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 to retrieve the trusted keyring from function parameters, and to perform signature verification by calling 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 can be provided by its serial, or by its special ID defined in verification.h, if the serial is zero (not a valid value). bpf_verify_pkcs7_signature() also accepts key lookup-specific flags, passed to lookup_user_key() when the helper searches the keyring by its serial.
While passing the keyring serial to bpf_verify_pkcs7_signature() provides great flexibility, it seems suboptimal in terms of security guarantees, as even if the eBPF program is assumed to be trusted, that serial might come from untrusted user space not choosing one that the system administrator approves to enforce a mandatory policy. The same goal could be instead more easily achieved by setting a hardcoded keyring ID in the signed eBPF program, to be passed to bpf_verify_pkcs7_signature().
bpf_verify_pkcs7_signature() can be called only from sleepable programs, because of memory allocation (with lookup flag KEY_LOOKUP_CREATE) and crypto operations. For example, the lsm.s/bpf attach point is suitable, fexit/array_map_update_elem is not.
A test was added to check the ability of bpf_verify_pkcs7_signature() to verify PKCS#7 signatures from the session keyring, a newly-created keyring, and from the primary and secondary keyring (taking the tcp_bic.ko kernel module for the verification). The test does not fail if that kernel module is not found (needs support from the CI).
A consideration was made on whether bpf_verify_pkcs7_signature() should be a simple wrapper, doing as little as possible, or whether it could have more complex logic. Having a simple and flexible wrapper requires two additional helpers, bpf_lookup_user_key() and bpf_key_put(), to search and acquire a key reference, pass that key to the wrapper, and release the reference. More care is also required on the eBPF verifier side, to ensure that an eBPF program always releases an acquired reference.
While that gives eBPF developers the greatest flexibility to use the helpers as necessary, it does not match the security of the solution of retrieving the key and using it within the same function, as for example in security/keys/keyctl.c. The risk is that an eBPF program requests a key for a purpose, and then uses the key in a different way with one of the available key-related helpers (to be added in the future).
struct key is not like a file descriptor, carrying permissions requested during an open, that can be revalidated at the time a read or write is performed. It is more close to a struct inode, the function using the key cannot know reliably which permission was requested at lookup time.
For that reason, the key lookup and usage cannot be separated, as the kernel will guarantee (also to other MAC mechanisms) that once a key has been requested with a specific purpose, it will be used accordingly, beyond the control of eBFP programs.
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 makes available for new eBPF helpers some key-related definitions. Patch 3 fixes the helper prototype regular expression to accept unsigned as type prefix. Patch 4 introduces the bpf_verify_pkcs7_signature() helper and patch 5 adds the corresponding test.
Changelog
v5: - Move KEY_LOOKUP_ to include/linux/key.h for validation of bpf_verify_pkcs7_signature() parameter - Remove bpf_lookup_user_key() and bpf_key_put() helpers, and the corresponding tests - Replace struct key parameter of bpf_verify_pkcs7_signature() with the keyring serial and lookup flags - Call lookup_user_key() and key_put() in bpf_verify_pkcs7_signature() code, to ensure that the retrieved key is used according to the permission requested at lookup time - Clarified keyring precedence in the description of bpf_verify_pkcs7_signature() (suggested by John) - Remove newline in the second argument of ASSERT_ - Fix helper prototype regular expression in bpf_doc.py
v4: - Remove bpf_request_key_by_id(), don't return an invalid pointer that other helpers can use - Pass the keyring ID (without ULONG_MAX, suggested by Alexei) to bpf_verify_pkcs7_signature() - Introduce bpf_lookup_user_key() and bpf_key_put() helpers (suggested by Alexei) - Add lookup_key_norelease test, to ensure that the verifier blocks eBPF programs which don't decrement the key reference count - Parse raw PKCS#7 signature instead of module-style signature in the verify_pkcs7_signature test (suggested by Alexei) - Parse kernel module in user space and pass raw PKCS#7 signature to the eBPF program for signature verification
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 (5): bpf: Export bpf_dynptr_get_size() KEYS: Move KEY_LOOKUP_ to include/linux/key.h scripts: Handle unsigned type prefix in bpf_doc.py bpf: Add bpf_verify_pkcs7_signature() helper selftests/bpf: Add test for bpf_verify_pkcs7_signature() helper
include/linux/bpf.h | 1 + include/linux/key.h | 3 + include/uapi/linux/bpf.h | 24 ++ kernel/bpf/bpf_lsm.c | 63 +++ kernel/bpf/helpers.c | 2 +- scripts/bpf_doc.py | 2 +- security/keys/internal.h | 2 - tools/include/uapi/linux/bpf.h | 24 ++ tools/testing/selftests/bpf/Makefile | 14 +- tools/testing/selftests/bpf/config | 2 + .../bpf/prog_tests/verify_pkcs7_sig.c | 359 ++++++++++++++++++ .../bpf/progs/test_verify_pkcs7_sig.c | 79 ++++ .../testing/selftests/bpf/verify_sig_setup.sh | 104 +++++ 13 files changed, 672 insertions(+), 7 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