From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753941AbaIOQP6 (ORCPT ); Mon, 15 Sep 2014 12:15:58 -0400 Received: from cam-admin0.cambridge.arm.com ([217.140.96.50]:41380 "EHLO cam-admin0.cambridge.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753272AbaIOQP5 (ORCPT ); Mon, 15 Sep 2014 12:15:57 -0400 Date: Mon, 15 Sep 2014 17:15:07 +0100 From: Will Deacon To: Daniel Borkmann Cc: Z Lim , Catalin Marinas , "David S. Miller" , Alexei Starovoitov , "linux-arm-kernel@lists.infradead.org" , LKML Subject: Re: [PATCH arm64-next v2] net: bpf: arm64: address randomize and write protect JIT code Message-ID: <20140915161507.GD29744@arm.com> References: <1410543308-27449-1-git-send-email-dborkman@redhat.com> <5416EF28.8020900@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <5416EF28.8020900@redhat.com> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Sep 15, 2014 at 02:52:40PM +0100, Daniel Borkmann wrote: > On 09/13/2014 06:32 AM, Z Lim wrote: > > On Fri, Sep 12, 2014 at 10:35 AM, Daniel Borkmann wrote: > >> This is the ARM64 variant for 314beb9bcab ("x86: bpf_jit_comp: secure bpf > >> jit against spraying attacks"). > >> > >> Thanks to commit 11d91a770f1f ("arm64: Add CONFIG_DEBUG_SET_MODULE_RONX > >> support") which added necessary infrastructure, we can now implement > >> RO marking of eBPF generated JIT image pages and randomize start offset > >> for the JIT code, so that it does not reside directly on a page boundary > >> anymore. Likewise, the holes are filled with illegal instructions. > >> > >> This is basically the ARM64 variant of what we already have in ARM via > >> commit 55309dd3d4cd ("net: bpf: arm: address randomize and write protect > >> JIT code"). Moreover, this commit also presents a merge resolution due to > >> conflicts with commit 60a3b2253c41 ("net: bpf: make eBPF interpreter images > >> read-only") as we don't use kfree() in bpf_jit_free() anymore to release > >> the locked bpf_prog structure, but instead bpf_prog_unlock_free() through > >> a different allocator. > >> > >> JIT tested on aarch64 with BPF test suite. > >> > >> Reference: http://mainisusuallyafunction.blogspot.com/2012/11/attacking-hardened-linux-systems-with.html > >> Signed-off-by: Daniel Borkmann > >> Cc: Zi Shen Lim > >> Cc: Will Deacon > >> Cc: Catalin Marinas > >> Cc: David S. Miller > >> Cc: Alexei Starovoitov > >> --- > >> v1 -> v2: > >> - Use brk insn as suggested by Catalin, thanks a lot for > >> your feedback! Rest unchanged. > >> Note: > >> - This patch depends on net-next being merged to mainline due > >> to the mentioned merge conflict. > >> > >> arch/arm64/net/bpf_jit_comp.c | 38 +++++++++++++++++++++++++++++--------- > >> 1 file changed, 29 insertions(+), 9 deletions(-) > >> > >> diff --git a/arch/arm64/net/bpf_jit_comp.c b/arch/arm64/net/bpf_jit_comp.c > >> index 7ae3354..4b71779 100644 > >> --- a/arch/arm64/net/bpf_jit_comp.c > >> +++ b/arch/arm64/net/bpf_jit_comp.c > >> @@ -19,7 +19,6 @@ > >> #define pr_fmt(fmt) "bpf_jit: " fmt > >> > >> #include > >> -#include > >> #include > >> #include > >> #include > >> @@ -119,6 +118,15 @@ static inline int bpf2a64_offset(int bpf_to, int bpf_from, > >> return to - from; > >> } > >> > >> +static void jit_fill_hole(void *area, unsigned int size) > >> +{ > >> + /* We use brk #0x100 to trigger a fault. */ > >> + u32 *ptr, fill_ins = 0xd4202000; > > > > Missed this on first round of review, I think we also need > > cpu_to_le32(...) here. > > Just wondering ... so that would also hold true in case I build/run my > kernel in big-endian (CPU_BIG_ENDIAN)? Yes, the instruction stream is always little endian so the conversion is required. Will > >> + /* We are guaranteed to have aligned memory. */ > >> + for (ptr = area; size >= sizeof(u32); size -= sizeof(u32)) > >> + *ptr++ = fill_ins; > >> +} > >> + > > [...] > > > > Thanks Daniel. >