Re[4]: [Gc] atomic_ops: bugfix for Solaris x86/amd64

Ivan Maidanski ivmai at mail.ru
Wed Feb 3 12:29:23 PST 2010


Hi!
Ian Wienand <ian at wienand.org> wrote:
> 2010/2/2 Ivan Maidanski <ivmai at mail.ru>:
> > No, my patch deals nothing with sparc.
> 
> Yeah, I read it a bit closer after I sent that email :)  Sorry
> 
> > I see the problem of [1]: If you compile for old sparc (I mean not v9) then use -D AO_NO_SPARC_V9.
> > If you can't replicate that problem then you are compiling on sparcv9, right? (OS means nothing here, only the compiler and CPU matter.)
> 
> Yeah, I guess I must be.  I admit I'm a little fuzzy on  all the
> different types of SPARC

I'm not an expert in the area neither...

> 
> This kind of makes more sense with [2] now.
> 
> So I should probably add -D AO_NO_SPARC_V9 to the Debian SPARC build,
> because they're not 64-bit, right?

AFAIK, v9 is always 64-bit (so it's ok to -D AO_NO_SPARC_V9 if you know the target is 32-bit).
On the other hand, such things should be auto-detected (either by configure or using the predefined macros). (The second variant is preferred.)

I know there are 2 predefined macros (__sparcv9 and __sparc) but I'm not sure AO_NO_SPARC_V9 could be replaced with any if them. Does somebody on the ML have a strong opinion about it?

If you are able to solve the problem thru configurea.ac, you can try to contribute.

> 
> [2] http://lists.debian.org/debian-sparc/2010/01/msg00002.html

Bye.


More information about the Gc mailing list