Re[2]: [PATCH] fork_init: fix division by zero

[Yuri Tikhonov <yur@emcraft.com>]

=0D=0A Hello Al,

On Wednesday, December 10, 2008 you wrote:

> On Wed, Dec 10, 2008 at 01:01:13PM +0300, Yuri Tikhonov wrote:

>> >> +     max_threads =3D mempages * PAGE_SIZE / (8 * THREAD_SIZE);
>> >                       ^^^^^^^^^^^^^^^^^^^^
>> >> +#endif
>>=20
>> > Can't this overflow, e.g. on 32-bit machines with HIGHMEM?
>>=20
>>  The multiplier here is not PAGE_SIZE, but [PAGE_SIZE / (8 *=20
>> THREAD_SIZE)], and this value is expected to be rather small (2, 4, or=
=20
>> so).

> x * y / z is parsed as (x * y) / z, not x * (y / z).

 Here we believe in preprocessor: since all PAGE_SIZE, 8, and=20
THREAD_SIZE are the constants we expect it will calculate this.

 E.g. here is the result from this line as produced by cross-gcc=20
4.2.2:

        lis     r9,0
        rlwinm  r29,r29,2,16,29
        stw     r29,0(r9)

 As you see - only rotate-left, i.e. multiplication to the constant.

 In any case, adding braces as follows probably would be better:

+     max_threads =3D mempages * (PAGE_SIZE / (8 * THREAD_SIZE));

 Right ?

> Only assignment operators (and ?:, in a sense that a ? b : c ? d : e is
> parsed as a ? b : (c ? d : e)) are right-to-left.  The rest is left-to-ri=
ght.



 Regards, Yuri

 --
 Yuri Tikhonov, Senior Software Engineer
 Emcraft Systems, www.emcraft.com