[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20191119160343.GA2440@quack2.suse.cz>
Date: Tue, 19 Nov 2019 17:03:43 +0100
From: Jan Kara <jack@...e.cz>
To: Chengguang Xu <cgxu519@...ernel.net>
Cc: jack@...e.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] quota: remove unnecessary check in dquot_add_inodes()
and dquot_add_space()
On Sun 17-11-19 21:20:28, Chengguang Xu wrote:
> After passed grace time we treat softlimit as hardlimit,
> so we don't have to compare desire usage with softlimit
> in this place.
>
> Signed-off-by: Chengguang Xu <cgxu519@...ernel.net>
Thanks for the patch! I guess you mean that when dqb_itime is set, we are
sure that softlimit is exceeded. You are right but the benefit of your
change is not big and I prefer to keep the current "defensive" condition.
Honza
> ---
> fs/quota/dquot.c | 2 --
> 1 file changed, 2 deletions(-)
>
> diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
> index 6e826b454082..97740077afac 100644
> --- a/fs/quota/dquot.c
> +++ b/fs/quota/dquot.c
> @@ -1283,7 +1283,6 @@ static int dquot_add_inodes(struct dquot *dquot, qsize_t inodes,
> }
>
> if (dquot->dq_dqb.dqb_isoftlimit &&
> - newinodes > dquot->dq_dqb.dqb_isoftlimit &&
> dquot->dq_dqb.dqb_itime &&
> ktime_get_real_seconds() >= dquot->dq_dqb.dqb_itime &&
> !ignore_hardlimit(dquot)) {
> @@ -1333,7 +1332,6 @@ static int dquot_add_space(struct dquot *dquot, qsize_t space,
> }
>
> if (dquot->dq_dqb.dqb_bsoftlimit &&
> - tspace > dquot->dq_dqb.dqb_bsoftlimit &&
> dquot->dq_dqb.dqb_btime &&
> ktime_get_real_seconds() >= dquot->dq_dqb.dqb_btime &&
> !ignore_hardlimit(dquot)) {
> --
> 2.21.0
>
>
>
--
Jan Kara <jack@...e.com>
SUSE Labs, CR
Powered by blists - more mailing lists