Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: Notesfiles $Revision: 1.6.2.17 $; site uiucdcsb.UUCP Path: utzoo!watmath!clyde!cbosgd!ihnp4!inuxc!pur-ee!uiucdcsb!grunwald From: grunwald@uiucdcsb.UUCP Newsgroups: net.lang.c Subject: Re: C stack frame sizes Message-ID: <9000033@uiucdcsb.UUCP> Date: Sat, 1-Dec-84 19:04:00 EST Article-I.D.: uiucdcsb.9000033 Posted: Sat Dec 1 19:04:00 1984 Date-Received: Tue, 4-Dec-84 04:48:35 EST References: <18092@arizona.UUCP> Lines: 18 Nf-ID: #R:arizona:-1809200:uiucdcsb:9000033:000:561 Nf-From: uiucdcsb!grunwald Dec 1 18:04:00 1984 /* Written 9:27 pm Nov 30, 1984 by malcolm@ecn-ee in uiucdcsb:net.lang.c */ ... I wonder what this type of programming style would do to a Berkeley style RISC machine? Malcolm /* End of text from uiucdcsb:net.lang.c */ On the pryamid, which is essentially a Berkeley style RISC machine, the array would be stored in real memory, not registers. Same goes for structs and (I think) unions. They only keep the "normal stuff" in registers. So, if you defined: int *i, j[1000], *k; both "i" and "k" would be in registers and j would be in memory.