Fix is now in CVS. Op vr 03-10-2003, om 19:44 schreef Jeroen Asselman: > Hi, > > 'good' news. I have been able to reproduce the bug with those extra > assertions. I am now checking when this exactly happens to find the > cleanest solution. > > Op vr 03-10-2003, om 12:52 schreef [EMAIL PROTECTED]: > > > Ah -- I suspect that 99.99% of all users are using gcc for their > > > compiler. In that case, I now understand why I'm having problems, > > > and they are not. My compiler can't even be told to all this with > > > any switches - accesses to address zero are forbidden by the OS, > > > so it wouldn't do any good to give the compiler a switch to allow > > > it. > > > > > > Once you add the assertion code to test for u->name == null, I think > > > you will see it happening on your test machine too. > > > > Well as I said in another mail yesterday evening. I added those assertions > > in parq.c but they didn't trigger a problem for me. I haven't heard it > > triggered the problem for anyone else. So the problem must be lying > > somewhere else. > > > > Perhaps you can join the gtk-gnutella irc channel at freenode. I am around > > there in the evening most of the time. Europe time zone. > > > > I also would still like to see the changes you made. Perhaps you are > > missing something obvious? -- Jeroen Asselman <[EMAIL PROTECTED]>
signature.asc
Description: Dit berichtdeel is digitaal ondertekend
