Re: [PATCH] mm/slub: mark racy accesses on slab->slabs

[Chengming Zhou]

On 2024/3/21 10:48, linke li wrote:
> Sorry for a late reply, I just found this because of my bad email client.
> 
>> There is another unmarked access of "slab->slabs" in the show_slab_objects(),
>> which you can change too.
> 
> Yes, I think show_slab_objects() has a similar situation. Should I
> consider to submit a V2 patch for this?

Yes, I think so.

> 
>> I'm not sure that it's really safe to access "slab->slabs" here without any protection?
>> Although it should be no problem in practice, alternative choice maybe putting partial
>> slabs count in the kmem_cache_cpu struct.
> 
> I think it is ok, because it seems that slab->slabs in slub_percpu_partial
> and show_slab_objects() are just used for showing some infomation.
> 
> I noticed Paul summarized some of these strategies in access-marking.txt[1]

Ok, thanks.

> 
> Quote from it:
> 
> "Use of the data_race() Macro
> ----------------------------
> 
> Here are some situations where data_race() should be used instead of
> READ_ONCE() and WRITE_ONCE():
> 
> 1.	Data-racy loads from shared variables whose values are used only
> 	for diagnostic purposes.
> 
> 2.	Data-racy reads whose values are checked against marked reload.
> 
> 3.	Reads whose values feed into error-tolerant heuristics.
> 
> 4.	Writes setting values that feed into error-tolerant heuristics.
> "
> 
> Thanks,
> Linke
> 
> [1]https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/tools/memory-model/Documentation/access-marking.txt
>