Re: [GIT PULL] asm-generic fixes

[Linus Torvalds <torvalds@linux-foundation.org>]

On Tue, 23 Jun 2009, Arnd Bergmann wrote:
> @@ -71,7 +75,7 @@ static unsigned int do_csum(const unsigned char *buff, int len)
>  		if (count) {
>  			unsigned long carry = 0;
>  			do {
> -				unsigned long w = *(unsigned long *) buff;
> +				unsigned long w = *(unsigned int *) buff;
>  				count--;
>  				buff += 4;
>  				result += carry;

I don't think this is sufficient.

You might need to make 'result', 'carry', and 'w' be 'unsigned int' too.

Why? Because the final folding is only doen from 32-bit to 16 bit, we 
don't do the whole 64-bit to 32-bit to 16-bit chain.

Now, it's possible (even likely) that even with a 64-bit word, we'll never 
actually do large enough areas that 'result' would ever have very many 
bits set in the 32+ bit region, and since we do end up folding to 16 bits 
twice (once after the loop and once at the end), it _probably_ gets things 
right in most cases. But I doubt "probably" is strong enough. Somebody 
should check.

Or just see arch/alpha/lib/checksum.c, which does the whole 64-bit case.
Maybe lib/checksum.c should be lib/checksum_{32,64}.c.

		Linus