On Sat, Sep 09, 2006 at 12:57:13PM +0200, Steinar H. Gunderson wrote: > I'm a bit stomped on the bug, though; the issue seems to be that > qUncompress() returns QByteArray* instead of QByteArray, which g++ refuses to > make a new QByteArray out of. The relevant snippet is this one: > > 11194 QByteArray *sipRes; > 11195 > 11196 sipRes = new QByteArray(qUncompress(a0,a1)); > 11197 > 11198 return > sipConvertFromNewInstance(sipRes,sipClass_QByteArray,NULL); > > (I might be deceived, though -- this file is autogenerated, and I'm looking at > it on an i386 since I don't have a suitable arm available yet... perhaps it > might be differently generated on arm?)
I was indeed right; line 11194 on arm reads "QByteArray sipRes;" (ie. missing an indirection). These files are generated by sip4-qt3, so I tried to download an build it to debug the issue -- but once I build it myself, the problem goes away. No other platforms exhibit this bug, so I guess it's either due to some (now fixed) gcc problem or a cosmic ray while building sip4-qt3. :-) I've asked Steve to schedule a binNMU for sip4-qt3 on arm, which will hopefully fix this issue. (I haven't actually done a full compile with the fixed sip, because that takes _ages_ on arm, but if it doesn't work, we'll take it from there.) /* Steinar */ -- Homepage: http://www.sesse.net/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]