nfs: page group syncing in write path
authorWeston Andros Adamson <dros@primarydata.com>
Thu, 15 May 2014 15:56:47 +0000 (11:56 -0400)
committerTrond Myklebust <trond.myklebust@primarydata.com>
Thu, 29 May 2014 15:11:45 +0000 (11:11 -0400)
commit20633f042fd0907300069714b98aaf607a8b5bf8
tree33f23c3c0043376a6636367ec64c2962131c7dd9
parent67d0338edd71db9a4f406d8778f7c525d31e9f7f
nfs: page group syncing in write path

Operations that modify state for a whole page must be syncronized across
all requests within a page group. In the write path, this is calling
end_page_writeback and removing the head request from an inode.
Both of these operations should not be called until all requests
in a page group have reached the point where they would call them.

This patch should have no effect yet since all page groups currently
have one request, but will come into play when pg_test functions are
modified to split pages into sub-page regions.

Signed-off-by: Weston Andros Adamson <dros@primarydata.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
fs/nfs/pagelist.c
fs/nfs/write.c
include/linux/nfs_page.h