Btrfs: use the inode's mapping mask for allocating pages
Johannes pointed out we were allocating only kernel pages for doing writes,
which is kind of a big deal if you are on 32bit and have more than a gig of ram.
So fix our allocations to use the mapping's gfp but still clear __GFP_FS so we
don't re-enter. Thanks,
Reported-by: Johannes Weiner <jweiner@redhat.com>
Signed-off-by: Josef Bacik <josef@redhat.com>
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index de569af..f2e9282 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1069,6 +1069,7 @@
int i;
unsigned long index = pos >> PAGE_CACHE_SHIFT;
struct inode *inode = fdentry(file)->d_inode;
+ gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
int err = 0;
int faili = 0;
u64 start_pos;
@@ -1080,7 +1081,7 @@
again:
for (i = 0; i < num_pages; i++) {
pages[i] = find_or_create_page(inode->i_mapping, index + i,
- GFP_NOFS);
+ mask);
if (!pages[i]) {
faili = i - 1;
err = -ENOMEM;
