[GIT PULL] RAS for 4.2

[Borislav Petkov <bp@alien8.de>]

Hi Ingo,

please queue for 4.2.

Thanks.

@Len, @Rafael: Guys, I'm sending this through tip even though strictly
speaking it is ACPI.

I also am guessing to the point of being almost right, though, that
you're secretly happy I'm doing that because it is one less issue you
have to deal with. :-) :-)

Let me know if you still want to pick that up though.

Thanks.

---
The following changes since commit b787f68c36d49bb1d9236f403813641efa74a031:

  Linux 4.1-rc1 (2015-04-26 17:59:10 -0700)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras.git tags/ras_for_4.2

for you to fetch changes up to 6fe9e7c26a97105645fd24f264f1b94e21aade3e:

  GHES: Make NMI handler have a single reader (2015-04-27 21:35:33 +0200)

----------------------------------------------------------------
GHES: Seriously speedup and cleanup NMI handler (Jiri Kosina and Borislav Petkov)

This is the result of us seeing this during boot

[   24.332560] INFO: NMI handler (ghes_notify_nmi) took too long to run: 3.265 msecs
[   24.332567] INFO: NMI handler (ghes_notify_nmi) took too long to run: 5.946 msecs
[   24.332568] INFO: NMI handler (ghes_notify_nmi) took too long to run: 5.948 msecs

and a report of people running perf stat and the machine softlocking.
"hest_disable" was helping in this case, which made us look at that
particular NMI handler. It was grabbing a lock each time it is run and
on each CPU. But this is not needed as the GHES sources are global and
they need only a single reader.

This patchset does that and cleans up the handler in the process.

----------------------------------------------------------------
Borislav Petkov (4):
      GHES: Carve out error queueing in a separate function
      GHES: Carve out the panic functionality
      GHES: Panic right after detection
      GHES: Elliminate double-loop in the NMI handler

Jiri Kosina (1):
      GHES: Make NMI handler have a single reader

 drivers/acpi/apei/ghes.c | 108 ++++++++++++++++++++++++-----------------------
 1 file changed, 55 insertions(+), 53 deletions(-)

diff --git a/drivers/acpi/apei/ghes.c b/drivers/acpi/apei/ghes.c
index e82d0976a5d0..2bfd53cbfe80 100644
--- a/drivers/acpi/apei/ghes.c
+++ b/drivers/acpi/apei/ghes.c
@@ -729,10 +729,10 @@ static struct llist_head ghes_estatus_llist;
 static struct irq_work ghes_proc_irq_work;
 
 /*
- * NMI may be triggered on any CPU, so ghes_nmi_lock is used for
- * mutual exclusion.
+ * NMI may be triggered on any CPU, so ghes_in_nmi is used for
+ * having only one concurrent reader.
  */
-static DEFINE_RAW_SPINLOCK(ghes_nmi_lock);
+static atomic_t ghes_in_nmi = ATOMIC_INIT(0);
 
 static LIST_HEAD(ghes_nmi);
 
@@ -797,73 +797,75 @@ static void ghes_print_queued_estatus(void)
 	}
 }
 
+/* Save estatus for further processing in IRQ context */
+static void __process_error(struct ghes *ghes)
+{
+#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
+	u32 len, node_len;
+	struct ghes_estatus_node *estatus_node;
+	struct acpi_hest_generic_status *estatus;
+
+	if (ghes_estatus_cached(ghes->estatus))
+		return;
+
+	len = cper_estatus_len(ghes->estatus);
+	node_len = GHES_ESTATUS_NODE_LEN(len);
+
+	estatus_node = (void *)gen_pool_alloc(ghes_estatus_pool, node_len);
+	if (!estatus_node)
+		return;
+
+	estatus_node->ghes = ghes;
+	estatus_node->generic = ghes->generic;
+	estatus = GHES_ESTATUS_FROM_NODE(estatus_node);
+	memcpy(estatus, ghes->estatus, len);
+	llist_add(&estatus_node->llnode, &ghes_estatus_llist);
+#endif
+}
+
+static void __ghes_panic(struct ghes *ghes)
+{
+	oops_begin();
+	ghes_print_queued_estatus();
+	__ghes_print_estatus(KERN_EMERG, ghes->generic, ghes->estatus);
+
+	/* reboot to log the error! */
+	if (panic_timeout == 0)
+		panic_timeout = ghes_panic_timeout;
+	panic("Fatal hardware error!");
+}
+
 static int ghes_notify_nmi(unsigned int cmd, struct pt_regs *regs)
 {
-	struct ghes *ghes, *ghes_global = NULL;
-	int sev, sev_global = -1;
-	int ret = NMI_DONE;
+	struct ghes *ghes;
+	int sev, ret = NMI_DONE;
+
+	if (!atomic_add_unless(&ghes_in_nmi, 1, 1))
+		return ret;
 
-	raw_spin_lock(&ghes_nmi_lock);
 	list_for_each_entry_rcu(ghes, &ghes_nmi, list) {
 		if (ghes_read_estatus(ghes, 1)) {
 			ghes_clear_estatus(ghes);
 			continue;
 		}
-		sev = ghes_severity(ghes->estatus->error_severity);
-		if (sev > sev_global) {
-			sev_global = sev;
-			ghes_global = ghes;
-		}
-		ret = NMI_HANDLED;
-	}
-
-	if (ret == NMI_DONE)
-		goto out;
 
-	if (sev_global >= GHES_SEV_PANIC) {
-		oops_begin();
-		ghes_print_queued_estatus();
-		__ghes_print_estatus(KERN_EMERG, ghes_global->generic,
-				     ghes_global->estatus);
-		/* reboot to log the error! */
-		if (panic_timeout == 0)
-			panic_timeout = ghes_panic_timeout;
-		panic("Fatal hardware error!");
-	}
+		sev = ghes_severity(ghes->estatus->error_severity);
+		if (sev >= GHES_SEV_PANIC)
+			__ghes_panic(ghes);
 
-	list_for_each_entry_rcu(ghes, &ghes_nmi, list) {
-#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
-		u32 len, node_len;
-		struct ghes_estatus_node *estatus_node;
-		struct acpi_hest_generic_status *estatus;
-#endif
 		if (!(ghes->flags & GHES_TO_CLEAR))
 			continue;
-#ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
-		if (ghes_estatus_cached(ghes->estatus))
-			goto next;
-		/* Save estatus for further processing in IRQ context */
-		len = cper_estatus_len(ghes->estatus);
-		node_len = GHES_ESTATUS_NODE_LEN(len);
-		estatus_node = (void *)gen_pool_alloc(ghes_estatus_pool,
-						      node_len);
-		if (estatus_node) {
-			estatus_node->ghes = ghes;
-			estatus_node->generic = ghes->generic;
-			estatus = GHES_ESTATUS_FROM_NODE(estatus_node);
-			memcpy(estatus, ghes->estatus, len);
-			llist_add(&estatus_node->llnode, &ghes_estatus_llist);
-		}
-next:
-#endif
+
+		__process_error(ghes);
 		ghes_clear_estatus(ghes);
+
+		ret = NMI_HANDLED;
 	}
+
 #ifdef CONFIG_ARCH_HAVE_NMI_SAFE_CMPXCHG
 	irq_work_queue(&ghes_proc_irq_work);
 #endif
-
-out:
-	raw_spin_unlock(&ghes_nmi_lock);
+	atomic_dec(&ghes_in_nmi);
 	return ret;
 }
 

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--