The purpose of this InfiniBand SRP initiator patch series is as follows: - Make the SRP initiator driver better suited for use in a H.A. setup. Add fast_io_fail_tmo, dev_loss_tmo and reconnect_delay parameters. With the default values of these parameters failover happens significantly faster. The dev_loss mechanism can be disabled which makes it possible to avoid device removal which is necessary when e.g. using initiator side mirroring. - Improve performance by making the queue size configurable. - Make it possible to figure out which SCSI host corresponds to which SRP initiator port by making the SGID (source GID) available in sysfs. The changes since the previous version of this patch series are as follows (see also http://thread.gmane.org/gmane.linux.drivers.rdma/17693): - Renamed the "can_queue" parameter into "queue_size". - Corrected the title of the introductory e-mail - changed kernel version "3.12" into "3.13". - Corrected the description of /sys/class/srp_remote_ports/port-<h>:<n>/state. - Corrected sysfs kernel version and date for the newly introduced sysfs attributes. - Fixed a hard to trigger race condition that could be triggered only with identical values of reconnect_delay and fast_io_fail_tmo and that could cause failback not to occur (see also rport_fast_io_fail_timedout()). Note: I don't think it's useful for anyone to set reconnect_delay identical to fast_io_fail_tmo. The individual patches in this series are: 0001-IB-srp-Make-transport-layer-retry-count-configurable.patch 0002-IB-srp-Keep-rport-as-long-as-the-IB-transport-layer.patch 0003-scsi_transport_srp-Add-transport-layer-error-handlin.patch 0004-IB-srp-Use-SRP-transport-layer-error-recovery.patch 0005-IB-srp-Start-timers-if-a-transport-layer-error-occur.patch 0006-scsi_transport_srp-Add-periodic-reconnect-support.patch 0007-IB-srp-Add-periodic-reconnect-functionality.patch 0008-IB-srp-Export-sgid-to-sysfs.patch 0009-IB-srp-Introduce-srp_alloc_req_data.patch 0010-IB-srp-Make-queue-size-configurable.patch The full diff compared with the previous version is as follows: diff --git a/Documentation/ABI/stable/sysfs-driver-ib_srp b/Documentation/ABI/stable/sysfs-driver-ib_srp index ab8efd5..b9688de 100644 --- a/Documentation/ABI/stable/sysfs-driver-ib_srp +++ b/Documentation/ABI/stable/sysfs-driver-ib_srp @@ -63,6 +63,10 @@ Description: Interface for making ib_srp connect to a new target. over multiple CPU's. * tl_retry_count, a number in the range 2..7 specifying the IB RC retry count. + * queue_size, the maximum number of commands that the + initiator is allowed to queue per SCSI host. The default + value for this parameter is 62. The lowest supported value + is 2. What: /sys/class/infiniband_srp/srp-<hca>-<port_number>/ibdev Date: January 2, 2006 @@ -156,8 +160,8 @@ Description: InfiniBand service ID used for establishing communication with the SRP target. What: /sys/class/scsi_host/host<n>/sgid -Date: December 1, 2013 -KernelVersion: 3.12 +Date: February 1, 2014 +KernelVersion: 3.13 Contact: linux-rdma@xxxxxxxxxxxxxxx Description: InfiniBand GID of the source port used for communication with the SRP target. diff --git a/Documentation/ABI/stable/sysfs-transport-srp b/Documentation/ABI/stable/sysfs-transport-srp index 21bd480..ec7af69 100644 --- a/Documentation/ABI/stable/sysfs-transport-srp +++ b/Documentation/ABI/stable/sysfs-transport-srp @@ -6,8 +6,8 @@ Description: Instructs an SRP initiator to disconnect from a target and to remove all LUNs imported from that target. What: /sys/class/srp_remote_ports/port-<h>:<n>/dev_loss_tmo -Date: December 1, 2013 -KernelVersion: 3.12 +Date: February 1, 2014 +KernelVersion: 3.13 Contact: linux-scsi@xxxxxxxxxxxxxxx, linux-rdma@xxxxxxxxxxxxxxx Description: Number of seconds the SCSI layer will wait after a transport layer error has been observed before removing a target port. @@ -15,8 +15,8 @@ Description: Number of seconds the SCSI layer will wait after a transport will disable the dev_loss timer. What: /sys/class/srp_remote_ports/port-<h>:<n>/fast_io_fail_tmo -Date: December 1, 2013 -KernelVersion: 3.12 +Date: February 1, 2014 +KernelVersion: 3.13 Contact: linux-scsi@xxxxxxxxxxxxxxx, linux-rdma@xxxxxxxxxxxxxxx Description: Number of seconds the SCSI layer will wait after a transport layer error has been observed before failing I/O. Zero means @@ -31,8 +31,8 @@ Description: 16-byte local SRP port identifier in hexadecimal format. An example: 4c:49:4e:55:58:20:56:49:4f:00:00:00:00:00:00:00. What: /sys/class/srp_remote_ports/port-<h>:<n>/reconnect_delay -Date: December 1, 2013 -KernelVersion: 3.12 +Date: February 1, 2014 +KernelVersion: 3.13 Contact: linux-scsi@xxxxxxxxxxxxxxx, linux-rdma@xxxxxxxxxxxxxxx Description: Number of seconds the SCSI layer will wait after a reconnect attempt failed before retrying. Setting this attribute to @@ -45,14 +45,14 @@ Contact: linux-scsi@xxxxxxxxxxxxxxx Description: Role of the remote port. Either "SRP Initiator" or "SRP Target". What: /sys/class/srp_remote_ports/port-<h>:<n>/state -Date: December 1, 2013 -KernelVersion: 3.12 +Date: February 1, 2014 +KernelVersion: 3.13 Contact: linux-scsi@xxxxxxxxxxxxxxx, linux-rdma@xxxxxxxxxxxxxxx Description: State of the transport layer used for communication with the remote port. "running" if the transport layer is operational; "blocked" if a transport layer error has been encountered but - the fail_io_fast_tmo timer has not yet fired; "fail-fast" - after the fail_io_fast_tmo timer has fired and before the + the fast_io_fail_tmo timer has not yet fired; "fail-fast" + after the fast_io_fail_tmo timer has fired and before the "dev_loss_tmo" timer has fired; "lost" after the "dev_loss_tmo" timer has fired and before the port is finally removed. diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index e158f59..b4bd903 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -2579,7 +2579,7 @@ enum { SRP_OPT_SG_TABLESIZE = 1 << 11, SRP_OPT_COMP_VECTOR = 1 << 12, SRP_OPT_TL_RETRY_COUNT = 1 << 13, - SRP_OPT_CAN_QUEUE = 1 << 14, + SRP_OPT_QUEUE_SIZE = 1 << 14, SRP_OPT_USE_FAST_REG = 1 << 15, SRP_OPT_ALL = (SRP_OPT_ID_EXT | SRP_OPT_IOC_GUID | @@ -2603,7 +2603,7 @@ static const match_table_t srp_opt_tokens = { { SRP_OPT_SG_TABLESIZE, "sg_tablesize=%u" }, { SRP_OPT_COMP_VECTOR, "comp_vector=%u" }, { SRP_OPT_TL_RETRY_COUNT, "tl_retry_count=%u" }, - { SRP_OPT_CAN_QUEUE, "can_queue=%d" }, + { SRP_OPT_QUEUE_SIZE, "queue_size=%d" }, { SRP_OPT_USE_FAST_REG, "use_fast_reg=%d" }, { SRP_OPT_ERR, NULL } }; @@ -2699,9 +2699,9 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target) target->scsi_host->max_sectors = token; break; - case SRP_OPT_CAN_QUEUE: + case SRP_OPT_QUEUE_SIZE: if (match_int(args, &token) || token < 1) { - pr_warn("bad can_queue parameter '%s'\n", p); + pr_warn("bad queue_size parameter '%s'\n", p); goto out; } target->scsi_host->can_queue = token; @@ -2816,7 +2816,7 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target) if (target->scsi_host->cmd_per_lun > target->scsi_host->can_queue && (opt_mask & SRP_OPT_MAX_CMD_PER_LUN)) - pr_warn("cmd_per_lun = %d > can_queue = %d\n", + pr_warn("cmd_per_lun = %d > queue_size = %d\n", target->scsi_host->cmd_per_lun, target->scsi_host->can_queue); diff --git a/drivers/scsi/scsi_transport_srp.c b/drivers/scsi/scsi_transport_srp.c index 8f66ed4..8b9cb22 100644 --- a/drivers/scsi/scsi_transport_srp.c +++ b/drivers/scsi/scsi_transport_srp.c @@ -423,7 +423,8 @@ static void rport_fast_io_fail_timedout(struct work_struct *work) dev_name(&rport->dev), dev_name(&shost->shost_gendev)); mutex_lock(&rport->mutex); - __rport_fail_io_fast(rport); + if (rport->state == SRP_RPORT_BLOCKED) + __rport_fail_io_fast(rport); mutex_unlock(&rport->mutex); } -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
