libbpf: Replace strncpy() with strnlen()+memcpy() in skel_map_create()
PATCH 1/1 libbpf: Replace strncpy() with strnlen()+memcpy() in skel_map_create()
Replace the deprecated[1] strncpy() with strnlen() on the source followed by memcpy(). Normally strscpy() would be used in this case, but skel_internal.h is shared between kernel and userspace tools, and strscpy() is not available in the userspace build context.
The source map_name is a NUL-terminated C string (the only caller passes a 12 character string literal). The destination attr.map_name is char[BPF_OBJ_NAME_LEN] (16 bytes) in union bpf_attr, passed to the bpf() syscall. The kernel's bpf_obj_name_cpy() requires a NUL terminator within the 16-byte field, rejecting names that use all 16 bytes. Valid names are therefore at most 15 characters.
The attr is pre-zeroed with memset() at the top of the function, so the byte at position 15 is always NUL. The copy is bounded to sizeof(attr.map_name) - 1 (15 bytes) to guarantee NUL-termination is preserved. This is safe because the kernel would reject a 16-byte unterminated name anyway, and the only in-tree caller passes "__loader.map" (12 characters).
While the original strncpy() would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.
---
Cc: Andrii Nakryiko <[email protected]>
Cc: Eduard Zingerman <[email protected]>
Cc: Alexei Starovoitov <[email protected]>
Cc: Daniel Borkmann <[email protected]>
Cc: Martin KaFai Lau <[email protected]>
Cc: Song Liu <[email protected]>
Cc: Yonghong Song <[email protected]>
Cc: John Fastabend <[email protected]>
Cc: KP Singh <[email protected]>
Cc: Stanislav Fomichev <[email protected]>
Cc: Hao Luo <[email protected]>
Cc: Jiri Olsa <[email protected]>
Cc: <[email protected]>
--- tools/lib/bpf/skel_internal.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
tools/lib/bpf/skel_internal.h +2 -1
@@ -243,7 +243,8 @@ static inline int skel_map_create(enum bpf_map_type map_type, attr.excl_prog_hash = (unsigned long ) excl_prog_hash; attr.excl_prog_hash_size = excl_prog_hash_sz; -strncpy (attr.map_name, map_name,sizeof (attr.map_name)); +memcpy (attr.map_name, map_name, +strnlen (map_name,sizeof (attr.map_name) -1 )); attr.key_size = key_size; attr.value_size = value_size; attr.max_entries = max_entries;
Signature · 1 line
2.34.1
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.
Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546
Signature · 1 line
Kees Cook
Quoted 9 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.
Quoted 6 lines
Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546
IMHO, it's a flaky test failure, likely caused by timer-related activity when selftests/bpf
are running concurrently.
As for your patch, replacing strncpy() makes sense, but this also changes semantics
for overlong map names: they are silently truncated to 15 bytes. I'd prefer an explicit length check and rejecting the overlong names, instead of normalizing
them by truncation.
Sun Jian
Quoted 9 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.
Quoted 6 lines
Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546
Don't fix what is not broken.
pw-bot: cr
Quoted 10 lines · Alexei Starovoitov
Quoted 9 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.
Quoted 8 lines
Quoted 6 lines
Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546Don't fix what is not broken.
strncpy is broken. ;) This is one of the 6 remaining uses of strncpy in the kernel. But it needs a v3. I'll add explicit rejection of over-long strings and validate that the bpf test was a flake, as suggested in the other reply.
-Kees
Signature · 1 line
Kees Cook
Quoted 18 lines · Kees Cook
Quoted 14 lines · Alexei Starovoitov
Quoted 13 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546Don't fix what is not broken.
strncpy is broken. ;) This is one of the 6 remaining uses of strncpy in the kernel. But it needs a v3. I'll add explicit rejection of over-long strings and validate that the bpf test was a flake, as suggested in the other reply.
Don't. it's not a kernel.
Quoted 19 lines · Alexei Starovoitov
Quoted 18 lines · Kees Cook
Quoted 14 lines · Alexei Starovoitov
Quoted 13 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546Don't fix what is not broken.
strncpy is broken. ;) This is one of the 6 remaining uses of strncpy in the kernel. But it needs a v3. I'll add explicit rejection of over-long strings and validate that the bpf test was a flake, as suggested in the other reply.
Don't. it's not a kernel.
This is a very terse reply, so I'm not sure what you're trying to tell me.
Do you mean to say that this header is not used by the Linux kernel build? If so, that's a misunderstanding. I found this case due to 0day complaining about it after I removed the kernel's strncpy implementations:
https://lore.kernel.org/all/[email protected]/
https://lore.kernel.org/all/[email protected]/
(I had originally ignored tools/ since normally the files there are not included in kernel builds, but this didn't seem to be the case for this BPF header.)
-Kees
Signature · 1 line
Kees Cook
Quoted 30 lines · Kees Cook
Quoted 19 lines · Alexei Starovoitov
Quoted 18 lines · Kees Cook
Quoted 14 lines · Alexei Starovoitov
Quoted 13 lines · Kees Cook
Quoted 6 lines · Kees Cook
While the original
strncpy()would have copied a full 16 bytes from an overlong name (producing an unterminated field that the syscall rejects), but this wasn't a reachable state. This replacement will instead always truncate to 15 bytes and keeps the NUL terminator, which should have no behavioral changes with the present code and avoids potential issues with future over-long string literals.Hm, I got a failure report, but it *seems* unrelated? But nothing else fails that way recently, so I will try a v2 with the "unterminated at 16 bytes" behavior restored and see if it passes...
Quoted 8 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
test_progs_no_alu32-x86_64-llvm-21:
https://github.com/kernel-patches/bpf/actions/runs/23472955268/job/68300440546Don't fix what is not broken.
strncpy is broken. ;) This is one of the 6 remaining uses of strncpy in the kernel. But it needs a v3. I'll add explicit rejection of over-long strings and validate that the bpf test was a flake, as suggested in the other reply.
Don't. it's not a kernel.
Quoted 8 lines
This is a very terse reply, so I'm not sure what you're trying to tell me.
Do you mean to say that this header is not used by the Linux kernel build? If so, that's a misunderstanding. I found this case due to 0day complaining about it after I removed the kernel's strncpy implementations:
https://lore.kernel.org/all/[email protected]/
https://lore.kernel.org/all/[email protected]/
skel_internal.h is dual purpose.
In that case it's used by "kernel light skeleton".
but mainly used by "user space light skeleton".
I guess it's a grey line whether it can be considered kernel or not.
Quoted 3 lines
(I had originally ignored
tools/since normally the files there are not included in kernel builds, but this didn't seem to be the case for this BPF header.)
And please keep ignoring it.
It's not broken. Don't fix it.
Quoted 2 lines · Alexei Starovoitov
And please keep ignoring it.
It's not broken. Don't fix it.
Okay. I'm confused, though: what's your plan for when strncpy is removed from the kernel?
Signature · 1 line
Kees Cook
Quoted 7 lines · Kees Cook
Quoted 2 lines · Alexei Starovoitov
And please keep ignoring it.
It's not broken. Don't fix it.Okay. I'm confused, though: what's your plan for when strncpy is removed from the kernel?
when it gets to the point to be actually removed then we will deal with it. Until then it's a witch-hunt.memcpy(,, strnlen()) is no better than strncpy.