On Mon, 2 Jul 2012, Christian Hägele wrote:

> 1. Currently we are using libCurl 7.23.1. As we tested the latest libcurl we
> found a serious regression Bug. (Detailed description:
> http://curl.haxx.se/bug/view.cgi?id=3538625)

Right. I'll try to assist there but it's really hard for me as I don't have
either end of the problematic case here...

> 2. There were a heap corruption when compiling libcurl with UNICODE defined
> and using unicode characters in username. I attached a patch which should
> resolve these issues. (The patch is against nightly build of 20120627)

I just had a look at it, and something stood out that I'm thinking might not
be 100% correct:

'domlen' is calculated to be the difference between two pointers, but yet it
decodes the domain name from utf8 which could lead the length amount to be
wrong on line 411 "ntlm->identity.Domain[domlen] = TEXT('\0')", can't it?

(I edited white space slightly so possibly it isn't exactly 411 in your
version.)

-- 
  / daniel.haxx.se

-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette: http://curl.haxx.se/mail/etiquette.html
Received on 2012-07-02