1. */ #include @@ -81,7 +82,7 @@ #include "3w-9xxx.h" /* Globals */ -#define TW_DRIVER_VERSION "2.26.02.002" +#define TW_DRIVER_VERSION "2.26.02.003" static TW_Device_Extension *twa_device_extension_list[TW_MAX_SLOT]; static unsigned int twa_device_extension_count; static int twa_major = -1; @@ -1805,6 +1806,8 @@ static int twa_scsiop_execute_scsi(TW_Device_Extension *tw_dev, int request_id, if (tw_dev->srb[request_id]->request_bufflen < TW_MIN_SGL_LENGTH) { command_packet->sg_list[0].address = tw_dev->generic_buffer_phys[request_id]; command_packet->sg_list[0].length = TW_MIN_SGL_LENGTH; + if (tw_dev->srb[request_id]->sc_data_direction == DMA_TO_DEVICE || tw_dev->srb[request_id]->sc_data_direction == DMA_BIDIRECTIONAL) + memcpy(tw_dev->generic_buffer_virt[request_id], tw_dev->srb[request_id]->request_buffer, tw_dev->srb[request_id]->request_bufflen); } else { buffaddr = twa_map_scsi_single_data(tw_dev, request_id); if (buffaddr == 0) @@ -1823,6 +1826,12 @@ static int twa_scsiop_execute_scsi(TW_Device_Extension *tw_dev, int request_id, if (tw_dev->srb[request_id]->use_sg > 0) { if ((tw_dev->srb[request_id]->use_sg == 1) && (tw_dev->srb[request_id]->request_bufflen < TW_MIN_SGL_LENGTH)) { + if (tw_dev->srb[request_id]->sc_data_direction == DMA_TO_DEVICE || tw_dev->srb[request_id]->sc_data_direction == DMA_BIDIRECTIONAL) { + struct scatterlist *sg = (struct scatterlist *)tw_dev->srb[request_id]->request_buffer; + char *buf = kmap_atomic(sg->page, KM_IRQ0) + sg->offset; + memcpy(tw_dev->generic_buffer_virt[request_id], buf, sg->length); + kunmap_atomic(buf - sg->offset, KM_IRQ0); + } command_packet->sg_list[0].address = tw_dev->generic_buffer_phys[request_id]; command_packet->sg_list[0].length = TW_MIN_SGL_LENGTH; } else { @@ -1888,11 +1897,20 @@ out: /* This function completes an execute scsi operation */ static void twa_scsiop_execute_scsi_complete(TW_Device_Extension *tw_dev, int request_id) { - /* Copy the response if too small */ - if ((tw_dev->srb[request_id]->request_buffer) && (tw_dev->srb[request_id]->request_bufflen < TW_MIN_SGL_LENGTH)) { - memcpy(tw_dev->srb[request_id]->request_buffer, - tw_dev->generic_buffer_virt[request_id], - tw_dev->srb[request_id]->request_bufflen); + if (tw_dev->srb[request_id]->request_bufflen < TW_MIN_SGL_LENGTH && + (tw_dev->srb[request_id]->sc_data_direction == DMA_FROM_DEVICE || + tw_dev->srb[request_id]->sc_data_direction == DMA_BIDIRECTIONAL)) { + if (tw_dev->srb[request_id]->use_sg == 0) { + memcpy(tw_dev->srb[request_id]->request_buffer, + tw_dev->generic_buffer_virt[request_id], + tw_dev->srb[request_id]->request_bufflen); + } + if (tw_dev->srb[request_id]->use_sg == 1) { + struct scatterlist *sg = (struct scatterlist *)tw_dev->srb[request_id]->request_buffer; + char *buf = kmap_atomic(sg->page, KM_IRQ0) + sg->offset; + memcpy(buf, tw_dev->generic_buffer_virt[request_id], sg->length); + kunmap_atomic(buf - sg->offset, KM_IRQ0); + } } } /* End twa_scsiop_execute_scsi_complete() */ -- cgit v1.1 From 37be6eeb4990c05fc7dd683ceaf1501d46ebe9a4 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Fri, 9 Sep 2005 18:38:27 -0500 Subject: [SCSI] SAS transport class: fixup prototype of sas_host_setup Signed-off-by: James Bottomley --- drivers/scsi/scsi_transport_sas.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/scsi_transport_sas.c b/drivers/scsi/scsi_transport_sas.c index ac4a53a..ff724bbe 100644 --- a/drivers/scsi/scsi_transport_sas.c +++ b/drivers/scsi/scsi_transport_sas.c @@ -156,7 +156,8 @@ sas_bitfield_name_search(linkspeed, sas_linkspeed_names) * SAS host attributes */ -static int sas_host_setup(struct device *dev) +static int sas_host_setup(struct transport_container *tc, struct device *dev, + struct class_device *cdev) { struct Scsi_Host *shost = dev_to_shost(dev); struct sas_host_attrs *sas_host = to_sas_host_attrs(shost); -- cgit v1.1 From 218fba0004390b0101e681f6db1b8920c1109e54 Mon Sep 17 00:00:00 2001 From: Andrew Vasquez Date: Wed, 31 Aug 2005 15:18:35 -0700 Subject: [SCSI] fc_transport: Generalize WWN to u64 interger conversions. On some platforms the hard-casting of 8 byte node_name and port_name arrays to an u64 would cause unaligned-access warnings. Generalize the conversions with a transport helper function which performs consistent shifting of WWN bytes. Signed-off-by: Andrew Vasquez Signed-off-by: James Bottomley --- include/scsi/scsi_transport_fc.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h index 70ad163..115db05 100644 --- a/include/scsi/scsi_transport_fc.h +++ b/include/scsi/scsi_transport_fc.h @@ -439,4 +439,12 @@ int fc_remote_port_block(struct fc_rport *rport); void fc_remote_port_unblock(struct fc_rport *rport); int scsi_is_fc_rport(const struct device *); +static inline u64 wwn_to_u64(u8 *wwn) +{ + return (u64)wwn[0] << 56 | (u64)wwn[1] << 48 | + (u64)wwn[2] << 40 | (u64)wwn[3] << 32 | + (u64)wwn[4] << 24 | (u64)wwn[5] << 16 | + (u64)wwn[6] << 8 | (u64)wwn[7]; +} + #endif /* SCSI_TRANSPORT_FC_H */ -- cgit v1.1 From f8b02a85ebbf5eed63163ca9ed915bf8c47309c5 Mon Sep 17 00:00:00 2001 From: Andrew Vasquez Date: Wed, 31 Aug 2005 15:21:20 -0700 Subject: [SCSI] qla2xxx: use wwn_to_u64() transport helper Signed-off-by: Andrew Vasquez Signed-off-by: James Bottomley --- drivers/scsi/qla2xxx/qla_attr.c | 18 ++++++++---------- drivers/scsi/qla2xxx/qla_init.c | 4 ++-- 2 files changed, 10 insertions(+), 12 deletions(-) diff --git a/drivers/scsi/qla2xxx/qla_attr.c b/drivers/scsi/qla2xxx/qla_attr.c index fe0fce7..fc25cd8 100644 --- a/drivers/scsi/qla2xxx/qla_attr.c +++ b/drivers/scsi/qla2xxx/qla_attr.c @@ -360,16 +360,16 @@ qla2x00_get_starget_node_name(struct scsi_target *starget) struct Scsi_Host *host = dev_to_shost(starget->dev.parent); scsi_qla_host_t *ha = to_qla_host(host); fc_port_t *fcport; - uint64_t node_name = 0; + u64 node_name = 0; list_for_each_entry(fcport, &ha->fcports, list) { if (starget->id == fcport->os_target_id) { - node_name = *(uint64_t *)fcport->node_name; + node_name = wwn_to_u64(fcport->node_name); break; } } - fc_starget_node_name(starget) = be64_to_cpu(node_name); + fc_starget_node_name(starget) = node_name; } static void @@ -378,16 +378,16 @@ qla2x00_get_starget_port_name(struct scsi_target *starget) struct Scsi_Host *host = dev_to_shost(starget->dev.parent); scsi_qla_host_t *ha = to_qla_host(host); fc_port_t *fcport; - uint64_t port_name = 0; + u64 port_name = 0; list_for_each_entry(fcport, &ha->fcports, list) { if (starget->id == fcport->os_target_id) { - port_name = *(uint64_t *)fcport->port_name; + port_name = wwn_to_u64(fcport->port_name); break; } } - fc_starget_port_name(starget) = be64_to_cpu(port_name); + fc_starget_port_name(starget) = port_name; } static void @@ -460,9 +460,7 @@ struct fc_function_template qla2xxx_transport_functions = { void qla2x00_init_host_attr(scsi_qla_host_t *ha) { - fc_host_node_name(ha->host) = - be64_to_cpu(*(uint64_t *)ha->init_cb->node_name); - fc_host_port_name(ha->host) = - be64_to_cpu(*(uint64_t *)ha->init_cb->port_name); + fc_host_node_name(ha->host) = wwn_to_u64(ha->init_cb->node_name); + fc_host_port_name(ha->host) = wwn_to_u64(ha->init_cb->port_name); fc_host_supported_classes(ha->host) = FC_COS_CLASS3; } diff --git a/drivers/scsi/qla2xxx/qla_init.c b/drivers/scsi/qla2xxx/qla_init.c index c619583..3e9b641 100644 --- a/drivers/scsi/qla2xxx/qla_init.c +++ b/drivers/scsi/qla2xxx/qla_init.c @@ -2066,8 +2066,8 @@ qla2x00_reg_remote_port(scsi_qla_host_t *ha, fc_port_t *fcport) return; } - rport_ids.node_name = be64_to_cpu(*(uint64_t *)fcport->node_name); - rport_ids.port_name = be64_to_cpu(*(uint64_t *)fcport->port_name); + rport_ids.node_name = wwn_to_u64(fcport->node_name); + rport_ids.port_name = wwn_to_u64(fcport->port_name); rport_ids.port_id = fcport->d_id.b.domain << 16 | fcport->d_id.b.area << 8 | fcport->d_id.b.al_pa; rport_ids.roles = FC_RPORT_ROLE_UNKNOWN; -- cgit v1.1 From f631b4be76355dc3bf49563c706a9fb938993bde Mon Sep 17 00:00:00 2001 From: Andrew Vasquez Date: Wed, 31 Aug 2005 15:23:12 -0700 Subject: [SCSI] lpfc: use wwn_to_u64() transport helper Signed-off-by: Andrew Vasquez Acked-by: Smart, James Signed-off-by: James Bottomley --- drivers/scsi/lpfc/lpfc_attr.c | 22 ++++++++++------------ drivers/scsi/lpfc/lpfc_hbadisc.c | 7 ++----- drivers/scsi/lpfc/lpfc_hw.h | 17 +++++++++++------ drivers/scsi/lpfc/lpfc_init.c | 7 ++----- 4 files changed, 25 insertions(+), 28 deletions(-) diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c index 0e089a4..86eaf6d 100644 --- a/drivers/scsi/lpfc/lpfc_attr.c +++ b/drivers/scsi/lpfc/lpfc_attr.c @@ -966,21 +966,21 @@ static void lpfc_get_host_fabric_name (struct Scsi_Host *shost) { struct lpfc_hba *phba = (struct lpfc_hba*)shost->hostdata[0]; - u64 nodename; + u64 node_name; spin_lock_irq(shost->host_lock); if ((phba->fc_flag & FC_FABRIC) || ((phba->fc_topology == TOPOLOGY_LOOP) && (phba->fc_flag & FC_PUBLIC_LOOP))) - memcpy(&nodename, &phba->fc_fabparam.nodeName, sizeof(u64)); + node_name = wwn_to_u64(phba->fc_fabparam.nodeName.wwn); else /* fabric is local port if there is no F/FL_Port */ - memcpy(&nodename, &phba->fc_nodename, sizeof(u64)); + node_name = wwn_to_u64(phba->fc_nodename.wwn); spin_unlock_irq(shost->host_lock); - fc_host_fabric_name(shost) = be64_to_cpu(nodename); + fc_host_fabric_name(shost) = node_name; } @@ -1103,21 +1103,20 @@ lpfc_get_starget_node_name(struct scsi_target *starget) { struct Scsi_Host *shost = dev_to_shost(starget->dev.parent); struct lpfc_hba *phba = (struct lpfc_hba *) shost->hostdata[0]; - uint64_t node_name = 0; + u64 node_name = 0; struct lpfc_nodelist *ndlp = NULL; spin_lock_irq(shost->host_lock); /* Search the mapped list for this target ID */ list_for_each_entry(ndlp, &phba->fc_nlpmap_list, nlp_listp) { if (starget->id == ndlp->nlp_sid) { - memcpy(&node_name, &ndlp->nlp_nodename, - sizeof(struct lpfc_name)); + node_name = wwn_to_u64(ndlp->nlp_nodename.wwn); break; } } spin_unlock_irq(shost->host_lock); - fc_starget_node_name(starget) = be64_to_cpu(node_name); + fc_starget_node_name(starget) = node_name; } static void @@ -1125,21 +1124,20 @@ lpfc_get_starget_port_name(struct scsi_target *starget) { struct Scsi_Host *shost = dev_to_shost(starget->dev.parent); struct lpfc_hba *phba = (struct lpfc_hba *) shost->hostdata[0]; - uint64_t port_name = 0; + u64 port_name = 0; struct lpfc_nodelist *ndlp = NULL; spin_lock_irq(shost->host_lock); /* Search the mapped list for this target ID */ list_for_each_entry(ndlp, &phba->fc_nlpmap_list, nlp_listp) { if (starget->id == ndlp->nlp_sid) { - memcpy(&port_name, &ndlp->nlp_portname, - sizeof(struct lpfc_name)); + port_name = wwn_to_u64(ndlp->nlp_portname.wwn); break; } } spin_unlock_irq(shost->host_lock); - fc_starget_port_name(starget) = be64_to_cpu(port_name); + fc_starget_port_name(starget) = port_name; } static void diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c index 0a8269d..4fb8eb0 100644 --- a/drivers/scsi/lpfc/lpfc_hbadisc.c +++ b/drivers/scsi/lpfc/lpfc_hbadisc.c @@ -1017,13 +1017,10 @@ lpfc_register_remote_port(struct lpfc_hba * phba, struct fc_rport *rport; struct lpfc_rport_data *rdata; struct fc_rport_identifiers rport_ids; - uint64_t wwn; /* Remote port has reappeared. Re-register w/ FC transport */ - memcpy(&wwn, &ndlp->nlp_nodename, sizeof(uint64_t)); - rport_ids.node_name = be64_to_cpu(wwn); - memcpy(&wwn, &ndlp->nlp_portname, sizeof(uint64_t)); - rport_ids.port_name = be64_to_cpu(wwn); + rport_ids.node_name = wwn_to_u64(ndlp->nlp_nodename.wwn); + rport_ids.port_name = wwn_to_u64(ndlp->nlp_portname.wwn); rport_ids.port_id = ndlp->nlp_DID; rport_ids.roles = FC_RPORT_ROLE_UNKNOWN; if (ndlp->nlp_type & NLP_FCP_TARGET) diff --git a/drivers/scsi/lpfc/lpfc_hw.h b/drivers/scsi/lpfc/lpfc_hw.h index 21591cb..047a87c 100644 --- a/drivers/scsi/lpfc/lpfc_hw.h +++ b/drivers/scsi/lpfc/lpfc_hw.h @@ -262,12 +262,14 @@ struct lpfc_sli_ct_request { #define FF_FRAME_SIZE 2048 struct lpfc_name { + union { + struct { #ifdef __BIG_ENDIAN_BITFIELD - uint8_t nameType:4; /* FC Word 0, bit 28:31 */ - uint8_t IEEEextMsn:4; /* FC Word 0, bit 24:27, bit 8:11 of IEEE ext */ + uint8_t nameType:4; /* FC Word 0, bit 28:31 */ + uint8_t IEEEextMsn:4; /* FC Word 0, bit 24:27, bit 8:11 of IEEE ext */ #else /* __LITTLE_ENDIAN_BITFIELD */ - uint8_t IEEEextMsn:4; /* FC Word 0, bit 24:27, bit 8:11 of IEEE ext */ - uint8_t nameType:4; /* FC Word 0, bit 28:31 */ + uint8_t IEEEextMsn:4; /* FC Word 0, bit 24:27, bit 8:11 of IEEE ext */ + uint8_t nameType:4; /* FC Word 0, bit 28:31 */ #endif #define NAME_IEEE 0x1 /* IEEE name - nameType */ @@ -276,8 +278,11 @@ struct lpfc_name { #define NAME_IP_TYPE 0x4 /* IP address */ #define NAME_CCITT_TYPE 0xC #define NAME_CCITT_GR_TYPE 0xE - uint8_t IEEEextLsb; /* FC Word 0, bit 16:23, IEEE extended Lsb */ - uint8_t IEEE[6]; /* FC IEEE address */ + uint8_t IEEEextLsb; /* FC Word 0, bit 16:23, IEEE extended Lsb */ + uint8_t IEEE[6]; /* FC IEEE address */ + }; + uint8_t wwn[8]; + }; }; struct csp { diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c index 6f3cb59..454058f 100644 --- a/drivers/scsi/lpfc/lpfc_init.c +++ b/drivers/scsi/lpfc/lpfc_init.c @@ -1333,7 +1333,6 @@ lpfc_pci_probe_one(struct pci_dev *pdev, const struct pci_device_id *pid) unsigned long bar0map_len, bar2map_len; int error = -ENODEV, retval; int i; - u64 wwname; if (pci_enable_device(pdev)) goto out; @@ -1524,10 +1523,8 @@ lpfc_pci_probe_one(struct pci_dev *pdev, const struct pci_device_id *pid) * Must done after lpfc_sli_hba_setup() */ - memcpy(&wwname, &phba->fc_nodename, sizeof(u64)); - fc_host_node_name(host) = be64_to_cpu(wwname); - memcpy(&wwname, &phba->fc_portname, sizeof(u64)); - fc_host_port_name(host) = be64_to_cpu(wwname); + fc_host_node_name(host) = wwn_to_u64(phba->fc_nodename.wwn); + fc_host_port_name(host) = wwn_to_u64(phba->fc_portname.wwn); fc_host_supported_classes(host) = FC_COS_CLASS3; memset(fc_host_supported_fc4s(host), 0, -- cgit v1.1 From b70d37bf61f278f9d9adf17c52af6b2d0ae7800c Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 26 Jul 2005 10:30:40 -0400 Subject: [SCSI] Fix module removal/device add race This patch (as546) fixes an oops-causing failure to check the return code from scsi_device_get. The call can return an error if the LLD is being unloaded from memory. Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_scan.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index 76577fa..a0975c7 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -870,8 +870,12 @@ static int scsi_probe_and_add_lun(struct scsi_target *starget, out_free_sdev: if (res == SCSI_SCAN_LUN_PRESENT) { if (sdevp) { - scsi_device_get(sdev); - *sdevp = sdev; + if (scsi_device_get(sdev) == 0) { + *sdevp = sdev; + } else { + __scsi_remove_device(sdev); + res = SCSI_SCAN_NO_RESPONSE; + } } } else { if (sdev->host->hostt->slave_destroy) -- cgit v1.1 From 146f7262ee0ec7fc6882f06e5fcb13883308073c Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Sat, 10 Sep 2005 12:44:09 -0500 Subject: [SCSI] Alter the scsi_add_device() API to conform to what users expect The original API returned either an ERR_PTR() or a refcounted sdev. Unfortunately, if it's successful, you need to do a scsi_device_put() on the sdev otherwise the refcounting is wrong. Everyone seems to expect that scsi_add_device() should be callable without doing the ref put, so alter the API so it is (we still have __scsi_add_device with the original behaviour). The only actual caller that needs altering is the one in firewire ... not because it gets this right, but because it acts on the error if one is returned. Acked-by: Stefan Richter Signed-off-by: James Bottomley --- drivers/ieee1394/sbp2.c | 8 ++++---- drivers/scsi/scsi_scan.c | 13 +++++++++++++ include/scsi/scsi_device.h | 4 ++-- 3 files changed, 19 insertions(+), 6 deletions(-) diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c index 627af50..de88218 100644 --- a/drivers/ieee1394/sbp2.c +++ b/drivers/ieee1394/sbp2.c @@ -790,7 +790,7 @@ static void sbp2_host_reset(struct hpsb_host *host) static int sbp2_start_device(struct scsi_id_instance_data *scsi_id) { struct sbp2scsi_host_info *hi = scsi_id->hi; - struct scsi_device *sdev; + int error; SBP2_DEBUG("sbp2_start_device"); @@ -939,10 +939,10 @@ alloc_fail: sbp2_max_speed_and_size(scsi_id); /* Add this device to the scsi layer now */ - sdev = scsi_add_device(scsi_id->scsi_host, 0, scsi_id->ud->id, 0); - if (IS_ERR(sdev)) { + error = scsi_add_device(scsi_id->scsi_host, 0, scsi_id->ud->id, 0); + if (error) { SBP2_ERR("scsi_add_device failed"); - return PTR_ERR(sdev); + return error; } return 0; diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index a0975c7..b86f170 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -1264,6 +1264,19 @@ struct scsi_device *__scsi_add_device(struct Scsi_Host *shost, uint channel, } EXPORT_SYMBOL(__scsi_add_device); +int scsi_add_device(struct Scsi_Host *host, uint channel, + uint target, uint lun) +{ + struct scsi_device *sdev = + __scsi_add_device(host, channel, target, lun, NULL); + if (IS_ERR(sdev)) + return PTR_ERR(sdev); + + scsi_device_put(sdev); + return 0; +} +EXPORT_SYMBOL(scsi_add_device); + void scsi_rescan_device(struct device *dev) { struct scsi_driver *drv; diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h index da63722..c0e4c67 100644 --- a/include/scsi/scsi_device.h +++ b/include/scsi/scsi_device.h @@ -178,8 +178,8 @@ static inline struct scsi_target *scsi_target(struct scsi_device *sdev) extern struct scsi_device *__scsi_add_device(struct Scsi_Host *, uint, uint, uint, void *hostdata); -#define scsi_add_device(host, channel, target, lun) \ - __scsi_add_device(host, channel, target, lun, NULL) +extern int scsi_add_device(struct Scsi_Host *host, uint channel, + uint target, uint lun); extern void scsi_remove_device(struct scsi_device *); extern int scsi_device_cancel(struct scsi_device *, int); -- cgit v1.1 From 6f16b5359ceb96780eac4178393b0e8a3c8aa1ea Mon Sep 17 00:00:00 2001 From: Mike Christie Date: Sat, 10 Sep 2005 16:45:35 -0500 Subject: [SCSI] set error value when failing commands in prep_fn set DID_NO_CONNECT for the BLKPREP_KILL case and correct a few BLKPREP_DEFER cases that weren't checking for the need to plug the queue. Signed-Off-By: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index d8d9848..863bb64 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1146,7 +1146,7 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) if (unlikely(!scsi_device_online(sdev))) { printk(KERN_ERR "scsi%d (%d:%d): rejecting I/O to offline device\n", sdev->host->host_no, sdev->id, sdev->lun); - return BLKPREP_KILL; + goto kill; } if (unlikely(sdev->sdev_state != SDEV_RUNNING)) { /* OK, we're not in a running state don't prep @@ -1156,7 +1156,7 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) * at all allowed down */ printk(KERN_ERR "scsi%d (%d:%d): rejecting I/O to dead device\n", sdev->host->host_no, sdev->id, sdev->lun); - return BLKPREP_KILL; + goto kill; } /* OK, we only allow special commands (i.e. not * user initiated ones */ @@ -1188,11 +1188,11 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) if(unlikely(specials_only) && !(req->flags & REQ_SPECIAL)) { if(specials_only == SDEV_QUIESCE || specials_only == SDEV_BLOCK) - return BLKPREP_DEFER; + goto defer; printk(KERN_ERR "scsi%d (%d:%d): rejecting I/O to device being removed\n", sdev->host->host_no, sdev->id, sdev->lun); - return BLKPREP_KILL; + goto kill; } @@ -1210,7 +1210,7 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) cmd->tag = req->tag; } else { blk_dump_rq_flags(req, "SCSI bad req"); - return BLKPREP_KILL; + goto kill; } /* note the overloading of req->special. When the tag @@ -1248,8 +1248,13 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) * required). */ ret = scsi_init_io(cmd); - if (ret) /* BLKPREP_KILL return also releases the command */ - return ret; + switch(ret) { + case BLKPREP_KILL: + /* BLKPREP_KILL return also releases the command */ + goto kill; + case BLKPREP_DEFER: + goto defer; + } /* * Initialize the actual SCSI command for this request. @@ -1259,7 +1264,7 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) if (unlikely(!drv->init_command(cmd))) { scsi_release_buffers(cmd); scsi_put_command(cmd); - return BLKPREP_KILL; + goto kill; } } else { memcpy(cmd->cmnd, req->cmd, sizeof(cmd->cmnd)); @@ -1290,6 +1295,9 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) if (sdev->device_busy == 0) blk_plug_device(q); return BLKPREP_DEFER; + kill: + req->errors = DID_NO_CONNECT << 16; + return BLKPREP_KILL; } /* -- cgit v1.1 From 70c83e110541e894fa16aee0f57bcf4207b33e7b Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sun, 11 Sep 2005 09:37:19 +0900 Subject: [SCSI] scsi: Error handler description document This patch adds Documentation/scsi/scs_eh.txt. I've chosen plain text over DocBook as most other scsi docs are in plain text and it's more accessible. Signed-off-by: Tejun Heo Signed-off-by: James Bottomley --- Documentation/scsi/00-INDEX | 2 + Documentation/scsi/scsi_eh.txt | 479 +++++++++++++++++++++++++++++++++++++++++ 2 files changed, 481 insertions(+) create mode 100644 Documentation/scsi/scsi_eh.txt diff --git a/Documentation/scsi/00-INDEX b/Documentation/scsi/00-INDEX index f9cb5bd..fef92eb 100644 --- a/Documentation/scsi/00-INDEX +++ b/Documentation/scsi/00-INDEX @@ -60,6 +60,8 @@ scsi.txt - short blurb on using SCSI support as a module. scsi_mid_low_api.txt - info on API between SCSI layer and low level drivers +scsi_eh.txt + - info on SCSI midlayer error handling infrastructure st.txt - info on scsi tape driver sym53c500_cs.txt diff --git a/Documentation/scsi/scsi_eh.txt b/Documentation/scsi/scsi_eh.txt new file mode 100644 index 0000000..534a509 --- /dev/null +++ b/Documentation/scsi/scsi_eh.txt @@ -0,0 +1,479 @@ + +SCSI EH +====================================== + + This document describes SCSI midlayer error handling infrastructure. +Please refer to Documentation/scsi/scsi_mid_low_api.txt for more +information regarding SCSI midlayer. + +TABLE OF CONTENTS + +[1] How SCSI commands travel through the midlayer and to EH + [1-1] struct scsi_cmnd + [1-2] How do scmd's get completed? + [1-2-1] Completing a scmd w/ scsi_done + [1-2-2] Completing a scmd w/ timeout + [1-3] How EH takes over +[2] How SCSI EH works + [2-1] EH through fine-grained callbacks + [2-1-1] Overview + [2-1-2] Flow of scmds through EH + [2-1-3] Flow of control + [2-2] EH through hostt->eh_strategy_handler() + [2-2-1] Pre hostt->eh_strategy_handler() SCSI midlayer conditions + [2-2-2] Post hostt->eh_strategy_handler() SCSI midlayer conditions + [2-2-3] Things to consider + + +[1] How SCSI commands travel through the midlayer and to EH + +[1-1] struct scsi_cmnd + + Each SCSI command is represented with struct scsi_cmnd (== scmd). A +scmd has two list_head's to link itself into lists. The two are +scmd->list and scmd->eh_entry. The former is used for free list or +per-device allocated scmd list and not of much interest to this EH +discussion. The latter is used for completion and EH lists and unless +otherwise stated scmds are always linked using scmd->eh_entry in this +discussion. + + +[1-2] How do scmd's get completed? + + Once LLDD gets hold of a scmd, either the LLDD will complete the +command by calling scsi_done callback passed from midlayer when +invoking hostt->queuecommand() or SCSI midlayer will time it out. + + +[1-2-1] Completing a scmd w/ scsi_done + + For all non-EH commands, scsi_done() is the completion callback. It +does the following. + + 1. Delete timeout timer. If it fails, it means that timeout timer + has expired and is going to finish the command. Just return. + + 2. Link scmd to per-cpu scsi_done_q using scmd->en_entry + + 3. Raise SCSI_SOFTIRQ + + SCSI_SOFTIRQ handler scsi_softirq calls scsi_decide_disposition() to +determine what to do with the command. scsi_decide_disposition() +looks at the scmd->result value and sense data to determine what to do +with the command. + + - SUCCESS + scsi_finish_command() is invoked for the command. The + function does some maintenance choirs and notify completion by + calling scmd->done() callback, which, for fs requests, would + be HLD completion callback - sd:sd_rw_intr, sr:rw_intr, + st:st_intr. + + - NEEDS_RETRY + - ADD_TO_MLQUEUE + scmd is requeued to blk queue. + + - otherwise + scsi_eh_scmd_add(scmd, 0) is invoked for the command. See + [1-3] for details of this funciton. + + +[1-2-2] Completing a scmd w/ timeout + + The timeout handler is scsi_times_out(). When a timeout occurs, this +function + + 1. invokes optional hostt->eh_timedout() callback. Return value can + be one of + + - EH_HANDLED + This indicates that eh_timedout() dealt with the timeout. The + scmd is passed to __scsi_done() and thus linked into per-cpu + scsi_done_q. Normal command completion described in [1-2-1] + follows. + + - EH_RESET_TIMER + This indicates that more time is required to finish the + command. Timer is restarted. This action is counted as a + retry and only allowed scmd->allowed + 1(!) times. Once the + limit is reached, action for EH_NOT_HANDLED is taken instead. + + *NOTE* This action is racy as the LLDD could finish the scmd + after the timeout has expired but before it's added back. In + such cases, scsi_done() would think that timeout has occurred + and return without doing anything. We lose completion and the + command will time out again. + + - EH_NOT_HANDLED + This is the same as when eh_timedout() callback doesn't exist. + Step #2 is taken. + + 2. scsi_eh_scmd_add(scmd, SCSI_EH_CANCEL_CMD) is invoked for the + command. See [1-3] for more information. + + +[1-3] How EH takes over + + scmds enter EH via scsi_eh_scmd_add(), which does the following. + + 1. Turns on scmd->eh_eflags as requested. It's 0 for error + completions and SCSI_EH_CANCEL_CMD for timeouts. + + 2. Links scmd->eh_entry to shost->eh_cmd_q + + 3. Sets SHOST_RECOVERY bit in shost->shost_state + + 4. Increments shost->host_failed + + 5. Wakes up SCSI EH thread if shost->host_busy == shost->host_failed + + As can be seen above, once any scmd is added to shost->eh_cmd_q, +SHOST_RECOVERY shost_state bit is turned on. This prevents any new +scmd to be issued from blk queue to the host; eventually, all scmds on +the host either complete normally, fail and get added to eh_cmd_q, or +time out and get added to shost->eh_cmd_q. + + If all scmds either complete or fail, the number of in-flight scmds +becomes equal to the number of failed scmds - i.e. shost->host_busy == +shost->host_failed. This wakes up SCSI EH thread. So, once woken up, +SCSI EH thread can expect that all in-flight commands have failed and +are linked on shost->eh_cmd_q. + + Note that this does not mean lower layers are quiescent. If a LLDD +completed a scmd with error status, the LLDD and lower layers are +assumed to forget about the scmd at that point. However, if a scmd +has timed out, unless hostt->eh_timedout() made lower layers forget +about the scmd, which currently no LLDD does, the command is still +active as long as lower layers are concerned and completion could +occur at any time. Of course, all such completions are ignored as the +timer has already expired. + + We'll talk about how SCSI EH takes actions to abort - make LLDD +forget about - timed out scmds later. + + +[2] How SCSI EH works + + LLDD's can implement SCSI EH actions in one of the following two +ways. + + - Fine-grained EH callbacks + LLDD can implement fine-grained EH callbacks and let SCSI + midlayer drive error handling and call appropriate callbacks. + This will be dicussed further in [2-1]. + + - eh_strategy_handler() callback + This is one big callback which should perform whole error + handling. As such, it should do all choirs SCSI midlayer + performs during recovery. This will be discussed in [2-2]. + + Once recovery is complete, SCSI EH resumes normal operation by +calling scsi_restart_operations(), which + + 1. Checks if door locking is needed and locks door. + + 2. Clears SHOST_RECOVERY shost_state bit + + 3. Wakes up waiters on shost->host_wait. This occurs if someone + calls scsi_block_when_processing_errors() on the host. + (*QUESTION* why is it needed? All operations will be blocked + anyway after it reaches blk queue.) + + 4. Kicks queues in all devices on the host in the asses + + +[2-1] EH through fine-grained callbacks + +[2-1-1] Overview + + If eh_strategy_handler() is not present, SCSI midlayer takes charge +of driving error handling. EH's goals are two - make LLDD, host and +device forget about timed out scmds and make them ready for new +commands. A scmd is said to be recovered if the scmd is forgotten by +lower layers and lower layers are ready to process or fail the scmd +again. + + To achieve these goals, EH performs recovery actions with increasing +severity. Some actions are performed by issueing SCSI commands and +others are performed by invoking one of the following fine-grained +hostt EH callbacks. Callbacks may be omitted and omitted ones are +considered to fail always. + +int (* eh_abort_handler)(struct scsi_cmnd *); +int (* eh_device_reset_handler)(struct scsi_cmnd *); +int (* eh_bus_reset_handler)(struct scsi_cmnd *); +int (* eh_host_reset_handler)(struct scsi_cmnd *); + + Higher-severity actions are taken only when lower-severity actions +cannot recover some of failed scmds. Also, note that failure of the +highest-severity action means EH failure and results in offlining of +all unrecovered devices. + + During recovery, the following rules are followed + + - Recovery actions are performed on failed scmds on the to do list, + eh_work_q. If a recovery action succeeds for a scmd, recovered + scmds are removed from eh_work_q. + + Note that single recovery action on a scmd can recover multiple + scmds. e.g. resetting a device recovers all failed scmds on the + device. + + - Higher severity actions are taken iff eh_work_q is not empty after + lower severity actions are complete. + + - EH reuses failed scmds to issue commands for recovery. For + timed-out scmds, SCSI EH ensures that LLDD forgets about a scmd + before reusing it for EH commands. + + When a scmd is recovered, the scmd is moved from eh_work_q to EH +local eh_done_q using scsi_eh_finish_cmd(). After all scmds are +recovered (eh_work_q is empty), scsi_eh_flush_done_q() is invoked to +either retry or error-finish (notify upper layer of failure) recovered +scmds. + + scmds are retried iff its sdev is still online (not offlined during +EH), REQ_FAILFAST is not set and ++scmd->retries is less than +scmd->allowed. + + +[2-1-2] Flow of scmds through EH + + 1. Error completion / time out + ACTION: scsi_eh_scmd_add() is invoked for scmd + - set scmd->eh_eflags + - add scmd to shost->eh_cmd_q + - set SHOST_RECOVERY + - shost->host_failed++ + LOCKING: shost->host_lock + + 2. EH starts + ACTION: move all scmds to EH's local eh_work_q. shost->eh_cmd_q + is cleared. + LOCKING: shost->host_lock (not strictly necessary, just for + consistency) + + 3. scmd recovered + ACTION: scsi_eh_finish_cmd() is invoked to EH-finish scmd + - shost->host_failed-- + - clear scmd->eh_eflags + - scsi_setup_cmd_retry() + - move from local eh_work_q to local eh_done_q + LOCKING: none + + 4. EH completes + ACTION: scsi_eh_flush_done_q() retries scmds or notifies upper + layer of failure. + - scmd is removed from eh_done_q and scmd->eh_entry is cleared + - if retry is necessary, scmd is requeued using + scsi_queue_insert() + - otherwise, scsi_finish_command() is invoked for scmd + LOCKING: queue or finish function performs appropriate locking + + +[2-1-3] Flow of control + + EH through fine-grained callbacks start from scsi_unjam_host(). + +<> + + 1. Lock shost->host_lock, splice_init shost->eh_cmd_q into local + eh_work_q and unlock host_lock. Note that shost->eh_cmd_q is + cleared by this action. + + 2. Invoke scsi_eh_get_sense. + + <> + + This action is taken for each error-completed + (!SCSI_EH_CANCEL_CMD) commands without valid sense data. Most + SCSI transports/LLDDs automatically acquire sense data on + command failures (autosense). Autosense is recommended for + performance reasons and as sense information could get out of + sync inbetween occurrence of CHECK CONDITION and this action. + + Note that if autosense is not supported, scmd->sense_buffer + contains invalid sense data when error-completing the scmd + with scsi_done(). scsi_decide_disposition() always returns + FAILED in such cases thus invoking SCSI EH. When the scmd + reaches here, sense data is acquired and + scsi_decide_disposition() is called again. + + 1. Invoke scsi_request_sense() which issues REQUEST_SENSE + command. If fails, no action. Note that taking no action + causes higher-severity recovery to be taken for the scmd. + + 2. Invoke scsi_decide_disposition() on the scmd + + - SUCCESS + scmd->retries is set to scmd->allowed preventing + scsi_eh_flush_done_q() from retrying the scmd and + scsi_eh_finish_cmd() is invoked. + + - NEEDS_RETRY + scsi_eh_finish_cmd() invoked + + - otherwise + No action. + + 3. If !list_empty(&eh_work_q), invoke scsi_eh_abort_cmds(). + + <> + + This action is taken for each timed out command. + hostt->eh_abort_handler() is invoked for each scmd. The + handler returns SUCCESS if it has succeeded to make LLDD and + all related hardware forget about the scmd. + + If a timedout scmd is successfully aborted and the sdev is + either offline or ready, scsi_eh_finish_cmd() is invoked for + the scmd. Otherwise, the scmd is left in eh_work_q for + higher-severity actions. + + Note that both offline and ready status mean that the sdev is + ready to process new scmds, where processing also implies + immediate failing; thus, if a sdev is in one of the two + states, no further recovery action is needed. + + Device readiness is tested using scsi_eh_tur() which issues + TEST_UNIT_READY command. Note that the scmd must have been + aborted successfully before reusing it for TEST_UNIT_READY. + + 4. If !list_empty(&eh_work_q), invoke scsi_eh_ready_devs() + + <> + + This function takes four increasingly more severe measures to + make failed sdevs ready for new commands. + + 1. Invoke scsi_eh_stu() + + <> + + For each sdev which has failed scmds with valid sense data + of which scsi_check_sense()'s verdict is FAILED, + START_STOP_UNIT command is issued w/ start=1. Note that + as we explicitly choose error-completed scmds, it is known + that lower layers have forgotten about the scmd and we can + reuse it for STU. + + If STU succeeds and the sdev is either offline or ready, + all failed scmds on the sdev are EH-finished with + scsi_eh_finish_cmd(). + + *NOTE* If hostt->eh_abort_handler() isn't implemented or + failed, we may still have timed out scmds at this point + and STU doesn't make lower layers forget about those + scmds. Yet, this function EH-finish all scmds on the sdev + if STU succeeds leaving lower layers in an inconsistent + state. It seems that STU action should be taken only when + a sdev has no timed out scmd. + + 2. If !list_empty(&eh_work_q), invoke scsi_eh_bus_device_reset(). + + <> + + This action is very similar to scsi_eh_stu() except that, + instead of issuing STU, hostt->eh_device_reset_handler() + is used. Also, as we're not issuing SCSI commands and + resetting clears all scmds on the sdev, there is no need + to choose error-completed scmds. + + 3. If !list_empty(&eh_work_q), invoke scsi_eh_bus_reset() + + <> + + hostt->eh_bus_reset_handler() is invoked for each channel + with failed scmds. If bus reset succeeds, all failed + scmds on all ready or offline sdevs on the channel are + EH-finished. + + 4. If !list_empty(&eh_work_q), invoke scsi_eh_host_reset() + + <> + + This is the last resort. hostt->eh_host_reset_handler() + is invoked. If host reset succeeds, all failed scmds on + all ready or offline sdevs on the host are EH-finished. + + 5. If !list_empty(&eh_work_q), invoke scsi_eh_offline_sdevs() + + <> + + Take all sdevs which still have unrecovered scmds offline + and EH-finish the scmds. + + 5. Invoke scsi_eh_flush_done_q(). + + <> + + At this point all scmds are recovered (or given up) and + put on eh_done_q by scsi_eh_finish_cmd(). This function + flushes eh_done_q by either retrying or notifying upper + layer of failure of the scmds. + + +[2-2] EH through hostt->eh_strategy_handler() + + hostt->eh_strategy_handler() is invoked in the place of +scsi_unjam_host() and it is responsible for whole recovery process. +On completion, the handler should have made lower layers forget about +all failed scmds and either ready for new commands or offline. Also, +it should perform SCSI EH maintenance choirs to maintain integrity of +SCSI midlayer. IOW, of the steps described in [2-1-2], all steps +except for #1 must be implemented by eh_strategy_handler(). + + +[2-2-1] Pre hostt->eh_strategy_handler() SCSI midlayer conditions + + The following conditions are true on entry to the handler. + + - Each failed scmd's eh_flags field is set appropriately. + + - Each failed scmd is linked on scmd->eh_cmd_q by scmd->eh_entry. + + - SHOST_RECOVERY is set. + + - shost->host_failed == shost->host_busy + + +[2-2-2] Post hostt->eh_strategy_handler() SCSI midlayer conditions + + The following conditions must be true on exit from the handler. + + - shost->host_failed is zero. + + - Each scmd's eh_eflags field is cleared. + + - Each scmd is in such a state that scsi_setup_cmd_retry() on the + scmd doesn't make any difference. + + - shost->eh_cmd_q is cleared. + + - Each scmd->eh_entry is cleared. + + - Either scsi_queue_insert() or scsi_finish_command() is called on + each scmd. Note that the handler is free to use scmd->retries and + ->allowed to limit the number of retries. + + +[2-2-3] Things to consider + + - Know that timed out scmds are still active on lower layers. Make + lower layers forget about them before doing anything else with + those scmds. + + - For consistency, when accessing/modifying shost data structure, + grab shost->host_lock. + + - On completion, each failed sdev must have forgotten about all + active scmds. + + - On completion, each failed sdev must be ready for new commands or + offline. + + +-- +Tejun Heo +htejun@gmail.com +11th September 2005 -- cgit v1.1