[PATCH] swsusp: read speedup
Implement async reads for swsusp resuming.
Crufty old PIII testbox:
15.7 MB/s -> 20.3 MB/s
Sony Vaio:
14.6 MB/s -> 33.3 MB/s
I didn't implement the post-resume bio_set_pages_dirty(). I don't really
understand why resume needs to run set_page_dirty() against these pages.
It might be a worry that this code modifies PG_Uptodate, PG_Error and
PG_Locked against the image pages. Can this possibly affect the resumed-into
kernel? Hopefully not, if we're atomically restoring its mem_map?
Cc: Pavel Machek <pavel@ucw.cz>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Jens Axboe <axboe@suse.de>
Cc: Laurent Riffard <laurent.riffard@free.fr>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c
index 75d4886..591301a 100644
--- a/kernel/power/snapshot.c
+++ b/kernel/power/snapshot.c
@@ -314,7 +314,7 @@
* and we count them using unsafe_pages
*/
-static inline void *alloc_image_page(gfp_t gfp_mask, int safe_needed)
+static void *alloc_image_page(gfp_t gfp_mask, int safe_needed)
{
void *res;
@@ -828,13 +828,16 @@
}
if (!handle->offset)
handle->buffer = buffer;
+ handle->sync_read = 1;
if (handle->prev < handle->page) {
if (!handle->prev) {
- error = load_header(handle, (struct swsusp_info *)buffer);
+ error = load_header(handle,
+ (struct swsusp_info *)buffer);
if (error)
return error;
} else if (handle->prev <= nr_meta_pages) {
- handle->pbe = unpack_orig_addresses(buffer, handle->pbe);
+ handle->pbe = unpack_orig_addresses(buffer,
+ handle->pbe);
if (!handle->pbe) {
error = prepare_image(handle);
if (error)
@@ -842,10 +845,12 @@
handle->pbe = pagedir_nosave;
handle->last_pbe = NULL;
handle->buffer = get_buffer(handle);
+ handle->sync_read = 0;
}
} else {
handle->pbe = handle->pbe->next;
handle->buffer = get_buffer(handle);
+ handle->sync_read = 0;
}
handle->prev = handle->page;
}
