CVE-2024-47728

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> bpf: Zero former ARG_PTR_TO_{LONG,INT} args in case of error<br /> <br /> For all non-tracing helpers which formerly had ARG_PTR_TO_{LONG,INT} as input<br /> arguments, zero the value for the case of an error as otherwise it could leak<br /> memory. For tracing, it is not needed given CAP_PERFMON can already read all<br /> kernel memory anyway hence bpf_get_func_arg() and bpf_get_func_ret() is skipped<br /> in here.<br /> <br /> Also, the MTU helpers mtu_len pointer value is being written but also read.<br /> Technically, the MEM_UNINIT should not be there in order to always force init.<br /> Removing MEM_UNINIT needs more verifier rework though: MEM_UNINIT right now<br /> implies two things actually: i) write into memory, ii) memory does not have<br /> to be initialized. If we lift MEM_UNINIT, it then becomes: i) read into memory,<br /> ii) memory must be initialized. This means that for bpf_*_check_mtu() we&amp;#39;re<br /> readding the issue we&amp;#39;re trying to fix, that is, it would then be able to<br /> write back into things like .rodata BPF maps. Follow-up work will rework the<br /> MEM_UNINIT semantics such that the intent can be better expressed. For now<br /> just clear the *mtu_len on error path which can be lifted later again.