mm: compaction: do not mark unmovable pageblocks as skipped in async compaction

[firefly-linux-kernel-4.4.55.git] / mm / mlock.c

diff --git a/mm/mlock.c b/mm/mlock.c
index c59c420fd6e16e81c10d0524dd980f38c09b2afa..10819ed4df3e6b34846c01e5cd4de84d5a33fe62 100644 (file)
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -298,10 +298,12 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
 {
        int i;
        int nr = pagevec_count(pvec);
-       int delta_munlocked = -nr;
+       int delta_munlocked;
        struct pagevec pvec_putback;
        int pgrescued = 0;
 
+       pagevec_init(&pvec_putback, 0);
+
        /* Phase 1: page isolation */
        spin_lock_irq(&zone->lru_lock);
        for (i = 0; i < nr; i++) {
@@ -330,18 +332,21 @@ skip_munlock:
                        /*
                         * We won't be munlocking this page in the next phase
                         * but we still need to release the follow_page_mask()
-                        * pin.
+                        * pin. We cannot do it under lru_lock however. If it's
+                        * the last pin, __page_cache_release would deadlock.
                         */
+                       pagevec_add(&pvec_putback, pvec->pages[i]);
                        pvec->pages[i] = NULL;
-                       put_page(page);
-                       delta_munlocked++;
                }
        }
+       delta_munlocked = -nr + pagevec_count(&pvec_putback);
        __mod_zone_page_state(zone, NR_MLOCK, delta_munlocked);
        spin_unlock_irq(&zone->lru_lock);
 
+       /* Now we can release pins of pages that we are not munlocking */
+       pagevec_release(&pvec_putback);
+
        /* Phase 2: page munlock */
-       pagevec_init(&pvec_putback, 0);
        for (i = 0; i < nr; i++) {
                struct page *page = pvec->pages[i];
 
@@ -704,19 +709,21 @@ SYSCALL_DEFINE2(mlock, unsigned long, start, size_t, len)
 
        lru_add_drain_all();    /* flush pagevec */
 
-       down_write(&current->mm->mmap_sem);
        len = PAGE_ALIGN(len + (start & ~PAGE_MASK));
        start &= PAGE_MASK;
 
-       locked = len >> PAGE_SHIFT;
-       locked += current->mm->locked_vm;
-
        lock_limit = rlimit(RLIMIT_MEMLOCK);
        lock_limit >>= PAGE_SHIFT;
+       locked = len >> PAGE_SHIFT;
+
+       down_write(&current->mm->mmap_sem);
+
+       locked += current->mm->locked_vm;
 
        /* check against resource limits */
        if ((locked <= lock_limit) || capable(CAP_IPC_LOCK))
                error = do_mlock(start, len, 1);
+
        up_write(&current->mm->mmap_sem);
        if (!error)
                error = __mm_populate(start, len, 0);
@@ -727,11 +734,13 @@ SYSCALL_DEFINE2(munlock, unsigned long, start, size_t, len)
 {
        int ret;
 
-       down_write(&current->mm->mmap_sem);
        len = PAGE_ALIGN(len + (start & ~PAGE_MASK));
        start &= PAGE_MASK;
+
+       down_write(&current->mm->mmap_sem);
        ret = do_mlock(start, len, 0);
        up_write(&current->mm->mmap_sem);
+
        return ret;
 }
 
@@ -776,12 +785,12 @@ SYSCALL_DEFINE1(mlockall, int, flags)
        if (flags & MCL_CURRENT)
                lru_add_drain_all();    /* flush pagevec */
 
-       down_write(&current->mm->mmap_sem);
-
        lock_limit = rlimit(RLIMIT_MEMLOCK);
        lock_limit >>= PAGE_SHIFT;
 
        ret = -ENOMEM;
+       down_write(&current->mm->mmap_sem);
+
        if (!(flags & MCL_CURRENT) || (current->mm->total_vm <= lock_limit) ||
            capable(CAP_IPC_LOCK))
                ret = do_mlockall(flags);