PM: hibernate: swap: defer linking the next map page

Browse source

Delay allocating and linking the next swap_map_page until another image
page actually needs to be recorded.

The previous code linked and wrote a new swap map page as soon as the
current one became full. When the image size was an exact multiple of
MAP_PAGE_ENTRIES, that left an empty final map page that existed only to
terminate the on-disk chain.

Instead, keep a full map page in memory and only allocate the next map page
when the next image page arrives. This preserves the resume chain while
avoiding an unnecessary swap slot allocation and write for the empty
swap map page.

Signed-off-by: Haesung Kim <mattkim513@gmail.com>
Link: https://patch.msgid.link/20260714072627.3165744-1-mattkim513@gmail.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

...

This commit is contained in:

Haesung Kim 2026-07-14 16:26:27 +09:00 • committed by Rafael J. Wysocki

parent 21d5c4cee3

commit 032270c308

1 changed files with 20 additions and 10 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download patch file Download diff file Expand all files Collapse all files

30

kernel/power/swap.c

View file