[prev in list] [next in list] [prev in thread] [next in thread] 

List:       pykde
Subject:    Re: [PyKDE] sip crash on configure.py standard MacPython 2.3.5
From:       Phil Thompson <phil () riverbankcomputing ! co ! uk>
Date:       2006-11-18 13:26:37
Message-ID: 200611181326.37242.phil () riverbankcomputing ! co ! uk
[Download RAW message or body]

On Saturday 18 November 2006 10:17 am, Ulrich Berning wrote:
> Phil Thompson schrieb:
> >On Friday 17 November 2006 9:09 pm, Patrick Stinson wrote:
> >>I thought I posted this once before, but I couldn't find it from
> >>searching, so...
> >>
> >>The latest sip snapshot crashed for me during configure.py from the
> >>latest PyQt4 snapshot on Tiger using the default python
> >>implementation. The C++ code was generated properly.
> >>
> >>Exception:  EXC_BAD_ACCESS (0x0001)
> >>Codes:      KERN_INVALID_ADDRESS (0x0001) at 0x53617465
> >>
> >>Thread 0 Crashed:
> >>0   libSystem.B.dylib 	0x9000c5f8 __vfprintf + 6311
> >>1   libSystem.B.dylib 	0x900e4530 vfprintf_l + 94
> >>2   libSystem.B.dylib 	0x900107b9 fprintf + 73
> >>3   sip               	0x000178bd apiArgument + 619
> >>4   sip               	0x000179b4 apiCtor + 135
> >>5   sip               	0x00017d75 generateAPI + 269
> >>6   sip               	0x000035e9 main + 2090
> >>7   sip               	0x000025de _start + 216
> >>8   sip               	0x00002505 start + 41
> >
> >You did post it before - but I couldn't (and still can't) reproduce the
> >problem.
> >
> >Can you do a bit of debugging? At first I would need to know which ctor
> > and which of the ctor's arguments was causing the problem.
> >
> >Phil
>
> For me, it crashes only on HP-UX and on Solaris. It crashes when
> generating the API files.
>
> I have searched for it and just found the bug. In sipgen/transform.c in
> addDefaultCopyCtor() starting at line 1043 a structure for a default
> public copy ctor is allocated and initialized, but
> copyct->pysig.args[0].name gets never initialized. This causes the crash
> later in sipgen/export.c in apiArgument() at line 298 (The line numbers
> are based on snapshot-20061117).

I should have found that myself - yet another beer I owe you.

Will be fixed in tonight's snapshot.

Phil

_______________________________________________
PyKDE mailing list    PyKDE@mats.imk.fraunhofer.de
http://mats.imk.fraunhofer.de/mailman/listinfo/pykde
[prev in list] [next in list] [prev in thread] [next in thread]