CVE-2025-37893

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> LoongArch: BPF: Fix off-by-one error in build_prologue()<br /> <br /> Vincent reported that running BPF progs with tailcalls on LoongArch<br /> causes kernel hard lockup. Debugging the issues shows that the JITed<br /> image missing a jirl instruction at the end of the epilogue.<br /> <br /> There are two passes in JIT compiling, the first pass set the flags and<br /> the second pass generates JIT code based on those flags. With BPF progs<br /> mixing bpf2bpf and tailcalls, build_prologue() generates N insns in the<br /> first pass and then generates N+1 insns in the second pass. This makes<br /> epilogue_offset off by one and we will jump to some unexpected insn and<br /> cause lockup. Fix this by inserting a nop insn.