On 2014-09-16 13:49:20 +0300, Heikki Linnakangas wrote:
> I used http://create.stephan-brumme.com/crc32/#slicing-by-8-overview as
> reference - you can probably see the similarity. Any implementation is going
> to look more or less the same, though; there aren't that many ways to write
> the implementation.
True.
I think I see what's the problem causing Amit's test to fail. Amit, did
you use the powerpc machine?
Heikki, you swap bytes unconditionally - afaics that's wrong on big
endian systems. My patch had:
+ static inline uint32 swab32(const uint32 x);
+ static inline uint32 swab32(const uint32 x){
+ return ((x & (uint32)0x000000ffUL) << 24) |
+ ((x & (uint32)0x0000ff00UL) << 8) |
+ ((x & (uint32)0x00ff0000UL) >> 8) |
+ ((x & (uint32)0xff000000UL) >> 24);
+ }
+
+ #if defined __BIG_ENDIAN__
+ #define cpu_to_be32(x)
+ #else
+ #define cpu_to_be32(x) swab32(x)
+ #endif
I guess yours needs something similar. I personally like the cpu_to_be*
naming - it imo makes it pretty clear what happens.
Greetings,
Andres Freund
-- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services