From mboxrd@z Thu Jan 1 00:00:00 1970 From: Viresh Kumar Subject: Re: [PATCH] cpufreq: powernv: Fixes initialization of chip and chip mask Date: Wed, 20 Apr 2016 15:20:56 +0530 Message-ID: <20160420095056.GE12019@vireshk-i7> References: <1461144758-17810-1-git-send-email-shilpa.bhat@linux.vnet.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-pa0-f47.google.com ([209.85.220.47]:32920 "EHLO mail-pa0-f47.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932257AbcDTJvA (ORCPT ); Wed, 20 Apr 2016 05:51:00 -0400 Received: by mail-pa0-f47.google.com with SMTP id zm5so16385144pac.0 for ; Wed, 20 Apr 2016 02:51:00 -0700 (PDT) Content-Disposition: inline In-Reply-To: <1461144758-17810-1-git-send-email-shilpa.bhat@linux.vnet.ibm.com> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Shilpasri G Bhat Cc: rjw@rjwysocki.net, linux-pm@vger.kernel.org, linuxppc-dev@ozlabs.org, linux-kernel@vger.kernel.org On 20-04-16, 15:02, Shilpasri G Bhat wrote: > commit 735366fc4077 ("cpufreq: powernv: Call throttle_check() on > receiving OCC_THROTTLE") used cpumask_of_node() as the chip mask. But > this mask contains only online cpus. This breaks a setup where cpufreq > is initialized with few offline cores and made online later. So this > patch fixes this bug by scanning all the possible cpus and sets the > cpu in the chip mask. It also fixes the chip discovery with > non-contiguous cpu mask. This patch creates a list of chips > 'powernv_chip_list' to replace the chip array for cleaner > initialization. > > Signed-off-by: Shilpasri G Bhat > Reviewed-by: Gautham R. Shenoy > --- > drivers/cpufreq/powernv-cpufreq.c | 80 +++++++++++++++++++++++---------------- > 1 file changed, 48 insertions(+), 32 deletions(-) You have made your patch less readable by mixing two things here. Yes you prefer/need the list way of doing things for the new stuff, but that should have been done separately. Right now, I have to read it very carefully to see which line did the real change you are talking about. So, please split this up into multiple patches. First one just moving to the list instead of array. -- viresh