------- Comment #14 from Jerry_V_DeLisle at rl dot gov 2008-01-30 18:04 ------- The pr19925 segfault is close by the segfault here but not the same. I am working on a patch for that one. What is becoming clear is that we are not handling expansion of constructors very well yet.
Also, I will add that the original problem with this PR shows that on ppc64, the expression types for the constructors are showing as EXPR_FUNCTION whereas on x86-64 where the test case passes, those same expressions are EXPR_CONSTANT or EXPR_VARIABLE. I fear that a pointer related problem is resulting in cobbering some things and it could be just about anywhere that happens. I am thinking we have to go back and scrutinize the matchers and work down through until we see where the corruption is happening. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34828
