Re: [PATCH 2/4] x86: convert threshold_bank.cpus from atomic_t to refcount_t

[Kees Cook]

On Mon, Feb 20, 2017 at 3:17 AM, Borislav Petkov <bp@xxxxxxxxx> wrote:
> On Mon, Feb 20, 2017 at 01:06:19PM +0200, Elena Reshetova wrote:
>> refcount_t type and corresponding API should be
>> used instead of atomic_t when the variable is used as
>> a reference counter. This allows to avoid accidental
>> refcounter overflows that might lead to use-after-free
>> situations.
>>
>> Signed-off-by: Elena Reshetova <elena.reshetova@xxxxxxxxx>
>> Signed-off-by: Hans Liljestrand <ishkamiel@xxxxxxxxx>
>> Signed-off-by: Kees Cook <keescook@xxxxxxxxxxxx>
>> Signed-off-by: David Windsor <dwindsor@xxxxxxxxx>
>
> That SOB chain tells me that you wrote the patch and Hans, Kees and
> David handled it in some way and the last one - David - is sending it to
> me. It doesn't look like that though.

Perhaps the least inaccurate form of this might be:


Inspired by atomic protections in PaX/grsecurity.

Suggested-by: Kees Cook <keescook@xxxxxxxxxxxx>
Reviewed-by: David Windsor <dwindsor@xxxxxxxxx>
Reviewed-by: Hans Liljestrand <ishkamiel@xxxxxxxxx>
Signed-off-by: Elena Reshetova <elena.reshetova@xxxxxxxxx>


As this is something I'd suggested we implement based on the work in
PaX/grsecurity, David took the first (and continuing) stab at
conversions, Hans did more, and Elena has been doing even more along
with the heavy-lifting of keeping the series organized. That way the
first SoB is still the author, the last SoB is still the email sender,
and everyone's name is mentioned.

Or just:


Inspired by atomic protections in PaX/grsecurity, based on work from
David Windsor, Hans Liljestrand, and myself.

Signed-off-by: Elena Reshetova <elena.reshetova@xxxxxxxxx>


I'm not picky -- I just want to see the conversion to refcount_t
happen, and everyone in Elena's SoB list has done work on it...

-Kees

-- 
Kees Cook
Pixel Security