Re: FYI: My current suspend bigdiff

Re: FYI: My current suspend bigdiff

[Pavel Machek]

Hi!

> >Well... it could work with scsi. You can try it, but be carefull. [If
> >it goes wrong it might eat your data.]
> 
> Thats why I use xfs on my main system to test... And I tried with libata 
> and it won't work as it complains that the "katad" process cannot be 
> stopped, so swsusp immediatly comes back.

I do not know how much more support is needed to allow powermanagment
for libata, but this one should be easy...

[Hmm, I hope it compiles, I certainly do not use libata for now.]
								Pavel

--- clean/drivers/scsi/libata-core.c	2003-11-28 17:06:39.000000000 +0100
+++ linux/drivers/scsi/libata-core.c	2003-11-30 18:16:02.000000000 +0100
@@ -34,6 +34,7 @@
 #include <linux/delay.h>
 #include <linux/timer.h>
 #include <linux/interrupt.h>
+#include <linux/suspend.h>
 #include <scsi/scsi.h>
 #include "scsi.h"
 #include "hosts.h"
@@ -2564,6 +2565,8 @@
 
         while (1) {
 		cond_resched();
+		if (current->flags & PF_FREEZE)
+			refrigerator(PF_IOTHREAD);
 
 		timeout = ata_thread_iter(ap);
 


-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: FYI: My current suspend bigdiff

[Jens Axboe]

On Sun, Nov 30 2003, Pavel Machek wrote:
> Hi!
> 
> > >Well... it could work with scsi. You can try it, but be carefull. [If
> > >it goes wrong it might eat your data.]
> > 
> > Thats why I use xfs on my main system to test... And I tried with libata 
> > and it won't work as it complains that the "katad" process cannot be 
> > stopped, so swsusp immediatly comes back.
> 
> I do not know how much more support is needed to allow powermanagment
> for libata, but this one should be easy...
> 
> [Hmm, I hope it compiles, I certainly do not use libata for now.]
> 								Pavel
> 
> --- clean/drivers/scsi/libata-core.c	2003-11-28 17:06:39.000000000 +0100
> +++ linux/drivers/scsi/libata-core.c	2003-11-30 18:16:02.000000000 +0100
> @@ -34,6 +34,7 @@
>  #include <linux/delay.h>
>  #include <linux/timer.h>
>  #include <linux/interrupt.h>
> +#include <linux/suspend.h>
>  #include <scsi/scsi.h>
>  #include "scsi.h"
>  #include "hosts.h"
> @@ -2564,6 +2565,8 @@
>  
>          while (1) {
>  		cond_resched();
> +		if (current->flags & PF_FREEZE)
> +			refrigerator(PF_IOTHREAD);

Curious - can't this be put some place a bit smarter, so we don't have
to update every single "driver" with a kernel thread out there (like in
the scheduler)? Seems pretty fragile to rely on this. Plus, when you
change this in the future there'll be N drivers to update again.

-- 
Jens Axboe

Re: FYI: My current suspend bigdiff

[Prakash K. Cheemplavam]

Pavel Machek wrote:
> Hi!
> 
> 
>>>Well... it could work with scsi. You can try it, but be carefull. [If
>>>it goes wrong it might eat your data.]
>>
>>Thats why I use xfs on my main system to test... And I tried with libata 
>>and it won't work as it complains that the "katad" process cannot be 
>>stopped, so swsusp immediatly comes back.
> 
> 
> I do not know how much more support is needed to allow powermanagment
> for libata, but this one should be easy...

Uhm, Jeff already fixed it in libata using the same call. Can both fixes 
"hurt" each other or are the safe?


> +		if (current->flags & PF_FREEZE)
> +			refrigerator(PF_IOTHREAD);

Re: FYI: My current suspend bigdiff

[Jeff Garzik]

Jens Axboe wrote:
> Curious - can't this be put some place a bit smarter, so we don't have
> to update every single "driver" with a kernel thread out there (like in
> the scheduler)? Seems pretty fragile to rely on this. Plus, when you
> change this in the future there'll be N drivers to update again.


Agreed.  I had to do this simple and obvious (and template-able) patch 
for 8139too and another driver whose name I forget ATM...

	Jeff

Re: FYI: My current suspend bigdiff

[Jeff Garzik]

Prakash K. Cheemplavam wrote:
> Pavel Machek wrote:
> 
>> Hi!
>>
>>
>>>> Well... it could work with scsi. You can try it, but be carefull. [If
>>>> it goes wrong it might eat your data.]
>>>
>>>
>>> Thats why I use xfs on my main system to test... And I tried with 
>>> libata and it won't work as it complains that the "katad" process 
>>> cannot be stopped, so swsusp immediatly comes back.
>>
>>
>>
>> I do not know how much more support is needed to allow powermanagment
>> for libata, but this one should be easy...
> 
> 
> Uhm, Jeff already fixed it in libata using the same call. Can both fixes 
> "hurt" each other or are the safe?


As you suspect, you want only one of the fixes.

I would probably prefer Pavel's patch over mine, as he knows the suspend 
subsystem better than me :)

	Jeff

Re: FYI: My current suspend bigdiff

[Pavel Machek]

Hi!

> >>>>Well... it could work with scsi. You can try it, but be carefull. [If
> >>>>it goes wrong it might eat your data.]
> >>>
> >>>
> >>>Thats why I use xfs on my main system to test... And I tried with 
> >>>libata and it won't work as it complains that the "katad" process 
> >>>cannot be stopped, so swsusp immediatly comes back.
> >>
> >>
> >>
> >>I do not know how much more support is needed to allow powermanagment
> >>for libata, but this one should be easy...
> >
> >
> >Uhm, Jeff already fixed it in libata using the same call. Can both fixes 
> >"hurt" each other or are the safe?
> 
> 
> As you suspect, you want only one of the fixes.
> 
> I would probably prefer Pavel's patch over mine, as he knows the suspend 
> subsystem better than me :)

Well, as I did not even try to compile it, take Jeff's patch.
								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: FYI: My current suspend bigdiff

[Benjamin Herrenschmidt]

> As you suspect, you want only one of the fixes.
> 
> I would probably prefer Pavel's patch over mine, as he knows the suspend 
> subsystem better than me :)

Actually, you want more than that...

You want something similar to what I did for IDE, that is pipe the
suspend/resume requests down the request queue & block the queue,
though with additional crap to deal with the libata probe thread
and the scsi error mgmnt.

I didn't yet have time for it, but basically, what is needed is
to make a scsi version of the suspend/resume code that is in the
IDE code, and then hook that on sd, sg, etc...

Then, libata would have to get in the loop as HW driver, and also
by properly translating the flush cache / standby requests

Ben.

FYI: My current suspend bigdiff

[Pavel Machek]

Hi!

This is how my current "bigdiff" looks... [This is not for
application, but if you want to make S3 working, it might help you].


						Pavel

%patch
Index: linux/drivers/input/serio/i8042.c
===================================================================
--- linux.orig/drivers/input/serio/i8042.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/serio/i8042.c	2003-11-22 16:56:05.000000000 +0100
@@ -18,6 +18,7 @@
 #include <linux/reboot.h>
 #include <linux/init.h>
 #include <linux/serio.h>
+#include <linux/sysdev.h>
 
 #include <asm/io.h>
 
@@ -398,18 +399,15 @@
  * desired.
  */
 	
-static int __init i8042_controller_init(void)
+static int i8042_controller_init(void)
 {
-
 /*
  * Test the i8042. We need to know if it thinks it's working correctly
  * before doing anything else.
  */
 
 	i8042_flush();
-
 	if (i8042_reset) {
-
 		unsigned char param;
 
 		if (i8042_command(&param, I8042_CMD_CTL_TEST)) {
@@ -783,6 +781,32 @@
 	values->mux = index;
 }
 
+static int i8042_resume_port(struct serio *port)
+{
+	struct serio_dev *dev = port->dev;
+	if (dev && dev->resume)
+		dev->resume(port);
+}
+
+static int i8042_resume(struct sys_device *dev)
+{
+	if (i8042_controller_init())
+		printk(KERN_ERR "i8042: resume failed\n");
+	i8042_resume_port(&i8042_aux_port);
+	i8042_resume_port(&i8042_kbd_port);
+	return 0;
+}
+
+static struct sysdev_class kbc_sysclass = {
+	set_kset_name("i8042"),
+	.resume = i8042_resume,
+};
+
+static struct sys_device device_i8042 = {
+	.id	= 0,
+	.cls	= &kbc_sysclass,
+};
+
 int __init i8042_init(void)
 {
 	int i;
@@ -819,6 +843,14 @@
 
 	register_reboot_notifier(&i8042_notifier);
 
+	{
+		int error = sysdev_class_register(&kbc_sysclass);
+		if (!error)
+			error = sys_device_register(&device_i8042);
+		if (error)
+			printk(KERN_CRIT "Unable to register i8042 to driver model\n");
+	}
+
 	return 0;
 }
 
Index: linux/drivers/input/keyboard/atkbd.c
===================================================================
--- linux.orig/drivers/input/keyboard/atkbd.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/keyboard/atkbd.c	2003-11-24 10:31:28.000000000 +0100
@@ -145,7 +145,7 @@
 	unsigned char set;
 	unsigned char release;
 	int lastkey;
-	volatile signed char ack;
+	atomic_t ack;
 	unsigned char emul;
 	unsigned short id;
 	unsigned char write;
@@ -214,10 +214,10 @@
 
 	switch (code) {
 		case ATKBD_RET_ACK:
-			atkbd->ack = 1;
+			atomic_set(&atkbd->ack, 1);
 			goto out;
 		case ATKBD_RET_NAK:
-			atkbd->ack = -1;
+			atomic_set(&atkbd->ack, 2);
 			goto out;
 	}
 
@@ -294,7 +294,7 @@
 static int atkbd_sendbyte(struct atkbd *atkbd, unsigned char byte)
 {
 	int timeout = 20000; /* 200 msec */
-	atkbd->ack = 0;
+	atomic_set(&atkbd->ack, 0);
 
 #ifdef ATKBD_DEBUG
 	printk(KERN_DEBUG "atkbd.c: Sent: %02x\n", byte);
@@ -302,9 +302,9 @@
 	if (serio_write(atkbd->serio, byte))
 		return -1;
 
-	while (!atkbd->ack && timeout--) udelay(10);
+	while (!atomic_read(&atkbd->ack) && timeout--) udelay(10);
 
-	return -(atkbd->ack <= 0);
+	return -(atomic_read(&atkbd->ack) != 1);
 }
 
 /*
@@ -582,6 +582,37 @@
 	kfree(atkbd);
 }
 
+int atkbd_hwinit(struct atkbd *atkbd)
+{
+	if (atkbd->write) {
+		if (atkbd_probe(atkbd)) {
+			return -EIO;
+		}
+		
+		atkbd->set = atkbd_set_3(atkbd);
+		atkbd_enable(atkbd);
+
+	} else {
+		atkbd->set = 2;
+		atkbd->id = 0xab00;
+	}
+
+	if (atkbd->set == 4) {
+		atkbd->dev.ledbit[0] |= BIT(LED_COMPOSE) | BIT(LED_SUSPEND) | BIT(LED_SLEEP) | BIT(LED_MUTE) | BIT(LED_MISC);
+		sprintf(atkbd->name, "AT Set 2 Extended keyboard");
+	} else
+		sprintf(atkbd->name, "AT %s Set %d keyboard",
+			atkbd->translated ? "Translated" : "Raw", atkbd->set);
+	return 0;
+}
+
+static void atkbd_resume(struct serio *serio)
+{
+	struct atkbd *atkbd = serio->private;
+	if (atkbd_hwinit(atkbd))
+		printk(KERN_ERR "Ouch, keyboard no longer there after resume?\n");
+}
+
 /*
  * atkbd_connect() is called when the serio module finds and interface
  * that isn't handled yet by an appropriate device driver. We check if
@@ -641,29 +672,11 @@
 		return;
 	}
 
-	if (atkbd->write) {
-
-		if (atkbd_probe(atkbd)) {
-			serio_close(serio);
-			kfree(atkbd);
-			return;
-		}
-		
-		atkbd->set = atkbd_set_3(atkbd);
-		atkbd_enable(atkbd);
-
-	} else {
-		atkbd->set = 2;
-		atkbd->id = 0xab00;
+	if (atkbd_hwinit(atkbd)) {
+		serio_close(serio);
+		kfree(atkbd);
+		return;
 	}
-
-	if (atkbd->set == 4) {
-		atkbd->dev.ledbit[0] |= BIT(LED_COMPOSE) | BIT(LED_SUSPEND) | BIT(LED_SLEEP) | BIT(LED_MUTE) | BIT(LED_MISC);
-		sprintf(atkbd->name, "AT Set 2 Extended keyboard");
-	} else
-		sprintf(atkbd->name, "AT %s Set %d keyboard",
-			atkbd->translated ? "Translated" : "Raw", atkbd->set);
-
 	sprintf(atkbd->phys, "%s/input0", serio->phys);
 
 	if (atkbd->set == 3)
@@ -688,10 +701,11 @@
 }
 
 
-static struct serio_dev atkbd_dev = {
+struct serio_dev atkbd_dev = {
 	.interrupt =	atkbd_interrupt,
 	.connect =	atkbd_connect,
 	.disconnect =	atkbd_disconnect,
+	.resume =	atkbd_resume,
 	.cleanup =	atkbd_cleanup,
 };
 
Index: linux/include/linux/serio.h
===================================================================
--- linux.orig/include/linux/serio.h	2003-11-24 11:30:30.000000000 +0100
+++ linux/include/linux/serio.h	2003-11-22 16:54:24.000000000 +0100
@@ -50,6 +50,7 @@
 			unsigned int, struct pt_regs *);
 	void (*connect)(struct serio *, struct serio_dev *dev);
 	void (*disconnect)(struct serio *);
+	void (*resume)(struct serio *);
 	void (*cleanup)(struct serio *);
 
 	struct list_head node;
Index: linux/drivers/input/power.c
===================================================================
--- linux.orig/drivers/input/power.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/power.c	2003-07-23 00:18:13.000000000 +0200
@@ -72,6 +72,18 @@
 				break;
 			case KEY_POWER:
 				/* Hum power down the machine. */
+				{
+					char *argv[2] = {NULL, NULL};
+					char *envp[3] = {NULL, NULL, NULL};
+
+					argv[0] = "/sbin/poweroff";
+
+					/* minimal command environment */
+					envp[0] = "HOME=/";
+					envp[1] = "PATH=/sbin:/bin:/usr/sbin:/usr/bin";
+	
+					call_usermodehelper(argv[0], argv, envp, 0);
+				}
 				break;
 			default:	
 				return;
Index: linux/drivers/acpi/thermal.c
===================================================================
--- linux.orig/drivers/acpi/thermal.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/acpi/thermal.c	2003-11-24 10:45:49.000000000 +0100
@@ -467,6 +467,7 @@
 	if (result)
 		return_VALUE(result);
 
+	printk(KERN_EMERG "Critical temperature reached (%d C), shutting down.\n", tz->temperature);
 	acpi_bus_generate_event(device, ACPI_THERMAL_NOTIFY_CRITICAL, tz->trips.critical.flags.enabled);
 
 	acpi_thermal_call_usermode(ACPI_THERMAL_PATH_POWEROFF);
Index: linux/arch/i386/kernel/cpu/mtrr/main.c
===================================================================
--- linux.orig/arch/i386/kernel/cpu/mtrr/main.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/arch/i386/kernel/cpu/mtrr/main.c	2003-10-23 00:42:26.000000000 +0200
@@ -588,6 +588,7 @@
 {
 	int i;
 
+#if 0
 	for (i = 0; i < num_var_ranges; i++) {
 		if (mtrr_state[i].lsize) 
 			set_mtrr(i,
@@ -596,6 +597,8 @@
 				 mtrr_state[i].ltype);
 	}
 	kfree(mtrr_state);
+#endif
+
 	return 0;
 }
 
Index: linux/arch/x86_64/kernel/time.c
===================================================================
--- linux.orig/arch/x86_64/kernel/time.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/arch/x86_64/kernel/time.c	2003-10-09 00:19:18.000000000 +0200
@@ -22,7 +22,7 @@
 #include <linux/time.h>
 #include <linux/ioport.h>
 #include <linux/module.h>
-#include <linux/device.h>
+#include <linux/sysdev.h>
 #include <linux/bcd.h>
 #include <asm/pgtable.h>
 #include <asm/vsyscall.h>
@@ -75,7 +75,7 @@
  * timer interrupt has happened already, but vxtime.trigger wasn't updated yet.
  * This is not a problem, because jiffies hasn't updated either. They are bound
  * together by xtime_lock.
-         */
+ */
 
 static inline unsigned int do_gettimeoffset_tsc(void)
 {
@@ -626,7 +626,17 @@
 	return 0;
 }
 
-void __init pit_init(void)
+static struct sysdev_class rtc_sysclass = {
+	set_kset_name("rtc"),
+};
+
+/* XXX this driverfs stuff should probably go elsewhere later -john */
+static struct sys_device device_i8253 = {
+	.id		= 0,
+	.cls	= &rtc_sysclass,
+};
+
+int __init pit_init(void)
 {
 	unsigned long flags;
 
@@ -635,6 +645,11 @@
 	outb_p(LATCH & 0xff, 0x40);	/* LSB */
 	outb_p(LATCH >> 8, 0x40);	/* MSB */
 	spin_unlock_irqrestore(&i8253_lock, flags);
+
+	int error = sysdev_class_register(&rtc_sysclass);
+	if (!error)
+		error = sys_device_register(&device_i8253);
+	return error;
 }
 
 int __init time_setup(char *str)
@@ -678,8 +693,8 @@
 		cpu_khz = hpet_calibrate_tsc();
 		timename = "HPET";
 	} else {
-	pit_init();
-	cpu_khz = pit_calibrate_tsc();
+		pit_init();
+		cpu_khz = pit_calibrate_tsc();
 		timename = "PIT";
 	}
 
Index: linux/include/linux/suspend.h
===================================================================
--- linux.orig/include/linux/suspend.h	2003-11-24 11:30:30.000000000 +0100
+++ linux/include/linux/suspend.h	2003-10-18 23:06:22.000000000 +0200
@@ -45,31 +45,43 @@
 /* mm/page_alloc.c */
 extern void drain_local_pages(void);
 
+/* kernel/power/swsusp.c */
+extern int software_suspend(void);
+
 extern unsigned int nr_copy_pages __nosavedata;
 extern suspend_pagedir_t *pagedir_nosave __nosavedata;
-#endif /* CONFIG_PM */
-
-#ifdef CONFIG_SOFTWARE_SUSPEND
-
-extern unsigned char software_suspend_enabled;
 
-extern void software_suspend(void);
 #else	/* CONFIG_SOFTWARE_SUSPEND */
-static inline void software_suspend(void)
+static inline int software_suspend(void)
 {
 	printk("Warning: fake suspend called\n");
+	return -EPERM;
 }
+#define software_resume()		do { } while(0)
 #endif	/* CONFIG_SOFTWARE_SUSPEND */
 
 
 #ifdef CONFIG_PM
 extern void refrigerator(unsigned long);
+extern int freeze_processes(void);
+extern void thaw_processes(void);
+
+extern int pm_prepare_console(void);
+extern void pm_restore_console(void);
 
 #else
 static inline void refrigerator(unsigned long flag)
 {
 
 }
+static inline int freeze_processes(void)
+{
+	return 0;
+}
+static inline void thaw_processes(void)
+{
+
+}
 #endif	/* CONFIG_PM */
 
 #endif /* _LINUX_SWSUSP_H */
Index: linux/kernel/power/swsusp.c
===================================================================
--- linux.orig/kernel/power/swsusp.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/kernel/power/swsusp.c	2003-10-23 00:59:17.000000000 +0200
@@ -73,9 +73,8 @@
 
 extern void do_magic(int resume);
 
-#define NORESUME		1
-#define RESUME_SPECIFIED	2
-
+#define NORESUME               1
+#define RESUME_SPECIFIED       2
 
 #define __ADDRESS(x)  ((unsigned long) phys_to_virt(x))
 #define ADDRESS(x) __ADDRESS((x) << PAGE_SHIFT)
@@ -227,6 +226,7 @@
 static void read_swapfiles(void) /* This is called before saving image */
 {
 	int i, len;
+	char buff[sizeof(resume_file)], *sname;
 	
 	len=strlen(resume_file);
 	root_swap = 0xFFFF;
@@ -245,8 +245,11 @@
 					swapfile_used[i] = SWAPFILE_IGNORED;				  
 			} else {
 	  			/* we ignore all swap devices that are not the resume_file */
-				if (1) {
-// FIXME				if(resume_device == swap_info[i].swap_device) {
+				sname = d_path(swap_info[i].swap_file->f_dentry,
+					       swap_info[i].swap_file->f_vfsmnt,
+					       buff,
+					       sizeof(buff));
+				if (!strcmp(sname, resume_file)) {
 					swapfile_used[i] = SWAPFILE_SUSPEND;
 					root_swap = i;
 				} else {
@@ -283,8 +286,8 @@
  *    would happen on next reboot -- corrupting data.
  *
  *    Note: The buffer we allocate to use to write the suspend header is
- *    not freed; its not needed since system is going down anyway
- *    (plus it causes oops and I'm lazy^H^H^H^Htoo busy).
+ *    not freed; its not needed since the system is going down anyway
+ *    (plus it causes an oops and I'm lazy^H^H^H^Htoo busy).
  */
 static int write_suspend_image(void)
 {
@@ -488,33 +491,6 @@
 	printk("|\n");
 }
 
-/* Make disk drivers accept operations, again */
-static void drivers_unsuspend(void)
-{
-	device_resume();
-}
-
-/* Called from process context */
-static int drivers_suspend(void)
-{
-	return device_suspend(4);
-}
-
-#define RESUME_PHASE1 1 /* Called from interrupts disabled */
-#define RESUME_PHASE2 2 /* Called with interrupts enabled */
-#define RESUME_ALL_PHASES (RESUME_PHASE1 | RESUME_PHASE2)
-static void drivers_resume(int flags)
-{
-	if (flags & RESUME_PHASE1) {
-		device_resume();
-	}
-  	if (flags & RESUME_PHASE2) {
-#ifdef SUSPEND_CONSOLE
-		update_screen(fg_console);	/* Hmm, is this the problem? */
-#endif
-	}
-}
-
 static int suspend_prepare_image(void)
 {
 	struct sysinfo i;
@@ -569,7 +545,7 @@
 
 static void suspend_save_image(void)
 {
-	drivers_unsuspend();
+	device_resume();
 
 	lock_swapdevices();
 	write_suspend_image();
@@ -615,6 +591,7 @@
 	mb();
 	spin_lock_irq(&suspend_pagedir_lock);	/* Done to disable interrupts */ 
 
+	device_power_down(4);
 	PRINTK( "Waiting for DMAs to settle down...\n");
 	mdelay(1000);	/* We do not want some readahead with DMA to corrupt our memory, right?
 			   Do it with disabled interrupts for best effect. That way, if some
@@ -630,8 +607,10 @@
 
 	PRINTK( "Freeing prev allocated pagedir\n" );
 	free_suspend_pagedir((unsigned long) pagedir_save);
+	device_power_up();
 	spin_unlock_irq(&suspend_pagedir_lock);
-	drivers_resume(RESUME_ALL_PHASES);
+	device_resume();
+	update_screen(fg_console);	/* Hmm, is this the problem? */
 
 	PRINTK( "Fixing swap signatures... " );
 	mark_swapfiles(((swp_entry_t) {0}), MARK_SWAP_RESUME);
@@ -672,7 +651,9 @@
 {
 	int is_problem;
 	read_swapfiles();
+	device_power_down(4);
 	is_problem = suspend_prepare_image();
+	device_power_up();
 	spin_unlock_irq(&suspend_pagedir_lock);
 	if (!is_problem) {
 		kernel_fpu_end();	/* save_processor_state() does kernel_fpu_begin, and we need to revert it in order to pass in_atomic() checks */
@@ -694,11 +675,22 @@
 	mark_swapfiles(((swp_entry_t) {0}), MARK_SWAP_RESUME);
 }
 
-static void do_software_suspend(void)
+/*
+ * This is main interface to the outside world. It needs to be
+ * called from process context.
+ */
+int software_suspend(void)
 {
+	int res;
+	if (!software_suspend_enabled)
+		return -EAGAIN;
+
+	software_suspend_enabled = 0;
+	might_sleep();
+
 	if (arch_prepare_suspend()) {
 		printk("%sArchitecture failed to prepare\n", name_suspend);
-		return;
+		return -EPERM;
 	}		
 	if (pm_prepare_console())
 		printk( "%sCan't allocate a console... proceeding\n", name_suspend);
@@ -716,7 +708,7 @@
 		blk_run_queues();
 
 		/* Save state of all device drivers, and stop them. */		   
-		if(drivers_suspend()==0)
+		if ((res = device_suspend(4))==0)
 			/* If stopping device drivers worked, we proceed basically into
 			 * suspend_save_image.
 			 *
@@ -728,24 +720,12 @@
 			 */
 			do_magic(0);
 		thaw_processes();
-	}
+	} else
+		res = -EBUSY;
 	software_suspend_enabled = 1;
 	MDELAY(1000);
 	pm_restore_console();
-}
-
-/*
- * This is main interface to the outside world. It needs to be
- * called from process context.
- */
-void software_suspend(void)
-{
-	if(!software_suspend_enabled)
-		return;
-
-	software_suspend_enabled = 0;
-	might_sleep();
-	do_software_suspend();
+	return res;
 }
 
 /* More restore stuff */
@@ -915,7 +895,7 @@
 
 extern dev_t __init name_to_dev_t(const char *line);
 
-static int __read_suspend_image(struct block_device *bdev, union diskpage *cur, int noresume)
+static int __init __read_suspend_image(struct block_device *bdev, union diskpage *cur, int noresume)
 {
 	swp_entry_t next;
 	int i, nr_pgdir_pages;
@@ -1091,6 +1071,7 @@
 	printk( "resuming from %s\n", resume_file);
 	if (read_suspend_image(resume_file, 0))
 		goto read_failure;
+	device_suspend(4);
 	do_magic(1);
 	panic("This never returns");
 
Index: linux/kernel/sys.c
===================================================================
--- linux.orig/kernel/sys.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/kernel/sys.c	2003-10-26 13:09:00.000000000 +0100
@@ -472,13 +472,11 @@
 
 #ifdef CONFIG_SOFTWARE_SUSPEND
 	case LINUX_REBOOT_CMD_SW_SUSPEND:
-		if (!software_suspend_enabled) {
+		{
+			int ret = software_suspend();
 			unlock_kernel();
-			return -EAGAIN;
+			return ret;
 		}
-		software_suspend();
-		do_exit(0);
-		break;
 #endif
 
 	default:
Index: linux/Documentation/power/swsusp.txt
===================================================================
--- linux.orig/Documentation/power/swsusp.txt	2003-11-24 11:30:30.000000000 +0100
+++ linux/Documentation/power/swsusp.txt	2003-10-26 12:49:22.000000000 +0100
@@ -17,13 +17,30 @@
 You need to append resume=/dev/your_swap_partition to kernel command
 line. Then you suspend by echo 4 > /proc/acpi/sleep.
 
-[Notice. Rest docs is pretty outdated (see date!) It should be safe to
-use swsusp on ext3/reiserfs these days.]
+Pavel's unreliable guide to swsusp mess
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 
+
+There are currently two versions of swap suspend in the kernel, the old
+"Pavel's" version in kernel/power/swsusp.c and the new "Patrick's"
+version in kernel/power/pmdisk.c. They provide the same functionality;
+the old version looks ugly but was tested, while the new version looks
+nicer but did not receive so much testing. echo 4 > /proc/acpi/sleep
+calls the old version, echo disk > /sys/power/state calls the new one.
+
+[In the future, when the new version is stable enough, two things can
+happen:
+
+* the new version is moved into swsusp.c, and swsusp is renamed to swap
+  suspend (Pavel prefers this)
+
+* pmdisk is kept as is and swsusp.c is removed from the kernel]
+
 
 
 Article about goals and implementation of Software Suspend for Linux
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Author: G‚ábor Kuti
-Last revised: 2002-04-08
+Last revised: 2003-10-20 by Pavel Machek
 
 Idea and goals to achieve
 
@@ -36,84 +53,23 @@
 interrupt our programs so processes that are calculating something for a long
 time shouldn't need to be written interruptible.
 
-On desk machines the power saving function isn't as important as it is in
-laptops but we really may benefit from the second one. Nowadays the number of
-desk machines supporting suspend function in their APM is going up but there
-are (and there will still be for a long time) machines that don't even support
-APM of any kind. On the other hand it is reported that using APM's suspend
-some irqs (e.g. ATA disk irq) is lost and it is annoying for the user until
-the Linux kernel resets the device.
-
-So I started thinking about implementing Software Suspend which doesn't need
-any APM support and - since it uses pretty near only high-level routines - is
-supposed to be architecture independent code.
-
 Using the code
 
-The code is experimental right now - testers, extra eyes are welcome. To
-compile this support into the kernel, you need CONFIG_EXPERIMENTAL, 
-and then CONFIG_SOFTWARE_SUSPEND in menu General Setup to be  enabled. It
-cannot be used as a module and I don't think it will ever be needed.
-
-You have two ways to use this code. The first one is if you've compiled in
-sysrq support then you may press Sysrq-D to request suspend. The other way
-is with a patched SysVinit (my patch is against 2.76 and available at my
-home page). You might call 'swsusp' or 'shutdown -z <time>'. Next way is to
-echo 4 > /proc/acpi/sleep.
+You have two ways to use this code. The first one is is with a patched
+SysVinit (my patch is against 2.76 and available at my home page). You
+might call 'swsusp' or 'shutdown -z <time>'. Next way is to echo 4 >
+/proc/acpi/sleep.
 
 Either way it saves the state of the machine into active swaps and then
 reboots.  You must explicitly specify the swap partition to resume from with
 ``resume='' kernel option. If signature is found it loads and restores saved
 state. If the option ``noresume'' is specified as a boot parameter, it skips
-the resuming.  Warning! Look at section ``Things to implement'' to see what
-isn't yet implemented.  Also I strongly suggest you to list all active swaps
-in /etc/fstab. Firstly because you don't have to specify anything to resume
-and secondly if you have more than one swap area you can't decide which one
-has the 'root' signature. 
+the resuming.
 
 In the meantime while the system is suspended you should not touch any of the
 hardware!
 
 About the code
-Goals reached
-
-The code can be downloaded from
-http://falcon.sch.bme.hu/~seasons/linux/. It mainly works but there are still
-some of XXXs, TODOs, FIXMEs in the code which seem not to be too important. It
-should work all right except for the problems listed in ``Things to
-implement''. Notes about the code are really welcome.
-
-How the code works
-
-When suspending is triggered it immediately wakes up process bdflush. Bdflush
-checks whether we have anything in our run queue tq_bdflush. Since we queued up
-function do_software_suspend, it is called. Here we shrink everything including
-dcache, inodes, buffers and memory (here mainly processes are swapped out). We
-count how many pages we need to duplicate (we have to be atomical!) then we
-create an appropriate sized page directory. It will point to the original and
-the new (copied) address of the page. We get the free pages by
-__get_free_pages() but since it changes state we have to be able to track it
-later so it also flips in a bit in page's flags (a new Nosave flag). We
-duplicate pages and then mark them as used (so atomicity is ensured). After
-this we write out the image to swaps, do another sync and the machine may
-reboot. We also save registers to stack.
-
-By resuming an ``inverse'' method is executed. The image if exists is loaded,
-loadling is either triggered by ``resume='' kernel option.  We
-change our task to bdflush (it is needed because if we don't do this init does
-an oops when it is waken up later) and then pages are copied back to their
-original location. We restore registers, free previously allocated memory,
-activate memory context and task information. Here we should restore hardware
-state but even without this the machine is restored and processes are continued
-to work. I think hardware state should be restored by some list (using
-notify_chain) and probably by some userland program (run-parts?) for users'
-pleasure. Check out my patch at the same location for the sysvinit patch.
-
-WARNINGS!
-- It does not like pcmcia cards. And this is logical: pcmcia cards need
-  cardmgr to be initialized. they are not initialized during singleuser boot,
-  but "resumed" kernel does expect them to be initialized. That leads to
-  armagedon. You should eject any pcmcia cards before suspending.
 
 Things to implement
 - SMP support. I've done an SMP support but since I don't have access to a kind
@@ -122,34 +78,14 @@
   interrupts AFAIK..
 - We should only make a copy of data related to kernel segment, since any
   process data won't be changed.
-- Hardware state restoring.  Now there's support for notifying via the notify
-  chain, event handlers are welcome. Some devices may have microcodes loaded
-  into them. We should have event handlers for them as well.
-- We should support other architectures (There are really only some arch
-  related functions..)
-- We should also restore original state of swaps if the ``noresume'' kernel
-  option is specified.. Or do we need such a feature to save state for some
-  other time? Do we need some kind of ``several saved states''?  (Linux-HA
-  people?). There's been some discussion about checkpointing on linux-future.
 - Should make more sanity checks. Or are these enough?
 
 Not so important ideas for implementing
 
 - If a real time process is running then don't suspend the machine.
-- Support for power.conf file as in Solaris, autoshutdown, special
-  devicetypes support, maybe in sysctl.
-- Introduce timeout for SMP locking. But first locking ought to work :O
-- Pre-detect if we don't have enough swap space or free it instead of
-  calling panic.
 - Support for adding/removing hardware while suspended?
 - We should not free pages at the beginning so aggressively, most of them
   go there anyway..
-- If X is active while suspending then by resuming calling svgatextmode
-  corrupts the virtual console of X.. (Maybe this has been fixed AFAIK).
-
-Drivers we support
-- IDE disks are okay
-- vesafb
 
 Drivers that need support
 - pc_keyb -- perhaps we can wait for vojtech's input patches
Index: linux/Documentation/power/video.txt
===================================================================
--- linux.orig/Documentation/power/video.txt	2003-11-24 11:30:30.000000000 +0100
+++ linux/Documentation/power/video.txt	2003-10-26 12:49:28.000000000 +0100
@@ -0,0 +1,36 @@
+
+		Video issues with S3 resume
+		~~~~~~~~~~~~~~~~~~~~~~~~~~~
+		     2003, Pavel Machek
+
+During S3 resume, hardware needs to be reinitialized. For most
+devices, this is easy, and kernel driver knows how to do
+it. Unfortunately there's one exception: video card. Those are usually
+initialized by BIOS, and kernel does not have enough information to
+boot video card. (Kernel usually does not even contain video card
+driver -- vesafb and vgacon are widely used).
+
+This is not problem for swsusp, because during swsusp resume, BIOS is
+run normally so video card is normally initialized.
+
+There are three types of systems where video works after S3 resume:
+
+* systems where video state is preserved over S3. (HP Omnibook xe3)
+
+* systems that initialize video card into vga text mode and where BIOS
+  works well enough to be able to set video mode. Use
+  acpi_sleep=s3_mode on these. (Toshiba 4030cdt)
+
+* systems where it is possible to call video bios during S3
+  resume. Unfortunately, it is not correct to call video BIOS at that
+  point, but it happens to work on some machines. Use
+  acpi_sleep=s3_bios (Athlon64 desktop system)
+
+Now, if you pass acpi_sleep=something, and it does not work with your
+bios, you'll get hard crash during resume. Be carefull.
+
+You may have system where none of above works. At that point you
+either invent another ugly hack that works, or write proper driver for
+your video card (good luck getting docs :-(). Maybe suspending from X
+(proper X, knowing your hardware, not XF68_FBcon) might have better
+chance of working.
Index: linux/arch/i386/kernel/acpi/wakeup.S
===================================================================
--- linux.orig/arch/i386/kernel/acpi/wakeup.S	2003-11-24 11:30:30.000000000 +0100
+++ linux/arch/i386/kernel/acpi/wakeup.S	2003-10-26 12:49:44.000000000 +0100
@@ -193,11 +193,6 @@
 	# and restore the stack ... but you need gdt for this to work
 	movl	saved_context_esp, %esp
 
-	movw	$0x0e00 + 'W', 0xb8018
-	outl	%eax, $0x80
-	outl	%eax, $0x80
-	movw	$0x0e00 + 'O', 0xb8018
-
 	movl	%cs:saved_magic, %eax
 	cmpl	$0x12345678, %eax
 	jne	bogus_magic
@@ -205,9 +200,6 @@
 	# jump to place where we left off
 	movl	saved_eip,%eax
 	movw	$0x0e00 + 'x', 0xb8018
-	outl	%eax, $0x80
-	outl	%eax, $0x80
-	movw	$0x0e00 + '!', 0xb801a
 	jmp	*%eax
 
 bogus_magic:
Index: linux/arch/i386/kernel/time.c
===================================================================
--- linux.orig/arch/i386/kernel/time.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/arch/i386/kernel/time.c	2003-10-26 13:08:50.000000000 +0100
@@ -280,16 +280,37 @@
 	unsigned long retval;
 
 	spin_lock(&rtc_lock);
-
 	retval = mach_get_cmos_time();
-
 	spin_unlock(&rtc_lock);
 
 	return retval;
 }
 
+static long clock_cmos_diff;
+
+static int pit_suspend(struct sys_device *dev, u32 state)
+{
+	/*
+	 * Estimate time zone so that set_time can update the clock
+	 */
+	clock_cmos_diff = -get_cmos_time();
+	clock_cmos_diff += get_seconds();
+	return 0;
+}
+
+static int pit_resume(struct sys_device *dev)
+{
+	write_seqlock_irq(&xtime_lock);
+	xtime.tv_sec = get_cmos_time() + clock_cmos_diff;
+	xtime.tv_nsec = 0; 
+	write_sequnlock_irq(&xtime_lock);
+	return 0;
+}
+
 static struct sysdev_class pit_sysclass = {
 	set_kset_name("pit"),
+	.resume = pit_resume,
+	.suspend = pit_suspend,
 };
 
 /* XXX this driverfs stuff should probably go elsewhere later -john */
Index: linux/kernel/power/process.c
===================================================================
--- linux.orig/kernel/power/process.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/kernel/power/process.c	2003-10-26 12:50:56.000000000 +0100
@@ -49,10 +49,11 @@
 	pr_debug("%s entered refrigerator\n", current->comm);
 	printk("=");
 	current->flags &= ~PF_FREEZE;
-	if (flag)
-		flush_signals(current); /* We have signaled a kernel thread, which isn't normal behaviour
-					   and that may lead to 100%CPU sucking because those threads
-					   just don't manage signals. */
+
+	spin_lock_irq(&current->sighand->siglock);
+	recalc_sigpending(); /* We sent fake signal, clean it up */
+	spin_unlock_irq(&current->sighand->siglock);
+
 	current->flags |= PF_FROZEN;
 	while (current->flags & PF_FROZEN)
 		schedule();
Index: linux/drivers/input/keyboard/98kbd.c
===================================================================
--- linux.orig/drivers/input/keyboard/98kbd.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/keyboard/98kbd.c	2003-11-24 11:29:07.000000000 +0100
@@ -100,7 +100,7 @@
 	char phys[32];
 	unsigned char cmdbuf[4];
 	unsigned char cmdcnt;
-	signed char ack;
+	atomic_t ack;
 	unsigned char shift;
 	struct {
 		unsigned char scancode;
@@ -118,10 +118,10 @@
 
 	switch (data) {
 		case KBD98_RET_ACK:
-			kbd98->ack = 1;
+			atomic_set(&kbd98->ack, 1);
 			return;
 		case KBD98_RET_NAK:
-			kbd98->ack = -1;
+			atomic_set(&kbd98->ack, 2);
 			return;
 	}
 
@@ -220,14 +220,14 @@
 static int kbd98_sendbyte(struct kbd98 *kbd98, unsigned char byte)
 {
 	int timeout = 10000; /* 100 msec */
-	kbd98->ack = 0;
+	atomic_set(&kbd98->ack, 0);
 
 	if (serio_write(kbd98->serio, byte))
 		return -1;
 
-	while (!kbd98->ack && timeout--) udelay(10);
+	while (!atomic_read(&kbd98->ack) && timeout--) udelay(10);
 
-	return -(kbd98->ack <= 0);
+	return -(atomic_read(&kbd98->ack) != 1);
 }
 
 /*
Index: linux/drivers/input/mouse/psmouse-base.c
===================================================================
--- linux.orig/drivers/input/mouse/psmouse-base.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/mouse/psmouse-base.c	2003-11-24 11:29:30.000000000 +0100
@@ -39,8 +39,8 @@
 #define PSMOUSE_LOGITECH_SMARTSCROLL	1
 
 static int psmouse_noext;
-int psmouse_resolution;
-unsigned int psmouse_rate = 60;
+static int psmouse_resolution;
+static unsigned int psmouse_rate = 60;
 int psmouse_smartscroll = PSMOUSE_LOGITECH_SMARTSCROLL;
 unsigned int psmouse_resetafter;
 
@@ -121,13 +121,14 @@
 	if (psmouse->acking) {
 		switch (data) {
 			case PSMOUSE_RET_ACK:
-				psmouse->ack = 1;
+				atomic_set(&psmouse->ack, 1);
 				break;
 			case PSMOUSE_RET_NAK:
-				psmouse->ack = -1;
+				atomic_set(&psmouse->ack, 2);
 				break;
 			default:
-				psmouse->ack = 1;	/* Workaround for mice which don't ACK the Get ID command */
+				/* Workaround for mice which don't ACK the Get ID command */
+				atomic_set(&psmouse->ack, 1);
 				if (psmouse->cmdcnt)
 					psmouse->cmdbuf[--psmouse->cmdcnt] = data;
 				break;
@@ -197,7 +198,7 @@
 static int psmouse_sendbyte(struct psmouse *psmouse, unsigned char byte)
 {
 	int timeout = 10000; /* 100 msec */
-	psmouse->ack = 0;
+	atomic_set(&psmouse->ack, 0);
 	psmouse->acking = 1;
 
 	if (serio_write(psmouse->serio, byte)) {
@@ -205,9 +206,9 @@
 		return -1;
 	}
 
-	while (!psmouse->ack && timeout--) udelay(10);
+	while (!atomic_read(&psmouse->ack) && timeout--) udelay(10);
 
-	return -(psmouse->ack <= 0);
+	return -(atomic_read(&psmouse->ack) != 1);
 }
 
 /*
@@ -439,8 +440,7 @@
 		param[0] = 2;
 	else if (psmouse_resolution >= 50)
 		param[0] = 1;
-	else if (psmouse_resolution)
-		param[0] = 0;
+	else	param[0] = 0;
 
         psmouse_command(psmouse, param, PSMOUSE_CMD_SETRES);
 }
@@ -532,20 +532,11 @@
 static int psmouse_pm_callback(struct pm_dev *dev, pm_request_t request, void *data)
 {
 	struct psmouse *psmouse = dev->data;
-	struct serio_dev *ser_dev = psmouse->serio->dev;
-
-	synaptics_disconnect(psmouse);
-
-	/* We need to reopen the serio port to reinitialize the i8042 controller */
-	serio_close(psmouse->serio);
-	serio_open(psmouse->serio, ser_dev);
-
-	/* Probe and re-initialize the mouse */
-	psmouse_probe(psmouse);
-	psmouse_initialize(psmouse);
-	synaptics_pt_init(psmouse);
-	psmouse_activate(psmouse);
 
+	if (request == PM_RESUME) {
+		psmouse->state = PSMOUSE_IGNORE;
+		serio_reconnect(psmouse->serio);
+	}
 	return 0;
 }
 
Index: linux/drivers/input/mouse/psmouse.h
===================================================================
--- linux.orig/drivers/input/mouse/psmouse.h	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/mouse/psmouse.h	2003-11-24 11:29:43.000000000 +0100
@@ -37,7 +37,7 @@
 	unsigned long last;
 	unsigned char state;
 	char acking;
-	volatile char ack;
+	atomic_t ack;	/* This is being accessed without locking, at least make sure we do not run into alignment problems */
 	char error;
 	char devname[64];
 	char phys[32];
Index: linux/drivers/input/serio/serio.c
===================================================================
--- linux.orig/drivers/input/serio/serio.c	2003-11-24 11:30:30.000000000 +0100
+++ linux/drivers/input/serio/serio.c	2003-11-24 11:30:06.000000000 +0100
@@ -83,6 +83,7 @@
 }
 
 #define SERIO_RESCAN	1
+#define SERIO_RECONNECT	2
 
 static DECLARE_WAIT_QUEUE_HEAD(serio_wait);
 static DECLARE_COMPLETION(serio_exited);
@@ -111,6 +112,38 @@
 	}
 }
 
+static void serio_invalidate_pending_events(struct serio *serio)
+{
+	struct serio_event *event;
+
+	list_for_each_entry(event, &serio_event_list, node)
+		if (event->serio == serio)
+			event->serio = NULL;
+}
+
+static void serio_queue_event(struct serio *serio, int event_type)
+{
+	struct serio_event *event;
+
+	if ((event = kmalloc(sizeof(struct serio_event), GFP_ATOMIC))) {
+		event->type = event_type;
+		event->serio = serio;
+
+		list_add_tail(&event->node, &serio_event_list);
+		wake_up(&serio_wait);
+	}
+}
+
+void serio_rescan(struct serio *serio)
+{
+	serio_queue_event(serio, SERIO_RESCAN);
+}
+
+void serio_reconnect(struct serio *serio)
+{
+	serio_queue_event(serio, SERIO_RECONNECT);
+}
+
 static int serio_thread(void *nothing)
 {
 	lock_kernel();
@@ -130,20 +163,6 @@
 	complete_and_exit(&serio_exited, 0);
 }
 
-void serio_rescan(struct serio *serio)
-{
-	struct serio_event *event;
-
-	if (!(event = kmalloc(sizeof(struct serio_event), GFP_ATOMIC)))
-		return;
-
-	event->type = SERIO_RESCAN;
-	event->serio = serio;
-
-	list_add_tail(&event->node, &serio_event_list);
-	wake_up(&serio_wait);
-}
-
 irqreturn_t serio_interrupt(struct serio *serio,
 		unsigned char data, unsigned int flags, struct pt_regs *regs)
 {

%diffstat
 Documentation/power/swsusp.txt     |  104 ++++++---------------------
 Documentation/power/video.txt      |   36 +++++++++
 arch/i386/kernel/acpi/wakeup.S     |    8 --
 arch/i386/kernel/cpu/mtrr/main.c   |    3 
 arch/i386/kernel/suspend_asm.S     |   94 ------------------------
 arch/i386/kernel/time.c            |   25 ++++++
 arch/x86_64/kernel/time.c          |   25 +++++-
 drivers/acpi/thermal.c             |    1 
 drivers/base/power.c               |  140 -------------------------------------
 drivers/input/keyboard/98kbd.c     |   12 +--
 drivers/input/keyboard/atkbd.c     |   72 +++++++++++--------
 drivers/input/mouse/psmouse-base.c |   37 +++------
 drivers/input/mouse/psmouse.h      |    2 
 drivers/input/power.c              |   12 +++
 drivers/input/serio/i8042.c        |   40 +++++++++-
 drivers/input/serio/serio.c        |   47 ++++++++----
 include/linux/serio.h              |    1 
 include/linux/suspend.h            |   26 +++++-
 kernel/power/process.c             |    9 +-
 kernel/power/swsusp.c              |   91 +++++++++---------------
 kernel/sys.c                       |    8 --
 21 files changed, 317 insertions(+), 476 deletions(-)


-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]