From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pa0-x229.google.com (mail-pa0-x229.google.com [IPv6:2607:f8b0:400e:c03::229]) (using TLSv1 with cipher ECDHE-RSA-RC4-SHA (128/128 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id CE5D51400D8 for ; Thu, 15 May 2014 12:22:45 +1000 (EST) Received: by mail-pa0-f41.google.com with SMTP id lj1so403696pab.14 for ; Wed, 14 May 2014 19:22:42 -0700 (PDT) Date: Wed, 14 May 2014 19:22:38 -0700 From: Brian Norris To: Linux Kernel Subject: roundup_pow_of_two() may not handle 64-bit integers Message-ID: <20140515022238.GL28907@ld-irv-0074> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Howells , Paul Mackerras , Andrew Morton , Brian Norris , linuxppc-dev@lists.ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hi, I'm looking to use roundup_pow_of_two() (actually, order_base_2()) from , but it seems that it only supports 64-bit integers if your toolchain uses a 64-bit 'unsigned long' type. This is strange, considering that ilog2() is explicitly designed for 32-bit or 64-bit compatibility. I also note that there is at least one location in which this limitation currently might be problematic: in pnv_pci_ioda2_set_bypass() (arch/powerpc/platforms/powernv/pci-ioda.c). It looks like this could be a problem if using large amounts of DRAM on a 32-bit PPC build, with 64-bit physical addresses. (There may be other cases like this one, but I haven't closely studied all callers of roundup_pow_of_two().) I'm thinking of cooking a patch to improve roundup_pow_of_two() (and thus order_base_2()), but I'd like to solicit comments on the basic problem first. Regards, Brian P.S. And of course, rounddown_pow_of_two() has the same issue.