Subject: Re: [boost] [lockfree] Review
From: Tim Blechmann (tim_at_[hidden])
Date: 2011-07-31 09:21:06
> Anyhow, I think this should be straightforward. I didn't look at it in
> detail, but it looked like you only need a specialization with size 8
> types that does a _InterlockedCompareExchange64 for most everything.
> Efficient loads & stores are a bit tricky in that SSE2 is not a
> requirement for 32-bit Windows. Without it, I think we need to resort
> FILD/FISTP, which is a pain.
iirc, sse2 intrinsics are not guaranteed to be atomic, so sometimes memory
access has to be emulated via CAS.
> Well, I hope that at least atomic<> gets it right. Nonaligned accesses are
> not guaranteed to be atomic. In this case I guess, the atomic<>
> implementation ends up using a uint64_t. In VC++ __alignof == 8, but that
> doesn't mean that autos of type uint64_t are necessarily aligned. In our
> case I would assume that you get memory from the OS to be at least 16 or 8
> byte aligned and the alignment of class types should be fine.
this is the reason, why atomic<>::is_lock_free() is not a static member
> E.g.: consider this on x86 with an older GCC
> void foo()
> atomic<tagged_node_ptr<X> > x;
> Is x aligned, here? I don't recall the ABI, but I believe it doesn't
> guarantee anything beyond 4-byte alignment for ESP on entry. So to align x
> properly in the stack frame, the stack must be dynamically aligned (or
> some interprocedural optimization may help) -- but I don't think older
> GCCs do that.
dynamic memory allocation makes it even worse, because you can use placement new
to put the data structure to virtually any memory location :/
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk