Re: [PATCH] pstore: avoid recursive spinlocks in theoops_in_progress case

[Anton Vorontsov]

On Tue, Sep 18, 2012 at 01:43:44AM +0800, Chuansheng Liu wrote:
> Like 8250 driver, when pstore is registered as a console,
> to avoid recursive spinlocks when panic happening, change the
> spin_lock_irqsave to spin_trylock_irqsave when oops_in_progress
> is true.
> 
> Signed-off-by: liu chuansheng <chuansheng.liu@xxxxxxxxx>
> ---
>  fs/pstore/platform.c |    7 ++++++-
>  1 files changed, 6 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c
> index 29996e8..cbcb5fc 100644
> --- a/fs/pstore/platform.c
> +++ b/fs/pstore/platform.c
> @@ -164,7 +164,12 @@ static void pstore_console_write(struct console *con, const char *s, unsigned c)
>  
>  		if (c > psinfo->bufsize)
>  			c = psinfo->bufsize;
> -		spin_lock_irqsave(&psinfo->buf_lock, flags);
> +
> +		if (oops_in_progress) {
> +			if (!spin_trylock_irqsave(&psinfo->buf_lock, flags))
> +				break;

Mm... why break?

Like 8250, in this case it's better still log the console.
That is, something along these lines:

bool locked = 1;

if (oops_in_progress)
	locked = spin_trylock_irqsave(&psinfo->buf_lock, flags);
else
	spin_lock_irqsave(&psinfo->buf_lock, flags);

memcpy(psinfo->buf, s, c);
psinfo->write(PSTORE_TYPE_CONSOLE, 0, NULL, 0, c, psinfo);

if (locked)
	spin_unlock_irqrestore(&psinfo->buf_lock, flags);


Thanks,

Anton.

> +		} else
> +			spin_lock_irqsave(&psinfo->buf_lock, flags);
>  		memcpy(psinfo->buf, s, c);
>  		psinfo->write(PSTORE_TYPE_CONSOLE, 0, NULL, 0, c, psinfo);
>  		spin_unlock_irqrestore(&psinfo->buf_lock, flags);
> -- 
> 1.7.0.4
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/