Re: [PATCH] Fix a race in pid generation that causes pids to be reused immediately.

[Linus Torvalds]

On Wed, Jun 9, 2010 at 10:55 PM, Salman Qazi <sqazi@xxxxxxxxxx> wrote:
> On Wed, Jun 9, 2010 at 10:17 PM, Michel Lespinasse <walken@xxxxxxxxxx> wrote:
>>
>> As shown by the three compares version, the question here does not depend on
>> the max_pid value. So, we can replace max_pid with any integer >= max_pid
>> without changing the result.

I like the logical argument, and cannot fault it.

>> This is nice because, replacing max_pid with
>> 2^32, the expression simplifies to:
>>
>> (unsigned)(a - base) < (unsigned)(b - base)
>>
>> I think I like this form after all :)

Sounds right to me.

> I don't mind doing this, if nobody disagrees.  Should be fine with a
> comment on top explaining the intention?

Yes. And lots of testing (including, very much, the whole max_pid
roll-over, of course). The logic of it all seems impeccable, but
nothing beats testing.

Thanks,
             Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/