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=-10.3 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,MENTIONS_GIT_HOSTING, NICE_REPLY_A,SPF_HELO_NONE,SPF_PASS,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 E1169C433ED for ; Fri, 16 Apr 2021 16:00:32 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 8D815613BB for ; Fri, 16 Apr 2021 16:00:32 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 8D815613BB Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=arm.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 1E59E6B0070; Fri, 16 Apr 2021 12:00:32 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 196DB6B0071; Fri, 16 Apr 2021 12:00:32 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 0362D6B0072; Fri, 16 Apr 2021 12:00:31 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0042.hostedemail.com [216.40.44.42]) by kanga.kvack.org (Postfix) with ESMTP id DBE106B0070 for ; Fri, 16 Apr 2021 12:00:31 -0400 (EDT) Received: from smtpin38.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay04.hostedemail.com (Postfix) with ESMTP id 951DA3620 for ; Fri, 16 Apr 2021 16:00:31 +0000 (UTC) X-FDA: 78038692662.38.FA8D5CC Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by imf18.hostedemail.com (Postfix) with ESMTP id F00A2200098C for ; Fri, 16 Apr 2021 16:00:31 +0000 (UTC) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id DCBA011B3; Fri, 16 Apr 2021 09:00:29 -0700 (PDT) Received: from [192.168.1.179] (unknown [172.31.20.19]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 098F13F99C; Fri, 16 Apr 2021 09:00:27 -0700 (PDT) Subject: Re: [PATCH v1 3/5] mm: ptdump: Provide page size to notepage() To: Christophe Leroy , Benjamin Herrenschmidt , Paul Mackerras , Michael Ellerman , akpm@linux-foundation.org Cc: linux-arch@vger.kernel.org, linux-s390@vger.kernel.org, x86@kernel.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, linux-riscv@lists.infradead.org, linuxppc-dev@lists.ozlabs.org, linux-arm-kernel@lists.infradead.org References: <1ef6b954fb7b0f4dfc78820f1e612d2166c13227.1618506910.git.christophe.leroy@csgroup.eu> <41819925-3ee5-4771-e98b-0073e8f095cf@arm.com> <1102cda1-b00f-b6ef-6bf3-22068cc11510@arm.com> <6ff4816b-8ff6-19de-73a2-3fcadc003ccd@csgroup.eu> <10adad00-14de-61b6-ce2a-bdde23a34bcf@csgroup.eu> From: Steven Price Message-ID: Date: Fri, 16 Apr 2021 17:00:22 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.7.1 MIME-Version: 1.0 In-Reply-To: <10adad00-14de-61b6-ce2a-bdde23a34bcf@csgroup.eu> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-GB X-Rspamd-Server: rspam03 X-Rspamd-Queue-Id: F00A2200098C X-Stat-Signature: jib8mxjrzfmd9jnwriqwzj1a1919prye Received-SPF: none (arm.com>: No applicable sender policy available) receiver=imf18; identity=mailfrom; envelope-from=""; helo=foss.arm.com; client-ip=217.140.110.172 X-HE-DKIM-Result: none/none X-HE-Tag: 1618588831-453883 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 16/04/2021 16:15, Christophe Leroy wrote: >=20 >=20 > Le 16/04/2021 =C3=A0 17:04, Christophe Leroy a =C3=A9crit=C2=A0: >> >> >> Le 16/04/2021 =C3=A0 16:40, Christophe Leroy a =C3=A9crit=C2=A0: >>> >>> >>> Le 16/04/2021 =C3=A0 15:00, Steven Price a =C3=A9crit=C2=A0: >>>> On 16/04/2021 12:08, Christophe Leroy wrote: >>>>> >>>>> >>>>> Le 16/04/2021 =C3=A0 12:51, Steven Price a =C3=A9crit=C2=A0: >>>>>> On 16/04/2021 11:38, Christophe Leroy wrote: >>>>>>> >>>>>>> >>>>>>> Le 16/04/2021 =C3=A0 11:28, Steven Price a =C3=A9crit=C2=A0: >>>>>>>> To be honest I don't fully understand why powerpc requires the=20 >>>>>>>> page_size - it appears to be using it purely to find "holes" in=20 >>>>>>>> the calls to note_page(), but I haven't worked out why such=20 >>>>>>>> holes would occur. >>>>>>> >>>>>>> I was indeed introduced for KASAN. We have a first commit=20 >>>>>>> https://github.com/torvalds/linux/commit/cabe8138 which uses page= =20 >>>>>>> size to detect whether it is a KASAN like stuff. >>>>>>> >>>>>>> Then came https://github.com/torvalds/linux/commit/b00ff6d8c as a= =20 >>>>>>> fix. I can't remember what the problem was exactly, something=20 >>>>>>> around the use of hugepages for kernel memory, came as part of=20 >>>>>>> the series=20 >>>>>>> https://patchwork.ozlabs.org/project/linuxppc-dev/cover/cover.158= 9866984.git.christophe.leroy@csgroup.eu/=20 >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Ah, that's useful context. So it looks like powerpc took a=20 >>>>>> different route to reducing the KASAN output to x86. >>>>>> >>>>>> Given the generic ptdump code has handling for KASAN already it=20 >>>>>> should be possible to drop that from the powerpc arch code, which=20 >>>>>> I think means we don't actually need to provide page size to=20 >>>>>> notepage(). Hopefully that means more code to delete ;) >>>>>> >>>>> >>>>> Yes ... and no. >>>>> >>>>> It looks like the generic ptdump handles the case when several=20 >>>>> pgdir entries points to the same kasan_early_shadow_pte. But it=20 >>>>> doesn't take into account the powerpc case where we have regular=20 >>>>> page tables where several (if not all) PTEs are pointing to the=20 >>>>> kasan_early_shadow_page . >>>> >>>> I'm not sure I follow quite how powerpc is different here. But could= =20 >>>> you have a similar check for PTEs against kasan_early_shadow_pte as=20 >>>> the other levels already have? >>>> >>>> I'm just worried that page_size isn't well defined in this interface= =20 >>>> and it's going to cause problems in the future. >>>> >>> >>> I'm trying. I reverted the two commits b00ff6d8c and cabe8138. >>> >>> At the moment, I don't get exactly what I expect: For linear memory I= =20 >>> get one line for each 8M page whereas before reverting the patches I=20 >>> got one 16M line and one 112M line. >>> >>> And for KASAN shadow area I get two lines for the 2x 8M pages=20 >>> shadowing linear mem then I get one 4M line for each PGDIR entry=20 >>> pointing to kasan_early_shadow_pte. >>> >>> 0xf8000000-0xf87fffff 0x07000000=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 8M=C2=A0=C2=A0 huge=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 r= w =20 >>> present >>> 0xf8800000-0xf8ffffff 0x07800000=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 8M=C2=A0=C2=A0 huge=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 r= w =20 >>> present >>> 0xf9000000-0xf93fffff 0x01430000=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 4M=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 r =20 >>> present >> ... >>> 0xfec00000-0xfeffffff 0x01430000=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0 4M=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 r =20 >>> present >>> >>> Any idea ? >>> >> >> >> I think the different with other architectures is here: >> >> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0} else if (flag !=3D st->current_flags |= | level !=3D st->level || >> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 add= r >=3D st->marker[1].start_address || >> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 pa = !=3D st->last_pa + PAGE_SIZE) { >> >> >> In addition to the checks everyone do, powerpc also checks "pa !=3D=20 >> st->last_pa + PAGE_SIZE". >> And it is definitely for that test that page_size argument add been=20 >> added. >=20 > By replacing that test by (pa - st->start_pa !=3D addr -=20 > st->start_address) it works again. So we definitely don't need the real= =20 > page size. Yes that should work. Thanks for figuring it out! >=20 >> >> I see that other architectures except RISCV don't dump the physical=20 >> address. But even RISCV doesn't include that check. Yes not having the physical address certainly simplifies things -=20 although I can see why that can be handy to see. The disadvantage is=20 that user space or vmalloc()'d memory will produce a lot of output=20 because the physical addresses are unlikely to be contiguous. And for=20 most uses you don't need the information. >> That physical address dump was added by commit aaa229529244=20 >> ("powerpc/mm: Add physical address to Linux page table dump")=20 >> [https://github.com/torvalds/linux/commit/aaa2295] >> >> How do other architectures deal with the problem described by the=20 >> commit log of that patch ? AFAIK other architectures are "broken" in this regard. In practice I=20 don't think it often causes an issue though. Steve