From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ozlabs.org (ozlabs.org [103.22.144.67]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 3rmgr2026FzDqpc for ; Sat, 9 Jul 2016 16:07:30 +1000 (AEST) From: Michael Ellerman To: Kees Cook , Christoph Lameter Cc: "kernel-hardening\@lists.openwall.com" , Jan Kara , Catalin Marinas , Will Deacon , Linux-MM , sparclinux , linux-ia64@vger.kernel.org, Andrea Arcangeli , linux-arch , "x86\@kernel.org" , Russell King , PaX Team , Borislav Petkov , Mathias Krause , Fenghua Yu , Rik van Riel , David Rientjes , Tony Luck , Andy Lutomirski , Joonsoo Kim , Dmitry Vyukov , Laura Abbott , Brad Spengler , Ard Biesheuvel , LKML , Pekka Enberg , Case y Sc hauf ler , Andrew Morton , "linuxppc-dev\@lists.ozlabs.org" , "David S. Miller" , "linux-arm-kernel\@lists.infradead.org" Subject: Re: [kernel-hardening] Re: [PATCH 9/9] mm: SLUB hardened usercopy support In-Reply-To: <8737njpd37.fsf@@concordia.ellerman.id.au> References: <577f7e55.4668420a.84f17.5cb9SMTPIN_ADDED_MISSING@mx.google.com> <8737njpd37.fsf@@concordia.ellerman.id.au> Date: Sat, 09 Jul 2016 16:07:29 +1000 Message-ID: <87ziprny3i.fsf@@concordia.ellerman.id.au> MIME-Version: 1.0 Content-Type: text/plain List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Michael Ellerman writes: > Kees Cook writes: > >> On Fri, Jul 8, 2016 at 1:41 PM, Kees Cook wrote: >>> So, as found already, the position in the usercopy check needs to be >>> bumped down by red_left_pad, which is what Michael's fix does, so I'll >>> include it in the next version. >> >> Actually, after some offline chats, I think this is better, since it >> makes sure the ptr doesn't end up somewhere weird before we start the >> calculations. This leaves the pointer as-is, but explicitly handles >> the redzone on the offset instead, with no wrapping, etc: >> >> /* Find offset within object. */ >> offset = (ptr - page_address(page)) % s->size; >> >> + /* Adjust for redzone and reject if within the redzone. */ >> + if (s->flags & SLAB_RED_ZONE) { >> + if (offset < s->red_left_pad) >> + return s->name; >> + offset -= s->red_left_pad; >> + } >> + >> /* Allow address range falling entirely within object size. */ >> if (offset <= s->object_size && n <= s->object_size - offset) >> return NULL; > > That fixes the case for me in kstrndup(), which allows the system to boot. Ugh, no it doesn't, booted the wrong kernel. I don't see the oops in strndup_user(), but instead get: usercopy: kernel memory overwrite attempt detected to d000000003610028 (cfq_io_cq) (88 bytes) CPU: 11 PID: 1 Comm: systemd Not tainted 4.7.0-rc3-00098-g09d9556ae5d1-dirty #65 Call Trace: [c0000001fb087bf0] [c0000000009bdbe8] dump_stack+0xb0/0xf0 (unreliable) [c0000001fb087c30] [c00000000029cf44] __check_object_size+0x74/0x320 [c0000001fb087cb0] [c00000000005d4d0] copy_from_user+0x60/0xd4 [c0000001fb087cf0] [c0000000008b38f4] __get_filter+0x74/0x160 [c0000001fb087d30] [c0000000008b408c] sk_attach_filter+0x2c/0xc0 [c0000001fb087d60] [c000000000871c34] sock_setsockopt+0x954/0xc00 [c0000001fb087dd0] [c00000000086ac44] SyS_setsockopt+0x134/0x150 [c0000001fb087e30] [c000000000009260] system_call+0x38/0x108 Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009 cheers From mboxrd@z Thu Jan 1 00:00:00 1970 From: mpe@ellerman.id.au (Michael Ellerman) Date: Sat, 09 Jul 2016 16:07:29 +1000 Subject: [kernel-hardening] Re: [PATCH 9/9] mm: SLUB hardened usercopy support In-Reply-To: <8737njpd37.fsf@@concordia.ellerman.id.au> References: <577f7e55.4668420a.84f17.5cb9SMTPIN_ADDED_MISSING@mx.google.com> <8737njpd37.fsf@@concordia.ellerman.id.au> Message-ID: <87ziprny3i.fsf@@concordia.ellerman.id.au> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Michael Ellerman writes: > Kees Cook writes: > >> On Fri, Jul 8, 2016 at 1:41 PM, Kees Cook wrote: >>> So, as found already, the position in the usercopy check needs to be >>> bumped down by red_left_pad, which is what Michael's fix does, so I'll >>> include it in the next version. >> >> Actually, after some offline chats, I think this is better, since it >> makes sure the ptr doesn't end up somewhere weird before we start the >> calculations. This leaves the pointer as-is, but explicitly handles >> the redzone on the offset instead, with no wrapping, etc: >> >> /* Find offset within object. */ >> offset = (ptr - page_address(page)) % s->size; >> >> + /* Adjust for redzone and reject if within the redzone. */ >> + if (s->flags & SLAB_RED_ZONE) { >> + if (offset < s->red_left_pad) >> + return s->name; >> + offset -= s->red_left_pad; >> + } >> + >> /* Allow address range falling entirely within object size. */ >> if (offset <= s->object_size && n <= s->object_size - offset) >> return NULL; > > That fixes the case for me in kstrndup(), which allows the system to boot. Ugh, no it doesn't, booted the wrong kernel. I don't see the oops in strndup_user(), but instead get: usercopy: kernel memory overwrite attempt detected to d000000003610028 (cfq_io_cq) (88 bytes) CPU: 11 PID: 1 Comm: systemd Not tainted 4.7.0-rc3-00098-g09d9556ae5d1-dirty #65 Call Trace: [c0000001fb087bf0] [c0000000009bdbe8] dump_stack+0xb0/0xf0 (unreliable) [c0000001fb087c30] [c00000000029cf44] __check_object_size+0x74/0x320 [c0000001fb087cb0] [c00000000005d4d0] copy_from_user+0x60/0xd4 [c0000001fb087cf0] [c0000000008b38f4] __get_filter+0x74/0x160 [c0000001fb087d30] [c0000000008b408c] sk_attach_filter+0x2c/0xc0 [c0000001fb087d60] [c000000000871c34] sock_setsockopt+0x954/0xc00 [c0000001fb087dd0] [c00000000086ac44] SyS_setsockopt+0x134/0x150 [c0000001fb087e30] [c000000000009260] system_call+0x38/0x108 Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009 cheers From mboxrd@z Thu Jan 1 00:00:00 1970 Reply-To: kernel-hardening@lists.openwall.com From: Michael Ellerman In-Reply-To: <8737njpd37.fsf@@concordia.ellerman.id.au> References: <577f7e55.4668420a.84f17.5cb9SMTPIN_ADDED_MISSING@mx.google.com> <8737njpd37.fsf@@concordia.ellerman.id.au> Date: Sat, 09 Jul 2016 16:07:29 +1000 Message-ID: <87ziprny3i.fsf@@concordia.ellerman.id.au> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [kernel-hardening] Re: [PATCH 9/9] mm: SLUB hardened usercopy support To: Kees Cook , Christoph Lameter Cc: "kernel-hardening@lists.openwall.com" , Jan Kara , Catalin Marinas , Will Deacon , Linux-MM , sparclinux , linux-ia64@vger.kernel.org, Andrea Arcangeli , linux-arch , "x86@kernel.org" , Russell King , PaX Team , Borislav Petkov , Mathias Krause , Fenghua Yu , Rik van Riel , David Rientjes , Tony Luck , Andy Lutomirski , Joonsoo Kim , Dmitry Vyukov , Laura Abbott , Brad Spengler , Ard Biesheuvel , LKML , Pekka Enberg , Case y Sc hauf ler , Andrew Morton , "linuxppc-dev@lists.ozlabs.org" , "David S. Miller" , "linux-arm-kernel@lists.infradead.org" List-ID: Michael Ellerman writes: > Kees Cook writes: > >> On Fri, Jul 8, 2016 at 1:41 PM, Kees Cook wrote: >>> So, as found already, the position in the usercopy check needs to be >>> bumped down by red_left_pad, which is what Michael's fix does, so I'll >>> include it in the next version. >> >> Actually, after some offline chats, I think this is better, since it >> makes sure the ptr doesn't end up somewhere weird before we start the >> calculations. This leaves the pointer as-is, but explicitly handles >> the redzone on the offset instead, with no wrapping, etc: >> >> /* Find offset within object. */ >> offset = (ptr - page_address(page)) % s->size; >> >> + /* Adjust for redzone and reject if within the redzone. */ >> + if (s->flags & SLAB_RED_ZONE) { >> + if (offset < s->red_left_pad) >> + return s->name; >> + offset -= s->red_left_pad; >> + } >> + >> /* Allow address range falling entirely within object size. */ >> if (offset <= s->object_size && n <= s->object_size - offset) >> return NULL; > > That fixes the case for me in kstrndup(), which allows the system to boot. Ugh, no it doesn't, booted the wrong kernel. I don't see the oops in strndup_user(), but instead get: usercopy: kernel memory overwrite attempt detected to d000000003610028 (cfq_io_cq) (88 bytes) CPU: 11 PID: 1 Comm: systemd Not tainted 4.7.0-rc3-00098-g09d9556ae5d1-dirty #65 Call Trace: [c0000001fb087bf0] [c0000000009bdbe8] dump_stack+0xb0/0xf0 (unreliable) [c0000001fb087c30] [c00000000029cf44] __check_object_size+0x74/0x320 [c0000001fb087cb0] [c00000000005d4d0] copy_from_user+0x60/0xd4 [c0000001fb087cf0] [c0000000008b38f4] __get_filter+0x74/0x160 [c0000001fb087d30] [c0000000008b408c] sk_attach_filter+0x2c/0xc0 [c0000001fb087d60] [c000000000871c34] sock_setsockopt+0x954/0xc00 [c0000001fb087dd0] [c00000000086ac44] SyS_setsockopt+0x134/0x150 [c0000001fb087e30] [c000000000009260] system_call+0x38/0x108 Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009 cheers