From patchwork Tue Jun 7 17:30:31 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jim Rees X-Patchwork-Id: 858302 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by demeter1.kernel.org (8.14.4/8.14.3) with ESMTP id p57HTn1Q019764 for ; Tue, 7 Jun 2011 17:30:34 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756769Ab1FGRae (ORCPT ); Tue, 7 Jun 2011 13:30:34 -0400 Received: from int-mailstore01.merit.edu ([207.75.116.232]:50747 "EHLO int-mailstore01.merit.edu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755935Ab1FGRad (ORCPT ); Tue, 7 Jun 2011 13:30:33 -0400 Received: from localhost (localhost.localdomain [127.0.0.1]) by int-mailstore01.merit.edu (Postfix) with ESMTP id 5E618305A9CC; Tue, 7 Jun 2011 13:30:33 -0400 (EDT) X-Virus-Scanned: amavisd-new at int-mailstore01.merit.edu Received: from int-mailstore01.merit.edu ([127.0.0.1]) by localhost (int-mailstore01.merit.edu [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id eZ83jECKwS9k; Tue, 7 Jun 2011 13:30:32 -0400 (EDT) Received: from merit.edu (host-17.subnet-17.med.umich.edu [141.214.17.17]) by int-mailstore01.merit.edu (Postfix) with ESMTPSA id 9F84B3055B74; Tue, 7 Jun 2011 13:30:32 -0400 (EDT) Date: Tue, 7 Jun 2011 13:30:31 -0400 From: Jim Rees To: Benny Halevy Cc: linux-nfs@vger.kernel.org, peter honeyman Subject: [PATCH 38/88] pnfsblock: merge rw extents Message-ID: References: MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: Sender: linux-nfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-nfs@vger.kernel.org X-Greylist: IP, sender and recipient auto-whitelisted, not delayed by milter-greylist-4.2.6 (demeter1.kernel.org [140.211.167.41]); Tue, 07 Jun 2011 17:30:34 +0000 (UTC) From: Fred Isaman Signed-off-by: Fred Isaman Signed-off-by: Benny Halevy --- fs/nfs/blocklayout/extents.c | 48 ++++++++++++++++++++++++++++++++++++++--- 1 files changed, 44 insertions(+), 4 deletions(-) diff --git a/fs/nfs/blocklayout/extents.c b/fs/nfs/blocklayout/extents.c index 1719a67..a05ee2a 100644 --- a/fs/nfs/blocklayout/extents.c +++ b/fs/nfs/blocklayout/extents.c @@ -773,12 +773,43 @@ _prep_new_extent(struct pnfs_block_extent *new, new->be_inval = orig->be_inval; } +/* Tries to merge be with extent in front of it in list. + * Frees storage if not used. + */ +static struct pnfs_block_extent * +_front_merge(struct pnfs_block_extent *be, struct list_head *head, + struct pnfs_block_extent *storage) +{ + struct pnfs_block_extent *prev; + + if (!storage) + goto no_merge; + if (&be->be_node == head || be->be_node.prev == head) + goto no_merge; + prev = list_entry(be->be_node.prev, struct pnfs_block_extent, be_node); + if ((prev->be_f_offset + prev->be_length != be->be_f_offset) || + !extents_consistent(prev, be)) + goto no_merge; + _prep_new_extent(storage, prev, prev->be_f_offset, + prev->be_length + be->be_length, prev->be_state); + list_replace(&prev->be_node, &storage->be_node); + put_extent(prev); + list_del(&be->be_node); + put_extent(be); + return storage; + + no_merge: + kfree(storage); + return be; +} + static u64 set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length) { - u64 rv = 0; + u64 rv = offset + length; struct pnfs_block_extent *be, *e1, *e2, *e3, *new, *old; struct pnfs_block_extent *children[3]; + struct pnfs_block_extent *merge1 = NULL, *merge2 = NULL; int i = 0, j; dprintk("%s(%llu, %llu)\n", __func__, offset, length); @@ -792,7 +823,6 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length) spin_lock(&bl->bl_ext_lock); be = find_get_extent_locked(bl, offset); - print_bl_extent(be); rv = be->be_f_offset + be->be_length; if (be->be_state != PNFS_BLOCK_INVALID_DATA) { spin_unlock(&bl->bl_ext_lock); @@ -805,13 +835,15 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length) PNFS_BLOCK_INVALID_DATA); children[i++] = e1; kref_get(&e1->be_refcnt); + print_bl_extent(e1); } else - kfree(e1); + merge1 = e1; _prep_new_extent(e2, be, offset, min(length, be->be_f_offset + be->be_length - offset), PNFS_BLOCK_READWRITE_DATA); children[i++] = e2; kref_get(&e2->be_refcnt); + print_bl_extent(e2); if (offset + length < be->be_f_offset + be->be_length) { _prep_new_extent(e3, be, e2->be_f_offset + e2->be_length, be->be_f_offset + be->be_length - @@ -819,8 +851,9 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length) PNFS_BLOCK_INVALID_DATA); children[i++] = e3; kref_get(&e3->be_refcnt); + print_bl_extent(e3); } else - kfree(e3); + merge2 = e3; /* Remove be from list, and insert the e* */ /* We don't get refs on e*, since this list is the base reference @@ -831,11 +864,18 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length) new = children[0]; list_replace(&be->be_node, &new->be_node); put_extent(be); + new = _front_merge(new, &bl->bl_extents[RW_EXTENT], merge1); for (j = 1; j < i; j++) { old = new; new = children[j]; list_add(&new->be_node, &old->be_node); } + if (merge2) { + /* This is a HACK, should just create a _back_merge function */ + new = list_entry(new->be_node.next, + struct pnfs_block_extent, be_node); + new = _front_merge(new, &bl->bl_extents[RW_EXTENT], merge2); + } spin_unlock(&bl->bl_ext_lock); /* Since we removed the base reference above, be is now scheduled for