mm/kmemleak.c: make cond_resched() rate-limiting more efficient
authorAndrew Morton <akpm@linux-foundation.org>
Thu, 14 Dec 2017 23:32:31 +0000 (15:32 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 15 Dec 2017 00:00:48 +0000 (16:00 -0800)
Commit bde5f6bc68db ("kmemleak: add scheduling point to
kmemleak_scan()") tries to rate-limit the frequency of cond_resched()
calls, but does it in a way which might incur an expensive division
operation in the inner loop.  Simplify this.

Fixes: bde5f6bc68db5 ("kmemleak: add scheduling point to kmemleak_scan()")
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Yisheng Xie <xieyisheng1@huawei.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/kmemleak.c

index 3d47817..d73c142 100644 (file)
@@ -1523,7 +1523,7 @@ static void kmemleak_scan(void)
                        if (page_count(page) == 0)
                                continue;
                        scan_block(page, page + 1, NULL);
-                       if (!(pfn % (MAX_SCAN_SIZE / sizeof(*page))))
+                       if (!(pfn & 63))
                                cond_resched();
                }
        }