Prev: KProbes support for MIPS
Next: linux-next: Tree for August 4
From: KAMEZAWA Hiroyuki on 4 Aug 2010 01:10

Fixed description and points of hooks. maybe much clearer.

==
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com>

At taking memory snapshot in hibernate_snapshot(), all (directly called)
memory allocator uses GFP_ATOMIC. And it seems swap-misusage during
hibernation never occurs.

But, from pessimistic point of view, there is no guarantee to trust
any page allcation doesn't have __GFP_WAIT. It's better to have an indication
"we enter hibernation, don't use swap!".

This patch tries to freeze new-swap-allocation during hibernation.
(We can trust all user processes are freezed, then, dont't take care of swapin)

By this, no updates will be happen to swap_map[] among hibernate_snapshot()
to save_image(). swap is thawed when swsusp_free() is called.
We can trust swap-corruption will never happen without any doubts.

Changelog: 2010-08-04
- Fixed hibernation_freeze_swap/thaw_swap call points.
- Rewrite the all description.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu(a)jp.fujitsu.com>
---
include/linux/swap.h | 8 +++-
kernel/power/hibernate.c | 1
kernel/power/snapshot.c | 1
kernel/power/swap.c | 6 +--
mm/swapfile.c | 94 ++++++++++++++++++++++++++++++++++++-----------
5 files changed, 84 insertions(+), 26 deletions(-)

Index: mmotm-0727/include/linux/swap.h
===================================================================
--- mmotm-0727.orig/include/linux/swap.h
+++ mmotm-0727/include/linux/swap.h
@@ -316,7 +316,6 @@ extern long nr_swap_pages;
extern long total_swap_pages;
extern void si_swapinfo(struct sysinfo *);
extern swp_entry_t get_swap_page(void);
-extern swp_entry_t get_swap_page_of_type(int);
extern int valid_swaphandles(swp_entry_t, unsigned long *);
extern int add_swap_count_continuation(swp_entry_t, gfp_t);
extern void swap_shmem_alloc(swp_entry_t);
@@ -333,6 +332,13 @@ extern int reuse_swap_page(struct page *
extern int try_to_free_swap(struct page *);
struct backing_dev_info;

+#ifdef CONFIG_HIBERNATION
+void hibernation_freeze_swap(void);
+void hibernation_thaw_swap(void);
+swp_entry_t get_swap_for_hibernation(int type);
+void swap_free_for_hibernation(swp_entry_t val);
+#endif
+
/* linux/mm/thrash.c */
extern struct mm_struct *swap_token_mm;
extern void grab_swap_token(struct mm_struct *);
Index: mmotm-0727/mm/swapfile.c
===================================================================
--- mmotm-0727.orig/mm/swapfile.c
+++ mmotm-0727/mm/swapfile.c
@@ -47,6 +47,8 @@ long nr_swap_pages;
long total_swap_pages;
static int least_priority;

+static bool swap_for_hibernation;
+
static const char Bad_file[] = "Bad swap file entry ";
static const char Unused_file[] = "Unused swap file entry ";
static const char Bad_offset[] = "Bad swap offset entry ";
@@ -449,6 +451,8 @@ swp_entry_t get_swap_page(void)
spin_lock(&swap_lock);
if (nr_swap_pages <= 0)
goto noswap;
+ if (swap_for_hibernation)
+ goto noswap;
nr_swap_pages--;

for (type = swap_list.next; type >= 0 && wrapped < 2; type = next) {
@@ -481,28 +485,6 @@ noswap:
return (swp_entry_t) {0};
}

-/* The only caller of this function is now susupend routine */
-swp_entry_t get_swap_page_of_type(int type)
-{
- struct swap_info_struct *si;
- pgoff_t offset;
-
- spin_lock(&swap_lock);
- si = swap_info[type];
- if (si && (si->flags & SWP_WRITEOK)) {
- nr_swap_pages--;
- /* This is called for allocating swap entry, not cache */
- offset = scan_swap_map(si, 1);
- if (offset) {
- spin_unlock(&swap_lock);
- return swp_entry(type, offset);
- }
- nr_swap_pages++;
- }
- spin_unlock(&swap_lock);
- return (swp_entry_t) {0};
-}
-
static struct swap_info_struct *swap_info_get(swp_entry_t entry)
{
struct swap_info_struct *p;
@@ -762,6 +744,74 @@ int mem_cgroup_count_swap_user(swp_entry
#endif

#ifdef CONFIG_HIBERNATION
+
+static pgoff_t hibernation_offset[MAX_SWAPFILES];
+/*
+ * Once hibernation starts to use swap, we freeze swap_map[]. Otherwise,
+ * saved swap_map[] image to the disk will be an incomplete because it's
+ * changing without synchronization with hibernation snap shot.
+ * At resume, we just make swap_for_hibernation=false. We can forget
+ * used maps easily.
+ */
+void hibernation_freeze_swap(void)
+{
+ int i;
+
+ spin_lock(&swap_lock);
+
+ printk(KERN_INFO "PM: Freeze Swap\n");
+ swap_for_hibernation = true;
+ for (i = 0; i < MAX_SWAPFILES; i++)
+ hibernation_offset[i] = 1;
+ spin_unlock(&swap_lock);
+}
+
+void hibernation_thaw_swap(void)
+{
+ spin_lock(&swap_lock);
+ if (swap_for_hibernation) {
+ printk(KERN_INFO "PM: Thaw Swap\n");
+ swap_for_hibernation = false;
+ }
+ spin_unlock(&swap_lock);
+}
+
+/*
+ * Because updateing swap_map[] can make not-saved-status-change,
+ * we use our own easy allocator.
+ * Please see kernel/power/swap.c, Used swaps are recorded into
+ * RB-tree.
+ */
+swp_entry_t get_swap_for_hibernation(int type)
+{
+ pgoff_t off;
+ swp_entry_t val = {0};
+ struct swap_info_struct *si;
+
+ spin_lock(&swap_lock);
+
+ si = swap_info[type];
+ if (!si || !(si->flags & SWP_WRITEOK))
+ goto done;
+
+ for (off = hibernation_offset[type]; off < si->max; ++off) {
+ if (!si->swap_map[off])
+ break;
+ }
+ if (off < si->max) {
+ val = swp_entry(type, off);
+ hibernation_offset[type] = off + 1;
+ }
+done:
+ spin_unlock(&swap_lock);
+ return val;
+}
+
+void swap_free_for_hibernation(swp_entry_t ent)
+{
+ /* Nothing to do */
+}
+
/*
* Find the swap type that corresponds to given device (if any).
*
Index: mmotm-0727/kernel/power/swap.c
===================================================================
--- mmotm-0727.orig/kernel/power/swap.c
+++ mmotm-0727/kernel/power/swap.c
@@ -135,10 +135,10 @@ sector_t alloc_swapdev_block(int swap)
{
unsigned long offset;

- offset = swp_offset(get_swap_page_of_type(swap));
+ offset = swp_offset(get_swap_for_hibernation(swap));
if (offset) {
if (swsusp_extents_insert(offset))
- swap_free(swp_entry(swap, offset));
+ swap_free_for_hibernation(swp_entry(swap, offset));
else
return swapdev_block(swap, offset);
}
@@ -162,7 +162,7 @@ void free_all_swap_pages(int swap)
ext = container_of(node, struct swsusp_extent, node);
rb_erase(node, &swsusp_extents);
for (offset = ext->start; offset <= ext->end; offset++)
- swap_free(swp_entry(swap, offset));
+ swap_free_for_hibernation(swp_entry(swap, offset));

kfree(ext);
}
Index: mmotm-0727/kernel/power/hibernate.c
===================================================================
--- mmotm-0727.orig/kernel/power/hibernate.c
+++ mmotm-0727/kernel/power/hibernate.c
@@ -338,6 +338,7 @@ int hibernation_snapshot(int platform_mo
goto Close;

suspend_console();
+ hibernation_freeze_swap();
saved_mask = clear_gfp_allowed_mask(GFP_IOFS);
error = dpm_suspend_start(PMSG_FREEZE);
if (error)
Index: mmotm-0727/kernel/power/snapshot.c
===================================================================
--- mmotm-0727.orig/kernel/power/snapshot.c
+++ mmotm-0727/kernel/power/snapshot.c
@@ -1086,6 +1086,7 @@ void swsusp_free(void)
buffer = NULL;
alloc_normal = 0;
alloc_highmem = 0;
+ hibernation_thaw_swap();
}

/* Helper functions used for the shrinking of memory. */

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: KProbes support for MIPS
Next: linux-next: Tree for August 4