From patchwork Mon Mar 29 05:29:22 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ilya Lipnitskiy X-Patchwork-Id: 412831 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=-15.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=unavailable 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 E25DDC433DB for ; Mon, 29 Mar 2021 05:30:19 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id B6AD861960 for ; Mon, 29 Mar 2021 05:30:19 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230331AbhC2F3s (ORCPT ); Mon, 29 Mar 2021 01:29:48 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:39478 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230334AbhC2F32 (ORCPT ); Mon, 29 Mar 2021 01:29:28 -0400 Received: from mail-pj1-x1035.google.com (mail-pj1-x1035.google.com [IPv6:2607:f8b0:4864:20::1035]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 53E41C061574; Sun, 28 Mar 2021 22:29:27 -0700 (PDT) Received: by mail-pj1-x1035.google.com with SMTP id w8so5450497pjf.4; Sun, 28 Mar 2021 22:29:27 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=nwW3wqGlwvgCl7fK73mXwZYZab2mQtuORLVwBoAbxgY=; b=qDos5l8euTHvWyys+1Vbt06YDZopfjoD9EcakCMdojFTtzovx7hLCAEQS4vRpkUyvJ x3qeS8Q/L6n4GeatU2IQfCR2Nd5/5HSoiJ8o3UdYXI0qrrN/7V7nTLFdtBe/khMr8wmH 0ZNw72Q39445uDG9f+62zDK+4UImr0LprtNmr2Z846j90bPrwpO68riVvWqjIW4cpaV9 6TV3P0JrRSfeQrmxC+5S+Rr1Wi+Ocn2rNO7fpLsdex50yJQvrT/wO8w9L5OBeWLTp+1g xUPeIR2ehpP2L77FbwuM3689Asewnp2fGIRdu2rUuzqQXAarSam9HKVByVYSDMUyCD1y 14Zw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=nwW3wqGlwvgCl7fK73mXwZYZab2mQtuORLVwBoAbxgY=; b=Fy4WFuUaEjg2Ha04vhqX1quuJtEBzkWCR3BJg3pHVdxoQjEf6XyN6mWW4fZsUW49a1 ncnwoqmVPqzD3A4/G6PGRuMwBj4o7RNUlIXcFkZsL/9s+C0w/m7Fw9cOO9Sgp48vxS3b IhGr+yshqNh9PBb5y1JZ26Sx2awYsbndEY0yizVxlNABZc1yIX9I+mVrew8mxOipfhYz J6tdbLH6dI2ya1OxZRlXCrSCm/IzrpJsyukiUguE0ltd8fsj29Lt5TAEwjpsP7xswSRo pU6les0edH7NaUCnjCsHZM3AHhxVJ7vCmILz2dVvPms9QujU2oR7xEK6xj5LeTxRQl+F i5hw== X-Gm-Message-State: AOAM530qGO/Q/iTxfLT4WH41ILsonUqAGvpt94yDxWdkP7Kyvrhn8LmH JYGGH9cayyGKeKbJKAzEfYc= X-Google-Smtp-Source: ABdhPJxSp0W3ypZlEfQWK22Bf8igEvCO5VyssScOVHbZclh1AP0AIw0oNX+mIFRaapWFIcMO4Retdw== X-Received: by 2002:a17:90b:94c:: with SMTP id dw12mr24734883pjb.119.1616995766893; Sun, 28 Mar 2021 22:29:26 -0700 (PDT) Received: from z640-arch.lan ([2602:61:7344:f100::678]) by smtp.gmail.com with ESMTPSA id y9sm15897460pgc.9.2021.03.28.22.29.26 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Sun, 28 Mar 2021 22:29:26 -0700 (PDT) From: Ilya Lipnitskiy To: Andrew Morton , linux-mm@kvack.org, linux-kernel@vger.kernel.org Cc: Ilya Lipnitskiy , "Eric W. Biederman" , stable@vger.kernel.org Subject: [PATCH v2] mm: fix race by making init_zero_pfn() early_initcall Date: Sun, 28 Mar 2021 22:29:22 -0700 Message-Id: <20210329052922.1130493-1-ilya.lipnitskiy@gmail.com> X-Mailer: git-send-email 2.31.0 In-Reply-To: <20210329052426.1130040-1-ilya.lipnitskiy@gmail.com> References: <20210329052426.1130040-1-ilya.lipnitskiy@gmail.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: stable@vger.kernel.org There are code paths that rely on zero_pfn to be fully initialized before core_initcall. For example, wq_sysfs_init() is a core_initcall function that eventually results in a call to kernel_execve, which causes a page fault with a subsequent mmput. If zero_pfn is not initialized by then it may not get cleaned up properly and result in an error: BUG: Bad rss-counter state mm:(ptrval) type:MM_ANONPAGES val:1 Here is an analysis of the race as seen on a MIPS device. On this particular MT7621 device (Ubiquiti ER-X), zero_pfn is PFN 0 until initialized, at which point it becomes PFN 5120: 1. wq_sysfs_init calls into kobject_uevent_env at core_initcall: [<80340dc8>] kobject_uevent_env+0x7e4/0x7ec [<8033f8b8>] kset_register+0x68/0x88 [<803cf824>] bus_register+0xdc/0x34c [<803cfac8>] subsys_virtual_register+0x34/0x78 [<8086afb0>] wq_sysfs_init+0x1c/0x4c [<80001648>] do_one_initcall+0x50/0x1a8 [<8086503c>] kernel_init_freeable+0x230/0x2c8 [<8066bca0>] kernel_init+0x10/0x100 [<80003038>] ret_from_kernel_thread+0x14/0x1c 2. kobject_uevent_env() calls call_usermodehelper_exec() which executes kernel_execve asynchronously. 3. Memory allocations in kernel_execve cause a page fault, bumping the MM reference counter: [<8015adb4>] add_mm_counter_fast+0xb4/0xc0 [<80160d58>] handle_mm_fault+0x6e4/0xea0 [<80158aa4>] __get_user_pages.part.78+0x190/0x37c [<8015992c>] __get_user_pages_remote+0x128/0x360 [<801a6d9c>] get_arg_page+0x34/0xa0 [<801a7394>] copy_string_kernel+0x194/0x2a4 [<801a880c>] kernel_execve+0x11c/0x298 [<800420f4>] call_usermodehelper_exec_async+0x114/0x194 4. In case zero_pfn has not been initialized yet, zap_pte_range does not decrement the MM_ANONPAGES RSS counter and the BUG message is triggered shortly afterwards when __mmdrop checks the ref counters: [<800285e8>] __mmdrop+0x98/0x1d0 [<801a6de8>] free_bprm+0x44/0x118 [<801a86a8>] kernel_execve+0x160/0x1d8 [<800420f4>] call_usermodehelper_exec_async+0x114/0x194 [<80003198>] ret_from_kernel_thread+0x14/0x1c To avoid races such as described above, initialize init_zero_pfn at early_initcall level. Depending on the architecture, ZERO_PAGE is either constant or gets initialized even earlier, at paging_init, so there is no issue with initializing zero_pfn earlier. ML discussion: https://lore.kernel.org/lkml/CALCv0x2YqOXEAy2Q=hafjhHCtTHVodChv1qpM=niAXOpqEbt7w@mail.gmail.com/ Signed-off-by: Ilya Lipnitskiy Cc: "Eric W. Biederman" Cc: stable@vger.kernel.org --- mm/memory.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/memory.c b/mm/memory.c index 46ef306375bd..a8bbc4fc121f 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -166,7 +166,7 @@ static int __init init_zero_pfn(void) zero_pfn = page_to_pfn(ZERO_PAGE(0)); return 0; } -core_initcall(init_zero_pfn); +early_initcall(init_zero_pfn); void mm_trace_rss_stat(struct mm_struct *mm, int member, long count) {