Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   cgel.zte@gmail.com
To:     axboe@kernel.dk, bsingharora@gmail.com, akpm@linux-foundation.org,
        iamjoonsoo.kim@lge.com
Cc:     linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
        Yang Yang <yang.yang29@zte.com.cn>,
        Ran Xiaokai <ran.xiaokai@zte.com.cn>,
        wangyong <wang.yong12@zte.com.cn>
Subject: [PATCH RESEND] block: count submission time as thrashing delay for delayacct
Date:   Mon, 12 Sep 2022 08:10:54 +0000
Message-Id: <20220912081053.17691-1-yang.yang29@zte.com.cn>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Yang Yang (CGEL ZTE) <yang.yang29@zte.com.cn>

PSI tracks the time system waiting for refaulting pages(or called
thrashing) spent on submitting the IO[1]. Delayacct also tracks
thrashing, but we failed to update it accordingly.

And we then have the ability to detect re-entrance of thrashing
accounting for delayacct[2], so we complete the accounting now.
This will make the thrashing accounting more accurate specifically
when device is congested or the submitting cgroup IO-throttled.

[1] commit b8e24a9300b0 ("block: annotate refault stalls from IO submission")
[2] commit 317bdccd67b2 ("delayacct: support re-entrance detection of thrashing accounting")

Signed-off-by: Yang Yang (CGEL ZTE) <yang.yang29@zte.com.cn>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Ran Xiaokai (CGEL ZTE) <ran.xiaokai@zte.com.cn>
Cc: wangyong (CGEL ZTE) <wang.yong12@zte.com.cn>
---
 block/blk-core.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/block/blk-core.c b/block/blk-core.c
index a0d1104c5590..d722ead0557f 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -41,6 +41,7 @@
 #include <linux/part_stat.h>
 #include <linux/sched/sysctl.h>
 #include <linux/blk-crypto.h>
+#include <linux/delayacct.h>
 
 #define CREATE_TRACE_POINTS
 #include <trace/events/block.h>
@@ -831,16 +832,19 @@ void submit_bio(struct bio *bio)
 
 	/*
 	 * If we're reading data that is part of the userspace workingset, count
-	 * submission time as memory stall.  When the device is congested, or
-	 * the submitting cgroup IO-throttled, submission can be a significant
+	 * submission time as memory stall and delay.  When the device is congested,
+	 * or the submitting cgroup IO-throttled, submission can be a significant
 	 * part of overall IO time.
 	 */
 	if (unlikely(bio_op(bio) == REQ_OP_READ &&
 	    bio_flagged(bio, BIO_WORKINGSET))) {
 		unsigned long pflags;
+		bool in_thrashing;
 
+		delayacct_thrashing_start(&in_thrashing);
 		psi_memstall_enter(&pflags);
 		submit_bio_noacct(bio);
+		delayacct_thrashing_end(&in_thrashing);
 		psi_memstall_leave(&pflags);
 		return;
 	}
-- 
2.25.1