Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!shelby!polya!Polya.Stanford.EDU!ham From: ham@Neon.Stanford.EDU (Peter R. Ham) Newsgroups: gnu.gcc.bug Subject: allocating floating point constants on the stack Message-ID: Date: 20 Sep 89 17:21:51 GMT Sender: USENET News System Distribution: gnu Organization: Stanford University Lines: 15 I'm using gcc as a cross compiler from a pmax to an m68k. The m68k uses a home brewed, byte-swapped vax floating point format. When the gcc optimizer pushes floating point constants (doubles) on the stack using immediate values, I lose, since it pushes the hosts format on the stack and not the targets. Is there a way to disable this particular optimization? Where is the optimization performed, so I can comment it out? What are the other instances where gcc converts doubles to integers before assembly? -- Peter Ham PO Box 3430 (h)(415) 324-9645 MS Computer Science Student Stanford, CA ham@polya.stanford.edu Stanford University 94309 (o)(415) 723-2067