Re: [RFC][PATCH] sched,livepatch: Untangle cond_resched() and live-patching

[Miroslav Benes]

>  void __klp_sched_try_switch(void)
>  {
> -	if (likely(!klp_patch_pending(current)))
> -		return;
> -
>  	/*
>  	 * This function is called from cond_resched() which is called in many
>  	 * places throughout the kernel.  Using the klp_mutex here might
> @@ -377,14 +365,14 @@ void __klp_sched_try_switch(void)
>  	 * klp_try_switch_task().  Thanks to task_call_func() they won't be
>  	 * able to switch this task while it's running.
>  	 */
> -	preempt_disable();
> +	lockdep_assert_preemption_disabled();
>  
>  	/*
>  	 * Make sure current didn't get patched between the above check and
>  	 * preempt_disable().
>  	 */
>  	if (unlikely(!klp_patch_pending(current)))
> -		goto out;
> +		return;

It does not look correct. We just make sure that 
klp_patch_pending(current) did not change here. It would be highly 
unlikely. However, we should keep the likely way out (the first removed 
condition above). So let's also s/unlikely/likely/ here.

And the comments in the function should be updated as well.

Miroslav