change error checks in .bat files throughout project

Author: ndgrigorian

conda-recipe/bld.bat

@@ -25,9 +25,9 @@ rem fix the issue with IntelLLVM integration with cmake on Windows
 if EXIST "%PLATFORM_DIR%" (
   dir "%PLATFORM_DIR%\%FN%"
   copy /Y "%PLATFORM_DIR%\%FN%" .
-  if errorlevel 1 exit 1
+  if %ERRORLEVEL% neq 0 exit 1
   copy /Y ".github\workflows\Windows-IntelLLVM_%PATCHED_CMAKE_VERSION%.cmake" "%PLATFORM_DIR%\%FN%"
-  if errorlevel 1 exit 1
+  if %ERRORLEVEL% neq 0 exit 1
 )
 
 set "CC=icx"

conda-recipe/run_test.bat

@@ -1,10 +1,10 @@
 @echo on
 
 "%PYTHON%" -c "import dpctl; print(dpctl.__version__)"
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1
 
 "%PYTHON%" -m dpctl -f
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1
 
 python -m pytest -q -ra --disable-warnings --pyargs dpctl -vv
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1

libsyclinterface/dbg_build.bat

@@ -3,7 +3,7 @@ dpcpp.exe --version >nul 2>&1
 if errorlevel 1 (
     set ERRORLEVEL=
     call "%ONEAPI_ROOT%\compiler\latest\env\vars.bat"
-    if errorlevel 1 exit 1
+    if %ERRORLEVEL% neq 0 exit 1
 )
 @REM conda uses %ERRORLEVEL% but FPGA scripts can set it. So it should be reseted.
 set ERRORLEVEL=
@@ -31,13 +31,13 @@ cmake -G Ninja ^
     "-DCMAKE_LINKER:PATH=%DPCPP_HOME%\bin\lld-link.exe" ^
     "-DDPCTL_BUILD_CAPI_TESTS=ON" ^
     ".."
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1
 
 ninja -n
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1
 @REM ninja check
 @REM IF %ERRORLEVEL% NEQ 0 exit /b 1
 ninja install
-if errorlevel 1 exit 1
+if %ERRORLEVEL% neq 0 exit 1
 
 cd ..