Re: [RFC PATCH 2/3] mm/slub: unify all sl[au]b parameters with "slab_$param"

[Vlastimil Babka]

On 12/2/23 05:23, Song, Xiongwei wrote:
> 
> 
>> -----Original Message-----
>> From: Kees Cook <keescook@xxxxxxxxxxxx>
>> Sent: Saturday, December 2, 2023 1:48 AM
>> To: sxwjean@xxxxxx
>> Cc: vbabka@xxxxxxx; 42.hyeyoo@xxxxxxxxx; cl@xxxxxxxxx; linux-mm@xxxxxxxxx;
>> penberg@xxxxxxxxxx; rientjes@xxxxxxxxxx; iamjoonsoo.kim@xxxxxxx;
>> roman.gushchin@xxxxxxxxx; corbet@xxxxxxx; arnd@xxxxxxxx; akpm@linux-
>> foundation.org; gregkh@xxxxxxxxxxxxxxxxxxx; linux-doc@xxxxxxxxxxxxxxx; linux-
>> kernel@xxxxxxxxxxxxxxx; Song, Xiongwei <Xiongwei.Song@xxxxxxxxxxxxx>
>> Subject: Re: [RFC PATCH 2/3] mm/slub: unify all sl[au]b parameters with "slab_$param"
>> 
>> CAUTION: This email comes from a non Wind River email account!
>> Do not click links or open attachments unless you recognize the sender and know the
>> content is safe.
>> 
>> On Fri, Dec 01, 2023 at 11:15:04AM +0800, sxwjean@xxxxxx wrote:
>> > From: Xiongwei Song <xiongwei.song@xxxxxxxxxxxxx>
>> >
>> > Since the SLAB allocator has been removed, so we need to clean up the
>> > sl[au]b_$params. However, the "slab/SLAB" terms should be keep for
>> > long-term rather than "slub/SLUB". Hence, we should use "slab_$param"
>> > as the primary prefix, which is pointed out by Vlastimil Babka. For more
>> > information please see [1].
>> >
>> > This patch is changing the following slab parameters
>> > - slub_max_order
>> > - slub_min_order
>> > - slub_min_objects
>> > - slub_debug
>> > to
>> > - slab_max_order
>> > - slab_min_order
>> > - slab_min_objects
>> > - slab_debug
>> > as the primary slab parameters in
>> > Documentation/admin-guide/kernel-parameters.txt and source, and rename all
>> > setup functions of them too. Meanwhile, "slub_$params" can also be passed
>> > by command line, which is to keep backward compatibility. Also mark all
>> > "slub_$params" as legacy.
>> 
>> *If* we do this, I think the old names need to be recognized, perhaps

Yes, they are already recognized by the patch.

>> with a boot pr_warn() about their deprecation/renaming for several
>> releases (likely across LTSes). I think it's not a good idea to
>> wholesale rename these with no warning. That's going to cause a lot of
>> surprises and broken userspace...
> 
> Oh, yes, that's a good idea. Will update.

I'd wait for a while with the warnings, no need to rush.

> Regards,
> Xiongwei
>> 
>> --
>> Kees Cook