On Sat, Oct 22, 2011 at 05:05:05AM -0400, David Miller wrote:
> From: Jurij Smakov <jurij@xxxxxxxxx>
> Date: Sat, 22 Oct 2011 10:00:24 +0100
>
> > Is something conceptually wrong with just doing this:
> >
> > asm volatile ("save; flush; restore");
> >
> > It fixes the problem for freebsd (will test on linux later today), and
> > does not rely on compiler doing (or not doing) the "right"
> > optimization.
>
> I would not do a "naked" save here, please just use the construct
> I showed you with the %o7 register clobber.
Yes, I missed that save is not allocating any memory, which is
wrong. What I actually had in mind was this:
asm volatile ("save %sp, -96, %sp; flushw; restore");
My concern with the approach you are proposing is that compiler
may (in the future, perhaps) just optimize out the helper function,
making arrangements to save %o7 in one of the unused registers, thus
removing allocation of another register window and breaking it again.
Best regards,
--
Jurij Smakov jurij@xxxxxxxxx
Key: http://www.wooyd.org/pgpkey/ KeyID: C99E03CC
--
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
