[PATCH AUTOSEL 5.3 65/89] perf/core: Rework memory accounting in perf_mmap()

[Sasha Levin]

From: Song Liu <songliubraving@xxxxxx>

[ Upstream commit d44248a41337731a111374822d7d4451b64e73e4 ]

perf_mmap() always increases user->locked_vm. As a result, "extra" could
grow bigger than "user_extra", which doesn't make sense. Here is an
example case:

(Note: Assume "user_lock_limit" is very small.)

  | # of perf_mmap calls |vma->vm_mm->pinned_vm|user->locked_vm|
  | 0                    | 0                   | 0             |
  | 1                    | user_extra          | user_extra    |
  | 2                    | 3 * user_extra      | 2 * user_extra|
  | 3                    | 6 * user_extra      | 3 * user_extra|
  | 4                    | 10 * user_extra     | 4 * user_extra|

Fix this by maintaining proper user_extra and extra.

Reviewed-By: Hechao Li <hechaol@xxxxxx>
Reported-by: Hechao Li <hechaol@xxxxxx>
Signed-off-by: Song Liu <songliubraving@xxxxxx>
Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Cc: <kernel-team@xxxxxx>
Cc: Jie Meng <jmeng@xxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Link: https://lkml.kernel.org/r/20190904214618.3795672-1-songliubraving@xxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
 kernel/events/core.c | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

diff --git a/kernel/events/core.c b/kernel/events/core.c
index 0463c1151baeb..a0120bdbce177 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -5585,7 +5585,8 @@ static void perf_mmap_close(struct vm_area_struct *vma)
 	 * undo the VM accounting.
 	 */
 
-	atomic_long_sub((size >> PAGE_SHIFT) + 1, &mmap_user->locked_vm);
+	atomic_long_sub((size >> PAGE_SHIFT) + 1 - mmap_locked,
+			&mmap_user->locked_vm);
 	atomic64_sub(mmap_locked, &vma->vm_mm->pinned_vm);
 	free_uid(mmap_user);
 
@@ -5729,8 +5730,20 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma)
 
 	user_locked = atomic_long_read(&user->locked_vm) + user_extra;
 
-	if (user_locked > user_lock_limit)
+	if (user_locked <= user_lock_limit) {
+		/* charge all to locked_vm */
+	} else if (atomic_long_read(&user->locked_vm) >= user_lock_limit) {
+		/* charge all to pinned_vm */
+		extra = user_extra;
+		user_extra = 0;
+	} else {
+		/*
+		 * charge locked_vm until it hits user_lock_limit;
+		 * charge the rest from pinned_vm
+		 */
 		extra = user_locked - user_lock_limit;
+		user_extra -= extra;
+	}
 
 	lock_limit = rlimit(RLIMIT_MEMLOCK);
 	lock_limit >>= PAGE_SHIFT;
-- 
2.20.1