Re: [PATCH -tip v2] x86/kprobes: Do not decode opcode in resume_execution()

[Masami Hiramatsu]

On Thu, 31 Dec 2020 17:09:23 +0100
Borislav Petkov <bp@xxxxxxxxx> wrote:

> On Fri, Dec 18, 2020 at 11:12:05PM +0900, Masami Hiramatsu wrote:
> > @@ -467,8 +489,8 @@ static int arch_copy_kprobe(struct kprobe *p)
> >  	 */
> >  	len = prepare_boost(buf, p, &insn);
> >  
> > -	/* Check whether the instruction modifies Interrupt Flag or not */
> > -	p->ainsn.if_modifier = is_IF_modifier(buf);
> > +	/* Analyze the opcode and set resume flags */
> > +	set_resume_flags(p, &insn);
> 
> So this function wants to be called something like analyze_insn() or so
> then? set_resume_flags() is kinda misleading as a name.

Hrm, I meant setting the flags used in the resume_execution() afterwards.
Since the instruction itself (not only opcode but also oprands) was
also analyzed in other places, so I like the set_resume_flags() for it.

Thank you,

-- 
Masami Hiramatsu <mhiramat@xxxxxxxxxx>