Re: [patch] ips.c spin lock 64 bit issues

Jes Sorensen (jes@trained-monkey.org)
Thu, 16 Aug 2001 08:55:22 -0400


>>>>> "Richard" == Richard B Johnson <root@chaos.analogic.com> writes:

Richard> I don't think that "unsigned long" is the correct data type.

Richard> Isn't there a data type that means "the largest unsigned
Richard> integer type that fits into a register on the target..."? I was
Richard> told that that's what "size_t" means. If so, all the flags
Richard> variables <everywhere> should be changed to this type. If not,
Richard> then somebody should define a "flags_t" type. With the new
Richard> 64-bit machines, this is going to bite over and over again
Richard> until something like this is done.

Face it, unsigned long *is* the correct data type. The API has been
specified like this for years, no reason to change it. long us
guaranteed to be able to hold a pointer on Linux, ie. the largest
natural data type.

And no, you are more likely to find a 32 bit size_t on some systems and
we do not need yet another obscure data type flags_t when unsigned long
does the job fine as it is.

Jes
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/