[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20181007232736.3780-20-martin@omnibond.com>
Date: Sun, 7 Oct 2018 23:27:36 +0000
From: Martin Brandenburg <martin@...ibond.com>
To: devel@...ts.orangefs.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, hubcap@...ibond.com
Cc: Martin Brandenburg <martin@...ibond.com>
Subject: [PATCH 19/19] orangefs: do writepages_work if a single page must be written
Otherwise the next page can't possibly be an append and it'll
just sit there and write pages one by one until it flushes the
saved region at the very end.
Signed-off-by: Martin Brandenburg <martin@...ibond.com>
---
fs/orangefs/inode.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 20950f3f758a..cd1263c45bb2 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -307,6 +307,10 @@ static int orangefs_writepages_callback(struct page *page,
wr = (struct orangefs_write_request *)page_private(page);
if (wr->len != PAGE_SIZE) {
+ if (ow->npages) {
+ orangefs_writepages_work(ow, wbc);
+ ow->npages = 0;
+ }
ret = orangefs_writepage_locked(page, wbc);
mapping_set_error(page->mapping, ret);
unlock_page(page);
@@ -335,6 +339,10 @@ static int orangefs_writepages_callback(struct page *page,
}
done:
if (ret == -1) {
+ if (ow->npages) {
+ orangefs_writepages_work(ow, wbc);
+ ow->npages = 0;
+ }
ret = orangefs_writepage_locked(page, wbc);
mapping_set_error(page->mapping, ret);
unlock_page(page);
--
2.19.0
Powered by blists - more mailing lists