Re: [BUG] Use of probe_kernel_address() in task_rcu_dereference() without checking return value

[Oleg Nesterov]

On 08/30, Linus Torvalds wrote:
>
> Side note: that code had better not be performance-critical, because
> "probe_kernel_address()" is actually really really slow.

Yes, please see

	[PATCH 2/3] introduce probe_slab_address()
	https://lore.kernel.org/lkml/20141027195425.GC11736@xxxxxxxxxx/

I sent 5 years ago ;) Do you think

	/*
	 * Same as probe_kernel_address(), but @addr must be the valid pointer
	 * to a slab object, potentially freed/reused/unmapped.
	 */
	#ifdef CONFIG_DEBUG_PAGEALLOC
	#define probe_slab_address(addr, retval)	\
		probe_kernel_address(addr, retval)
	#else
	#define probe_slab_address(addr, retval)	\
		({							\
			(retval) = *(typeof(retval) *)(addr);		\
			0;						\
		})
	#endif

can work?

Oleg.