Merge rsync://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6

Conflicts:

	include/linux/kernel.h

6 files changed, 127 insertions, 81 deletions

diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig
index ce0dfb8f4a4e..ae44a70aae8a 100644
--- a/kernel/power/Kconfig
+++ b/kernel/power/Kconfig
@@ -36,6 +36,24 @@ config PM_DEBUG
 	code. This is helpful when debugging and reporting various PM bugs, 
 	like suspend support.
 
+config PM_TRACE
+	bool "Suspend/resume event tracing"
+	depends on PM && PM_DEBUG && X86_32 && EXPERIMENTAL
+	default n
+	---help---
+	This enables some cheesy code to save the last PM event point in the
+	RTC across reboots, so that you can debug a machine that just hangs
+	during suspend (or more commonly, during resume).
+
+	To use this debugging feature you should attempt to suspend the machine,
+	then reboot it, then run
+
+		dmesg -s 1000000 | grep 'hash matches'
+
+	CAUTION: this option will cause your machine's real-time clock to be
+	set to an invalid time after a resume.
+
+
 config SOFTWARE_SUSPEND
 	bool "Software Suspend"
 	depends on PM && SWAP && (X86 && (!SMP || SUSPEND_SMP)) || ((FRV || PPC32) && !SMP)
@@ -82,18 +100,6 @@ config PM_STD_PARTITION
 	  suspended image to. It will simply pick the first available swap 
 	  device.
 
-config SWSUSP_ENCRYPT
-	bool "Encrypt suspend image"
-	depends on SOFTWARE_SUSPEND && CRYPTO=y && (CRYPTO_AES=y || CRYPTO_AES_586=y || CRYPTO_AES_X86_64=y)
-	default ""
-	---help---
-	  To prevent data gathering from swap after resume you can encrypt
-	  the suspend image with a temporary key that is deleted on
-	  resume.
-
-	  Note that the temporary key is stored unencrypted on disk while the
-	  system is suspended.
-
 config SUSPEND_SMP
 	bool
 	depends on HOTPLUG_CPU && X86 && PM
diff --git a/kernel/power/disk.c b/kernel/power/disk.c
index 81d4d982f3f0..e13e74067845 100644
--- a/kernel/power/disk.c
+++ b/kernel/power/disk.c
@@ -231,7 +231,7 @@ static int software_resume(void)
 late_initcall(software_resume);
 
 
-static char * pm_disk_modes[] = {
+static const char * const pm_disk_modes[] = {
 	[PM_DISK_FIRMWARE]	= "firmware",
 	[PM_DISK_PLATFORM]	= "platform",
 	[PM_DISK_SHUTDOWN]	= "shutdown",
diff --git a/kernel/power/main.c b/kernel/power/main.c
index 0a907f0dc56b..6d295c776794 100644
--- a/kernel/power/main.c
+++ b/kernel/power/main.c
@@ -15,7 +15,7 @@
 #include <linux/errno.h>
 #include <linux/init.h>
 #include <linux/pm.h>
-
+#include <linux/console.h>
 
 #include "power.h"
 
@@ -145,7 +145,7 @@ static void suspend_finish(suspend_state_t state)
 
 
 
-static char *pm_states[PM_SUSPEND_MAX] = {
+static const char * const pm_states[PM_SUSPEND_MAX] = {
 	[PM_SUSPEND_STANDBY]	= "standby",
 	[PM_SUSPEND_MEM]	= "mem",
 #ifdef CONFIG_SOFTWARE_SUSPEND
@@ -262,7 +262,7 @@ static ssize_t state_show(struct subsystem * subsys, char * buf)
 static ssize_t state_store(struct subsystem * subsys, const char * buf, size_t n)
 {
 	suspend_state_t state = PM_SUSPEND_STANDBY;
-	char ** s;
+	const char * const *s;
 	char *p;
 	int error;
 	int len;
diff --git a/kernel/power/power.h b/kernel/power/power.h
index f06f12f21767..57a792982fb9 100644
--- a/kernel/power/power.h
+++ b/kernel/power/power.h
@@ -55,7 +55,7 @@ struct snapshot_handle {
 	unsigned int	page;
 	unsigned int	page_offset;
 	unsigned int	prev;
-	struct pbe	*pbe;
+	struct pbe	*pbe, *last_pbe;
 	void		*buffer;
 	unsigned int	buf_offset;
 };
diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c
index 3eeedbb13b78..24c96f354231 100644
--- a/kernel/power/snapshot.c
+++ b/kernel/power/snapshot.c
@@ -150,6 +150,10 @@ int restore_highmem(void)
 	}
 	return 0;
 }
+#else
+static inline unsigned int count_highmem_pages(void) {return 0;}
+static inline int save_highmem(void) {return 0;}
+static inline int restore_highmem(void) {return 0;}
 #endif
 
 static int pfn_is_nosave(unsigned long pfn)
@@ -293,62 +297,29 @@ static inline void create_pbe_list(struct pbe *pblist, unsigned int nr_pages)
 	}
 }
 
-/**
- *	On resume it is necessary to trace and eventually free the unsafe
- *	pages that have been allocated, because they are needed for I/O
- *	(on x86-64 we likely will "eat" these pages once again while
- *	creating the temporary page translation tables)
- */
-
-struct eaten_page {
-	struct eaten_page *next;
-	char padding[PAGE_SIZE - sizeof(void *)];
-};
-
-static struct eaten_page *eaten_pages = NULL;
-
-static void release_eaten_pages(void)
-{
-	struct eaten_page *p, *q;
-
-	p = eaten_pages;
-	while (p) {
-		q = p->next;
-		/* We don't want swsusp_free() to free this page again */
-		ClearPageNosave(virt_to_page(p));
-		free_page((unsigned long)p);
-		p = q;
-	}
-	eaten_pages = NULL;
-}
+static unsigned int unsafe_pages;
 
 /**
  *	@safe_needed - on resume, for storing the PBE list and the image,
  *	we can only use memory pages that do not conflict with the pages
- *	which had been used before suspend.
+ *	used before suspend.
  *
  *	The unsafe pages are marked with the PG_nosave_free flag
- *
- *	Allocated but unusable (ie eaten) memory pages should be marked
- *	so that swsusp_free() can release them
+ *	and we count them using unsafe_pages
  */
 
 static inline void *alloc_image_page(gfp_t gfp_mask, int safe_needed)
 {
 	void *res;
 
+	res = (void *)get_zeroed_page(gfp_mask);
 	if (safe_needed)
-		do {
+		while (res && PageNosaveFree(virt_to_page(res))) {
+			/* The page is unsafe, mark it for swsusp_free() */
+			SetPageNosave(virt_to_page(res));
+			unsafe_pages++;
 			res = (void *)get_zeroed_page(gfp_mask);
-			if (res && PageNosaveFree(virt_to_page(res))) {
-				/* This is for swsusp_free() */
-				SetPageNosave(virt_to_page(res));
-				((struct eaten_page *)res)->next = eaten_pages;
-				eaten_pages = res;
-			}
-		} while (res && PageNosaveFree(virt_to_page(res)));
-	else
-		res = (void *)get_zeroed_page(gfp_mask);
+		}
 	if (res) {
 		SetPageNosave(virt_to_page(res));
 		SetPageNosaveFree(virt_to_page(res));
@@ -374,7 +345,8 @@ unsigned long get_safe_page(gfp_t gfp_mask)
  *	On each page we set up a list of struct_pbe elements.
  */
 
-struct pbe *alloc_pagedir(unsigned int nr_pages, gfp_t gfp_mask, int safe_needed)
+static struct pbe *alloc_pagedir(unsigned int nr_pages, gfp_t gfp_mask,
+				 int safe_needed)
 {
 	unsigned int num;
 	struct pbe *pblist, *pbe;
@@ -642,6 +614,8 @@ static int mark_unsafe_pages(struct pbe *pblist)
 			return -EFAULT;
 	}
 
+	unsafe_pages = 0;
+
 	return 0;
 }
 
@@ -719,42 +693,99 @@ static inline struct pbe *unpack_orig_addresses(unsigned long *buf,
 }
 
 /**
- *	create_image - use metadata contained in the PBE list
+ *	prepare_image - use metadata contained in the PBE list
  *	pointed to by pagedir_nosave to mark the pages that will
  *	be overwritten in the process of restoring the system
- *	memory state from the image and allocate memory for
- *	the image avoiding these pages
+ *	memory state from the image ("unsafe" pages) and allocate
+ *	memory for the image
+ *
+ *	The idea is to allocate the PBE list first and then
+ *	allocate as many pages as it's needed for the image data,
+ *	but not to assign these pages to the PBEs initially.
+ *	Instead, we just mark them as allocated and create a list
+ *	of "safe" which will be used later
  */
 
-static int create_image(struct snapshot_handle *handle)
+struct safe_page {
+	struct safe_page *next;
+	char padding[PAGE_SIZE - sizeof(void *)];
+};
+
+static struct safe_page *safe_pages;
+
+static int prepare_image(struct snapshot_handle *handle)
 {
 	int error = 0;
-	struct pbe *p, *pblist;
+	unsigned int nr_pages = nr_copy_pages;
+	struct pbe *p, *pblist = NULL;
 
 	p = pagedir_nosave;
 	error = mark_unsafe_pages(p);
 	if (!error) {
-		pblist = alloc_pagedir(nr_copy_pages, GFP_ATOMIC, 1);
+		pblist = alloc_pagedir(nr_pages, GFP_ATOMIC, 1);
 		if (pblist)
 			copy_page_backup_list(pblist, p);
 		free_pagedir(p, 0);
 		if (!pblist)
 			error = -ENOMEM;
 	}
-	if (!error)
-		error = alloc_data_pages(pblist, GFP_ATOMIC, 1);
+	safe_pages = NULL;
+	if (!error && nr_pages > unsafe_pages) {
+		nr_pages -= unsafe_pages;
+		while (nr_pages--) {
+			struct safe_page *ptr;
+
+			ptr = (struct safe_page *)get_zeroed_page(GFP_ATOMIC);
+			if (!ptr) {
+				error = -ENOMEM;
+				break;
+			}
+			if (!PageNosaveFree(virt_to_page(ptr))) {
+				/* The page is "safe", add it to the list */
+				ptr->next = safe_pages;
+				safe_pages = ptr;
+			}
+			/* Mark the page as allocated */
+			SetPageNosave(virt_to_page(ptr));
+			SetPageNosaveFree(virt_to_page(ptr));
+		}
+	}
 	if (!error) {
-		release_eaten_pages();
 		pagedir_nosave = pblist;
 	} else {
-		pagedir_nosave = NULL;
 		handle->pbe = NULL;
-		nr_copy_pages = 0;
-		nr_meta_pages = 0;
+		swsusp_free();
 	}
 	return error;
 }
 
+static void *get_buffer(struct snapshot_handle *handle)
+{
+	struct pbe *pbe = handle->pbe, *last = handle->last_pbe;
+	struct page *page = virt_to_page(pbe->orig_address);
+
+	if (PageNosave(page) && PageNosaveFree(page)) {
+		/*
+		 * We have allocated the "original" page frame and we can
+		 * use it directly to store the read page
+		 */
+		pbe->address = 0;
+		if (last && last->next)
+			last->next = NULL;
+		return (void *)pbe->orig_address;
+	}
+	/*
+	 * The "original" page frame has not been allocated and we have to
+	 * use a "safe" page frame to store the read page
+	 */
+	pbe->address = (unsigned long)safe_pages;
+	safe_pages = safe_pages->next;
+	if (last)
+		last->next = pbe;
+	handle->last_pbe = pbe;
+	return (void *)pbe->address;
+}
+
 /**
  *	snapshot_write_next - used for writing the system memory snapshot.
  *
@@ -799,15 +830,16 @@ int snapshot_write_next(struct snapshot_handle *handle, size_t count)
 		} else if (handle->prev <= nr_meta_pages) {
 			handle->pbe = unpack_orig_addresses(buffer, handle->pbe);
 			if (!handle->pbe) {
-				error = create_image(handle);
+				error = prepare_image(handle);
 				if (error)
 					return error;
 				handle->pbe = pagedir_nosave;
-				handle->buffer = (void *)handle->pbe->address;
+				handle->last_pbe = NULL;
+				handle->buffer = get_buffer(handle);
 			}
 		} else {
 			handle->pbe = handle->pbe->next;
-			handle->buffer = (void *)handle->pbe->address;
+			handle->buffer = get_buffer(handle);
 		}
 		handle->prev = handle->page;
 	}
diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c
index c4016cbbd3e0..17f669c83012 100644
--- a/kernel/power/swsusp.c
+++ b/kernel/power/swsusp.c
@@ -67,9 +67,9 @@ unsigned int count_highmem_pages(void);
 int save_highmem(void);
 int restore_highmem(void);
 #else
-static int save_highmem(void) { return 0; }
-static int restore_highmem(void) { return 0; }
-static unsigned int count_highmem_pages(void) { return 0; }
+static inline int save_highmem(void) { return 0; }
+static inline int restore_highmem(void) { return 0; }
+static inline unsigned int count_highmem_pages(void) { return 0; }
 #endif
 
 /**
@@ -175,6 +175,12 @@ void free_all_swap_pages(int swap, struct bitmap_page *bitmap)
  */
 
 #define SHRINK_BITE	10000
+static inline unsigned long __shrink_memory(long tmp)
+{
+	if (tmp > SHRINK_BITE)
+		tmp = SHRINK_BITE;
+	return shrink_all_memory(tmp);
+}
 
 int swsusp_shrink_memory(void)
 {
@@ -192,15 +198,17 @@ int swsusp_shrink_memory(void)
 			PAGES_FOR_IO;
 		tmp = size;
 		for_each_zone (zone)
-			if (!is_highmem(zone))
+			if (!is_highmem(zone) && populated_zone(zone)) {
 				tmp -= zone->free_pages;
+				tmp += zone->lowmem_reserve[ZONE_NORMAL];
+			}
 		if (tmp > 0) {
-			tmp = shrink_all_memory(SHRINK_BITE);
+			tmp = __shrink_memory(tmp);
 			if (!tmp)
 				return -ENOMEM;
 			pages += tmp;
 		} else if (size > image_size / PAGE_SIZE) {
-			tmp = shrink_all_memory(SHRINK_BITE);
+			tmp = __shrink_memory(size - (image_size / PAGE_SIZE));
 			pages += tmp;
 		}
 		printk("\b%c", p[i++%4]);