mm: retry page fault when blocking on disk transfer

This change reduces mmap_sem hold times that are caused by waiting for
disk transfers when accessing file mapped VMAs.

It introduces the VM_FAULT_ALLOW_RETRY flag, which indicates that the call
site wants mmap_sem to be released if blocking on a pending disk transfer.
In that case, filemap_fault() returns the VM_FAULT_RETRY status bit and
do_page_fault() will then re-acquire mmap_sem and retry the page fault.

It is expected that the retry will hit the same page which will now be
cached, and thus it will complete with a low mmap_sem hold time.

Tests:

- microbenchmark: thread A mmaps a large file and does random read accesses
  to the mmaped area - achieves about 55 iterations/s. Thread B does
  mmap/munmap in a loop at a separate location - achieves 55 iterations/s
  before, 15000 iterations/s after.

- We are seeing related effects in some applications in house, which show
  significant performance regressions when running without this change.

[akpm@linux-foundation.org: fix warning & crash]
Signed-off-by: Michel Lespinasse <walken@google.com>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Reviewed-by: Wu Fengguang <fengguang.wu@intel.com>
Cc: Ying Han <yinghan@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

2 files changed, 23 insertions, 3 deletions

diff --git a/mm/filemap.c b/mm/filemap.c
index 8ed709a83eb..33f81252a74 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -612,6 +612,19 @@ void __lock_page_nosync(struct page *page)
 							TASK_UNINTERRUPTIBLE);
 }
 
+int __lock_page_or_retry(struct page *page, struct mm_struct *mm,
+			 unsigned int flags)
+{
+	if (!(flags & FAULT_FLAG_ALLOW_RETRY)) {
+		__lock_page(page);
+		return 1;
+	} else {
+		up_read(&mm->mmap_sem);
+		wait_on_page_locked(page);
+		return 0;
+	}
+}
+
 /**
  * find_get_page - find and get a page reference
  * @mapping: the address_space to search
@@ -1550,7 +1563,8 @@ retry_find:
 			goto no_cached_page;
 	}
 
-	lock_page(page);
+	if (!lock_page_or_retry(page, vma->vm_mm, vmf->flags))
+		return ret | VM_FAULT_RETRY;
 
 	/* Did it get truncated? */
 	if (unlikely(page->mapping != mapping)) {
diff --git a/mm/memory.c b/mm/memory.c
index 92cc54e9413..714c4438d88 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2627,6 +2627,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
 	struct page *page, *swapcache = NULL;
 	swp_entry_t entry;
 	pte_t pte;
+	int locked;
 	struct mem_cgroup *ptr = NULL;
 	int exclusive = 0;
 	int ret = 0;
@@ -2677,8 +2678,12 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma,
 		goto out_release;
 	}
 
-	lock_page(page);
+	locked = lock_page_or_retry(page, mm, flags);
 	delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
+	if (!locked) {
+		ret |= VM_FAULT_RETRY;
+		goto out_release;
+	}
 
 	/*
 	 * Make sure try_to_free_swap or reuse_swap_page or swapoff did not
@@ -2927,7 +2932,8 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma,
 	vmf.page = NULL;
 
 	ret = vma->vm_ops->fault(vma, &vmf);
-	if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))
+	if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE |
+			    VM_FAULT_RETRY)))
 		return ret;
 
 	if (unlikely(PageHWPoison(vmf.page))) {