linux-kernel - Re: [RFC PATCH 1/2] genirq: Add chip hooks for taking CPUs on/off line.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <alpine.LFD.2.00.1103192050400.2787@localhost6.localdomain6>
Date:	Sat, 19 Mar 2011 21:51:13 +0100 (CET)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	David Daney <ddaney@...iumnetworks.com>
cc:	linux-mips@...ux-mips.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH 1/2] genirq: Add chip hooks for taking CPUs on/off
 line.

On Fri, 18 Mar 2011, David Daney wrote:
> --- a/include/linux/irqdesc.h
> +++ b/include/linux/irqdesc.h
> @@ -178,6 +178,12 @@ static inline int irq_has_action(unsigned int irq)
>  	return desc->action != NULL;
>  }
>  
> +/* Test to see if the irq is currently enabled */
> +static inline int irq_desc_is_enabled(struct irq_desc *desc)
> +{
> +	return desc->depth == 0;
> +}

That want's to go into kernel/irq/internal.h

>  #ifndef CONFIG_GENERIC_HARDIRQS_NO_COMPAT
>  static inline int irq_balancing_disabled(unsigned int irq)
>  {
> diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
> index c9c0601..40736f7 100644
> --- a/kernel/irq/chip.c
> +++ b/kernel/irq/chip.c
> @@ -689,3 +689,38 @@ void irq_modify_status(unsigned int irq, unsigned long clr, unsigned long set)
>  
>  	irq_put_desc_unlock(desc, flags);
>  }
> +
> +void irq_cpu_online(unsigned int irq)

Odd function name. It does not reflect that this is for per cpu
interrupts. So something like irq_xxx_per_cpu_irq(irq)
might be a bit more descriptive.

> +{

So that's called on the cpu which goes online, right?

I wonder whether we can add any sanity check to verify this.

Though I would not worry too much about it. Calling that from a cpu
which is not going offline should have enough nasty side effects that
it's noticed during development. :)

> +	unsigned long flags;
> +	struct irq_chip *chip;
> +	struct irq_desc *desc = irq_to_desc(irq);

Needs to check !desc

> +	raw_spin_lock_irqsave(&desc->lock, flags);
> +
> +	chip = irq_data_get_irq_chip(&desc->irq_data);
> +
> +	if (chip && chip->irq_cpu_online)
> +		chip->irq_cpu_online(&desc->irq_data,
> +				     irq_desc_is_enabled(desc));
> +
> +	raw_spin_unlock_irqrestore(&desc->lock, flags);
> +}
> +
> +void irq_cpu_offline(unsigned int irq)
> +{
> +	unsigned long flags;
> +	struct irq_chip *chip;
> +	struct irq_desc *desc = irq_to_desc(irq);

See above. 

Style nit: I prefer ordering:

+	struct irq_desc *desc = irq_to_desc(irq);
+	struct irq_chip *chip;
+	unsigned long flags;

For some reason, probably because I'm used to it, that's easier to
parse. But don't worry about that, I'll turn it around before sticking
it into git. :)

Otherwise I'm fine with the approach itself. 

Though one question remains: should we just iterate over the irq space
and call the online/offline callbacks when available instead of having
the arch code do the iteration.

Thanks,

	tglx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/