[POWERPC] EEH: Add clarifying messages.
authorLinas Vepstas <linas@austin.ibm.com>
Mon, 19 Mar 2007 19:52:04 +0000 (14:52 -0500)
committerPaul Mackerras <paulus@samba.org>
Thu, 22 Mar 2007 11:52:50 +0000 (22:52 +1100)
There are multiple code patchs tht resuls in a "permanent
failure"; when examining rare events, it can be hard to see
which was taken. This patch adds printk's to assist.

Signed-off-by: Linas Vepstas <linas@austin.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
arch/powerpc/platforms/pseries/eeh_driver.c

index a4c0bf84ef2e96644f40f6cbe4eaa69c40a8fb53..550fad2199adf998fa0d0bd371fedf5a19c680e9 100644 (file)
@@ -367,8 +367,10 @@ struct pci_dn * handle_eeh_events (struct eeh_event *event)
         */
        if ((event->state == pci_channel_io_perm_failure) &&
            ((event->time_unavail <= 0) ||
-            (event->time_unavail > MAX_WAIT_FOR_RECOVERY*1000)))
+            (event->time_unavail > MAX_WAIT_FOR_RECOVERY*1000))) {
+               printk(KERN_WARNING "EEH: Permanent failure\n");
                goto hard_fail;
+       }
 
        eeh_slot_error_detail(frozen_pdn, 1 /* Temporary Error */);
        printk(KERN_WARNING
@@ -390,8 +392,10 @@ struct pci_dn * handle_eeh_events (struct eeh_event *event)
         */
        if (result == PCI_ERS_RESULT_NONE) {
                rc = eeh_reset_device(frozen_pdn, frozen_bus);
-               if (rc)
+               if (rc) {
+                       printk(KERN_WARNING "EEH: Unable to reset, rc=%d\n", rc);
                        goto hard_fail;
+               }
        }
 
        /* If all devices reported they can proceed, then re-enable MMIO */
@@ -417,21 +421,27 @@ struct pci_dn * handle_eeh_events (struct eeh_event *event)
        }
 
        /* If any device has a hard failure, then shut off everything. */
-       if (result == PCI_ERS_RESULT_DISCONNECT)
+       if (result == PCI_ERS_RESULT_DISCONNECT) {
+               printk(KERN_WARNING "EEH: Device driver gave up\n");
                goto hard_fail;
+       }
 
        /* If any device called out for a reset, then reset the slot */
        if (result == PCI_ERS_RESULT_NEED_RESET) {
                rc = eeh_reset_device(frozen_pdn, NULL);
-               if (rc)
+               if (rc) {
+                       printk(KERN_WARNING "EEH: Cannot reset, rc=%d\n", rc);
                        goto hard_fail;
+               }
                result = PCI_ERS_RESULT_NONE;
                pci_walk_bus(frozen_bus, eeh_report_reset, &result);
        }
 
        /* All devices should claim they have recovered by now. */
-       if (result != PCI_ERS_RESULT_RECOVERED)
+       if (result != PCI_ERS_RESULT_RECOVERED) {
+               printk(KERN_WARNING "EEH: Not recovered\n");
                goto hard_fail;
+       }
 
        /* Tell all device drivers that they can resume operations */
        pci_walk_bus(frozen_bus, eeh_report_resume, NULL);