[PATCH] remap_file_pages() speedup

Avoid taking down_write(mmap_sem) unless we really need it.

Seems that the only reason we're taking it for writing is to protect
vma->vm_flags.
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

mm/fremap.c

@@ -161,6 +161,7 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size,
 	unsigned long end = start + size;
 	struct vm_area_struct *vma;
 	int err = -EINVAL;
+	int has_write_lock = 0;
 
 	if (__prot)
 		return err;
@@ -181,7 +182,8 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size,
 #endif
 
 	/* We need down_write() to change vma->vm_flags. */
-	down_write(&mm->mmap_sem);
+	down_read(&mm->mmap_sem);
+ retry:
 	vma = find_vma(mm, start);
 
 	/*
@@ -200,6 +202,12 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size,
 		/* Must set VM_NONLINEAR before any pages are populated. */
 		if (pgoff != linear_page_index(vma, start) &&
 		    !(vma->vm_flags & VM_NONLINEAR)) {
+			if (!has_write_lock) {
+				up_read(&mm->mmap_sem);
+				down_write(&mm->mmap_sem);
+				has_write_lock = 1;
+				goto retry;
+			}
 			mapping = vma->vm_file->f_mapping;
 			spin_lock(&mapping->i_mmap_lock);
 			flush_dcache_mmap_lock(mapping);
@@ -212,8 +220,6 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size,
 			spin_unlock(&mapping->i_mmap_lock);
 		}
 
-		/* ->populate can take a long time, so downgrade the lock. */
-		downgrade_write(&mm->mmap_sem);
 		err = vma->vm_ops->populate(vma, start, size,
 					    vma->vm_page_prot,
 					    pgoff, flags & MAP_NONBLOCK);
@@ -223,10 +229,11 @@ asmlinkage long sys_remap_file_pages(unsigned long start, unsigned long size,
 		 * it after ->populate completes, and that would prevent
 		 * downgrading the lock.  (Locks can't be upgraded).
 		 */
+	}
+	if (likely(!has_write_lock))
 		up_read(&mm->mmap_sem);
-	} else {
+	else
 		up_write(&mm->mmap_sem);
-	}
 
 	return err;
 }