From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: Xu Kuohai <xukuohai@huaweicloud.com>
Cc: Yonghong Song <yonghong.song@linux.dev>,
Eduard Zingerman <eddyz87@gmail.com>,
bpf@vger.kernel.org, netdev@vger.kernel.org,
linux-security-module@vger.kernel.org,
linux-kselftest@vger.kernel.org,
Alexei Starovoitov <ast@kernel.org>,
Andrii Nakryiko <andrii@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
Martin KaFai Lau <martin.lau@linux.dev>,
Song Liu <song@kernel.org>,
John Fastabend <john.fastabend@gmail.com>,
KP Singh <kpsingh@kernel.org>,
Stanislav Fomichev <sdf@google.com>, Hao Luo <haoluo@google.com>,
Jiri Olsa <jolsa@kernel.org>,
Matt Bobrowski <mattbobrowski@google.com>,
Brendan Jackman <jackmanb@chromium.org>,
Paul Moore <paul@paul-moore.com>,
James Morris <jmorris@namei.org>,
"Serge E . Hallyn" <serge@hallyn.com>,
Khadija Kamran <kamrankhadijadj@gmail.com>,
Casey Schaufler <casey@schaufler-ca.com>,
Ondrej Mosnacek <omosnace@redhat.com>,
Kees Cook <keescook@chromium.org>,
John Johansen <john.johansen@canonical.com>,
Lukas Bulwahn <lukas.bulwahn@gmail.com>,
Roberto Sassu <roberto.sassu@huawei.com>,
Shung-Hsi Yu <shung-hsi.yu@suse.com>
Subject: Re: [PATCH bpf-next v3 07/11] bpf: Fix a false rejection caused by AND operation
Date: Fri, 26 Apr 2024 13:36:37 -0700 [thread overview]
Message-ID: <CAEf4BzZTzftrOCFsfBd81sHDBpmNK+4Jefqa3SSS6NiuncO0tQ@mail.gmail.com> (raw)
In-Reply-To: <576c7c44-d1b4-42c8-8b6e-2e6b93d7547a@huaweicloud.com>
On Tue, Apr 23, 2024 at 7:26 PM Xu Kuohai <xukuohai@huaweicloud.com> wrote:
>
> On 4/24/2024 5:55 AM, Yonghong Song wrote:
> >
> > On 4/20/24 1:33 AM, Xu Kuohai wrote:
> >> On 4/20/2024 7:00 AM, Eduard Zingerman wrote:
> >>> On Thu, 2024-04-11 at 20:27 +0800, Xu Kuohai wrote:
> >>>> From: Xu Kuohai <xukuohai@huawei.com>
> >>>>
> >>>> With lsm return value check, the no-alu32 version test_libbpf_get_fd_by_id_opts
> >>>> is rejected by the verifier, and the log says:
> >>>>
> >>>> 0: R1=ctx() R10=fp0
> >>>> ; int BPF_PROG(check_access, struct bpf_map *map, fmode_t fmode) @ test_libbpf_get_fd_by_id_opts.c:27
> >>>> 0: (b7) r0 = 0 ; R0_w=0
> >>>> 1: (79) r2 = *(u64 *)(r1 +0)
> >>>> func 'bpf_lsm_bpf_map' arg0 has btf_id 916 type STRUCT 'bpf_map'
> >>>> 2: R1=ctx() R2_w=trusted_ptr_bpf_map()
> >>>> ; if (map != (struct bpf_map *)&data_input) @ test_libbpf_get_fd_by_id_opts.c:29
> >>>> 2: (18) r3 = 0xffff9742c0951a00 ; R3_w=map_ptr(map=data_input,ks=4,vs=4)
> >>>> 4: (5d) if r2 != r3 goto pc+4 ; R2_w=trusted_ptr_bpf_map() R3_w=map_ptr(map=data_input,ks=4,vs=4)
> >>>> ; int BPF_PROG(check_access, struct bpf_map *map, fmode_t fmode) @ test_libbpf_get_fd_by_id_opts.c:27
> >>>> 5: (79) r0 = *(u64 *)(r1 +8) ; R0_w=scalar() R1=ctx()
> >>>> ; if (fmode & FMODE_WRITE) @ test_libbpf_get_fd_by_id_opts.c:32
> >>>> 6: (67) r0 <<= 62 ; R0_w=scalar(smax=0x4000000000000000,umax=0xc000000000000000,smin32=0,smax32=umax32=0,var_off=(0x0; 0xc000000000000000))
> >>>> 7: (c7) r0 s>>= 63 ; R0_w=scalar(smin=smin32=-1,smax=smax32=0)
> >>>> ; @ test_libbpf_get_fd_by_id_opts.c:0
> >>>> 8: (57) r0 &= -13 ; R0_w=scalar(smax=0x7ffffffffffffff3,umax=0xfffffffffffffff3,smax32=0x7ffffff3,umax32=0xfffffff3,var_off=(0x0; 0xfffffffffffffff3))
> >>>> ; int BPF_PROG(check_access, struct bpf_map *map, fmode_t fmode) @ test_libbpf_get_fd_by_id_opts.c:27
> >>>> 9: (95) exit
[...]
>
> As suggested by Eduard, this patch makes a special case for source
> or destination register of '&=' operation being in range [-1, 0].
>
> Meaning that one of the '&=' operands is either:
> - all ones, in which case the counterpart is the result of the operation;
> - all zeros, in which case zero is the result of the operation.
>
> And MIN and MAX values could be derived based on above two observations.
>
> [0] https://lore.kernel.org/bpf/e62e2971301ca7f2e9eb74fc500c520285cad8f5.camel@gmail.com/
> [1] https://github.com/llvm/llvm-project/blob/4523a267829c807f3fc8fab8e5e9613985a51565/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
>
> Suggested-by: Eduard Zingerman <eddyz87@gmail.com>
> Signed-off-by: Xu Kuohai <xukuohai@huawei.com>
>
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 640747b53745..30c551d39329 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -13374,6 +13374,24 @@ static void scalar32_min_max_and(struct bpf_reg_state *dst_reg,
> dst_reg->u32_min_value = var32_off.value;
> dst_reg->u32_max_value = min(dst_reg->u32_max_value, umax_val);
>
> + /* Special case: src_reg is known and dst_reg is in range [-1, 0] */
> + if (src_known &&
> + dst_reg->s32_min_value == -1 && dst_reg->s32_max_value == 0 &&
> + dst_reg->smin_value == -1 && dst_reg->smax_value == 0) {
please keep if () condition aligned across multiple lines, it's super
confusing this way
> + dst_reg->s32_min_value = min_t(s32, src_reg->s32_min_value, 0);
> + dst_reg->s32_max_value = max_t(s32, src_reg->s32_min_value, 0);
do we need to update tnum parts as well (or reset and re-derive, probably)?
btw, can't we support src being a range here? the idea is that dst_reg
either all ones or all zeros. For and it means that it either stays
all zero, or will be *exactly equal* to src, right? So I think the
logic would be:
a) if [s32_min, s32_max] is on the same side of zero, then resulting
range would be [min(s32_min, 0), max(s32_max, 0)], just like you have
here
b) if [s32_min, s32_max] contains zero, then resulting range will be
exactly [s32_min, s32_max]
Or did I make a mistake above?
> + return;
> + }
> +
> + /* Special case: dst_reg is known and src_reg is in range [-1, 0] */
> + if (dst_known &&
> + src_reg->s32_min_value == -1 && src_reg->s32_max_value == 0 &&
> + src_reg->smin_value == -1 && src_reg->smax_value == 0) {
> + dst_reg->s32_min_value = min_t(s32, dst_reg->s32_min_value, 0);
> + dst_reg->s32_max_value = max_t(s32, dst_reg->s32_min_value, 0);
> + return;
> + }
> +
> /* Safe to set s32 bounds by casting u32 result into s32 when u32
> * doesn't cross sign boundary. Otherwise set s32 bounds to unbounded.
> */
[...]
next prev parent reply other threads:[~2024-04-26 20:36 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-11 12:27 [PATCH bpf-next v3 00/11] Add check for bpf lsm return value Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 01/11] bpf, lsm: Annotate lsm hook return value range Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 02/11] bpf, lsm: Add helper to read " Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 03/11] bpf, lsm: Check bpf lsm hook return values in verifier Xu Kuohai
2024-04-13 11:44 ` Eduard Zingerman
2024-04-11 12:27 ` [PATCH bpf-next v3 04/11] bpf, lsm: Add bpf lsm disabled hook list Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 05/11] bpf: Avoid progs for different hooks calling each other with tail call Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 06/11] bpf: Fix compare error in function retval_range_within Xu Kuohai
2024-04-12 8:53 ` Shung-Hsi Yu
2024-04-25 23:41 ` Andrii Nakryiko
2024-04-26 8:08 ` Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 07/11] bpf: Fix a false rejection caused by AND operation Xu Kuohai
2024-04-19 23:00 ` Eduard Zingerman
2024-04-20 8:33 ` Xu Kuohai
2024-04-23 21:55 ` Yonghong Song
2024-04-24 2:25 ` Xu Kuohai
2024-04-24 22:06 ` Yonghong Song
2024-04-25 2:42 ` Xu Kuohai
2024-04-25 16:28 ` Yonghong Song
2024-04-26 7:43 ` Xu Kuohai
2024-04-26 20:36 ` Andrii Nakryiko [this message]
2024-04-28 15:15 ` Xu Kuohai
2024-04-29 20:58 ` Andrii Nakryiko
2024-04-29 22:18 ` Eduard Zingerman
2024-04-30 3:56 ` Xu Kuohai
2024-04-30 3:54 ` Xu Kuohai
2024-04-29 21:56 ` Eduard Zingerman
2024-04-11 12:27 ` [PATCH bpf-next v3 08/11] selftests/bpf: Avoid load failure for token_lsm.c Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 09/11] selftests/bpf: Add return value checks for failed tests Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 10/11] selftests/bpf: Add test for lsm tail call Xu Kuohai
2024-04-11 12:27 ` [PATCH bpf-next v3 11/11] selftests/bpf: Add verifier tests for bpf lsm Xu Kuohai
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAEf4BzZTzftrOCFsfBd81sHDBpmNK+4Jefqa3SSS6NiuncO0tQ@mail.gmail.com \
--to=andrii.nakryiko@gmail.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=casey@schaufler-ca.com \
--cc=daniel@iogearbox.net \
--cc=eddyz87@gmail.com \
--cc=haoluo@google.com \
--cc=jackmanb@chromium.org \
--cc=jmorris@namei.org \
--cc=john.fastabend@gmail.com \
--cc=john.johansen@canonical.com \
--cc=jolsa@kernel.org \
--cc=kamrankhadijadj@gmail.com \
--cc=keescook@chromium.org \
--cc=kpsingh@kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=lukas.bulwahn@gmail.com \
--cc=martin.lau@linux.dev \
--cc=mattbobrowski@google.com \
--cc=netdev@vger.kernel.org \
--cc=omosnace@redhat.com \
--cc=paul@paul-moore.com \
--cc=roberto.sassu@huawei.com \
--cc=sdf@google.com \
--cc=serge@hallyn.com \
--cc=shung-hsi.yu@suse.com \
--cc=song@kernel.org \
--cc=xukuohai@huaweicloud.com \
--cc=yonghong.song@linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).