On Thu, Oct 09, 2003 at 12:09:18AM +0900, Toru Nishimura wrote:
> Izumi Tsutsui asked about fixing endian issue in if_wmreg.h
> 
> > With a quick glance, structure of wm(4) actually smells Intel-ism ;-p
> > uint8_t/uint16_t values in 32 bit DMA descriptors always cause
> > endian issue (and might cause race conditions on RISC CPUs.)
> >
> > How about the attached patch?
> > 
> > typedef struct wiseman_rxdesc {
> > ... 
> > +#if BYTE_ORDER == LITTLE_ENDIGN
> >     	uint16_t	wrx_len;	/* buffer length */
> >     	uint16_t	wrx_cksum;	/* checksum (starting at PCSS) */
> > +#else
> > +	uint16_t	wrx_cksum;	/* checksum (starting at PCSS) */
> > +	uint16_t	wrx_len;	/* buffer length */
> > +#endif
> 
> It can be written in endian-neutral way by declaring a 32bit quantity of
> unsigned long and do explicit bitmasking and shifting.  The solution
> is cheap and elegant, removing endian puzzles forever.
> 

Generated code will be less efficient thou.

	Andrey