Re: [PATCH v6 10/20] x86/split_lock: Handle #AC exception for split lock

[Thomas Gleixner]

On Wed, 3 Apr 2019, Fenghua Yu wrote:
> +dotraplinkage void do_alignment_check(struct pt_regs *regs, long error_code)
> +{
> +	unsigned int trapnr = X86_TRAP_AC;
> +	char str[] = "alignment check";
> +	int signr = SIGBUS;
> +
> +	RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
> +
> +	/*
> +	 * WARN*()s end up here; fix them up before we call the
> +	 * notifier chain.
> +	 */

How exactly is WARN*() ending up here?

> +	if (!user_mode(regs) && fixup_bug(regs, trapnr))

And that fixup_bug() check does what?

int fixup_bug(struct pt_regs *regs, int trapnr)
{
 	if (trapnr != X86_TRAP_UD)
                return 0;

Copy and paste from do_error_trap() ....

> +		return;
> +
> +	if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) ==
> +		       NOTIFY_STOP)
> +		return;
> +
> +	cond_local_irq_enable(regs);
> +	if (!user_mode(regs) &&
> +	    static_cpu_has(X86_FEATURE_SPLIT_LOCK_DETECT)) {
> +		/*
> +		 * Only split lock can generate #AC from kernel at this point.
> +		 * Warn and disable split lock detection on this CPU. The
> +		 * faulting instruction will be executed without generating
> +		 * another #AC fault. User needs to check the warning and
> +		 * fix the split lock issue in the faulting instruction.

  "User needs to check the warning and fix the issue ..."

I'm looking forward to all the fixes from Joe Users.

Please remove that sentence. It's useless. Users report warnings if at all
and the kernel developers who actually look at them surely don't need an
advice like that.

Thanks,

	tglx