Re: forward list patch

[François Dumont <francois dot cppdevs at free dot fr>]

On 10/24/2010 12:12 PM, Paolo Carlini wrote:
> Hi,
>    
> Do you have n3126 at hand? Please fetch it from the ISO web site, if you
> didn't already.
I have it now.
>   I see that 23.2.1/11 explicitly says that swap can
> invalidate end(), I wonder if it should explicitly say something similar
> about before_begin?!? After all, aren't the two otherwise very similar?
> What do you think?
Yes, end and before begin looks similar. However, with current libstdc++ 
implementation there is a major difference. end has no relation with the 
forward list but before begin is tighly link to it with no way to 
associate it with any other forward list instance.
>   Because, really, that looping, etc, part of swap
> seems just crazy to me: it's a very, very basic design idea in the
> standard (even more now in the move semantics era) that swap is a very
> fast constant time operation, is just about exchanging a few pointers.
> If we cannot manage to implement it like that, either the standard
> should be slightly tweaked somewhere, or... we should try harder ;)
>    
I agree that it is not something you would expect in a swap 
implementation. At least not in the normal form and the whole Standard 
is written in such a way that swap is easy to write in a performant way. 
However the safe mode is not Standard and it is not surprising that as 
long as you keep links between iterators and sequences then some 
operations become more difficult to implement.

So here is an other proposition. This time I made the forward list 
_M_swap method much closer to the basic version. If you check the basic 
version in debug.cc you will see that there are already 4 for loops to 
update the sequence reference for all safe iterators. In the forward 
list _M_swap version I have the same 4 for loops except that at the same 
time I look for before begin iterators and move them back to the list 
they should belong to.

At the same time your remark about end and before begin similarities 
made me realized that I also miss to change _M_invalidate_all so that it 
do not invalidate before begin iterator.

François

Attachment: ChangeLog.entry
Description: Text document

Attachment: forward_list.patch
Description: Text document