Linux Serial subsystem development
 help / color / mirror / Atom feed
From: Jiri Slaby <jirislaby@kernel.org>
To: Andy Chiu <andy.chiu@sifive.com>,
	linux-serial@vger.kernel.org, linux-riscv@lists.infradead.org,
	palmer@dabbelt.com, gregkh@linuxfoundation.org
Cc: paul.walmsley@sifive.com
Subject: Re: [PATCH] tty: serial: manage irq with spin_lock_irqsave in SiFive console
Date: Thu, 1 Feb 2024 08:21:44 +0100	[thread overview]
Message-ID: <2f1b1262-8c32-4dd5-8052-d47682c0b1f6@kernel.org> (raw)
In-Reply-To: <20240201065932.19899-1-andy.chiu@sifive.com>

On 01. 02. 24, 7:59, Andy Chiu wrote:
> It is not safe to call spin_lock() with irq disabled on RT-linux.
> Instead, transfer the code segment to spin_lock_irqsave to make it work
> on both RT and non-RT linux.

Hi,

have you investigated what is protected by the local_irq_save() in 
there? The lock is not always taken, OTOH the interrupts are always 
disabled.

I believe the fix is not as easy as is presented below.

> Signed-off-by: Andy Chiu <andy.chiu@sifive.com>
> ---
>   drivers/tty/serial/sifive.c | 8 +++-----
>   1 file changed, 3 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/tty/serial/sifive.c b/drivers/tty/serial/sifive.c
> index fa4c9336924f..3f0ddf8bfa7b 100644
> --- a/drivers/tty/serial/sifive.c
> +++ b/drivers/tty/serial/sifive.c
> @@ -788,13 +788,12 @@ static void sifive_serial_console_write(struct console *co, const char *s,
>   	if (!ssp)
>   		return;
>   
> -	local_irq_save(flags);
>   	if (ssp->port.sysrq)
>   		locked = 0;
>   	else if (oops_in_progress)
> -		locked = spin_trylock(&ssp->port.lock);
> +		locked = spin_trylock_irqsave(&ssp->port.lock, flags);
>   	else
> -		spin_lock(&ssp->port.lock);
> +		spin_lock_irqsave(&ssp->port.lock, flags);
>   
>   	ier = __ssp_readl(ssp, SIFIVE_SERIAL_IE_OFFS);
>   	__ssp_writel(0, SIFIVE_SERIAL_IE_OFFS, ssp);
> @@ -804,8 +803,7 @@ static void sifive_serial_console_write(struct console *co, const char *s,
>   	__ssp_writel(ier, SIFIVE_SERIAL_IE_OFFS, ssp);
>   
>   	if (locked)
> -		spin_unlock(&ssp->port.lock);
> -	local_irq_restore(flags);
> +		spin_unlock_irqrestore(&ssp->port.lock, flags);
>   }
>   
>   static int __init sifive_serial_console_setup(struct console *co, char *options)

-- 
js
suse labs


  reply	other threads:[~2024-02-01  7:21 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-01  6:59 [PATCH] tty: serial: manage irq with spin_lock_irqsave in SiFive console Andy Chiu
2024-02-01  7:21 ` Jiri Slaby [this message]
2024-02-26  6:22   ` Andy Chiu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=2f1b1262-8c32-4dd5-8052-d47682c0b1f6@kernel.org \
    --to=jirislaby@kernel.org \
    --cc=andy.chiu@sifive.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=linux-serial@vger.kernel.org \
    --cc=palmer@dabbelt.com \
    --cc=paul.walmsley@sifive.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox