From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.1 required=3.0 tests=BAYES_00,DKIM_INVALID, DKIM_SIGNED,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,NICE_REPLY_A,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_SANE_1 autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 591B8C433E0 for ; Mon, 8 Mar 2021 08:37:35 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id DB514650FD for ; Mon, 8 Mar 2021 08:37:34 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org DB514650FD Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 3895B8D000D; Mon, 8 Mar 2021 03:37:34 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 312308D000A; Mon, 8 Mar 2021 03:37:34 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 165C28D000D; Mon, 8 Mar 2021 03:37:34 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0080.hostedemail.com [216.40.44.80]) by kanga.kvack.org (Postfix) with ESMTP id E820D8D000A for ; Mon, 8 Mar 2021 03:37:33 -0500 (EST) Received: from smtpin30.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay05.hostedemail.com (Postfix) with ESMTP id A795018039006 for ; Mon, 8 Mar 2021 08:37:33 +0000 (UTC) X-FDA: 77896053186.30.3D6F3FF Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [63.128.21.124]) by imf26.hostedemail.com (Postfix) with ESMTP id E3BBD407F8F3 for ; Mon, 8 Mar 2021 08:37:30 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1615192652; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=VSeZk3InN/9j9fIXKRvmBsyCrTSCQ5Gd7zaa1tggq28=; b=YCBMDaWTyTqyPEwWHRpFGcYD/dCouA21QRN5PqpKNHYdul0MGmQak5JliHTbbHpmqju42m nmEHig6ugBmlTpIEZzz5X5pIaQeQXkqJB7H4RGP5GSKMeQLkD5ISMIMreDJen2g4+1VuTh ayzCAc/sLE1R1RNKGWRsjtyKudNvw7U= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-150-pNJuItIwPwCPGRE46QMh_g-1; Mon, 08 Mar 2021 03:37:25 -0500 X-MC-Unique: pNJuItIwPwCPGRE46QMh_g-1 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 3FF3E804333; Mon, 8 Mar 2021 08:37:24 +0000 (UTC) Received: from [10.36.113.123] (ovpn-113-123.ams2.redhat.com [10.36.113.123]) by smtp.corp.redhat.com (Postfix) with ESMTP id C94C861F2B; Mon, 8 Mar 2021 08:37:21 +0000 (UTC) To: Anshuman Khandual , linux-mm@kvack.org Cc: Russell King , Catalin Marinas , Will Deacon , Andrew Morton , Mike Rapoport , linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org References: <1615174073-10520-1-git-send-email-anshuman.khandual@arm.com> From: David Hildenbrand Organization: Red Hat GmbH Subject: Re: [RFC] mm: Enable generic pfn_valid() to handle early sections with memmap holes Message-ID: <745496f5-e099-8780-e42e-f347b55e8476@redhat.com> Date: Mon, 8 Mar 2021 09:37:20 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.8.0 MIME-Version: 1.0 In-Reply-To: <1615174073-10520-1-git-send-email-anshuman.khandual@arm.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-Stat-Signature: 1j5j1jtu4w9q1pfr9ka3z1u4pp73rjek X-Rspamd-Server: rspam02 X-Rspamd-Queue-Id: E3BBD407F8F3 Received-SPF: none (redhat.com>: No applicable sender policy available) receiver=imf26; identity=mailfrom; envelope-from=""; helo=us-smtp-delivery-124.mimecast.com; client-ip=63.128.21.124 X-HE-DKIM-Result: pass/pass X-HE-Tag: 1615192650-348673 Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On 08.03.21 04:27, Anshuman Khandual wrote: > Platforms like arm and arm64 have redefined pfn_valid() because their e= arly > memory sections might have contained memmap holes caused by memblock ar= eas > tagged with MEMBLOCK_NOMAP, which should be skipped while validating a = pfn > for struct page backing. This scenario could be captured with a new opt= ion > CONFIG_HAVE_EARLY_SECTION_MEMMAP_HOLES and then generic pfn_valid() can= be > improved to accommodate such platforms. This reduces overall code footp= rint > and also improves maintainability. >=20 > Commit 4f5b0c178996 ("arm, arm64: move free_unused_memmap() to generic = mm") > had used CONFIG_HAVE_ARCH_PFN_VALID to gate free_unused_memmap(), which= in > turn had expanded its scope to new platforms like arc and m68k. Rather = lets > restrict back the scope for free_unused_memmap() to arm and arm64 platf= orms > using this new config option i.e CONFIG_HAVE_EARLY_SECTION_MEMMAP. >=20 > While here, it exports the symbol memblock_is_map_memory() to build dri= vers > that depend on pfn_valid() but does not have the required visibility. A= fter > this new config is in place, just drop CONFIG_HAVE_ARCH_PFN_VALID from = both > arm and arm64 platforms. >=20 > Cc: Russell King > Cc: Catalin Marinas > Cc: Will Deacon > Cc: Andrew Morton > Cc: Mike Rapoport > Cc: David Hildenbrand > Cc: linux-arm-kernel@lists.infradead.org > Cc: linux-kernel@vger.kernel.org > Cc: linux-mm@kvack.org > Suggested-by: David Hildenbrand > Signed-off-by: Anshuman Khandual > --- > This applies on 5.12-rc2 along with arm64 pfn_valid() fix patches [1] a= nd > has been lightly tested on the arm64 platform. The idea to represent th= is > unique situation on the arm and arm64 platforms with a config option wa= s > proposed by David H during an earlier discussion [2]. This still does n= ot > build on arm platform due to pfn_valid() resolution errors. Nonetheless > wanted to get some early feedback whether the overall approach here, is > acceptable or not. It might make sense to keep the arm variant for now. The arm64 variant=20 is where the magic happens and where we missed updates when working on=20 the generic variant. The generic variant really only applies to 64bit targets where we have=20 SPARSEMEM. See x86 as an example. [...] > /* > diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h > index 47946cec7584..93532994113f 100644 > --- a/include/linux/mmzone.h > +++ b/include/linux/mmzone.h > @@ -1409,8 +1409,23 @@ static inline int pfn_section_valid(struct mem_s= ection *ms, unsigned long pfn) > } > #endif > =20 > +bool memblock_is_map_memory(phys_addr_t addr); > + > #ifndef CONFIG_HAVE_ARCH_PFN_VALID > static inline int pfn_valid(unsigned long pfn) > +{ > + phys_addr_t addr =3D PFN_PHYS(pfn); > + > + /* > + * Ensure the upper PAGE_SHIFT bits are clear in the > + * pfn. Else it might lead to false positives when > + * some of the upper bits are set, but the lower bits > + * match a valid pfn. > + */ > + if (PHYS_PFN(addr) !=3D pfn) > + return 0; I think this should be fine for other archs as well. > + > +#ifdef CONFIG_SPARSEMEM Why do we need the ifdef now? If that's to cover the arm case, then=20 please consider the arm64 case only for now. > { > struct mem_section *ms; > =20 > @@ -1423,7 +1438,14 @@ static inline int pfn_valid(unsigned long pfn) > * Traditionally early sections always returned pfn_valid() for > * the entire section-sized span. > */ > - return early_section(ms) || pfn_section_valid(ms, pfn); > + if (early_section(ms)) > + return IS_ENABLED(CONFIG_HAVE_EARLY_SECTION_MEMMAP_HOLES) ? > + memblock_is_map_memory(pfn << PAGE_SHIFT) : 1; > + > + return pfn_section_valid(ms, pfn); > +} > +#endif > + return 1; > } > #endif > =20 > diff --git a/mm/Kconfig b/mm/Kconfig > index 24c045b24b95..0ec20f661b3f 100644 > --- a/mm/Kconfig > +++ b/mm/Kconfig > @@ -135,6 +135,16 @@ config HAVE_FAST_GUP > config ARCH_KEEP_MEMBLOCK > bool > =20 > +config HAVE_EARLY_SECTION_MEMMAP_HOLES > + depends on ARCH_KEEP_MEMBLOCK && SPARSEMEM_VMEMMAP > + def_bool n > + help > + Early sections on certain platforms might have portions which are > + not backed with struct page mapping as their memblock entries are > + marked with MEMBLOCK_NOMAP. When subscribed, this option enables > + specific handling for those memory sections in certain situations > + such as pfn_valid(). > + > # Keep arch NUMA mapping infrastructure post-init. > config NUMA_KEEP_MEMINFO > bool > diff --git a/mm/memblock.c b/mm/memblock.c > index afaefa8fc6ab..d9fa2e62ab7a 100644 > --- a/mm/memblock.c > +++ b/mm/memblock.c > @@ -1744,6 +1744,7 @@ bool __init_memblock memblock_is_map_memory(phys_= addr_t addr) > return false; > return !memblock_is_nomap(&memblock.memory.regions[i]); > } > +EXPORT_SYMBOL(memblock_is_map_memory); > =20 > int __init_memblock memblock_search_pfn_nid(unsigned long pfn, > unsigned long *start_pfn, unsigned long *end_pfn) > @@ -1926,7 +1927,7 @@ static void __init free_unused_memmap(void) > unsigned long start, end, prev_end =3D 0; > int i; > =20 > - if (!IS_ENABLED(CONFIG_HAVE_ARCH_PFN_VALID) || > + if (!IS_ENABLED(CONFIG_HAVE_EARLY_SECTION_MEMMAP_HOLES) || > IS_ENABLED(CONFIG_SPARSEMEM_VMEMMAP)) > return; > =20 >=20 With commit 1f90a3477df3ff1a91e064af554cdc887c8f9e5e Author: Dan Williams Date: Thu Feb 25 17:17:05 2021 -0800 mm: teach pfn_to_online_page() about ZONE_DEVICE section collisions (still in -next I think) You'll also have to take care of pfn_to_online_page(). --=20 Thanks, David / dhildenb