scsi: lpfc: devloss timeout race condition caused null pointer reference

A race condition between the context of devloss timeout handler and I/O
completion caused devloss timeout handler de-referencing pointer that had
been released.

Added the check in lpfc_sli_validate_fcp_iocb() on LPFC_IO_ON_TXCMPLQ to
capture the race condition of I/O completion and devloss timeout handler
attemption for aborting the I/O. Also, added check on lpfc_cmd->rdata
pointer before de-referenceing lpfc_cmd->rdata->pnode.

Also, added protection in lpfc_sli_abort_iocb() routine on driver performed
FCP I/O FLUSHING already under way before proceeding to aborting I/Os.

Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index c31c43f43553ff6f1f98c7e09533a4202f7c2c27..5e4cad6a531aa0a5756c90b600429278f650553e 100644 (file)
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -4538,6 +4538,11 @@ lpfc_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *cmnd)
        int err;
 
        rdata = lpfc_rport_data_from_scsi_device(cmnd->device);
+
+       /* sanity check on references */
+       if (unlikely(!rdata) || unlikely(!rport))
+               goto out_fail_command;
+
        err = fc_remote_port_chkready(rport);
        if (err) {
                cmnd->result = err;

diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index b55873d502128a63b3ddf372900155ff9538c1cd..72500dcb13a9028367577bc190358cb9c839e076 100644 (file)
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -11091,10 +11091,11 @@ lpfc_sli_validate_fcp_iocb(struct lpfc_iocbq *iocbq, struct lpfc_vport *vport,
        struct lpfc_scsi_buf *lpfc_cmd;
        int rc = 1;
 
-       if (!(iocbq->iocb_flag &  LPFC_IO_FCP))
+       if (iocbq->vport != vport)
                return rc;
 
-       if (iocbq->vport != vport)
+       if (!(iocbq->iocb_flag &  LPFC_IO_FCP) ||
+           !(iocbq->iocb_flag & LPFC_IO_ON_TXCMPLQ))
                return rc;
 
        lpfc_cmd = container_of(iocbq, struct lpfc_scsi_buf, cur_iocbq);
@@ -11104,13 +11105,13 @@ lpfc_sli_validate_fcp_iocb(struct lpfc_iocbq *iocbq, struct lpfc_vport *vport,
 
        switch (ctx_cmd) {
        case LPFC_CTX_LUN:
-               if ((lpfc_cmd->rdata->pnode) &&
+               if ((lpfc_cmd->rdata) && (lpfc_cmd->rdata->pnode) &&
                    (lpfc_cmd->rdata->pnode->nlp_sid == tgt_id) &&
                    (scsilun_to_int(&lpfc_cmd->fcp_cmnd->fcp_lun) == lun_id))
                        rc = 0;
                break;
        case LPFC_CTX_TGT:
-               if ((lpfc_cmd->rdata->pnode) &&
+               if ((lpfc_cmd->rdata) && (lpfc_cmd->rdata->pnode) &&
                    (lpfc_cmd->rdata->pnode->nlp_sid == tgt_id))
                        rc = 0;
                break;
@@ -11225,6 +11226,10 @@ lpfc_sli_abort_iocb(struct lpfc_vport *vport, struct lpfc_sli_ring *pring,
        int errcnt = 0, ret_val = 0;
        int i;
 
+       /* all I/Os are in process of being flushed */
+       if (phba->hba_flag & HBA_FCP_IOQ_FLUSH)
+               return errcnt;
+
        for (i = 1; i <= phba->sli.last_iotag; i++) {
                iocbq = phba->sli.iocbq_lookup[i];