Re: [PATCH] x86: Make mwait_usable() respect "idle=nomwait" kernelparameter

From: Srivatsa S. Bhat
Date: Tue Apr 03 2012 - 07:50:09 EST


On 04/02/2012 07:36 PM, Srivatsa S. Bhat wrote:

> mwait_usable() returns 1 even if the "idle=nomwait" kernel parameter is passed.
> Fix it by adding a check for boot_option_idle_override == IDLE_NOMWAIT and
> returning 0 if it is set.
>
> Before applying the patch (dmesg snippet):
> [ 0.000000] Command line: [...] idle=nomwait
> [ 0.000000] Kernel command line: [...] idle=nomwait
> [ 0.000000] RCU dyntick-idle grace-period acceleration is enabled.
> [ 0.140606] using mwait in idle threads. <======= mwait being used
> [ 4.303986] cpuidle: using governor ladder
> [ 4.308232] cpuidle: using governor menu
>
> After applying the patch:
> [ 0.000000] Command line: [...] idle=nomwait
> [ 0.000000] Kernel command line: [...] idle=nomwait
> [ 0.000000] RCU dyntick-idle grace-period acceleration is enabled.
> [ 4.264100] cpuidle: using governor ladder
> [ 4.268342] cpuidle: using governor menu
>
> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@xxxxxxxxxxxxxxxxxx>
> ---
>
> arch/x86/kernel/process.c | 3 +++
> 1 files changed, 3 insertions(+), 0 deletions(-)
>
> diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
> index a33afaa..945fbf0 100644
> --- a/arch/x86/kernel/process.c
> +++ b/arch/x86/kernel/process.c
> @@ -621,6 +621,9 @@ int mwait_usable(const struct cpuinfo_x86 *c)
> if (boot_option_idle_override == IDLE_FORCE_MWAIT)
> return 1;
>
> + if (boot_option_idle_override == IDLE_NOMWAIT)
> + return 0;
> +
> if (c->cpuid_level < MWAIT_INFO)
> return 0;
>
>


I realized that actually more stuff is broken than what the above patch fixes.

So here is the updated patch:

---

From: Srivatsa S. Bhat <srivatsa.bhat@xxxxxxxxxxxxxxxxxx>
Subject: [v2] x86: Make mwait_usable() heed to "idle=" kernel parameters properly

The checks that exist in mwait_usable() for "idle=" kernel parameters
are insufficient. As a result, mwait_usable() can return 1 even if
"idle=nomwait" or "idle=poll" or "idle=halt" parameters are passed.

Of these cases, incorrect handling of idle=nomwait is a universal problem
since mwait can get used for usual CPU idling. However the rest of the
cases are problematic only during CPU Hotplug (offline) because, in the
CPU offline path, the function mwait_play_dead() is called, which might
result in mwait being used in the offline CPUs, if mwait_usable() happens
to return 1.

Fix these issues by checking for the boot time "idle=" kernel parameter
properly in mwait_usable().

The first issue (usual cpu idling) is demonstrated below:

Before applying the patch (dmesg snippet):
[ 0.000000] Command line: [...] idle=nomwait
[ 0.000000] Kernel command line: [...] idle=nomwait
[ 0.000000] RCU dyntick-idle grace-period acceleration is enabled.
[ 0.140606] using mwait in idle threads. <======= mwait being used
[ 4.303986] cpuidle: using governor ladder
[ 4.308232] cpuidle: using governor menu

After applying the patch:
[ 0.000000] Command line: [...] idle=nomwait
[ 0.000000] Kernel command line: [...] idle=nomwait
[ 0.000000] RCU dyntick-idle grace-period acceleration is enabled.
[ 4.264100] cpuidle: using governor ladder
[ 4.268342] cpuidle: using governor menu

Signed-off-by: Deepthi Dharwar <deepthi@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@xxxxxxxxxxxxxxxxxx>
---

arch/x86/kernel/process.c | 8 ++++++++
1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index a33afaa..b526c4e 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -618,9 +618,17 @@ int mwait_usable(const struct cpuinfo_x86 *c)
{
u32 eax, ebx, ecx, edx;

+ /* Use mwait if idle=mwait boot option is given */
if (boot_option_idle_override == IDLE_FORCE_MWAIT)
return 1;

+ /*
+ * Any idle= boot option other than idle=mwait means that we must not
+ * use mwait. Eg: idle=halt or idle=poll or idle=nomwait
+ */
+ if (boot_option_idle_override != IDLE_NO_OVERRIDE)
+ return 0;
+
if (c->cpuid_level < MWAIT_INFO)
return 0;



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