[SCSI] mptfc: fix fibre channel infinite request/response loop

While doing board reset testing I was able to put the system in
an infinite request/response loop between the scsi layer and
mptscsih_qcmd() by aborting the reset.  This patch installs
a "SETUP RESET" handler which calls fc_remote_port_delete()
for all registered rports.  This blocks the target which
prevents the loop.  Additionally, should the reset fail to
complete, the transport will now terminate i/o to the target.

Signed-off-by: Michael Reed <mdr@sgi.com>
Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>

1 files changed, 34 insertions, 2 deletions

diff --git a/drivers/message/fusion/mptfc.c b/drivers/message/fusion/mptfc.c
index 918aca0146ff..770df553047a 100644
--- a/drivers/message/fusion/mptfc.c
+++ b/drivers/message/fusion/mptfc.c
@@ -580,10 +580,10 @@ mptfc_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 #ifdef DMPT_DEBUG_FC
         if (unlikely(err)) {
                 dfcprintk ((MYIOC_s_INFO_FMT
-                        "mptfc_qcmd.%d: %d:%d, mptscsih_qcmd returns non-zero.\n",
+                        "mptfc_qcmd.%d: %d:%d, mptscsih_qcmd returns non-zero, (%x).\n",
                         ((MPT_SCSI_HOST *) SCpnt->device->host->hostdata)->ioc->name,
                         ((MPT_SCSI_HOST *) SCpnt->device->host->hostdata)->ioc->sh->host_no,
-                        SCpnt->device->id,SCpnt->device->lun));
+                        SCpnt->device->id,SCpnt->device->lun,err));
         }
 #endif
         return err;
@@ -873,6 +873,31 @@ mptfc_init_host_attr(MPT_ADAPTER *ioc,int portnum)
 }
 
 static void
+mptfc_setup_reset(void *arg)
+{
+        MPT_ADAPTER             *ioc = (MPT_ADAPTER *)arg;
+        u64                     pn;
+        struct mptfc_rport_info *ri;
+
+        /* reset about to happen, delete (block) all rports */
+        list_for_each_entry(ri, &ioc->fc_rports, list) {
+                if (ri->flags & MPT_RPORT_INFO_FLAGS_REGISTERED) {
+                        ri->flags &= ~MPT_RPORT_INFO_FLAGS_REGISTERED;
+                        fc_remote_port_delete(ri->rport);       /* won't sleep */
+                        ri->rport = NULL;
+
+                        pn = (u64)ri->pg0.WWPN.High << 32 |
+                             (u64)ri->pg0.WWPN.Low;
+                        dfcprintk ((MYIOC_s_INFO_FMT
+                                "mptfc_setup_reset.%d: %llx deleted\n",
+                                ioc->name,
+                                ioc->sh->host_no,
+                                (unsigned long long)pn));
+                }
+        }
+}
+
+static void
 mptfc_rescan_devices(void *arg)
 {
         MPT_ADAPTER             *ioc = (MPT_ADAPTER *)arg;
@@ -999,6 +1024,7 @@ mptfc_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
         spin_lock_init(&ioc->fc_rescan_work_lock);
         INIT_WORK(&ioc->fc_rescan_work, mptfc_rescan_devices,(void *)ioc);
+        INIT_WORK(&ioc->fc_setup_reset_work, mptfc_setup_reset, (void *)ioc);
 
         spin_lock_irqsave(&ioc->FreeQlock, flags);
 
@@ -1221,6 +1247,12 @@ mptfc_ioc_reset(MPT_ADAPTER *ioc, int reset_phase)
                 reset_phase==MPT_IOC_PRE_RESET ? "pre" : "post")));
 
         if (reset_phase == MPT_IOC_SETUP_RESET) {
+                spin_lock_irqsave(&ioc->fc_rescan_work_lock, flags);
+                if (ioc->fc_rescan_work_q) {
+                        queue_work(ioc->fc_rescan_work_q,
+                                   &ioc->fc_setup_reset_work);
+                }
+                spin_unlock_irqrestore(&ioc->fc_rescan_work_lock, flags);
         }
 
         else if (reset_phase == MPT_IOC_PRE_RESET) {