[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200408112152.GP21484@bombadil.infradead.org>
Date: Wed, 8 Apr 2020 04:21:52 -0700
From: Matthew Wilcox <willy@...radead.org>
To: Tetsuhiro Kohada <Kohada.Tetsuhiro@...mitsubishielectric.co.jp>
Cc: Mori.Takahiro@...mitsubishielectric.co.jp,
motai.hirotaka@...mitsubishielectric.co.jp,
Namjae Jeon <namjae.jeon@...sung.com>,
Sungjong Seo <sj1557.seo@...sung.com>,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] exfat: replace 'time_ms' with 'time_10ms'
Please leave at least 24 hours between sending new versions so that
you can collect all feedback relating to your change, and we don't see
discussion fragment between different threads.
> @@ -84,10 +84,10 @@ void exfat_get_entry_time(struct exfat_sb_info *sbi, struct timespec64 *ts,
> t >> 11, (t >> 5) & 0x003F, (t & 0x001F) << 1);
>
>
> - /* time_ms field represent 0 ~ 199(1990 ms) */
> - if (time_ms) {
> - ts->tv_sec += time_ms / 100;
> - ts->tv_nsec = (time_ms % 100) * 10 * NSEC_PER_MSEC;
> + /* time_10ms field represent 0 ~ 199cs(1990 ms) */
> + if (time_10ms) {
> + ts->tv_sec += (time_10ms * 10) / 1000;
> + ts->tv_nsec = (time_10ms * 10) % 1000 * NSEC_PER_MSEC;
I find this more confusing than the original.
ts->tv_sec += time_10ms / 100;
ts->tv_nsec = (time_10ms % 100) * 10 * NSEC_PER_MSEC;
is easier to understand for me, not least because I don't need to worry
about the operator precedence between % and *.
Powered by blists - more mailing lists