Re: [Linux-ia64] rwlock and atomic_sub on ia64

[Noah Romer <nromer@lsil.com>]

Jes Sorensen wrote:
>>>>>>"Noah" = Noah Romer <nromer@lsil.com> writes:
>>>>>>
> 
> Noah> This is a kernel driver (linux/drivers/message/fusion). The
> Noah> version in 2.4.7 isn't up to date (and doesn't contain the code
> Noah> that uses rwlocks), but it's there, finally.
> 
> Ok, I saw Dave Miller checked it into his CVS tree yesterday or so.
> 
> Noah> Yep, but, for some reason, I've never gotten any complaints from
> Noah> the compiler until I tried compiling on ia64. It's not a big
> Noah> issue (especially since the subtracted value in this instance
> Noah> should never be big enough that signed/unsigned will make a
> Noah> diff), as I can just cast it to int. It just struck me as
> Noah> strange that the compiler would now decide to issue a
> Noah> warning. The rwlock bit is what's got me hung up.
> 
> The reason for this is most likely the compiler versions. For
> Linux/ia64 you have to use at least gcc-2.96 to compile the
> kernel. This compiler is a lot more picky about data types than the
> older versions such as egcs-1.1.2/gcc-2.95 etc.

Ok, I wondered if it had anything to do w/ the newer compiler that came 
w/ Redhat's ia64 beta. Most of my i386 compiles have been done w/, um, 
egcs 2.91.66. I guess it's time to update those systems.

> I'll leave the question about rwsem's to David ;)
> 
> Cheers
> Jes

Thanks.
-- 
Noah Romer
Driver Developer, CM gopher and Linux Whipping Boy
Storage Components Firmware
LSI Logic Corp.