Re: [PATCH] printk: replacing the raw_spin_lock/unlock with raw_spin_lock_irqsave/irqrestore

[Kay Sievers]

On Mon, Jul 2, 2012 at 12:36 PM, Liu, Chuansheng
<chuansheng.liu@xxxxxxxxx> wrote:
> From: liu chuansheng <chuansheng.liu@xxxxxxxxx>
> Subject: [PATCH] printk: replacing the raw_spin_lock/unlock with raw_spin_lock_irqsave/irqrestore
>
> In function devkmsg_read/writev/llseek/poll/open()..., the function
> raw_spin_lock/unlock is used, there is potential deadlock case happening.
> CPU1: thread1 doing the cat /dev/kmsg:
>         raw_spin_lock(&logbuf_lock);
>         while (user->seq == log_next_seq) {
> when thread1 run here, at this time one interrupt is coming on CPU1 and running
> based on this thread,if the interrupt handle called the printk which need the
> logbuf_lock spin also, it will cause deadlock.
>
> So we should use raw_spin_lock_irq_save/irqrestore here.

They are all handling a system call, isn't raw_spin_lock_irq() without
the flag restoration good enough then?

Thanks,
Kay
--
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/