[PATCH] x86: default-disable MWAIT-based idle driver for CPUs without ARAT

[PATCH] x86: default-disable MWAIT-based idle driver for CPUs without ARAT

[Jan Beulich]

[-- Attachment #1: Type: text/plain, Size: 1102 bytes --]

Without ARAT, and apparently only when using HPET broadcast mode as
replacement, CPUs occasionally fail to wake up, causing the system to
(transiently) hang. Until the reason is understood, disable the driver
on such systems.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

--- a/xen/arch/x86/cpu/mwait-idle.c
+++ b/xen/arch/x86/cpu/mwait-idle.c
@@ -71,7 +71,7 @@
 # define pr_debug(fmt...)
 #endif
 
-static __initdata bool_t no_mwait_idle;
+static __initdata s8 no_mwait_idle = -1;
 invbool_param("mwait-idle", no_mwait_idle);
 
 static unsigned int mwait_substates;
@@ -500,6 +500,13 @@ int __init mwait_idle_init(struct notifi
 	if (pm_idle_save)
 		return -ENODEV;
 
+	/* XXX The no-ARAT case is supposedly being taken care of, but at
+	 * least some systems without ARAT hang for some reason, apparently
+	 * only when using HPET broadcast mode (PIT broadcast mode seems to
+	 * be fine). */
+	if (no_mwait_idle < 0 && boot_cpu_has(X86_FEATURE_ARAT))
+		no_mwait_idle = 0;
+
 	err = mwait_idle_probe();
 	if (!err) {
 		if (!boot_cpu_has(X86_FEATURE_ARAT))




[-- Attachment #2: x86-mwait-idle-no-ARAT.patch --]
[-- Type: text/plain, Size: 1166 bytes --]

x86: default-disable MWAIT-based idle driver for CPUs without ARAT

Without ARAT, and apparently only when using HPET broadcast mode as
replacement, CPUs occasionally fail to wake up, causing the system to
(transiently) hang. Until the reason is understood, disable the driver
on such systems.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

--- a/xen/arch/x86/cpu/mwait-idle.c
+++ b/xen/arch/x86/cpu/mwait-idle.c
@@ -71,7 +71,7 @@
 # define pr_debug(fmt...)
 #endif
 
-static __initdata bool_t no_mwait_idle;
+static __initdata s8 no_mwait_idle = -1;
 invbool_param("mwait-idle", no_mwait_idle);
 
 static unsigned int mwait_substates;
@@ -500,6 +500,13 @@ int __init mwait_idle_init(struct notifi
 	if (pm_idle_save)
 		return -ENODEV;
 
+	/* XXX The no-ARAT case is supposedly being taken care of, but at
+	 * least some systems without ARAT hang for some reason, apparently
+	 * only when using HPET broadcast mode (PIT broadcast mode seems to
+	 * be fine). */
+	if (no_mwait_idle < 0 && boot_cpu_has(X86_FEATURE_ARAT))
+		no_mwait_idle = 0;
+
 	err = mwait_idle_probe();
 	if (!err) {
 		if (!boot_cpu_has(X86_FEATURE_ARAT))

[-- Attachment #3: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH] x86: default-disable MWAIT-based idle driver for CPUs without ARAT

[Keir Fraser]

On 26/09/2012 15:00, "Jan Beulich" <JBeulich@suse.com> wrote:

> Without ARAT, and apparently only when using HPET broadcast mode as
> replacement, CPUs occasionally fail to wake up, causing the system to
> (transiently) hang. Until the reason is understood, disable the driver
> on such systems.
> 
> Signed-off-by: Jan Beulich <jbeulich@suse.com>

Acked-by: Keir Fraser <keir@xen.org>

> --- a/xen/arch/x86/cpu/mwait-idle.c
> +++ b/xen/arch/x86/cpu/mwait-idle.c
> @@ -71,7 +71,7 @@
>  # define pr_debug(fmt...)
>  #endif
>  
> -static __initdata bool_t no_mwait_idle;
> +static __initdata s8 no_mwait_idle = -1;
>  invbool_param("mwait-idle", no_mwait_idle);
>  
>  static unsigned int mwait_substates;
> @@ -500,6 +500,13 @@ int __init mwait_idle_init(struct notifi
> if (pm_idle_save)
> return -ENODEV;
>  
> + /* XXX The no-ARAT case is supposedly being taken care of, but at
> +  * least some systems without ARAT hang for some reason, apparently
> +  * only when using HPET broadcast mode (PIT broadcast mode seems to
> +  * be fine). */
> + if (no_mwait_idle < 0 && boot_cpu_has(X86_FEATURE_ARAT))
> +  no_mwait_idle = 0;
> +
> err = mwait_idle_probe();
> if (!err) {
> if (!boot_cpu_has(X86_FEATURE_ARAT))
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel