[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20160317211439.GA21968@intel.com>
Date: Thu, 17 Mar 2016 14:14:40 -0700
From: "Luck, Tony" <tony.luck@...el.com>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: LKML <linux-kernel@...r.kernel.org>,
Harry Junior <harryjr@...look.fr>, x86@...nel.org,
Peter Zijlstra <peterz@...radead.org>,
Joe Lawrence <joe.lawrence@...atus.com>,
Borislav Petkov <bp@...en8.de>,
Ben Hutchings <ben@...adent.org.uk>
Subject: Re: [PATCH V2] x86/irq: Cure live lock in fixup_irqs()
This version also passes repeated iterations of:
# echo 0 | tee /sys/devices/system/cpu/cpu*/online
# echo 1 | tee /sys/devices/system/cpu/cpu*/online
Some of the time taken in the offline cycle may be
a more than O(N) chasing of IRQs onto an ever smaller
available set of online cpus. There are over 4000
lines like this:
[ 218.220045] Broke affinity for irq 66
on the console for each offline pass.
Also a couple of these:
[ 221.951171] IRQ fixup: irq 68 move in progress, old vector 79
for each pass offlining all but one cpu.
Tested-by: Tony Luck <tony.luck@...el.com>
-Tony
Powered by blists - more mailing lists