Closed renatosilva closed 9 years ago
That's odd. Your code looks OK. And matches the unit test for that entrypoint.
It's also weird that it's crashing inside libssp-0.dll
. This is for some sort of stack protection.
How are you building this code? What's the gcc/g++ version?
Interesting, log file is actually set successfully and reports this libssp crash just like GDB above. Debug output from the application also mentions stack smashing detected. I build both the application and Dr. MinGW with MSYS2/MinGW-w64. Here are my compiler versions:
gcc.exe (Rev5, Built by MSYS2 project) 4.9.2
g++.exe (Rev5, Built by MSYS2 project) 4.9.2
Yes, I suspected that much -- something is causing a buffer overflow, but I don't thing it is due to the APIENTRY calling convention.
You built Dr.MinGW? So you are not using my binaries? (I hope you're not building exchdnl.c with -D_UNICODE ....)
In that case, does mingw32-make check
work?
I have found the solution, there was another function pointer definition in the code that also needed to be updated. I confirm that SetLogFileNameA
works for me.
I can't make the new
SetLogFileNameA
function work. I'm trying to call it like this:Starting with revision 318048b I get: