Re: [PATCH 19/60] microblaze_v4: checksum support

[Michal Simek]

Hi Segher,

you find bug because there is microblaze asm code in generic folder. 
Big thanks. I'll fix it and I'll look at parameters too.

Michal

> +static inline unsigned int
> +csum_tcpudp_nofold(unsigned long saddr, unsigned long daddr, unsigned 
> short len,
> +		unsigned short proto, unsigned int sum)
> +{
> +	__asm__("add %0, %4, %1\n\t"
> +		"addc %0, %4, %2\n\t"
> +		"addc %0, %4, %3\n\t"
> +		"addc %0, %4, r0\n\t"
> +		: "=d" (sum)
> +		: "d" (saddr), "d" (daddr), "d" (len + proto),
> +		"0"(sum));

"sum" should use an earlyclobber, i.e.  "=&d"(sum)  , since some
inputs are used after %0 is first written to.

Also, you can use "+" instead of "=" to say the argument is both
input and output, and get rid of %4, if you like.


Segher