Re: [RFC PATCH v3 06/24] x86/cet: Control protection exception handler

[Yu-cheng Yu]

On Fri, 2018-08-31 at 17:01 +0200, Jann Horn wrote:

> Is there a reason why all the code in this patch isn't #ifdef'ed
> away
> on builds that don't support CET? It looks like the CET handler is
> hooked up to the IDT conditionally, but the handler code is always
> built?

Yes, in idt.c, it should have been:

#ifdef CONFIG_X86_64
	INTG(X86_TRAP_CP,		control_protection),
#endif

I will fix it.

> > +dotraplinkage void
> > +do_control_protection(struct pt_regs *regs, long error_code)
> > +{
> > +ÂÂÂÂÂÂÂstruct task_struct *tsk;
> > +
> > +ÂÂÂÂÂÂÂRCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't
> > wake RCU");
> > +ÂÂÂÂÂÂÂif (notify_die(DIE_TRAP, "control protection fault", regs,
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂerror_code, X86_TRAP_CP, SIGSEGV) ==
> > NOTIFY_STOP)
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂreturn;
> > +ÂÂÂÂÂÂÂcond_local_irq_enable(regs);
> > +
> > +ÂÂÂÂÂÂÂif (!user_mode(regs))
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂdie("kernel control protection fault", regs,
> > error_code);
> > +
> > +ÂÂÂÂÂÂÂif (!static_cpu_has(X86_FEATURE_SHSTK) &&
> > +ÂÂÂÂÂÂÂÂÂÂÂ!static_cpu_has(X86_FEATURE_IBT))
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂWARN_ONCE(1, "CET is disabled but got control "
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ"protection fault\n");
> > +
> > +ÂÂÂÂÂÂÂtsk = current;
> > +ÂÂÂÂÂÂÂtsk->thread.error_code = error_code;
> > +ÂÂÂÂÂÂÂtsk->thread.trap_nr = X86_TRAP_CP;
> > +
> > +ÂÂÂÂÂÂÂif (show_unhandled_signals && unhandled_signal(tsk,
> > SIGSEGV) &&
> > +ÂÂÂÂÂÂÂÂÂÂÂprintk_ratelimit()) {
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂunsigned int max_err;
> > +
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂmax_err = ARRAY_SIZE(control_protection_err) - 1;
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂif ((error_code < 0) || (error_code > max_err))
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂerror_code = 0;
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂpr_info("%s[%d] control protection ip:%lx sp:%lx
> > error:%lx(%s)",
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂtsk->comm, task_pid_nr(tsk),
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂregs->ip, regs->sp, error_code,
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂcontrol_protection_err[error_code]);
> > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂprint_vma_addr(" in ", regs->ip);
> Shouldn't this be using KERN_CONT, like other callers of
> print_vma_addr(), to get the desired output?

I will change it.