On 2024/10/31 08:27, Quentin Monnet wrote: > 2024-10-30 17:47 UTC+0800 ~ Leon Hwang <hffilwlqm@xxxxxxxxx> >> From: Leon Hwang <leon.hwang@xxxxxxxxx> >> >> This patch addresses the bpftool issue "Wrong callq address displayed"[0]. >> >> The issue stemmed from an incorrect program counter (PC) value used during >> disassembly with LLVM or libbfd. To calculate the correct address for >> relative calls, the PC argument must reflect the actual address in the >> kernel. >> >> [0] https://github.com/libbpf/bpftool/issues/109 >> >> Fixes: e1947c750ffe ("bpftool: Refactor disassembler for JIT-ed programs") >> Signed-off-by: Leon Hwang <leon.hwang@xxxxxxxxx> >> --- >> tools/bpf/bpftool/jit_disasm.c | 6 +++--- >> 1 file changed, 3 insertions(+), 3 deletions(-) >> >> diff --git a/tools/bpf/bpftool/jit_disasm.c b/tools/bpf/bpftool/jit_disasm.c >> index 7b8d9ec89ebd3..fe8fabba4b05f 100644 >> --- a/tools/bpf/bpftool/jit_disasm.c >> +++ b/tools/bpf/bpftool/jit_disasm.c >> @@ -114,8 +114,7 @@ disassemble_insn(disasm_ctx_t *ctx, unsigned char *image, ssize_t len, int pc) >> char buf[256]; >> int count; >> >> - count = LLVMDisasmInstruction(*ctx, image + pc, len - pc, pc, >> - buf, sizeof(buf)); >> + count = LLVMDisasmInstruction(*ctx, image, len, pc, buf, sizeof(buf)); >> if (json_output) >> printf_json(buf); >> else >> @@ -360,7 +359,8 @@ int disasm_print_insn(unsigned char *image, ssize_t len, int opcodes, >> printf("%4x:" DISASM_SPACER, pc); >> } >> >> - count = disassemble_insn(&ctx, image, len, pc); >> + count = disassemble_insn(&ctx, image + pc, len - pc, >> + func_ksym + pc); > > Thanks a lot for looking into this! Your patch does solve the issue for > the LLVM disassembler (nice!), but it breaks the libbfd one: > > > $ ./bpftool version | grep features > features: libbfd > # ./bpftool prog dump j id 111 op > int xdp_redirect_map_0(struct xdp_md * xdp): > bpf_prog_a8f6f9c4be77b94c_xdp_redirect_map_0: > ; return bpf_redirect_map(&tx_port, 0, 0); > 0: Address 0xffffffffc01ae950 is out of bounds. > > I don't think we can change the PC in the case of libbfd, as far as I > can tell it needs to point to the first instruction to disassemble. Two > of the arguments we pass to disassemble_insn(), image and len, are > ignored by the libbfd disassembler; so it leaves only the ctx argument > that we can maybe update to pass the func_ksym, but I haven't found how > to do that yet (if possible at all). > > Thanks, > Quentin > Hi Quentin, After diving into the details of libbfd, I’ve found a way to correct the callq address. By adjusting the relative addresses using func_ksym within a custom info->print_addr_func, we can achieve accurate results. Here’s the updated patch:
