lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4B02EE1D.6020302@sgi.com>
Date:	Tue, 17 Nov 2009 10:40:29 -0800
From:	Mike Travis <travis@....com>
To:	Hidetoshi Seto <seto.hidetoshi@...fujitsu.com>,
	Ingo Molnar <mingo@...e.hu>
CC:	Thomas Gleixner <tglx@...utronix.de>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Heiko Carstens <heiko.carstens@...ibm.com>,
	Roland Dreier <rdreier@...co.com>,
	Randy Dunlap <rdunlap@...otime.net>, Tejun Heo <tj@...nel.org>,
	Andi Kleen <andi@...stfloor.org>,
	Greg Kroah-Hartman <gregkh@...e.de>,
	Yinghai Lu <yhlu.kernel@...il.com>,
	"H. Peter Anvin" <hpa@...or.com>,
	David Rientjes <rientjes@...gle.com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Rusty Russell <rusty@...tcorp.com.au>,
	Jack Steiner <steiner@....com>,
	Frederic Weisbecker <fweisbec@...il.com>, x86@...nel.org,
	linux-kernel@...r.kernel.org
Subject: [PATCH] x86, mce: rework output of MCE banks ownership information

Author: Hidetoshi Seto <seto.hidetoshi@...fujitsu.com>

The output of MCE banks ownership information on boot tend
to be long on new processor which has many banks:

  CPU 1 MCA banks SHD:0 SHD:1 CMCI:2 CMCI:3 CMCI:5 SHD:6 SHD:7 SHD:8 SHD:9 SHD:12 SHD:13 SHD:14 SHD:15 SHD:16 SHD:17 SHD:18 SHD:19 SHD:20 SHD:21

This message can fill up the console output when the number
of cpus is large.

This patch suppress this info message on boot, and introduce
debug message in shorter format instead, like:

  CPU 1 MCE banks map: ssCC PCss ssPP ssss ssss ss

  where: s: shared, C: checked by cmci, P: checked by poll.

This patch still keep the info when ownership is updated.
E.g. if a cpu take over the ownership from hot-removed cpu,
both message will be shown:

  CPU 1 MCE banks map updated: CMCI:6 CMCI:7 CMCI:10 CMCI:11
  CPU 1 MCE banks map: ssCC PCCC ssPP ssCC ssss ss

v2:
  - stop changing the level of message on update
  - change the number of banks message on boot to debug level

Signed-off-by: Hidetoshi Seto <seto.hidetoshi@...fujitsu.com>

  - Modified to not use pr_cont().

Signed-off-by: Mike Travis <travis@....com>
---
 arch/x86/kernel/cpu/mcheck/mce.c       |    6 +--
 arch/x86/kernel/cpu/mcheck/mce_intel.c |   63 +++++++++++++++++++++++++++------
 2 files changed, 55 insertions(+), 14 deletions(-)

--- linux.orig/arch/x86/kernel/cpu/mcheck/mce.c
+++ linux/arch/x86/kernel/cpu/mcheck/mce.c
@@ -1215,11 +1215,11 @@
 
 	b = cap & MCG_BANKCNT_MASK;
 	if (!banks)
-		printk(KERN_INFO "mce: CPU supports %d MCE banks\n", b);
+		pr_debug("mce: CPU supports %d MCE banks\n", b);
 
 	if (b > MAX_NR_BANKS) {
-		printk(KERN_WARNING
-		       "MCE: Using only %u machine check banks out of %u\n",
+		pr_warning(
+			"MCE: Using only %u machine check banks out of %u\n",
 			MAX_NR_BANKS, b);
 		b = MAX_NR_BANKS;
 	}
--- linux.orig/arch/x86/kernel/cpu/mcheck/mce_intel.c
+++ linux/arch/x86/kernel/cpu/mcheck/mce_intel.c
@@ -64,14 +64,50 @@
 	mce_notify_irq();
 }
 
-static void print_update(char *type, int *hdr, int num)
+#define	MCE_MSG_LEN	120
+
+#ifdef DEBUG_KERNEL
+static void print_banks_map(int banks, char *buf)
+{
+	int i, n;
+
+	n = snprintf(buf, MCE_MSG_LEN, "CPU %d MCE banks map:",
+							smp_processor_id());
+	for (i = 0; i < banks; i++) {
+		n += snprintf(&buf[n], MCE_MSG_LEN - n,
+			"%s%s", (i % 4) ? "" : " ",
+			test_bit(i, __get_cpu_var(mce_banks_owned)) ? "C" :
+			test_bit(i, __get_cpu_var(mce_poll_banks)) ? "P" : "s");
+	}
+
+	/* (indicate if message buffer overflowed) */
+	pr_debug("%s%s\n", buf, n < MCE_MSG_LEN ? "" : "..." );
+}
+
+static void print_update(char *type, int *hdr, int num, char *buf)
+{
+	int n = *hdr;
+
+	if (n == 0)
+		n = snprintf(buf, MCE_MSG_LEN,
+			"CPU %d MCE banks map updated:", smp_processor_id());
+
+	n += snprintf(&buf[n], MCE_MSG_LEN - n, " %s:%d", type, num);
+	*hdr = n;
+}
+
+#else /* !DEBUG_KERNEL */
+
+static inline void print_banks_map(int banks, char *buf)
+{
+}
+
+static inline void print_update(char *type, int *hdr, int num, char *buf)
 {
-	if (*hdr == 0)
-		printk(KERN_INFO "CPU %d MCA banks", smp_processor_id());
-	*hdr = 1;
-	printk(KERN_CONT " %s:%d", type, num);
 }
 
+#endif
+
 /*
  * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks
  * on this CPU. Use the algorithm recommended in the SDM to discover shared
@@ -83,8 +119,10 @@
 	unsigned long flags;
 	int hdr = 0;
 	int i;
+	char buf[MCE_MSG_LEN];
 
 	spin_lock_irqsave(&cmci_discover_lock, flags);
+
 	for (i = 0; i < banks; i++) {
 		u64 val;
 
@@ -95,8 +133,8 @@
 
 		/* Already owned by someone else? */
 		if (val & CMCI_EN) {
-			if (test_and_clear_bit(i, owned) || boot)
-				print_update("SHD", &hdr, i);
+			if (test_and_clear_bit(i, owned) && !boot)
+				print_update("SHD", &hdr, i, buf);
 			__clear_bit(i, __get_cpu_var(mce_poll_banks));
 			continue;
 		}
@@ -107,16 +145,19 @@
 
 		/* Did the enable bit stick? -- the bank supports CMCI */
 		if (val & CMCI_EN) {
-			if (!test_and_set_bit(i, owned) || boot)
-				print_update("CMCI", &hdr, i);
+			if (!test_and_set_bit(i, owned) && !boot)
+				print_update("CMCI", &hdr, i, buf);
 			__clear_bit(i, __get_cpu_var(mce_poll_banks));
 		} else {
 			WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks)));
 		}
 	}
-	spin_unlock_irqrestore(&cmci_discover_lock, flags);
 	if (hdr)
-		printk(KERN_CONT "\n");
+		pr_debug("%s%s\n", buf, hdr < MCE_MSG_LEN ? "" : "...");
+	if (hdr || boot)
+		print_banks_map(banks, buf);
+
+	spin_unlock_irqrestore(&cmci_discover_lock, flags);
 }
 
 /*

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ