be2iscsi: Fix IOPOLL implementation

OS not responding when running 2 port traffic on 72 CPUs system.

be2iscsi IRQs gets affined to CPU0 when irqbalancer is disabled.
be_iopoll processing completions in BLOCK_IOPOLL_SOFTIRQ hogged CPU0.

1. Use budget to exit the polling loop. beiscsi_process_cq didn't honour
   it.
2. Rearming of EQ is done only after iopoll completes.

[mkp: Fixed up blk_iopoll -> irq_poll transition]

Signed-off-by: Jitendra Bhivare <jitendra.bhivare@avagotech.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

diff --git a/drivers/scsi/be2iscsi/be_cmds.c b/drivers/scsi/be2iscsi/be_cmds.c
index dbe62c0..14a1c71 100644 (file)
--- a/drivers/scsi/be2iscsi/be_cmds.c
+++ b/drivers/scsi/be2iscsi/be_cmds.c
@@ -546,7 +546,7 @@ int beiscsi_process_mcc(struct beiscsi_hba *phba)
        }
 
        if (num)
-               hwi_ring_cq_db(phba, phba->ctrl.mcc_obj.cq.id, num, 1, 0);
+               hwi_ring_cq_db(phba, phba->ctrl.mcc_obj.cq.id, num, 1);
 
        spin_unlock_bh(&phba->ctrl.mcc_cq_lock);
        return status;

diff --git a/drivers/scsi/be2iscsi/be_iscsi.c b/drivers/scsi/be2iscsi/be_iscsi.c
index 52aab12..8bf7379 100644 (file)
--- a/drivers/scsi/be2iscsi/be_iscsi.c
+++ b/drivers/scsi/be2iscsi/be_iscsi.c
@@ -1298,7 +1298,7 @@ static void beiscsi_flush_cq(struct beiscsi_hba *phba)
        for (i = 0; i < phba->num_cpus; i++) {
                pbe_eq = &phwi_context->be_eq[i];
                irq_poll_disable(&pbe_eq->iopoll);
-               beiscsi_process_cq(pbe_eq);
+               beiscsi_process_cq(pbe_eq, BE2_MAX_NUM_CQ_PROC);
                irq_poll_enable(&pbe_eq->iopoll);
        }
 }

diff --git a/drivers/scsi/be2iscsi/be_main.c b/drivers/scsi/be2iscsi/be_main.c
index bdedcbb..9c3a7b2 100644 (file)
--- a/drivers/scsi/be2iscsi/be_main.c
+++ b/drivers/scsi/be2iscsi/be_main.c
@@ -29,6 +29,7 @@
 #include <linux/iscsi_boot_sysfs.h>
 #include <linux/module.h>
 #include <linux/bsg-lib.h>
+#include <linux/irq_poll.h>
 
 #include <scsi/libiscsi.h>
 #include <scsi/scsi_bsg_iscsi.h>
@@ -895,31 +896,17 @@ static irqreturn_t be_isr_mcc(int irq, void *dev_id)
 static irqreturn_t be_isr_msix(int irq, void *dev_id)
 {
        struct beiscsi_hba *phba;
-       struct be_eq_entry *eqe = NULL;
        struct be_queue_info *eq;
-       struct be_queue_info *cq;
-       unsigned int num_eq_processed;
        struct be_eq_obj *pbe_eq;
 
        pbe_eq = dev_id;
        eq = &pbe_eq->q;
-       cq = pbe_eq->cq;
-       eqe = queue_tail_node(eq);
 
        phba = pbe_eq->phba;
-       num_eq_processed = 0;
-       while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32]
-                               & EQE_VALID_MASK) {
-               irq_poll_sched(&pbe_eq->iopoll);
 
-               AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
-               queue_tail_inc(eq);
-               eqe = queue_tail_node(eq);
-               num_eq_processed++;
-       }
-
-       if (num_eq_processed)
-               hwi_ring_eq_db(phba, eq->id, 1, num_eq_processed, 0, 1);
+       /* disable interrupt till iopoll completes */
+       hwi_ring_eq_db(phba, eq->id, 1, 0, 0, 1);
+       irq_poll_sched(&pbe_eq->iopoll);
 
        return IRQ_HANDLED;
 }
@@ -996,6 +983,7 @@ static irqreturn_t be_isr(int irq, void *dev_id)
                return IRQ_NONE;
 }
 
+
 static int beiscsi_init_irqs(struct beiscsi_hba *phba)
 {
        struct pci_dev *pcidev = phba->pcidev;
@@ -1070,7 +1058,7 @@ free_msix_irqs:
 
 void hwi_ring_cq_db(struct beiscsi_hba *phba,
                           unsigned int id, unsigned int num_processed,
-                          unsigned char rearm, unsigned char event)
+                          unsigned char rearm)
 {
        u32 val = 0;
 
@@ -2042,7 +2030,7 @@ static void  beiscsi_process_mcc_isr(struct beiscsi_hba *phba)
 
                if (num_processed >= 32) {
                        hwi_ring_cq_db(phba, mcc_cq->id,
-                                       num_processed, 0, 0);
+                                       num_processed, 0);
                        num_processed = 0;
                }
                if (mcc_compl->flags & CQE_FLAGS_ASYNC_MASK) {
@@ -2060,24 +2048,25 @@ static void  beiscsi_process_mcc_isr(struct beiscsi_hba *phba)
        }
 
        if (num_processed > 0)
-               hwi_ring_cq_db(phba, mcc_cq->id, num_processed, 1, 0);
+               hwi_ring_cq_db(phba, mcc_cq->id, num_processed, 1);
 
 }
 
 /**
  * beiscsi_process_cq()- Process the Completion Queue
  * @pbe_eq: Event Q on which the Completion has come
+ * @budget: Max number of events to processed
  *
  * return
  *     Number of Completion Entries processed.
  **/
-unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq)
+unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq, int budget)
 {
        struct be_queue_info *cq;
        struct sol_cqe *sol;
        struct dmsg_cqe *dmsg;
+       unsigned int total = 0;
        unsigned int num_processed = 0;
-       unsigned int tot_nump = 0;
        unsigned short code = 0, cid = 0;
        uint16_t cri_index = 0;
        struct beiscsi_conn *beiscsi_conn;
@@ -2128,12 +2117,12 @@ unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq)
                beiscsi_ep = ep->dd_data;
                beiscsi_conn = beiscsi_ep->conn;
 
-               if (num_processed >= 32) {
-                       hwi_ring_cq_db(phba, cq->id,
-                                       num_processed, 0, 0);
-                       tot_nump += num_processed;
+               /* replenish cq */
+               if (num_processed == 32) {
+                       hwi_ring_cq_db(phba, cq->id, 32, 0);
                        num_processed = 0;
                }
+               total++;
 
                switch (code) {
                case SOL_CMD_COMPLETE:
@@ -2178,7 +2167,13 @@ unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq)
                                    "BM_%d : Ignoring %s[%d] on CID : %d\n",
                                    cqe_desc[code], code, cid);
                        break;
+               case CXN_KILLED_HDR_DIGEST_ERR:
                case SOL_CMD_KILLED_DATA_DIGEST_ERR:
+                       beiscsi_log(phba, KERN_ERR,
+                                   BEISCSI_LOG_CONFIG | BEISCSI_LOG_IO,
+                                   "BM_%d : Cmd Notification %s[%d] on CID : %d\n",
+                                   cqe_desc[code], code,  cid);
+                       break;
                case CMD_KILLED_INVALID_STATSN_RCVD:
                case CMD_KILLED_INVALID_R2T_RCVD:
                case CMD_CXN_KILLED_LUN_INVALID:
@@ -2204,7 +2199,6 @@ unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq)
                case CXN_KILLED_PDU_SIZE_EXCEEDS_DSL:
                case CXN_KILLED_BURST_LEN_MISMATCH:
                case CXN_KILLED_AHS_RCVD:
-               case CXN_KILLED_HDR_DIGEST_ERR:
                case CXN_KILLED_UNKNOWN_HDR:
                case CXN_KILLED_STALE_ITT_TTT_RCVD:
                case CXN_KILLED_INVALID_ITT_TTT_RCVD:
@@ -2239,13 +2233,12 @@ proc_next_cqe:
                queue_tail_inc(cq);
                sol = queue_tail_node(cq);
                num_processed++;
+               if (total == budget)
+                       break;
        }
 
-       if (num_processed > 0) {
-               tot_nump += num_processed;
-               hwi_ring_cq_db(phba, cq->id, num_processed, 1, 0);
-       }
-       return tot_nump;
+       hwi_ring_cq_db(phba, cq->id, num_processed, 1);
+       return total;
 }
 
 void beiscsi_process_all_cqs(struct work_struct *work)
@@ -2272,7 +2265,7 @@ void beiscsi_process_all_cqs(struct work_struct *work)
                spin_lock_irqsave(&phba->isr_lock, flags);
                pbe_eq->todo_cq = false;
                spin_unlock_irqrestore(&phba->isr_lock, flags);
-               beiscsi_process_cq(pbe_eq);
+               beiscsi_process_cq(pbe_eq, BE2_MAX_NUM_CQ_PROC);
        }
 
        /* rearm EQ for further interrupts */
@@ -2281,20 +2274,36 @@ void beiscsi_process_all_cqs(struct work_struct *work)
 
 static int be_iopoll(struct irq_poll *iop, int budget)
 {
-       unsigned int ret;
+       unsigned int ret, num_eq_processed;
        struct beiscsi_hba *phba;
        struct be_eq_obj *pbe_eq;
+       struct be_eq_entry *eqe = NULL;
+       struct be_queue_info *eq;
 
+       num_eq_processed = 0;
        pbe_eq = container_of(iop, struct be_eq_obj, iopoll);
-       ret = beiscsi_process_cq(pbe_eq);
+       phba = pbe_eq->phba;
+       eq = &pbe_eq->q;
+       eqe = queue_tail_node(eq);
+
+       while (eqe->dw[offsetof(struct amap_eq_entry, valid) / 32] &
+                       EQE_VALID_MASK) {
+               AMAP_SET_BITS(struct amap_eq_entry, valid, eqe, 0);
+               queue_tail_inc(eq);
+               eqe = queue_tail_node(eq);
+               num_eq_processed++;
+       }
+
+       hwi_ring_eq_db(phba, eq->id, 1, num_eq_processed, 0, 1);
+
+       ret = beiscsi_process_cq(pbe_eq, budget);
        pbe_eq->cq_count += ret;
        if (ret < budget) {
-               phba = pbe_eq->phba;
                irq_poll_complete(iop);
                beiscsi_log(phba, KERN_INFO,
                            BEISCSI_LOG_CONFIG | BEISCSI_LOG_IO,
-                           "BM_%d : rearm pbe_eq->q.id =%d\n",
-                           pbe_eq->q.id);
+                           "BM_%d : rearm pbe_eq->q.id =%d ret %d\n",
+                           pbe_eq->q.id, ret);
                hwi_ring_eq_db(phba, pbe_eq->q.id, 0, 0, 1, 1);
        }
        return ret;

diff --git a/drivers/scsi/be2iscsi/be_main.h b/drivers/scsi/be2iscsi/be_main.h
index f89861b..fabade3 100644 (file)
--- a/drivers/scsi/be2iscsi/be_main.h
+++ b/drivers/scsi/be2iscsi/be_main.h
@@ -63,6 +63,7 @@
 #define BE2_SGE                        32
 #define BE2_DEFPDU_HDR_SZ      64
 #define BE2_DEFPDU_DATA_SZ     8192
+#define BE2_MAX_NUM_CQ_PROC    512
 
 #define MAX_CPUS               64
 #define BEISCSI_MAX_NUM_CPUS   7
@@ -848,9 +849,9 @@ void beiscsi_free_mgmt_task_handles(struct beiscsi_conn *beiscsi_conn,
 
 void hwi_ring_cq_db(struct beiscsi_hba *phba,
                     unsigned int id, unsigned int num_processed,
-                    unsigned char rearm, unsigned char event);
+                    unsigned char rearm);
 
-unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq);
+unsigned int beiscsi_process_cq(struct be_eq_obj *pbe_eq, int budget);
 
 static inline bool beiscsi_error(struct beiscsi_hba *phba)
 {