Xref: utzoo news.software.b:3734 unix-pc.general:4236 Path: utzoo!utstat!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!zaphod.mps.ohio-state.edu!usc!apple!uokmax!servalan!rmtodd From: rmtodd@servalan.uucp (Richard Todd) Newsgroups: news.software.b,unix-pc.general Subject: Re: Cnews and gcc 1.36 on UNIXPC Message-ID: <1989Dec3.053129.1227@servalan.uucp> Date: 3 Dec 89 05:31:29 GMT References: <1989Dec2.022346.436@ivucsb.sba.ca.us> <1989Dec2.235241.6935@utzoo.uucp> Reply-To: rmtodd@servalan.UUCP (Richard Todd) Organization: Ministry of SIlly Walks Lines: 21 In article <1989Dec2.235241.6935@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: >In article <1989Dec2.022346.436@ivucsb.sba.ca.us> todd@ivucsb.sba.ca.us (Todd Day) writes: >>Usually, I have no trouble with Cnews when using the stock UNIXPC cc >>compiler. However, when I tried using gcc1.36, Cnews would claim that >>all articles are duplicates (even happens during the regression test!). >I'd suspect problems with the interface to the dbm library. As I >understand it -- I haven't used gcc significantly -- gcc does *not* >use pcc-compatible conventions for calling struct-valued functions >unless told to. The dbm interface includes some such functions. Yep, that's exactly what happens. All history-file-related functions (like checking whether articles are dupes or not) will not work if you forget to use -fpcc-struct-return on the gcc command line. With that, everything should work fine; a friend of mine got C News running on an Encore Multimax with gcc, and as I recall we didn't have to specify any flags to gcc except for -fpcc-struct-return and -traditional. -- Richard Todd rmtodd@uokmax.ecn.uoknor.edu rmtodd@chinet.chi.il.us rmtodd@servalan.uucp Motorola Skates On Intel's Head! Brought to you by Super Global Mega Corp .com