[PyQt] Segfault in SIP 4.19.6

[Scott Talbert]

On Sun, 14 Jan 2018, Phil Thompson wrote:

> On 13 Jan 2018, at 7:22 pm, Scott Talbert <swt at techie.net> wrote:
>>
>> I'm seeing a segmentation fault in SIP 4.19.6 when running sip on wxPython Phoenix.
>>
>> Command line:
>> /usr/bin/sip -w -o -g -I /home/fedora/swt2c/wxPython-4.0.0rc1.dev3597+5f1ee01/src -I /home/fedora/swt2c/wxPython-4.0.0rc1.dev3597+5f1ee01/sip/gen -c /tmp/tmpATp7q5 -b sip/cpp/_core.sbf -X pycode_core:wx/core.py sip/gen/_core.sip
>>
>> Stack trace:
>> #0  prEnumMemberScope (fp=fp at entry=0x55b4139e5ed0, emd=<optimized out>)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/gencode.c:15541
>> #1  0x000055b4120b427c in generateCastZero (fp=fp at entry=0x55b4139e5ed0,
>>    ad=<optimized out>, ad=<optimized out>)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/gencode.c:7588
>> #2  0x000055b4120b4a14 in generateVirtualHandler (
>>    mod=mod at entry=0x55b4139e6180, vhd=vhd at entry=0x55b41578c8e0,
>>    fp=fp at entry=0x55b4139e5ed0)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/gencode.c:8115
>> #3  0x000055b4120c51ee in generateCpp (pt=pt at entry=0x7fff082c53a0,
>>    mod=0x55b4139e6180, codeDir=codeDir at entry=0x7fff082c737a "/tmp/tmpATp7q5",
>>    srcSuffix=srcSuffix at entry=0x55b4120e7e67 ".cpp", parts=parts at entry=0,
>>    needed_qualifiers=needed_qualifiers at entry=0x0, xsl=0x0, py_debug=0)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/gencode.c:1563
>> #4  0x000055b4120c93e1 in generateCode (pt=0x7fff082c53a0,
>>    codeDir=0x7fff082c737a "/tmp/tmpATp7q5",
>>    buildFile=0x7fff082c738c "sip/cpp/_core.sbf", docFile=<optimized out>,
>>    srcSuffix=0x55b4120e7e67 ".cpp", except=<optimized out>, trace=0,
>>    releaseGIL=1, parts=0, needed_qualifiers=0x0, xsl=0x0, consModule=0x0,
>>    docs=1, py_debug=0)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/gencode.c:351
>> #5  0x000055b4120ab40a in main (argc=15, argv=0x7fff082c56b8)
>>    at /usr/src/debug/sip-4.19.6-4.fc28.x86_64/sipgen/main.c:285
>
> I think it's fixed in tonight's snapshot.

Thanks Phil!