From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754375AbbJSPkS (ORCPT ); Mon, 19 Oct 2015 11:40:18 -0400 Received: from mail-pa0-f46.google.com ([209.85.220.46]:36367 "EHLO mail-pa0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752684AbbJSPkO convert rfc822-to-8bit (ORCPT ); Mon, 19 Oct 2015 11:40:14 -0400 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8BIT To: Viresh Kumar , "Arnd Bergmann" From: Michael Turquette In-Reply-To: <20151008074828.GF4570@linux> Cc: "Greg Kroah-Hartman" , "Stephen Boyd" , "Rafael J. Wysocki" , "Len Brown" , "Pavel Machek" , "Bartlomiej Zolnierkiewicz" , "Nishanth Menon" , "Dmitry Torokhov" , linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org References: <5706174.BUsqiXljxZ@wuerfel> <20151007125735.GA8170@kroah.com> <20151007130324.GA4557@linux> <5308760.2G5KJFF9Pi@wuerfel> <20151008074828.GF4570@linux> Message-ID: <20151019154009.20687.41178@quantum> User-Agent: alot/0.3.6 Subject: Re: [PATCH] PM / OPP: fix debugfs files for 64-bit Date: Mon, 19 Oct 2015 08:40:09 -0700 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Quoting Viresh Kumar (2015-10-08 00:48:28) > On 07-10-15, 21:12, Arnd Bergmann wrote: > > I think it clearly makes sense to have a fixed length for each of these > > members: > > > > either 32 bit is enough to represent all possible values, then > > there is no need to make them 'long' on 64-bit architectures, or 32 bit > > is not enough and then the code is broken on 32-bit architectures today > > and should be fixed. > > I agree. > > But I am not 100% sure why it was done this way to start with. > Probably this is the logic behind that: > - Max clock rate supported by a u32 is ~ 4.295 GHz > - People expected that, we will not reach this rate for 32 bit systems > but for 64 bit ones. > - If above is true, then making it u64 for all will generate not very > optimized code for 32bit systems, as we need to fetch two 32bit > values everytime then. > - And making it u32 for 64 bit systems wouldn't be great as well, as > we need to mask out half of the read value. > > Ofcourse, Mike and Stephen can correct me here :) I chose unsigned long for the common clock framework _implemenation_, because the long-standing clk.h _api_ returns this type for clk_get_rate and passes this type in for clk_set_rate and clk_round_rate. > > > In my patch, I assumed that if 32-bit architectures work fine today, then > > we don't need more range on 64-bit architectures either. > > The problem here is that we haven't fixed it properly. > - clock framework expects it to be unsigned long > - DT is sending a 64 bit value in Hz > - But we are storing and exposing it in u32 Don't forget cpufreq is using unsigned int for KHz. > > That's weird, isn't it? Yes it is. > > So, either we update clock API and other similar APIs to u64 or u32 > (which may not be the right thing to do), Or we keep it unsigned long > here as well and add debugfs_create_ulong(). This comes up every now and then. I'm still trying to figure out if sub-Hertz quantities should be considered (e.g. representing freqs in milliHertz). Regards, Mike > > -- > viresh