On 10/18/24 18:14, Karan Tilak Kumar wrote:
Modify IO path to use FDLS.
Add helper functions to process IOs.
Remove unused template functions.
Cleanup obsolete code.
Refactor old function definitions.
Reviewed-by: Sesidhar Baddela <sebaddel@xxxxxxxxx>
Reviewed-by: Arulprabhu Ponnusamy <arulponn@xxxxxxxxx>
Reviewed-by: Gian Carlo Boffa <gcboffa@xxxxxxxxx>
Reviewed-by: Arun Easi <aeasi@xxxxxxxxx>
Signed-off-by: Karan Tilak Kumar <kartilak@xxxxxxxxx>
---
Changes between v2 and v3:
Replace fnic->host with fnic->lport->host to prevent compilation
errors.
Changes between v1 and v2:
Replace pr_err with printk.
Incorporate review comments by Hannes:
Restore usage of tagset iterators.
---
drivers/scsi/fnic/fnic.h | 20 +-
drivers/scsi/fnic/fnic_io.h | 3 +
drivers/scsi/fnic/fnic_main.c | 5 +-
drivers/scsi/fnic/fnic_scsi.c | 846 +++++++++++++++++++--------------
drivers/scsi/fnic/fnic_stats.h | 2 -
5 files changed, 513 insertions(+), 363 deletions(-)
diff --git a/drivers/scsi/fnic/fnic.h b/drivers/scsi/fnic/fnic.h
index 4f38cbae1994..c6fe9eec9a0c 100644
--- a/drivers/scsi/fnic/fnic.h
+++ b/drivers/scsi/fnic/fnic.h
@@ -40,6 +40,7 @@
#define FNIC_DFLT_IO_REQ 256 /* Default scsi_cmnd tag map entries */
#define FNIC_DFLT_QUEUE_DEPTH 256
#define FNIC_STATS_RATE_LIMIT 4 /* limit rate at which stats are pulled up */
+#define LUN0_DELAY_TIME 9
/*
* Tag bits used for special requests.
@@ -472,7 +473,6 @@ int fnic_set_intr_mode_msix(struct fnic *fnic);
void fnic_free_intr(struct fnic *fnic);
int fnic_request_intr(struct fnic *fnic);
-int fnic_send(struct fc_lport *, struct fc_frame *);
void fnic_free_wq_buf(struct vnic_wq *wq, struct vnic_wq_buf *buf);
void fnic_handle_frame(struct work_struct *work);
void fnic_tport_event_handler(struct work_struct *work);
@@ -489,11 +489,9 @@ int fnic_abort_cmd(struct scsi_cmnd *);
int fnic_device_reset(struct scsi_cmnd *);
int fnic_eh_host_reset_handler(struct scsi_cmnd *sc);
int fnic_host_reset(struct Scsi_Host *shost);
-int fnic_reset(struct Scsi_Host *);
-void fnic_scsi_cleanup(struct fc_lport *);
-void fnic_scsi_abort_io(struct fc_lport *);
-void fnic_empty_scsi_cleanup(struct fc_lport *);
-void fnic_exch_mgr_reset(struct fc_lport *, u32, u32);
+void fnic_reset(struct Scsi_Host *shost);
+int fnic_issue_fc_host_lip(struct Scsi_Host *shost);
+void fnic_scsi_fcpio_reset(struct fnic *fnic);
int fnic_wq_copy_cmpl_handler(struct fnic *fnic, int copy_work_to_do, unsigned int cq_index);
int fnic_wq_cmpl_handler(struct fnic *fnic, int);
int fnic_flogi_reg_handler(struct fnic *fnic, u32);
@@ -505,7 +503,8 @@ const char *fnic_state_to_str(unsigned int state);
void fnic_mq_map_queues_cpus(struct Scsi_Host *host);
void fnic_log_q_error(struct fnic *fnic);
void fnic_handle_link_event(struct fnic *fnic);
-
+void fnic_stats_debugfs_init(struct fnic *fnic);
+void fnic_stats_debugfs_remove(struct fnic *fnic);
int fnic_is_abts_pending(struct fnic *, struct scsi_cmnd *);
void fnic_handle_fip_frame(struct work_struct *work);
@@ -526,5 +525,12 @@ int fnic_get_desc_by_devid(struct pci_dev *pdev, char **desc,
void fnic_fdls_link_status_change(struct fnic *fnic, int linkup);
void fnic_delete_fcp_tports(struct fnic *fnic);
void fnic_flush_tport_event_list(struct fnic *fnic);
+int fnic_count_ioreqs_wq(struct fnic *fnic, u32 hwq, u32 portid);
+unsigned int fnic_count_ioreqs(struct fnic *fnic, u32 portid);
+unsigned int fnic_count_all_ioreqs(struct fnic *fnic);
+unsigned int fnic_count_lun_ioreqs_wq(struct fnic *fnic, u32 hwq,
+ struct scsi_device *device);
+unsigned int fnic_count_lun_ioreqs(struct fnic *fnic,
+ struct scsi_device *device);
#endif /* _FNIC_H_ */
diff --git a/drivers/scsi/fnic/fnic_io.h b/drivers/scsi/fnic/fnic_io.h
index 6fe642cb387b..0d974e040ab7 100644
--- a/drivers/scsi/fnic/fnic_io.h
+++ b/drivers/scsi/fnic/fnic_io.h
@@ -7,6 +7,7 @@
#define _FNIC_IO_H_
#include <scsi/fc/fc_fcp.h>
+#include "fnic_fdls.h"
#define FNIC_DFLT_SG_DESC_CNT 32
#define FNIC_MAX_SG_DESC_CNT 256 /* Maximum descriptors per sgl */
@@ -41,6 +42,8 @@ enum fnic_ioreq_state {
};
struct fnic_io_req {
+ struct fnic_iport_s *iport;
+ struct fnic_tport_s *tport;
struct host_sg_desc *sgl_list; /* sgl list */
void *sgl_list_alloc; /* sgl list address used for free */
dma_addr_t sense_buf_pa; /* dma address for sense buffer*/
diff --git a/drivers/scsi/fnic/fnic_main.c b/drivers/scsi/fnic/fnic_main.c
index 33c63d7a9787..dadbe47c0bbb 100644
--- a/drivers/scsi/fnic/fnic_main.c
+++ b/drivers/scsi/fnic/fnic_main.c
@@ -85,9 +85,6 @@ module_param(fnic_max_qdepth, uint, S_IRUGO|S_IWUSR);
MODULE_PARM_DESC(fnic_max_qdepth, "Queue depth to report for each LUN");
static struct libfc_function_template fnic_transport_template = {
- .fcp_abort_io = fnic_empty_scsi_cleanup,
- .fcp_cleanup = fnic_empty_scsi_cleanup,
- .exch_mgr_reset = fnic_exch_mgr_reset
};
The template is now empty; can't you delete it completely?
struct workqueue_struct *fnic_fip_queue;
@@ -162,7 +159,7 @@ static struct fc_function_template fnic_fc_functions = {
.show_starget_port_id = 1,
.show_rport_dev_loss_tmo = 1,
.set_rport_dev_loss_tmo = fnic_set_rport_dev_loss_tmo,
- .issue_fc_host_lip = fnic_reset,
+ .issue_fc_host_lip = fnic_issue_fc_host_lip,
.get_fc_host_stats = fnic_get_stats,
.reset_fc_host_stats = fnic_reset_host_stats,
.dd_fcrport_size = sizeof(struct fc_rport_libfc_priv),
diff --git a/drivers/scsi/fnic/fnic_scsi.c b/drivers/scsi/fnic/fnic_scsi.c
index 74298f9a34e5..af729637ef2d 100644
--- a/drivers/scsi/fnic/fnic_scsi.c
+++ b/drivers/scsi/fnic/fnic_scsi.c
@@ -25,9 +25,12 @@
#include <scsi/fc/fc_fcoe.h>
#include <scsi/libfc.h>
#include <scsi/fc_frame.h>
+#include <scsi/scsi_transport_fc.h>
#include "fnic_io.h"
#include "fnic.h"
+static void fnic_cleanup_io(struct fnic *fnic, int exclude_id);
+
const char *fnic_state_str[] = {
[FNIC_IN_FC_MODE] = "FNIC_IN_FC_MODE",
[FNIC_IN_FC_TRANS_ETH_MODE] = "FNIC_IN_FC_TRANS_ETH_MODE",
@@ -65,6 +68,18 @@ static const char *fcpio_status_str[] = {
[FCPIO_LUNMAP_CHNG_PEND] = "FCPIO_LUNHMAP_CHNG_PEND",
};
+enum terminate_io_return {
+ TERM_SUCCESS = 0,
+ TERM_NO_SC = 1,
+ TERM_IO_REQ_NOT_FOUND,
+ TERM_ANOTHER_PORT,
+ TERM_GSTATE,
+ TERM_IO_BLOCKED,
+ TERM_OUT_OF_WQ_DESC,
+ TERM_TIMED_OUT,
+ TERM_MISC,
+};
+
const char *fnic_state_to_str(unsigned int state)
{
if (state >= ARRAY_SIZE(fnic_state_str) || !fnic_state_str[state])
@@ -90,8 +105,6 @@ static const char *fnic_fcpio_status_to_str(unsigned int status)
return fcpio_status_str[status];
}
-static void fnic_cleanup_io(struct fnic *fnic);
-
/*
* Unmap the data buffer and sense buffer for an io_req,
* also unmap and free the device-private scatter/gather list.
@@ -114,6 +127,80 @@ static void fnic_release_ioreq_buf(struct fnic *fnic,
SCSI_SENSE_BUFFERSIZE, DMA_FROM_DEVICE);
}
+int fnic_count_ioreqs_wq(struct fnic *fnic, u32 hwq, u32 portid)
+{
+ unsigned long flags = 0;
+ int i = 0, count = 0;
+
+ spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
+ for (i = 0; i != fnic->sw_copy_wq[hwq].ioreq_table_size; ++i) {
+ if (fnic->sw_copy_wq[hwq].io_req_table[i] != NULL &&
+ (!portid
+ || fnic->sw_copy_wq[hwq].io_req_table[i]->port_id == portid))
+ count++;
+ }
+ spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
+
+ return count;
+}
+
+unsigned int fnic_count_ioreqs(struct fnic *fnic, u32 portid)
+{
+ int i;
+ unsigned int count = 0;
+
+ for (i = 0; i < fnic->wq_copy_count; i++)
+ count += fnic_count_ioreqs_wq(fnic, i, portid);
+
+ return count;
+}
+
+unsigned int fnic_count_all_ioreqs(struct fnic *fnic)
+{
+ return fnic_count_ioreqs(fnic, 0);
+}
+
+unsigned int fnic_count_lun_ioreqs_wq(struct fnic *fnic, u32 hwq,
+ struct scsi_device *device)
+{
+ struct fnic_io_req *io_req;
+ int i;
+ unsigned int count = 0;
+ unsigned long flags = 0;
+
+ spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
+ for (i = 0; i != fnic->sw_copy_wq[hwq].ioreq_table_size; ++i) {
+ io_req = fnic->sw_copy_wq[hwq].io_req_table[i];
+
+ if (io_req != NULL) {
+ struct scsi_cmnd *sc =
+ scsi_host_find_tag(fnic->lport->host, io_req->tag);
+
Be careful. 'scsi_host_find_tag()' finds requests on _all_ hwqs.
But here you are interested in tags for a specific hwq only.
+ if (!sc)
+ continue;
+
+ if (sc->device == device)
+ count++;
+ }
+ }
+ spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
+
+ return count;
+}
+
And now this is pretty pointless. Please use the block layer
blk_mq_tagset_busy_iter() to iterate over busy/active tags.
+unsigned int fnic_count_lun_ioreqs(struct fnic *fnic,
+ struct scsi_device *device)
+{
+ int hwq;
+ unsigned int count = 0;
+
+ /*count if any pending IOs on this lun */
+ for (hwq = 0; hwq < fnic->wq_copy_count; hwq++)
+ count += fnic_count_lun_ioreqs_wq(fnic, hwq, device);
+
+ return count;
+}
+
Same here.
/* Free up Copy Wq descriptors. Called with copy_wq lock held */
static int free_wq_copy_descs(struct fnic *fnic, struct vnic_wq_copy *wq, unsigned int hwq)
{
@@ -179,12 +266,11 @@ int fnic_fw_reset_handler(struct fnic *fnic)
struct vnic_wq_copy *wq = &fnic->hw_copy_wq[0];
int ret = 0;
unsigned long flags;
+ unsigned int ioreq_count;
/* indicate fwreset to io path */
fnic_set_state_flags(fnic, FNIC_FLAGS_FWRESET);
-
- fnic_free_txq(&fnic->frame_queue);
- fnic_free_txq(&fnic->tx_queue);
+ ioreq_count = fnic_count_all_ioreqs(fnic);
/* wait for io cmpl */
while (atomic_read(&fnic->in_flight))
@@ -231,10 +317,10 @@ int fnic_flogi_reg_handler(struct fnic *fnic, u32 fc_id)
{
struct vnic_wq_copy *wq = &fnic->hw_copy_wq[0];
enum fcpio_flogi_reg_format_type format;
- struct fc_lport *lp = fnic->lport;
u8 gw_mac[ETH_ALEN];
int ret = 0;
unsigned long flags;
+ struct fnic_iport_s *iport = &fnic->iport;
spin_lock_irqsave(&fnic->wq_copy_lock[0], flags);
@@ -246,28 +332,23 @@ int fnic_flogi_reg_handler(struct fnic *fnic, u32 fc_id)
goto flogi_reg_ioreq_end;
}
- if (fnic->ctlr.map_dest) {
- eth_broadcast_addr(gw_mac);
- format = FCPIO_FLOGI_REG_DEF_DEST;
- } else {
- memcpy(gw_mac, fnic->ctlr.dest_addr, ETH_ALEN);
- format = FCPIO_FLOGI_REG_GW_DEST;
- }
+ memcpy(gw_mac, fnic->iport.fcfmac, ETH_ALEN);
+ format = FCPIO_FLOGI_REG_GW_DEST;
- if ((fnic->config.flags & VFCF_FIP_CAPABLE) && !fnic->ctlr.map_dest) {
+ if (fnic->config.flags & VFCF_FIP_CAPABLE) {
fnic_queue_wq_copy_desc_fip_reg(wq, SCSI_NO_TAG,
fc_id, gw_mac,
- fnic->data_src_addr,
- lp->r_a_tov, lp->e_d_tov);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "FLOGI FIP reg issued fcid %x src %pM dest %pM\n",
- fc_id, fnic->data_src_addr, gw_mac);
+ fnic->iport.fpma,
+ iport->r_a_tov, iport->e_d_tov);
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "FLOGI FIP reg issued fcid: 0x%x src %p dest %p\n",
+ fc_id, fnic->iport.fpma, gw_mac);
} else {
fnic_queue_wq_copy_desc_flogi_reg(wq, SCSI_NO_TAG,
format, fc_id, gw_mac);
FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
- "FLOGI reg issued fcid 0x%x map %d dest 0x%p\n",
- fc_id, fnic->ctlr.map_dest, gw_mac);
+ "FLOGI reg issued fcid 0x%x dest %p\n",
+ fc_id, gw_mac);
}
atomic64_inc(&fnic->fnic_stats.fw_stats.active_fw_reqs);
@@ -295,13 +376,17 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic,
{
struct scatterlist *sg;
struct fc_rport *rport = starget_to_rport(scsi_target(sc->device));
- struct fc_rport_libfc_priv *rp = rport->dd_data;
struct host_sg_desc *desc;
struct misc_stats *misc_stats = &fnic->fnic_stats.misc_stats;
unsigned int i;
int flags;
u8 exch_flags;
struct scsi_lun fc_lun;
+ struct fnic_tport_s *tport;
+ struct rport_dd_data_s *rdd_data;
+
+ rdd_data = rport->dd_data;
+ tport = rdd_data->tport;
if (sg_count) {
/* For each SGE, create a device desc entry */
@@ -356,7 +441,7 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic,
exch_flags = 0;
if ((fnic->config.flags & VFCF_FCP_SEQ_LVL_ERR) &&
- (rp->flags & FC_RP_FLAGS_RETRY))
+ (tport->tgt_flags & FDLS_FC_RP_FLAGS_RETRY))
exch_flags |= FCPIO_ICMND_SRFLAG_RETRY;
fnic_queue_wq_copy_desc_icmnd_16(wq, mqtag,
@@ -371,8 +456,8 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic,
sc->cmnd, sc->cmd_len,
scsi_bufflen(sc),
fc_lun.scsi_lun, io_req->port_id,
- rport->maxframe_size, rp->r_a_tov,
- rp->e_d_tov);
+ tport->max_payload_size,
+ tport->r_a_tov, tport->e_d_tov);
atomic64_inc(&fnic->fnic_stats.fw_stats.active_fw_reqs);
if (atomic64_read(&fnic->fnic_stats.fw_stats.active_fw_reqs) >
@@ -388,10 +473,10 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
struct request *const rq = scsi_cmd_to_rq(sc);
uint32_t mqtag = 0;
void (*done)(struct scsi_cmnd *) = scsi_done;
- struct fc_lport *lp = shost_priv(sc->device->host);
struct fc_rport *rport;
struct fnic_io_req *io_req = NULL;
- struct fnic *fnic = lport_priv(lp);
+ struct fnic *fnic = *((struct fnic **) shost_priv(sc->device->host));
+ struct fnic_iport_s *iport = NULL;
struct fnic_stats *fnic_stats = &fnic->fnic_stats;
struct vnic_wq_copy *wq;
int ret = 1;
@@ -400,32 +485,14 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
unsigned long flags = 0;
unsigned long ptr;
int io_lock_acquired = 0;
- struct fc_rport_libfc_priv *rp;
uint16_t hwq = 0;
-
- mqtag = blk_mq_unique_tag(rq);
- spin_lock_irqsave(&fnic->fnic_lock, flags);
-
- if (unlikely(fnic_chk_state_flags_locked(fnic, FNIC_FLAGS_IO_BLOCKED))) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
- "fnic IO blocked flags: 0x%lx. Returning SCSI_MLQUEUE_HOST_BUSY\n",
- fnic->state_flags);
- return SCSI_MLQUEUE_HOST_BUSY;
- }
-
- if (unlikely(fnic_chk_state_flags_locked(fnic, FNIC_FLAGS_FWRESET))) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
- "fnic flags: 0x%lx. Returning SCSI_MLQUEUE_HOST_BUSY\n",
- fnic->state_flags);
- return SCSI_MLQUEUE_HOST_BUSY;
- }
+ struct fnic_tport_s *tport = NULL;
+ struct rport_dd_data_s *rdd_data;
+ uint16_t lun0_delay = 0;
rport = starget_to_rport(scsi_target(sc->device));
if (!rport) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
"returning DID_NO_CONNECT for IO as rport is NULL\n");
sc->result = DID_NO_CONNECT << 16;
done(sc);
@@ -434,50 +501,95 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
ret = fc_remote_port_chkready(rport);
if (ret) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
"rport is not ready\n");
- atomic64_inc(&fnic_stats->misc_stats.rport_not_ready);
sc->result = ret;
done(sc);
return 0;
}
- rp = rport->dd_data;
- if (!rp || rp->rp_state == RPORT_ST_DELETE) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "rport 0x%x removed, returning DID_NO_CONNECT\n",
- rport->port_id);
+ mqtag = blk_mq_unique_tag(rq);
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
+ iport = &fnic->iport;
- atomic64_inc(&fnic_stats->misc_stats.rport_not_ready);
- sc->result = DID_NO_CONNECT<<16;
+ if (iport->state != FNIC_IPORT_STATE_READY) {
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "returning DID_NO_CONNECT for IO as iport state: %d\n",
+ iport->state);
+ sc->result = DID_NO_CONNECT << 16;
done(sc);
return 0;
}
- if (rp->rp_state != RPORT_ST_READY) {
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "rport 0x%x in state 0x%x, returning DID_IMM_RETRY\n",
- rport->port_id, rp->rp_state);
+ /* fc_remote_port_add() may have added the tport to
+ * fc_transport but dd_data not yet set
+ */
+ rdd_data = rport->dd_data;
+ tport = rdd_data->tport;
+ if (!tport || (rdd_data->iport != iport)) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "dd_data not yet set in SCSI for rport portid: 0x%x\n",
+ rport->port_id);
+ tport = fnic_find_tport_by_fcid(iport, rport->port_id);
+ if (!tport) {
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "returning DID_BUS_BUSY for IO as tport not found for: 0x%x\n",
+ rport->port_id);
+ sc->result = DID_BUS_BUSY << 16;
+ done(sc);
+ return 0;
+ }
+
+ /* Re-assign same params as in fnic_fdls_add_tport */
+ rport->maxframe_size = FNIC_FC_MAX_PAYLOAD_LEN;
+ rport->supported_classes =
+ FC_COS_CLASS3 | FC_RPORT_ROLE_FCP_TARGET;
+ /* the dd_data is allocated by fctransport of size dd_fcrport_size */
+ rdd_data = rport->dd_data;
+ rdd_data->tport = tport;
+ rdd_data->iport = iport;
+ tport->rport = rport;
+ tport->flags |= FNIC_FDLS_SCSI_REGISTERED;
+ }
- sc->result = DID_IMM_RETRY << 16;
+ if ((tport->state != FDLS_TGT_STATE_READY)
+ && (tport->state != FDLS_TGT_STATE_ADISC)) {
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "returning DID_NO_CONNECT for IO as tport state: %d\n",
+ tport->state);
+ sc->result = DID_NO_CONNECT << 16;
done(sc);
return 0;
}
- if (lp->state != LPORT_ST_READY || !(lp->link_up)) {
+ atomic_inc(&fnic->in_flight);
+ atomic_inc(&tport->in_flight);
+
+ if (unlikely(fnic_chk_state_flags_locked(fnic, FNIC_FLAGS_IO_BLOCKED))) {
+ atomic_dec(&fnic->in_flight);
+ atomic_dec(&tport->in_flight);
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ return SCSI_MLQUEUE_HOST_BUSY;
+ }
+
+ if (unlikely(fnic_chk_state_flags_locked(fnic, FNIC_FLAGS_FWRESET))) {
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
- "state not ready: %d/link not up: %d Returning HOST_BUSY\n",
- lp->state, lp->link_up);
+ "fnic flags FW reset: 0x%lx. Returning SCSI_MLQUEUE_HOST_BUSY\n",
+ fnic->state_flags);
return SCSI_MLQUEUE_HOST_BUSY;
}
- atomic_inc(&fnic->in_flight);
+ if (!tport->lun0_delay) {
+ lun0_delay = 1;
+ tport->lun0_delay++;
+ }
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+
fnic_priv(sc)->state = FNIC_IOREQ_NOT_INITED;
fnic_priv(sc)->flags = FNIC_NO_FLAGS;
@@ -499,6 +611,7 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
goto out;
}
+ io_req->tport = tport;
/* Determine the type of scatter/gather list we need */
io_req->sgl_cnt = sg_count;
io_req->sgl_type = FNIC_SGL_CACHE_DFLT;
@@ -575,6 +688,7 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
mempool_free(io_req, fnic->io_req_pool);
}
atomic_dec(&fnic->in_flight);
+ atomic_dec(&tport->in_flight);
return ret;
} else {
atomic64_inc(&fnic_stats->io_stats.active_ios);
@@ -602,6 +716,14 @@ int fnic_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *sc)
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
atomic_dec(&fnic->in_flight);
+ atomic_dec(&tport->in_flight);
+
+ if (lun0_delay) {
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "LUN0 delay\n");
+ mdelay(LUN0_DELAY_TIME);
+ }
+
return ret;
}
@@ -625,7 +747,7 @@ static int fnic_fcpio_fw_reset_cmpl_handler(struct fnic *fnic,
atomic64_inc(&reset_stats->fw_reset_completions);
/* Clean up all outstanding io requests */
- fnic_cleanup_io(fnic);
+ fnic_cleanup_io(fnic, SCSI_NO_TAG);
atomic64_set(&fnic->fnic_stats.fw_stats.active_fw_reqs, 0);
atomic64_set(&fnic->fnic_stats.io_stats.active_ios, 0);
@@ -646,12 +768,6 @@ static int fnic_fcpio_fw_reset_cmpl_handler(struct fnic *fnic,
"reset failed with header status: %s\n",
fnic_fcpio_status_to_str(hdr_status));
- /*
- * Unable to change to eth mode, cannot send out flogi
- * Change state to fc mode, so that subsequent Flogi
- * requests from libFC will cause more attempts to
- * reset the firmware. Free the cached flogi
- */
fnic->state = FNIC_IN_FC_MODE;
atomic64_inc(&reset_stats->fw_reset_failures);
ret = -1;
@@ -664,15 +780,14 @@ static int fnic_fcpio_fw_reset_cmpl_handler(struct fnic *fnic,
ret = -1;
}
- /* Thread removing device blocks till firmware reset is complete */
- if (fnic->remove_wait)
- complete(fnic->remove_wait);
+ if (fnic->fw_reset_done)
+ complete(fnic->fw_reset_done);
/*
* If fnic is being removed, or fw reset failed
* free the flogi frame. Else, send it out
*/
- if (fnic->remove_wait || ret) {
+ if (ret) {
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
fnic_free_txq(&fnic->tx_queue);
goto reset_cmpl_handler_end;
@@ -711,12 +826,12 @@ static int fnic_fcpio_flogi_reg_cmpl_handler(struct fnic *fnic,
/* Check flogi registration completion status */
if (!hdr_status) {
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "flog reg succeeded\n");
+ "FLOGI reg succeeded\n");
fnic->state = FNIC_IN_FC_MODE;
} else {
FNIC_SCSI_DBG(KERN_DEBUG,
fnic->lport->host, fnic->fnic_num,
- "fnic flogi reg :failed %s\n",
+ "fnic flogi reg failed: %s\n",
fnic_fcpio_status_to_str(hdr_status));
fnic->state = FNIC_IN_ETH_MODE;
ret = -1;
@@ -1023,15 +1138,6 @@ static void fnic_fcpio_icmnd_cmpl_handler(struct fnic *fnic, unsigned int cq_ind
jiffies_to_msecs(jiffies - start_time)),
desc, cmd_trace, fnic_flags_and_state(sc));
- if (sc->sc_data_direction == DMA_FROM_DEVICE) {
- fnic->lport->host_stats.fcp_input_requests++;
- fnic->fcp_input_bytes += xfer_len;
- } else if (sc->sc_data_direction == DMA_TO_DEVICE) {
- fnic->lport->host_stats.fcp_output_requests++;
- fnic->fcp_output_bytes += xfer_len;
- } else
- fnic->lport->host_stats.fcp_control_requests++;
-
/* Call SCSI completion function to complete the IO */
scsi_done(sc);
@@ -1414,8 +1520,8 @@ static bool fnic_cleanup_io_iter(struct scsi_cmnd *sc, void *data)
struct request *const rq = scsi_cmd_to_rq(sc);
struct fnic *fnic = data;
struct fnic_io_req *io_req;
- unsigned long flags = 0;
unsigned long start_time = 0;
+ unsigned long flags;
struct fnic_stats *fnic_stats = &fnic->fnic_stats;
uint16_t hwq = 0;
int tag;
@@ -1439,7 +1545,7 @@ static bool fnic_cleanup_io_iter(struct scsi_cmnd *sc, void *data)
}
if ((fnic_priv(sc)->flags & FNIC_DEVICE_RESET) &&
- !(fnic_priv(sc)->flags & FNIC_DEV_RST_DONE)) {
+ !(fnic_priv(sc)->flags & FNIC_DEV_RST_DONE)) {
/*
* We will be here only when FW completes reset
* without sending completions for outstanding ios.
@@ -1449,6 +1555,7 @@ static bool fnic_cleanup_io_iter(struct scsi_cmnd *sc, void *data)
complete(io_req->dr_done);
else if (io_req && io_req->abts_done)
complete(io_req->abts_done);
+
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
return true;
} else if (fnic_priv(sc)->flags & FNIC_DEVICE_RESET) {
@@ -1458,19 +1565,19 @@ static bool fnic_cleanup_io_iter(struct scsi_cmnd *sc, void *data)
fnic_priv(sc)->io_req = NULL;
io_req->sc = NULL;
+ start_time = io_req->start_time;
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
/*
* If there is a scsi_cmnd associated with this io_req, then
* free the corresponding state
*/
- start_time = io_req->start_time;
fnic_release_ioreq_buf(fnic, io_req, sc);
mempool_free(io_req, fnic->io_req_pool);
sc->result = DID_TRANSPORT_DISRUPTED << 16;
FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
- "mqtag:0x%x tag: 0x%x sc:0x%p duration = %lu DID_TRANSPORT_DISRUPTED\n",
+ "mqtag: 0x%x tag: 0x%x sc: 0x%p duration = %lu DID_TRANSPORT_DISRUPTED\n",
mqtag, tag, sc, (jiffies - start_time));
if (atomic64_read(&fnic->io_cmpl_skip))
@@ -1479,23 +1586,60 @@ static bool fnic_cleanup_io_iter(struct scsi_cmnd *sc, void *data)
atomic64_inc(&fnic_stats->io_stats.io_completions);
FNIC_TRACE(fnic_cleanup_io,
- sc->device->host->host_no, tag, sc,
- jiffies_to_msecs(jiffies - start_time),
- 0, ((u64)sc->cmnd[0] << 32 |
- (u64)sc->cmnd[2] << 24 |
- (u64)sc->cmnd[3] << 16 |
- (u64)sc->cmnd[4] << 8 | sc->cmnd[5]),
- fnic_flags_and_state(sc));
-
+ sc->device->host->host_no, tag, sc,
+ jiffies_to_msecs(jiffies - start_time),
+ 0, ((u64) sc->cmnd[0] << 32 |
+ (u64) sc->cmnd[2] << 24 |
+ (u64) sc->cmnd[3] << 16 |
+ (u64) sc->cmnd[4] << 8 | sc->cmnd[5]),
+ (((u64) fnic_priv(sc)->flags << 32) | fnic_priv(sc)->
+ state));
+
+ /* Complete the command to SCSI */
scsi_done(sc);
-
return true;
}
-static void fnic_cleanup_io(struct fnic *fnic)
+static void fnic_cleanup_io(struct fnic *fnic, int exclude_id)
{
+ unsigned int io_count = 0;
+ unsigned long flags;
+ struct fnic_io_req *io_req = NULL;
+ struct scsi_cmnd *sc = NULL;
+
+ io_count = fnic_count_all_ioreqs(fnic);
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "Outstanding ioreq count: %d active io count: %lld Waiting\n",
+ io_count,
+ atomic64_read(&fnic->fnic_stats.io_stats.active_ios));
+
scsi_host_busy_iter(fnic->lport->host,
- fnic_cleanup_io_iter, fnic);
+ fnic_cleanup_io_iter, fnic);
+
+ /* with sg3utils device reset, SC needs to be retrieved from ioreq */
+ spin_lock_irqsave(&fnic->wq_copy_lock[0], flags);
+ io_req = fnic->sw_copy_wq[0].io_req_table[fnic->fnic_max_tag_id];
+ if (io_req) {
+ sc = io_req->sc;
+ if (sc) {
+ if ((fnic_priv(sc)->flags & FNIC_DEVICE_RESET)
+ && !(fnic_priv(sc)->flags & FNIC_DEV_RST_DONE)) {
+ fnic_priv(sc)->flags |= FNIC_DEV_RST_DONE;
+ if (io_req && io_req->dr_done)
+ complete(io_req->dr_done);
+ }
+ }
+ }
+ spin_unlock_irqrestore(&fnic->wq_copy_lock[0], flags);
+
+ while ((io_count = fnic_count_all_ioreqs(fnic))) {
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "Outstanding ioreq count: %d active io count: %lld Waiting\n",
+ io_count,
+ atomic64_read(&fnic->fnic_stats.io_stats.active_ios));
+
+ schedule_timeout(msecs_to_jiffies(100));
+ }
}
void fnic_wq_copy_cleanup_handler(struct vnic_wq_copy *wq,
@@ -1567,10 +1711,13 @@ static inline int fnic_queue_abort_io_req(struct fnic *fnic, int tag,
struct vnic_wq_copy *wq = &fnic->hw_copy_wq[hwq];
struct misc_stats *misc_stats = &fnic->fnic_stats.misc_stats;
unsigned long flags;
+ struct fnic_tport_s *tport = io_req->tport;
spin_lock_irqsave(&fnic->fnic_lock, flags);
if (unlikely(fnic_chk_state_flags_locked(fnic,
FNIC_FLAGS_IO_BLOCKED))) {
+ atomic_dec(&fnic->in_flight);
+ atomic_dec(&tport->in_flight);
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
return 1;
} else
@@ -1585,6 +1732,7 @@ static inline int fnic_queue_abort_io_req(struct fnic *fnic, int tag,
if (!vnic_wq_copy_desc_avail(wq)) {
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
atomic_dec(&fnic->in_flight);
+ atomic_dec(&tport->in_flight);
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
"fnic_queue_abort_io_req: failure: no descriptors\n");
atomic64_inc(&misc_stats->abts_cpwq_alloc_failures);
@@ -1619,20 +1767,24 @@ static bool fnic_rport_abort_io_iter(struct scsi_cmnd *sc, void *data)
struct fnic *fnic = iter_data->fnic;
int abt_tag = 0;
struct fnic_io_req *io_req;
- unsigned long flags;
struct reset_stats *reset_stats = &fnic->fnic_stats.reset_stats;
struct terminate_stats *term_stats = &fnic->fnic_stats.term_stats;
struct scsi_lun fc_lun;
enum fnic_ioreq_state old_ioreq_state;
uint16_t hwq = 0;
+ unsigned long flags;
abt_tag = blk_mq_unique_tag(rq);
hwq = blk_mq_unique_tag_to_hwq(abt_tag);
- spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
+ if (!sc) {
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "sc is NULL abt_tag: 0x%x hwq: %d\n", abt_tag, hwq);
+ return true;
+ }
+ spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
io_req = fnic_priv(sc)->io_req;
-
if (!io_req || io_req->port_id != iter_data->port_id) {
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
return true;
@@ -1655,37 +1807,38 @@ static bool fnic_rport_abort_io_iter(struct scsi_cmnd *sc, void *data)
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
return true;
}
+
if (io_req->abts_done) {
shost_printk(KERN_ERR, fnic->lport->host,
- "fnic_rport_exch_reset: io_req->abts_done is set "
- "state is %s\n",
+ "fnic_rport_exch_reset: io_req->abts_done is set state is %s\n",
fnic_ioreq_state_to_str(fnic_priv(sc)->state));
}
if (!(fnic_priv(sc)->flags & FNIC_IO_ISSUED)) {
shost_printk(KERN_ERR, fnic->lport->host,
- "rport_exch_reset "
- "IO not yet issued %p tag 0x%x flags "
- "%x state %d\n",
- sc, abt_tag, fnic_priv(sc)->flags, fnic_priv(sc)->state);
+ "rport_exch_reset IO not yet issued %p abt_tag 0x%x",
+ sc, abt_tag);
+ shost_printk(KERN_ERR, fnic->lport->host,
+ "flags %x state %d\n", fnic_priv(sc)->flags,
+ fnic_priv(sc)->state);
}
old_ioreq_state = fnic_priv(sc)->state;
fnic_priv(sc)->state = FNIC_IOREQ_ABTS_PENDING;
fnic_priv(sc)->abts_status = FCPIO_INVALID_CODE;
+
if (fnic_priv(sc)->flags & FNIC_DEVICE_RESET) {
atomic64_inc(&reset_stats->device_reset_terminates);
abt_tag |= FNIC_TAG_DEV_RST;
}
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "fnic_rport_exch_reset dev rst sc 0x%p\n", sc);
- BUG_ON(io_req->abts_done);
-
+ "fnic_rport_exch_reset: dev rst sc 0x%p\n", sc);
+ WARN_ON_ONCE(io_req->abts_done);
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
"fnic_rport_reset_exch: Issuing abts\n");
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
- /* Now queue the abort command to firmware */
+ /* Queue the abort command to firmware */
int_to_scsilun(sc->device->lun, &fc_lun);
if (fnic_queue_abort_io_req(fnic, abt_tag,
@@ -1714,11 +1867,14 @@ static bool fnic_rport_abort_io_iter(struct scsi_cmnd *sc, void *data)
atomic64_inc(&term_stats->terminates);
iter_data->term_cnt++;
}
+
return true;
}
void fnic_rport_exch_reset(struct fnic *fnic, u32 port_id)
{
+ unsigned int io_count = 0;
+ unsigned long flags;
struct terminate_stats *term_stats = &fnic->fnic_stats.term_stats;
struct fnic_rport_abort_io_iter_data iter_data = {
.fnic = fnic,
@@ -1726,53 +1882,75 @@ void fnic_rport_exch_reset(struct fnic *fnic, u32 port_id)
.term_cnt = 0,
};
- FNIC_SCSI_DBG(KERN_DEBUG,
- fnic->lport->host, fnic->fnic_num,
- "fnic_rport_exch_reset called portid 0x%06x\n",
- port_id);
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "fnic rport exchange reset for tport: 0x%06x\n",
+ port_id);
if (fnic->in_remove)
return;
+ io_count = fnic_count_ioreqs(fnic, port_id);
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "Starting terminates: rport:0x%x portid-io-count: %d active-io-count: %lld\n",
+ port_id, io_count,
+ atomic64_read(&fnic->fnic_stats.io_stats.active_ios));
+
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
+ atomic_inc(&fnic->in_flight);
+ if (unlikely(fnic_chk_state_flags_locked(fnic, FNIC_FLAGS_IO_BLOCKED))) {
+ atomic_dec(&fnic->in_flight);
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ return;
+ }
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+
scsi_host_busy_iter(fnic->lport->host, fnic_rport_abort_io_iter,
&iter_data);
+
if (iter_data.term_cnt > atomic64_read(&term_stats->max_terminates))
atomic64_set(&term_stats->max_terminates, iter_data.term_cnt);
+ atomic_dec(&fnic->in_flight);
+
+ while ((io_count = fnic_count_ioreqs(fnic, port_id)))
+ schedule_timeout(msecs_to_jiffies(1000));
+
That really feels overly complex.
You send the aborts with the previous 'scsi_host_busy_iter()' call, and
now you wait for the I/Os to complete.
But you already have an atomic counter for I/Os in flight to the tport;
can't you just wait for that counter to drop to zero?
+ FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
+ "rport: 0x%x remaining portid-io-count: %d ",
+ port_id, io_count);
}
void fnic_terminate_rport_io(struct fc_rport *rport)
{
- struct fc_rport_libfc_priv *rdata;
- struct fc_lport *lport;
- struct fnic *fnic;
+ struct fnic_tport_s *tport;
+ struct rport_dd_data_s *rdd_data;
+ struct fnic_iport_s *iport = NULL;
+ struct fnic *fnic = NULL;
if (!rport) {
- printk(KERN_ERR "fnic_terminate_rport_io: rport is NULL\n");
+ pr_err("rport is NULL\n");
return;
}
- rdata = rport->dd_data;
- if (!rdata) {
- printk(KERN_ERR "fnic_terminate_rport_io: rdata is NULL\n");
- return;
- }
- lport = rdata->local_port;
-
- if (!lport) {
- printk(KERN_ERR "fnic_terminate_rport_io: lport is NULL\n");
- return;
+ rdd_data = rport->dd_data;
+ if (rdd_data) {
+ tport = rdd_data->tport;
+ if (!tport) {
+ pr_err(
+ "term rport io called after tport is deleted. Returning 0x%8x\n",
+ rport->port_id);
+ } else {
+ pr_err(
+ "term rport io called after tport is set 0x%8x\n",
+ rport->port_id);
+ pr_err(
+ "tport maybe rediscovered\n");
+
+ iport = (struct fnic_iport_s *) tport->iport;
+ fnic = iport->fnic;
+ fnic_rport_exch_reset(fnic, rport->port_id);
+ }
}
- fnic = lport_priv(lport);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "wwpn 0x%llx, wwnn0x%llx, rport 0x%p, portid 0x%06x\n",
- rport->port_name, rport->node_name, rport,
- rport->port_id);
-
- if (fnic->in_remove)
- return;
-
- fnic_rport_exch_reset(fnic, rport->port_id);
}
/*
@@ -1783,10 +1961,12 @@ void fnic_terminate_rport_io(struct fc_rport *rport)
int fnic_abort_cmd(struct scsi_cmnd *sc)
{
struct request *const rq = scsi_cmd_to_rq(sc);
- struct fc_lport *lp;
+ struct fnic_iport_s *iport;
+ struct fnic_tport_s *tport;
struct fnic *fnic;
struct fnic_io_req *io_req = NULL;
struct fc_rport *rport;
+ struct rport_dd_data_s *rdd_data;
unsigned long flags;
unsigned long start_time = 0;
int ret = SUCCESS;
@@ -1806,11 +1986,11 @@ int fnic_abort_cmd(struct scsi_cmnd *sc)
fc_block_scsi_eh(sc);
/* Get local-port, check ready and link up */
- lp = shost_priv(sc->device->host);
-
- fnic = lport_priv(lp);
+ fnic = *((struct fnic **) shost_priv(sc->device->host));
spin_lock_irqsave(&fnic->fnic_lock, flags);
+ iport = &fnic->iport;
+
fnic_stats = &fnic->fnic_stats;
abts_stats = &fnic->fnic_stats.abts_stats;
term_stats = &fnic->fnic_stats.term_stats;
@@ -1821,7 +2001,43 @@ int fnic_abort_cmd(struct scsi_cmnd *sc)
fnic_priv(sc)->flags = FNIC_NO_FLAGS;
- if (lp->state != LPORT_ST_READY || !(lp->link_up)) {
+ rdd_data = rport->dd_data;
+ tport = rdd_data->tport;
+
+ if (!tport) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "Abort cmd called after tport delete! rport fcid: 0x%x",
+ rport->port_id);
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "lun: %llu hwq: 0x%x mqtag: 0x%x Op: 0x%x flags: 0x%x\n",
+ sc->device->lun, hwq, mqtag,
+ sc->cmnd[0], fnic_priv(sc)->flags);
+ ret = FAILED;
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ goto fnic_abort_cmd_end;
+ }
+
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "Abort cmd called rport fcid: 0x%x lun: %llu hwq: 0x%x mqtag: 0x%x",
+ rport->port_id, sc->device->lun, hwq, mqtag);
+
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "Op: 0x%x flags: 0x%x\n",
+ sc->cmnd[0],
+ fnic_priv(sc)->flags);
+
+ if (iport->state != FNIC_IPORT_STATE_READY) {
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "iport NOT in READY state");
+ ret = FAILED;
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ goto fnic_abort_cmd_end;
+ }
+
+ if ((tport->state != FDLS_TGT_STATE_READY) &&
+ (tport->state != FDLS_TGT_STATE_ADISC)) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "tport state: %d\n", tport->state);
ret = FAILED;
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
goto fnic_abort_cmd_end;
@@ -1843,6 +2059,7 @@ int fnic_abort_cmd(struct scsi_cmnd *sc)
spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
io_req = fnic_priv(sc)->io_req;
if (!io_req) {
+ ret = FAILED;
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
goto fnic_abort_cmd_end;
}
@@ -1893,7 +2110,6 @@ int fnic_abort_cmd(struct scsi_cmnd *sc)
if (fc_remote_port_chkready(rport) == 0)
task_req = FCPIO_ITMF_ABT_TASK;
else {
- atomic64_inc(&fnic_stats->misc_stats.rport_not_ready);
task_req = FCPIO_ITMF_ABT_TASK_TERM;
}
@@ -2027,6 +2243,7 @@ static inline int fnic_queue_dr_io_req(struct fnic *fnic,
unsigned long flags;
uint16_t hwq = 0;
uint32_t tag = 0;
+ struct fnic_tport_s *tport = io_req->tport;
tag = io_req->tag;
hwq = blk_mq_unique_tag_to_hwq(tag);
@@ -2037,8 +2254,10 @@ static inline int fnic_queue_dr_io_req(struct fnic *fnic,
FNIC_FLAGS_IO_BLOCKED))) {
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
return FAILED;
- } else
+ } else {
atomic_inc(&fnic->in_flight);
+ atomic_inc(&tport->in_flight);
+ }
spin_unlock_irqrestore(&fnic->fnic_lock, flags);
spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
@@ -2072,6 +2291,7 @@ static inline int fnic_queue_dr_io_req(struct fnic *fnic,
lr_io_req_end:
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
atomic_dec(&fnic->in_flight);
+ atomic_inc(&tport->in_flight);
return ret;
}
@@ -2274,11 +2494,11 @@ static int fnic_clean_pending_aborts(struct fnic *fnic,
int fnic_device_reset(struct scsi_cmnd *sc)
{
struct request *rq = scsi_cmd_to_rq(sc);
- struct fc_lport *lp;
struct fnic *fnic;
struct fnic_io_req *io_req = NULL;
struct fc_rport *rport;
int status;
+ int count = 0;
int ret = FAILED;
unsigned long flags;
unsigned long start_time = 0;
@@ -2289,31 +2509,61 @@ int fnic_device_reset(struct scsi_cmnd *sc)
DECLARE_COMPLETION_ONSTACK(tm_done);
bool new_sc = 0;
uint16_t hwq = 0;
+ struct fnic_iport_s *iport = NULL;
+ struct rport_dd_data_s *rdd_data;
+ struct fnic_tport_s *tport;
+ u32 old_soft_reset_count;
+ u32 old_link_down_cnt;
+ int exit_dr = 0;
/* Wait for rport to unblock */
fc_block_scsi_eh(sc);
/* Get local-port, check ready and link up */
- lp = shost_priv(sc->device->host);
+ fnic = *((struct fnic **) shost_priv(sc->device->host));
+ iport = &fnic->iport;
- fnic = lport_priv(lp);
fnic_stats = &fnic->fnic_stats;
reset_stats = &fnic->fnic_stats.reset_stats;
atomic64_inc(&reset_stats->device_resets);
rport = starget_to_rport(scsi_target(sc->device));
+
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "fcid: 0x%x lun: 0x%llx hwq: %d mqtag: 0x%x flags: 0x%x Device reset\n",
+ "fcid: 0x%x lun: %llu hwq: %d mqtag: 0x%x flags: 0x%x Device reset\n",
rport->port_id, sc->device->lun, hwq, mqtag,
fnic_priv(sc)->flags);
- if (lp->state != LPORT_ST_READY || !(lp->link_up))
+ rdd_data = rport->dd_data;
+ tport = rdd_data->tport;
+ if (!tport) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "Dev rst called after tport delete! rport fcid: 0x%x lun: %llu\n",
+ rport->port_id, sc->device->lun);
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ goto fnic_device_reset_end;
+ }
+
+ if (iport->state != FNIC_IPORT_STATE_READY) {
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "iport NOT in READY state");
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
goto fnic_device_reset_end;
+ }
+
+ if ((tport->state != FDLS_TGT_STATE_READY) &&
+ (tport->state != FDLS_TGT_STATE_ADISC)) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "tport state: %d\n", tport->state);
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ goto fnic_device_reset_end;
+ }
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
Please check if returning FAST_IO_FAIL here wouldn't be a better option.
Most of the time a device reset is triggered by a command timeout, which
typically happens due to a transport issue (eg link is down or something).
So returning 'FAILED' will just escalate to host reset, and that can
take for a really long time trying to abort all commands.
/* Check if remote port up */
if (fc_remote_port_chkready(rport)) {
- atomic64_inc(&fnic_stats->misc_stats.rport_not_ready);
goto fnic_device_reset_end;
}
@@ -2352,6 +2602,7 @@ int fnic_device_reset(struct scsi_cmnd *sc)
io_req->port_id = rport->port_id;
io_req->tag = mqtag;
fnic_priv(sc)->io_req = io_req;
+ io_req->tport = tport;
io_req->sc = sc;
if (fnic->sw_copy_wq[hwq].io_req_table[blk_mq_unique_tag_to_tag(mqtag)] != NULL)
@@ -2383,6 +2634,11 @@ int fnic_device_reset(struct scsi_cmnd *sc)
fnic_priv(sc)->flags |= FNIC_DEV_RST_ISSUED;
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
+ old_link_down_cnt = iport->fnic->link_down_cnt;
+ old_soft_reset_count = fnic->soft_reset_count;
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+
/*
* Wait on the local completion for LUN reset. The io_req may be
* freed while we wait since we hold no lock.
@@ -2390,6 +2646,24 @@ int fnic_device_reset(struct scsi_cmnd *sc)
wait_for_completion_timeout(&tm_done,
msecs_to_jiffies(FNIC_LUN_RESET_TIMEOUT));
+ /*
+ * Wake up can be due to the following reasons:
+ * 1) The device reset completed from target.
+ * 2) Device reset timed out.
+ * 3) A link-down/host_reset may have happened in between.
+ * 4) The device reset was aborted and io_req->dr_done was called.
+ */
+
+ exit_dr = 0;
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
+ if ((old_link_down_cnt != fnic->link_down_cnt) ||
+ (fnic->reset_in_progress) ||
+ (fnic->soft_reset_count != old_soft_reset_count) ||
+ (iport->state != FNIC_IPORT_STATE_READY))
+ exit_dr = 1;
+
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+
spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
io_req = fnic_priv(sc)->io_req;
if (!io_req) {
@@ -2398,6 +2672,13 @@ int fnic_device_reset(struct scsi_cmnd *sc)
"io_req is null mqtag 0x%x sc 0x%p\n", mqtag, sc);
goto fnic_device_reset_end;
}
+
+ if (exit_dr) {
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "Host reset called for fnic. Exit device reset\n");
+ io_req->dr_done = NULL;
+ goto fnic_device_reset_clean;
+ }
io_req->dr_done = NULL;
status = fnic_priv(sc)->lr_status;
@@ -2411,50 +2692,8 @@ int fnic_device_reset(struct scsi_cmnd *sc)
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
"Device reset timed out\n");
fnic_priv(sc)->flags |= FNIC_DEV_RST_TIMED_OUT;
- spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
int_to_scsilun(sc->device->lun, &fc_lun);
- /*
- * Issue abort and terminate on device reset request.
- * If q'ing of terminate fails, retry it after a delay.
- */
- while (1) {
- spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
- if (fnic_priv(sc)->flags & FNIC_DEV_RST_TERM_ISSUED) {
- spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
- break;
- }
- spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
- if (fnic_queue_abort_io_req(fnic,
- mqtag | FNIC_TAG_DEV_RST,
- FCPIO_ITMF_ABT_TASK_TERM,
- fc_lun.scsi_lun, io_req, hwq)) {
- wait_for_completion_timeout(&tm_done,
- msecs_to_jiffies(FNIC_ABT_TERM_DELAY_TIMEOUT));
- } else {
- spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
- fnic_priv(sc)->flags |= FNIC_DEV_RST_TERM_ISSUED;
- fnic_priv(sc)->state = FNIC_IOREQ_ABTS_PENDING;
- io_req->abts_done = &tm_done;
- spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "Abort and terminate issued on Device reset mqtag 0x%x sc 0x%p\n",
- mqtag, sc);
- break;
- }
- }
- while (1) {
- spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
- if (!(fnic_priv(sc)->flags & FNIC_DEV_RST_DONE)) {
- spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
- wait_for_completion_timeout(&tm_done,
- msecs_to_jiffies(FNIC_LUN_RESET_TIMEOUT));
- break;
- } else {
- io_req = fnic_priv(sc)->io_req;
- io_req->abts_done = NULL;
- goto fnic_device_reset_clean;
- }
- }
+ goto fnic_device_reset_clean;
} else {
spin_unlock_irqrestore(&fnic->wq_copy_lock[hwq], flags);
}
@@ -2480,8 +2719,7 @@ int fnic_device_reset(struct scsi_cmnd *sc)
spin_lock_irqsave(&fnic->wq_copy_lock[hwq], flags);
io_req = fnic_priv(sc)->io_req;
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "Device reset failed"
- " since could not abort all IOs\n");
+ "Device reset failed: Cannot abort all IOs\n");
goto fnic_device_reset_clean;
}
@@ -2507,6 +2745,15 @@ int fnic_device_reset(struct scsi_cmnd *sc)
mempool_free(io_req, fnic->io_req_pool);
}
+ /*
+ * If link-event is seen while LUN reset is issued we need
+ * to complete the LUN reset here
+ */
+ if (!new_sc) {
+ sc->result = DID_RESET << 16;
+ scsi_done(sc);
+ }
+
fnic_device_reset_end:
FNIC_TRACE(fnic_device_reset, sc->device->host->host_no, rq->tag, sc,
jiffies_to_msecs(jiffies - start_time),
@@ -2520,6 +2767,17 @@ int fnic_device_reset(struct scsi_cmnd *sc)
mutex_unlock(&fnic->sgreset_mutex);
}
+ while ((ret == SUCCESS) && fnic_count_lun_ioreqs(fnic, sc->device)) {
+ if (count >= 2) {
+ ret = FAILED;
+ break;
+ }
+ FNIC_SCSI_DBG(KERN_ERR, fnic->lport->host, fnic->fnic_num,
+ "Cannot clean up all IOs for the LUN\n");
+ schedule_timeout(msecs_to_jiffies(1000));
+ count++;
+ }
+
FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
"Returning from device reset %s\n",
(ret == SUCCESS) ?
@@ -2531,50 +2789,54 @@ int fnic_device_reset(struct scsi_cmnd *sc)
return ret;
}
-/* Clean up all IOs, clean up libFC local port */
-int fnic_reset(struct Scsi_Host *shost)
+static void fnic_post_flogo_linkflap(struct fnic *fnic)
+{
+ unsigned long flags;
+
+ fnic_fdls_link_status_change(fnic, 0);
+ spin_lock_irqsave(&fnic->fnic_lock, flags);
+
+ if (fnic->link_status) {
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+ fnic_fdls_link_status_change(fnic, 1);
+ return;
+ }
+ spin_unlock_irqrestore(&fnic->fnic_lock, flags);
+}
+
+/* Logout from all the targets and simulate link flap */
+void fnic_reset(struct Scsi_Host *shost)
{
- struct fc_lport *lp;
struct fnic *fnic;
- int ret = 0;
struct reset_stats *reset_stats;
- lp = shost_priv(shost);
- fnic = lport_priv(lp);
+ fnic = *((struct fnic **) shost_priv(shost));
reset_stats = &fnic->fnic_stats.reset_stats;
FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
- "Issuing fnic reset\n");
+ "Issuing fnic reset\n");
atomic64_inc(&reset_stats->fnic_resets);
-
- /*
- * Reset local port, this will clean up libFC exchanges,
- * reset remote port sessions, and if link is up, begin flogi
- */
- ret = fc_lport_reset(lp);
+ fnic_post_flogo_linkflap(fnic);
FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
- "Returning from fnic reset with: %s\n",
- (ret == 0) ? "SUCCESS" : "FAILED");
+ "Returning from fnic reset");
- if (ret == 0)
- atomic64_inc(&reset_stats->fnic_reset_completions);
- else
- atomic64_inc(&reset_stats->fnic_reset_failures);
+ atomic64_inc(&reset_stats->fnic_reset_completions);
+}
+
+int fnic_issue_fc_host_lip(struct Scsi_Host *shost)
+{
+ int ret = 0;
+ struct fnic *fnic = *((struct fnic **) shost_priv(shost));
+
+ FNIC_SCSI_DBG(KERN_INFO, fnic->lport->host, fnic->fnic_num,
+ "FC host lip issued");
+ ret = fnic_host_reset(shost);
return ret;
}
-/*
- * SCSI Error handling calls driver's eh_host_reset if all prior
- * error handling levels return FAILED. If host reset completes
- * successfully, and if link is up, then Fabric login begins.
- *
- * Host Reset is the highest level of error recovery. If this fails, then
- * host is offlined by SCSI.
- *
- */
int fnic_host_reset(struct Scsi_Host *shost)
{
int ret = SUCCESS;
@@ -2637,122 +2899,6 @@ int fnic_host_reset(struct Scsi_Host *shost)
return ret;
}
-/*
- * This fxn is called from libFC when host is removed
- */
-void fnic_scsi_abort_io(struct fc_lport *lp)
-{
- int err = 0;
- unsigned long flags;
- enum fnic_state old_state;
- struct fnic *fnic = lport_priv(lp);
- DECLARE_COMPLETION_ONSTACK(remove_wait);
-
- /* Issue firmware reset for fnic, wait for reset to complete */
-retry_fw_reset:
- spin_lock_irqsave(&fnic->fnic_lock, flags);
- if (unlikely(fnic->state == FNIC_IN_FC_TRANS_ETH_MODE) &&
- fnic->link_events) {
- /* fw reset is in progress, poll for its completion */
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- schedule_timeout(msecs_to_jiffies(100));
- goto retry_fw_reset;
- }
-
- fnic->remove_wait = &remove_wait;
- old_state = fnic->state;
- fnic->state = FNIC_IN_FC_TRANS_ETH_MODE;
- fnic_update_mac_locked(fnic, fnic->ctlr.ctl_src_addr);
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
-
- err = fnic_fw_reset_handler(fnic);
- if (err) {
- spin_lock_irqsave(&fnic->fnic_lock, flags);
- if (fnic->state == FNIC_IN_FC_TRANS_ETH_MODE)
- fnic->state = old_state;
- fnic->remove_wait = NULL;
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- return;
- }
-
- /* Wait for firmware reset to complete */
- wait_for_completion_timeout(&remove_wait,
- msecs_to_jiffies(FNIC_RMDEVICE_TIMEOUT));
-
- spin_lock_irqsave(&fnic->fnic_lock, flags);
- fnic->remove_wait = NULL;
- FNIC_SCSI_DBG(KERN_DEBUG, fnic->lport->host, fnic->fnic_num,
- "fnic_scsi_abort_io %s\n",
- (fnic->state == FNIC_IN_ETH_MODE) ?
- "SUCCESS" : "FAILED");
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
-
-}
-
-/*
- * This fxn called from libFC to clean up driver IO state on link down
- */
-void fnic_scsi_cleanup(struct fc_lport *lp)
-{
- unsigned long flags;
- enum fnic_state old_state;
- struct fnic *fnic = lport_priv(lp);
-
- /* issue fw reset */
-retry_fw_reset:
- spin_lock_irqsave(&fnic->fnic_lock, flags);
- if (unlikely(fnic->state == FNIC_IN_FC_TRANS_ETH_MODE)) {
- /* fw reset is in progress, poll for its completion */
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- schedule_timeout(msecs_to_jiffies(100));
- goto retry_fw_reset;
- }
- old_state = fnic->state;
- fnic->state = FNIC_IN_FC_TRANS_ETH_MODE;
- fnic_update_mac_locked(fnic, fnic->ctlr.ctl_src_addr);
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
-
- if (fnic_fw_reset_handler(fnic)) {
- spin_lock_irqsave(&fnic->fnic_lock, flags);
- if (fnic->state == FNIC_IN_FC_TRANS_ETH_MODE)
- fnic->state = old_state;
- spin_unlock_irqrestore(&fnic->fnic_lock, flags);
- }
-
-}
-
-void fnic_empty_scsi_cleanup(struct fc_lport *lp)
-{
-}
-
-void fnic_exch_mgr_reset(struct fc_lport *lp, u32 sid, u32 did)
-{
- struct fnic *fnic = lport_priv(lp);
-
- /* Non-zero sid, nothing to do */
- if (sid)
- goto call_fc_exch_mgr_reset;
-
- if (did) {
- fnic_rport_exch_reset(fnic, did);
- goto call_fc_exch_mgr_reset;
- }
-
- /*
- * sid = 0, did = 0
- * link down or device being removed
- */
- if (!fnic->in_remove)
- fnic_scsi_cleanup(lp);
- else
- fnic_scsi_abort_io(lp);
-
- /* call libFC exch mgr reset to reset its exchanges */
-call_fc_exch_mgr_reset:
- fc_exch_mgr_reset(lp, sid, did);
-
-}
-
static bool fnic_abts_pending_iter(struct scsi_cmnd *sc, void *data)
{
struct request *const rq = scsi_cmd_to_rq(sc);
diff --git a/drivers/scsi/fnic/fnic_stats.h b/drivers/scsi/fnic/fnic_stats.h
index 9d7f98c452dd..1f1a1ec90a23 100644
--- a/drivers/scsi/fnic/fnic_stats.h
+++ b/drivers/scsi/fnic/fnic_stats.h
@@ -127,6 +127,4 @@ struct stats_debug_info {
};
int fnic_get_stats_data(struct stats_debug_info *, struct fnic_stats *);
-void fnic_stats_debugfs_init(struct fnic *);
-void fnic_stats_debugfs_remove(struct fnic *);
#endif /* _FNIC_STATS_H_ */
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@xxxxxxx +49 911 74053 688
SUSE Software Solutions GmbH, Frankenstr. 146, 90461 Nürnberg
HRB 36809 (AG Nürnberg), GF: I. Totev, A. McDonald, W. Knoblich
