[PATCH] [BZ 18034] [AArch64] Lazy TLSDESC relocation data race fix

[Torvald Riegel]

On Tue, 2015-06-09 at 00:10 -0300, Alexandre Oliva wrote:
> On May 26, 2015, Torvald Riegel <triegel@redhat.com> wrote:
> 
> > As a result, you can drop the volatile qualification I
> > believe (I haven't checked, but it seems this was a pre-memory-model way
> > to avoid compiler reordering

I meant that I haven't checked what the intent behind it was, and why it
was chosen instead of other alternatives.

> That's correct.  It was the only portable way available back then to
> force stores and loads to take place in specific orders.

An alternative would have been to do what Boehm's libatomic-ops did:
Have functions/macros for all atomic accesses, and implement them with
asm volatile.