Re: [PATCH v3 2/2] slab: __GFP_ZERO is incompatible with a constructor

[Guenter Roeck <linux@roeck-us.net>]

On 08/04/2018 02:28 AM, Geert Uytterhoeven wrote:
> On Sat, Aug 4, 2018 at 12:34 AM Matthew Wilcox <willy@infradead.org> wrote:
>> On Fri, Aug 03, 2018 at 02:22:57PM -0700, Guenter Roeck wrote:
>>> On Thu, Apr 12, 2018 at 12:13:22PM -0700, Matthew Wilcox wrote:
>>>> From: Matthew Wilcox <mawilcox@microsoft.com>
>>>> __GFP_ZERO requests that the object be initialised to all-zeroes,
>>>> while the purpose of a constructor is to initialise an object to a
>>>> particular pattern.  We cannot do both.  Add a warning to catch any
>>>> users who mistakenly pass a __GFP_ZERO flag when allocating a slab with
>>>> a constructor.
>>>>
>>>> Fixes: d07dbea46405 ("Slab allocators: support __GFP_ZERO in all allocators")
>>>> Signed-off-by: Matthew Wilcox <mawilcox@microsoft.com>
>>>> Acked-by: Johannes Weiner <hannes@cmpxchg.org>
>>>> Acked-by: Vlastimil Babka <vbabka@suse.cz>
>>>> Acked-by: Michal Hocko <mhocko@suse.com>
>>>
>>> Seen with v4.18-rc7-139-gef46808 and v4.18-rc7-178-g0b5b1f9a78b5 when
>>> booting sh4 images in qemu:
>>
>> Thanks!  It's under discussion here:
>>
>> https://marc.info/?t\x153301426900002&r=1&w=2
> 
> and https://www.spinics.net/lists/linux-sh/msg53298.html
> 
>> also reported here with a bogus backtrace:
>>
>> https://marc.info/?l=linux-sh&m\x153305755505935&w=2
>>
>> Short version: It's a bug that's been present since 2009 and nobody
>> noticed until now.  And nobody's quite sure what the effect of this
>> bug is.

Though now it is making a lot of noise :-).

I just found two more 0-day bugs, so maybe improved testing and log messages
such as the one encountered here do help a bit.

Guenter