From: Prarit Bhargava <prarit@redhat.com>
Subject: Re: [RHEL5 PATCH]: Regression: Add panic on unrecovered NMI
Date: Mon, 08 Jan 2007 10:34:47 -0500
Bugzilla: 220829
Message-Id: <45A26497.40604@redhat.com>
Changelog: x86: Add panic on unrecovered NMI




Resubmitting after dzickus' suggestion to backport 2 additional changes.


Add panic on unrecovered NMI.  Regression from RHEL4.

Backport of

http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=8da5adda91df3d2fcc5300e68da491694c9af019

http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=29cbc78b90a73ad80f2f58ba2927956cf663abed

http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=c41c5cd3b20a2d81c30498f13b1527847a8fdf69

Compiled and tested against latest RHEL5.

Resolves BZ 220829.


diff -urNp linux-2.6.18.x86_64.orig/arch/i386/kernel/nmi.c linux-2.6.18.x86_64/arch/i386/kernel/nmi.c
--- linux-2.6.18.x86_64.orig/arch/i386/kernel/nmi.c	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/arch/i386/kernel/nmi.c	2007-01-09 14:34:55.000000000 -0500
@@ -28,7 +28,7 @@
 #include "mach_traps.h"
 
 unsigned int nmi_watchdog = NMI_NONE;
-extern int unknown_nmi_panic;
+int unknown_nmi_panic;
 static unsigned int nmi_hz = HZ;
 static unsigned int nmi_perfctr_msr;	/* the MSR to reset in NMI handler */
 static unsigned int nmi_p4_cccr_val;
diff -urNp linux-2.6.18.x86_64.orig/arch/i386/kernel/traps.c linux-2.6.18.x86_64/arch/i386/kernel/traps.c
--- linux-2.6.18.x86_64.orig/arch/i386/kernel/traps.c	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/arch/i386/kernel/traps.c	2007-01-09 14:34:55.000000000 -0500
@@ -58,6 +58,8 @@
 
 #include "mach_traps.h"
 
+int panic_on_unrecovered_nmi;
+
 asmlinkage int system_call(void);
 
 struct desc_struct default_ldt[] = { { 0, 0 }, { 0, 0 }, { 0, 0 },
@@ -729,10 +731,14 @@ gp_in_kernel:
 
 static void mem_parity_error(unsigned char reason, struct pt_regs * regs)
 {
-	printk(KERN_EMERG "Uhhuh. NMI received. Dazed and confused, but trying "
-			"to continue\n");
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
+		"CPU %d.\n", reason, smp_processor_id());
 	printk(KERN_EMERG "You probably have a hardware problem with your RAM "
 			"chips\n");
+	if (panic_on_unrecovered_nmi)
+                panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 
 	/* Clear and disable the memory parity error line. */
 	clear_mem_error(reason);
@@ -757,10 +763,13 @@ static void unknown_nmi_error(unsigned c
 		return;
 	}
 #endif
-	printk("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n",
-		reason, smp_processor_id());
-	printk("Dazed and confused, but trying to continue\n");
-	printk("Do you have a strange power saving mode enabled?\n");
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
+		"CPU %d.\n", reason, smp_processor_id());
+	printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
+	if (panic_on_unrecovered_nmi)
+                panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 }
 
 static DEFINE_SPINLOCK(nmi_print_lock);
diff -urNp linux-2.6.18.x86_64.orig/arch/i386/kernel/traps-xen.c linux-2.6.18.x86_64/arch/i386/kernel/traps-xen.c
--- linux-2.6.18.x86_64.orig/arch/i386/kernel/traps-xen.c 2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/arch/i386/kernel/traps-xen.c	2007-01-09 14:34:55.000000000 -0500
@@ -58,6 +58,8 @@
 
 #include "mach_traps.h"
 
+int panic_on_unrecovered_nmi;
+
 asmlinkage int system_call(void);
 
 struct desc_struct default_ldt[] = { { 0, 0 }, { 0, 0 }, { 0, 0 },
@@ -703,10 +705,14 @@ gp_in_kernel:
 
 static void mem_parity_error(unsigned char reason, struct pt_regs * regs)
 {
-	printk(KERN_EMERG "Uhhuh. NMI received. Dazed and confused, but trying "
-			"to continue\n");
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
+		"CPU %d.\n", reason, smp_processor_id());
 	printk(KERN_EMERG "You probably have a hardware problem with your RAM "
 			"chips\n");
+	if (panic_on_unrecovered_nmi)
+                panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 
 	/* Clear and disable the memory parity error line. */
 	clear_mem_error(reason);
@@ -731,10 +737,13 @@ static void unknown_nmi_error(unsigned c
 		return;
 	}
 #endif
-	printk("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n",
-		reason, smp_processor_id());
-	printk("Dazed and confused, but trying to continue\n");
-	printk("Do you have a strange power saving mode enabled?\n");
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x on "
+		"CPU %d.\n", reason, smp_processor_id());
+	printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
+	if (panic_on_unrecovered_nmi)
+                panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 }
 
 static DEFINE_SPINLOCK(nmi_print_lock);
diff -urNp linux-2.6.18.x86_64.orig/arch/x86_64/kernel/nmi.c linux-2.6.18.x86_64/arch/x86_64/kernel/nmi.c
--- linux-2.6.18.x86_64.orig/arch/x86_64/kernel/nmi.c	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/arch/x86_64/kernel/nmi.c	2007-01-09 14:34:55.000000000 -0500
@@ -28,6 +28,8 @@
 #include <asm/mce.h>
 #include <asm/intel_arch_perfmon.h>
 
+int unknown_nmi_panic;
+int panic_on_unrecovered_nmi;
 /*
  * lapic_nmi_owner tracks the ownership of the lapic NMI hardware:
  * - it may be reserved by some other driver, or not
diff -urNp linux-2.6.18.x86_64.orig/arch/x86_64/kernel/traps.c linux-2.6.18.x86_64/arch/x86_64/kernel/traps.c
--- linux-2.6.18.x86_64.orig/arch/x86_64/kernel/traps.c	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/arch/x86_64/kernel/traps.c	2007-01-09 14:34:55.000000000 -0500
@@ -734,8 +734,15 @@ asmlinkage void __kprobes do_general_pro
 static __kprobes void
 mem_parity_error(unsigned char reason, struct pt_regs * regs)
 {
-	printk("Uhhuh. NMI received. Dazed and confused, but trying to continue\n");
-	printk("You probably have a hardware problem with your RAM chips\n");
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
+		reason);
+	printk(KERN_EMERG "You probably have a hardware problem with your "
+		"RAM chips\n");
+
+	if (panic_on_unrecovered_nmi)
+		panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 
 	/* Clear and disable the memory parity error line. */
 	reason = (reason & 0xf) | 4;
@@ -758,9 +765,15 @@ io_check_error(unsigned char reason, str
 
 static __kprobes void
 unknown_nmi_error(unsigned char reason, struct pt_regs * regs)
-{	printk("Uhhuh. NMI received for unknown reason %02x.\n", reason);
-	printk("Dazed and confused, but trying to continue\n");
-	printk("Do you have a strange power saving mode enabled?\n");
+{
+	printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n",
+		reason);
+	printk(KERN_EMERG "Do you have a strange power saving mode enabled?\n");
+
+	if (panic_on_unrecovered_nmi)
+		panic("NMI: Not continuing");
+
+	printk(KERN_EMERG "Dazed and confused, but trying to continue\n");
 }
 
 /* Runs on IST stack. This code must keep interrupts off all the time.
diff -urNp linux-2.6.18.x86_64.orig/include/asm-i386/nmi.h linux-2.6.18.x86_64/include/asm-i386/nmi.h
--- linux-2.6.18.x86_64.orig/include/asm-i386/nmi.h	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/include/asm-i386/nmi.h	2007-01-09 14:34:55.000000000 -0500
@@ -39,4 +39,6 @@ extern unsigned int nmi_watchdog;
 #define NMI_LOCAL_APIC	2
 #define NMI_INVALID	3
 
+extern int unknown_nmi_panic;
+
 #endif /* ASM_NMI_H */
diff -urNp linux-2.6.18.x86_64.orig/include/linux/kernel.h linux-2.6.18.x86_64/include/linux/kernel.h
--- linux-2.6.18.x86_64.orig/include/linux/kernel.h	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/include/linux/kernel.h	2007-01-09 14:34:55.000000000 -0500
@@ -187,6 +187,7 @@ extern void bust_spinlocks(int yes);
 extern int oops_in_progress;		/* If set, an oops, panic(), BUG() or die() is in progress */
 extern int panic_timeout;
 extern int panic_on_oops;
+extern int panic_on_unrecovered_nmi;
 extern int tainted;
 extern const char *print_tainted(void);
 extern void add_taint(unsigned);
diff -urNp linux-2.6.18.x86_64.orig/include/linux/sysctl.h linux-2.6.18.x86_64/include/linux/sysctl.h
--- linux-2.6.18.x86_64.orig/include/linux/sysctl.h	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/include/linux/sysctl.h	2007-01-09 14:34:55.000000000 -0500
@@ -154,6 +154,7 @@ enum
 	KERN_COMPAT_LOG=73,	/* int: print compat layer  messages */
 	KERN_MAX_LOCK_DEPTH=74,
 	KERN_KDUMP_ON_INIT=75,	/* int: ia64 kdump with INIT */
+ 	KERN_PANIC_ON_NMI=76, /* int: whether we will panic on an unrecovered */
 };
 
 
diff -urNp linux-2.6.18.x86_64.orig/kernel/sysctl.c linux-2.6.18.x86_64/kernel/sysctl.c
--- linux-2.6.18.x86_64.orig/kernel/sysctl.c	2007-01-09 14:34:35.000000000 -0500
+++ linux-2.6.18.x86_64/kernel/sysctl.c	2007-01-09 14:35:09.000000000 -0500
@@ -52,6 +52,10 @@
 extern int proc_nr_files(ctl_table *table, int write, struct file *filp,
                      void __user *buffer, size_t *lenp, loff_t *ppos);
 
+#ifdef CONFIG_X86
+#include <asm/nmi.h>
+#endif
+
 #if defined(CONFIG_SYSCTL)
 
 /* External variables not in a header file. */
@@ -75,7 +79,6 @@ extern int percpu_pagelist_fraction;
 extern int compat_log;
 
 #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86)
-int unknown_nmi_panic;
 extern int proc_unknown_nmi_panic(ctl_table *, int, struct file *,
 				  void __user *, size_t *, loff_t *);
 #endif
@@ -689,6 +692,14 @@ static ctl_table kern_table[] = {
 #endif
 #if defined(CONFIG_X86)
 	{
+		.ctl_name	= KERN_PANIC_ON_NMI,
+		.procname	= "panic_on_unrecovered_nmi",
+		.data		= &panic_on_unrecovered_nmi,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec,
+	},
+	{
 		.ctl_name	= KERN_BOOTLOADER_TYPE,
 		.procname	= "bootloader_type",
 		.data		= &bootloader_type,