Since upgrading to build v3.3.1.1076 on Windows, the x86 build of my application always fails to start with “Initialization error 4”.
All options are the same and this used to work with prior versions of VMProtect.
The x64 build works fine.
Any ideas?
I updated my subscription in order to upgrade to the latest build 1097, and the issue persists.
With build 1097, the “Initialization error 4” is gone, but now the x86 exe simply refuses to start - it looks like it starts but then immediately closes itself again.
And again, the x64 build works fine.
I can make my x86 application available.
Disregard this.
This was caused by a memory overwrite of 4 bytes in the application’s startup code. I guess this corrupted something in a way that made VMProtect “unhappy”.
Did you find an application that does it?
There is definitely something wrong with VMProtect and x86 builds.
I built a small test case which is literally the following C++ code:
#include “VMProtectSDK.h”
int DeactivateLicense(void)
{
VMProtectBeginUltra(“DeactivateLicense”);
VMProtectEnd();
return 0;
}
int main(void)
{
const int result = DeactivateLicense();
printf(“exit code: %d\n”, result);
return result;
}
The resulting executable runs fine on its own.
When using VMProtect and running the resulting executable, VMProtect will detect a file modification and close the app.
Where can I submit the application in question so the developers can take a look at it?
When using VMProtect and running the resulting executable, VMProtect will detect a file modification and close the app.
… because the file was modified in the memory? It seems something changes your protected file at the runtime and VMProtect detects it and shows the error.
The file is not modified in memory. The code I posted above is the exact C++ code - there is nothing else in there.
The x64 application runs fine (and VMProtect doesn’t detect any modification), but x86 fails.
Can I send you the compiled x86 executable along with my VMProtect options? It’s only around 100KB.
OK. Please send all files (the original binary + MAP + VMP files too) to info@vmpsoft.com
Done, thank you for your help!
Can you confirm yet if this is indeed a bug in VMProtect along with a possible ETA for a fix?
I have quite a few customers who need to use the x86 version of my product.
Please try the 1099 build
I can confirm that the issue is fixed in the 1099 build.