On Fri, Dec 04, 2015 at 12:17:08PM +0100, Dmitry Vyukov wrote: > Hello, > > UBSAN reports the following undefined behavior: > > UBSAN: Undefined behaviour in kernel/bpf/core.c:336:2 > shift exponent 2835 is to large for 32-bit type 'unsigned int' > CPU: 1 PID: 14227 Comm: syzkaller_execu Not tainted 4.4.0-rc3+ #142 > Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 > 0000000000000001 ffff88003892f898 ffffffff82c747b8 0000000041b58ab3 > ffffffff878cbc05 ffffffff82c74706 ffff88003892f860 ffff88003892f9a0 > 0000000000000000 0000000000000b13 ffffffff88178de2 0000000000000020 > Call Trace: > [<ffffffff82d684f0>] __ubsan_handle_shift_out_of_bounds+0x294/0x2e5 > lib/ubsan.c:417 > [<ffffffff8160a408>] __bpf_prog_run+0x8f48/0x9ac0 kernel/bpf/core.c:336 > [< inline >] seccomp_run_filters kernel/seccomp.c:198 > [< inline >] __seccomp_phase1_filter kernel/seccomp.c:588 > [<ffffffff8156ddfb>] seccomp_phase1+0x1cb/0x990 kernel/seccomp.c:667 > [<ffffffff8100651f>] syscall_trace_enter_phase1+0x28f/0x4e0 > arch/x86/entry/common.c:132 > [<ffffffff8691b939>] tracesys+0xd/0x44 arch/x86/entry/entry_64.S:240
is it with some random seccomp program? If normal libseccomp generates such programs than it needs to be fixed. > Such shifts have undefined behavior according to C standard and behave > differently on different archs. I guess we don't want to rely on any > kind of undefined behavior in bpf/seccomp. And generally want to > completely define results of all operations in bpf. bpf is an engine and we're not going to slow down each shift operation by extra run-time checks or masks. In other words bpf shift instruction == shift in C. Both undefined with for large operands. If seccomp is relying on undefined behavior, it should be fixed. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
