x86 idle: remove mwait_idle() and "idle=mwait" cmdline param

mwait_idle() is a C1-only idle loop intended to be more efficient than HLT, starting on Pentium-4 HT-enabled processors. But mwait_idle() has been replaced by the more general mwait_idle_with_hints(), which handles both C1 and deeper C-states. ACPI processor_idle and intel_idle use only mwait_idle_with_hints(), and no longer use mwait_idle(). Here we simplify the x86 native idle code by removing mwait_idle(), and the "idle=mwait" bootparam used to invoke it. Since Linux 3.0 there has been a boot-time warning when "idle=mwait" was invoked saying it would be removed in 2012. This removal was also noted in the (now removed:-) feature-removal-schedule.txt. After this change, kernels configured with (CONFIG_ACPI=n && CONFIG_INTEL_IDLE=n) when run on hardware that supports MWAIT will simply use HLT. If MWAIT is desired on those systems, cpuidle and the cpuidle drivers above can be enabled. Signed-off-by: Len Brown <len.brown@intel.com> Cc: x86@kernel.org

x86 idle: remove mwait_idle() and "idle=mwait" cmdline param
mwait_idle() is a C1-only idle loop intended to be more efficient than HLT, starting on Pentium-4 HT-enabled processors. But mwait_idle() has been replaced by the more general mwait_idle_with_hints(), which handles both C1 and deeper C-states. ACPI processor_idle and intel_idle use only mwait_idle_with_hints(), and no longer use mwait_idle(). Here we simplify the x86 native idle code by removing mwait_idle(), and the "idle=mwait" bootparam used to invoke it. Since Linux 3.0 there has been a boot-time warning when "idle=mwait" was invoked saying it would be removed in 2012. This removal was also noted in the (now removed:-) feature-removal-schedule.txt. After this change, kernels configured with (CONFIG_ACPI=n && CONFIG_INTEL_IDLE=n) when run on hardware that supports MWAIT will simply use HLT. If MWAIT is desired on those systems, cpuidle and the cpuidle drivers above can be enabled. Signed-off-by: Len Brown <len.brown@intel.com> Cc: x86@kernel.org
Len Brown · Eric Lee · Eric Lee · Eric Lee · Eric Lee · Eric Lee
1 parent 6a377ddc4e
Showing 5 changed files with 4 additions and 87 deletions Side-by-side Diff
Documentation/kernel-parameters.txt
arch/x86/include/asm/processor.h
arch/x86/kernel/process.c
arch/x86/kernel/smpboot.c
drivers/acpi/processor_idle.c
@@ -1039,16 +1039,11 @@
 			Claim all unknown PCI IDE storage controllers.
  
 	idle=		[X86]
-			Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
+			Format: idle=poll, idle=halt, idle=nomwait
 			Poll forces a polling idle loop that can slightly
 			improve the performance of waking up a idle CPU, but
 			will use a lot of power and make the system run hot.
 			Not recommended.
-			idle=mwait: On systems which support MONITOR/MWAIT but
-			the kernel chose to not use it because it doesn't save
-			as much power as a normal idle loop, use the
-			MONITOR/MWAIT idle loop anyways. Performance should be
-			the same as idle=poll.
 			idle=halt: Halt is forced to be used for CPU idle.
 			In such case C2/C3 won't be used again.
 			idle=nomwait: Disable mwait for CPU C-states
@@ -725,7 +725,7 @@
 extern bool			amd_e400_c1e_detected;
  
 enum idle_boot_override {IDLE_NO_OVERRIDE=0, IDLE_HALT, IDLE_NOMWAIT,
-			 IDLE_POLL, IDLE_FORCE_MWAIT};
+			 IDLE_POLL};
  
 extern void enable_sep_cpu(void);
 extern int sysenter_setup(void);
@@ -421,27 +421,6 @@
 	}
 }
  
-/* Default MONITOR/MWAIT with no hints, used for default C1 state */
-static void mwait_idle(void)
-{
-	if (!need_resched()) {
-		trace_power_start_rcuidle(POWER_CSTATE, 1, smp_processor_id());
-		trace_cpu_idle_rcuidle(1, smp_processor_id());
-		if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR))
-			clflush((void *)&current_thread_info()->flags);
-
-		__monitor((void *)&current_thread_info()->flags, 0, 0);
-		smp_mb();
-		if (!need_resched())
-			__sti_mwait(0, 0);
-		else
-			local_irq_enable();
-		trace_power_end_rcuidle(smp_processor_id());
-		trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id());
-	} else
-		local_irq_enable();
-}
-
 /*
  * On SMP it's slightly faster (but much more power-consuming!)
  * to poll the ->work.need_resched flag instead of waiting for the
@@ -458,53 +437,6 @@
 	trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id());
 }
  
-/*
- * mwait selection logic:
- *
- * It depends on the CPU. For AMD CPUs that support MWAIT this is
- * wrong. Family 0x10 and 0x11 CPUs will enter C1 on HLT. Powersavings
- * then depend on a clock divisor and current Pstate of the core. If
- * all cores of a processor are in halt state (C1) the processor can
- * enter the C1E (C1 enhanced) state. If mwait is used this will never
- * happen.
- *
- * idle=mwait overrides this decision and forces the usage of mwait.
- */
-
-#define MWAIT_INFO			0x05
-#define MWAIT_ECX_EXTENDED_INFO		0x01
-#define MWAIT_EDX_C1			0xf0
-
-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;
-
-	cpuid(MWAIT_INFO, &eax, &ebx, &ecx, &edx);
-	/* Check, whether EDX has extended info about MWAIT */
-	if (!(ecx & MWAIT_ECX_EXTENDED_INFO))
-		return 1;
-
-	/*
-	 * edx enumeratios MONITOR/MWAIT extensions. Check, whether
-	 * C1  supports MWAIT
-	 */
-	return (edx & MWAIT_EDX_C1);
-}
-
 bool amd_e400_c1e_detected;
 EXPORT_SYMBOL(amd_e400_c1e_detected);
  
@@ -576,13 +508,7 @@
 	if (pm_idle)
 		return;
  
-	if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) {
-		/*
-		 * One CPU supports mwait => All CPUs supports mwait
-		 */
-		pr_info("using mwait in idle threads\n");
-		pm_idle = mwait_idle;
-	} else if (cpu_has_amd_erratum(amd_erratum_400)) {
+	if (cpu_has_amd_erratum(amd_erratum_400)) {
 		/* E400: APIC timer interrupt does not wake up CPU from C1e */
 		pr_info("using AMD E400 aware idle routine\n");
 		pm_idle = amd_e400_idle;
@@ -606,9 +532,6 @@
 		pr_info("using polling idle threads\n");
 		pm_idle = poll_idle;
 		boot_option_idle_override = IDLE_POLL;
-	} else if (!strcmp(str, "mwait")) {
-		boot_option_idle_override = IDLE_FORCE_MWAIT;
-		WARN_ONCE(1, "\"idle=mwait\" will be removed in 2012\n");
 	} else if (!strcmp(str, "halt")) {
 		/*
 		 * When the boot option of idle=halt is added, halt is
@@ -1369,7 +1369,7 @@
 	void *mwait_ptr;
 	struct cpuinfo_x86 *c = __this_cpu_ptr(&cpu_info);
  
-	if (!(this_cpu_has(X86_FEATURE_MWAIT) && mwait_usable(c)))
+	if (!this_cpu_has(X86_FEATURE_MWAIT))
 		return;
 	if (!this_cpu_has(X86_FEATURE_CLFLSH))
 		return;
@@ -84,7 +84,6 @@
 static int disabled_by_idle_boot_param(void)
 {
 	return boot_option_idle_override == IDLE_POLL ||
-		boot_option_idle_override == IDLE_FORCE_MWAIT ||
 		boot_option_idle_override == IDLE_HALT;
 }
...	...	@@ -1039,16 +1039,11 @@
1039	1039	Claim all unknown PCI IDE storage controllers.
1040	1040
1041	1041	idle= [X86]
1042		- Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
	1042	+ Format: idle=poll, idle=halt, idle=nomwait
1043	1043	Poll forces a polling idle loop that can slightly
1044	1044	improve the performance of waking up a idle CPU, but
1045	1045	will use a lot of power and make the system run hot.
1046	1046	Not recommended.
1047		- idle=mwait: On systems which support MONITOR/MWAIT but
1048		- the kernel chose to not use it because it doesn't save
1049		- as much power as a normal idle loop, use the
1050		- MONITOR/MWAIT idle loop anyways. Performance should be
1051		- the same as idle=poll.
1052	1047	idle=halt: Halt is forced to be used for CPU idle.
1053	1048	In such case C2/C3 won't be used again.
1054	1049	idle=nomwait: Disable mwait for CPU C-states
...	...	@@ -725,7 +725,7 @@
725	725	extern bool amd_e400_c1e_detected;
726	726
727	727	enum idle_boot_override {IDLE_NO_OVERRIDE=0, IDLE_HALT, IDLE_NOMWAIT,
728		- IDLE_POLL, IDLE_FORCE_MWAIT};
	728	+ IDLE_POLL};
729	729
730	730	extern void enable_sep_cpu(void);
731	731	extern int sysenter_setup(void);
...	...	@@ -421,27 +421,6 @@
421	421	}
422	422	}
423	423
424		-/* Default MONITOR/MWAIT with no hints, used for default C1 state */
425		-static void mwait_idle(void)
426		-{
427		- if (!need_resched()) {
428		- trace_power_start_rcuidle(POWER_CSTATE, 1, smp_processor_id());
429		- trace_cpu_idle_rcuidle(1, smp_processor_id());
430		- if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR))
431		- clflush((void *)&current_thread_info()->flags);
432		-
433		- __monitor((void *)&current_thread_info()->flags, 0, 0);
434		- smp_mb();
435		- if (!need_resched())
436		- __sti_mwait(0, 0);
437		- else
438		- local_irq_enable();
439		- trace_power_end_rcuidle(smp_processor_id());
440		- trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id());
441		- } else
442		- local_irq_enable();
443		-}
444		-
445	424	/*
446	425	* On SMP it's slightly faster (but much more power-consuming!)
447	426	* to poll the ->work.need_resched flag instead of waiting for the
...	...	@@ -458,53 +437,6 @@
458	437	trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id());
459	438	}
460	439
461		-/*
462		- * mwait selection logic:
463		- *
464		- * It depends on the CPU. For AMD CPUs that support MWAIT this is
465		- * wrong. Family 0x10 and 0x11 CPUs will enter C1 on HLT. Powersavings
466		- * then depend on a clock divisor and current Pstate of the core. If
467		- * all cores of a processor are in halt state (C1) the processor can
468		- * enter the C1E (C1 enhanced) state. If mwait is used this will never
469		- * happen.
470		- *
471		- * idle=mwait overrides this decision and forces the usage of mwait.
472		- */
473		-
474		-#define MWAIT_INFO 0x05
475		-#define MWAIT_ECX_EXTENDED_INFO 0x01
476		-#define MWAIT_EDX_C1 0xf0
477		-
478		-int mwait_usable(const struct cpuinfo_x86 *c)
479		-{
480		- u32 eax, ebx, ecx, edx;
481		-
482		- /* Use mwait if idle=mwait boot option is given */
483		- if (boot_option_idle_override == IDLE_FORCE_MWAIT)
484		- return 1;
485		-
486		- /*
487		- * Any idle= boot option other than idle=mwait means that we must not
488		- * use mwait. Eg: idle=halt or idle=poll or idle=nomwait
489		- */
490		- if (boot_option_idle_override != IDLE_NO_OVERRIDE)
491		- return 0;
492		-
493		- if (c->cpuid_level < MWAIT_INFO)
494		- return 0;
495		-
496		- cpuid(MWAIT_INFO, &eax, &ebx, &ecx, &edx);
497		- /* Check, whether EDX has extended info about MWAIT */
498		- if (!(ecx & MWAIT_ECX_EXTENDED_INFO))
499		- return 1;
500		-
501		- /*
502		- * edx enumeratios MONITOR/MWAIT extensions. Check, whether
503		- * C1 supports MWAIT
504		- */
505		- return (edx & MWAIT_EDX_C1);
506		-}
507		-
508	440	bool amd_e400_c1e_detected;
509	441	EXPORT_SYMBOL(amd_e400_c1e_detected);
510	442
...	...	@@ -576,13 +508,7 @@
576	508	if (pm_idle)
577	509	return;
578	510
579		- if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) {
580		- /*
581		- * One CPU supports mwait => All CPUs supports mwait
582		- */
583		- pr_info("using mwait in idle threads\n");
584		- pm_idle = mwait_idle;
585		- } else if (cpu_has_amd_erratum(amd_erratum_400)) {
	511	+ if (cpu_has_amd_erratum(amd_erratum_400)) {
586	512	/* E400: APIC timer interrupt does not wake up CPU from C1e */
587	513	pr_info("using AMD E400 aware idle routine\n");
588	514	pm_idle = amd_e400_idle;
...	...	@@ -606,9 +532,6 @@
606	532	pr_info("using polling idle threads\n");
607	533	pm_idle = poll_idle;
608	534	boot_option_idle_override = IDLE_POLL;
609		- } else if (!strcmp(str, "mwait")) {
610		- boot_option_idle_override = IDLE_FORCE_MWAIT;
611		- WARN_ONCE(1, "\"idle=mwait\" will be removed in 2012\n");
612	535	} else if (!strcmp(str, "halt")) {
613	536	/*
614	537	* When the boot option of idle=halt is added, halt is
...	...	@@ -1369,7 +1369,7 @@
1369	1369	void *mwait_ptr;
1370	1370	struct cpuinfo_x86 *c = __this_cpu_ptr(&cpu_info);
1371	1371
1372		- if (!(this_cpu_has(X86_FEATURE_MWAIT) && mwait_usable(c)))
	1372	+ if (!this_cpu_has(X86_FEATURE_MWAIT))
1373	1373	return;
1374	1374	if (!this_cpu_has(X86_FEATURE_CLFLSH))
1375	1375	return;
...	...	@@ -84,7 +84,6 @@
84	84	static int disabled_by_idle_boot_param(void)
85	85	{
86	86	return boot_option_idle_override == IDLE_POLL \|\|
87		- boot_option_idle_override == IDLE_FORCE_MWAIT \|\|
88	87	boot_option_idle_override == IDLE_HALT;
89	88	}
90	89