| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 25 Oct 2018 18:09:20 +0900
From: Prashant Bhole <bhole_prashant_q7@....ntt.co.jp>
To: Naresh Kamboju <naresh.kamboju@...aro.org>, liu.song.a23@...il.com
Cc: ast@...nel.org, Daniel Borkmann <daniel@...earbox.net>,
jakub.kicinski@...ronome.com,
"David S. Miller" <davem@...emloft.net>,
quentin.monnet@...ronome.com, netdev@...r.kernel.org,
"open list:KERNEL SELFTEST FRAMEWORK"
<linux-kselftest@...r.kernel.org>
Subject: Re: [PATCH bpf-next 6/6] selftests/bpf: test_verifier, check
bpf_map_lookup_elem access in bpf prog
On 10/25/2018 5:54 PM, Naresh Kamboju wrote:
> On Tue, 9 Oct 2018 at 12:32, Song Liu <liu.song.a23@...il.com> wrote:
>>
>> On Mon, Oct 8, 2018 at 6:07 PM Prashant Bhole
>> <bhole_prashant_q7@....ntt.co.jp> wrote:
>>>
>>> map_lookup_elem isn't supported by certain map types like:
>>> - BPF_MAP_TYPE_PROG_ARRAY
>>> - BPF_MAP_TYPE_STACK_TRACE
>>> - BPF_MAP_TYPE_XSKMAP
>>> - BPF_MAP_TYPE_SOCKMAP/BPF_MAP_TYPE_SOCKHASH
>>> Let's add verfier tests to check whether verifier prevents
>>> bpf_map_lookup_elem call on above programs from bpf program.
>>>
>>> Signed-off-by: Prashant Bhole <bhole_prashant_q7@....ntt.co.jp>
>>> Acked-by: Alexei Starovoitov <ast@...nel.org>
>> Acked-by: Song Liu <songliubraving@...com>
>>
>>> ---
>>> tools/testing/selftests/bpf/test_verifier.c | 121 +++++++++++++++++++-
>>> 1 file changed, 120 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c
>>> index 65ae44c85d27..cf4cd32b6772 100644
>>> --- a/tools/testing/selftests/bpf/test_verifier.c
>>> +++ b/tools/testing/selftests/bpf/test_verifier.c
>>> @@ -48,7 +48,7 @@
>>>
>>> #define MAX_INSNS BPF_MAXINSNS
>>> #define MAX_FIXUPS 8
>>> -#define MAX_NR_MAPS 8
>>> +#define MAX_NR_MAPS 13
>>> #define POINTER_VALUE 0xcafe4all
>>> #define TEST_DATA_LEN 64
>>>
>>> @@ -65,6 +65,10 @@ struct bpf_test {
>>> int fixup_map_hash_48b[MAX_FIXUPS];
>>> int fixup_map_hash_16b[MAX_FIXUPS];
>>> int fixup_map_array_48b[MAX_FIXUPS];
>>> + int fixup_map_sockmap[MAX_FIXUPS];
>>> + int fixup_map_sockhash[MAX_FIXUPS];
>>> + int fixup_map_xskmap[MAX_FIXUPS];
>>> + int fixup_map_stacktrace[MAX_FIXUPS];
>>> int fixup_prog1[MAX_FIXUPS];
>>> int fixup_prog2[MAX_FIXUPS];
>>> int fixup_map_in_map[MAX_FIXUPS];
>>> @@ -4541,6 +4545,85 @@ static struct bpf_test tests[] = {
>>> .errstr = "invalid access to packet",
>>> .prog_type = BPF_PROG_TYPE_SCHED_CLS,
>>> },
>>> + {
>>> + "prevent map lookup in sockmap",
>>> + .insns = {
>>> + BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
>>> + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
>>> + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
>>> + BPF_LD_MAP_FD(BPF_REG_1, 0),
>>> + BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
>>> + BPF_FUNC_map_lookup_elem),
>>> + BPF_EXIT_INSN(),
>>> + },
>>> + .fixup_map_sockmap = { 3 },
>>> + .result = REJECT,
>>> + .errstr = "cannot pass map_type 15 into func bpf_map_lookup_elem",
>>> + .prog_type = BPF_PROG_TYPE_SOCK_OPS,
>>> + },
>>> + {
>>> + "prevent map lookup in sockhash",
>>> + .insns = {
>>> + BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
>>> + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
>>> + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
>>> + BPF_LD_MAP_FD(BPF_REG_1, 0),
>>> + BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
>>> + BPF_FUNC_map_lookup_elem),
>>> + BPF_EXIT_INSN(),
>>> + },
>>> + .fixup_map_sockhash = { 3 },
>>> + .result = REJECT,
>>> + .errstr = "cannot pass map_type 18 into func bpf_map_lookup_elem",
>>> + .prog_type = BPF_PROG_TYPE_SOCK_OPS,
>>> + },
>>> + {
>>> + "prevent map lookup in xskmap",
>>> + .insns = {
>>> + BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
>>> + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
>>> + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
>>> + BPF_LD_MAP_FD(BPF_REG_1, 0),
>>> + BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
>>> + BPF_FUNC_map_lookup_elem),
>>> + BPF_EXIT_INSN(),
>>> + },
>>> + .fixup_map_xskmap = { 3 },
>>> + .result = REJECT,
>>> + .errstr = "cannot pass map_type 17 into func bpf_map_lookup_elem",
>>> + .prog_type = BPF_PROG_TYPE_XDP,
>>> + },
>>> + {
>>> + "prevent map lookup in stack trace",
>>> + .insns = {
>>> + BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
>>> + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
>>> + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
>>> + BPF_LD_MAP_FD(BPF_REG_1, 0),
>>> + BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
>>> + BPF_FUNC_map_lookup_elem),
>>> + BPF_EXIT_INSN(),
>>> + },
>>> + .fixup_map_stacktrace = { 3 },
>>> + .result = REJECT,
>>> + .errstr = "cannot pass map_type 7 into func bpf_map_lookup_elem",
>>> + .prog_type = BPF_PROG_TYPE_PERF_EVENT,
>>> + },
>>> + {
>>> + "prevent map lookup in prog array",
>>> + .insns = {
>>> + BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
>>> + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
>>> + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
>>> + BPF_LD_MAP_FD(BPF_REG_1, 0),
>>> + BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
>>> + BPF_FUNC_map_lookup_elem),
>>> + BPF_EXIT_INSN(),
>>> + },
>>> + .fixup_prog2 = { 3 },
>>> + .result = REJECT,
>>> + .errstr = "cannot pass map_type 3 into func bpf_map_lookup_elem",
>>> + },
>>> {
>>> "valid map access into an array with a constant",
>>> .insns = {
>>> @@ -13515,6 +13598,10 @@ static void do_test_fixup(struct bpf_test *test, enum bpf_map_type prog_type,
>>> int *fixup_map_hash_48b = test->fixup_map_hash_48b;
>>> int *fixup_map_hash_16b = test->fixup_map_hash_16b;
>>> int *fixup_map_array_48b = test->fixup_map_array_48b;
>>> + int *fixup_map_sockmap = test->fixup_map_sockmap;
>>> + int *fixup_map_sockhash = test->fixup_map_sockhash;
>>> + int *fixup_map_xskmap = test->fixup_map_xskmap;
>>> + int *fixup_map_stacktrace = test->fixup_map_stacktrace;
>>> int *fixup_prog1 = test->fixup_prog1;
>>> int *fixup_prog2 = test->fixup_prog2;
>>> int *fixup_map_in_map = test->fixup_map_in_map;
>>> @@ -13603,6 +13690,38 @@ static void do_test_fixup(struct bpf_test *test, enum bpf_map_type prog_type,
>>> fixup_percpu_cgroup_storage++;
>>> } while (*fixup_percpu_cgroup_storage);
>>> }
>>> + if (*fixup_map_sockmap) {
>>> + map_fds[9] = create_map(BPF_MAP_TYPE_SOCKMAP, sizeof(int),
>>> + sizeof(int), 1);
>>> + do {
>>> + prog[*fixup_map_sockmap].imm = map_fds[9];
>>> + fixup_map_sockmap++;
>>> + } while (*fixup_map_sockmap);
>>> + }
>>> + if (*fixup_map_sockhash) {
>>> + map_fds[10] = create_map(BPF_MAP_TYPE_SOCKHASH, sizeof(int),
>>> + sizeof(int), 1);
>>> + do {
>>> + prog[*fixup_map_sockhash].imm = map_fds[10];
>>> + fixup_map_sockhash++;
>>> + } while (*fixup_map_sockhash);
>>> + }
>>> + if (*fixup_map_xskmap) {
>>> + map_fds[11] = create_map(BPF_MAP_TYPE_XSKMAP, sizeof(int),
>>> + sizeof(int), 1);
>>> + do {
>>> + prog[*fixup_map_xskmap].imm = map_fds[11];
>>> + fixup_map_xskmap++;
>>> + } while (*fixup_map_xskmap);
>>> + }
>
> selftests: bpf: test_verifier sockmap, sockhash, xskmap failed on
> mainline and next
> (from 4.19.0-rc7-next-20181011 to till date )
> Are we missing any pre-required kernel configs ?
sockmap/hashmap is dependent on CONFIG_BPF_STREAM_PARSER and xskmap is
dependent on CONFIG_XDP_SOCKETS.
Reference: include/linux/bpf_types.h
-Prashant
>
> Test log,
> ------------
> selftests: bpf: test_verifier
> <>
> #274/p prevent map lookup in sockmap Failed to create hash map
> 'Invalid argument'!
> FAIL
> Unexpected error message!
> EXP: cannot pass map_type 15 into func bpf_map_lookup_elem
> RES: fd -1 is not pointing to valid bpf_map
> fd -1 is not pointing to valid bpf_map
> #275/p prevent map lookup in sockhash Failed to create hash map
> 'Invalid argument'!
> FAIL
> Unexpected error message!
> EXP: cannot pass map_type 18 into func bpf_map_lookup_elem
> RES: fd -1 is not pointing to valid bpf_map
> fd -1 is not pointing to valid bpf_map
> #276/p prevent map lookup in xskmap Failed to create hash map 'Invalid
> argument'!
> FAIL
> Unexpected error message!
> EXP: cannot pass map_type 17 into func bpf_map_lookup_elem
> RES: fd -1 is not pointing to valid bpf_map
> fd -1 is not pointing to valid bpf_map
> <>
> Summary: 962 PASSED, 0 SKIPPED, 3 FAILED
> not ok 1..1 selftests: bpf: test_verifier [FAIL]
> selftests: bpf_test_verifier [FAIL]
>
> -mainline results history,
> https://qa-reports.linaro.org/lkft/linux-next-oe/tests/kselftest/bpf_test_verifier
>
> -next results history,
> https://qa-reports.linaro.org/lkft/linux-next-oe/tests/kselftest/bpf_test_verifier
>
> Test case full log,
> https://lkft.validation.linaro.org/scheduler/job/461881#L1655
>
> Best regards
> Naresh Kamboju
>
>>> + if (*fixup_map_stacktrace) {
>>> + map_fds[12] = create_map(BPF_MAP_TYPE_STACK_TRACE, sizeof(u32),
>>> + sizeof(u64), 1);
>>> + do {
>>> + prog[*fixup_map_stacktrace].imm = map_fds[12];
>>> + fixup_map_stacktrace++;
>>> + } while (fixup_map_stacktrace);
>>> + }
>>> }
>>>
>>> static void do_test_single(struct bpf_test *test, bool unpriv,
>>> --
>>> 2.17.1
>>>
>>>
>
>
Powered by blists - more mailing lists