[tip:ras/core] x86/mce: Look in genpool instead of mcelog for pending error records

[tip-bot for Tony Luck]

Commit-ID:  5541c93cdf0cc0bb7f6065b43509171838665ea1
Gitweb:     http://git.kernel.org/tip/5541c93cdf0cc0bb7f6065b43509171838665ea1
Author:     Tony Luck <tony.luck@xxxxxxxxx>
AuthorDate: Sat, 30 Apr 2016 14:33:56 +0200
Committer:  Ingo Molnar <mingo@xxxxxxxxxx>
CommitDate: Tue, 3 May 2016 08:24:16 +0200

x86/mce: Look in genpool instead of mcelog for pending error records

A couple of issues here:

1) MCE_LOG_LEN is only 32 - so we may have more pending records than will
   fit in the buffer on high core count CPUs.

2) During a panic we may have a lot of duplicate records because multiple
   logical CPUs may have seen and logged the same error because some
   banks are shared.

Switch to using the genpool to look for the pending records. Squeeze out
duplicated records.

Signed-off-by: Tony Luck <tony.luck@xxxxxxxxx>
Signed-off-by: Borislav Petkov <bp@xxxxxxx>
Cc: Andy Lutomirski <luto@xxxxxxxxxxxxxx>
Cc: Ashok Raj <ashok.raj@xxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: Brian Gerst <brgerst@xxxxxxxxx>
Cc: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
Cc: H. Peter Anvin <hpa@xxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: linux-edac <linux-edac@xxxxxxxxxxxxxxx>
Link: http://lkml.kernel.org/r/1462019637-16474-7-git-send-email-bp@xxxxxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
---
 arch/x86/kernel/cpu/mcheck/mce-genpool.c  | 46 +++++++++++++++++++++++++++++++
 arch/x86/kernel/cpu/mcheck/mce-internal.h | 15 ++++++++++
 arch/x86/kernel/cpu/mcheck/mce.c          | 21 ++++++--------
 3 files changed, 70 insertions(+), 12 deletions(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce-genpool.c b/arch/x86/kernel/cpu/mcheck/mce-genpool.c
index 2658e2a..93d824e 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-genpool.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-genpool.c
@@ -26,6 +26,52 @@ static struct gen_pool *mce_evt_pool;
 static LLIST_HEAD(mce_event_llist);
 static char gen_pool_buf[MCE_POOLSZ];
 
+/*
+ * Compare the record "t" with each of the records on list "l" to see if
+ * an equivalent one is present in the list.
+ */
+static bool is_duplicate_mce_record(struct mce_evt_llist *t, struct mce_evt_llist *l)
+{
+	struct mce_evt_llist *node;
+	struct mce *m1, *m2;
+
+	m1 = &t->mce;
+
+	llist_for_each_entry(node, &l->llnode, llnode) {
+		m2 = &node->mce;
+
+		if (!mce_cmp(m1, m2))
+			return true;
+	}
+	return false;
+}
+
+/*
+ * The system has panicked - we'd like to peruse the list of MCE records
+ * that have been queued, but not seen by anyone yet.  The list is in
+ * reverse time order, so we need to reverse it. While doing that we can
+ * also drop duplicate records (these were logged because some banks are
+ * shared between cores or by all threads on a socket).
+ */
+struct llist_node *mce_gen_pool_prepare_records(void)
+{
+	struct llist_node *head;
+	LLIST_HEAD(new_head);
+	struct mce_evt_llist *node, *t;
+
+	head = llist_del_all(&mce_event_llist);
+	if (!head)
+		return NULL;
+
+	/* squeeze out duplicates while reversing order */
+	llist_for_each_entry_safe(node, t, head, llnode) {
+		if (!is_duplicate_mce_record(node, t))
+			llist_add(&node->llnode, &new_head);
+	}
+
+	return new_head.first;
+}
+
 void mce_gen_pool_process(void)
 {
 	struct llist_node *head;
diff --git a/arch/x86/kernel/cpu/mcheck/mce-internal.h b/arch/x86/kernel/cpu/mcheck/mce-internal.h
index 547720e..cd74a3f 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-internal.h
+++ b/arch/x86/kernel/cpu/mcheck/mce-internal.h
@@ -35,6 +35,7 @@ void mce_gen_pool_process(void);
 bool mce_gen_pool_empty(void);
 int mce_gen_pool_add(struct mce *mce);
 int mce_gen_pool_init(void);
+struct llist_node *mce_gen_pool_prepare_records(void);
 
 extern int (*mce_severity)(struct mce *a, int tolerant, char **msg, bool is_excp);
 struct dentry *mce_get_debugfs_dir(void);
@@ -81,3 +82,17 @@ static inline int apei_clear_mce(u64 record_id)
 #endif
 
 void mce_inject_log(struct mce *m);
+
+/*
+ * We consider records to be equivalent if bank+status+addr+misc all match.
+ * This is only used when the system is going down because of a fatal error
+ * to avoid cluttering the console log with essentially repeated information.
+ * In normal processing all errors seen are logged.
+ */
+static inline bool mce_cmp(struct mce *m1, struct mce *m2)
+{
+	return m1->bank != m2->bank ||
+		m1->status != m2->status ||
+		m1->addr != m2->addr ||
+		m1->misc != m2->misc;
+}
diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index d77f3b0..c356f47 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -161,7 +161,6 @@ void mce_log(struct mce *mce)
 	if (!mce_gen_pool_add(mce))
 		irq_work_queue(&mce_irq_work);
 
-	mce->finished = 0;
 	wmb();
 	for (;;) {
 		entry = mce_log_get_idx_check(mcelog.next);
@@ -194,7 +193,6 @@ void mce_log(struct mce *mce)
 	mcelog.entry[entry].finished = 1;
 	wmb();
 
-	mce->finished = 1;
 	set_bit(0, &mce_need_notify);
 }
 
@@ -337,7 +335,9 @@ static void wait_for_panic(void)
 
 static void mce_panic(const char *msg, struct mce *final, char *exp)
 {
-	int i, apei_err = 0;
+	int apei_err = 0;
+	struct llist_node *pending;
+	struct mce_evt_llist *l;
 
 	if (!fake_panic) {
 		/*
@@ -354,11 +354,10 @@ static void mce_panic(const char *msg, struct mce *final, char *exp)
 		if (atomic_inc_return(&mce_fake_panicked) > 1)
 			return;
 	}
+	pending = mce_gen_pool_prepare_records();
 	/* First print corrected ones that are still unlogged */
-	for (i = 0; i < MCE_LOG_LEN; i++) {
-		struct mce *m = &mcelog.entry[i];
-		if (!(m->status & MCI_STATUS_VAL))
-			continue;
+	llist_for_each_entry(l, pending, llnode) {
+		struct mce *m = &l->mce;
 		if (!(m->status & MCI_STATUS_UC)) {
 			print_mce(m);
 			if (!apei_err)
@@ -366,13 +365,11 @@ static void mce_panic(const char *msg, struct mce *final, char *exp)
 		}
 	}
 	/* Now print uncorrected but with the final one last */
-	for (i = 0; i < MCE_LOG_LEN; i++) {
-		struct mce *m = &mcelog.entry[i];
-		if (!(m->status & MCI_STATUS_VAL))
-			continue;
+	llist_for_each_entry(l, pending, llnode) {
+		struct mce *m = &l->mce;
 		if (!(m->status & MCI_STATUS_UC))
 			continue;
-		if (!final || memcmp(m, final, sizeof(struct mce))) {
+		if (!final || mce_cmp(m, final)) {
 			print_mce(m);
 			if (!apei_err)
 				apei_err = apei_write_mce(m);