Re: [PATCH] mm: make minimum slab alignment a runtime property

From: Andrey Konovalov <andreyknvl@gmail.com>
To: Hyeonggon Yoo <42.hyeyoo@gmail.com>
Cc: Peter Collingbourne <pcc@google.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Linux ARM <linux-arm-kernel@lists.infradead.org>,
	Linux Memory Management List <linux-mm@kvack.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	Pekka Enberg <penberg@kernel.org>,
	cl@linux.org, roman.gushchin@linux.dev,
	Joonsoo Kim <iamjoonsoo.kim@lge.com>,
	David Rientjes <rientjes@google.com>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Herbert Xu <herbert@gondor.apana.org.au>,
	Andrey Ryabinin <ryabinin.a.a@gmail.com>,
	Alexander Potapenko <glider@google.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	kasan-dev <kasan-dev@googlegroups.com>,
	Eric Biederman <ebiederm@xmission.com>,
	Kees Cook <keescook@chromium.org>
Subject: Re: [PATCH] mm: make minimum slab alignment a runtime property
Date: Fri, 22 Apr 2022 18:09:04 +0200	[thread overview]
Message-ID: <CA+fCnZdTPiH_jeiiHCqdTcUdcJ0qajQ0MvqHWTJ1er7w6ABq5A@mail.gmail.com> (raw)
In-Reply-To: <YmKiDt12Xb/KXX3z@hyeyoo>

On Fri, Apr 22, 2022 at 2:39 PM Hyeonggon Yoo <42.hyeyoo@gmail.com> wrote:
>
> > > kasan_hw_tags_enabled() is also false when kasan is just not initialized yet.
> > > What about writing a new helper something like kasan_is_disabled()
> > > instead?
> >
> > The decision of whether to enable KASAN is made early, before the slab
> > allocator is initialized (start_kernel -> smp_prepare_boot_cpu ->
> > kasan_init_hw_tags vs start_kernel -> mm_init -> kmem_cache_init). If
> > you think about it, this needs to be the case for KASAN to operate
> > correctly because it influences the behavior of the slab allocator via
> > the kasan_*poison* hooks. So I don't think we can end up calling this
> > function before then.
>
> Sounds not bad. I wanted to make sure the value of arch_slab_minaligned()
> is not changed during its execution.
>
> Just some part of me thought something like this would be more
> intuitive/robust.
>
> if (systems_supports_mte() && kasan_arg != KASAN_ARG_OFF)
>         return MTE_GRANULE_SIZE;
> else
>         return __alignof__(unsigned long long);

Hi Hyeonggon,

We could add and use kasan_hw_rags_requested(), which would return
(systems_supports_mte() && kasan_arg != KASAN_ARG_OFF).

However, I'm not sure we will get a fully static behavior:
systems_supports_mte() also only starts returning proper result at
some point during CPU bring-up if I'm not mistaken.

Thanks!