Message-ID: <1377261222.14021.23.camel@jlt4.sipsolutions.net>
Subject: Re: [sched_delayed] sched: RT throttling activated
From: Johannes Berg <johannes@sipsolutions.net>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Martin Mokrejs <mmokrejs@fold.natur.cuni.cz>, Theodore Tso <tytso@mit.edu>,
        Thomas Gleixner <tglx@linutronix.de>, mingo@redhat.com,
        LKML <linux-kernel@vger.kernel.org>
Date: Fri, 23 Aug 2013 14:33:42 +0200
In-Reply-To: <20130823121004.GW31370@twins.programming.kicks-ass.net>
References: <521722EE.70209@fold.natur.cuni.cz>
	 <20130823100913.GQ31370@twins.programming.kicks-ass.net>
	 <52173BBD.3050803@fold.natur.cuni.cz>
	 <20130823110254.GU31370@twins.programming.kicks-ass.net>
	 <521748FC.10809@fold.natur.cuni.cz>
	 <20130823121004.GW31370@twins.programming.kicks-ass.net>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 966
Lines: 28

On Fri, 2013-08-23 at 14:10 +0200, Peter Zijlstra wrote:

> >  2161  FF [irq/50-iwlwifi]
> 
> Oh a threaded interrupt, I presume you're not using "threadiqrs" since
> this is the only interrupt thread around and I see a
> 'request_threaded_irq()' call in
> drivers/net/wireless/iwlwifi/pcie/trans.c
> 
> And wow, why would that thing consume that much cpu. 

How much is "that much"?

> Johill, ever seen the iwlwifi interrupt go 'funny' and consume gobs of
> cpu-time?

Not really, no, though if you do a lot of data transfer it can actually
consume a fair bit of CPU - made better by two somewhat recent patches,
commits 01911dab97cb3e21e640aaca82689acec00ed848 and
68972c46f2975d3d61f9dc9f311f77bfc8a8b12b.

johannes

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