home *** CD-ROM | disk | FTP | other *** search
- Path: Norway.EU.net!usenet
- From: patrick.hanevold@login.eunet.no (Patrick Hanevold)
- Newsgroups: comp.sys.amiga.programmer
- Subject: Re: Ceck out this bug. What the *** is SASC/6.56 doing?
- Date: 15 Mar 1996 18:21:58 GMT
- Organization: EUnet Norway
- Message-ID: <1271.6647T1136T1048@login.eunet.no>
- References: <1120.6643T91T406@login.eunet.no> <4hvk1f$o0c@zeus.central.ntua.gr>
- <874.6644T381T2315@login.eunet.no> <4i17ga$doo@btmpjg.god.bel.alcatel.be>
- <885.6646T9T693@login.eunet.no> <4i6led$stj@zeus.central.ntua.gr>
- NNTP-Posting-Host: pc3.asker-pm2-1.eunet.no
- X-Newsreader: THOR 2.22 (Amiga;TCP/IP)
-
-
- >>OpenScreenTags() doesnt crash. exit(0) does.
- >>I guess it maby could mess up so exit(0) would crash.
- >>Stach check should be smart enough to fail if there was too litle stack
- >>left. I would guess none of the OS calls would need more then 4K since thats
- >>the default stack size.
-
- >exit(0) probably crashes cause it peeks a wrong return address from the stack
- >You are right OS calls need no more than 4k but when you call the OpenScreen-
- >Tags you are left with less than one kilobyte of stack.
-
- >I dont think a compiler can ever check if stack will overflow when a call
- >will be made to a function that wasn't compiled so to include code for
- >stack checking.
- The minimum stack should be 4K by default.
-
- >My guess is that if you had the OS source and you had compiled it with 'stack
- >checking' then your program wouldn't crash.
- Yups.
-
- -----------------------------------------------------------------------------
- Patrick Hanevold - VR developer - patrick.hanevold@login.eunet.no
- Silicon Graphics Power Onyx system with four R8000 processors and two
- Reality Engine 2s. 5 gigabytes of RAM. 200 gigabytes of fast disk
- (single-file optimized read rate of 250 megabytes/sec).
- --------------------- But heck, my Amy boots faster. :) ---------------------
-
-