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 Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9555EECAAA1 for ; Fri, 16 Sep 2022 19:39:22 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id D63558D0003; Fri, 16 Sep 2022 15:39:21 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id D12418D0001; Fri, 16 Sep 2022 15:39:21 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id C01AB8D0003; Fri, 16 Sep 2022 15:39:21 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0016.hostedemail.com [216.40.44.16]) by kanga.kvack.org (Postfix) with ESMTP id B23268D0001 for ; Fri, 16 Sep 2022 15:39:21 -0400 (EDT) Received: from smtpin05.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay02.hostedemail.com (Postfix) with ESMTP id 7E6C71207BB for ; Fri, 16 Sep 2022 19:39:21 +0000 (UTC) X-FDA: 79918962522.05.3FC9A38 Received: from casper.infradead.org (casper.infradead.org [90.155.50.34]) by imf02.hostedemail.com (Postfix) with ESMTP id 8CFD8800B5 for ; Fri, 16 Sep 2022 19:39:20 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=infradead.org; s=casper.20170209; h=In-Reply-To:Content-Type:MIME-Version: References:Message-ID:Subject:Cc:To:From:Date:Sender:Reply-To: Content-Transfer-Encoding:Content-ID:Content-Description; bh=ypkdyUP1JcI6s7cVs4ktcAodUubTPEOwr3HQWpEnFKY=; b=Pw3hxWjjBMOCUduKA/HxQ14yPx EtpjxuqhNVlO7xUAG1GiuC9mpTmG/QTVPro7HxagyGxN9VkBlmLvx7NMHh9wSumjLFYepfiELs9Cd L/F8LVbEUlI/SOg273mukg4YjcT9UvosbxETburGeC8B4GIgZIn7VUaChHcA6m/iOWjes4AkxKI+U bimitz6WRDlt12TsreTRphNaObbIwWrzmCYNmcmsdtqtqwKy36NMaDH9g55Q+UE+jHuQbArDIH+mt EHMkZ3fvkf3sOqfueGzEP/a22mk7pmGwjlYP7tZ7dsmsMc08W0SHQAwpqWR80JJbmUFra+g5P6dGf gh9aVtwA==; Received: from willy by casper.infradead.org with local (Exim 4.94.2 #2 (Red Hat Linux)) id 1oZGo8-002Xle-ME; Fri, 16 Sep 2022 19:15:24 +0000 Date: Fri, 16 Sep 2022 20:15:24 +0100 From: Matthew Wilcox To: Kees Cook Cc: Uladzislau Rezki , Andrew Morton , Yu Zhao , dev@der-flo.net, linux-mm@kvack.org, linux-hardening@vger.kernel.org, Peter Zijlstra , Ingo Molnar , linux-kernel@vger.kernel.org, x86@kernel.org, linux-perf-users@vger.kernel.org, linux-arch@vger.kernel.org Subject: Re: [PATCH 3/3] usercopy: Add find_vmap_area_try() to avoid deadlocks Message-ID: References: <20220916135953.1320601-1-keescook@chromium.org> <20220916135953.1320601-4-keescook@chromium.org> <202209160805.CA47B2D673@keescook> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <202209160805.CA47B2D673@keescook> ARC-Authentication-Results: i=1; imf02.hostedemail.com; dkim=pass header.d=infradead.org header.s=casper.20170209 header.b=Pw3hxWjj; spf=none (imf02.hostedemail.com: domain of willy@infradead.org has no SPF policy when checking 90.155.50.34) smtp.mailfrom=willy@infradead.org; dmarc=none ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1663357161; a=rsa-sha256; cv=none; b=nxeGgd9NV/lQzG19GZiYNGbhPY99Muz2UJF7jeSzjtysIo8n+K3fpszgbypVJdAGQmS1Wf Py24PhH0+0uXn5a0PAHIsvqfHHpNrxEHUg9GZDg2ZICsj6EnSAGmnP8GZQmKm0Yq8DC1Zn gjW4Bua+44kVM4qwoRlAEGd0GlXhm1s= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1663357161; h=from:from:sender: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: in-reply-to:in-reply-to:references:references:dkim-signature; bh=ypkdyUP1JcI6s7cVs4ktcAodUubTPEOwr3HQWpEnFKY=; b=0bniR8nb9ZpVEOEQ6c9SCaff1Go6E7FpV2XV2c4zdrqaAyqAU2wLU+mozICLiRPqPEX6Xm vdNrRH2CRFvsSV3LbhH/sG8gtWCs3eG3WEstN1lFd65BJizHcOtpSVRJnW6V2T4Z5bhqE3 Z/dvstw5q9eknRGzStuFEOtuudyhE5g= X-Rspam-User: X-Rspamd-Server: rspam07 X-Rspamd-Queue-Id: 8CFD8800B5 Authentication-Results: imf02.hostedemail.com; dkim=pass header.d=infradead.org header.s=casper.20170209 header.b=Pw3hxWjj; spf=none (imf02.hostedemail.com: domain of willy@infradead.org has no SPF policy when checking 90.155.50.34) smtp.mailfrom=willy@infradead.org; dmarc=none X-Stat-Signature: 4g4dk37n43sxxxzp6hcqwhxmyqcuetd5 X-HE-Tag: 1663357160-536996 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 Fri, Sep 16, 2022 at 08:09:16AM -0700, Kees Cook wrote: > On Fri, Sep 16, 2022 at 03:46:07PM +0100, Matthew Wilcox wrote: > > On Fri, Sep 16, 2022 at 06:59:57AM -0700, Kees Cook wrote: > > > The check_object_size() checks under CONFIG_HARDENED_USERCOPY need to be > > > more defensive against running from interrupt context. Use a best-effort > > > check for VMAP areas when running in interrupt context > > > > I had something more like this in mind: > > Yeah, I like -EAGAIN. I'd like to keep the interrupt test to choose lock > vs trylock, otherwise it's trivial to bypass the hardening test by having > all the other CPUs beating on the spinlock. I was thinking about this: +++ b/mm/vmalloc.c @@ -1844,12 +1844,19 @@ { struct vmap_area *va; - if (!spin_lock(&vmap_area_lock)) - return ERR_PTR(-EAGAIN); + /* + * It's safe to walk the rbtree under the RCU lock, but we may + * incorrectly find no vmap_area if the tree is being modified. + */ + rcu_read_lock(); va = __find_vmap_area(addr, &vmap_area_root); - spin_unlock(&vmap_area_lock); + if (!va && in_interrupt()) + va = ERR_PTR(-EAGAIN); + rcu_read_unlock(); - return va; + if (va) + return va; + return find_vmap_area(addr); } /*** Per cpu kva allocator ***/ ... but I don't think that works since vmap_areas aren't freed by RCU, and I think they're reused without going through an RCU cycle. So here's attempt #4, which actually compiles, and is, I think, what you had in mind. diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h index 096d48aa3437..2b7c52e76856 100644 --- a/include/linux/vmalloc.h +++ b/include/linux/vmalloc.h @@ -215,7 +215,7 @@ extern struct vm_struct *__get_vm_area_caller(unsigned long size, void free_vm_area(struct vm_struct *area); extern struct vm_struct *remove_vm_area(const void *addr); extern struct vm_struct *find_vm_area(const void *addr); -struct vmap_area *find_vmap_area(unsigned long addr); +struct vmap_area *find_vmap_area_try(unsigned long addr); static inline bool is_vm_area_hugepages(const void *addr) { diff --git a/mm/usercopy.c b/mm/usercopy.c index c1ee15a98633..e0fb605c1b38 100644 --- a/mm/usercopy.c +++ b/mm/usercopy.c @@ -173,7 +173,11 @@ static inline void check_heap_object(const void *ptr, unsigned long n, } if (is_vmalloc_addr(ptr)) { - struct vmap_area *area = find_vmap_area(addr); + struct vmap_area *area = find_vmap_area_try(addr); + + /* We may be in NMI context */ + if (area == ERR_PTR(-EAGAIN)) + return; if (!area) usercopy_abort("vmalloc", "no area", to_user, 0, n); diff --git a/mm/vmalloc.c b/mm/vmalloc.c index dd6cdb201195..c47b3b5d1c2d 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -1829,7 +1829,7 @@ static void free_unmap_vmap_area(struct vmap_area *va) free_vmap_area_noflush(va); } -struct vmap_area *find_vmap_area(unsigned long addr) +static struct vmap_area *find_vmap_area(unsigned long addr) { struct vmap_area *va; @@ -1840,6 +1840,26 @@ struct vmap_area *find_vmap_area(unsigned long addr) return va; } +/* + * The vmap_area_lock is not interrupt-safe, and we can end up here from + * NMI context, so it's not worth even trying to make it IRQ-safe. + */ +struct vmap_area *find_vmap_area_try(unsigned long addr) +{ + struct vmap_area *va; + + if (in_interrupt()) { + if (!spin_trylock(&vmap_area_lock)) + return ERR_PTR(-EAGAIN); + } else { + spin_lock(&vmap_area_lock); + } + va = __find_vmap_area(addr, &vmap_area_root); + spin_unlock(&vmap_area_lock); + + return va; +} + /*** Per cpu kva allocator ***/ /*