Resending with a fix of mktemp argument in verify_sig_setup.sh
One of the desirable features in security is the ability to restrict import of data to a given system based on data authenticity. If data import can be restricted, it would be possible to enforce a system-wide policy based on the signing keys the system owner trusts.
This feature is widely used in the kernel. For example, if the restriction is enabled, kernel modules can be plugged in only if they are signed with a key whose public part is in the primary or secondary keyring.
For eBPF, it can be useful as well. For example, it might be useful to authenticate data an eBPF program makes security decisions on.
After a discussion in the eBPF mailing list, it was decided that the stated goal should be accomplished by introducing a new helper: bpf_verify_pkcs7_signature(), dedicated to verify PKCS#7 signatures. More helpers will be introduced later, as necessary.
The job of bpf_verify_pkcs7_signature() is simply to call the corresponding signature verification function verify_pkcs7_signature(). Data and signature can be provided to the new helper with two dynamic pointers, to reduce the number of parameters. The keyring containing the signature verification key can be obtained with a new helper called bpf_request_key_by_id().
For now, keyrings can be obtained with an identifier defined in verification.h (except for the special value ULONG_MAX, used for testing). In the future, keyring can be searched also by their description. This functionality has not been included here in this patch set, as would require additional care for decrementing the reference count of the keyring. It could be added later.
While bpf_request_key_by_id() can be called from any program, bpf_verify_pkcs7_signature(), instead, must be called by a sleepable program, as it is doing crypto operations. For the latter, for example, lsm.s/bpf is suitable, fexit/array_map_update_elem is not.
The added test, which invokes both helpers, checks the ability of an eBPF program to verify module-style appended signatures, as produced by the kernel tool sign-file, currently used to sign kernel modules.
The patch set is organized as follows.
Patch 1 exports bpf_dynptr_get_size(), to obtain the real size of data carried by a dynamic pointer. Patch 2 introduces the bpf_request_key_by_id() helper. Patch 3 introduces the bpf_verify_pkcs7_signature() helper. Finally, patch 4 adds a test for both helpers.
Changelog
v3: - Rename bpf_verify_signature() back to bpf_verify_pkcs7_signature() to avoid managing different parameters for each signature verification function in one helper (suggested by Daniel) - Use dynamic pointers and export bpf_dynptr_get_size() (suggested by Alexei) - Introduce bpf_request_key_by_id() to give more flexibility to the caller of bpf_verify_pkcs7_signature() to retrieve the appropriate keyring (suggested by Alexei) - Fix test by reordering the gcc command line, always compile sign-file - Improve helper support check mechanism in the test
v2: - Rename bpf_verify_pkcs7_signature() to a more generic bpf_verify_signature() and pass the signature type (suggested by KP) - Move the helper and prototype declaration under #ifdef so that user space can probe for support for the helper (suggested by Daniel) - Describe better the keyring types (suggested by Daniel) - Include linux/bpf.h instead of vmlinux.h to avoid implicit or redeclaration - Make the test selfcontained (suggested by Alexei)
v1: - Don't define new map flag but introduce simple wrapper of verify_pkcs7_signature() (suggested by Alexei and KP)
Roberto Sassu (4): bpf: Export bpf_dynptr_get_size() bpf: Add bpf_request_key_by_id() helper bpf: Add bpf_verify_pkcs7_signature() helper selftests/bpf: Add test for bpf_verify_pkcs7_signature() helper
include/linux/bpf.h | 1 + include/uapi/linux/bpf.h | 25 ++ kernel/bpf/bpf_lsm.c | 60 +++++ kernel/bpf/helpers.c | 2 +- scripts/bpf_doc.py | 2 + tools/include/uapi/linux/bpf.h | 25 ++ tools/testing/selftests/bpf/Makefile | 14 +- tools/testing/selftests/bpf/config | 2 + .../bpf/prog_tests/verify_pkcs7_sig.c | 217 ++++++++++++++++++ .../bpf/progs/test_verify_pkcs7_sig.c | 168 ++++++++++++++ .../testing/selftests/bpf/verify_sig_setup.sh | 100 ++++++++ 11 files changed, 612 insertions(+), 4 deletions(-) create mode 100644 tools/testing/selftests/bpf/prog_tests/verify_pkcs7_sig.c create mode 100644 tools/testing/selftests/bpf/progs/test_verify_pkcs7_sig.c create mode 100755 tools/testing/selftests/bpf/verify_sig_setup.sh
Export bpf_dynptr_get_size(), so that kernel code dealing with eBPF dynamic pointers can obtain the real size of data carried by this data structure.
Cc: Joanne Koong joannelkoong@gmail.com Signed-off-by: Roberto Sassu roberto.sassu@huawei.com --- include/linux/bpf.h | 1 + kernel/bpf/helpers.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-)
diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 8e6092d0ea95..6eb03a0c9687 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -2420,5 +2420,6 @@ void bpf_dynptr_init(struct bpf_dynptr_kern *ptr, void *data, enum bpf_dynptr_type type, u32 offset, u32 size); void bpf_dynptr_set_null(struct bpf_dynptr_kern *ptr); int bpf_dynptr_check_size(u32 size); +u32 bpf_dynptr_get_size(struct bpf_dynptr_kern *ptr);
#endif /* _LINUX_BPF_H */ diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c index 225806a02efb..29e1810afaf6 100644 --- a/kernel/bpf/helpers.c +++ b/kernel/bpf/helpers.c @@ -1430,7 +1430,7 @@ static void bpf_dynptr_set_type(struct bpf_dynptr_kern *ptr, enum bpf_dynptr_typ ptr->size |= type << DYNPTR_TYPE_SHIFT; }
-static u32 bpf_dynptr_get_size(struct bpf_dynptr_kern *ptr) +u32 bpf_dynptr_get_size(struct bpf_dynptr_kern *ptr) { return ptr->size & DYNPTR_SIZE_MASK; }
Add the bpf_request_key_by_id() helper, so that an eBPF program can obtain a suitable key pointer to pass to the bpf_verify_pkcs7_signature() helper, to be introduced in a later patch.
The passed identifier can have the following values: 0 for the primary keyring (immutable keyring of system keys); 1 for both the primary and secondary keyring (where keys can be added only if they are vouched for by existing keys in those keyrings); 2 for the platform keyring (primarily used by the integrity subsystem to verify a kexec'ed kerned image and, possibly, the initramfs signature); ULONG_MAX for the session keyring (for testing purposes).
Signed-off-by: Roberto Sassu roberto.sassu@huawei.com --- include/uapi/linux/bpf.h | 17 +++++++++++++++++ kernel/bpf/bpf_lsm.c | 30 ++++++++++++++++++++++++++++++ scripts/bpf_doc.py | 2 ++ tools/include/uapi/linux/bpf.h | 17 +++++++++++++++++ 4 files changed, 66 insertions(+)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index f4009dbdf62d..dfd93e0e0759 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -5249,6 +5249,22 @@ union bpf_attr { * Pointer to the underlying dynptr data, NULL if the dynptr is * read-only, if the dynptr is invalid, or if the offset and length * is out of bounds. + * + * struct key *bpf_request_key_by_id(unsigned long id) + * Description + * Request a keyring by *id*. + * + * *id* can have the following values (some defined in + * verification.h): 0 for the primary keyring (immutable keyring of + * system keys); 1 for both the primary and secondary keyring + * (where keys can be added only if they are vouched for by + * existing keys in those keyrings); 2 for the platform keyring + * (primarily used by the integrity subsystem to verify a kexec'ed + * kerned image and, possibly, the initramfs signature); ULONG_MAX + * for the session keyring (for testing purposes). + * Return + * A non-NULL pointer if *id* is valid and not 0, a NULL pointer + * otherwise. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5455,6 +5471,7 @@ union bpf_attr { FN(dynptr_read), \ FN(dynptr_write), \ FN(dynptr_data), \ + FN(request_key_by_id), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index c1351df9f7ee..e1911812398b 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -16,6 +16,7 @@ #include <linux/bpf_local_storage.h> #include <linux/btf_ids.h> #include <linux/ima.h> +#include <linux/verification.h>
/* For every LSM hook that allows attachment of BPF programs, declare a nop * function where a BPF program can be attached. @@ -132,6 +133,31 @@ static const struct bpf_func_proto bpf_get_attach_cookie_proto = { .arg1_type = ARG_PTR_TO_CTX, };
+#ifdef CONFIG_KEYS +BTF_ID_LIST_SINGLE(bpf_request_key_by_id_btf_ids, struct, key) + +BPF_CALL_1(bpf_request_key_by_id, unsigned long, id) +{ + const struct cred *cred = current_cred(); + + if (id > (unsigned long)VERIFY_USE_PLATFORM_KEYRING && id != ULONG_MAX) + return (unsigned long)NULL; + + if (id == ULONG_MAX) + return (unsigned long)cred->session_keyring; + + return id; +} + +static const struct bpf_func_proto bpf_request_key_by_id_proto = { + .func = bpf_request_key_by_id, + .gpl_only = false, + .ret_type = RET_PTR_TO_BTF_ID_OR_NULL, + .ret_btf_id = &bpf_request_key_by_id_btf_ids[0], + .arg1_type = ARG_ANYTHING, +}; +#endif /* CONFIG_KEYS */ + static const struct bpf_func_proto * bpf_lsm_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) { @@ -158,6 +184,10 @@ bpf_lsm_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return prog->aux->sleepable ? &bpf_ima_file_hash_proto : NULL; case BPF_FUNC_get_attach_cookie: return bpf_prog_has_trampoline(prog) ? &bpf_get_attach_cookie_proto : NULL; +#ifdef CONFIG_KEYS + case BPF_FUNC_request_key_by_id: + return &bpf_request_key_by_id_proto; +#endif /* CONFIG_KEYS */ default: return tracing_prog_func_proto(func_id, prog); } diff --git a/scripts/bpf_doc.py b/scripts/bpf_doc.py index 855b937e7585..176917df0ac0 100755 --- a/scripts/bpf_doc.py +++ b/scripts/bpf_doc.py @@ -635,6 +635,7 @@ class PrinterHelpers(Printer): 'struct bpf_timer', 'struct mptcp_sock', 'struct bpf_dynptr', + 'struct key', ] known_types = { '...', @@ -686,6 +687,7 @@ class PrinterHelpers(Printer): 'struct bpf_timer', 'struct mptcp_sock', 'struct bpf_dynptr', + 'struct key', } mapped_types = { 'u8': '__u8', diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index f4009dbdf62d..dfd93e0e0759 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -5249,6 +5249,22 @@ union bpf_attr { * Pointer to the underlying dynptr data, NULL if the dynptr is * read-only, if the dynptr is invalid, or if the offset and length * is out of bounds. + * + * struct key *bpf_request_key_by_id(unsigned long id) + * Description + * Request a keyring by *id*. + * + * *id* can have the following values (some defined in + * verification.h): 0 for the primary keyring (immutable keyring of + * system keys); 1 for both the primary and secondary keyring + * (where keys can be added only if they are vouched for by + * existing keys in those keyrings); 2 for the platform keyring + * (primarily used by the integrity subsystem to verify a kexec'ed + * kerned image and, possibly, the initramfs signature); ULONG_MAX + * for the session keyring (for testing purposes). + * Return + * A non-NULL pointer if *id* is valid and not 0, a NULL pointer + * otherwise. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5455,6 +5471,7 @@ union bpf_attr { FN(dynptr_read), \ FN(dynptr_write), \ FN(dynptr_data), \ + FN(request_key_by_id), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper
On Tue, Jun 14, 2022 at 03:06:19PM +0200, Roberto Sassu wrote:
Add the bpf_request_key_by_id() helper, so that an eBPF program can obtain a suitable key pointer to pass to the bpf_verify_pkcs7_signature() helper, to be introduced in a later patch.
The passed identifier can have the following values: 0 for the primary keyring (immutable keyring of system keys); 1 for both the primary and secondary keyring (where keys can be added only if they are vouched for by existing keys in those keyrings); 2 for the platform keyring (primarily used by the integrity subsystem to verify a kexec'ed kerned image and, possibly, the initramfs signature); ULONG_MAX for the session keyring (for testing purposes).
Signed-off-by: Roberto Sassu roberto.sassu@huawei.com
include/uapi/linux/bpf.h | 17 +++++++++++++++++ kernel/bpf/bpf_lsm.c | 30 ++++++++++++++++++++++++++++++ scripts/bpf_doc.py | 2 ++ tools/include/uapi/linux/bpf.h | 17 +++++++++++++++++ 4 files changed, 66 insertions(+)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index f4009dbdf62d..dfd93e0e0759 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -5249,6 +5249,22 @@ union bpf_attr {
Pointer to the underlying dynptr data, NULL if the dynptr isread-only, if the dynptr is invalid, or if the offset and lengthis out of bounds.
- struct key *bpf_request_key_by_id(unsigned long id)
- Description
Request a keyring by *id*.
*id* can have the following values (some defined in
verification.h): 0 for the primary keyring (immutable keyring of
system keys); 1 for both the primary and secondary keyring
(where keys can be added only if they are vouched for by
existing keys in those keyrings); 2 for the platform keyring
(primarily used by the integrity subsystem to verify a kexec'ed
kerned image and, possibly, the initramfs signature); ULONG_MAX
for the session keyring (for testing purposes).
It's never ok to add something like this to uapi 'for testing purposes'. If it's not useful in general it should not be a part of api.
- Return
A non-NULL pointer if *id* is valid and not 0, a NULL pointer*/
otherwise.#define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5455,6 +5471,7 @@ union bpf_attr { FN(dynptr_read), \ FN(dynptr_write), \ FN(dynptr_data), \
- FN(request_key_by_id), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index c1351df9f7ee..e1911812398b 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -16,6 +16,7 @@ #include <linux/bpf_local_storage.h> #include <linux/btf_ids.h> #include <linux/ima.h> +#include <linux/verification.h> /* For every LSM hook that allows attachment of BPF programs, declare a nop
- function where a BPF program can be attached.
@@ -132,6 +133,31 @@ static const struct bpf_func_proto bpf_get_attach_cookie_proto = { .arg1_type = ARG_PTR_TO_CTX, }; +#ifdef CONFIG_KEYS +BTF_ID_LIST_SINGLE(bpf_request_key_by_id_btf_ids, struct, key)
+BPF_CALL_1(bpf_request_key_by_id, unsigned long, id) +{
- const struct cred *cred = current_cred();
- if (id > (unsigned long)VERIFY_USE_PLATFORM_KEYRING && id != ULONG_MAX)
return (unsigned long)NULL;- if (id == ULONG_MAX)
return (unsigned long)cred->session_keyring;- return id;
It needs to do a proper lookup. Why cannot it do lookup_user_key ? The helper needs 'flags' arg too. Please think hard of extensibility and long term usefulness of api. At present this api feels like it was 'let me just hack something quickly'. Not ok.
From: Alexei Starovoitov alexei.starovoitov@gmail.com Sent: Friday, June 17, 2022 5:46 AM
Adding in CC the keyring mailing list and David.
Sort summary: we are adding an eBPF helper, to let eBPF programs verify PKCS#7 signatures. The helper simply calls verify_pkcs7_signature().
The problem is how to pass the key for verification.
For hardcoded keyring IDs, it is easy, pass 0, 1 or 2 for respectively the built-in, secondary and platform keyring.
If you want to pass another keyring, you need to do a lookup, which returns a key with reference count increased.
While in the kernel you can call key_put() to decrease the reference count, that is not guaranteed with an eBPF program, if the developer forgets about it. What probably is necessary, is to add the capability to the verifier to check whether the reference count is decreased, or adding a callback mechanism to call automatically key_put() when the eBPF program is terminated.
Is there an alternative solution?
Thanks
Roberto
HUAWEI TECHNOLOGIES Duesseldorf GmbH, HRB 56063 Managing Director: Li Peng, Yang Xi, Li He
On Tue, Jun 14, 2022 at 03:06:19PM +0200, Roberto Sassu wrote:
Add the bpf_request_key_by_id() helper, so that an eBPF program can obtain a suitable key pointer to pass to the bpf_verify_pkcs7_signature() helper, to be introduced in a later patch.
The passed identifier can have the following values: 0 for the primary keyring (immutable keyring of system keys); 1 for both the primary and secondary keyring (where keys can be added only if they are vouched for by existing keys in those keyrings); 2 for the platform keyring (primarily used by the integrity subsystem to verify a kexec'ed kerned image and, possibly, the initramfs signature); ULONG_MAX for the session keyring (for testing purposes).
Signed-off-by: Roberto Sassu roberto.sassu@huawei.com
include/uapi/linux/bpf.h | 17 +++++++++++++++++ kernel/bpf/bpf_lsm.c | 30 ++++++++++++++++++++++++++++++ scripts/bpf_doc.py | 2 ++ tools/include/uapi/linux/bpf.h | 17 +++++++++++++++++ 4 files changed, 66 insertions(+)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index f4009dbdf62d..dfd93e0e0759 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -5249,6 +5249,22 @@ union bpf_attr {
Pointer to the underlying dynptr data, NULL if the dynptr isread-only, if the dynptr is invalid, or if the offset and lengthis out of bounds.
- struct key *bpf_request_key_by_id(unsigned long id)
- Description
Request a keyring by *id*.
*id* can have the following values (some defined in
verification.h): 0 for the primary keyring (immutable keyringof
system keys); 1 for both the primary and secondary keyring
(where keys can be added only if they are vouched for by
existing keys in those keyrings); 2 for the platform keyring
(primarily used by the integrity subsystem to verify akexec'ed
kerned image and, possibly, the initramfs signature);ULONG_MAX
for the session keyring (for testing purposes).It's never ok to add something like this to uapi 'for testing purposes'. If it's not useful in general it should not be a part of api.
- Return
A non-NULL pointer if *id* is valid and not 0, a NULL pointer*/
otherwise.#define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5455,6 +5471,7 @@ union bpf_attr { FN(dynptr_read), \ FN(dynptr_write), \ FN(dynptr_data), \
- FN(request_key_by_id), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index c1351df9f7ee..e1911812398b 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -16,6 +16,7 @@ #include <linux/bpf_local_storage.h> #include <linux/btf_ids.h> #include <linux/ima.h> +#include <linux/verification.h>
/* For every LSM hook that allows attachment of BPF programs, declare a
nop
- function where a BPF program can be attached.
@@ -132,6 +133,31 @@ static const struct bpf_func_proto
bpf_get_attach_cookie_proto = {
.arg1_type = ARG_PTR_TO_CTX, };
+#ifdef CONFIG_KEYS +BTF_ID_LIST_SINGLE(bpf_request_key_by_id_btf_ids, struct, key)
+BPF_CALL_1(bpf_request_key_by_id, unsigned long, id) {
- const struct cred *cred = current_cred();
- if (id > (unsigned long)VERIFY_USE_PLATFORM_KEYRING && id !=
ULONG_MAX)
return (unsigned long)NULL;- if (id == ULONG_MAX)
return (unsigned long)cred->session_keyring;- return id;
It needs to do a proper lookup. Why cannot it do lookup_user_key ? The helper needs 'flags' arg too. Please think hard of extensibility and long term usefulness of api. At present this api feels like it was 'let me just hack something quickly'. Not ok.
On Fri, Jun 17, 2022 at 2:11 AM Roberto Sassu roberto.sassu@huawei.com wrote:
From: Alexei Starovoitov alexei.starovoitov@gmail.com Sent: Friday, June 17, 2022 5:46 AM
Adding in CC the keyring mailing list and David.
Sort summary: we are adding an eBPF helper, to let eBPF programs verify PKCS#7 signatures. The helper simply calls verify_pkcs7_signature().
The problem is how to pass the key for verification.
For hardcoded keyring IDs, it is easy, pass 0, 1 or 2 for respectively the built-in, secondary and platform keyring.
If you want to pass another keyring, you need to do a lookup, which returns a key with reference count increased.
While in the kernel you can call key_put() to decrease the reference count, that is not guaranteed with an eBPF program, if the developer forgets about it. What probably is necessary, is to add the capability to the verifier to check whether the reference count is decreased, or adding a callback mechanism to call automatically key_put() when the eBPF program is terminated.
Nothing special here. See acquire/release logic in the verifier and relevant helpers. Like bpf_sk_lookup_tcp and others.
Is there an alternative solution?
Thanks
Roberto
HUAWEI TECHNOLOGIES Duesseldorf GmbH, HRB 56063 Managing Director: Li Peng, Yang Xi, Li He
Please remove this footer from your emails.
Add the bpf_verify_pkcs7_signature() helper, to give eBPF security modules the ability to check the validity of a signature against supplied data, by using system-provided keys as trust anchor.
The new helper makes it possible to enforce mandatory policies, as eBPF programs might be allowed to make security decisions only based on data sources the system administrator approves.
The caller should provide both the data to be verified and the signature as eBPF dynamic pointers (to minimize the number of parameters). The caller should also provide a keyring pointer, that can be currently obtained with the bpf_request_key_by_id() helper. In the future, it will be possible to search a keyring by its description.
Signed-off-by: Roberto Sassu roberto.sassu@huawei.com Reported-by: kernel test robot lkp@intel.com (cast warning) --- include/uapi/linux/bpf.h | 8 ++++++++ kernel/bpf/bpf_lsm.c | 30 ++++++++++++++++++++++++++++++ tools/include/uapi/linux/bpf.h | 8 ++++++++ 3 files changed, 46 insertions(+)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index dfd93e0e0759..e6d9158e4981 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -5265,6 +5265,13 @@ union bpf_attr { * Return * A non-NULL pointer if *id* is valid and not 0, a NULL pointer * otherwise. + * + * long bpf_verify_pkcs7_signature(struct bpf_dynptr *data_ptr, struct bpf_dynptr *sig_ptr, struct key *trusted_keys) + * Description + * Verify the PKCS#7 signature *sig* against the supplied *data* + * with keys in *trusted_keys*. + * Return + * 0 on success, a negative value on error. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5472,6 +5479,7 @@ union bpf_attr { FN(dynptr_write), \ FN(dynptr_data), \ FN(request_key_by_id), \ + FN(verify_pkcs7_signature), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index e1911812398b..6c0d71b8c3b6 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -156,6 +156,31 @@ static const struct bpf_func_proto bpf_request_key_by_id_proto = { .ret_btf_id = &bpf_request_key_by_id_btf_ids[0], .arg1_type = ARG_ANYTHING, }; + +#ifdef CONFIG_SYSTEM_DATA_VERIFICATION +BPF_CALL_3(bpf_verify_pkcs7_signature, struct bpf_dynptr_kern *, data_ptr, + struct bpf_dynptr_kern *, sig_ptr, struct key *, trusted_keys) +{ + return verify_pkcs7_signature(data_ptr->data, + bpf_dynptr_get_size(data_ptr), + sig_ptr->data, + bpf_dynptr_get_size(sig_ptr), + trusted_keys, + VERIFYING_UNSPECIFIED_SIGNATURE, NULL, + NULL); +} + +static const struct bpf_func_proto bpf_verify_pkcs7_signature_proto = { + .func = bpf_verify_pkcs7_signature, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_DYNPTR | DYNPTR_TYPE_LOCAL, + .arg2_type = ARG_PTR_TO_DYNPTR | DYNPTR_TYPE_LOCAL, + .arg3_type = ARG_PTR_TO_BTF_ID_OR_NULL, + .arg3_btf_id = &bpf_request_key_by_id_btf_ids[0], + .allowed = bpf_ima_inode_hash_allowed, +}; +#endif /* CONFIG_SYSTEM_DATA_VERIFICATION */ #endif /* CONFIG_KEYS */
static const struct bpf_func_proto * @@ -187,6 +212,11 @@ bpf_lsm_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) #ifdef CONFIG_KEYS case BPF_FUNC_request_key_by_id: return &bpf_request_key_by_id_proto; +#ifdef CONFIG_SYSTEM_DATA_VERIFICATION + case BPF_FUNC_verify_pkcs7_signature: + return prog->aux->sleepable ? + &bpf_verify_pkcs7_signature_proto : NULL; +#endif /* CONFIG_SYSTEM_DATA_VERIFICATION */ #endif /* CONFIG_KEYS */ default: return tracing_prog_func_proto(func_id, prog); diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index dfd93e0e0759..e6d9158e4981 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -5265,6 +5265,13 @@ union bpf_attr { * Return * A non-NULL pointer if *id* is valid and not 0, a NULL pointer * otherwise. + * + * long bpf_verify_pkcs7_signature(struct bpf_dynptr *data_ptr, struct bpf_dynptr *sig_ptr, struct key *trusted_keys) + * Description + * Verify the PKCS#7 signature *sig* against the supplied *data* + * with keys in *trusted_keys*. + * Return + * 0 on success, a negative value on error. */ #define __BPF_FUNC_MAPPER(FN) \ FN(unspec), \ @@ -5472,6 +5479,7 @@ union bpf_attr { FN(dynptr_write), \ FN(dynptr_data), \ FN(request_key_by_id), \ + FN(verify_pkcs7_signature), \ /* */
/* integer value in 'imm' field of BPF_CALL instruction selects which helper
Ensure that signature verification is performed successfully from an eBPF program, with the new bpf_verify_pkcs7_signature() helper.
Generate a testing signature key and compile sign-file from scripts/, so that the test is selfcontained.
Signed-off-by: Roberto Sassu roberto.sassu@huawei.com --- tools/testing/selftests/bpf/Makefile | 14 +- tools/testing/selftests/bpf/config | 2 + .../bpf/prog_tests/verify_pkcs7_sig.c | 217 ++++++++++++++++++ .../bpf/progs/test_verify_pkcs7_sig.c | 168 ++++++++++++++ .../testing/selftests/bpf/verify_sig_setup.sh | 100 ++++++++ 5 files changed, 498 insertions(+), 3 deletions(-) create mode 100644 tools/testing/selftests/bpf/prog_tests/verify_pkcs7_sig.c create mode 100644 tools/testing/selftests/bpf/progs/test_verify_pkcs7_sig.c create mode 100755 tools/testing/selftests/bpf/verify_sig_setup.sh
diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index 8ad7a733a505..c689b63ecac2 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -14,6 +14,7 @@ BPFTOOLDIR := $(TOOLSDIR)/bpf/bpftool APIDIR := $(TOOLSINCDIR)/uapi GENDIR := $(abspath ../../../../include/generated) GENHDR := $(GENDIR)/autoconf.h +HOSTPKG_CONFIG := pkg-config
ifneq ($(wildcard $(GENHDR)),) GENFLAGS := -DHAVE_GENHDR @@ -75,7 +76,7 @@ TEST_PROGS := test_kmod.sh \ test_xsk.sh
TEST_PROGS_EXTENDED := with_addr.sh \ - with_tunnels.sh ima_setup.sh \ + with_tunnels.sh ima_setup.sh verify_sig_setup.sh \ test_xdp_vlan.sh test_bpftool.py
# Compile but not part of 'make run_tests' @@ -84,7 +85,7 @@ TEST_GEN_PROGS_EXTENDED = test_sock_addr test_skb_cgroup_id_user \ test_lirc_mode2_user xdping test_cpp runqslower bench bpf_testmod.ko \ xdpxceiver xdp_redirect_multi
-TEST_CUSTOM_PROGS = $(OUTPUT)/urandom_read +TEST_CUSTOM_PROGS = $(OUTPUT)/urandom_read $(OUTPUT)/sign-file
# Emit succinct information message describing current building step # $1 - generic step name (e.g., CC, LINK, etc); @@ -180,6 +181,12 @@ $(OUTPUT)/urandom_read: urandom_read.c urandom_read_aux.c $(OUTPUT)/liburandom_r liburandom_read.so $(LDLIBS) \ -Wl,-rpath=. -Wl,--build-id=sha1 -o $@
+$(OUTPUT)/sign-file: ../../../../scripts/sign-file.c + $(call msg,SIGN-FILE,,$@) + $(Q)$(CC) $(shell $(HOSTPKG_CONFIG)--cflags libcrypto 2> /dev/null) \ + $< -o $@ \ + $(shell $(HOSTPKG_CONFIG) --libs libcrypto 2> /dev/null || echo -lcrypto) + $(OUTPUT)/bpf_testmod.ko: $(VMLINUX_BTF) $(wildcard bpf_testmod/Makefile bpf_testmod/*.[ch]) $(call msg,MOD,,$@) $(Q)$(RM) bpf_testmod/bpf_testmod.ko # force re-compilation @@ -502,7 +509,8 @@ TRUNNER_EXTRA_SOURCES := test_progs.c cgroup_helpers.c trace_helpers.c \ cap_helpers.c TRUNNER_EXTRA_FILES := $(OUTPUT)/urandom_read $(OUTPUT)/bpf_testmod.ko \ $(OUTPUT)/liburandom_read.so \ - ima_setup.sh \ + $(OUTPUT)/sign-file \ + ima_setup.sh verify_sig_setup.sh \ $(wildcard progs/btf_dump_test_case_*.c) TRUNNER_BPF_BUILD_RULE := CLANG_BPF_BUILD_RULE TRUNNER_BPF_CFLAGS := $(BPF_CFLAGS) $(CLANG_CFLAGS) -DENABLE_ATOMICS_TESTS diff --git a/tools/testing/selftests/bpf/config b/tools/testing/selftests/bpf/config index 3b3edc0fc8a6..e4013c459f15 100644 --- a/tools/testing/selftests/bpf/config +++ b/tools/testing/selftests/bpf/config @@ -57,3 +57,5 @@ CONFIG_FPROBE=y CONFIG_IKCONFIG=y CONFIG_IKCONFIG_PROC=y CONFIG_MPTCP=y +CONFIG_MODULE_SIG=y +CONFIG_KEYS=y diff --git a/tools/testing/selftests/bpf/prog_tests/verify_pkcs7_sig.c b/tools/testing/selftests/bpf/prog_tests/verify_pkcs7_sig.c new file mode 100644 index 000000000000..c88bc563a71b --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/verify_pkcs7_sig.c @@ -0,0 +1,217 @@ +// SPDX-License-Identifier: GPL-2.0 + +/* + * Copyright (C) 2022 Huawei Technologies Duesseldorf GmbH + * + * Author: Roberto Sassu roberto.sassu@huawei.com + */ + +#include <stdio.h> +#include <errno.h> +#include <stdlib.h> +#include <unistd.h> +#include <endian.h> +#include <limits.h> +#include <sys/stat.h> +#include <sys/wait.h> +#include <linux/keyctl.h> +#include <test_progs.h> + +#include "test_verify_pkcs7_sig.skel.h" + +#define MAX_DATA_SIZE 4096 +#define LOG_BUF_SIZE 16384 + +struct data { + u8 payload[MAX_DATA_SIZE]; +}; + +static int _run_setup_process(const char *setup_dir, const char *cmd) +{ + int child_pid, child_status; + + child_pid = fork(); + if (child_pid == 0) { + execlp("./verify_sig_setup.sh", "./verify_sig_setup.sh", cmd, + setup_dir, NULL); + exit(errno); + + } else if (child_pid > 0) { + waitpid(child_pid, &child_status, 0); + return WEXITSTATUS(child_status); + } + + return -EINVAL; +} + +static int populate_data_item(const char *tmp_dir, struct data *data_item) +{ + struct stat st; + char signed_file_template[] = "/tmp/signed_fileXXXXXX"; + char path[PATH_MAX]; + int ret, fd, child_status, child_pid; + + fd = mkstemp(signed_file_template); + if (fd == -1) + return -errno; + + ret = write(fd, "test", 4); + + close(fd); + + if (ret != 4) { + ret = -EIO; + goto out; + } + + child_pid = fork(); + + if (child_pid == -1) { + ret = -errno; + goto out; + } + + if (child_pid == 0) { + snprintf(path, sizeof(path), "%s/signing_key.pem", tmp_dir); + + return execlp("./sign-file", "./sign-file", "sha256", + path, path, signed_file_template, NULL); + } + + waitpid(child_pid, &child_status, 0); + + ret = WEXITSTATUS(child_status); + if (ret) + goto out; + + ret = stat(signed_file_template, &st); + if (ret == -1) { + ret = -errno; + goto out; + } + + if (st.st_size > sizeof(data_item->payload) - sizeof(u32)) { + ret = -EINVAL; + goto out; + } + + *(u32 *)data_item->payload = __cpu_to_be32(st.st_size); + + fd = open(signed_file_template, O_RDONLY); + if (fd == -1) { + ret = -errno; + goto out; + } + + ret = read(fd, data_item->payload + sizeof(u32), st.st_size); + + close(fd); + + if (ret != st.st_size) { + ret = -EIO; + goto out; + } + + ret = 0; +out: + unlink(signed_file_template); + return ret; +} + +void test_verify_pkcs7_sig(void) +{ + char tmp_dir_template[] = "/tmp/verify_sigXXXXXX"; + char *tmp_dir; + char *buf = NULL; + struct test_verify_pkcs7_sig *skel = NULL; + struct bpf_map *map; + struct data data; + u32 saved_len; + int ret, zero = 0; + + LIBBPF_OPTS(bpf_object_open_opts, opts); + + /* Trigger creation of session keyring. */ + syscall(__NR_request_key, "keyring", "_uid.0", NULL, + KEY_SPEC_SESSION_KEYRING); + + tmp_dir = mkdtemp(tmp_dir_template); + if (!ASSERT_OK_PTR(tmp_dir, "mkdtemp")) + return; + + ret = _run_setup_process(tmp_dir, "setup"); + if (!ASSERT_OK(ret, "_run_setup_process")) + goto close_prog; + + buf = malloc(LOG_BUF_SIZE); + if (!ASSERT_OK_PTR(buf, "malloc")) + goto close_prog; + + opts.kernel_log_buf = buf; + opts.kernel_log_size = LOG_BUF_SIZE; + opts.kernel_log_level = 1; + + skel = test_verify_pkcs7_sig__open_opts(&opts); + if (!ASSERT_OK_PTR(skel, "test_verify_pkcs7_sig__open_opts")) + goto close_prog; + + ret = test_verify_pkcs7_sig__load(skel); + + if (ret < 0 && strstr(buf, "unknown func bpf_verify_pkcs7_signature")) { + printf( + "%s:SKIP:bpf_verify_pkcs7_signature() helper not supported\n", + __func__); + test__skip(); + goto close_prog; + } + + if (!ASSERT_OK(ret, "test_verify_pkcs7_sig__load\n")) + goto close_prog; + + ret = test_verify_pkcs7_sig__attach(skel); + if (!ASSERT_OK(ret, "test_verify_pkcs7_sig__attach\n")) + goto close_prog; + + map = bpf_object__find_map_by_name(skel->obj, "data_input"); + if (!ASSERT_OK_PTR(map, "data_input not found")) + goto close_prog; + + ret = populate_data_item(tmp_dir, &data); + if (!ASSERT_OK(ret, "populate_data_item\n")) + goto close_prog; + + skel->bss->monitored_pid = getpid(); + skel->bss->keyring_id = ULONG_MAX; + + ret = bpf_map_update_elem(bpf_map__fd(map), &zero, &data, BPF_ANY); + if (!ASSERT_OK(ret, "bpf_map_update_elem\n")) + goto close_prog; + + skel->bss->monitored_pid = getpid(); + /* Search the verification key in the primary keyring (should fail). */ + skel->bss->keyring_id = 0; + + ret = bpf_map_update_elem(bpf_map__fd(map), &zero, &data, BPF_ANY); + if (!ASSERT_LT(ret, 0, "bpf_map_update_elem data_input\n")) + goto close_prog; + + saved_len = *(__u32 *)data.payload; + *(__u32 *)data.payload = sizeof(data.payload); + ret = bpf_map_update_elem(bpf_map__fd(map), &zero, &data, BPF_ANY); + if (!ASSERT_LT(ret, 0, "bpf_map_update_elem data_input\n")) + goto close_prog; + + *(__u32 *)data.payload = saved_len; + data.payload[sizeof(__u32)] = 'a'; + ret = bpf_map_update_elem(bpf_map__fd(map), &zero, &data, BPF_ANY); + ASSERT_LT(ret, 0, "bpf_map_update_elem data_input\n"); +close_prog: + _run_setup_process(tmp_dir, "cleanup"); + free(buf); + + if (!skel) + return; + + skel->bss->monitored_pid = 0; + test_verify_pkcs7_sig__destroy(skel); +} diff --git a/tools/testing/selftests/bpf/progs/test_verify_pkcs7_sig.c b/tools/testing/selftests/bpf/progs/test_verify_pkcs7_sig.c new file mode 100644 index 000000000000..824379fa4a2c --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_verify_pkcs7_sig.c @@ -0,0 +1,168 @@ +// SPDX-License-Identifier: GPL-2.0 + +/* + * Copyright (C) 2022 Huawei Technologies Duesseldorf GmbH + * + * Author: Roberto Sassu roberto.sassu@huawei.com + */ + +#include <errno.h> +#include <stdlib.h> +#include <limits.h> +#include <linux/bpf.h> +#include <bpf/bpf_helpers.h> +#include <bpf/bpf_tracing.h> +#include <bpf/bpf_endian.h> + +#define MAX_DATA_SIZE 4096 + +#ifdef __BIG_ENDIAN__ +#define be32_to_cpu(x) (x) +#else +#define be32_to_cpu(x) ___bpf_swab32(x) +#endif + +#define VERIFY_USE_SECONDARY_KEYRING (1UL) + +/* In stripped ARM and x86-64 modules, ~ is surprisingly rare. */ +#define MODULE_SIG_STRING "~Module signature appended~\n" + +typedef __u8 u8; +typedef __u16 u16; +typedef __u32 u32; +typedef __u64 u64; + +enum pkey_id_type { + PKEY_ID_PGP, /* OpenPGP generated key ID */ + PKEY_ID_X509, /* X.509 arbitrary subjectKeyIdentifier */ + PKEY_ID_PKCS7, /* Signature in PKCS#7 message */ +}; + +/* + * Module signature information block. + * + * The constituents of the signature section are, in order: + * + * - Signer's name + * - Key identifier + * - Signature data + * - Information block + */ +struct module_signature { + u8 algo; /* Public-key crypto algorithm [0] */ + u8 hash; /* Digest algorithm [0] */ + u8 id_type; /* Key identifier type [PKEY_ID_PKCS7] */ + u8 signer_len; /* Length of signer's name [0] */ + u8 key_id_len; /* Length of key identifier [0] */ + u8 __pad[3]; + __be32 sig_len; /* Length of signature data */ +}; + +struct key; + +u32 monitored_pid; +unsigned long keyring_id; + +struct data { + u8 payload[MAX_DATA_SIZE]; +}; + +struct { + __uint(type, BPF_MAP_TYPE_ARRAY); + __uint(max_entries, 1); + __type(key, __u32); + __type(value, struct data); +} data_input SEC(".maps"); + +char _license[] SEC("license") = "GPL"; + +static int mod_check_sig(const struct module_signature *ms, size_t file_len) +{ + if (!ms) + return -ENOENT; + + if (be32_to_cpu(ms->sig_len) >= file_len - sizeof(*ms)) + return -EBADMSG; + + if (ms->id_type != PKEY_ID_PKCS7) + return -ENOPKG; + + if (ms->algo != 0 || + ms->hash != 0 || + ms->signer_len != 0 || + ms->key_id_len != 0 || + ms->__pad[0] != 0 || + ms->__pad[1] != 0 || + ms->__pad[2] != 0) + return -EBADMSG; + + return 0; +} + +SEC("lsm.s/bpf") +int BPF_PROG(bpf, int cmd, union bpf_attr *attr, unsigned int size) +{ + const size_t marker_len = sizeof(MODULE_SIG_STRING) - 1; + char marker[sizeof(MODULE_SIG_STRING) - 1]; + struct bpf_dynptr data_ptr, sig_ptr; + struct module_signature ms; + struct data *data_val; + struct key *trusted_keys; + u32 modlen; + u32 sig_len; + u64 value; + u8 *mod; + u32 pid; + int ret, zero = 0; + + pid = bpf_get_current_pid_tgid() >> 32; + if (pid != monitored_pid) + return 0; + + data_val = bpf_map_lookup_elem(&data_input, &zero); + if (!data_val) + return 0; + + bpf_probe_read(&value, sizeof(value), &attr->value); + + bpf_copy_from_user(data_val, sizeof(struct data), + (void *)(unsigned long)value); + + modlen = be32_to_cpu(*(u32 *)data_val->payload); + mod = data_val->payload + sizeof(u32); + + if (modlen > sizeof(struct data) - sizeof(u32)) + return -EINVAL; + + if (modlen <= marker_len) + return -ENOENT; + + modlen &= sizeof(struct data) - 1; + bpf_probe_read(marker, marker_len, (char *)mod + modlen - marker_len); + + if (bpf_strncmp(marker, marker_len, MODULE_SIG_STRING)) + return -ENOENT; + + modlen -= marker_len; + + if (modlen <= sizeof(ms)) + return -EBADMSG; + + bpf_probe_read(&ms, sizeof(ms), (char *)mod + (modlen - sizeof(ms))); + + ret = mod_check_sig(&ms, modlen); + if (ret) + return ret; + + sig_len = be32_to_cpu(ms.sig_len); + modlen -= sig_len + sizeof(ms); + + modlen &= 0x3ff; + sig_len &= 0x3ff; + + bpf_dynptr_from_mem(mod, modlen, 0, &data_ptr); + bpf_dynptr_from_mem(mod + modlen, sig_len, 0, &sig_ptr); + trusted_keys = bpf_request_key_by_id(keyring_id); + + return bpf_verify_pkcs7_signature(&data_ptr, &sig_ptr, trusted_keys); +} diff --git a/tools/testing/selftests/bpf/verify_sig_setup.sh b/tools/testing/selftests/bpf/verify_sig_setup.sh new file mode 100755 index 000000000000..e459362260f4 --- /dev/null +++ b/tools/testing/selftests/bpf/verify_sig_setup.sh @@ -0,0 +1,100 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 + +set -e +set -u +set -o pipefail + +VERBOSE="${SELFTESTS_VERBOSE:=0}" +LOG_FILE="$(mktemp /tmp/verify_sig_setup.log.XXXX)" + +x509_genkey_content="\ +[ req ] +default_bits = 2048 +distinguished_name = req_distinguished_name +prompt = no +string_mask = utf8only +x509_extensions = myexts + +[ req_distinguished_name ] +CN = eBPF Signature Verification Testing Key + +[ myexts ] +basicConstraints=critical,CA:FALSE +keyUsage=digitalSignature +subjectKeyIdentifier=hash +authorityKeyIdentifier=keyid +" + +usage() +{ + echo "Usage: $0 <setup|cleanup <existing_tmp_dir>" + exit 1 +} + +setup() +{ + local tmp_dir="$1" + + echo "${x509_genkey_content}" > ${tmp_dir}/x509.genkey + + openssl req -new -nodes -utf8 -sha256 -days 36500 \ + -batch -x509 -config ${tmp_dir}/x509.genkey \ + -outform PEM -out ${tmp_dir}/signing_key.pem \ + -keyout ${tmp_dir}/signing_key.pem 2>&1 + + openssl x509 -in ${tmp_dir}/signing_key.pem -out \ + ${tmp_dir}/signing_key.der -outform der + + cat ${tmp_dir}/signing_key.der | keyctl padd asymmetric ebpf_testing_key @s +} + +cleanup() { + local tmp_dir="$1" + + keyctl unlink $(keyctl search @s asymmetric ebpf_testing_key) @s + rm -rf ${tmp_dir} +} + +catch() +{ + local exit_code="$1" + local log_file="$2" + + if [[ "${exit_code}" -ne 0 ]]; then + cat "${log_file}" >&3 + fi + + rm -f "${log_file}" + exit ${exit_code} +} + +main() +{ + [[ $# -ne 2 ]] && usage + + local action="$1" + local tmp_dir="$2" + + [[ ! -d "${tmp_dir}" ]] && echo "Directory ${tmp_dir} doesn't exist" && exit 1 + + if [[ "${action}" == "setup" ]]; then + setup "${tmp_dir}" + elif [[ "${action}" == "cleanup" ]]; then + cleanup "${tmp_dir}" + else + echo "Unknown action: ${action}" + exit 1 + fi +} + +trap 'catch "$?" "${LOG_FILE}"' EXIT + +if [[ "${VERBOSE}" -eq 0 ]]; then + # Save the stderr to 3 so that we can output back to + # it incase of an error. + exec 3>&2 1>"${LOG_FILE}" 2>&1 +fi + +main "$@" +rm -f "${LOG_FILE}"
On Tue, Jun 14, 2022 at 03:06:21PM +0200, Roberto Sassu wrote:
- if (child_pid == 0) {
snprintf(path, sizeof(path), "%s/signing_key.pem", tmp_dir);return execlp("./sign-file", "./sign-file", "sha256",path, path, signed_file_template, NULL);
Please use sign_only option, so it saves the signature and doesn't do 'struct module_signature' append. Parsing of that is unnecessary for the purpose of the helper. Checking MODULE_SIG_STRING is unnecessary, etc, etc. Long term we won't be following mod sig approach anyway. bpf maps and progs will have a different format.
- }
- waitpid(child_pid, &child_status, 0);
- ret = WEXITSTATUS(child_status);
- if (ret)
goto out;- ret = stat(signed_file_template, &st);
- if (ret == -1) {
ret = -errno;goto out;- }
- if (st.st_size > sizeof(data_item->payload) - sizeof(u32)) {
ret = -EINVAL;goto out;- }
- *(u32 *)data_item->payload = __cpu_to_be32(st.st_size);
- fd = open(signed_file_template, O_RDONLY);
- if (fd == -1) {
ret = -errno;goto out;- }
- ret = read(fd, data_item->payload + sizeof(u32), st.st_size);
- close(fd);
- if (ret != st.st_size) {
ret = -EIO;goto out;- }
- ret = 0;
+out:
- unlink(signed_file_template);
- return ret;
+}
+void test_verify_pkcs7_sig(void) +{
- char tmp_dir_template[] = "/tmp/verify_sigXXXXXX";
- char *tmp_dir;
- char *buf = NULL;
- struct test_verify_pkcs7_sig *skel = NULL;
- struct bpf_map *map;
- struct data data;
- u32 saved_len;
- int ret, zero = 0;
- LIBBPF_OPTS(bpf_object_open_opts, opts);
- /* Trigger creation of session keyring. */
- syscall(__NR_request_key, "keyring", "_uid.0", NULL,
KEY_SPEC_SESSION_KEYRING);
My understanding that user space can receive a specific id here. It should pass it to bpf prog via global variable and prog should use that id instead of max_ulong hack.
linux-kselftest-mirror@lists.linaro.org
-
Alexei Starovoitov -
Roberto Sassu