Re: gcc 4.5 and TLS
On Sat, Jun 05, 2010 at 12:21:47PM +1000, Finn Thain wrote:
> On Sat, 5 Jun 2010, Richard Zidlicky wrote:
> > I am not sure what happens if the futex is inside a "misaligned struct"
> > - would that be handled with an attribute of the futex?
> Is this problem confined to m68k? (All architectures seem to be subject to
> the same rule for futex offsets.)
it was unique to m68k last I looked. Would be interesting to look at it again
- I think I did read that another of the new arches had this problem but can not
recall which one.
Traditionally even if there were other arches we got most of the impact because
none of those tried to run xemacs, apl, smalltalk or any of the other packages
that break on the naive alignment assumption. Otoh so far everything was fixable
(sometimes with lots of ugly debugging) without api breakage.