From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e28smtp07.in.ibm.com (e28smtp07.in.ibm.com [122.248.162.7]) (using TLSv1 with cipher CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 8D2C31A0010 for ; Mon, 11 May 2015 18:54:21 +1000 (AEST) Received: from /spool/local by e28smtp07.in.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Mon, 11 May 2015 14:24:18 +0530 Received: from d28relay03.in.ibm.com (d28relay03.in.ibm.com [9.184.220.60]) by d28dlp03.in.ibm.com (Postfix) with ESMTP id 104B21258056 for ; Mon, 11 May 2015 14:26:26 +0530 (IST) Received: from d28av05.in.ibm.com (d28av05.in.ibm.com [9.184.220.67]) by d28relay03.in.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id t4B8sGCb39518396 for ; Mon, 11 May 2015 14:24:16 +0530 Received: from d28av05.in.ibm.com (localhost [127.0.0.1]) by d28av05.in.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id t4B8sFJn032747 for ; Mon, 11 May 2015 14:24:15 +0530 From: "Aneesh Kumar K.V" To: "Kirill A. Shutemov" Subject: Re: [PATCH V3] powerpc/thp: Serialize pmd clear against a linux page table walk. In-Reply-To: <20150511074631.GA10974@node.dhcp.inet.fi> References: <1431325561-21396-1-git-send-email-aneesh.kumar@linux.vnet.ibm.com> <20150511074631.GA10974@node.dhcp.inet.fi> Date: Mon, 11 May 2015 14:24:14 +0530 Message-ID: <87twvj4hqh.fsf@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain Cc: aarcange@redhat.com, linux-kernel@vger.kernel.org, linux-mm@kvack.org, paulus@samba.org, akpm@linux-foundation.org, linuxppc-dev@lists.ozlabs.org, kirill.shutemov@linux.intel.com List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , "Kirill A. Shutemov" writes: > On Mon, May 11, 2015 at 11:56:01AM +0530, Aneesh Kumar K.V wrote: >> Serialize against find_linux_pte_or_hugepte which does lock-less >> lookup in page tables with local interrupts disabled. For huge pages >> it casts pmd_t to pte_t. Since format of pte_t is different from >> pmd_t we want to prevent transit from pmd pointing to page table >> to pmd pointing to huge page (and back) while interrupts are disabled. >> We clear pmd to possibly replace it with page table pointer in >> different code paths. So make sure we wait for the parallel >> find_linux_pte_or_hugepage to finish. >> >> Without this patch, a find_linux_pte_or_hugepte running in parallel to >> __split_huge_zero_page_pmd or do_huge_pmd_wp_page_fallback or zap_huge_pmd >> can run into the above issue. With __split_huge_zero_page_pmd and >> do_huge_pmd_wp_page_fallback we clear the hugepage pte before inserting >> the pmd entry with a regular pgtable address. Such a clear need to >> wait for the parallel find_linux_pte_or_hugepte to finish. >> >> With zap_huge_pmd, we can run into issues, with a hugepage pte >> getting zapped due to a MADV_DONTNEED while other cpu fault it >> in as small pages. >> >> Reported-by: Kirill A. Shutemov >> Signed-off-by: Aneesh Kumar K.V > > Reviewed-by: Kirill A. Shutemov > > CC: stable@ ? Yes, We also need to pick, dac5657067919161eb3273ca787d8ae9814801e7 691e95fd7396905a38d98919e9c150dbc3ea21a3 7d6e7f7ffaba4e013c7a0589140431799bc17985 But that may need me to a backport, because we have dependencies in kvm and a cherry-pick may not work. Will work with Michael Ellerman to find out what needs to be done. -aneesh