From: Jeff Layton <jlayton-vpEMnDpepFuMZCB2o+C8xQ@public.gmane.org> To: bfields-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org Cc: linux-nfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Al Viro <viro-3bDd1+5oDREiFSDQTTA3OLVCufUGDwFn@public.gmane.org> Subject: [PATCH v6 00/19] nfsd: open file caching Date: Tue, 20 Oct 2015 13:33:33 -0400 [thread overview] Message-ID: <1445362432-18869-1-git-send-email-jeff.layton@primarydata.com> (raw) v6: - rename delayed_fput infrastructure to global_fput - drop list_lru_rotate patch, rework LRU list code to use LRU_ROTATE and a new NFSD_FILE_REFERENCED flag - rework fsnotify_mark handling to be done with separate allocation - bugfixes v5: - switch to using flush_delayed_fput instead of __fput_sync - hash on inode->i_ino instead of inode pointer - add /proc/fs/nfsd/file_cache_stats file to track stats on the hash - eliminate extra fh_verify in nfsd_file_acquire v4: - squash some of the patches down into one patch to reduce churn - close cached open files after unlink instead of before - don't just close files after nfsd does an unlink, must do it after any vfs-layer unlink. Use fsnotify to handle that. - use a SRCU notifier chain for setlease - add patch to allow non-kthreads to do a fput_sync v3: - open files are now hashed on inode pointer instead of fh - eliminate the recurring workqueue job in favor of shrinker/LRU and notifier from lease setting code - have nfsv4 use the cache as well - removal of raparms cache v2: - changelog cleanups and clarifications - allow COMMIT to use cached open files - tracepoints for nfsd_file cache - proactively close open files prior to REMOVE, or a RENAME over a positive dentry This is the sixth posting of this patchset. For those just tuning in now, the basic idea here is to allow nfsd to keep a cache of open file descriptions, primarily to help NFSv3 workloads but also to clean up some nastiness in the NFSv4 file handling. Of course, we can't keep files open indefinitely, so much of the new infrastructure is geared toward allowing the files to be closed on demand, in the following cases: - nfsd_files are released whenever the inode's link count goes to zero - they are released when userspace wants to set a lease on the inode - when a shrinker callback occurs for the nfsd_file cache - when filesystems are unexported or nfsd is shut down, we clean out the cache This allows nfsd to keep the file open basically indefinitely. The tricky part is the setlease case. To handle that properly we have to allow userland processes to queue the fputs involved to a workqueue. This means allowing non-kthreads to queue fputs to the delayed_fput infrastructure. Al, can you look and weigh in? Do the delayed_fput/global_fput changes look reasonable? Jeff Layton (19): nfsd: move include of state.h from trace.c to trace.h fs: have flush_delayed_fput flush the workqueue job fs: add a kerneldoc header to fput fs: rename "delayed_fput" infrastructure to "fput_global" fs: add fput_global fsnotify: export several symbols locks: create a new notifier chain for lease attempts sunrpc: add a new cache_detail operation for when a cache is flushed nfsd: add a new struct file caching facility to nfsd nfsd: keep some rudimentary stats on nfsd_file cache nfsd: allow filecache open to skip fh_verify check nfsd: hook up nfsd_write to the new nfsd_file cache nfsd: hook up nfsd_read to the nfsd_file cache nfsd: hook nfsd_commit up to the nfsd_file cache nfsd: convert nfs4_file->fi_fds array to use nfsd_files nfsd: have nfsd_test_lock use the nfsd_file cache nfsd: convert fi_deleg_file and ls_file fields to nfsd_file nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache nfsd: rip out the raparms cache fs/file_table.c | 94 ++++-- fs/locks.c | 37 +++ fs/nfsd/Kconfig | 2 + fs/nfsd/Makefile | 3 +- fs/nfsd/export.c | 14 + fs/nfsd/filecache.c | 712 +++++++++++++++++++++++++++++++++++++++++++ fs/nfsd/filecache.h | 44 +++ fs/nfsd/nfs3proc.c | 2 +- fs/nfsd/nfs4layouts.c | 12 +- fs/nfsd/nfs4proc.c | 32 +- fs/nfsd/nfs4state.c | 174 +++++------ fs/nfsd/nfs4xdr.c | 16 +- fs/nfsd/nfsctl.c | 10 + fs/nfsd/nfsproc.c | 2 +- fs/nfsd/nfssvc.c | 16 +- fs/nfsd/state.h | 10 +- fs/nfsd/trace.c | 2 - fs/nfsd/trace.h | 137 +++++++++ fs/nfsd/vfs.c | 269 ++++------------ fs/nfsd/vfs.h | 11 +- fs/nfsd/xdr4.h | 15 +- fs/notify/group.c | 2 + fs/notify/inode_mark.c | 1 + fs/notify/mark.c | 4 + include/linux/file.h | 3 +- include/linux/fs.h | 1 + include/linux/sunrpc/cache.h | 1 + init/main.c | 2 +- net/sunrpc/cache.c | 3 + 29 files changed, 1249 insertions(+), 382 deletions(-) create mode 100644 fs/nfsd/filecache.c create mode 100644 fs/nfsd/filecache.h -- 2.4.3 -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org More majordomo info at http://vger.kernel.org/majordomo-info.html
WARNING: multiple messages have this Message-ID (diff)
From: Jeff Layton <jlayton@poochiereds.net> To: bfields@fieldses.org Cc: linux-nfs@vger.kernel.org, linux-fsdevel@vger.kernel.org, Al Viro <viro@ZenIV.linux.org.uk> Subject: [PATCH v6 00/19] nfsd: open file caching Date: Tue, 20 Oct 2015 13:33:33 -0400 [thread overview] Message-ID: <1445362432-18869-1-git-send-email-jeff.layton@primarydata.com> (raw) v6: - rename delayed_fput infrastructure to global_fput - drop list_lru_rotate patch, rework LRU list code to use LRU_ROTATE and a new NFSD_FILE_REFERENCED flag - rework fsnotify_mark handling to be done with separate allocation - bugfixes v5: - switch to using flush_delayed_fput instead of __fput_sync - hash on inode->i_ino instead of inode pointer - add /proc/fs/nfsd/file_cache_stats file to track stats on the hash - eliminate extra fh_verify in nfsd_file_acquire v4: - squash some of the patches down into one patch to reduce churn - close cached open files after unlink instead of before - don't just close files after nfsd does an unlink, must do it after any vfs-layer unlink. Use fsnotify to handle that. - use a SRCU notifier chain for setlease - add patch to allow non-kthreads to do a fput_sync v3: - open files are now hashed on inode pointer instead of fh - eliminate the recurring workqueue job in favor of shrinker/LRU and notifier from lease setting code - have nfsv4 use the cache as well - removal of raparms cache v2: - changelog cleanups and clarifications - allow COMMIT to use cached open files - tracepoints for nfsd_file cache - proactively close open files prior to REMOVE, or a RENAME over a positive dentry This is the sixth posting of this patchset. For those just tuning in now, the basic idea here is to allow nfsd to keep a cache of open file descriptions, primarily to help NFSv3 workloads but also to clean up some nastiness in the NFSv4 file handling. Of course, we can't keep files open indefinitely, so much of the new infrastructure is geared toward allowing the files to be closed on demand, in the following cases: - nfsd_files are released whenever the inode's link count goes to zero - they are released when userspace wants to set a lease on the inode - when a shrinker callback occurs for the nfsd_file cache - when filesystems are unexported or nfsd is shut down, we clean out the cache This allows nfsd to keep the file open basically indefinitely. The tricky part is the setlease case. To handle that properly we have to allow userland processes to queue the fputs involved to a workqueue. This means allowing non-kthreads to queue fputs to the delayed_fput infrastructure. Al, can you look and weigh in? Do the delayed_fput/global_fput changes look reasonable? Jeff Layton (19): nfsd: move include of state.h from trace.c to trace.h fs: have flush_delayed_fput flush the workqueue job fs: add a kerneldoc header to fput fs: rename "delayed_fput" infrastructure to "fput_global" fs: add fput_global fsnotify: export several symbols locks: create a new notifier chain for lease attempts sunrpc: add a new cache_detail operation for when a cache is flushed nfsd: add a new struct file caching facility to nfsd nfsd: keep some rudimentary stats on nfsd_file cache nfsd: allow filecache open to skip fh_verify check nfsd: hook up nfsd_write to the new nfsd_file cache nfsd: hook up nfsd_read to the nfsd_file cache nfsd: hook nfsd_commit up to the nfsd_file cache nfsd: convert nfs4_file->fi_fds array to use nfsd_files nfsd: have nfsd_test_lock use the nfsd_file cache nfsd: convert fi_deleg_file and ls_file fields to nfsd_file nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache nfsd: rip out the raparms cache fs/file_table.c | 94 ++++-- fs/locks.c | 37 +++ fs/nfsd/Kconfig | 2 + fs/nfsd/Makefile | 3 +- fs/nfsd/export.c | 14 + fs/nfsd/filecache.c | 712 +++++++++++++++++++++++++++++++++++++++++++ fs/nfsd/filecache.h | 44 +++ fs/nfsd/nfs3proc.c | 2 +- fs/nfsd/nfs4layouts.c | 12 +- fs/nfsd/nfs4proc.c | 32 +- fs/nfsd/nfs4state.c | 174 +++++------ fs/nfsd/nfs4xdr.c | 16 +- fs/nfsd/nfsctl.c | 10 + fs/nfsd/nfsproc.c | 2 +- fs/nfsd/nfssvc.c | 16 +- fs/nfsd/state.h | 10 +- fs/nfsd/trace.c | 2 - fs/nfsd/trace.h | 137 +++++++++ fs/nfsd/vfs.c | 269 ++++------------ fs/nfsd/vfs.h | 11 +- fs/nfsd/xdr4.h | 15 +- fs/notify/group.c | 2 + fs/notify/inode_mark.c | 1 + fs/notify/mark.c | 4 + include/linux/file.h | 3 +- include/linux/fs.h | 1 + include/linux/sunrpc/cache.h | 1 + init/main.c | 2 +- net/sunrpc/cache.c | 3 + 29 files changed, 1249 insertions(+), 382 deletions(-) create mode 100644 fs/nfsd/filecache.c create mode 100644 fs/nfsd/filecache.h -- 2.4.3
next reply other threads:[~2015-10-20 17:33 UTC|newest] Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top 2015-10-20 17:33 Jeff Layton [this message] 2015-10-20 17:33 ` [PATCH v6 00/19] nfsd: open file caching Jeff Layton 2015-10-20 17:33 ` [PATCH v6 03/19] fs: add a kerneldoc header to fput Jeff Layton 2015-10-20 17:33 ` [PATCH v6 04/19] fs: rename "delayed_fput" infrastructure to "fput_global" Jeff Layton 2015-10-20 17:33 ` [PATCH v6 06/19] fsnotify: export several symbols Jeff Layton 2015-10-20 17:33 ` [PATCH v6 07/19] locks: create a new notifier chain for lease attempts Jeff Layton 2015-10-20 17:33 ` [PATCH v6 15/19] nfsd: convert nfs4_file->fi_fds array to use nfsd_files Jeff Layton 2015-10-20 17:33 ` [PATCH v6 19/19] nfsd: rip out the raparms cache Jeff Layton [not found] ` <1445362432-18869-1-git-send-email-jeff.layton-7I+n7zu2hftEKMMhf/gKZA@public.gmane.org> 2015-10-20 17:33 ` [PATCH v6 01/19] nfsd: move include of state.h from trace.c to trace.h Jeff Layton 2015-10-20 17:33 ` Jeff Layton [not found] ` <1445362432-18869-2-git-send-email-jeff.layton-7I+n7zu2hftEKMMhf/gKZA@public.gmane.org> 2015-10-21 15:51 ` J. Bruce Fields 2015-10-21 15:51 ` J. Bruce Fields 2015-10-20 17:33 ` [PATCH v6 02/19] fs: have flush_delayed_fput flush the workqueue job Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 05/19] fs: add fput_global Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 08/19] sunrpc: add a new cache_detail operation for when a cache is flushed Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 09/19] nfsd: add a new struct file caching facility to nfsd Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 10/19] nfsd: keep some rudimentary stats on nfsd_file cache Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 11/19] nfsd: allow filecache open to skip fh_verify check Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 12/19] nfsd: hook up nfsd_write to the new nfsd_file cache Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 13/19] nfsd: hook up nfsd_read to the " Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 14/19] nfsd: hook nfsd_commit up " Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 16/19] nfsd: have nfsd_test_lock use " Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 17/19] nfsd: convert fi_deleg_file and ls_file fields to nfsd_file Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-20 17:33 ` [PATCH v6 18/19] nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache Jeff Layton 2015-10-20 17:33 ` Jeff Layton 2015-10-21 15:57 ` [PATCH v6 00/19] nfsd: open file caching J. Bruce Fields 2015-10-21 15:57 ` J. Bruce Fields 2015-10-22 21:19 ` J. Bruce Fields 2015-10-22 21:19 ` J. Bruce Fields 2015-10-23 0:21 ` Jeff Layton
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=1445362432-18869-1-git-send-email-jeff.layton@primarydata.com \ --to=jlayton-vpemndpepfumzcb2o+c8xq@public.gmane.org \ --cc=bfields-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org \ --cc=linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \ --cc=linux-nfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \ --cc=viro-3bDd1+5oDREiFSDQTTA3OLVCufUGDwFn@public.gmane.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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.