From: David Howells <dhowells@redhat.com>
To: linux-fsdevel@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, linux-cachefs@redhat.com,
linux-nfs@vger.kernel.org
Subject: [PATCH 00/17] Fix assorted FS-Cache problems
Date: Wed, 08 Feb 2012 21:16:40 +0000 [thread overview]
Message-ID: <20120208211640.15607.58537.stgit@warthog.procyon.org.uk> (raw)
Following the mail is a series of patches to fix a number of FS-Cache problems,
including a number of oopses. The patches can also be found at:
http://git.kernel.org/?p=linux/kernel/git/dhowells/linux-fscache.git;a=shortlog;h=refs/heads/fixes
The patches are as follows:
(*) A patch to correctly mark cached netfs pages. This fixes an oops that
shows up in the page allocator with PG_private_2 still being set on pages.
(*) A patch to downgrade memory allocation levels in the cache to not try so
hard and to be more willing to abort with ENOMEM. It's a cache - it
doesn't matter if we can't store something.
(*) A patch to check that there are no read operations outstanding on a
cookie when it is relinquished.
(*) A patch to make conditional some debugging prints.
(*) A patch to make cookie relinquishment log a warning and wait for any
outstanding reads.
(*) A patch to fix up operation state handling and accounting. This fixes a
BUG in __fscache_relinquish_cookie() that occurs when cache operation
completion gets out of sync with the netfs releasing a cookie back to
fscache.
(*) Four patches to provide proper invalidation facilities so that the netfs
can request a cache object be invalidated rather than retiring the cookie
it has and then getting a new one.
The first patch provides the hooks and management in the FS-Cache module.
The second patch splits do_sys_truncate() to create a vfs_truncate() for
cachefiles's invalidation to call.
The third patch makes cachefiles honour invalidation requests.
The fourth patch makes NFS use the invalidation call.
These fix a BUG() in __fscache_read_or_alloc_pages() that occurs when an
object that has operations outstanding on it gets its cookie removed.
(*) Add some missing calls to note completion of retrieval operations in
cachefiles. These cause an assertion failure BUG in
fscache_put_operation() when it tries to destroy an operation that is
still live.
(*) Three patches to clean up object event IDs in FS-Cache and to stop masking
off bits that needn't be masked off.
(*) A patch to limit the number of I/O errors reported on a cache to 1. This
prevents a lot of noise being generated from a busy cache when it is
aborted.
(*) A patch to handle an unexpected state in exclusive op submission in
FS-Cache. This fixes a BUG that occurs in fscache_submit_exclusive_op()
if it sees the cache in an aborted state due to an I/O error having
occurred.
(*) A patch to make nfs_migrate_page() wait for FS-Cache to finish with a
page. This fixes a bad page state BUG in the memory allocator due to NFS
trying to migrate a page that FS-Cache is doing stuff to.
David
---
David Howells (17):
NFS: nfs_migrate_page() does not wait for FS-Cache to finish with a page
FS-Cache: Exclusive op submission can BUG if there's been an I/O error
FS-Cache: Limit the number of I/O error reports for a cache
FS-Cache: Don't mask off the object event mask when printing it
FS-Cache: Initialise the object event mask with the calculated mask
FS-Cache: Convert the object event ID #defines into an enum
CacheFiles: Add missing retrieval completions
NFS: Use FS-Cache invalidation
CacheFiles: Implement invalidation
VFS: Make more complete truncate operation available to CacheFiles
FS-Cache: Provide proper invalidation
FS-Cache: Fix operation state management and accounting
FS-Cache: Make cookie relinquishment wait for outstanding reads
CacheFiles: Make some debugging statements conditional
FS-Cache: Check that there are no read ops when cookie relinquished
CacheFiles: Downgrade the requirements passed to the allocator
CacheFiles: Fix the marking of cached pages
Documentation/filesystems/caching/backend-api.txt | 38 +++++
Documentation/filesystems/caching/netfs-api.txt | 46 +++++-
Documentation/filesystems/caching/object.txt | 23 ++-
Documentation/filesystems/caching/operations.txt | 2
fs/cachefiles/interface.c | 57 +++++++-
fs/cachefiles/internal.h | 2
fs/cachefiles/key.c | 2
fs/cachefiles/namei.c | 3
fs/cachefiles/rdwr.c | 114 +++++++++------
fs/cachefiles/xattr.c | 2
fs/fscache/cache.c | 8 +
fs/fscache/cookie.c | 78 ++++++++++
fs/fscache/internal.h | 12 ++
fs/fscache/object-list.c | 2
fs/fscache/object.c | 98 ++++++++++++-
fs/fscache/operation.c | 134 ++++++++++++++----
fs/fscache/page.c | 158 +++++++++++++++++----
fs/fscache/stats.c | 17 ++
fs/nfs/fscache.h | 20 +++
fs/nfs/inode.c | 20 ++-
fs/nfs/nfs4proc.c | 2
fs/nfs/write.c | 4 -
fs/open.c | 50 ++++---
include/linux/fs.h | 1
include/linux/fscache-cache.h | 71 ++++++++-
include/linux/fscache.h | 50 ++++++-
26 files changed, 813 insertions(+), 201 deletions(-)
next reply other threads:[~2012-02-08 21:16 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-02-08 21:16 David Howells [this message]
2012-02-08 21:16 ` [PATCH 01/17] CacheFiles: Fix the marking of cached pages David Howells
2012-02-08 21:17 ` [PATCH 02/17] CacheFiles: Downgrade the requirements passed to the allocator David Howells
2012-02-08 21:17 ` [PATCH 03/17] FS-Cache: Check that there are no read ops when cookie relinquished David Howells
2012-02-08 21:17 ` [PATCH 04/17] CacheFiles: Make some debugging statements conditional David Howells
2012-02-08 21:17 ` [PATCH 05/17] FS-Cache: Make cookie relinquishment wait for outstanding reads David Howells
2012-02-08 21:17 ` [PATCH 06/17] FS-Cache: Fix operation state management and accounting David Howells
2012-02-08 21:17 ` [PATCH 07/17] FS-Cache: Provide proper invalidation David Howells
2012-02-08 21:18 ` [PATCH 08/17] VFS: Make more complete truncate operation available to CacheFiles David Howells
2012-02-08 21:18 ` [PATCH 09/17] CacheFiles: Implement invalidation David Howells
2012-02-08 21:18 ` [PATCH 10/17] NFS: Use FS-Cache invalidation David Howells
2012-02-08 21:18 ` [PATCH 11/17] CacheFiles: Add missing retrieval completions David Howells
2012-02-08 21:18 ` [PATCH 12/17] FS-Cache: Convert the object event ID #defines into an enum David Howells
2012-02-08 21:18 ` [PATCH 13/17] FS-Cache: Initialise the object event mask with the calculated mask David Howells
2012-02-08 21:19 ` [PATCH 14/17] FS-Cache: Don't mask off the object event mask when printing it David Howells
2012-02-08 21:19 ` [PATCH 15/17] FS-Cache: Limit the number of I/O error reports for a cache David Howells
2012-02-08 21:19 ` [PATCH 16/17] FS-Cache: Exclusive op submission can BUG if there's been an I/O error David Howells
2012-02-08 21:19 ` [PATCH 17/17] NFS: nfs_migrate_page() does not wait for FS-Cache to finish with a page David Howells
2012-07-21 16:57 ` Jonathan Nieder
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20120208211640.15607.58537.stgit@warthog.procyon.org.uk \
--to=dhowells@redhat.com \
--cc=linux-cachefs@redhat.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).