From:   "Sungjong Seo" <sj1557.seo@samsung.com>
To:     <linkinjeon@kernel.org>
Cc:     <linux-fsdevel@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
        <sj1557.seo@samsung.com>
In-Reply-To: <HK2PR04MB3891563FC310AE5E70896932817B9@HK2PR04MB3891.apcprd04.prod.outlook.com>
Subject: RE: [PATCH] exfat: fix missing REQ_SYNC in exfat_update_bhs()
Date:   Tue, 21 Dec 2021 21:34:58 +0900
Message-ID: <002001d7f667$2a968190$7fc384b0$@samsung.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Thread-Index: AQHJ22daV37VIoTUQqGa996KyhXcpAHh1MBHrEn4ssA=
Content-Language: ko
Content-Type: text/plain; charset="utf-8"
CMS-TYPE: 101P
DLP-Filter: Pass
References: <CGME20211220015249epcas1p4870b0ee2aeedbc4ca3b0b6e39361d4bb@epcas1p4.samsung.com>
        <HK2PR04MB3891563FC310AE5E70896932817B9@HK2PR04MB3891.apcprd04.prod.outlook.com>
Precedence: bulk

> If 'dirsync' is enabled, all directory updates within the
> filesystem should be done synchronously. exfat_update_bh()
> does as this, but exfat_update_bhs() does not.
> 
> Signed-off-by: Yuezhang.Mo <Yuezhang.Mo@sony.com>
> Reviewed-by: Andy.Wu <Andy.Wu@sony.com>
> Reviewed-by: Aoyama, Wataru <wataru.aoyama@sony.com>
> Reviewed-by: Kobayashi, Kento <Kento.A.Kobayashi@sony.com>
> ---
>  fs/exfat/misc.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/exfat/misc.c b/fs/exfat/misc.c
> index d34e6193258d..d5bd8e6d9741 100644
> --- a/fs/exfat/misc.c
> +++ b/fs/exfat/misc.c
> @@ -10,6 +10,7 @@
>  #include <linux/fs.h>
>  #include <linux/slab.h>
>  #include <linux/buffer_head.h>
> +#include <linux/blk_types.h>
> 
>  #include "exfat_raw.h"
>  #include "exfat_fs.h"
> @@ -180,7 +181,7 @@ int exfat_update_bhs(struct buffer_head **bhs, int
> nr_bhs, int sync)
>  		set_buffer_uptodate(bhs[i]);
>  		mark_buffer_dirty(bhs[i]);
>  		if (sync)
> -			write_dirty_buffer(bhs[i], 0);
> +			write_dirty_buffer(bhs[i], REQ_SYNC);

I think there is no problem in terms of functionality related to
"synchronously" in the original code. However, REQ_SYNC could affect
I/O scheduling, and exfat_update_bh() already requests I/O with this
flag by calling sync_dirty_buffer(). And it is desirable for two
functions to have the same concept for I/O requests.

So, the original code does not seem like a bug,
but this patch looks useful.
Thanks.

Reviewed-by: Sungjong Seo <sj1557.seo@samsung.com>

>  	}
> 
>  	for (i = 0; i < nr_bhs && sync; i++) {
> --
> 2.25.1