Re: [Bug-gnubg] Fwd: Bug#396445: gnubg: segfault adding match to postgres database

[Russ Allbery]

Christian Anthon <address@hidden> writes:

> Hi Russ,
> could you tell us what the code looks like around the failing line

>> | #5  0x00000000005078f0 in PyGameStats (sc=0xd91380, fIsMatch=0, nMatchTo=7)
>> |     at gnubgmodule.c:1558

> since I don't know what revision of gnubgmodule.c you have.

> In my version it says,

> PyObject* d =
>       Py_BuildValue("{s:f,s:f,s:f,s:f,s:f}",
>                     "luck", sc->arLuck[side][0],
>                     "luck-cost", sc->arLuck[side][1],
>                      "actual-result", sc->arActualResult[side],
>                      "luck-adjusted-result", sc->arLuckAdj[side]);

> and clearly Py_BuildValue expects five argument pairs, but only receives
> four.

Aha!  Yup, same problem here.

-- 
Russ Allbery (address@hidden)             <http://www.eyrie.org/~eagle/>