labath added a comment.
In D73767#1851418 <https://reviews.llvm.org/D73767#1851418>, @mgorny wrote:
> The question is, does this break any of the platform hacks? ;-)
I think it /may/ break one. See inline comment.
BTW, I think the best way to clean a lot of this up would be to take the tests
which build executables linking to liblldb, and rewrite them as c++ unit tests
(at least those that aren't skipped/xfailed everywhere), as cmake knows best
how to link to the thing it has built. Then, a lot of this stuff can be just
deleted...
================
Comment at: lldb/packages/Python/lldbsuite/test/lldbtest.py:1680
self.dylibPath] if self.dylibPath in os.environ else None
- lib_dir = os.environ["LLDB_LIB_DIR"]
+ lib_dir = configuration.lldb_libs_dir
if existing_library_path:
----------------
I'm not sure if this is right, because (despite the name) `LLDB_LIB_DIR` points
to the "bin" dir. But then again, this whole function seems pretty wrong, so
it's hard to say what should it really do here.
If you don't need this part (the previous patch didn't have it) and want to
make this safe-ish for cherry-picking, maybe you could leave this part out for
now.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D73767/new/
https://reviews.llvm.org/D73767
_______________________________________________
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
