From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pf0-f199.google.com (mail-pf0-f199.google.com [209.85.192.199]) by kanga.kvack.org (Postfix) with ESMTP id D5CF36B0069 for ; Thu, 5 Jan 2017 07:22:16 -0500 (EST) Received: by mail-pf0-f199.google.com with SMTP id i88so796125248pfk.3 for ; Thu, 05 Jan 2017 04:22:16 -0800 (PST) Received: from NAM03-DM3-obe.outbound.protection.outlook.com (mail-dm3nam03on0080.outbound.protection.outlook.com. [104.47.41.80]) by mx.google.com with ESMTPS id d85si75851180pfb.163.2017.01.05.04.22.15 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Thu, 05 Jan 2017 04:22:16 -0800 (PST) Date: Thu, 5 Jan 2017 13:22:00 +0100 From: Robert Richter Subject: Re: [PATCH 2/2] arm64: mm: enable CONFIG_HOLES_IN_ZONE for NUMA Message-ID: <20170105122200.GV4930@rric.localdomain> References: <1481706707-6211-1-git-send-email-ard.biesheuvel@linaro.org> <1481706707-6211-3-git-send-email-ard.biesheuvel@linaro.org> <20170104132831.GD18193@arm.com> <20170104140223.GF18193@arm.com> <20170105112407.GU4930@rric.localdomain> <20170105120819.GH679@arm.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline In-Reply-To: <20170105120819.GH679@arm.com> Sender: owner-linux-mm@kvack.org List-ID: To: Will Deacon Cc: Ard Biesheuvel , "linux-arm-kernel@lists.infradead.org" , "linux-kernel@vger.kernel.org" , "linux-mm@kvack.org" , Catalin Marinas , Andrew Morton , Hanjun Guo , Yisheng Xie , James Morse On 05.01.17 12:08:20, Will Deacon wrote: > On Thu, Jan 05, 2017 at 12:24:07PM +0100, Robert Richter wrote: > > On 04.01.17 14:02:23, Will Deacon wrote: > > > Using early_pfn_valid feels like a bodge to me, since having pfn_valid > > > return false for something that early_pfn_valid says is valid (and is > > > therefore initialised in the memmap) makes the NOMAP semantics even more > > > confusing. > > > > The concern I have had with HOLES_IN_ZONE is that it enables > > pfn_valid_within() for arm64. This means that each pfn of a section is > > checked which is done only once for the section otherwise. With up to > > 2^18 pages per section we traverse the memblock list by that factor > > more often. There could be a performance regression. > > There could be, but we're trying to fix a bug here. I wouldn't have > thought that walking over pfns like that is done very often. The bug happens on a small number of machines depending on the memory layout. The fix affects all systems. And right know the impact is unclear. > > I haven't numbers yet, since the fix causes another kernel crash. And, > > this is the next problem I have. The crash doesn't happen otherwise. So, > > either it uncovers another bug or the fix is incomplete. Though the > > changes look like it should work. This needs more investigation. > > I really can't see how the fix causes a crash, and I couldn't reproduce > it on any of my boards, nor could any of the Linaro folk afaik. Are you > definitely running mainline with just these two patches from Ard? Yes, just both patches applied. Various other solutions were working. -Robert -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org