[lldb] Add a cmake warning about the python/swig incompatibility

[labath]

Summary

Raise awareness of the fact that some versions of swig and python (and
build types) just don't mix.

One day this will be a reason to require swig>=4.0, but this version is
too hot off the press right now..

[amccarth]

If I recall correctly, the non-debug builds still had the problem, they just didn't have the assertion that made it obvious.

[labath]

In D88967#2317522, @amccarth wrote:

If I recall correctly, the non-debug builds still had the problem, they just didn't have the assertion that made it obvious.

Is that problem only theoretical (like, "you shouldn't be doing that") or does it have some practical consequences (crashes, incorrect operation, etc.)?

I'm certain that it is a "theoretical" problem, but AFAICT it doesn't have any practical consequences. I've been running python-3.8+swig-3.0.12 for quite some time now any I didn't observe see any issues. And I know a lot of other people are running a similar combo too..

I think the "benignness" of this assertion failure was the reason it took so long to find/fix this bug.

[amccarth]

In D88967#2317545, @labath wrote:

In D88967#2317522, @amccarth wrote:

If I recall correctly, the non-debug builds still had the problem, they just didn't have the assertion that made it obvious.

Is that problem only theoretical (like, "you shouldn't be doing that") or does it have some practical consequences (crashes, incorrect operation, etc.)?

My memory isn't that detailed. I think it was an argument being passed in as a single value but being referenced as though it was an array or list. Or vice versa.

[JDevlieghere]

Neat!

[labath]

In D88967#2317851, @amccarth wrote:

In D88967#2317545, @labath wrote:

In D88967#2317522, @amccarth wrote:

If I recall correctly, the non-debug builds still had the problem, they just didn't have the assertion that made it obvious.

Is that problem only theoretical (like, "you shouldn't be doing that") or does it have some practical consequences (crashes, incorrect operation, etc.)?

My memory isn't that detailed. I think it was an argument being passed in as a single value but being referenced as though it was an array or list. Or vice versa.

If we're talking about the same thing then it's about passing None instead of an empty list/dictionary. And it seems this is benign because the way python checks for emptyness also returns zero for None values.

lldb/cmake/modules/FindPythonAndSwig.cmake

 else()
   find_package(SWIG 2.0)
   if (SWIG_FOUND)
       FindPython3()
   else()
     message(STATUS "SWIG 2 or later is required for Python support in LLDB but could not be found")
   endif()
 
+  get_property(MULTI_CONFIG GLOBAL PROPERTY GENERATOR_IS_MULTI_CONFIG)
+  if ("${Python3_VERSION}" VERSION_GREATER_EQUAL "3.7" AND
+      "${SWIG_VERSION}" VERSION_LESS "4.0" AND WIN32 AND (
+      ${MULTI_CONFIG} OR (${uppercase_CMAKE_BUILD_TYPE} STREQUAL "DEBUG")))
+    # Technically this can happen with non-Windows builds too, but we are not
+    # able to detect whether Python was built with assertions, and only Windows
+    # has the requirement that Debug LLDB must use Debug Python.
+    message(WARNING "Debug builds of LLDB are likely to be unusable due to "
+      "<https://github.com/swig/swig/issues/1321>. Please use SWIG >= 4.0.")
+  endif()
+
   include(FindPackageHandleStandardArgs)
   find_package_handle_standard_args(PythonAndSwig
                                     FOUND_VAR
                                       PYTHONANDSWIG_FOUND
                                     REQUIRED_VARS
                                       Python3_LIBRARIES
                                       Python3_INCLUDE_DIRS
                                       Python3_EXECUTABLE
                                       SWIG_EXECUTABLE)
 endif()