From patchwork Wed Jul 25 23:59:15 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Omar Sandoval X-Patchwork-Id: 10544941 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id CBB9A1822 for ; Wed, 25 Jul 2018 23:59:35 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id BA69A2A9AC for ; Wed, 25 Jul 2018 23:59:35 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id AECF22A9CE; Wed, 25 Jul 2018 23:59:35 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.9 required=2.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,MAILING_LIST_MULTI,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 3021B2A9AC for ; Wed, 25 Jul 2018 23:59:35 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728587AbeGZBNh (ORCPT ); Wed, 25 Jul 2018 21:13:37 -0400 Received: from mail-pl0-f68.google.com ([209.85.160.68]:39538 "EHLO mail-pl0-f68.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728486AbeGZBNg (ORCPT ); Wed, 25 Jul 2018 21:13:36 -0400 Received: by mail-pl0-f68.google.com with SMTP id m1-v6so3945878plt.6 for ; Wed, 25 Jul 2018 16:59:32 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=osandov-com.20150623.gappssmtp.com; s=20150623; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=jejP2yrNUB7zHbhkc1Ji+iKVjXdsuDSSW0aNb0+t3w4=; b=rSaiZYTBpJvtTfER7UyiV0vgtVPMHJvNbL0d4NHd5473ZGsxdd+dYdRIj67JFVkkyg Sjbsmzvmn0TZEhSBKGKWDMpD25h0vS4uhjn9vuNu0g1Ivv1z4VvRQXP5Z4dU99R58TPV ZOIpm+jukziTkScJrGF++3tdsbgIK5Bq8JVjHIsQLmTPdS9BadKWIS/l/+p3ZmCE80Aa X4KoF1xmDIzZJ+OgfWj584/TShSJ4BsUcsxl7fUHSofEUIiBgoBR7KGEWarIom7Hr/9j fNJsVR6QWMXu0r+vLWgkNGFoq8GWm2IuFsF4sX+d7vRLq8vX6r6eA2tYtP3p5do3ieCZ cznQ== 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; bh=jejP2yrNUB7zHbhkc1Ji+iKVjXdsuDSSW0aNb0+t3w4=; b=nh2k5p8gCfKCE2vn567FdyDAOdLs+rvO6DfleOQP4tVyeGA4qvAkd9ZhSQbon5LiI7 ewb+uQH2Vc4xm5Q6yUmDPvd+uDY73wGeiSDb/doMidUWkVi4oNDfMMtL1QTJvtt8Fw2o 4lIQmE6NUuM81dTw8IAND5u/biR2pcDBvg/qoHEY5Y+z0FKOusc3w1ueoUiQLaykfv1R RJanUcMwEGkcMxKhx/S01INEXiycXEWklRAGv+HysG96ozh+uuQGPHai0kotLH+eyXOf esPQZ0yC6bmD4htZJjHhNpRLZZFrHsAMOwiUMM5zw/z+8l1XgqdddqCagqLSV/Tj8GdD isBw== X-Gm-Message-State: AOUpUlGgWwV47Vg+BSKEKkfdgteAsNlsulxq7m0fBu2s4ox5CbhZMKqI RJLWmd82rXxeJ5Z0FOg73HycJw== X-Google-Smtp-Source: AAOMgpccLSWJN9v6HHbhkX6ZmvWKOSmfWL2cGtcOoKvSutpphfb9usi+bvhnJGyT0Q+u60aS+Ty/NA== X-Received: by 2002:a17:902:ab90:: with SMTP id f16-v6mr23534126plr.182.1532563172107; Wed, 25 Jul 2018 16:59:32 -0700 (PDT) Received: from vader.thefacebook.com ([2620:10d:c090:180::1:8d38]) by smtp.gmail.com with ESMTPSA id 65-v6sm23188753pfq.81.2018.07.25.16.59.31 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Wed, 25 Jul 2018 16:59:31 -0700 (PDT) From: Omar Sandoval To: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, Andrew Morton Cc: Alexey Dobriyan , Eric Biederman , James Morse , Bhupesh Sharma , kernel-team@fb.com Subject: [PATCH v4 4/9] proc/kcore: fix memory hotplug vs multiple opens race Date: Wed, 25 Jul 2018 16:59:15 -0700 Message-Id: <0eed0c15cf94c6869fcefccae46759d702d7118f.1532563124.git.osandov@fb.com> X-Mailer: git-send-email 2.18.0 In-Reply-To: References: Sender: linux-fsdevel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-fsdevel@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP From: Omar Sandoval There's a theoretical race condition that will cause /proc/kcore to miss a memory hotplug event: CPU0 CPU1 // hotplug event 1 kcore_need_update = 1 open_kcore() open_kcore() kcore_update_ram() kcore_update_ram() // Walk RAM // Walk RAM __kcore_update_ram() __kcore_update_ram() kcore_need_update = 0 // hotplug event 2 kcore_need_update = 1 kcore_need_update = 0 Note that CPU1 set up the RAM kcore entries with the state after hotplug event 1 but cleared the flag for hotplug event 2. The RAM entries will therefore be stale until there is another hotplug event. This is an extremely unlikely sequence of events, but the fix makes the synchronization saner, anyways: we serialize the entire update sequence, which means that whoever clears the flag will always succeed in replacing the kcore list. Signed-off-by: Omar Sandoval --- fs/proc/kcore.c | 93 +++++++++++++++++++++++-------------------------- 1 file changed, 44 insertions(+), 49 deletions(-) diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index ae43a97d511d..95aa988c5b5d 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -98,53 +98,15 @@ static size_t get_kcore_size(int *nphdr, size_t *elf_buflen) return size + *elf_buflen; } -static void free_kclist_ents(struct list_head *head) -{ - struct kcore_list *tmp, *pos; - - list_for_each_entry_safe(pos, tmp, head, list) { - list_del(&pos->list); - kfree(pos); - } -} -/* - * Replace all KCORE_RAM/KCORE_VMEMMAP information with passed list. - */ -static void __kcore_update_ram(struct list_head *list) -{ - int nphdr; - size_t size; - struct kcore_list *tmp, *pos; - LIST_HEAD(garbage); - - down_write(&kclist_lock); - if (xchg(&kcore_need_update, 0)) { - list_for_each_entry_safe(pos, tmp, &kclist_head, list) { - if (pos->type == KCORE_RAM - || pos->type == KCORE_VMEMMAP) - list_move(&pos->list, &garbage); - } - list_splice_tail(list, &kclist_head); - } else - list_splice(list, &garbage); - proc_root_kcore->size = get_kcore_size(&nphdr, &size); - up_write(&kclist_lock); - - free_kclist_ents(&garbage); -} - - #ifdef CONFIG_HIGHMEM /* * If no highmem, we can assume [0...max_low_pfn) continuous range of memory * because memory hole is not as big as !HIGHMEM case. * (HIGHMEM is special because part of memory is _invisible_ from the kernel.) */ -static int kcore_update_ram(void) +static int kcore_ram_list(struct list_head *head) { - LIST_HEAD(head); struct kcore_list *ent; - int ret = 0; ent = kmalloc(sizeof(*ent), GFP_KERNEL); if (!ent) @@ -152,9 +114,8 @@ static int kcore_update_ram(void) ent->addr = (unsigned long)__va(0); ent->size = max_low_pfn << PAGE_SHIFT; ent->type = KCORE_RAM; - list_add(&ent->list, &head); - __kcore_update_ram(&head); - return ret; + list_add(&ent->list, head); + return 0; } #else /* !CONFIG_HIGHMEM */ @@ -253,11 +214,10 @@ kclist_add_private(unsigned long pfn, unsigned long nr_pages, void *arg) return 1; } -static int kcore_update_ram(void) +static int kcore_ram_list(struct list_head *list) { int nid, ret; unsigned long end_pfn; - LIST_HEAD(head); /* Not inialized....update now */ /* find out "max pfn" */ @@ -269,15 +229,50 @@ static int kcore_update_ram(void) end_pfn = node_end; } /* scan 0 to max_pfn */ - ret = walk_system_ram_range(0, end_pfn, &head, kclist_add_private); - if (ret) { - free_kclist_ents(&head); + ret = walk_system_ram_range(0, end_pfn, list, kclist_add_private); + if (ret) return -ENOMEM; + return 0; +} +#endif /* CONFIG_HIGHMEM */ + +static int kcore_update_ram(void) +{ + LIST_HEAD(list); + LIST_HEAD(garbage); + int nphdr; + size_t size; + struct kcore_list *tmp, *pos; + int ret = 0; + + down_write(&kclist_lock); + if (!xchg(&kcore_need_update, 0)) + goto out; + + ret = kcore_ram_list(&list); + if (ret) { + /* Couldn't get the RAM list, try again next time. */ + WRITE_ONCE(kcore_need_update, 1); + list_splice_tail(&list, &garbage); + goto out; + } + + list_for_each_entry_safe(pos, tmp, &kclist_head, list) { + if (pos->type == KCORE_RAM || pos->type == KCORE_VMEMMAP) + list_move(&pos->list, &garbage); + } + list_splice_tail(&list, &kclist_head); + + proc_root_kcore->size = get_kcore_size(&nphdr, &size); + +out: + up_write(&kclist_lock); + list_for_each_entry_safe(pos, tmp, &garbage, list) { + list_del(&pos->list); + kfree(pos); } - __kcore_update_ram(&head); return ret; } -#endif /* CONFIG_HIGHMEM */ /*****************************************************************************/ /*