U++ framework: Coffee corner

Home » Community » Coffee corner » Clang vs. GCC

Show: Today's Messages :: Polls :: Message Navigator
E-mail to friend

Re: Clang vs. GCC [message #25010 is a reply to message #25007]

Sun, 07 February 2010 11:49

gprentice
Messages: 260
Registered: November 2005
Location: New Zealand

Experienced Member

Yep it's a clang bug and a dark corner of C++.

As mentioned in that article, the standard defines a pseudo destructor (5.2.4 ) one form of which looks like this
::opt nested-name-specifier opt type-name :: ~ type-name

where the first :: and the nested-name-specifier are optional and type-name is a non-class type. The only effect is the evaluation of the post-fix expression before the arrow.

There's no such thing as a constructor for a fundamental type but the standard defines (5.2.3 / 2) that the expression T() for simple type specifier T creates an rvalue of the specified type whose value is determined by default initialization.

Graeme

Report message to a moderator

[Message index]

		Clang vs. GCC By: dolik.rce on Sat, 06 February 2010 23:10
		Re: Clang vs. GCC By: Novo on Sun, 07 February 2010 02:01
		Re: Clang vs. GCC By: mirek on Sun, 07 February 2010 09:55
		Re: Clang vs. GCC By: mirek on Sun, 07 February 2010 10:13
		Re: Clang vs. GCC By: gprentice on Sun, 07 February 2010 11:49
		Re: Clang vs. GCC By: Didier on Sun, 07 February 2010 12:11
		Re: Clang vs. GCC By: mirek on Sun, 07 February 2010 14:35
		Re: Clang vs. GCC By: Didier on Sun, 07 February 2010 14:47
		Re: Clang vs. GCC By: dolik.rce on Sun, 07 February 2010 15:35
		Re: Clang vs. GCC By: Sgifan on Wed, 24 February 2010 08:53

Previous Topic:	U++ for newbees
Next Topic:	Note about how classic OOP with C++ fails efficiency

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Wed Jun 24 15:27:56 GMT+2 2026

Total time taken to generate the page: 0.00662 seconds