"Anthony Williams" <[EMAIL PROTECTED]> wrote in message ae72m2$6vg$[EMAIL PROTECTED]">news:ae72m2$6vg$[EMAIL PROTECTED]... > "Anthony Williams" <[EMAIL PROTECTED]> wrote in message > ae6skn$sc$[EMAIL PROTECTED]">news:ae6skn$sc$[EMAIL PROTECTED]... > > "Bo Berglund" <[EMAIL PROTECTED]> wrote in message > > [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > > > On Tue, 11 Jun 2002 16:08:31 +0000 (UTC), "Anthony Williams" > > > <[EMAIL PROTECTED]> wrote: > > > > > > >I've just updated my cvsnt installation on my w2k box to build 57d, and > > it > > > >seems that the pserver server end is broken --- any attempt to log in > > gives > > > >an access violation in the server process. OTOH, I can happily talk to > a > > > >freshly-built pserver running on Solaris 8, and SSPI works fine on the > > w2k > > > >box. It also works fine as a client for either protocol. > > > > > > > >Is this a known problem? Is there a patch? > > > > > > > >Anthony > > > >-- > > > > > > Did you copy in the fresh protocol dll:s too? > > > > > > Yes. I just did a complete uninstall, and then reinstalled it, and the > > problem persists. > > > Just to add to the description: > > The problem lies in server.c switch_to_user --- the call to putenv for > LOGNAME (env="LOGNAME=EUROPE\ANTHWIL" in my case) causes the access > violation. I haven't worked out the cause or a fix yet.
However, the use of cvs_putenv in the CVS version fixes the problem. Note that in pserver.c, the char array (username_buffer) in get_pserver_username needs to be a static variable (as in my second patch), not an auto variable (as in CVS, and in my first patch). Anthony -- Anthony Williams Software Engineer, Nortel Networks Optical Components Ltd The opinions expressed in this message are not necessarily those of my employer _______________________________________________ Cvsnt mailing list [EMAIL PROTECTED] http://www.cvsnt.org/cgi-bin/mailman/listinfo/cvsnt
