Received-SPF: pass (google.com: domain of linux-ext4+bounces-2186-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
From: "Ritesh Harjani (IBM)" <ritesh.list@gmail.com>
To: linux-ext4@vger.kernel.org,
	linux-xfs@vger.kernel.org
Cc: linux-fsdevel@vger.kernel.org,
	Matthew Wilcox <willy@infradead.org>,
	"Darrick J . Wong" <djwong@kernel.org>,
	Ojaswin Mujoo <ojaswin@linux.ibm.com>,
	Ritesh Harjani <ritesh.list@gmail.com>,
	Jan Kara <jack@suse.cz>
Subject: [RFCv3 6/7] iomap: Optimize iomap_read_folio
Date: Thu, 25 Apr 2024 18:58:50 +0530
Message-ID: <a01641c22af0856fa2b19ab00a6660706056666d.1714046808.git.ritesh.list@gmail.com>
In-Reply-To: <cover.1714046808.git.ritesh.list@gmail.com>
References: <cover.1714046808.git.ritesh.list@gmail.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

iomap_readpage_iter() handles "uptodate blocks" and "not uptodate blocks"
within a folio separately. This makes iomap_read_folio() to call into
->iomap_begin() to request for extent mapping even though it might already
have an extent which is not fully processed.

This happens when we either have a large folio or with bs < ps. In these
cases we can have sub blocks which can be uptodate (say for e.g. due to
previous writes). With iomap_read_folio_iter(), this is handled more
efficiently by not calling ->iomap_begin() call until all the sub blocks
with the current folio are processed.

Signed-off-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
cc: Ojaswin Mujoo <ojaswin@linux.ibm.com>
---
 fs/iomap/buffered-io.c | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 9f79c82d1f73..0a4269095ae2 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -444,6 +444,24 @@ static loff_t iomap_readpage_iter(const struct iomap_iter *iter,
 	return pos - orig_pos + plen;
 }
 
+static loff_t iomap_read_folio_iter(const struct iomap_iter *iter,
+		struct iomap_readpage_ctx *ctx)
+{
+	struct folio *folio = ctx->cur_folio;
+	size_t pos = offset_in_folio(folio, iter->pos);
+	loff_t length = min_t(loff_t, folio_size(folio) - pos,
+			      iomap_length(iter));
+	loff_t done, ret;
+
+	for (done = 0; done < length; done += ret) {
+		ret = iomap_readpage_iter(iter, ctx, done);
+		if (ret <= 0)
+			return ret;
+	}
+
+	return done;
+}
+
 int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops)
 {
 	struct iomap_iter iter = {
@@ -459,7 +477,7 @@ int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops)
 	trace_iomap_readpage(iter.inode, 1);
 
 	while ((ret = iomap_iter(&iter, ops)) > 0)
-		iter.processed = iomap_readpage_iter(&iter, &ctx, 0);
+		iter.processed = iomap_read_folio_iter(&iter, &ctx);
 
 	if (ret < 0)
 		folio_set_error(folio);
-- 
2.44.0