[SCSI] lpfc 8.1.2: Fixed a race condition in the PLOGI retry logic.

[linux-2.6-omap-h63xx.git] / drivers / scsi / lpfc / lpfc_nportdisc.c

diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index fbead786031f9ec948df36249d557d6e646a2d17..1c04ea353ffa8e024bc914e3d92519dd639b9eea 100644 (file)
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -1627,6 +1627,14 @@ lpfc_device_recov_npr_node(struct lpfc_hba * phba,
 {
        spin_lock_irq(phba->host->host_lock);
        ndlp->nlp_flag &= ~NLP_NPR_2B_DISC;
+       if (ndlp->nlp_flag & NLP_DELAY_TMO) {
+               ndlp->nlp_flag &= ~NLP_DELAY_TMO;
+               if (!list_empty(&ndlp->els_retry_evt.evt_listp))
+                       list_del_init(&ndlp->els_retry_evt.evt_listp);
+               spin_unlock_irq(phba->host->host_lock);
+               del_timer_sync(&ndlp->nlp_delayfunc);
+               return (ndlp->nlp_state);
+       }
        spin_unlock_irq(phba->host->host_lock);
        return (ndlp->nlp_state);
 }