Andreas Metzler wrote:
> On Wed, Oct 22, 2003 at 04:30:55PM +0200, Tom Kistner wrote:
>
>>I was asking Philip if he handles endianess in
>>any way, and he doesn't, so there are no collisions at compile time.
>
> Actually there is a little bit of endian-dependent code in exim, using
> this as test
>
> int i = 0x01020304;
> uschar *ctest = (uschar *)(&i);
> printf("Checking md5: %s-endian\n", (ctest[0] == 0x04)? "little" : "big");
>
> so having this globally available might make sense.
> cu andreas
No exim on a PDP-11? sniff...
uint32_t i;
uint8_t *ctest = (uint8_t *)&i;
ctest[0] = 0x04;
ctest[1] = 0x03;
ctest[2] = 0x02;
ctest[3] = 0x01;
i == 0x01020304 => little endian
i == 0x04030201 => big endian
i == 0x03040102 => pdp endian
;-)
Oliver
