LWN: Comments on "A look at C++14: Papers Part 2"

Swap

mathstuf — 2013-04-05T23:18:35+00:00

Can't you just treat them as integers of the same size for CAS?

A look at C++14: Papers Part 2

robert_s — 2013-04-05T17:44:48+00:00

Head already spinning...

...I mean whoever's job it is to make the final decisions.

Swap

heijo — 2013-04-03T11:04:50+00:00

Haswell CPUs with TSX.

A look at C++14: Papers Part 2

jwakely — 2013-04-03T08:33:09+00:00

Right. I don't remember ever seeing a formal proposal to add them, and there is strong opposition to them from some parts of the committee.

Swap

Ben_P — 2013-04-03T05:05:45+00:00

Atomic swaps on doubles (and floats) are made more tricky by the fact very often NaN != NaN.

A look at C++14: Papers Part 2

daniel — 2013-04-02T23:56:21+00:00

Designated initializers? Conspicuously absent.

A look at C++14: Papers Part 2

jwakely — 2013-04-02T23:13:27+00:00

If you mean www.open-std.org, it's noone's job and there is no open standards group, there's one volunteer who maintains the websites for several standards bodies.

There's a table at http://isocpp.org/blog/2013/03/pre-bristol-standards-pape... which categorizes the papers, the ones marked "Evolution" are probably the "interesting language" ones.

A look at C++14: Papers Part 2

robert_s — 2013-04-02T22:57:34+00:00

That's kinda the job of the open standards group.

Swap

wahern — 2013-04-02T20:31:38+00:00

Are there any architectures where there exists an atomic, wait-free swap between two memory locations?

I suspect that they're uncommon if not non-existent.

In any event, presumably such an operation was left out of the stdatomic.h libraries for C11 and C++11 for a reason.

Swap

scottwood — 2013-04-02T20:12:27+00:00

If non-atomic swap is a useful addition, why would you want to slow down swaps that don't need to be atomic by making it automatic, even if it's optional for the implementation? Atomic swap should be a specially requested operation (e.g. using compiler intrinsics), just like most other atomic operations.

Swap

tialaramex — 2013-04-02T09:17:24+00:00

I can imagine that there might be platforms on which (especially for larger or more sophisticated native C types like doubles) atomic swap is not practical.

So perhaps C could consider offering swap with no promises of atomicity, plus a set of reserved pre-processor variables which tell us whether atomicity is implemented for specific types. This would let you do

#if !defined(SWAP_UINT64_T_IS_ATOMIC)
#error This program requires an atomic swap for unsigned 64-bit integers but you don't have one. Too bad.
#endif

... or something more useful, but you get the idea.

Swap

proski — 2013-04-01T21:18:36+00:00

The swap operator looks interesting. I hope it would come to C as well. Perhaps it could be made atomic at least for some types.

A look at C++14: Papers Part 2

heijo — 2013-04-01T18:37:45+00:00

How about dividing the mass of mere library proposals from the actually interesting language proposals?