Re: [PATCH mm v3 00/34] kasan: switch tag-based modes to stack ring from per-object metadata

[Yu Zhao]

On Mon, Sep 12, 2022 at 2:06 PM Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> On Mon, 12 Sep 2022 11:39:07 +0200 Marco Elver <elver@xxxxxxxxxx> wrote:
>
> >
> > ...
> >
> > > Hi Andrew,
> > >
> > > Could you consider picking up this series into mm?
> > >
> > > Most of the patches have a Reviewed-by tag from Marco, and I've
> > > addressed the last few comments he had in v3.
> > >
> > > Thanks!
> >
> > I see them in -next, so they've been picked up?
>
> yup.
>
> > FWIW, my concerns have been addressed, so for patches that don't yet
> > have my Reviewed:
> >
> >
> > Acked-by: Marco Elver <elver@xxxxxxxxxx>
>
> Updated, thanks.

Hit the following with the latest mm-unstable. Please take a look. Thanks.

BUG: rwlock bad magic on CPU#0, swapper/0, ffffffdc589d8218
CPU: 0 PID: 0 Comm: swapper Not tainted 6.0.0-rc3-lockdep+ #36
Call trace:
 dump_backtrace+0xfc/0x14c
 show_stack+0x24/0x58
 dump_stack_lvl+0x7c/0xa0
 dump_stack+0x18/0x44
 rwlock_bug+0x88/0x8c
 do_raw_read_unlock+0x7c/0x90
 _raw_read_unlock_irqrestore+0x54/0xa0
 save_stack_info+0x100/0x118
 kasan_save_alloc_info+0x20/0x2c
 __kasan_slab_alloc+0x90/0x94
 early_kmem_cache_node_alloc+0x8c/0x1a8
 __kmem_cache_create+0x1ac/0x338
 create_boot_cache+0xac/0xec
 kmem_cache_init+0x8c/0x174
 mm_init+0x3c/0x78
 start_kernel+0x188/0x49c