hfsplus: remove ->writepage
authorChristoph Hellwig <hch@lst.de>
Fri, 2 Dec 2022 10:26:41 +0000 (11:26 +0100)
committerAndrew Morton <akpm@linux-foundation.org>
Mon, 12 Dec 2022 02:12:18 +0000 (18:12 -0800)
->writepage is a very inefficient method to write back data, and only
used through write_cache_pages or a a fallback when no ->migrate_folio
method is present.

Set ->migrate_folio to the generic buffer_head based helper, and stop
wiring up ->writepage for hfsplus_aops.

Link: https://lkml.kernel.org/r/20221202102644.770505-5-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
fs/hfsplus/inode.c

index aeab83e..d6572ad 100644 (file)
@@ -170,12 +170,12 @@ const struct address_space_operations hfsplus_aops = {
        .dirty_folio    = block_dirty_folio,
        .invalidate_folio = block_invalidate_folio,
        .read_folio     = hfsplus_read_folio,
-       .writepage      = hfsplus_writepage,
        .write_begin    = hfsplus_write_begin,
        .write_end      = generic_write_end,
        .bmap           = hfsplus_bmap,
        .direct_IO      = hfsplus_direct_IO,
        .writepages     = hfsplus_writepages,
+       .migrate_folio  = buffer_migrate_folio,
 };
 
 const struct dentry_operations hfsplus_dentry_operations = {