2022-02-21 23:19:06

by Matthew Wilcox

[permalink] [raw]
Subject: Re: [PATCH 2/5] mm/sl[auo]b: Do not export __ksize()

On Mon, Feb 21, 2022 at 10:53:33AM +0000, Hyeonggon Yoo wrote:
> Do not export __ksize(). Only kasan calls __ksize() directly.

We should probably delete (most of) the kernel-doc comment on __ksize,
leaving only the note that it's uninstrumented and only called by
kasan. Also, we should probably move the declaration of __ksize from
include/linux/slab.h to mm/slab.h since we don't want to grow new callers.


2022-02-23 05:24:48

by Hyeonggon Yoo

[permalink] [raw]
Subject: Re: [PATCH 2/5] mm/sl[auo]b: Do not export __ksize()

On Mon, Feb 21, 2022 at 03:46:01PM +0000, Matthew Wilcox wrote:
> On Mon, Feb 21, 2022 at 10:53:33AM +0000, Hyeonggon Yoo wrote:
> > Do not export __ksize(). Only kasan calls __ksize() directly.
>
> We should probably delete (most of) the kernel-doc comment on __ksize,
> leaving only the note that it's uninstrumented and only called by
> kasan. Also, we should probably move the declaration of __ksize from
> include/linux/slab.h to mm/slab.h since we don't want to grow new callers.

Seems reasonable. I'll update it in next version!

Thanks.

--
Hyeonggon

2022-02-24 01:03:59

by Vlastimil Babka

[permalink] [raw]
Subject: Re: [PATCH 2/5] mm/sl[auo]b: Do not export __ksize()

On 2/21/22 16:46, Matthew Wilcox wrote:
> On Mon, Feb 21, 2022 at 10:53:33AM +0000, Hyeonggon Yoo wrote:
>> Do not export __ksize(). Only kasan calls __ksize() directly.
>
> We should probably delete (most of) the kernel-doc comment on __ksize,
> leaving only the note that it's uninstrumented and only called by
> kasan. Also, we should probably move the declaration of __ksize from
> include/linux/slab.h to mm/slab.h since we don't want to grow new callers.

Agreed, thanks.