Messages in this thread | | | Date | Thu, 1 Apr 2021 10:54:22 +0900 | From | Masami Hiramatsu <> | Subject | Re: [RFC PATCH -tip 3/3] x86/kprobes,orc: Unwind optprobe trampoline correctly |
| |
On Wed, 31 Mar 2021 14:44:56 +0900 Masami Hiramatsu <mhiramat@kernel.org> wrote:
> +#ifdef CONFIG_UNWINDER_ORC > +unsigned long recover_optprobe_trampoline(unsigned long addr, unsigned long *sp) > +{ > + unsigned long offset, entry, probe_addr; > + struct optimized_kprobe *op; > + struct orc_entry *orc; > + > + entry = find_kprobe_optinsn_slot_entry(addr); > + if (!entry) > + return addr; > + > + offset = addr - entry; > + > + /* Decode arg1 and get the optprobe */ > + op = (void *)extract_set_arg1((void *)(entry + TMPL_MOVE_IDX)); > + if (!op) > + return addr; > + > + probe_addr = (unsigned long)op->kp.addr; > + > + if (offset < TMPL_END_IDX) {
Maybe I should add a comment here.
/* * Since this is on the trampoline code based on the template code, * ORC information on the template code can be used for adjusting * stack pointer. The template code may change the stack pointer to * store pt_regs. */
> + orc = orc_find((unsigned long)optprobe_template_func + offset); > + if (!orc || orc->sp_reg != ORC_REG_SP) > + return addr; > + /* > + * Since optprobe trampoline doesn't push caller on the stack, > + * need to decrement 1 stack entry size > + */ > + *sp += orc->sp_offset - sizeof(long); > + return probe_addr; > + } else {
/* * In this case, the address is on the instruction copied from probed * address, and jump instruction. Here the stack pointer is exactly * the same as the value where it was copied by the kprobes. */
> + return probe_addr + offset - TMPL_END_IDX; > + } > +} > +#endif > +
Thank you,
-- Masami Hiramatsu <mhiramat@kernel.org>
| |