From mboxrd@z Thu Jan 1 00:00:00 1970 From: Prarit Bhargava Subject: Re: [patch 3/4] Locally disable the softlockup watchdog rather than touching it Date: Wed, 28 Mar 2007 11:27:34 -0400 Message-ID: <460A8966.6090208@redhat.com> References: <20070327214919.800272641@goop.org> <20070327215828.085422178@goop.org> <460A6EC0.4020701@redhat.com> <460A7F57.9020006@goop.org> <460A80E6.9040003@redhat.com> <460A8820.3060708@goop.org> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <460A8820.3060708@goop.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: Jeremy Fitzhardinge Cc: virtualization@lists.osdl.org, Andrew Morton , Ingo Molnar , John Hawkes , Linux Kernel , Eric Dumazet List-Id: virtualization@lists.linuxfoundation.org Jeremy Fitzhardinge wrote: > Prarit Bhargava wrote: > = >> You don't have to do them all -- you could do one with (as in my >> previous patch -- which I'm not married to BTW ;) ) >> >> touch_cpu_softlockup_watchdog() >> >> and all with >> >> touch_softlockup_watchdog() >> = > > Well, I think changing the meaning of touch_softlockup_watchdog() for > all existing callers is wrong - even if you change most of them to refer > to the cpu-local function. = Hmmm .... it was suggested to me that I should mimic what = touch_nmi_watchdog() does. > There are definitely specific occasions on > which touching all CPUs is the right thing to do, but not in the general > case. > = Yep. That's why I have both a single cpu touch and the whole shebang :) > The only thing I really care about in my patches is ignoring stolen > time. It may be that fixing that is enough to fix the reported problems > with spurious watchdog messages on tickless idle CPUs. > > = > J > =