[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <01c442d2b3aff85e0e95bfefb44ac6e77eb3373f.camel@redhat.com>
Date: Tue, 24 Aug 2021 10:25:20 -0400
From: Jeff Layton <jlayton@...hat.com>
To: David Howells <dhowells@...hat.com>, linux-cachefs@...hat.com
Cc: Anna Schumaker <anna.schumaker@...app.com>,
Steve French <sfrench@...ba.org>,
Dominique Martinet <asmadeus@...ewreck.org>,
David Wysochanski <dwysocha@...hat.com>,
linux-afs@...ts.infradead.org, linux-nfs@...r.kernel.org,
linux-cifs@...r.kernel.org, ceph-devel@...r.kernel.org,
v9fs-developer@...ts.sourceforge.net,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 00/12] fscache: Some prep work for fscache rewrite
On Mon, 2021-06-21 at 22:44 +0100, David Howells wrote:
> Here are some patches that perform some preparatory work for the fscache
> rewrite that's being worked on. These include:
>
> (1) Always select netfs stats when enabling fscache stats since they're
> displayed through the same procfile.
>
> (2) Add a cookie debug ID that can be used in tracepoints instead of a
> pointer and cache it in the netfs_cache_resources struct rather than
> in the netfs_read_request struct to make it more available.
>
> (3) Use file_inode() in cachefiles rather than dereferencing file->f_inode
> directly.
>
> (4) Provide a procfile to display fscache cookies.
>
> (5) Remove the fscache and cachefiles histogram procfiles.
>
> (6) Remove the fscache object list procfile.
>
> (7) Avoid using %p in fscache and cachefiles as the value is hashed and
> not comparable to the register dump in an oops trace.
>
> (8) Fix the cookie hash function to actually achieve useful dispersion.
>
> (9) Fix fscache_cookie_put() so that it doesn't dereference the cookie
> pointer in the tracepoint after the refcount has been decremented
> (we're only allowed to do that if we decremented it to zero).
>
> (10) Use refcount_t rather than atomic_t for the fscache_cookie refcount.
>
> The patches can be found on this branch:
>
> http://git.kernel.org/cgit/linux/kernel/git/dhowells/linux-fs.git/log/?h=fscache-next
>
> David
> ---
> David Howells (12):
> fscache: Select netfs stats if fscache stats are enabled
> netfs: Move cookie debug ID to struct netfs_cache_resources
> cachefiles: Use file_inode() rather than accessing ->f_inode
> fscache: Add a cookie debug ID and use that in traces
> fscache: Procfile to display cookies
> fscache, cachefiles: Remove the histogram stuff
> fscache: Remove the object list procfile
> fscache: Change %p in format strings to something else
> cachefiles: Change %p in format strings to something else
> fscache: Fix cookie key hashing
> fscache: Fix fscache_cookie_put() to not deref after dec
> fscache: Use refcount_t for the cookie refcount instead of atomic_t
>
>
> fs/cachefiles/Kconfig | 19 --
> fs/cachefiles/Makefile | 2 -
> fs/cachefiles/bind.c | 2 -
> fs/cachefiles/interface.c | 6 +-
> fs/cachefiles/internal.h | 25 --
> fs/cachefiles/io.c | 6 +-
> fs/cachefiles/key.c | 2 +-
> fs/cachefiles/main.c | 7 -
> fs/cachefiles/namei.c | 61 ++---
> fs/cachefiles/proc.c | 114 --------
> fs/cachefiles/xattr.c | 4 +-
> fs/fscache/Kconfig | 24 --
> fs/fscache/Makefile | 2 -
> fs/fscache/cache.c | 11 +-
> fs/fscache/cookie.c | 201 +++++++++++----
> fs/fscache/fsdef.c | 3 +-
> fs/fscache/histogram.c | 87 -------
> fs/fscache/internal.h | 57 +---
> fs/fscache/main.c | 39 +++
> fs/fscache/netfs.c | 2 +-
> fs/fscache/object-list.c | 414 ------------------------------
> fs/fscache/object.c | 8 -
> fs/fscache/operation.c | 3 -
> fs/fscache/page.c | 6 -
> fs/fscache/proc.c | 20 +-
> include/linux/fscache-cache.h | 4 -
> include/linux/fscache.h | 4 +-
> include/linux/netfs.h | 2 +-
> include/trace/events/cachefiles.h | 68 ++---
> include/trace/events/fscache.h | 160 ++++++------
> include/trace/events/netfs.h | 2 +-
> 31 files changed, 367 insertions(+), 998 deletions(-)
> delete mode 100644 fs/cachefiles/proc.c
> delete mode 100644 fs/fscache/histogram.c
> delete mode 100644 fs/fscache/object-list.c
>
>
This all looks good (modulo a nitpicky changelog comment). You can add:
Reviewed-by: Jeff Layton <jlayton@...hat.com>
Powered by blists - more mailing lists