Re: 答复: [RFC] irq: Skip printing irq when desc->action is null even if any_count is not zero

[Neil Horman <nhorman@tuxdriver.com>]

On Thu, Jan 23, 2020 at 01:34:33PM +0100, Thomas Gleixner wrote:
> Chao,
> 
> "liuchao (CR)" <liuchao173@huawei.com> writes:
> > On Thu, Jan 23, 2020 at 03:29AM +0800, Neil Horman wrote:
> >> > I'm not opposed to suppress the output, but I really want the opinion
> >> > of the irqbalance maintainers on that.
> >
> > Irqbalance is an example. I mean, when this happens, users who cat /proc/interrupts 
> > may be confused about where the interrupt came from and what it was used for. 
> > People who use Linux may not understand the principle of this. They are not sure 
> > whether this is a problem of the system or not.
> 
> Well, this has been that way for 20+ years and so far nobody got
> confused. If it's not documented then we should do so.
> 
> >> Actually, irqbalance ignores the trailing irq name (or it should at least), so you
> >> should be able to drop that portion of /proc/irqbalance, though I cant speak for
> >> any other users of it.
> >
> > If irq isn't removed from /proc/interrups, it will still be parsed in
> > collect_full_irq_list and parse_proc_interrupts.
> 
> Sure, and why is that a problem? Again, this is really historic behaviour.
> 
> > irq_name is used in guess_arm_irq_hints.
> 
> That's a problem of guess_arm_irq_hints() then.
> 
> Again, I'm not against supressing such lines in general, but I want to
> make sure that no tool depends on that information.
> 
I think it probably makes sense to just keep it then.  I'm not sure I
see it as hurting anything to keep it around.

Neil

> Thanks,
> 
>         tglx
>