ACPI: Processor native C-states using MWAIT

Intel processors starting with the Core Duo support
support processor native C-state using the MWAIT instruction.
Refer: Intel Architecture Software Developer's Manual
http://www.intel.com/design/Pentium4/manuals/253668.htm

Platform firmware exports the support for Native C-state to OS using
ACPI _PDC and _CST methods.
Refer: Intel Processor Vendor-Specific ACPI: Interface Specification
http://www.intel.com/technology/iapc/acpi/downloads/302223.htm

With Processor Native C-state, we use 'MWAIT' instruction on the processor
to enter different C-states (C1, C2, C3).  We won't use the special IO
ports to enter C-state and no SMM mode etc required to enter C-state.
Overall this will mean better C-state support.

One major advantage of using MWAIT for all C-states is, with this and
"treat interrupt as break event" feature of MWAIT, we can now get accurate
timing for the time spent in C1, C2, ..  states.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Len Brown <len.brown@intel.com>

3 files changed, 151 insertions, 15 deletions

diff --git a/arch/i386/kernel/acpi/cstate.c b/arch/i386/kernel/acpi/cstate.c
index 25db49ef1770..20563e52c622 100644
--- a/arch/i386/kernel/acpi/cstate.c
+++ b/arch/i386/kernel/acpi/cstate.c
@@ -10,6 +10,7 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/acpi.h>
+#include <linux/cpu.h>
 
 #include <acpi/processor.h>
 #include <asm/acpi.h>
@@ -41,5 +42,124 @@ void acpi_processor_power_init_bm_check(struct acpi_processor_flags *flags,
                 flags->bm_check = 1;
         }
 }
-
 EXPORT_SYMBOL(acpi_processor_power_init_bm_check);
+
+/* The code below handles cstate entry with monitor-mwait pair on Intel*/
+
+struct cstate_entry_s {
+        struct {
+                unsigned int eax;
+                unsigned int ecx;
+        } states[ACPI_PROCESSOR_MAX_POWER];
+};
+static struct cstate_entry_s *cpu_cstate_entry; /* per CPU ptr */
+
+static short mwait_supported[ACPI_PROCESSOR_MAX_POWER];
+
+#define MWAIT_SUBSTATE_MASK     (0xf)
+#define MWAIT_SUBSTATE_SIZE     (4)
+
+#define CPUID_MWAIT_LEAF (5)
+#define CPUID5_ECX_EXTENSIONS_SUPPORTED (0x1)
+#define CPUID5_ECX_INTERRUPT_BREAK      (0x2)
+
+#define MWAIT_ECX_INTERRUPT_BREAK       (0x1)
+
+#define NATIVE_CSTATE_BEYOND_HALT       (2)
+
+int acpi_processor_ffh_cstate_probe(unsigned int cpu,
+                struct acpi_processor_cx *cx, struct acpi_power_register *reg)
+{
+        struct cstate_entry_s *percpu_entry;
+        struct cpuinfo_x86 *c = cpu_data + cpu;
+
+        cpumask_t saved_mask;
+        int retval;
+        unsigned int eax, ebx, ecx, edx;
+        unsigned int edx_part;
+        unsigned int cstate_type; /* C-state type and not ACPI C-state type */
+        unsigned int num_cstate_subtype;
+
+        if (!cpu_cstate_entry || c->cpuid_level < CPUID_MWAIT_LEAF )
+                return -1;
+
+        if (reg->bit_offset != NATIVE_CSTATE_BEYOND_HALT)
+                return -1;
+
+        percpu_entry = per_cpu_ptr(cpu_cstate_entry, cpu);
+        percpu_entry->states[cx->index].eax = 0;
+        percpu_entry->states[cx->index].ecx = 0;
+
+        /* Make sure we are running on right CPU */
+        saved_mask = current->cpus_allowed;
+        retval = set_cpus_allowed(current, cpumask_of_cpu(cpu));
+        if (retval)
+                return -1;
+
+        cpuid(CPUID_MWAIT_LEAF, &eax, &ebx, &ecx, &edx);
+
+        /* Check whether this particular cx_type (in CST) is supported or not */
+        cstate_type = (cx->address >> MWAIT_SUBSTATE_SIZE) + 1;
+        edx_part = edx >> (cstate_type * MWAIT_SUBSTATE_SIZE);
+        num_cstate_subtype = edx_part & MWAIT_SUBSTATE_MASK;
+
+        retval = 0;
+        if (num_cstate_subtype < (cx->address & MWAIT_SUBSTATE_MASK)) {
+                retval = -1;
+                goto out;
+        }
+
+        /* mwait ecx extensions INTERRUPT_BREAK should be supported for C2/C3 */
+        if (!(ecx & CPUID5_ECX_EXTENSIONS_SUPPORTED) ||
+            !(ecx & CPUID5_ECX_INTERRUPT_BREAK)) {
+                retval = -1;
+                goto out;
+        }
+        percpu_entry->states[cx->index].ecx = MWAIT_ECX_INTERRUPT_BREAK;
+
+        /* Use the hint in CST */
+        percpu_entry->states[cx->index].eax = cx->address;
+
+        if (!mwait_supported[cstate_type]) {
+                mwait_supported[cstate_type] = 1;
+                printk(KERN_DEBUG "Monitor-Mwait will be used to enter C-%d "
+                       "state\n", cx->type);
+        }
+
+out:
+        set_cpus_allowed(current, saved_mask);
+        return retval;
+}
+EXPORT_SYMBOL_GPL(acpi_processor_ffh_cstate_probe);
+
+void acpi_processor_ffh_cstate_enter(struct acpi_processor_cx *cx)
+{
+        unsigned int cpu = smp_processor_id();
+        struct cstate_entry_s *percpu_entry;
+
+        percpu_entry = per_cpu_ptr(cpu_cstate_entry, cpu);
+        mwait_idle_with_hints(percpu_entry->states[cx->index].eax,
+                              percpu_entry->states[cx->index].ecx);
+}
+EXPORT_SYMBOL_GPL(acpi_processor_ffh_cstate_enter);
+
+static int __init ffh_cstate_init(void)
+{
+        struct cpuinfo_x86 *c = &boot_cpu_data;
+        if (c->x86_vendor != X86_VENDOR_INTEL)
+                return -1;
+
+        cpu_cstate_entry = alloc_percpu(struct cstate_entry_s);
+        return 0;
+}
+
+static void __exit ffh_cstate_exit(void)
+{
+        if (cpu_cstate_entry) {
+                free_percpu(cpu_cstate_entry);
+                cpu_cstate_entry = NULL;
+        }
+}
+
+arch_initcall(ffh_cstate_init);
+__exitcall(ffh_cstate_exit);
diff --git a/arch/i386/kernel/process.c b/arch/i386/kernel/process.c
index b0a07801d9df..57d375900afb 100644
--- a/arch/i386/kernel/process.c
+++ b/arch/i386/kernel/process.c
@@ -236,20 +236,28 @@ EXPORT_SYMBOL_GPL(cpu_idle_wait);
  * We execute MONITOR against need_resched and enter optimized wait state
  * through MWAIT. Whenever someone changes need_resched, we would be woken
  * up from MWAIT (without an IPI).
+ *
+ * New with Core Duo processors, MWAIT can take some hints based on CPU
+ * capability.
  */
-static void mwait_idle(void)
+void mwait_idle_with_hints(unsigned long eax, unsigned long ecx)
 {
-        local_irq_enable();
-
-        while (!need_resched()) {
+        if (!need_resched()) {
                 __monitor((void *)&current_thread_info()->flags, 0, 0);
                 smp_mb();
-                if (need_resched())
-                        break;
-                __mwait(0, 0);
+                if (!need_resched())
+                        __mwait(eax, ecx);
         }
 }
 
+/* Default MONITOR/MWAIT with no hints, used for default C1 state */
+static void mwait_idle(void)
+{
+        local_irq_enable();
+        while (!need_resched())
+                mwait_idle_with_hints(0, 0);
+}
+
 void __devinit select_idle_routine(const struct cpuinfo_x86 *c)
 {
         if (cpu_has(c, X86_FEATURE_MWAIT)) {
diff --git a/arch/x86_64/kernel/process.c b/arch/x86_64/kernel/process.c
index 5e95b257ee26..49f7fac6229e 100644
--- a/arch/x86_64/kernel/process.c
+++ b/arch/x86_64/kernel/process.c
@@ -238,20 +238,28 @@ void cpu_idle (void)
  * We execute MONITOR against need_resched and enter optimized wait state
  * through MWAIT. Whenever someone changes need_resched, we would be woken
  * up from MWAIT (without an IPI).
+ *
+ * New with Core Duo processors, MWAIT can take some hints based on CPU
+ * capability.
  */
-static void mwait_idle(void)
+void mwait_idle_with_hints(unsigned long eax, unsigned long ecx)
 {
-        local_irq_enable();
-
-        while (!need_resched()) {
+        if (!need_resched()) {
                 __monitor((void *)&current_thread_info()->flags, 0, 0);
                 smp_mb();
-                if (need_resched())
-                        break;
-                __mwait(0, 0);
+                if (!need_resched())
+                        __mwait(eax, ecx);
         }
 }
 
+/* Default MONITOR/MWAIT with no hints, used for default C1 state */
+static void mwait_idle(void)
+{
+        local_irq_enable();
+        while (!need_resched())
+                mwait_idle_with_hints(0,0);
+}
+
 void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c)
 {
         static int printed;