]> err.no Git - linux-2.6/commitdiff
[PATCH] ppc64: RTAS error reporting restructuring
authorLinas Vepstas <linas@linas.org>
Fri, 4 Nov 2005 00:49:01 +0000 (18:49 -0600)
committerPaul Mackerras <paulus@samba.org>
Thu, 10 Nov 2005 00:33:14 +0000 (11:33 +1100)
05-eeh-slot-error-detail.patch

This patch encapsulates a section of code that reports the EEH event.
The new subroutine can be used in several places to report the error.

Signed-off-by: Linas Vepstas <linas@linas.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
arch/ppc64/kernel/eeh.c

index 926d3110dfd75963c10c3791a2416627e2ebb12e..b23159464abb1591eddcae62f663627f022b7e45 100644 (file)
@@ -397,6 +397,28 @@ void __init pci_addr_cache_build(void)
 /* --------------------------------------------------------------- */
 /* Above lies the PCI Address Cache. Below lies the EEH event infrastructure */
 
+void eeh_slot_error_detail (struct pci_dn *pdn, int severity)
+{
+       unsigned long flags;
+       int rc;
+
+       /* Log the error with the rtas logger */
+       spin_lock_irqsave(&slot_errbuf_lock, flags);
+       memset(slot_errbuf, 0, eeh_error_buf_size);
+
+       rc = rtas_call(ibm_slot_error_detail,
+                      8, 1, NULL, pdn->eeh_config_addr,
+                      BUID_HI(pdn->phb->buid),
+                      BUID_LO(pdn->phb->buid), NULL, 0,
+                      virt_to_phys(slot_errbuf),
+                      eeh_error_buf_size,
+                      severity);
+
+       if (rc == 0)
+               log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0);
+       spin_unlock_irqrestore(&slot_errbuf_lock, flags);
+}
+
 /**
  * eeh_register_notifier - Register to find out about EEH events.
  * @nb: notifier block to callback on events
@@ -454,9 +476,12 @@ static void eeh_panic(struct pci_dev *dev, int reset_state)
         * Since the panic_on_oops sysctl is used to halt the system
         * in light of potential corruption, we can use it here.
         */
-       if (panic_on_oops)
+       if (panic_on_oops) {
+               struct device_node *dn = pci_device_to_OF_node(dev);
+               eeh_slot_error_detail (PCI_DN(dn), 2 /* Permanent Error */);
                panic("EEH: MMIO failure (%d) on device:%s\n", reset_state,
                      pci_name(dev));
+       }
        else {
                __get_cpu_var(ignored_failures)++;
                printk(KERN_INFO "EEH: Ignored MMIO failure (%d) on device:%s\n",
@@ -539,7 +564,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev)
        int ret;
        int rets[3];
        unsigned long flags;
-       int rc, reset_state;
+       int reset_state;
        struct eeh_event  *event;
        struct pci_dn *pdn;
 
@@ -603,20 +628,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev)
 
        reset_state = rets[0];
 
-       spin_lock_irqsave(&slot_errbuf_lock, flags);
-       memset(slot_errbuf, 0, eeh_error_buf_size);
-
-       rc = rtas_call(ibm_slot_error_detail,
-                      8, 1, NULL, pdn->eeh_config_addr,
-                      BUID_HI(pdn->phb->buid),
-                      BUID_LO(pdn->phb->buid), NULL, 0,
-                      virt_to_phys(slot_errbuf),
-                      eeh_error_buf_size,
-                      1 /* Temporary Error */);
-
-       if (rc == 0)
-               log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0);
-       spin_unlock_irqrestore(&slot_errbuf_lock, flags);
+       eeh_slot_error_detail (pdn, 1 /* Temporary Error */);
 
        printk(KERN_INFO "EEH: MMIO failure (%d) on device: %s %s\n",
               rets[0], dn->name, dn->full_name);
@@ -783,6 +795,8 @@ void __init eeh_init(void)
        struct device_node *phb, *np;
        struct eeh_early_enable_info info;
 
+       spin_lock_init(&slot_errbuf_lock);
+
        np = of_find_node_by_path("/rtas");
        if (np == NULL)
                return;