--- /dev/null
+Contents:
+
+1) TCM Userspace Design
+ a) Background
+ b) Benefits
+ c) Design constraints
+ d) Implementation overview
+ i. Mailbox
+ ii. Command ring
+ iii. Data Area
+ e) Device discovery
+ f) Device events
+ g) Other contingencies
+2) Writing a user pass-through handler
+ a) Discovering and configuring TCMU uio devices
+ b) Waiting for events on the device(s)
+ c) Managing the command ring
+3) Command filtering and pass_level
+4) A final note
+
+
+TCM Userspace Design
+--------------------
+
+TCM is another name for LIO, an in-kernel iSCSI target (server).
+Existing TCM targets run in the kernel. TCMU (TCM in Userspace)
+allows userspace programs to be written which act as iSCSI targets.
+This document describes the design.
+
+The existing kernel provides modules for different SCSI transport
+protocols. TCM also modularizes the data storage. There are existing
+modules for file, block device, RAM or using another SCSI device as
+storage. These are called "backstores" or "storage engines". These
+built-in modules are implemented entirely as kernel code.
+
+Background:
+
+In addition to modularizing the transport protocol used for carrying
+SCSI commands ("fabrics"), the Linux kernel target, LIO, also modularizes
+the actual data storage as well. These are referred to as "backstores"
+or "storage engines". The target comes with backstores that allow a
+file, a block device, RAM, or another SCSI device to be used for the
+local storage needed for the exported SCSI LUN. Like the rest of LIO,
+these are implemented entirely as kernel code.
+
+These backstores cover the most common use cases, but not all. One new
+use case that other non-kernel target solutions, such as tgt, are able
+to support is using Gluster's GLFS or Ceph's RBD as a backstore. The
+target then serves as a translator, allowing initiators to store data
+in these non-traditional networked storage systems, while still only
+using standard protocols themselves.
+
+If the target is a userspace process, supporting these is easy. tgt,
+for example, needs only a small adapter module for each, because the
+modules just use the available userspace libraries for RBD and GLFS.
+
+Adding support for these backstores in LIO is considerably more
+difficult, because LIO is entirely kernel code. Instead of undertaking
+the significant work to port the GLFS or RBD APIs and protocols to the
+kernel, another approach is to create a userspace pass-through
+backstore for LIO, "TCMU".
+
+
+Benefits:
+
+In addition to allowing relatively easy support for RBD and GLFS, TCMU
+will also allow easier development of new backstores. TCMU combines
+with the LIO loopback fabric to become something similar to FUSE
+(Filesystem in Userspace), but at the SCSI layer instead of the
+filesystem layer. A SUSE, if you will.
+
+The disadvantage is there are more distinct components to configure, and
+potentially to malfunction. This is unavoidable, but hopefully not
+fatal if we're careful to keep things as simple as possible.
+
+Design constraints:
+
+- Good performance: high throughput, low latency
+- Cleanly handle if userspace:
+ 1) never attaches
+ 2) hangs
+ 3) dies
+ 4) misbehaves
+- Allow future flexibility in user & kernel implementations
+- Be reasonably memory-efficient
+- Simple to configure & run
+- Simple to write a userspace backend
+
+
+Implementation overview:
+
+The core of the TCMU interface is a memory region that is shared
+between kernel and userspace. Within this region is: a control area
+(mailbox); a lockless producer/consumer circular buffer for commands
+to be passed up, and status returned; and an in/out data buffer area.
+
+TCMU uses the pre-existing UIO subsystem. UIO allows device driver
+development in userspace, and this is conceptually very close to the
+TCMU use case, except instead of a physical device, TCMU implements a
+memory-mapped layout designed for SCSI commands. Using UIO also
+benefits TCMU by handling device introspection (e.g. a way for
+userspace to determine how large the shared region is) and signaling
+mechanisms in both directions.
+
+There are no embedded pointers in the memory region. Everything is
+expressed as an offset from the region's starting address. This allows
+the ring to still work if the user process dies and is restarted with
+the region mapped at a different virtual address.
+
+See target_core_user.h for the struct definitions.
+
+The Mailbox:
+
+The mailbox is always at the start of the shared memory region, and
+contains a version, details about the starting offset and size of the
+command ring, and head and tail pointers to be used by the kernel and
+userspace (respectively) to put commands on the ring, and indicate
+when the commands are completed.
+
+version - 1 (userspace should abort if otherwise)
+flags - none yet defined.
+cmdr_off - The offset of the start of the command ring from the start
+of the memory region, to account for the mailbox size.
+cmdr_size - The size of the command ring. This does *not* need to be a
+power of two.
+cmd_head - Modified by the kernel to indicate when a command has been
+placed on the ring.
+cmd_tail - Modified by userspace to indicate when it has completed
+processing of a command.
+
+The Command Ring:
+
+Commands are placed on the ring by the kernel incrementing
+mailbox.cmd_head by the size of the command, modulo cmdr_size, and
+then signaling userspace via uio_event_notify(). Once the command is
+completed, userspace updates mailbox.cmd_tail in the same way and
+signals the kernel via a 4-byte write(). When cmd_head equals
+cmd_tail, the ring is empty -- no commands are currently waiting to be
+processed by userspace.
+
+TCMU commands start with a common header containing "len_op", a 32-bit
+value that stores the length, as well as the opcode in the lowest
+unused bits. Currently only two opcodes are defined, TCMU_OP_PAD and
+TCMU_OP_CMD. When userspace encounters a command with PAD opcode, it
+should skip ahead by the bytes in "length". (The kernel inserts PAD
+entries to ensure each CMD entry fits contigously into the circular
+buffer.)
+
+When userspace handles a CMD, it finds the SCSI CDB (Command Data
+Block) via tcmu_cmd_entry.req.cdb_off. This is an offset from the
+start of the overall shared memory region, not the entry. The data
+in/out buffers are accessible via tht req.iov[] array. Note that
+each iov.iov_base is also an offset from the start of the region.
+
+TCMU currently does not support BIDI operations.
+
+When completing a command, userspace sets rsp.scsi_status, and
+rsp.sense_buffer if necessary. Userspace then increments
+mailbox.cmd_tail by entry.hdr.length (mod cmdr_size) and signals the
+kernel via the UIO method, a 4-byte write to the file descriptor.
+
+The Data Area:
+
+This is shared-memory space after the command ring. The organization
+of this area is not defined in the TCMU interface, and userspace
+should access only the parts referenced by pending iovs.
+
+
+Device Discovery:
+
+Other devices may be using UIO besides TCMU. Unrelated user processes
+may also be handling different sets of TCMU devices. TCMU userspace
+processes must find their devices by scanning sysfs
+class/uio/uio*/name. For TCMU devices, these names will be of the
+format:
+
+tcm-user/<hba_num>/<device_name>/<subtype>/<path>
+
+where "tcm-user" is common for all TCMU-backed UIO devices. <hba_num>
+and <device_name> allow userspace to find the device's path in the
+kernel target's configfs tree. Assuming the usual mount point, it is
+found at:
+
+/sys/kernel/config/target/core/user_<hba_num>/<device_name>
+
+This location contains attributes such as "hw_block_size", that
+userspace needs to know for correct operation.
+
+<subtype> will be a userspace-process-unique string to identify the
+TCMU device as expecting to be backed by a certain handler, and <path>
+will be an additional handler-specific string for the user process to
+configure the device, if needed. The name cannot contain ':', due to
+LIO limitations.
+
+For all devices so discovered, the user handler opens /dev/uioX and
+calls mmap():
+
+mmap(NULL, size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0)
+
+where size must be equal to the value read from
+/sys/class/uio/uioX/maps/map0/size.
+
+
+Device Events:
+
+If a new device is added or removed, a notification will be broadcast
+over netlink, using a generic netlink family name of "TCM-USER" and a
+multicast group named "config". This will include the UIO name as
+described in the previous section, as well as the UIO minor
+number. This should allow userspace to identify both the UIO device and
+the LIO device, so that after determining the device is supported
+(based on subtype) it can take the appropriate action.
+
+
+Other contingencies:
+
+Userspace handler process never attaches:
+
+- TCMU will post commands, and then abort them after a timeout period
+ (30 seconds.)
+
+Userspace handler process is killed:
+
+- It is still possible to restart and re-connect to TCMU
+ devices. Command ring is preserved. However, after the timeout period,
+ the kernel will abort pending tasks.
+
+Userspace handler process hangs:
+
+- The kernel will abort pending tasks after a timeout period.
+
+Userspace handler process is malicious:
+
+- The process can trivially break the handling of devices it controls,
+ but should not be able to access kernel memory outside its shared
+ memory areas.
+
+
+Writing a user pass-through handler (with example code)
+-------------------------------------------------------
+
+A user process handing a TCMU device must support the following:
+
+a) Discovering and configuring TCMU uio devices
+b) Waiting for events on the device(s)
+c) Managing the command ring: Parsing operations and commands,
+ performing work as needed, setting response fields (scsi_status and
+ possibly sense_buffer), updating cmd_tail, and notifying the kernel
+ that work has been finished
+
+First, consider instead writing a plugin for tcmu-runner. tcmu-runner
+implements all of this, and provides a higher-level API for plugin
+authors.
+
+TCMU is designed so that multiple unrelated processes can manage TCMU
+devices separately. All handlers should make sure to only open their
+devices, based opon a known subtype string.
+
+a) Discovering and configuring TCMU UIO devices:
+
+(error checking omitted for brevity)
+
+int fd, dev_fd;
+char buf[256];
+unsigned long long map_len;
+void *map;
+
+fd = open("/sys/class/uio/uio0/name", O_RDONLY);
+ret = read(fd, buf, sizeof(buf));
+close(fd);
+buf[ret-1] = '\0'; /* null-terminate and chop off the \n */
+
+/* we only want uio devices whose name is a format we expect */
+if (strncmp(buf, "tcm-user", 8))
+ exit(-1);
+
+/* Further checking for subtype also needed here */
+
+fd = open(/sys/class/uio/%s/maps/map0/size, O_RDONLY);
+ret = read(fd, buf, sizeof(buf));
+close(fd);
+str_buf[ret-1] = '\0'; /* null-terminate and chop off the \n */
+
+map_len = strtoull(buf, NULL, 0);
+
+dev_fd = open("/dev/uio0", O_RDWR);
+map = mmap(NULL, map_len, PROT_READ|PROT_WRITE, MAP_SHARED, dev_fd, 0);
+
+
+b) Waiting for events on the device(s)
+
+while (1) {
+ char buf[4];
+
+ int ret = read(dev_fd, buf, 4); /* will block */
+
+ handle_device_events(dev_fd, map);
+}
+
+
+c) Managing the command ring
+
+#include <linux/target_core_user.h>
+
+int handle_device_events(int fd, void *map)
+{
+ struct tcmu_mailbox *mb = map;
+ struct tcmu_cmd_entry *ent = (void *) mb + mb->cmdr_off + mb->cmd_tail;
+ int did_some_work = 0;
+
+ /* Process events from cmd ring until we catch up with cmd_head */
+ while (ent != (void *)mb + mb->cmdr_off + mb->cmd_head) {
+
+ if (tcmu_hdr_get_op(&ent->hdr) == TCMU_OP_CMD) {
+ uint8_t *cdb = (void *)mb + ent->req.cdb_off;
+ bool success = true;
+
+ /* Handle command here. */
+ printf("SCSI opcode: 0x%x\n", cdb[0]);
+
+ /* Set response fields */
+ if (success)
+ ent->rsp.scsi_status = SCSI_NO_SENSE;
+ else {
+ /* Also fill in rsp->sense_buffer here */
+ ent->rsp.scsi_status = SCSI_CHECK_CONDITION;
+ }
+ }
+ else {
+ /* Do nothing for PAD entries */
+ }
+
+ /* update cmd_tail */
+ mb->cmd_tail = (mb->cmd_tail + tcmu_hdr_get_len(&ent->hdr)) % mb->cmdr_size;
+ ent = (void *) mb + mb->cmdr_off + mb->cmd_tail;
+ did_some_work = 1;
+ }
+
+ /* Notify the kernel that work has been finished */
+ if (did_some_work) {
+ uint32_t buf = 0;
+
+ write(fd, &buf, 4);
+ }
+
+ return 0;
+}
+
+
+Command filtering and pass_level
+--------------------------------
+
+TCMU supports a "pass_level" option with valid values of 0 or 1. When
+the value is 0 (the default), nearly all SCSI commands received for
+the device are passed through to the handler. This allows maximum
+flexibility but increases the amount of code required by the handler,
+to support all mandatory SCSI commands. If pass_level is set to 1,
+then only IO-related commands are presented, and the rest are handled
+by LIO's in-kernel command emulation. The commands presented at level
+1 include all versions of:
+
+READ
+WRITE
+WRITE_VERIFY
+XDWRITEREAD
+WRITE_SAME
+COMPARE_AND_WRITE
+SYNCHRONIZE_CACHE
+UNMAP
+
+
+A final note
+------------
+
+Please be careful to return codes as defined by the SCSI
+specifications. These are different than some values defined in the
+scsi/scsi.h include file. For example, CHECK CONDITION's status code
+is 2, not 1.
isert_cmd->tx_desc.num_sge = 2;
}
- isert_init_send_wr(isert_conn, isert_cmd, send_wr, true);
+ isert_init_send_wr(isert_conn, isert_cmd, send_wr, false);
pr_debug("Posting SCSI Response IB_WR_SEND >>>>>>>>>>>>>>>>>>>>>>\n");
&isert_cmd->tx_desc.iscsi_header);
isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc);
isert_init_send_wr(isert_conn, isert_cmd,
- &isert_cmd->tx_desc.send_wr, true);
+ &isert_cmd->tx_desc.send_wr, false);
isert_cmd->rdma_wr.s_send_wr.next = &isert_cmd->tx_desc.send_wr;
wr->send_wr_num += 1;
}
accept_wait:
ret = down_interruptible(&isert_np->np_sem);
- if (max_accept > 5)
+ if (ret || max_accept > 5)
return -ENODEV;
spin_lock_bh(&np->np_thread_lock);
struct qla_tgt_cmd *cmd);
static void qlt_alloc_qfull_cmd(struct scsi_qla_host *vha,
struct atio_from_isp *atio, uint16_t status, int qfull);
+static void qlt_disable_vha(struct scsi_qla_host *vha);
/*
* Global Variables
*/
spin_unlock_irqrestore(&vha->hw->tgt.q_full_lock, flags);
}
-void qlt_24xx_atio_pkt_all_vps(struct scsi_qla_host *vha,
+static void qlt_24xx_atio_pkt_all_vps(struct scsi_qla_host *vha,
struct atio_from_isp *atio)
{
ql_dbg(ql_dbg_tgt, vha, 0xe072,
#if 0 /* FIXME: Re-enable Global event handling.. */
/* Global event */
atomic_inc(&ha->tgt.qla_tgt->tgt_global_resets_count);
- qlt_clear_tgt_db(ha->tgt.qla_tgt, 1);
+ qlt_clear_tgt_db(ha->tgt.qla_tgt);
if (!list_empty(&ha->tgt.qla_tgt->sess_list)) {
sess = list_entry(ha->tgt.qla_tgt->sess_list.next,
typeof(*sess), sess_list_entry);
}
/* ha->hardware_lock supposed to be held on entry */
-static void qlt_clear_tgt_db(struct qla_tgt *tgt, bool local_only)
+static void qlt_clear_tgt_db(struct qla_tgt *tgt)
{
struct qla_tgt_sess *sess;
mutex_lock(&vha->vha_tgt.tgt_mutex);
spin_lock_irqsave(&ha->hardware_lock, flags);
tgt->tgt_stop = 1;
- qlt_clear_tgt_db(tgt, true);
+ qlt_clear_tgt_db(tgt);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
mutex_unlock(&vha->vha_tgt.tgt_mutex);
mutex_unlock(&qla_tgt_mutex);
return -1;
}
-static inline void qlt_unmap_sg(struct scsi_qla_host *vha,
- struct qla_tgt_cmd *cmd)
+static void qlt_unmap_sg(struct scsi_qla_host *vha, struct qla_tgt_cmd *cmd)
{
struct qla_hw_data *ha = vha->hw;
- BUG_ON(!cmd->sg_mapped);
+ if (!cmd->sg_mapped)
+ return;
+
pci_unmap_sg(ha->pdev, cmd->sg, cmd->sg_cnt, cmd->dma_data_direction);
cmd->sg_mapped = 0;
return 0;
out_unmap_unlock:
- if (cmd->sg_mapped)
- qlt_unmap_sg(vha, cmd);
+ qlt_unmap_sg(vha, cmd);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return res;
return res;
out_unlock_free_unmap:
- if (cmd->sg_mapped)
- qlt_unmap_sg(vha, cmd);
+ qlt_unmap_sg(vha, cmd);
spin_unlock_irqrestore(&ha->hardware_lock, flags);
return res;
if (!ha_locked && !in_interrupt())
msleep(250); /* just in case */
- if (cmd->sg_mapped)
- qlt_unmap_sg(vha, cmd);
+ qlt_unmap_sg(vha, cmd);
vha->hw->tgt.tgt_ops->free_cmd(cmd);
}
return;
tfo = se_cmd->se_tfo;
cmd->cmd_sent_to_fw = 0;
- if (cmd->sg_mapped)
- qlt_unmap_sg(vha, cmd);
+ qlt_unmap_sg(vha, cmd);
if (unlikely(status != CTIO_SUCCESS)) {
switch (status & 0xFFFF) {
EXPORT_SYMBOL(qlt_lport_deregister);
/* Must be called under HW lock */
-void qlt_set_mode(struct scsi_qla_host *vha)
+static void qlt_set_mode(struct scsi_qla_host *vha)
{
struct qla_hw_data *ha = vha->hw;
}
/* Must be called under HW lock */
-void qlt_clear_mode(struct scsi_qla_host *vha)
+static void qlt_clear_mode(struct scsi_qla_host *vha)
{
struct qla_hw_data *ha = vha->hw;
*
* Disable Target Mode and reset the adapter
*/
-void
-qlt_disable_vha(struct scsi_qla_host *vha)
+static void qlt_disable_vha(struct scsi_qla_host *vha)
{
struct qla_hw_data *ha = vha->hw;
struct qla_tgt *tgt = vha->vha_tgt.qla_tgt;
struct qla_tgt *tgt;
void *pkt;
struct scatterlist *sg; /* cmd data buffer SG vector */
+ unsigned char *sense_buffer;
int seg_cnt;
int req_cnt;
uint16_t rq_result;
uint16_t scsi_status;
- unsigned char *sense_buffer;
int sense_buffer_len;
int residual;
int add_status_pkt;
extern struct qla_tgt_data qla_target;
-/*
- * Internal function prototypes
- */
-void qlt_disable_vha(struct scsi_qla_host *);
/*
* Function prototypes for qla_target.c logic used by qla2xxx LLD code.
extern void qlt_unreg_sess(struct qla_tgt_sess *);
extern void qlt_fc_port_added(struct scsi_qla_host *, fc_port_t *);
extern void qlt_fc_port_deleted(struct scsi_qla_host *, fc_port_t *);
-extern void qlt_set_mode(struct scsi_qla_host *ha);
-extern void qlt_clear_mode(struct scsi_qla_host *ha);
extern int __init qlt_init(void);
extern void qlt_exit(void);
extern void qlt_update_vp_map(struct scsi_qla_host *, int);
/*
* Exported symbols from qla_target.c LLD logic used by qla2xxx code..
*/
-extern void qlt_24xx_atio_pkt_all_vps(struct scsi_qla_host *,
- struct atio_from_isp *);
extern void qlt_response_pkt_all_vps(struct scsi_qla_host *, response_t *);
extern int qlt_rdy_to_xfer(struct qla_tgt_cmd *);
extern int qlt_xmit_response(struct qla_tgt_cmd *, int, uint8_t);
-extern int qlt_rdy_to_xfer_dif(struct qla_tgt_cmd *);
-extern int qlt_xmit_response_dif(struct qla_tgt_cmd *, int, uint8_t);
extern void qlt_xmit_tm_rsp(struct qla_tgt_mgmt_cmd *);
extern void qlt_free_mcmd(struct qla_tgt_mgmt_cmd *);
extern void qlt_free_cmd(struct qla_tgt_cmd *cmd);
pr_debug("fc_rport domain: port_id 0x%06x\n", nacl->nport_id);
node = btree_remove32(&lport->lport_fcport_map, nacl->nport_id);
- WARN_ON(node && (node != se_nacl));
+ if (WARN_ON(node && (node != se_nacl))) {
+ /*
+ * The nacl no longer matches what we think it should be.
+ * Most likely a new dynamic acl has been added while
+ * someone dropped the hardware lock. It clearly is a
+ * bug elsewhere, but this bit can't make things worse.
+ */
+ btree_insert32(&lport->lport_fcport_map, nacl->nport_id,
+ node, GFP_ATOMIC);
+ }
pr_debug("Removed from fcport_map: %p for WWNN: 0x%016LX, port_id: 0x%06x\n",
se_nacl, nacl->nport_wwnn, nacl->nport_id);
Say Y here to enable the TCM/pSCSI subsystem plugin for non-buffered
passthrough access to Linux/SCSI device
+config TCM_USER
+ tristate "TCM/USER Subsystem Plugin for Linux"
+ depends on UIO && NET
+ help
+ Say Y here to enable the TCM/USER subsystem plugin for a userspace
+ process to handle requests
+
source "drivers/target/loopback/Kconfig"
source "drivers/target/tcm_fc/Kconfig"
source "drivers/target/iscsi/Kconfig"
obj-$(CONFIG_TCM_IBLOCK) += target_core_iblock.o
obj-$(CONFIG_TCM_FILEIO) += target_core_file.o
obj-$(CONFIG_TCM_PSCSI) += target_core_pscsi.o
+obj-$(CONFIG_TCM_USER) += target_core_user.o
# Fabric modules
obj-$(CONFIG_LOOPBACK_TARGET) += loopback/
struct task_struct *p,
int mode)
{
- char buf[128];
/*
* mode == 1 signals iscsi_target_tx_thread() usage.
* mode == 0 signals iscsi_target_rx_thread() usage.
* both TX and RX kthreads are scheduled to run on the
* same CPU.
*/
- memset(buf, 0, 128);
- cpumask_scnprintf(buf, 128, conn->conn_cpumask);
set_cpus_allowed_ptr(p, conn->conn_cpumask);
}
if (conn->conn_tx_hash.tfm)
crypto_free_hash(conn->conn_tx_hash.tfm);
- if (conn->conn_cpumask)
- free_cpumask_var(conn->conn_cpumask);
+ free_cpumask_var(conn->conn_cpumask);
kfree(conn->conn_ops);
conn->conn_ops = NULL;
} else {
sess = se_sess->fabric_sess_ptr;
- if (sess->sess_ops->InitiatorName)
- rb += sprintf(page+rb, "InitiatorName: %s\n",
- sess->sess_ops->InitiatorName);
- if (sess->sess_ops->InitiatorAlias)
- rb += sprintf(page+rb, "InitiatorAlias: %s\n",
- sess->sess_ops->InitiatorAlias);
+ rb += sprintf(page+rb, "InitiatorName: %s\n",
+ sess->sess_ops->InitiatorName);
+ rb += sprintf(page+rb, "InitiatorAlias: %s\n",
+ sess->sess_ops->InitiatorAlias);
rb += sprintf(page+rb, "LIO Session ID: %u "
"ISID: 0x%02x %02x %02x %02x %02x %02x "
struct iscsi_cmd *cmd,
unsigned char *buf)
{
- int dump = 0, recovery = 0;
u32 data_sn = 0;
struct iscsi_conn *conn = cmd->conn;
struct iscsi_data *hdr = (struct iscsi_data *) buf;
pr_err("Command ITT: 0x%08x, received DataSN: 0x%08x"
" higher than expected 0x%08x.\n", cmd->init_task_tag,
be32_to_cpu(hdr->datasn), data_sn);
- recovery = 1;
goto recover;
} else if (be32_to_cpu(hdr->datasn) < data_sn) {
pr_err("Command ITT: 0x%08x, received DataSN: 0x%08x"
" lower than expected 0x%08x, discarding payload.\n",
cmd->init_task_tag, be32_to_cpu(hdr->datasn), data_sn);
- dump = 1;
goto dump;
}
if (iscsit_dump_data_payload(conn, payload_length, 1) < 0)
return DATAOUT_CANNOT_RECOVER;
- return (recovery || dump) ? DATAOUT_WITHIN_COMMAND_RECOVERY :
- DATAOUT_NORMAL;
+ return DATAOUT_WITHIN_COMMAND_RECOVERY;
}
static int iscsit_dataout_pre_datapduinorder_yes(
return 0;
fail:
np->np_socket = NULL;
- if (sock)
- sock_release(sock);
+ sock_release(sock);
return ret;
}
if (!IS_ERR(conn->conn_tx_hash.tfm))
crypto_free_hash(conn->conn_tx_hash.tfm);
- if (conn->conn_cpumask)
- free_cpumask_var(conn->conn_cpumask);
+ free_cpumask_var(conn->conn_cpumask);
kfree(conn->conn_ops);
iscsit_put_transport(conn->conn_transport);
kfree(conn);
conn = NULL;
- if (ret == -ENODEV)
- goto out;
/* Get another socket */
return 1;
}
if (conn->param_list)
intrname = iscsi_find_param_from_key(INITIATORNAME,
conn->param_list);
- strcpy(ls->last_intr_fail_name,
- (intrname ? intrname->value : "Unknown"));
+ strlcpy(ls->last_intr_fail_name,
+ (intrname ? intrname->value : "Unknown"),
+ sizeof(ls->last_intr_fail_name));
ls->last_intr_fail_ip_family = conn->login_family;
/*
* Locate the SAM Task Attr from struct scsi_cmnd *
*/
-static int tcm_loop_sam_attr(struct scsi_cmnd *sc)
-{
- if (sc->device->tagged_supported) {
- switch (sc->tag) {
- case HEAD_OF_QUEUE_TAG:
- return MSG_HEAD_TAG;
- case ORDERED_QUEUE_TAG:
- return MSG_ORDERED_TAG;
- default:
- break;
- }
- }
+static int tcm_loop_sam_attr(struct scsi_cmnd *sc, int tag)
+{
+ if (sc->device->tagged_supported &&
+ sc->device->ordered_tags && tag >= 0)
+ return MSG_ORDERED_TAG;
return MSG_SIMPLE_TAG;
}
rc = target_submit_cmd_map_sgls(se_cmd, tl_nexus->se_sess, sc->cmnd,
&tl_cmd->tl_sense_buf[0], tl_cmd->sc->device->lun,
- transfer_length, tcm_loop_sam_attr(sc),
+ transfer_length, tcm_loop_sam_attr(sc, tl_cmd->sc_cmd_tag),
sc->sc_data_direction, 0,
scsi_sglist(sc), scsi_sg_count(sc),
sgl_bidi, sgl_bidi_count,
}
tl_cmd->sc = sc;
- tl_cmd->sc_cmd_tag = sc->tag;
+ tl_cmd->sc_cmd_tag = sc->request->tag;
INIT_WORK(&tl_cmd->work, tcm_loop_submission_work);
queue_work(tcm_loop_workqueue, &tl_cmd->work);
return 0;
*/
tl_tpg = &tl_hba->tl_hba_tpgs[sc->device->id];
ret = tcm_loop_issue_tmr(tl_tpg, tl_nexus, sc->device->lun,
- sc->tag, TMR_ABORT_TASK);
+ sc->request->tag, TMR_ABORT_TASK);
return (ret == TMR_FUNCTION_COMPLETE) ? SUCCESS : FAILED;
}
struct tcm_loop_tpg, tl_se_tpg);
struct tcm_loop_hba *tl_hba = tl_tpg->tl_hba;
- atomic_inc(&tl_tpg->tl_tpg_port_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&tl_tpg->tl_tpg_port_count);
/*
* Add Linux/SCSI struct scsi_device by HCTL
*/
scsi_remove_device(sd);
scsi_device_put(sd);
- atomic_dec(&tl_tpg->tl_tpg_port_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&tl_tpg->tl_tpg_port_count);
pr_debug("TCM_Loop_ConfigFS: Port Unlink Successful\n");
}
if (tg_pt_id != tg_pt_gp->tg_pt_gp_id)
continue;
- atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_inc_mb(&tg_pt_gp->tg_pt_gp_ref_cnt);
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
found = true;
spin_lock(&dev->t10_alua.tg_pt_gps_lock);
- atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&tg_pt_gp->tg_pt_gp_ref_cnt);
break;
}
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
* every I_T nexus other than the I_T nexus on which the SET
* TARGET PORT GROUPS command
*/
- atomic_inc(&mem->tg_pt_gp_mem_ref_cnt);
- smp_mb__after_atomic();
+ atomic_inc_mb(&mem->tg_pt_gp_mem_ref_cnt);
spin_unlock(&tg_pt_gp->tg_pt_gp_lock);
spin_lock_bh(&port->sep_alua_lock);
spin_unlock_bh(&port->sep_alua_lock);
spin_lock(&tg_pt_gp->tg_pt_gp_lock);
- atomic_dec(&mem->tg_pt_gp_mem_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&mem->tg_pt_gp_mem_ref_cnt);
}
spin_unlock(&tg_pt_gp->tg_pt_gp_lock);
/*
core_alua_dump_state(tg_pt_gp->tg_pt_gp_alua_pending_state));
spin_lock(&dev->t10_alua.tg_pt_gps_lock);
atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
if (tg_pt_gp->tg_pt_gp_transition_complete)
*/
spin_lock(&dev->t10_alua.tg_pt_gps_lock);
atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
if (!explicit && tg_pt_gp->tg_pt_gp_implicit_trans_secs) {
spin_lock(&local_lu_gp_mem->lu_gp_mem_lock);
lu_gp = local_lu_gp_mem->lu_gp;
atomic_inc(&lu_gp->lu_gp_ref_cnt);
- smp_mb__after_atomic();
spin_unlock(&local_lu_gp_mem->lu_gp_mem_lock);
/*
* For storage objects that are members of the 'default_lu_gp',
l_tg_pt_gp->tg_pt_gp_alua_nacl = l_nacl;
rc = core_alua_do_transition_tg_pt(l_tg_pt_gp,
new_state, explicit);
- atomic_dec(&lu_gp->lu_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&lu_gp->lu_gp_ref_cnt);
return rc;
}
/*
lu_gp_mem_list) {
dev = lu_gp_mem->lu_gp_mem_dev;
- atomic_inc(&lu_gp_mem->lu_gp_mem_ref_cnt);
- smp_mb__after_atomic();
+ atomic_inc_mb(&lu_gp_mem->lu_gp_mem_ref_cnt);
spin_unlock(&lu_gp->lu_gp_lock);
spin_lock(&dev->t10_alua.tg_pt_gps_lock);
tg_pt_gp->tg_pt_gp_alua_port = NULL;
tg_pt_gp->tg_pt_gp_alua_nacl = NULL;
}
- atomic_inc(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_inc_mb(&tg_pt_gp->tg_pt_gp_ref_cnt);
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
/*
* core_alua_do_transition_tg_pt() will always return
new_state, explicit);
spin_lock(&dev->t10_alua.tg_pt_gps_lock);
- atomic_dec(&tg_pt_gp->tg_pt_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&tg_pt_gp->tg_pt_gp_ref_cnt);
if (rc)
break;
}
spin_unlock(&dev->t10_alua.tg_pt_gps_lock);
spin_lock(&lu_gp->lu_gp_lock);
- atomic_dec(&lu_gp_mem->lu_gp_mem_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&lu_gp_mem->lu_gp_mem_ref_cnt);
}
spin_unlock(&lu_gp->lu_gp_lock);
core_alua_dump_state(new_state));
}
- atomic_dec(&lu_gp->lu_gp_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&lu_gp->lu_gp_ref_cnt);
return rc;
}
DEF_DEV_ATTRIB(emulate_rest_reord);
SE_DEV_ATTR(emulate_rest_reord, S_IRUGO | S_IWUSR);
+DEF_DEV_ATTRIB(force_pr_aptpl);
+SE_DEV_ATTR(force_pr_aptpl, S_IRUGO | S_IWUSR);
+
DEF_DEV_ATTRIB_RO(hw_block_size);
SE_DEV_ATTR_RO(hw_block_size);
&target_core_dev_attrib_hw_pi_prot_type.attr,
&target_core_dev_attrib_pi_prot_format.attr,
&target_core_dev_attrib_enforce_pr_isids.attr,
+ &target_core_dev_attrib_force_pr_aptpl.attr,
&target_core_dev_attrib_is_nonrot.attr,
&target_core_dev_attrib_emulate_rest_reord.attr,
&target_core_dev_attrib_hw_block_size.attr,
{
unsigned char *i_fabric = NULL, *i_port = NULL, *isid = NULL;
unsigned char *t_fabric = NULL, *t_port = NULL;
- char *orig, *ptr, *arg_p, *opts;
+ char *orig, *ptr, *opts;
substring_t args[MAX_OPT_ARGS];
unsigned long long tmp_ll;
u64 sa_res_key = 0;
token = match_token(ptr, tokens, args);
switch (token) {
case Opt_initiator_fabric:
- i_fabric = match_strdup(&args[0]);
+ i_fabric = match_strdup(args);
if (!i_fabric) {
ret = -ENOMEM;
goto out;
}
break;
case Opt_initiator_node:
- i_port = match_strdup(&args[0]);
+ i_port = match_strdup(args);
if (!i_port) {
ret = -ENOMEM;
goto out;
}
break;
case Opt_initiator_sid:
- isid = match_strdup(&args[0]);
+ isid = match_strdup(args);
if (!isid) {
ret = -ENOMEM;
goto out;
}
break;
case Opt_sa_res_key:
- arg_p = match_strdup(&args[0]);
- if (!arg_p) {
- ret = -ENOMEM;
- goto out;
- }
- ret = kstrtoull(arg_p, 0, &tmp_ll);
+ ret = kstrtoull(args->from, 0, &tmp_ll);
if (ret < 0) {
- pr_err("kstrtoull() failed for"
- " sa_res_key=\n");
+ pr_err("kstrtoull() failed for sa_res_key=\n");
goto out;
}
sa_res_key = (u64)tmp_ll;
* PR APTPL Metadata for Target Port
*/
case Opt_target_fabric:
- t_fabric = match_strdup(&args[0]);
+ t_fabric = match_strdup(args);
if (!t_fabric) {
ret = -ENOMEM;
goto out;
}
break;
case Opt_target_node:
- t_port = match_strdup(&args[0]);
+ t_port = match_strdup(args);
if (!t_port) {
ret = -ENOMEM;
goto out;
if (port->sep_rtpi != rtpi)
continue;
- atomic_inc(&deve->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&deve->pr_ref_count);
spin_unlock_irq(&nacl->device_list_lock);
return deve;
return 0;
}
+int se_dev_set_force_pr_aptpl(struct se_device *dev, int flag)
+{
+ if ((flag != 0) && (flag != 1)) {
+ printk(KERN_ERR "Illegal value %d\n", flag);
+ return -EINVAL;
+ }
+ if (dev->export_count) {
+ pr_err("dev[%p]: Unable to set force_pr_aptpl while"
+ " export_count is %d\n", dev, dev->export_count);
+ return -EINVAL;
+ }
+
+ dev->dev_attrib.force_pr_aptpl = flag;
+ pr_debug("dev[%p]: SE Device force_pr_aptpl: %d\n", dev, flag);
+ return 0;
+}
+
int se_dev_set_is_nonrot(struct se_device *dev, int flag)
{
if ((flag != 0) && (flag != 1)) {
*
*
*/
-int core_dev_del_lun(
+void core_dev_del_lun(
struct se_portal_group *tpg,
- u32 unpacked_lun)
+ struct se_lun *lun)
{
- struct se_lun *lun;
-
- lun = core_tpg_pre_dellun(tpg, unpacked_lun);
- if (IS_ERR(lun))
- return PTR_ERR(lun);
-
- core_tpg_post_dellun(tpg, lun);
-
- pr_debug("%s_TPG[%u]_LUN[%u] - Deactivated %s Logical Unit from"
+ pr_debug("%s_TPG[%u]_LUN[%u] - Deactivating %s Logical Unit from"
" device object\n", tpg->se_tpg_tfo->get_fabric_name(),
- tpg->se_tpg_tfo->tpg_get_tag(tpg), unpacked_lun,
+ tpg->se_tpg_tfo->tpg_get_tag(tpg), lun->unpacked_lun,
tpg->se_tpg_tfo->get_fabric_name());
- return 0;
+ core_tpg_remove_lun(tpg, lun);
}
struct se_lun *core_get_lun_from_tpg(struct se_portal_group *tpg, u32 unpacked_lun)
spin_lock(&lun->lun_acl_lock);
list_add_tail(&lacl->lacl_list, &lun->lun_acl_list);
- atomic_inc(&lun->lun_acl_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&lun->lun_acl_count);
spin_unlock(&lun->lun_acl_lock);
pr_debug("%s_TPG[%hu]_LUN[%u->%u] - Added %s ACL for "
* Check to see if there are any existing persistent reservation APTPL
* pre-registrations that need to be enabled for this LUN ACL..
*/
- core_scsi3_check_aptpl_registration(lun->lun_se_dev, tpg, lun, lacl);
+ core_scsi3_check_aptpl_registration(lun->lun_se_dev, tpg, lun, nacl,
+ lacl->mapped_lun);
return 0;
}
spin_lock(&lun->lun_acl_lock);
list_del(&lacl->lacl_list);
- atomic_dec(&lun->lun_acl_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&lun->lun_acl_count);
spin_unlock(&lun->lun_acl_lock);
core_disable_device_list_for_node(lun, NULL, lacl->mapped_lun,
dev->dev_attrib.emulate_3pc = DA_EMULATE_3PC;
dev->dev_attrib.pi_prot_type = TARGET_DIF_TYPE0_PROT;
dev->dev_attrib.enforce_pr_isids = DA_ENFORCE_PR_ISIDS;
+ dev->dev_attrib.force_pr_aptpl = DA_FORCE_PR_APTPL;
dev->dev_attrib.is_nonrot = DA_IS_NONROT;
dev->dev_attrib.emulate_rest_reord = DA_EMULATE_REST_REORD;
dev->dev_attrib.max_unmap_lba_count = DA_MAX_UNMAP_LBA_COUNT;
struct se_node_acl, acl_group);
struct se_portal_group *se_tpg = se_nacl->se_tpg;
struct target_fabric_configfs *tf = se_tpg->se_tpg_wwn->wwn_tf;
- struct se_lun_acl *lacl;
+ struct se_lun_acl *lacl = NULL;
struct config_item *acl_ci;
struct config_group *lacl_cg = NULL, *ml_stat_grp = NULL;
char *buf;
out:
if (lacl_cg)
kfree(lacl_cg->default_groups);
+ kfree(lacl);
kfree(buf);
return ERR_PTR(ret);
}
tf->tf_ops.fabric_pre_unlink(se_tpg, lun);
}
- core_dev_del_lun(se_tpg, lun->unpacked_lun);
+ core_dev_del_lun(se_tpg, lun);
return 0;
}
GFP_KERNEL);
if (!port_stat_grp->default_groups) {
pr_err("Unable to allocate port_stat_grp->default_groups\n");
- errno = -ENOMEM;
- goto out;
+ kfree(lun_cg->default_groups);
+ return ERR_PTR(-ENOMEM);
}
target_stat_setup_port_default_groups(lun);
return &lun->lun_group;
-out:
- if (lun_cg)
- kfree(lun_cg->default_groups);
- return ERR_PTR(errno);
}
static void target_fabric_drop_lun(
* If the caller wants the TransportID Length, we set that value for the
* entire iSCSI Tarnsport ID now.
*/
- if (out_tid_len != NULL) {
- add_len = ((buf[2] >> 8) & 0xff);
- add_len |= (buf[3] & 0xff);
+ if (out_tid_len) {
+ /* The shift works thanks to integer promotion rules */
+ add_len = (buf[2] << 8) | buf[3];
tid_len = strlen(&buf[4]);
tid_len += 4; /* Add four bytes for iSCSI Transport ID header */
} else {
start = cmd->t_task_lba * dev->dev_attrib.block_size;
if (cmd->data_length)
- end = start + cmd->data_length;
+ end = start + cmd->data_length - 1;
else
end = LLONG_MAX;
}
struct fd_dev *fd_dev = FD_DEV(dev);
loff_t start = cmd->t_task_lba *
dev->dev_attrib.block_size;
- loff_t end = start + cmd->data_length;
+ loff_t end;
+
+ if (cmd->data_length)
+ end = start + cmd->data_length - 1;
+ else
+ end = LLONG_MAX;
vfs_fsync_range(fd_dev->fd_file, start, end, 1);
}
fd_dev->fbd_flags |= FBDF_HAS_SIZE;
break;
case Opt_fd_buffered_io:
- match_int(args, &arg);
+ ret = match_int(args, &arg);
+ if (ret)
+ goto out;
if (arg != 1) {
pr_err("bogus fd_buffered_io=%d value\n", arg);
ret = -EINVAL;
int se_dev_set_pi_prot_type(struct se_device *, int);
int se_dev_set_pi_prot_format(struct se_device *, int);
int se_dev_set_enforce_pr_isids(struct se_device *, int);
+int se_dev_set_force_pr_aptpl(struct se_device *, int);
int se_dev_set_is_nonrot(struct se_device *, int);
int se_dev_set_emulate_rest_reord(struct se_device *dev, int);
int se_dev_set_queue_depth(struct se_device *, u32);
int se_dev_set_optimal_sectors(struct se_device *, u32);
int se_dev_set_block_size(struct se_device *, u32);
struct se_lun *core_dev_add_lun(struct se_portal_group *, struct se_device *, u32);
-int core_dev_del_lun(struct se_portal_group *, u32);
+void core_dev_del_lun(struct se_portal_group *, struct se_lun *);
struct se_lun *core_get_lun_from_tpg(struct se_portal_group *, u32);
struct se_lun_acl *core_dev_init_initiator_node_lun_acl(struct se_portal_group *,
struct se_node_acl *, u32, int *);
struct se_lun *core_tpg_alloc_lun(struct se_portal_group *, u32);
int core_tpg_add_lun(struct se_portal_group *, struct se_lun *,
u32, struct se_device *);
-struct se_lun *core_tpg_pre_dellun(struct se_portal_group *, u32 unpacked_lun);
-int core_tpg_post_dellun(struct se_portal_group *, struct se_lun *);
+void core_tpg_remove_lun(struct se_portal_group *, struct se_lun *);
/* target_core_transport.c */
extern struct kmem_cache *se_tmr_req_cache;
*/
spin_lock(&dev->se_port_lock);
list_for_each_entry_safe(port, port_tmp, &dev->dev_sep_list, sep_list) {
- atomic_inc(&port->sep_tg_pt_ref_cnt);
- smp_mb__after_atomic();
+ atomic_inc_mb(&port->sep_tg_pt_ref_cnt);
spin_unlock(&dev->se_port_lock);
spin_lock_bh(&port->sep_alua_lock);
if (strcmp(nacl->initiatorname, nacl_tmp->initiatorname))
continue;
- atomic_inc(&deve_tmp->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&deve_tmp->pr_ref_count);
spin_unlock_bh(&port->sep_alua_lock);
/*
* Grab a configfs group dependency that is released
if (ret < 0) {
pr_err("core_scsi3_lunacl_depend"
"_item() failed\n");
- atomic_dec(&port->sep_tg_pt_ref_cnt);
- smp_mb__after_atomic();
- atomic_dec(&deve_tmp->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&port->sep_tg_pt_ref_cnt);
+ atomic_dec_mb(&deve_tmp->pr_ref_count);
goto out;
}
/*
nacl_tmp, deve_tmp, NULL,
sa_res_key, all_tg_pt, aptpl);
if (!pr_reg_atp) {
- atomic_dec(&port->sep_tg_pt_ref_cnt);
- smp_mb__after_atomic();
- atomic_dec(&deve_tmp->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&port->sep_tg_pt_ref_cnt);
+ atomic_dec_mb(&deve_tmp->pr_ref_count);
core_scsi3_lunacl_undepend_item(deve_tmp);
goto out;
}
spin_unlock_bh(&port->sep_alua_lock);
spin_lock(&dev->se_port_lock);
- atomic_dec(&port->sep_tg_pt_ref_cnt);
- smp_mb__after_atomic();
+ atomic_dec_mb(&port->sep_tg_pt_ref_cnt);
}
spin_unlock(&dev->se_port_lock);
spin_lock(&pr_tmpl->aptpl_reg_lock);
list_for_each_entry_safe(pr_reg, pr_reg_tmp, &pr_tmpl->aptpl_reg_list,
pr_reg_aptpl_list) {
+
if (!strcmp(pr_reg->pr_iport, i_port) &&
(pr_reg->pr_res_mapped_lun == deve->mapped_lun) &&
!(strcmp(pr_reg->pr_tport, t_port)) &&
struct se_device *dev,
struct se_portal_group *tpg,
struct se_lun *lun,
- struct se_lun_acl *lun_acl)
+ struct se_node_acl *nacl,
+ u32 mapped_lun)
{
- struct se_node_acl *nacl = lun_acl->se_lun_nacl;
- struct se_dev_entry *deve = nacl->device_list[lun_acl->mapped_lun];
+ struct se_dev_entry *deve = nacl->device_list[mapped_lun];
if (dev->dev_reservation_flags & DRF_SPC2_RESERVATIONS)
return 0;
if (dev->dev_attrib.enforce_pr_isids)
continue;
}
- atomic_inc(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_inc_mb(&pr_reg->pr_res_holders);
spin_unlock(&pr_tmpl->registration_lock);
return pr_reg;
}
if (strcmp(isid, pr_reg->pr_reg_isid))
continue;
- atomic_inc(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_inc_mb(&pr_reg->pr_res_holders);
spin_unlock(&pr_tmpl->registration_lock);
return pr_reg;
}
static void core_scsi3_put_pr_reg(struct t10_pr_registration *pr_reg)
{
- atomic_dec(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_dec_mb(&pr_reg->pr_res_holders);
}
static int core_scsi3_check_implicit_release(
configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&tpg->tpg_group.cg_item);
- atomic_dec(&tpg->tpg_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&tpg->tpg_pr_ref_count);
}
static int core_scsi3_nodeacl_depend_item(struct se_node_acl *nacl)
struct se_portal_group *tpg = nacl->se_tpg;
if (nacl->dynamic_node_acl) {
- atomic_dec(&nacl->acl_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&nacl->acl_pr_ref_count);
return;
}
configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&nacl->acl_group.cg_item);
- atomic_dec(&nacl->acl_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&nacl->acl_pr_ref_count);
}
static int core_scsi3_lunacl_depend_item(struct se_dev_entry *se_deve)
* For nacl->dynamic_node_acl=1
*/
if (!lun_acl) {
- atomic_dec(&se_deve->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&se_deve->pr_ref_count);
return;
}
nacl = lun_acl->se_lun_nacl;
configfs_undepend_item(tpg->se_tpg_tfo->tf_subsys,
&lun_acl->se_lun_group.cg_item);
- atomic_dec(&se_deve->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&se_deve->pr_ref_count);
}
static sense_reason_t
if (!i_str)
continue;
- atomic_inc(&tmp_tpg->tpg_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&tmp_tpg->tpg_pr_ref_count);
spin_unlock(&dev->se_port_lock);
if (core_scsi3_tpg_depend_item(tmp_tpg)) {
pr_err(" core_scsi3_tpg_depend_item()"
" for tmp_tpg\n");
- atomic_dec(&tmp_tpg->tpg_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&tmp_tpg->tpg_pr_ref_count);
ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
goto out_unmap;
}
spin_lock_irq(&tmp_tpg->acl_node_lock);
dest_node_acl = __core_tpg_get_initiator_node_acl(
tmp_tpg, i_str);
- if (dest_node_acl) {
- atomic_inc(&dest_node_acl->acl_pr_ref_count);
- smp_mb__after_atomic();
- }
+ if (dest_node_acl)
+ atomic_inc_mb(&dest_node_acl->acl_pr_ref_count);
spin_unlock_irq(&tmp_tpg->acl_node_lock);
if (!dest_node_acl) {
if (core_scsi3_nodeacl_depend_item(dest_node_acl)) {
pr_err("configfs_depend_item() failed"
" for dest_node_acl->acl_group\n");
- atomic_dec(&dest_node_acl->acl_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dest_node_acl->acl_pr_ref_count);
core_scsi3_tpg_undepend_item(tmp_tpg);
ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
goto out_unmap;
if (core_scsi3_lunacl_depend_item(dest_se_deve)) {
pr_err("core_scsi3_lunacl_depend_item()"
" failed\n");
- atomic_dec(&dest_se_deve->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dest_se_deve->pr_ref_count);
core_scsi3_nodeacl_undepend_item(dest_node_acl);
core_scsi3_tpg_undepend_item(dest_tpg);
ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
if (!dest_tf_ops)
continue;
- atomic_inc(&dest_se_tpg->tpg_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&dest_se_tpg->tpg_pr_ref_count);
spin_unlock(&dev->se_port_lock);
if (core_scsi3_tpg_depend_item(dest_se_tpg)) {
pr_err("core_scsi3_tpg_depend_item() failed"
" for dest_se_tpg\n");
- atomic_dec(&dest_se_tpg->tpg_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dest_se_tpg->tpg_pr_ref_count);
ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
goto out_put_pr_reg;
}
spin_lock_irq(&dest_se_tpg->acl_node_lock);
dest_node_acl = __core_tpg_get_initiator_node_acl(dest_se_tpg,
initiator_str);
- if (dest_node_acl) {
- atomic_inc(&dest_node_acl->acl_pr_ref_count);
- smp_mb__after_atomic();
- }
+ if (dest_node_acl)
+ atomic_inc_mb(&dest_node_acl->acl_pr_ref_count);
spin_unlock_irq(&dest_se_tpg->acl_node_lock);
if (!dest_node_acl) {
if (core_scsi3_nodeacl_depend_item(dest_node_acl)) {
pr_err("core_scsi3_nodeacl_depend_item() for"
" dest_node_acl\n");
- atomic_dec(&dest_node_acl->acl_pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dest_node_acl->acl_pr_ref_count);
dest_node_acl = NULL;
ret = TCM_INVALID_PARAMETER_LIST;
goto out;
if (core_scsi3_lunacl_depend_item(dest_se_deve)) {
pr_err("core_scsi3_lunacl_depend_item() failed\n");
- atomic_dec(&dest_se_deve->pr_ref_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dest_se_deve->pr_ref_count);
dest_se_deve = NULL;
ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
goto out;
sense_reason_t
target_scsi3_emulate_pr_out(struct se_cmd *cmd)
{
+ struct se_device *dev = cmd->se_dev;
unsigned char *cdb = &cmd->t_task_cdb[0];
unsigned char *buf;
u64 res_key, sa_res_key;
aptpl = (buf[17] & 0x01);
unreg = (buf[17] & 0x02);
}
+ /*
+ * If the backend device has been configured to force APTPL metadata
+ * write-out, go ahead and propigate aptpl=1 down now.
+ */
+ if (dev->dev_attrib.force_pr_aptpl)
+ aptpl = 1;
+
transport_kunmap_data_sg(cmd);
buf = NULL;
if (!buf)
return TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
- buf[0] = ((add_len << 8) & 0xff);
+ buf[0] = ((add_len >> 8) & 0xff);
buf[1] = (add_len & 0xff);
buf[2] |= 0x10; /* CRH: Compatible Reservation Hanlding bit. */
buf[2] |= 0x08; /* SIP_C: Specify Initiator Ports Capable bit */
se_tpg = pr_reg->pr_reg_nacl->se_tpg;
add_desc_len = 0;
- atomic_inc(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_inc_mb(&pr_reg->pr_res_holders);
spin_unlock(&pr_tmpl->registration_lock);
/*
* Determine expected length of $FABRIC_MOD specific
pr_warn("SPC-3 PRIN READ_FULL_STATUS ran"
" out of buffer: %d\n", cmd->data_length);
spin_lock(&pr_tmpl->registration_lock);
- atomic_dec(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_dec_mb(&pr_reg->pr_res_holders);
break;
}
/*
se_nacl, pr_reg, &format_code, &buf[off+4]);
spin_lock(&pr_tmpl->registration_lock);
- atomic_dec(&pr_reg->pr_res_holders);
- smp_mb__after_atomic();
+ atomic_dec_mb(&pr_reg->pr_res_holders);
/*
* Set the ADDITIONAL DESCRIPTOR LENGTH
*/
unsigned char *, u16, u32, int, int, u8);
extern int core_scsi3_check_aptpl_registration(struct se_device *,
struct se_portal_group *, struct se_lun *,
- struct se_lun_acl *);
+ struct se_node_acl *, u32);
extern void core_scsi3_free_pr_reg_from_nacl(struct se_device *,
struct se_node_acl *);
extern void core_scsi3_free_all_registrations(struct se_device *);
ret = -EINVAL;
goto out;
}
- match_int(args, &arg);
+ ret = match_int(args, &arg);
+ if (ret)
+ goto out;
pdv->pdv_host_id = arg;
pr_debug("PSCSI[%d]: Referencing SCSI Host ID:"
" %d\n", phv->phv_host_id, pdv->pdv_host_id);
pdv->pdv_flags |= PDF_HAS_VIRT_HOST_ID;
break;
case Opt_scsi_channel_id:
- match_int(args, &arg);
+ ret = match_int(args, &arg);
+ if (ret)
+ goto out;
pdv->pdv_channel_id = arg;
pr_debug("PSCSI[%d]: Referencing SCSI Channel"
" ID: %d\n", phv->phv_host_id,
pdv->pdv_flags |= PDF_HAS_CHANNEL_ID;
break;
case Opt_scsi_target_id:
- match_int(args, &arg);
+ ret = match_int(args, &arg);
+ if (ret)
+ goto out;
pdv->pdv_target_id = arg;
pr_debug("PSCSI[%d]: Referencing SCSI Target"
" ID: %d\n", phv->phv_host_id,
pdv->pdv_flags |= PDF_HAS_TARGET_ID;
break;
case Opt_scsi_lun_id:
- match_int(args, &arg);
+ ret = match_int(args, &arg);
+ if (ret)
+ goto out;
pdv->pdv_lun_id = arg;
pr_debug("PSCSI[%d]: Referencing SCSI LUN ID:"
" %d\n", phv->phv_host_id, pdv->pdv_lun_id);
}
/* reject any command that we don't have a handler for */
- if (!(cmd->se_cmd_flags & SCF_SCSI_DATA_CDB) && !cmd->execute_cmd)
+ if (!cmd->execute_cmd)
return TCM_UNSUPPORTED_SCSI_OPCODE;
if (cmd->se_cmd_flags & SCF_SCSI_DATA_CDB) {
}
EXPORT_SYMBOL(core_tmr_alloc_req);
-void core_tmr_release_req(
- struct se_tmr_req *tmr)
+void core_tmr_release_req(struct se_tmr_req *tmr)
{
struct se_device *dev = tmr->tmr_dev;
unsigned long flags;
- if (!dev) {
- kfree(tmr);
- return;
+ if (dev) {
+ spin_lock_irqsave(&dev->se_tmr_lock, flags);
+ list_del(&tmr->tmr_list);
+ spin_unlock_irqrestore(&dev->se_tmr_lock, flags);
}
- spin_lock_irqsave(&dev->se_tmr_lock, flags);
- list_del(&tmr->tmr_list);
- spin_unlock_irqrestore(&dev->se_tmr_lock, flags);
-
kfree(tmr);
}
bool remove = true;
/*
* TASK ABORTED status (TAS) bit support
- */
- if ((tmr_nacl &&
- (tmr_nacl != cmd->se_sess->se_node_acl)) && tas) {
+ */
+ if ((tmr_nacl && (tmr_nacl != cmd->se_sess->se_node_acl)) && tas) {
remove = false;
transport_send_task_abort(cmd);
}
struct se_tmr_req *tmr,
struct se_session *se_sess)
{
- struct se_cmd *se_cmd, *tmp_cmd;
+ struct se_cmd *se_cmd;
unsigned long flags;
int ref_tag;
spin_lock_irqsave(&se_sess->sess_cmd_lock, flags);
- list_for_each_entry_safe(se_cmd, tmp_cmd,
- &se_sess->sess_cmd_list, se_cmd_list) {
+ list_for_each_entry(se_cmd, &se_sess->sess_cmd_list, se_cmd_list) {
if (dev != se_cmd->se_dev)
continue;
#include <target/target_core_fabric.h>
#include "target_core_internal.h"
+#include "target_core_pr.h"
extern struct se_device *g_lun0_dev;
core_enable_device_list_for_node(lun, NULL, lun->unpacked_lun,
lun_access, acl, tpg);
+ /*
+ * Check to see if there are any existing persistent reservation
+ * APTPL pre-registrations that need to be enabled for this dynamic
+ * LUN ACL now..
+ */
+ core_scsi3_check_aptpl_registration(dev, tpg, lun, acl,
+ lun->unpacked_lun);
spin_lock(&tpg->tpg_lun_lock);
}
spin_unlock(&tpg->tpg_lun_lock);
continue;
spin_unlock(&tpg->tpg_lun_lock);
- core_dev_del_lun(tpg, lun->unpacked_lun);
+ core_dev_del_lun(tpg, lun);
spin_lock(&tpg->tpg_lun_lock);
}
spin_unlock(&tpg->tpg_lun_lock);
return 0;
}
-static void core_tpg_release_virtual_lun0(struct se_portal_group *se_tpg)
-{
- struct se_lun *lun = &se_tpg->tpg_virt_lun0;
-
- core_tpg_post_dellun(se_tpg, lun);
-}
-
int core_tpg_register(
struct target_core_fabric_ops *tfo,
struct se_wwn *se_wwn,
spin_unlock_irq(&se_tpg->acl_node_lock);
if (se_tpg->se_tpg_type == TRANSPORT_TPG_TYPE_NORMAL)
- core_tpg_release_virtual_lun0(se_tpg);
+ core_tpg_remove_lun(se_tpg, &se_tpg->tpg_virt_lun0);
se_tpg->se_tpg_fabric_ptr = NULL;
array_free(se_tpg->tpg_lun_list, TRANSPORT_MAX_LUNS_PER_TPG);
return 0;
}
-struct se_lun *core_tpg_pre_dellun(
- struct se_portal_group *tpg,
- u32 unpacked_lun)
-{
- struct se_lun *lun;
-
- if (unpacked_lun > (TRANSPORT_MAX_LUNS_PER_TPG-1)) {
- pr_err("%s LUN: %u exceeds TRANSPORT_MAX_LUNS_PER_TPG"
- "-1: %u for Target Portal Group: %u\n",
- tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
- TRANSPORT_MAX_LUNS_PER_TPG-1,
- tpg->se_tpg_tfo->tpg_get_tag(tpg));
- return ERR_PTR(-EOVERFLOW);
- }
-
- spin_lock(&tpg->tpg_lun_lock);
- lun = tpg->tpg_lun_list[unpacked_lun];
- if (lun->lun_status != TRANSPORT_LUN_STATUS_ACTIVE) {
- pr_err("%s Logical Unit Number: %u is not active on"
- " Target Portal Group: %u, ignoring request.\n",
- tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
- tpg->se_tpg_tfo->tpg_get_tag(tpg));
- spin_unlock(&tpg->tpg_lun_lock);
- return ERR_PTR(-ENODEV);
- }
- spin_unlock(&tpg->tpg_lun_lock);
-
- return lun;
-}
-
-int core_tpg_post_dellun(
+void core_tpg_remove_lun(
struct se_portal_group *tpg,
struct se_lun *lun)
{
spin_unlock(&tpg->tpg_lun_lock);
percpu_ref_exit(&lun->lun_ref);
-
- return 0;
}
if (ret != 0)
pr_err("Unable to load target_core_pscsi\n");
+ ret = request_module("target_core_user");
+ if (ret != 0)
+ pr_err("Unable to load target_core_user\n");
+
sub_api_initialized = 1;
}
list_for_each_entry_safe(cmd, cmd_tmp, &qf_cmd_list, se_qf_node) {
list_del(&cmd->se_qf_node);
- atomic_dec(&dev->dev_qf_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dev->dev_qf_count);
pr_debug("Processing %s cmd: %p QUEUE_FULL in work queue"
" context: %s\n", cmd->se_tfo->get_fabric_name(), cmd,
* Dormant to Active status.
*/
cmd->se_ordered_id = atomic_inc_return(&dev->dev_ordered_id);
- smp_mb__after_atomic();
pr_debug("Allocated se_ordered_id: %u for Task Attr: 0x%02x on %s\n",
cmd->se_ordered_id, cmd->sam_task_attr,
dev->transport->name);
cmd->t_task_cdb[0], cmd->se_ordered_id);
return false;
case MSG_ORDERED_TAG:
- atomic_inc(&dev->dev_ordered_sync);
- smp_mb__after_atomic();
+ atomic_inc_mb(&dev->dev_ordered_sync);
pr_debug("Added ORDERED for CDB: 0x%02x to ordered list, "
" se_ordered_id: %u\n",
/*
* For SIMPLE and UNTAGGED Task Attribute commands
*/
- atomic_inc(&dev->simple_cmds);
- smp_mb__after_atomic();
+ atomic_inc_mb(&dev->simple_cmds);
break;
}
return;
if (cmd->sam_task_attr == MSG_SIMPLE_TAG) {
- atomic_dec(&dev->simple_cmds);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dev->simple_cmds);
dev->dev_cur_ordered_id++;
pr_debug("Incremented dev->dev_cur_ordered_id: %u for"
" SIMPLE: %u\n", dev->dev_cur_ordered_id,
" HEAD_OF_QUEUE: %u\n", dev->dev_cur_ordered_id,
cmd->se_ordered_id);
} else if (cmd->sam_task_attr == MSG_ORDERED_TAG) {
- atomic_dec(&dev->dev_ordered_sync);
- smp_mb__after_atomic();
+ atomic_dec_mb(&dev->dev_ordered_sync);
dev->dev_cur_ordered_id++;
pr_debug("Incremented dev_cur_ordered_id: %u for ORDERED:"
if (cmd->se_cmd_flags & SCF_TRANSPORT_TASK_SENSE) {
trace_target_cmd_complete(cmd);
ret = cmd->se_tfo->queue_status(cmd);
- if (ret)
- goto out;
+ goto out;
}
switch (cmd->data_direction) {
{
spin_lock_irq(&dev->qf_cmd_lock);
list_add_tail(&cmd->se_qf_node, &cmd->se_dev->qf_cmd_list);
- atomic_inc(&dev->dev_qf_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&dev->dev_qf_count);
spin_unlock_irq(&cmd->se_dev->qf_cmd_lock);
schedule_work(&cmd->se_dev->qf_work_queue);
if (cmd->se_tfo->write_pending_status(cmd) != 0) {
cmd->transport_state |= CMD_T_ABORTED;
cmd->se_cmd_flags |= SCF_SEND_DELAYED_TAS;
- smp_mb__after_atomic();
return;
}
}
spin_unlock(&deve->ua_lock);
spin_unlock_irq(&nacl->device_list_lock);
- atomic_inc(&deve->ua_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&deve->ua_count);
return 0;
}
list_add_tail(&ua->ua_nacl_list, &deve->ua_list);
nacl->se_tpg->se_tpg_tfo->get_fabric_name(), unpacked_lun,
asc, ascq);
- atomic_inc(&deve->ua_count);
- smp_mb__after_atomic();
+ atomic_inc_mb(&deve->ua_count);
return 0;
}
list_del(&ua->ua_nacl_list);
kmem_cache_free(se_ua_cache, ua);
- atomic_dec(&deve->ua_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&deve->ua_count);
}
spin_unlock(&deve->ua_lock);
}
list_del(&ua->ua_nacl_list);
kmem_cache_free(se_ua_cache, ua);
- atomic_dec(&deve->ua_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&deve->ua_count);
}
spin_unlock(&deve->ua_lock);
spin_unlock_irq(&nacl->device_list_lock);
list_del(&ua->ua_nacl_list);
kmem_cache_free(se_ua_cache, ua);
- atomic_dec(&deve->ua_count);
- smp_mb__after_atomic();
+ atomic_dec_mb(&deve->ua_count);
}
spin_unlock(&deve->ua_lock);
spin_unlock_irq(&nacl->device_list_lock);
#ifndef TARGET_CORE_UA_H
+#define TARGET_CORE_UA_H
/*
* From spc4r17, Table D.1: ASC and ASCQ Assignement
--- /dev/null
+/*
+ * Copyright (C) 2013 Shaohua Li <shli@kernel.org>
+ * Copyright (C) 2014 Red Hat, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ */
+
+#include <linux/spinlock.h>
+#include <linux/module.h>
+#include <linux/idr.h>
+#include <linux/timer.h>
+#include <linux/parser.h>
+#include <scsi/scsi.h>
+#include <scsi/scsi_host.h>
+#include <linux/uio_driver.h>
+#include <net/genetlink.h>
+#include <target/target_core_base.h>
+#include <target/target_core_fabric.h>
+#include <target/target_core_backend.h>
+#include <linux/target_core_user.h>
+
+/*
+ * Define a shared-memory interface for LIO to pass SCSI commands and
+ * data to userspace for processing. This is to allow backends that
+ * are too complex for in-kernel support to be possible.
+ *
+ * It uses the UIO framework to do a lot of the device-creation and
+ * introspection work for us.
+ *
+ * See the .h file for how the ring is laid out. Note that while the
+ * command ring is defined, the particulars of the data area are
+ * not. Offset values in the command entry point to other locations
+ * internal to the mmap()ed area. There is separate space outside the
+ * command ring for data buffers. This leaves maximum flexibility for
+ * moving buffer allocations, or even page flipping or other
+ * allocation techniques, without altering the command ring layout.
+ *
+ * SECURITY:
+ * The user process must be assumed to be malicious. There's no way to
+ * prevent it breaking the command ring protocol if it wants, but in
+ * order to prevent other issues we must only ever read *data* from
+ * the shared memory area, not offsets or sizes. This applies to
+ * command ring entries as well as the mailbox. Extra code needed for
+ * this may have a 'UAM' comment.
+ */
+
+
+#define TCMU_TIME_OUT (30 * MSEC_PER_SEC)
+
+#define CMDR_SIZE (16 * 4096)
+#define DATA_SIZE (257 * 4096)
+
+#define TCMU_RING_SIZE (CMDR_SIZE + DATA_SIZE)
+
+static struct device *tcmu_root_device;
+
+struct tcmu_hba {
+ u32 host_id;
+};
+
+/* User wants all cmds or just some */
+enum passthru_level {
+ TCMU_PASS_ALL = 0,
+ TCMU_PASS_IO,
+ TCMU_PASS_INVALID,
+};
+
+#define TCMU_CONFIG_LEN 256
+
+struct tcmu_dev {
+ struct se_device se_dev;
+
+ char *name;
+ struct se_hba *hba;
+
+#define TCMU_DEV_BIT_OPEN 0
+#define TCMU_DEV_BIT_BROKEN 1
+ unsigned long flags;
+ enum passthru_level pass_level;
+
+ struct uio_info uio_info;
+
+ struct tcmu_mailbox *mb_addr;
+ size_t dev_size;
+ u32 cmdr_size;
+ u32 cmdr_last_cleaned;
+ /* Offset of data ring from start of mb */
+ size_t data_off;
+ size_t data_size;
+ /* Ring head + tail values. */
+ /* Must add data_off and mb_addr to get the address */
+ size_t data_head;
+ size_t data_tail;
+
+ wait_queue_head_t wait_cmdr;
+ /* TODO should this be a mutex? */
+ spinlock_t cmdr_lock;
+
+ struct idr commands;
+ spinlock_t commands_lock;
+
+ struct timer_list timeout;
+
+ char dev_config[TCMU_CONFIG_LEN];
+};
+
+#define TCMU_DEV(_se_dev) container_of(_se_dev, struct tcmu_dev, se_dev)
+
+#define CMDR_OFF sizeof(struct tcmu_mailbox)
+
+struct tcmu_cmd {
+ struct se_cmd *se_cmd;
+ struct tcmu_dev *tcmu_dev;
+
+ uint16_t cmd_id;
+
+ /* Can't use se_cmd->data_length when cleaning up expired cmds, because if
+ cmd has been completed then accessing se_cmd is off limits */
+ size_t data_length;
+
+ unsigned long deadline;
+
+#define TCMU_CMD_BIT_EXPIRED 0
+ unsigned long flags;
+};
+
+static struct kmem_cache *tcmu_cmd_cache;
+
+/* multicast group */
+enum tcmu_multicast_groups {
+ TCMU_MCGRP_CONFIG,
+};
+
+static const struct genl_multicast_group tcmu_mcgrps[] = {
+ [TCMU_MCGRP_CONFIG] = { .name = "config", },
+};
+
+/* Our generic netlink family */
+static struct genl_family tcmu_genl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = 0,
+ .name = "TCM-USER",
+ .version = 1,
+ .maxattr = TCMU_ATTR_MAX,
+ .mcgrps = tcmu_mcgrps,
+ .n_mcgrps = ARRAY_SIZE(tcmu_mcgrps),
+};
+
+static struct tcmu_cmd *tcmu_alloc_cmd(struct se_cmd *se_cmd)
+{
+ struct se_device *se_dev = se_cmd->se_dev;
+ struct tcmu_dev *udev = TCMU_DEV(se_dev);
+ struct tcmu_cmd *tcmu_cmd;
+ int cmd_id;
+
+ tcmu_cmd = kmem_cache_zalloc(tcmu_cmd_cache, GFP_KERNEL);
+ if (!tcmu_cmd)
+ return NULL;
+
+ tcmu_cmd->se_cmd = se_cmd;
+ tcmu_cmd->tcmu_dev = udev;
+ tcmu_cmd->data_length = se_cmd->data_length;
+
+ tcmu_cmd->deadline = jiffies + msecs_to_jiffies(TCMU_TIME_OUT);
+
+ idr_preload(GFP_KERNEL);
+ spin_lock_irq(&udev->commands_lock);
+ cmd_id = idr_alloc(&udev->commands, tcmu_cmd, 0,
+ USHRT_MAX, GFP_NOWAIT);
+ spin_unlock_irq(&udev->commands_lock);
+ idr_preload_end();
+
+ if (cmd_id < 0) {
+ kmem_cache_free(tcmu_cmd_cache, tcmu_cmd);
+ return NULL;
+ }
+ tcmu_cmd->cmd_id = cmd_id;
+
+ return tcmu_cmd;
+}
+
+static inline void tcmu_flush_dcache_range(void *vaddr, size_t size)
+{
+ unsigned long offset = (unsigned long) vaddr & ~PAGE_MASK;
+
+ size = round_up(size+offset, PAGE_SIZE);
+ vaddr -= offset;
+
+ while (size) {
+ flush_dcache_page(virt_to_page(vaddr));
+ size -= PAGE_SIZE;
+ }
+}
+
+/*
+ * Some ring helper functions. We don't assume size is a power of 2 so
+ * we can't use circ_buf.h.
+ */
+static inline size_t spc_used(size_t head, size_t tail, size_t size)
+{
+ int diff = head - tail;
+
+ if (diff >= 0)
+ return diff;
+ else
+ return size + diff;
+}
+
+static inline size_t spc_free(size_t head, size_t tail, size_t size)
+{
+ /* Keep 1 byte unused or we can't tell full from empty */
+ return (size - spc_used(head, tail, size) - 1);
+}
+
+static inline size_t head_to_end(size_t head, size_t size)
+{
+ return size - head;
+}
+
+#define UPDATE_HEAD(head, used, size) smp_store_release(&head, ((head % size) + used) % size)
+
+/*
+ * We can't queue a command until we have space available on the cmd ring *and* space
+ * space avail on the data ring.
+ *
+ * Called with ring lock held.
+ */
+static bool is_ring_space_avail(struct tcmu_dev *udev, size_t cmd_size, size_t data_needed)
+{
+ struct tcmu_mailbox *mb = udev->mb_addr;
+ size_t space;
+ u32 cmd_head;
+ size_t cmd_needed;
+
+ tcmu_flush_dcache_range(mb, sizeof(*mb));
+
+ cmd_head = mb->cmd_head % udev->cmdr_size; /* UAM */
+
+ /*
+ * If cmd end-of-ring space is too small then we need space for a NOP plus
+ * original cmd - cmds are internally contiguous.
+ */
+ if (head_to_end(cmd_head, udev->cmdr_size) >= cmd_size)
+ cmd_needed = cmd_size;
+ else
+ cmd_needed = cmd_size + head_to_end(cmd_head, udev->cmdr_size);
+
+ space = spc_free(cmd_head, udev->cmdr_last_cleaned, udev->cmdr_size);
+ if (space < cmd_needed) {
+ pr_debug("no cmd space: %u %u %u\n", cmd_head,
+ udev->cmdr_last_cleaned, udev->cmdr_size);
+ return false;
+ }
+
+ space = spc_free(udev->data_head, udev->data_tail, udev->data_size);
+ if (space < data_needed) {
+ pr_debug("no data space: %zu %zu %zu\n", udev->data_head,
+ udev->data_tail, udev->data_size);
+ return false;
+ }
+
+ return true;
+}
+
+static int tcmu_queue_cmd_ring(struct tcmu_cmd *tcmu_cmd)
+{
+ struct tcmu_dev *udev = tcmu_cmd->tcmu_dev;
+ struct se_cmd *se_cmd = tcmu_cmd->se_cmd;
+ size_t base_command_size, command_size;
+ struct tcmu_mailbox *mb;
+ struct tcmu_cmd_entry *entry;
+ int i;
+ struct scatterlist *sg;
+ struct iovec *iov;
+ int iov_cnt = 0;
+ uint32_t cmd_head;
+ uint64_t cdb_off;
+
+ if (test_bit(TCMU_DEV_BIT_BROKEN, &udev->flags))
+ return -EINVAL;
+
+ /*
+ * Must be a certain minimum size for response sense info, but
+ * also may be larger if the iov array is large.
+ *
+ * iovs = sgl_nents+1, for end-of-ring case, plus another 1
+ * b/c size == offsetof one-past-element.
+ */
+ base_command_size = max(offsetof(struct tcmu_cmd_entry,
+ req.iov[se_cmd->t_data_nents + 2]),
+ sizeof(struct tcmu_cmd_entry));
+ command_size = base_command_size
+ + round_up(scsi_command_size(se_cmd->t_task_cdb), TCMU_OP_ALIGN_SIZE);
+
+ WARN_ON(command_size & (TCMU_OP_ALIGN_SIZE-1));
+
+ spin_lock_irq(&udev->cmdr_lock);
+
+ mb = udev->mb_addr;
+ cmd_head = mb->cmd_head % udev->cmdr_size; /* UAM */
+ if ((command_size > (udev->cmdr_size / 2))
+ || tcmu_cmd->data_length > (udev->data_size - 1))
+ pr_warn("TCMU: Request of size %zu/%zu may be too big for %u/%zu "
+ "cmd/data ring buffers\n", command_size, tcmu_cmd->data_length,
+ udev->cmdr_size, udev->data_size);
+
+ while (!is_ring_space_avail(udev, command_size, tcmu_cmd->data_length)) {
+ int ret;
+ DEFINE_WAIT(__wait);
+
+ prepare_to_wait(&udev->wait_cmdr, &__wait, TASK_INTERRUPTIBLE);
+
+ pr_debug("sleeping for ring space\n");
+ spin_unlock_irq(&udev->cmdr_lock);
+ ret = schedule_timeout(msecs_to_jiffies(TCMU_TIME_OUT));
+ finish_wait(&udev->wait_cmdr, &__wait);
+ if (!ret) {
+ pr_warn("tcmu: command timed out\n");
+ return -ETIMEDOUT;
+ }
+
+ spin_lock_irq(&udev->cmdr_lock);
+
+ /* We dropped cmdr_lock, cmd_head is stale */
+ cmd_head = mb->cmd_head % udev->cmdr_size; /* UAM */
+ }
+
+ /* Insert a PAD if end-of-ring space is too small */
+ if (head_to_end(cmd_head, udev->cmdr_size) < command_size) {
+ size_t pad_size = head_to_end(cmd_head, udev->cmdr_size);
+
+ entry = (void *) mb + CMDR_OFF + cmd_head;
+ tcmu_flush_dcache_range(entry, sizeof(*entry));
+ tcmu_hdr_set_op(&entry->hdr, TCMU_OP_PAD);
+ tcmu_hdr_set_len(&entry->hdr, pad_size);
+
+ UPDATE_HEAD(mb->cmd_head, pad_size, udev->cmdr_size);
+
+ cmd_head = mb->cmd_head % udev->cmdr_size; /* UAM */
+ WARN_ON(cmd_head != 0);
+ }
+
+ entry = (void *) mb + CMDR_OFF + cmd_head;
+ tcmu_flush_dcache_range(entry, sizeof(*entry));
+ tcmu_hdr_set_op(&entry->hdr, TCMU_OP_CMD);
+ tcmu_hdr_set_len(&entry->hdr, command_size);
+ entry->cmd_id = tcmu_cmd->cmd_id;
+
+ /*
+ * Fix up iovecs, and handle if allocation in data ring wrapped.
+ */
+ iov = &entry->req.iov[0];
+ for_each_sg(se_cmd->t_data_sg, sg, se_cmd->t_data_nents, i) {
+ size_t copy_bytes = min((size_t)sg->length,
+ head_to_end(udev->data_head, udev->data_size));
+ void *from = kmap_atomic(sg_page(sg)) + sg->offset;
+ void *to = (void *) mb + udev->data_off + udev->data_head;
+
+ if (tcmu_cmd->se_cmd->data_direction == DMA_TO_DEVICE) {
+ memcpy(to, from, copy_bytes);
+ tcmu_flush_dcache_range(to, copy_bytes);
+ }
+
+ /* Even iov_base is relative to mb_addr */
+ iov->iov_len = copy_bytes;
+ iov->iov_base = (void *) udev->data_off + udev->data_head;
+ iov_cnt++;
+ iov++;
+
+ UPDATE_HEAD(udev->data_head, copy_bytes, udev->data_size);
+
+ /* Uh oh, we wrapped the buffer. Must split sg across 2 iovs. */
+ if (sg->length != copy_bytes) {
+ from += copy_bytes;
+ copy_bytes = sg->length - copy_bytes;
+
+ iov->iov_len = copy_bytes;
+ iov->iov_base = (void *) udev->data_off + udev->data_head;
+
+ if (se_cmd->data_direction == DMA_TO_DEVICE) {
+ to = (void *) mb + udev->data_off + udev->data_head;
+ memcpy(to, from, copy_bytes);
+ tcmu_flush_dcache_range(to, copy_bytes);
+ }
+
+ iov_cnt++;
+ iov++;
+
+ UPDATE_HEAD(udev->data_head, copy_bytes, udev->data_size);
+ }
+
+ kunmap_atomic(from);
+ }
+ entry->req.iov_cnt = iov_cnt;
+
+ /* All offsets relative to mb_addr, not start of entry! */
+ cdb_off = CMDR_OFF + cmd_head + base_command_size;
+ memcpy((void *) mb + cdb_off, se_cmd->t_task_cdb, scsi_command_size(se_cmd->t_task_cdb));
+ entry->req.cdb_off = cdb_off;
+ tcmu_flush_dcache_range(entry, sizeof(*entry));
+
+ UPDATE_HEAD(mb->cmd_head, command_size, udev->cmdr_size);
+ tcmu_flush_dcache_range(mb, sizeof(*mb));
+
+ spin_unlock_irq(&udev->cmdr_lock);
+
+ /* TODO: only if FLUSH and FUA? */
+ uio_event_notify(&udev->uio_info);
+
+ mod_timer(&udev->timeout,
+ round_jiffies_up(jiffies + msecs_to_jiffies(TCMU_TIME_OUT)));
+
+ return 0;
+}
+
+static int tcmu_queue_cmd(struct se_cmd *se_cmd)
+{
+ struct se_device *se_dev = se_cmd->se_dev;
+ struct tcmu_dev *udev = TCMU_DEV(se_dev);
+ struct tcmu_cmd *tcmu_cmd;
+ int ret;
+
+ tcmu_cmd = tcmu_alloc_cmd(se_cmd);
+ if (!tcmu_cmd)
+ return -ENOMEM;
+
+ ret = tcmu_queue_cmd_ring(tcmu_cmd);
+ if (ret < 0) {
+ pr_err("TCMU: Could not queue command\n");
+ spin_lock_irq(&udev->commands_lock);
+ idr_remove(&udev->commands, tcmu_cmd->cmd_id);
+ spin_unlock_irq(&udev->commands_lock);
+
+ kmem_cache_free(tcmu_cmd_cache, tcmu_cmd);
+ }
+
+ return ret;
+}
+
+static void tcmu_handle_completion(struct tcmu_cmd *cmd, struct tcmu_cmd_entry *entry)
+{
+ struct se_cmd *se_cmd = cmd->se_cmd;
+ struct tcmu_dev *udev = cmd->tcmu_dev;
+
+ if (test_bit(TCMU_CMD_BIT_EXPIRED, &cmd->flags)) {
+ /* cmd has been completed already from timeout, just reclaim data
+ ring space */
+ UPDATE_HEAD(udev->data_tail, cmd->data_length, udev->data_size);
+ return;
+ }
+
+ if (entry->rsp.scsi_status == SAM_STAT_CHECK_CONDITION) {
+ memcpy(se_cmd->sense_buffer, entry->rsp.sense_buffer,
+ se_cmd->scsi_sense_length);
+
+ UPDATE_HEAD(udev->data_tail, cmd->data_length, udev->data_size);
+ }
+ else if (se_cmd->data_direction == DMA_FROM_DEVICE) {
+ struct scatterlist *sg;
+ int i;
+
+ /* It'd be easier to look at entry's iovec again, but UAM */
+ for_each_sg(se_cmd->t_data_sg, sg, se_cmd->t_data_nents, i) {
+ size_t copy_bytes;
+ void *to;
+ void *from;
+
+ copy_bytes = min((size_t)sg->length,
+ head_to_end(udev->data_tail, udev->data_size));
+
+ to = kmap_atomic(sg_page(sg)) + sg->offset;
+ WARN_ON(sg->length + sg->offset > PAGE_SIZE);
+ from = (void *) udev->mb_addr + udev->data_off + udev->data_tail;
+ tcmu_flush_dcache_range(from, copy_bytes);
+ memcpy(to, from, copy_bytes);
+
+ UPDATE_HEAD(udev->data_tail, copy_bytes, udev->data_size);
+
+ /* Uh oh, wrapped the data buffer for this sg's data */
+ if (sg->length != copy_bytes) {
+ from = (void *) udev->mb_addr + udev->data_off + udev->data_tail;
+ WARN_ON(udev->data_tail);
+ to += copy_bytes;
+ copy_bytes = sg->length - copy_bytes;
+ tcmu_flush_dcache_range(from, copy_bytes);
+ memcpy(to, from, copy_bytes);
+
+ UPDATE_HEAD(udev->data_tail, copy_bytes, udev->data_size);
+ }
+
+ kunmap_atomic(to);
+ }
+
+ } else if (se_cmd->data_direction == DMA_TO_DEVICE) {
+ UPDATE_HEAD(udev->data_tail, cmd->data_length, udev->data_size);
+ } else {
+ pr_warn("TCMU: data direction was %d!\n", se_cmd->data_direction);
+ }
+
+ target_complete_cmd(cmd->se_cmd, entry->rsp.scsi_status);
+ cmd->se_cmd = NULL;
+
+ kmem_cache_free(tcmu_cmd_cache, cmd);
+}
+
+static unsigned int tcmu_handle_completions(struct tcmu_dev *udev)
+{
+ struct tcmu_mailbox *mb;
+ LIST_HEAD(cpl_cmds);
+ unsigned long flags;
+ int handled = 0;
+
+ if (test_bit(TCMU_DEV_BIT_BROKEN, &udev->flags)) {
+ pr_err("ring broken, not handling completions\n");
+ return 0;
+ }
+
+ spin_lock_irqsave(&udev->cmdr_lock, flags);
+
+ mb = udev->mb_addr;
+ tcmu_flush_dcache_range(mb, sizeof(*mb));
+
+ while (udev->cmdr_last_cleaned != ACCESS_ONCE(mb->cmd_tail)) {
+
+ struct tcmu_cmd_entry *entry = (void *) mb + CMDR_OFF + udev->cmdr_last_cleaned;
+ struct tcmu_cmd *cmd;
+
+ tcmu_flush_dcache_range(entry, sizeof(*entry));
+
+ if (tcmu_hdr_get_op(&entry->hdr) == TCMU_OP_PAD) {
+ UPDATE_HEAD(udev->cmdr_last_cleaned, tcmu_hdr_get_len(&entry->hdr), udev->cmdr_size);
+ continue;
+ }
+ WARN_ON(tcmu_hdr_get_op(&entry->hdr) != TCMU_OP_CMD);
+
+ spin_lock(&udev->commands_lock);
+ cmd = idr_find(&udev->commands, entry->cmd_id);
+ if (cmd)
+ idr_remove(&udev->commands, cmd->cmd_id);
+ spin_unlock(&udev->commands_lock);
+
+ if (!cmd) {
+ pr_err("cmd_id not found, ring is broken\n");
+ set_bit(TCMU_DEV_BIT_BROKEN, &udev->flags);
+ break;
+ }
+
+ tcmu_handle_completion(cmd, entry);
+
+ UPDATE_HEAD(udev->cmdr_last_cleaned, tcmu_hdr_get_len(&entry->hdr), udev->cmdr_size);
+
+ handled++;
+ }
+
+ if (mb->cmd_tail == mb->cmd_head)
+ del_timer(&udev->timeout); /* no more pending cmds */
+
+ spin_unlock_irqrestore(&udev->cmdr_lock, flags);
+
+ wake_up(&udev->wait_cmdr);
+
+ return handled;
+}
+
+static int tcmu_check_expired_cmd(int id, void *p, void *data)
+{
+ struct tcmu_cmd *cmd = p;
+
+ if (test_bit(TCMU_CMD_BIT_EXPIRED, &cmd->flags))
+ return 0;
+
+ if (!time_after(cmd->deadline, jiffies))
+ return 0;
+
+ set_bit(TCMU_CMD_BIT_EXPIRED, &cmd->flags);
+ target_complete_cmd(cmd->se_cmd, SAM_STAT_CHECK_CONDITION);
+ cmd->se_cmd = NULL;
+
+ kmem_cache_free(tcmu_cmd_cache, cmd);
+
+ return 0;
+}
+
+static void tcmu_device_timedout(unsigned long data)
+{
+ struct tcmu_dev *udev = (struct tcmu_dev *)data;
+ unsigned long flags;
+ int handled;
+
+ handled = tcmu_handle_completions(udev);
+
+ pr_warn("%d completions handled from timeout\n", handled);
+
+ spin_lock_irqsave(&udev->commands_lock, flags);
+ idr_for_each(&udev->commands, tcmu_check_expired_cmd, NULL);
+ spin_unlock_irqrestore(&udev->commands_lock, flags);
+
+ /*
+ * We don't need to wakeup threads on wait_cmdr since they have their
+ * own timeout.
+ */
+}
+
+static int tcmu_attach_hba(struct se_hba *hba, u32 host_id)
+{
+ struct tcmu_hba *tcmu_hba;
+
+ tcmu_hba = kzalloc(sizeof(struct tcmu_hba), GFP_KERNEL);
+ if (!tcmu_hba)
+ return -ENOMEM;
+
+ tcmu_hba->host_id = host_id;
+ hba->hba_ptr = tcmu_hba;
+
+ return 0;
+}
+
+static void tcmu_detach_hba(struct se_hba *hba)
+{
+ kfree(hba->hba_ptr);
+ hba->hba_ptr = NULL;
+}
+
+static struct se_device *tcmu_alloc_device(struct se_hba *hba, const char *name)
+{
+ struct tcmu_dev *udev;
+
+ udev = kzalloc(sizeof(struct tcmu_dev), GFP_KERNEL);
+ if (!udev)
+ return NULL;
+
+ udev->name = kstrdup(name, GFP_KERNEL);
+ if (!udev->name) {
+ kfree(udev);
+ return NULL;
+ }
+
+ udev->hba = hba;
+
+ init_waitqueue_head(&udev->wait_cmdr);
+ spin_lock_init(&udev->cmdr_lock);
+
+ idr_init(&udev->commands);
+ spin_lock_init(&udev->commands_lock);
+
+ setup_timer(&udev->timeout, tcmu_device_timedout,
+ (unsigned long)udev);
+
+ udev->pass_level = TCMU_PASS_ALL;
+
+ return &udev->se_dev;
+}
+
+static int tcmu_irqcontrol(struct uio_info *info, s32 irq_on)
+{
+ struct tcmu_dev *tcmu_dev = container_of(info, struct tcmu_dev, uio_info);
+
+ tcmu_handle_completions(tcmu_dev);
+
+ return 0;
+}
+
+/*
+ * mmap code from uio.c. Copied here because we want to hook mmap()
+ * and this stuff must come along.
+ */
+static int tcmu_find_mem_index(struct vm_area_struct *vma)
+{
+ struct tcmu_dev *udev = vma->vm_private_data;
+ struct uio_info *info = &udev->uio_info;
+
+ if (vma->vm_pgoff < MAX_UIO_MAPS) {
+ if (info->mem[vma->vm_pgoff].size == 0)
+ return -1;
+ return (int)vma->vm_pgoff;
+ }
+ return -1;
+}
+
+static int tcmu_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
+{
+ struct tcmu_dev *udev = vma->vm_private_data;
+ struct uio_info *info = &udev->uio_info;
+ struct page *page;
+ unsigned long offset;
+ void *addr;
+
+ int mi = tcmu_find_mem_index(vma);
+ if (mi < 0)
+ return VM_FAULT_SIGBUS;
+
+ /*
+ * We need to subtract mi because userspace uses offset = N*PAGE_SIZE
+ * to use mem[N].
+ */
+ offset = (vmf->pgoff - mi) << PAGE_SHIFT;
+
+ addr = (void *)(unsigned long)info->mem[mi].addr + offset;
+ if (info->mem[mi].memtype == UIO_MEM_LOGICAL)
+ page = virt_to_page(addr);
+ else
+ page = vmalloc_to_page(addr);
+ get_page(page);
+ vmf->page = page;
+ return 0;
+}
+
+static const struct vm_operations_struct tcmu_vm_ops = {
+ .fault = tcmu_vma_fault,
+};
+
+static int tcmu_mmap(struct uio_info *info, struct vm_area_struct *vma)
+{
+ struct tcmu_dev *udev = container_of(info, struct tcmu_dev, uio_info);
+
+ vma->vm_flags |= VM_DONTEXPAND | VM_DONTDUMP;
+ vma->vm_ops = &tcmu_vm_ops;
+
+ vma->vm_private_data = udev;
+
+ /* Ensure the mmap is exactly the right size */
+ if (vma_pages(vma) != (TCMU_RING_SIZE >> PAGE_SHIFT))
+ return -EINVAL;
+
+ return 0;
+}
+
+static int tcmu_open(struct uio_info *info, struct inode *inode)
+{
+ struct tcmu_dev *udev = container_of(info, struct tcmu_dev, uio_info);
+
+ /* O_EXCL not supported for char devs, so fake it? */
+ if (test_and_set_bit(TCMU_DEV_BIT_OPEN, &udev->flags))
+ return -EBUSY;
+
+ pr_debug("open\n");
+
+ return 0;
+}
+
+static int tcmu_release(struct uio_info *info, struct inode *inode)
+{
+ struct tcmu_dev *udev = container_of(info, struct tcmu_dev, uio_info);
+
+ clear_bit(TCMU_DEV_BIT_OPEN, &udev->flags);
+
+ pr_debug("close\n");
+
+ return 0;
+}
+
+static int tcmu_netlink_event(enum tcmu_genl_cmd cmd, const char *name, int minor)
+{
+ struct sk_buff *skb;
+ void *msg_header;
+ int ret = -ENOMEM;
+
+ skb = genlmsg_new(NLMSG_GOODSIZE, GFP_KERNEL);
+ if (!skb)
+ return ret;
+
+ msg_header = genlmsg_put(skb, 0, 0, &tcmu_genl_family, 0, cmd);
+ if (!msg_header)
+ goto free_skb;
+
+ ret = nla_put_string(skb, TCMU_ATTR_DEVICE, name);
+ if (ret < 0)
+ goto free_skb;
+
+ ret = nla_put_u32(skb, TCMU_ATTR_MINOR, minor);
+ if (ret < 0)
+ goto free_skb;
+
+ ret = genlmsg_end(skb, msg_header);
+ if (ret < 0)
+ goto free_skb;
+
+ ret = genlmsg_multicast(&tcmu_genl_family, skb, 0,
+ TCMU_MCGRP_CONFIG, GFP_KERNEL);
+
+ /* We don't care if no one is listening */
+ if (ret == -ESRCH)
+ ret = 0;
+
+ return ret;
+free_skb:
+ nlmsg_free(skb);
+ return ret;
+}
+
+static int tcmu_configure_device(struct se_device *dev)
+{
+ struct tcmu_dev *udev = TCMU_DEV(dev);
+ struct tcmu_hba *hba = udev->hba->hba_ptr;
+ struct uio_info *info;
+ struct tcmu_mailbox *mb;
+ size_t size;
+ size_t used;
+ int ret = 0;
+ char *str;
+
+ info = &udev->uio_info;
+
+ size = snprintf(NULL, 0, "tcm-user/%u/%s/%s", hba->host_id, udev->name,
+ udev->dev_config);
+ size += 1; /* for \0 */
+ str = kmalloc(size, GFP_KERNEL);
+ if (!str)
+ return -ENOMEM;
+
+ used = snprintf(str, size, "tcm-user/%u/%s", hba->host_id, udev->name);
+
+ if (udev->dev_config[0])
+ snprintf(str + used, size - used, "/%s", udev->dev_config);
+
+ info->name = str;
+
+ udev->mb_addr = vzalloc(TCMU_RING_SIZE);
+ if (!udev->mb_addr) {
+ ret = -ENOMEM;
+ goto err_vzalloc;
+ }
+
+ /* mailbox fits in first part of CMDR space */
+ udev->cmdr_size = CMDR_SIZE - CMDR_OFF;
+ udev->data_off = CMDR_SIZE;
+ udev->data_size = TCMU_RING_SIZE - CMDR_SIZE;
+
+ mb = udev->mb_addr;
+ mb->version = 1;
+ mb->cmdr_off = CMDR_OFF;
+ mb->cmdr_size = udev->cmdr_size;
+
+ WARN_ON(!PAGE_ALIGNED(udev->data_off));
+ WARN_ON(udev->data_size % PAGE_SIZE);
+
+ info->version = "1";
+
+ info->mem[0].name = "tcm-user command & data buffer";
+ info->mem[0].addr = (phys_addr_t) udev->mb_addr;
+ info->mem[0].size = TCMU_RING_SIZE;
+ info->mem[0].memtype = UIO_MEM_VIRTUAL;
+
+ info->irqcontrol = tcmu_irqcontrol;
+ info->irq = UIO_IRQ_CUSTOM;
+
+ info->mmap = tcmu_mmap;
+ info->open = tcmu_open;
+ info->release = tcmu_release;
+
+ ret = uio_register_device(tcmu_root_device, info);
+ if (ret)
+ goto err_register;
+
+ /* Other attributes can be configured in userspace */
+ dev->dev_attrib.hw_block_size = 512;
+ dev->dev_attrib.hw_max_sectors = 128;
+ dev->dev_attrib.hw_queue_depth = 128;
+
+ ret = tcmu_netlink_event(TCMU_CMD_ADDED_DEVICE, udev->uio_info.name,
+ udev->uio_info.uio_dev->minor);
+ if (ret)
+ goto err_netlink;
+
+ return 0;
+
+err_netlink:
+ uio_unregister_device(&udev->uio_info);
+err_register:
+ vfree(udev->mb_addr);
+err_vzalloc:
+ kfree(info->name);
+
+ return ret;
+}
+
+static int tcmu_check_pending_cmd(int id, void *p, void *data)
+{
+ struct tcmu_cmd *cmd = p;
+
+ if (test_bit(TCMU_CMD_BIT_EXPIRED, &cmd->flags))
+ return 0;
+ return -EINVAL;
+}
+
+static void tcmu_free_device(struct se_device *dev)
+{
+ struct tcmu_dev *udev = TCMU_DEV(dev);
+ int i;
+
+ del_timer_sync(&udev->timeout);
+
+ vfree(udev->mb_addr);
+
+ /* Upper layer should drain all requests before calling this */
+ spin_lock_irq(&udev->commands_lock);
+ i = idr_for_each(&udev->commands, tcmu_check_pending_cmd, NULL);
+ idr_destroy(&udev->commands);
+ spin_unlock_irq(&udev->commands_lock);
+ WARN_ON(i);
+
+ /* Device was configured */
+ if (udev->uio_info.uio_dev) {
+ tcmu_netlink_event(TCMU_CMD_REMOVED_DEVICE, udev->uio_info.name,
+ udev->uio_info.uio_dev->minor);
+
+ uio_unregister_device(&udev->uio_info);
+ kfree(udev->uio_info.name);
+ kfree(udev->name);
+ }
+
+ kfree(udev);
+}
+
+enum {
+ Opt_dev_config, Opt_dev_size, Opt_err, Opt_pass_level,
+};
+
+static match_table_t tokens = {
+ {Opt_dev_config, "dev_config=%s"},
+ {Opt_dev_size, "dev_size=%u"},
+ {Opt_pass_level, "pass_level=%u"},
+ {Opt_err, NULL}
+};
+
+static ssize_t tcmu_set_configfs_dev_params(struct se_device *dev,
+ const char *page, ssize_t count)
+{
+ struct tcmu_dev *udev = TCMU_DEV(dev);
+ char *orig, *ptr, *opts, *arg_p;
+ substring_t args[MAX_OPT_ARGS];
+ int ret = 0, token;
+ int arg;
+
+ opts = kstrdup(page, GFP_KERNEL);
+ if (!opts)
+ return -ENOMEM;
+
+ orig = opts;
+
+ while ((ptr = strsep(&opts, ",\n")) != NULL) {
+ if (!*ptr)
+ continue;
+
+ token = match_token(ptr, tokens, args);
+ switch (token) {
+ case Opt_dev_config:
+ if (match_strlcpy(udev->dev_config, &args[0],
+ TCMU_CONFIG_LEN) == 0) {
+ ret = -EINVAL;
+ break;
+ }
+ pr_debug("TCMU: Referencing Path: %s\n", udev->dev_config);
+ break;
+ case Opt_dev_size:
+ arg_p = match_strdup(&args[0]);
+ if (!arg_p) {
+ ret = -ENOMEM;
+ break;
+ }
+ ret = kstrtoul(arg_p, 0, (unsigned long *) &udev->dev_size);
+ kfree(arg_p);
+ if (ret < 0)
+ pr_err("kstrtoul() failed for dev_size=\n");
+ break;
+ case Opt_pass_level:
+ match_int(args, &arg);
+ if (arg >= TCMU_PASS_INVALID) {
+ pr_warn("TCMU: Invalid pass_level: %d\n", arg);
+ break;
+ }
+
+ pr_debug("TCMU: Setting pass_level to %d\n", arg);
+ udev->pass_level = arg;
+ break;
+ default:
+ break;
+ }
+ }
+
+ kfree(orig);
+ return (!ret) ? count : ret;
+}
+
+static ssize_t tcmu_show_configfs_dev_params(struct se_device *dev, char *b)
+{
+ struct tcmu_dev *udev = TCMU_DEV(dev);
+ ssize_t bl = 0;
+
+ bl = sprintf(b + bl, "Config: %s ",
+ udev->dev_config[0] ? udev->dev_config : "NULL");
+ bl += sprintf(b + bl, "Size: %zu PassLevel: %u\n",
+ udev->dev_size, udev->pass_level);
+
+ return bl;
+}
+
+static sector_t tcmu_get_blocks(struct se_device *dev)
+{
+ struct tcmu_dev *udev = TCMU_DEV(dev);
+
+ return div_u64(udev->dev_size - dev->dev_attrib.block_size,
+ dev->dev_attrib.block_size);
+}
+
+static sense_reason_t
+tcmu_execute_rw(struct se_cmd *se_cmd, struct scatterlist *sgl, u32 sgl_nents,
+ enum dma_data_direction data_direction)
+{
+ int ret;
+
+ ret = tcmu_queue_cmd(se_cmd);
+
+ if (ret != 0)
+ return TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
+ else
+ return TCM_NO_SENSE;
+}
+
+static sense_reason_t
+tcmu_pass_op(struct se_cmd *se_cmd)
+{
+ int ret = tcmu_queue_cmd(se_cmd);
+
+ if (ret != 0)
+ return TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
+ else
+ return TCM_NO_SENSE;
+}
+
+static struct sbc_ops tcmu_sbc_ops = {
+ .execute_rw = tcmu_execute_rw,
+ .execute_sync_cache = tcmu_pass_op,
+ .execute_write_same = tcmu_pass_op,
+ .execute_write_same_unmap = tcmu_pass_op,
+ .execute_unmap = tcmu_pass_op,
+};
+
+static sense_reason_t
+tcmu_parse_cdb(struct se_cmd *cmd)
+{
+ unsigned char *cdb = cmd->t_task_cdb;
+ struct tcmu_dev *udev = TCMU_DEV(cmd->se_dev);
+ sense_reason_t ret;
+
+ switch (udev->pass_level) {
+ case TCMU_PASS_ALL:
+ /* We're just like pscsi, then */
+ /*
+ * For REPORT LUNS we always need to emulate the response, for everything
+ * else, pass it up.
+ */
+ switch (cdb[0]) {
+ case REPORT_LUNS:
+ cmd->execute_cmd = spc_emulate_report_luns;
+ break;
+ case READ_6:
+ case READ_10:
+ case READ_12:
+ case READ_16:
+ case WRITE_6:
+ case WRITE_10:
+ case WRITE_12:
+ case WRITE_16:
+ case WRITE_VERIFY:
+ cmd->se_cmd_flags |= SCF_SCSI_DATA_CDB;
+ /* FALLTHROUGH */
+ default:
+ cmd->execute_cmd = tcmu_pass_op;
+ }
+ ret = TCM_NO_SENSE;
+ break;
+ case TCMU_PASS_IO:
+ ret = sbc_parse_cdb(cmd, &tcmu_sbc_ops);
+ break;
+ default:
+ pr_err("Unknown tcm-user pass level %d\n", udev->pass_level);
+ ret = TCM_CHECK_CONDITION_ABORT_CMD;
+ }
+
+ return ret;
+}
+
+static struct se_subsystem_api tcmu_template = {
+ .name = "user",
+ .inquiry_prod = "USER",
+ .inquiry_rev = TCMU_VERSION,
+ .owner = THIS_MODULE,
+ .transport_type = TRANSPORT_PLUGIN_VHBA_PDEV,
+ .attach_hba = tcmu_attach_hba,
+ .detach_hba = tcmu_detach_hba,
+ .alloc_device = tcmu_alloc_device,
+ .configure_device = tcmu_configure_device,
+ .free_device = tcmu_free_device,
+ .parse_cdb = tcmu_parse_cdb,
+ .set_configfs_dev_params = tcmu_set_configfs_dev_params,
+ .show_configfs_dev_params = tcmu_show_configfs_dev_params,
+ .get_device_type = sbc_get_device_type,
+ .get_blocks = tcmu_get_blocks,
+};
+
+static int __init tcmu_module_init(void)
+{
+ int ret;
+
+ BUILD_BUG_ON((sizeof(struct tcmu_cmd_entry) % TCMU_OP_ALIGN_SIZE) != 0);
+
+ tcmu_cmd_cache = kmem_cache_create("tcmu_cmd_cache",
+ sizeof(struct tcmu_cmd),
+ __alignof__(struct tcmu_cmd),
+ 0, NULL);
+ if (!tcmu_cmd_cache)
+ return -ENOMEM;
+
+ tcmu_root_device = root_device_register("tcm_user");
+ if (IS_ERR(tcmu_root_device)) {
+ ret = PTR_ERR(tcmu_root_device);
+ goto out_free_cache;
+ }
+
+ ret = genl_register_family(&tcmu_genl_family);
+ if (ret < 0) {
+ goto out_unreg_device;
+ }
+
+ ret = transport_subsystem_register(&tcmu_template);
+ if (ret)
+ goto out_unreg_genl;
+
+ return 0;
+
+out_unreg_genl:
+ genl_unregister_family(&tcmu_genl_family);
+out_unreg_device:
+ root_device_unregister(tcmu_root_device);
+out_free_cache:
+ kmem_cache_destroy(tcmu_cmd_cache);
+
+ return ret;
+}
+
+static void __exit tcmu_module_exit(void)
+{
+ transport_subsystem_release(&tcmu_template);
+ genl_unregister_family(&tcmu_genl_family);
+ root_device_unregister(tcmu_root_device);
+ kmem_cache_destroy(tcmu_cmd_cache);
+}
+
+MODULE_DESCRIPTION("TCM USER subsystem plugin");
+MODULE_AUTHOR("Shaohua Li <shli@kernel.org>");
+MODULE_AUTHOR("Andy Grover <agrover@redhat.com>");
+MODULE_LICENSE("GPL");
+
+module_init(tcmu_module_init);
+module_exit(tcmu_module_exit);
ft_sess_delete_all(tport);
lport = tport->lport;
BUG_ON(tport != lport->prov[FC_TYPE_FCP]);
- rcu_assign_pointer(lport->prov[FC_TYPE_FCP], NULL);
+ RCU_INIT_POINTER(lport->prov[FC_TYPE_FCP], NULL);
tpg = tport->tpg;
if (tpg) {
#define UIO_MAX_DEVICES (1U << MINORBITS)
-struct uio_device {
- struct module *owner;
- struct device *dev;
- int minor;
- atomic_t event;
- struct fasync_struct *async_queue;
- wait_queue_head_t wait;
- struct uio_info *info;
- struct kobject *map_dir;
- struct kobject *portio_dir;
-};
-
static int uio_major;
static struct cdev *uio_cdev;
static DEFINE_IDR(uio_idr);
#define MAX_UIO_PORT_REGIONS 5
-struct uio_device;
+struct uio_device {
+ struct module *owner;
+ struct device *dev;
+ int minor;
+ atomic_t event;
+ struct fasync_struct *async_queue;
+ wait_queue_head_t wait;
+ struct uio_info *info;
+ struct kobject *map_dir;
+ struct kobject *portio_dir;
+};
/**
* struct uio_info - UIO device capabilities
#define DA_EMULATE_ALUA 0
/* Enforce SCSI Initiator Port TransportID with 'ISID' for PR */
#define DA_ENFORCE_PR_ISIDS 1
+/* Force SPC-3 PR Activate Persistence across Target Power Loss */
+#define DA_FORCE_PR_APTPL 0
#define DA_STATUS_MAX_SECTORS_MIN 16
#define DA_STATUS_MAX_SECTORS_MAX 8192
/* By default don't report non-rotating (solid state) medium */
enum target_prot_type pi_prot_type;
enum target_prot_type hw_pi_prot_type;
int enforce_pr_isids;
+ int force_pr_aptpl;
int is_nonrot;
int emulate_rest_reord;
u32 hw_block_size;
struct config_group fabric_stat_group;
};
+static inline void atomic_inc_mb(atomic_t *v)
+{
+ smp_mb__before_atomic();
+ atomic_inc(v);
+ smp_mb__after_atomic();
+}
+
+static inline void atomic_dec_mb(atomic_t *v)
+{
+ smp_mb__before_atomic();
+ atomic_dec(v);
+ smp_mb__after_atomic();
+}
+
#endif /* TARGET_CORE_BASE_H */
header-y += synclink.h
header-y += sysctl.h
header-y += sysinfo.h
+header-y += target_core_user.h
header-y += taskstats.h
header-y += tcp.h
header-y += tcp_metrics.h
--- /dev/null
+#ifndef __TARGET_CORE_USER_H
+#define __TARGET_CORE_USER_H
+
+/* This header will be used by application too */
+
+#include <linux/types.h>
+#include <linux/uio.h>
+
+#ifndef __packed
+#define __packed __attribute__((packed))
+#endif
+
+#define TCMU_VERSION "1.0"
+
+/*
+ * Ring Design
+ * -----------
+ *
+ * The mmaped area is divided into three parts:
+ * 1) The mailbox (struct tcmu_mailbox, below)
+ * 2) The command ring
+ * 3) Everything beyond the command ring (data)
+ *
+ * The mailbox tells userspace the offset of the command ring from the
+ * start of the shared memory region, and how big the command ring is.
+ *
+ * The kernel passes SCSI commands to userspace by putting a struct
+ * tcmu_cmd_entry in the ring, updating mailbox->cmd_head, and poking
+ * userspace via uio's interrupt mechanism.
+ *
+ * tcmu_cmd_entry contains a header. If the header type is PAD,
+ * userspace should skip hdr->length bytes (mod cmdr_size) to find the
+ * next cmd_entry.
+ *
+ * Otherwise, the entry will contain offsets into the mmaped area that
+ * contain the cdb and data buffers -- the latter accessible via the
+ * iov array. iov addresses are also offsets into the shared area.
+ *
+ * When userspace is completed handling the command, set
+ * entry->rsp.scsi_status, fill in rsp.sense_buffer if appropriate,
+ * and also set mailbox->cmd_tail equal to the old cmd_tail plus
+ * hdr->length, mod cmdr_size. If cmd_tail doesn't equal cmd_head, it
+ * should process the next packet the same way, and so on.
+ */
+
+#define TCMU_MAILBOX_VERSION 1
+#define ALIGN_SIZE 64 /* Should be enough for most CPUs */
+
+struct tcmu_mailbox {
+ __u16 version;
+ __u16 flags;
+ __u32 cmdr_off;
+ __u32 cmdr_size;
+
+ __u32 cmd_head;
+
+ /* Updated by user. On its own cacheline */
+ __u32 cmd_tail __attribute__((__aligned__(ALIGN_SIZE)));
+
+} __packed;
+
+enum tcmu_opcode {
+ TCMU_OP_PAD = 0,
+ TCMU_OP_CMD,
+};
+
+/*
+ * Only a few opcodes, and length is 8-byte aligned, so use low bits for opcode.
+ */
+struct tcmu_cmd_entry_hdr {
+ __u32 len_op;
+} __packed;
+
+#define TCMU_OP_MASK 0x7
+
+static inline enum tcmu_opcode tcmu_hdr_get_op(struct tcmu_cmd_entry_hdr *hdr)
+{
+ return hdr->len_op & TCMU_OP_MASK;
+}
+
+static inline void tcmu_hdr_set_op(struct tcmu_cmd_entry_hdr *hdr, enum tcmu_opcode op)
+{
+ hdr->len_op &= ~TCMU_OP_MASK;
+ hdr->len_op |= (op & TCMU_OP_MASK);
+}
+
+static inline __u32 tcmu_hdr_get_len(struct tcmu_cmd_entry_hdr *hdr)
+{
+ return hdr->len_op & ~TCMU_OP_MASK;
+}
+
+static inline void tcmu_hdr_set_len(struct tcmu_cmd_entry_hdr *hdr, __u32 len)
+{
+ hdr->len_op &= TCMU_OP_MASK;
+ hdr->len_op |= len;
+}
+
+/* Currently the same as SCSI_SENSE_BUFFERSIZE */
+#define TCMU_SENSE_BUFFERSIZE 96
+
+struct tcmu_cmd_entry {
+ struct tcmu_cmd_entry_hdr hdr;
+
+ uint16_t cmd_id;
+ uint16_t __pad1;
+
+ union {
+ struct {
+ uint64_t cdb_off;
+ uint64_t iov_cnt;
+ struct iovec iov[0];
+ } req;
+ struct {
+ uint8_t scsi_status;
+ uint8_t __pad1;
+ uint16_t __pad2;
+ uint32_t __pad3;
+ char sense_buffer[TCMU_SENSE_BUFFERSIZE];
+ } rsp;
+ };
+
+} __packed;
+
+#define TCMU_OP_ALIGN_SIZE sizeof(uint64_t)
+
+enum tcmu_genl_cmd {
+ TCMU_CMD_UNSPEC,
+ TCMU_CMD_ADDED_DEVICE,
+ TCMU_CMD_REMOVED_DEVICE,
+ __TCMU_CMD_MAX,
+};
+#define TCMU_CMD_MAX (__TCMU_CMD_MAX - 1)
+
+enum tcmu_genl_attr {
+ TCMU_ATTR_UNSPEC,
+ TCMU_ATTR_DEVICE,
+ TCMU_ATTR_MINOR,
+ __TCMU_ATTR_MAX,
+};
+#define TCMU_ATTR_MAX (__TCMU_ATTR_MAX - 1)
+
+#endif
projects / cascardo / linux.git / commitdiff