[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230523085226.00006933.zbestahu@gmail.com>
Date: Tue, 23 May 2023 08:52:26 +0800
From: Yue Hu <zbestahu@...il.com>
To: Gao Xiang <hsiangkao@...ux.alibaba.com>
Cc: linux-erofs@...ts.ozlabs.org, LKML <linux-kernel@...r.kernel.org>,
Sandeep Dhavale <dhavale@...gle.com>, huyue2@...lpad.com,
zhangwen@...lpad.com
Subject: Re: [PATCH] erofs: use HIPRI by default if per-cpu kthreads are
enabled
On Mon, 22 May 2023 17:21:41 +0800
Gao Xiang <hsiangkao@...ux.alibaba.com> wrote:
> As Sandeep shown [1], high priority RT per-cpu kthreads are
> typically helpful for Android scenarios to minimize the scheduling
> latencies.
>
> Switch EROFS_FS_PCPU_KTHREAD_HIPRI on by default if
> EROFS_FS_PCPU_KTHREAD is on since it's the typical use cases for
> EROFS_FS_PCPU_KTHREAD.
>
> Also clean up unneeded sched_set_normal().
>
> [1] https://lore.kernel.org/r/CAB=BE-SBtO6vcoyLNA9F-9VaN5R0t3o_Zn+FW8GbO6wyUqFneQ@mail.gmail.com
> Cc: Sandeep Dhavale <dhavale@...gle.com>
> Signed-off-by: Gao Xiang <hsiangkao@...ux.alibaba.com>
> ---
> fs/erofs/Kconfig | 1 +
> fs/erofs/zdata.c | 2 --
> 2 files changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/fs/erofs/Kconfig b/fs/erofs/Kconfig
> index 704fb59577e0..f259d92c9720 100644
> --- a/fs/erofs/Kconfig
> +++ b/fs/erofs/Kconfig
> @@ -121,6 +121,7 @@ config EROFS_FS_PCPU_KTHREAD
> config EROFS_FS_PCPU_KTHREAD_HIPRI
> bool "EROFS high priority per-CPU kthread workers"
> depends on EROFS_FS_ZIP && EROFS_FS_PCPU_KTHREAD
> + default y
How about removing this config option?
> help
> This permits EROFS to configure per-CPU kthread workers to run
> at higher priority.
> diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
> index 92f3a01262cf..3ba505434f03 100644
> --- a/fs/erofs/zdata.c
> +++ b/fs/erofs/zdata.c
> @@ -367,8 +367,6 @@ static struct kthread_worker *erofs_init_percpu_worker(int cpu)
> return worker;
> if (IS_ENABLED(CONFIG_EROFS_FS_PCPU_KTHREAD_HIPRI))
> sched_set_fifo_low(worker->task);
> - else
> - sched_set_normal(worker->task, 0);
> return worker;
> }
>
Powered by blists - more mailing lists