64bit fix in float to int and int to float bit conversion.
Use `int` instead of `INT` for `INT` may be 64bit int. Better would
be to use `int32_t`, indeed, but this one is not supported by BCC.
(we will drop BCC55 support soon, though)
@REM -------
@REM make using mingw gnu compiler
@REM type mingwmake, and wait...
@REM do not edit - automatically generated from ProjectDefinition
@REM -------
@SET DEFINES=
@REM Kludge got Mercurial, cannot be implemented in Borland make
@FOR /F "tokens=*" %%i in ('hg root') do SET HGROOT=%%i
@IF "%HGROOT%" NEQ "" SET DEFINES=%DEFINES% "-DHGROOT=%HGROOT%"
@pushd ..\rules
@call find_mingw.bat
@popd
make.exe -N -f bc.mak %DEFINES% %USEMINGW_ARG% %*
@echo "***********************************"
@echo "Buildung stx/libjava/experiments
@echo "***********************************"
@pushd experiments
@call mingwmake %1 %2 || exit /b "%errorlevel%"
@popd
@echo "***********************************"
@echo "Buildung stx/libjava/tools
@echo "***********************************"
@pushd tools
@call mingwmake %1 %2 || exit /b "%errorlevel%"
@popd