Re: [Open64-devel] r3574 - in trunk/osprey/common/com: . MIPS NVISA SL ia64 loongson ppc32 x8664

2011-05-03 Thread Gilmore, Doug
Arggh, sorry I dropped Hui's message with the patch for the fix for this issue. I tried it out and ran our gcc/g++/gfortran sanity tests, which is cobbled from the tests in the testsuite directory of the GCC 4.2 release. I did see one error that I haven't had a chance to investigate: $ openf95

[Open64-devel] Code Review Request for stdcall

2011-05-03 Thread Feng Zhou
Hi, Can a gatekeeper review the attached patch related to stdcall for me please? Thank you very much. Background -- The problem is related to gcc's extension, namely the stdcall function attributes. It is exposed while using gnu m4 as the test case. According to gcc, stdcall

Re: [Open64-devel] r3574 - in trunk/osprey/common/com: . MIPS NVISA SL ia64 loongson ppc32 x8664

2011-05-03 Thread Gilmore, Doug
Is anyone looking into this? It appears to me that the compiler built with debugging is quite broken, since wgen42 will assert when compiling the following code: foo (int); void bar(void) { foo(-4); } The problem that when the WHIRL node is created for -4 via: WN *WN_CreateIntconst(OPERATO