Re: [v3] libstdc++/44413 (for ext/vstring)

[Paolo Carlini <paolo dot carlini at oracle dot com>]

On 06/09/2010 04:32 PM, Doug Semler wrote:
> I was not the original poster/requester/suggester/whatever and just
> jumped in the middle of the conversation :)
>
> Sorry if I gave the wrong impression --- the comment "I don't see it"
> was directed at the optimization attempt itself...I don't have any
> ideas myself ;-)
>   
I see ;-)
> That being said, my interest in the compiler is on the 32 bit *and* 64
> bit windows side (iow x86 family), and seeing something that hurts one
> side with only minor benefits on the other side makes me want to at
> least say something.
>   
You did the right thing, I overlooked the 32-bit case, in my desperate
attempt to improve - very marginally - the 64-bit case.

To be honest, having considered the issue a bit more, I have no idea how
to substantively improve the 64-bit case, considering that the length of
a string can be much larger than 2^32, and we have to return an int in
that case too. If you could get back to the original
poster/requester/suggester/whatever, and report back to us when you have
additional information, it would be great.

Thanks,
Paolo.