d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix a race in vm_page_dequeue_deferred().

Browse Source 
To detect the case where the page is already marked for a deferred
dequeue, we must read the "queue" and "aflags" fields in a
precise order.  Otherwise, a race with a concurrent
vm_page_dequeue_complete() could leave the page with PGA_DEQUEUE
set despite it already having been dequeued.  Fix the problem by
using vm_page_queue() to check the queue state, which correctly
handles the race.

Reviewed by:	kib
Tested by:	pho
MFC after:	3 days
Sponsored by:	Netflix
Differential Revision:	https://reviews.freebsd.org/D19039
This commit is contained in:
Mark Johnston 2019-02-03 18:38:58 +00:00
parent 634a8a8873
commit d0488e698f
1 changed files with 4 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
sys/vm/vm_page.c
View File

@ -3279,18 +3279,13 @@ vm_page_dequeue_complete(vm_page_t m)
void
vm_page_dequeue_deferred(vm_page_t m)
{
int queue;
uint8_t queue;
vm_page_assert_locked(m);
queue = atomic_load_8(&m->queue);
if (queue == PQ_NONE) {
KASSERT((m->aflags & PGA_QUEUE_STATE_MASK) == 0,
("page %p has queue state", m));
if ((queue = vm_page_queue(m)) == PQ_NONE)
return;
}
if ((m->aflags & PGA_DEQUEUE) == 0)
vm_page_aflag_set(m, PGA_DEQUEUE);
vm_page_aflag_set(m, PGA_DEQUEUE);
vm_pqbatch_submit_page(m, queue);
}
@ -3386,7 +3381,7 @@ vm_page_requeue(vm_page_t m)
{
vm_page_assert_locked(m);
KASSERT(m->queue != PQ_NONE,
KASSERT(vm_page_queue(m) != PQ_NONE,
("%s: page %p is not logically enqueued", __func__, m));
if ((m->aflags & PGA_REQUEUE) == 0)