Re: An invalid memory access was discovered by a fuzz test

[Daniel Borkmann <daniel@xxxxxxxxxxxxx>]

On 12/19/23 3:15 PM, Xin Liu wrote:
> Hi all:
>
> The issue occurred while reading an ELF file in libbpf.c during fuzzing
>
>      Using host libthread_db library "/usr/lib64/libthread_db.so.1".
>      0.000243187s DEBUG total counters = 7816
>      0.000346533s DEBUG binary maps to 400000-155f280, len = 18215552
>      0.000765462s DEBUG init_fuzzer:run_seed: running initial seed path="crash-sigsegv-b905489aaeb39555ff1245117f1efd1677195b9ac1437bfb18b8d2d04099704b"
>
>      Program received signal SIGSEGV, Segmentation fault.
>      0x0000000000958e97 in bpf_object.collect_prog_relos () at libbpf.c:4206
>      4206 in libbpf.c
>      (gdb) bt
>      #0 0x0000000000958e97 in bpf_object.collect_prog_relos () at libbpf.c:4206
>      #1 0x000000000094f9d6 in bpf_object.collect_relos () at libbpf.c:6706
>      #2 0x000000000092bef3 in bpf_object_open () at libbpf.c:7437
>      #3 0x000000000092c046 in bpf_object.open_mem () at libbpf.c:7497
>      #4 0x0000000000924afa in LLVMFuzzerTestOneInput () at fuzz/bpf-object-fuzzer.c:16
>      #5 0x000000000060be11 in testblitz_engine::fuzzer::Fuzzer::run_one ()
>      #6 0x000000000087ad92 in tracing::span::Span::in_scope ()
>      #7 0x00000000006078aa in testblitz_engine::fuzzer::util::walkdir ()
>      #8 0x00000000005f3217 in testblitz_engine::entrypoint::main::{{closure}} ()
>      #9 0x00000000005f2601 in main ()
>      (gdb)
>
> then, I checked the code and found that scn_data was null at this code(tools/lib/bpf/src/libbpf.c):
>
>      if (rel->r_offset % BPF_INSN_SZ || rel->r_offset >= scn_data->d_size) {
>      
> The scn_data is derived from the code above:
>      
>      scn = elf_sec_by_idx(obj, sec_idx);
>      scn_data = elf_sec_data(obj, scn);
>      
>      relo_sec_name = elf_sec_str(obj, shdr->sh_name);
>      sec_name = elf_sec_name(obj, scn);
>      if (!relo_sec_name || !sec_name)    // don't check whether scn_data is NULL
>      	return -EINVAL;
>
> Do sec_data and sec_name always occur together? Is it possible that scn_data is NULL but sec_name
> is not NULL? libbpf uses sec_name to determine if it’s a null pointer, Maybe we should do some
> check here.

Weird, is this based on a malformed elf given sec_idx comes from shdr->sh_info?
It probably makes sense to NULL check and then return with -LIBBPF_ERRNO__FORMAT
as we do elsewhere. Do you want to send a fix?

Thanks,
Daniel