delorie.com/archives/browse.cgi | search |
On Tue, Oct 23, 2001 at 10:41:36AM +1000, Robert Collins wrote: >My 2c on this is that this could be a lot worse than a malloc issue... >even though it is occuring at process exit. It's not a malloc issue. gcc's -O3 optimization is screwing up regparm parameter passing, specifically in fhandler_console::read which seems to need to call alloca for some reason, although I can't say exactly why. It's actually clobbering the EAX register almost immediately on entry into the function. AFAICT, EAX is the register that should hold the first argument. I knew that regparm was buggy with some gcc's. Apparently 2.95.3 is one of them. cgf
webmaster | delorie software privacy |
Copyright © 2019 by DJ Delorie | Updated Jul 2019 |