[PATCH] Avoid use of uninitialised spinlock in EEH.

If the kernel supports both G5 and pSeries, and CONFIG_EEH is enabled,
eeh_init() is (quite reasonably) never called when we boot on a G5. Yet
eeh_check_failure() still gets called. We should avoid doing that if
!eeh_subsystem_enabled.

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>

arch/powerpc/platforms/pseries/eeh.c

@@ -86,7 +86,8 @@ static int ibm_read_slot_reset_state;
 static int ibm_read_slot_reset_state2;
 static int ibm_slot_error_detail;
 
-static int eeh_subsystem_enabled;
+int eeh_subsystem_enabled;
+EXPORT_SYMBOL(eeh_subsystem_enabled);
 
 /* Lock to avoid races due to multiple reports of an error */
 static DEFINE_SPINLOCK(confirm_error_lock);

include/asm-powerpc/eeh.h

@@ -30,6 +30,8 @@ struct device_node;
 
 #ifdef CONFIG_EEH
 
+extern int eeh_subsystem_enabled;
+
 /* Values for eeh_mode bits in device_node */
 #define EEH_MODE_SUPPORTED	(1<<0)
 #define EEH_MODE_NOCHECK	(1<<1)
@@ -75,7 +77,7 @@ void eeh_remove_device(struct pci_dev *);
  * If this macro yields TRUE, the caller relays to eeh_check_failure()
  * which does further tests out of line.
  */
-#define EEH_POSSIBLE_ERROR(val, type)	((val) == (type)~0)
+#define EEH_POSSIBLE_ERROR(val, type)	((val) == (type)~0 && eeh_subsystem_enabled)
 
 /*
  * Reads from a device which has been isolated by EEH will return