[PATCH] suspend/resume SMP support

Using CPU hotplug to support suspend/resume SMP.  Both S3 and S4 use
disable/enable_nonboot_cpus API.  The S4 part is based on Pavel's original S4
SMP patch.

Signed-off-by: Li Shaohua<shaohua.li@intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
diff --git a/arch/i386/kernel/cpu/mcheck/k7.c b/arch/i386/kernel/cpu/mcheck/k7.c
index 8df52e8..c4abe76 100644
--- a/arch/i386/kernel/cpu/mcheck/k7.c
+++ b/arch/i386/kernel/cpu/mcheck/k7.c
@@ -69,7 +69,7 @@
 
 
 /* AMD K7 machine check is Intel like */
-void __init amd_mcheck_init(struct cpuinfo_x86 *c)
+void __devinit amd_mcheck_init(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
 	int i;
diff --git a/arch/i386/kernel/cpu/mcheck/mce.c b/arch/i386/kernel/cpu/mcheck/mce.c
index 7218a73..2cf25d2 100644
--- a/arch/i386/kernel/cpu/mcheck/mce.c
+++ b/arch/i386/kernel/cpu/mcheck/mce.c
@@ -16,7 +16,7 @@
 
 #include "mce.h"
 
-int mce_disabled __initdata = 0;
+int mce_disabled __devinitdata = 0;
 int nr_mce_banks;
 
 EXPORT_SYMBOL_GPL(nr_mce_banks);	/* non-fatal.o */
diff --git a/arch/i386/kernel/cpu/mcheck/p4.c b/arch/i386/kernel/cpu/mcheck/p4.c
index 8b16ceb..0abccb6 100644
--- a/arch/i386/kernel/cpu/mcheck/p4.c
+++ b/arch/i386/kernel/cpu/mcheck/p4.c
@@ -78,7 +78,7 @@
 }
 
 /* P4/Xeon Thermal regulation detect and init */
-static void __init intel_init_thermal(struct cpuinfo_x86 *c)
+static void __devinit intel_init_thermal(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
 	unsigned int cpu = smp_processor_id();
@@ -232,7 +232,7 @@
 }
 
 
-void __init intel_p4_mcheck_init(struct cpuinfo_x86 *c)
+void __devinit intel_p4_mcheck_init(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
 	int i;
diff --git a/arch/i386/kernel/cpu/mcheck/p6.c b/arch/i386/kernel/cpu/mcheck/p6.c
index 46640f8..f01b73f 100644
--- a/arch/i386/kernel/cpu/mcheck/p6.c
+++ b/arch/i386/kernel/cpu/mcheck/p6.c
@@ -80,7 +80,7 @@
 }
 
 /* Set up machine check reporting for processors with Intel style MCE */
-void __init intel_p6_mcheck_init(struct cpuinfo_x86 *c)
+void __devinit intel_p6_mcheck_init(struct cpuinfo_x86 *c)
 {
 	u32 l, h;
 	int i;
diff --git a/arch/i386/kernel/cpu/mcheck/winchip.c b/arch/i386/kernel/cpu/mcheck/winchip.c
index 753fa7a..7bae68f 100644
--- a/arch/i386/kernel/cpu/mcheck/winchip.c
+++ b/arch/i386/kernel/cpu/mcheck/winchip.c
@@ -23,7 +23,7 @@
 }
 
 /* Set up machine check reporting on the Winchip C6 series */
-void __init winchip_mcheck_init(struct cpuinfo_x86 *c)
+void __devinit winchip_mcheck_init(struct cpuinfo_x86 *c)
 {
 	u32 lo, hi;
 	machine_check_vector = winchip_machine_check;
diff --git a/drivers/acpi/Kconfig b/drivers/acpi/Kconfig
index 670fdb5..86c5252 100644
--- a/drivers/acpi/Kconfig
+++ b/drivers/acpi/Kconfig
@@ -55,7 +55,7 @@
 
 config ACPI_SLEEP
 	bool "Sleep States (EXPERIMENTAL)"
-	depends on X86
+	depends on X86 && (!SMP || SUSPEND_SMP)
 	depends on EXPERIMENTAL && PM
 	default y
 	---help---
diff --git a/include/linux/suspend.h b/include/linux/suspend.h
index 2bf0d5f..f2e96fd 100644
--- a/include/linux/suspend.h
+++ b/include/linux/suspend.h
@@ -58,7 +58,7 @@
 }
 #endif
 
-#ifdef CONFIG_SMP
+#ifdef CONFIG_SUSPEND_SMP
 extern void disable_nonboot_cpus(void);
 extern void enable_nonboot_cpus(void);
 #else
diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig
index 696387f..fdb3776 100644
--- a/kernel/power/Kconfig
+++ b/kernel/power/Kconfig
@@ -28,7 +28,7 @@
 
 config SOFTWARE_SUSPEND
 	bool "Software Suspend (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && PM && SWAP
+	depends on EXPERIMENTAL && PM && SWAP && (SUSPEND_SMP || !SMP)
 	---help---
 	  Enable the possibility of suspending the machine.
 	  It doesn't need APM.
@@ -72,3 +72,7 @@
 	  suspended image to. It will simply pick the first available swap 
 	  device.
 
+config SUSPEND_SMP
+	bool
+	depends on HOTPLUG_CPU && X86 && PM
+	default y
diff --git a/kernel/power/Makefile b/kernel/power/Makefile
index fbdc634..2f438d0 100644
--- a/kernel/power/Makefile
+++ b/kernel/power/Makefile
@@ -3,9 +3,9 @@
 EXTRA_CFLAGS	+=	-DDEBUG
 endif
 
-swsusp-smp-$(CONFIG_SMP)	+= smp.o
-
 obj-y				:= main.o process.o console.o pm.o
-obj-$(CONFIG_SOFTWARE_SUSPEND)	+= swsusp.o $(swsusp-smp-y) disk.o
+obj-$(CONFIG_SOFTWARE_SUSPEND)	+= swsusp.o disk.o
+
+obj-$(CONFIG_SUSPEND_SMP)	+= smp.o
 
 obj-$(CONFIG_MAGIC_SYSRQ)	+= poweroff.o
diff --git a/kernel/power/disk.c b/kernel/power/disk.c
index 02b6764..fb8de63 100644
--- a/kernel/power/disk.c
+++ b/kernel/power/disk.c
@@ -117,8 +117,8 @@
 {
 	device_resume();
 	platform_finish();
-	enable_nonboot_cpus();
 	thaw_processes();
+	enable_nonboot_cpus();
 	pm_restore_console();
 }
 
@@ -131,28 +131,35 @@
 
 	sys_sync();
 
+	disable_nonboot_cpus();
+
 	if (freeze_processes()) {
 		error = -EBUSY;
-		return error;
+		goto thaw;
 	}
 
 	if (pm_disk_mode == PM_DISK_PLATFORM) {
 		if (pm_ops && pm_ops->prepare) {
 			if ((error = pm_ops->prepare(PM_SUSPEND_DISK)))
-				return error;
+				goto thaw;
 		}
 	}
 
 	/* Free memory before shutting down devices. */
 	free_some_memory();
-
 	return 0;
+thaw:
+	thaw_processes();
+	enable_nonboot_cpus();
+	pm_restore_console();
+	return error;
 }
 
 static void unprepare_processes(void)
 {
-	enable_nonboot_cpus();
+	platform_finish();
 	thaw_processes();
+	enable_nonboot_cpus();
 	pm_restore_console();
 }
 
@@ -160,15 +167,9 @@
 {
 	int error;
 
-	disable_nonboot_cpus();
-	if ((error = device_suspend(PMSG_FREEZE))) {
+	if ((error = device_suspend(PMSG_FREEZE)))
 		printk("Some devices failed to suspend\n");
-		platform_finish();
-		enable_nonboot_cpus();
-		return error;
-	}
-
-	return 0;
+	return error;
 }
 
 /**
@@ -185,9 +186,9 @@
 	int error;
 
 	error = prepare_processes();
-	if (!error) {
-		error = prepare_devices();
-	}
+	if (error)
+		return error;
+	error = prepare_devices();
 
 	if (error) {
 		unprepare_processes();
@@ -250,7 +251,7 @@
 
 	if ((error = prepare_processes())) {
 		swsusp_close();
-		goto Cleanup;
+		goto Done;
 	}
 
 	pr_debug("PM: Reading swsusp image.\n");
diff --git a/kernel/power/main.c b/kernel/power/main.c
index 4cdebc9..c94cb9e 100644
--- a/kernel/power/main.c
+++ b/kernel/power/main.c
@@ -55,6 +55,13 @@
 
 	pm_prepare_console();
 
+	disable_nonboot_cpus();
+
+	if (num_online_cpus() != 1) {
+		error = -EPERM;
+		goto Enable_cpu;
+	}
+
 	if (freeze_processes()) {
 		error = -EAGAIN;
 		goto Thaw;
@@ -75,6 +82,8 @@
 		pm_ops->finish(state);
  Thaw:
 	thaw_processes();
+ Enable_cpu:
+	enable_nonboot_cpus();
 	pm_restore_console();
 	return error;
 }
@@ -113,6 +122,7 @@
 	if (pm_ops && pm_ops->finish)
 		pm_ops->finish(state);
 	thaw_processes();
+	enable_nonboot_cpus();
 	pm_restore_console();
 }
 
@@ -150,12 +160,6 @@
 		goto Unlock;
 	}
 
-	/* Suspend is hard to get right on SMP. */
-	if (num_online_cpus() != 1) {
-		error = -EPERM;
-		goto Unlock;
-	}
-
 	pr_debug("PM: Preparing system for %s sleep\n", pm_states[state]);
 	if ((error = suspend_prepare(state)))
 		goto Unlock;
diff --git a/kernel/power/smp.c b/kernel/power/smp.c
index 457c230..bbe2307 100644
--- a/kernel/power/smp.c
+++ b/kernel/power/smp.c
@@ -13,73 +13,52 @@
 #include <linux/interrupt.h>
 #include <linux/suspend.h>
 #include <linux/module.h>
+#include <linux/cpu.h>
 #include <asm/atomic.h>
 #include <asm/tlbflush.h>
 
-static atomic_t cpu_counter, freeze;
-
-
-static void smp_pause(void * data)
-{
-	struct saved_context ctxt;
-	__save_processor_state(&ctxt);
-	printk("Sleeping in:\n");
-	dump_stack();
-	atomic_inc(&cpu_counter);
-	while (atomic_read(&freeze)) {
-		/* FIXME: restore takes place at random piece inside this.
-		   This should probably be written in assembly, and
-		   preserve general-purpose registers, too
-
-		   What about stack? We may need to move to new stack here.
-
-		   This should better be ran with interrupts disabled.
-		 */
-		cpu_relax();
-		barrier();
-	}
-	atomic_dec(&cpu_counter);
-	__restore_processor_state(&ctxt);
-}
-
-static cpumask_t oldmask;
+/* This is protected by pm_sem semaphore */
+static cpumask_t frozen_cpus;
 
 void disable_nonboot_cpus(void)
 {
-	oldmask = current->cpus_allowed;
-	set_cpus_allowed(current, cpumask_of_cpu(0));
-	printk("Freezing CPUs (at %d)", raw_smp_processor_id());
-	current->state = TASK_INTERRUPTIBLE;
-	schedule_timeout(HZ);
-	printk("...");
-	BUG_ON(raw_smp_processor_id() != 0);
+	int cpu, error;
 
-	/* FIXME: for this to work, all the CPUs must be running
-	 * "idle" thread (or we deadlock). Is that guaranteed? */
-
-	atomic_set(&cpu_counter, 0);
-	atomic_set(&freeze, 1);
-	smp_call_function(smp_pause, NULL, 0, 0);
-	while (atomic_read(&cpu_counter) < (num_online_cpus() - 1)) {
-		cpu_relax();
-		barrier();
+	error = 0;
+	cpus_clear(frozen_cpus);
+	printk("Freezing cpus ...\n");
+	for_each_online_cpu(cpu) {
+		if (cpu == 0)
+			continue;
+		error = cpu_down(cpu);
+		if (!error) {
+			cpu_set(cpu, frozen_cpus);
+			printk("CPU%d is down\n", cpu);
+			continue;
+		}
+		printk("Error taking cpu %d down: %d\n", cpu, error);
 	}
-	printk("ok\n");
+	BUG_ON(smp_processor_id() != 0);
+	if (error)
+		panic("cpus not sleeping");
 }
 
 void enable_nonboot_cpus(void)
 {
-	printk("Restarting CPUs");
-	atomic_set(&freeze, 0);
-	while (atomic_read(&cpu_counter)) {
-		cpu_relax();
-		barrier();
+	int cpu, error;
+
+	printk("Thawing cpus ...\n");
+	for_each_cpu_mask(cpu, frozen_cpus) {
+		error = smp_prepare_cpu(cpu);
+		if (!error)
+			error = cpu_up(cpu);
+		if (!error) {
+			printk("CPU%d is up\n", cpu);
+			continue;
+		}
+		printk("Error taking cpu %d up: %d\n", cpu, error);
+		panic("Not enough cpus");
 	}
-	printk("...");
-	set_cpus_allowed(current, oldmask);
-	schedule();
-	printk("ok\n");
-
+	cpus_clear(frozen_cpus);
 }
 
-
diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c
index 53f9f87..339b5c3 100644
--- a/kernel/power/swsusp.c
+++ b/kernel/power/swsusp.c
@@ -1193,8 +1193,10 @@
 		return "version";
 	if (strcmp(swsusp_info.uts.machine,system_utsname.machine))
 		return "machine";
+#if 0
 	if(swsusp_info.cpus != num_online_cpus())
 		return "number of cpus";
+#endif
 	return NULL;
 }