From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dave Jones Subject: Re: [PATCH] Eliminate cpufreq_userspace scaling_setspeed deadlock Date: Tue, 20 Nov 2007 18:41:33 -0500 Message-ID: <20071120234132.GA9882@redhat.com> References: <20071026171821.GA20924@linux-os.sc.intel.com> <08DF4D958216244799FC84F3514D70F0B5555C@pdsmsx415.ccr.corp.intel.com> <20071107125638.GB27092@inferi.kami.home> <924EFEDD5F540B4284297C4DC59F3DEE15F13A@orsmsx423.amr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <924EFEDD5F540B4284297C4DC59F3DEE15F13A@orsmsx423.amr.corp.intel.com> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: cpufreq-bounces@lists.linux.org.uk Errors-To: cpufreq-bounces+glkc-cpufreq=m.gmane.org+glkc-cpufreq=m.gmane.org@lists.linux.org.uk To: "Pallipadi, Venkatesh" Cc: "Yu, Luming" , cpufreq , "Yu, Fenghua" On Wed, Nov 07, 2007 at 10:45:35PM -0800, Venki Pallipadi wrote: > > > >-----Original Message----- > >From: Mattia Dongili [mailto:malattia@linux.it] > >Sent: Wednesday, November 07, 2007 4:57 AM > >To: Yu, Luming > >Cc: Pallipadi, Venkatesh; Dave Jones; Yu, Fenghua; cpufreq > >Subject: Re: [PATCH] Eliminate cpufreq_userspace > >scaling_setspeed deadlock > > > >On Mon, Nov 05, 2007 at 04:58:18PM +0800, Yu, Luming wrote: > >> Based on my current testing results, this patch fixes my problem. > >> Please help push it upstream. > > > >BTW, this will break some userspace apps. I can think of cpufrequtils > >and some deamons that use the userspace governor to set frequencies. > > > > Yes. But, this is the only solution we have for the deadlock right now > and I think it should goto upstream as soon as possible, unless we have > some better alternative. Userspace breakage is better bargain than a > kernel deadlock. Isn't it :-). I'm torn over this patch. I don't see a better way to fix this either, but the potential for breaking userspace is there. I think I'm going to merge this, and see how things go in -mm for a while. Dave -- http://www.codemonkey.org.uk