* [PATCH v3 00/18] Allow readpage to return a locked page
@ 2020-10-16 16:04 Matthew Wilcox (Oracle)
2020-10-16 16:04 ` [PATCH v3 08/18] cifs: Tell the VFS that readpage was synchronous Matthew Wilcox (Oracle)
0 siblings, 1 reply; 2+ messages in thread
From: Matthew Wilcox (Oracle) @ 2020-10-16 16:04 UTC (permalink / raw)
To: linux-fsdevel
Cc: Matthew Wilcox (Oracle),
linux-mm, David Howells, Steve French, linux-cifs, Nicolas Pitre,
Tyler Hicks, ecryptfs, Theodore Ts'o, Andreas Dilger,
linux-ext4, Miklos Szeredi, Hans de Goede
I've dropped the conversion of readpage implementations to synchronous
from this patchset. I realised I'd neglected the requirement for making
the sleep killable, and that turns out to be more convoluted to fix.
So these patches add:
- An error-path bugfix for cachefiles.
- The ability for the filesystem to tell the caller of ->readpage
that the read was successful and the page was not unlocked. This is
a performance improvement for some scenarios that I think are rare.
- Mildly improved error handling for ext4.
v2: https://lore.kernel.org/linux-fsdevel/20201009143104.22673-1-willy@infradead.org/
v1: https://lore.kernel.org/linux-fsdevel/20200917151050.5363-1-willy@infradead.org/
Matthew Wilcox (Oracle) (18):
cachefiles: Handle readpage error correctly
swap: Call aops->readahead if appropriate
fs: Add AOP_UPDATED_PAGE return value
mm/filemap: Inline wait_on_page_read into its one caller
9p: Tell the VFS that readpage was synchronous
afs: Tell the VFS that readpage was synchronous
ceph: Tell the VFS that readpage was synchronous
cifs: Tell the VFS that readpage was synchronous
cramfs: Tell the VFS that readpage was synchronous
ecryptfs: Tell the VFS that readpage was synchronous
ext4: Tell the VFS that readpage was synchronous
ext4: Return error from ext4_readpage
fuse: Tell the VFS that readpage was synchronous
hostfs: Tell the VFS that readpage was synchronous
jffs2: Tell the VFS that readpage was synchronous
ubifs: Tell the VFS that readpage was synchronous
udf: Tell the VFS that readpage was synchronous
vboxsf: Tell the VFS that readpage was synchronous
Documentation/filesystems/locking.rst | 7 +++---
Documentation/filesystems/vfs.rst | 21 +++++++++++------
fs/9p/vfs_addr.c | 6 ++++-
fs/afs/file.c | 3 ++-
fs/buffer.c | 15 +++++++-----
fs/cachefiles/rdwr.c | 9 ++++++++
fs/ceph/addr.c | 9 ++++----
fs/cifs/file.c | 8 +++++--
fs/cramfs/inode.c | 5 ++--
fs/ecryptfs/mmap.c | 11 +++++----
fs/ext4/inline.c | 9 +++++---
fs/ext4/readpage.c | 24 +++++++++++--------
fs/fuse/file.c | 2 ++
fs/hostfs/hostfs_kern.c | 2 ++
fs/jffs2/file.c | 6 +++--
fs/ubifs/file.c | 16 ++++++++-----
fs/udf/file.c | 3 +--
fs/vboxsf/file.c | 2 ++
include/linux/fs.h | 5 ++++
mm/filemap.c | 33 +++++++++++++--------------
mm/page_io.c | 27 ++++++++++++++++++++--
mm/readahead.c | 3 ++-
22 files changed, 151 insertions(+), 75 deletions(-)
--
2.28.0
^ permalink raw reply [flat|nested] 2+ messages in thread
* [PATCH v3 08/18] cifs: Tell the VFS that readpage was synchronous
2020-10-16 16:04 [PATCH v3 00/18] Allow readpage to return a locked page Matthew Wilcox (Oracle)
@ 2020-10-16 16:04 ` Matthew Wilcox (Oracle)
0 siblings, 0 replies; 2+ messages in thread
From: Matthew Wilcox (Oracle) @ 2020-10-16 16:04 UTC (permalink / raw)
To: linux-fsdevel; +Cc: Matthew Wilcox (Oracle), linux-mm, Steve French, linux-cifs
The cifs readpage implementation was already synchronous, so use
AOP_UPDATED_PAGE to avoid cycling the page lock.
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
fs/cifs/file.c | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index be46fab4c96d..533b151a9143 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -4537,7 +4537,8 @@ static int cifs_readpage_worker(struct file *file, struct page *page,
/* send this page to the cache */
cifs_readpage_to_fscache(file_inode(file), page);
- rc = 0;
+ kunmap(page);
+ return AOP_UPDATED_PAGE;
io_error:
kunmap(page);
@@ -4677,7 +4678,10 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping,
* an error, we don't need to return it. cifs_write_end will
* do a sync write instead since PG_uptodate isn't set.
*/
- cifs_readpage_worker(file, page, &page_start);
+ int err = cifs_readpage_worker(file, page, &page_start);
+
+ if (err == AOP_UPDATED_PAGE)
+ goto out;
put_page(page);
oncethru = 1;
goto start;
--
2.28.0
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-10-16 16:05 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-16 16:04 [PATCH v3 00/18] Allow readpage to return a locked page Matthew Wilcox (Oracle)
2020-10-16 16:04 ` [PATCH v3 08/18] cifs: Tell the VFS that readpage was synchronous Matthew Wilcox (Oracle)
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).