Misha Koshelev mk144210@bcm.edu writes:
The previous patch was last week I think. This fixes a Valgrind error in the test. This is correct as vtResult is just hte _expected_ return type, and this way we check the actual return type (new this version) and then the validity of the pointer (same as old patch) before calling lstrcpyW so we don't pass NULL to lstrcpyW.
It seems to me what you really want to do is check that the call succeeded, instead of relying on the variant being cleared on error.