diff options
author | Oleg Nesterov | 2010-08-11 03:03:17 +0200 |
---|---|---|
committer | Linus Torvalds | 2010-08-11 17:59:20 +0200 |
commit | c52b0b91ba1f4b7ea90e20385c0a6df0ba54aed4 (patch) | |
tree | 6d92439150a3218da5807610991da4d70af56bf8 /kernel | |
parent | pids: fix a race in pid generation that causes pids to be reused immediately (diff) | |
download | kernel-qcow2-linux-c52b0b91ba1f4b7ea90e20385c0a6df0ba54aed4.tar.gz kernel-qcow2-linux-c52b0b91ba1f4b7ea90e20385c0a6df0ba54aed4.tar.xz kernel-qcow2-linux-c52b0b91ba1f4b7ea90e20385c0a6df0ba54aed4.zip |
pids: alloc_pidmap: remove the unnecessary boundary checks
alloc_pidmap() calculates max_scan so that if the initial offset != 0 we
inspect the first map->page twice. This is correct, we want to find the
unused bits < offset in this bitmap block. Add the comment.
But it doesn't make any sense to stop the find_next_offset() loop when we
are looking into this map->page for the second time. We have already
already checked the bits >= offset during the first attempt, it is fine to
do this again, no matter if we succeed this time or not.
Remove this hard-to-understand code. It optimizes the very unlikely case
when we are going to fail, but slows down the more likely case.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Salman Qazi <sqazi@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'kernel')
-rw-r--r-- | kernel/pid.c | 17 |
1 files changed, 7 insertions, 10 deletions
diff --git a/kernel/pid.c b/kernel/pid.c index fbbd5f6b6f2f..d55c6fb8d087 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -169,7 +169,12 @@ static int alloc_pidmap(struct pid_namespace *pid_ns) pid = RESERVED_PIDS; offset = pid & BITS_PER_PAGE_MASK; map = &pid_ns->pidmap[pid/BITS_PER_PAGE]; - max_scan = (pid_max + BITS_PER_PAGE - 1)/BITS_PER_PAGE - !offset; + /* + * If last_pid points into the middle of the map->page we + * want to scan this bitmap block twice, the second time + * we start with offset == 0 (or RESERVED_PIDS). + */ + max_scan = DIV_ROUND_UP(pid_max, BITS_PER_PAGE) - !offset; for (i = 0; i <= max_scan; ++i) { if (unlikely(!map->page)) { void *page = kzalloc(PAGE_SIZE, GFP_KERNEL); @@ -196,15 +201,7 @@ static int alloc_pidmap(struct pid_namespace *pid_ns) } offset = find_next_offset(map, offset); pid = mk_pid(pid_ns, map, offset); - /* - * find_next_offset() found a bit, the pid from it - * is in-bounds, and if we fell back to the last - * bitmap block and the final block was the same - * as the starting point, pid is before last_pid. - */ - } while (offset < BITS_PER_PAGE && pid < pid_max && - (i != max_scan || pid < last || - !((last+1) & BITS_PER_PAGE_MASK))); + } while (offset < BITS_PER_PAGE && pid < pid_max); } if (map < &pid_ns->pidmap[(pid_max-1)/BITS_PER_PAGE]) { ++map; |