jess@gn.ecn.purdue.edu (Jess M Holle) (06/26/91)
In hopes of finding some silverlining in the clouds of MPW, how does the speed and size of gcc (the MPW ported version) compare with Think and MPW C? I know that gcc doesn't produce debugger code and that it is slower yet at compilation than MPW C. I would just like to be able to produce fast, small code out of my finished source code with little or no changes. Is this possible? Does gcc work under System 7.0? Does it use MPW C's headers? If not, how do you get access to System 7.0 features (I'm assuming that it works w/ System 7.0)? Thanks in advance for your responses. Jess Holle
dorner@pequod.cso.uiuc.edu (Steve Dorner) (06/26/91)
In article <1991Jun25.181936.29476@gn.ecn.purdue.edu> jess@gn.ecn.purdue.edu (Jess M Holle) writes: >In hopes of finding some silverlining in the clouds of MPW, how does the >speed and size of gcc (the MPW ported version) compare with Think and >MPW C? The docs that come with the port indicate several things: 1) The compiler is a pig; 4M to run. 2) The code produced is about 10% faster, around the same size. 3) The compiler is designed as a drop-in replacement for the MPW C compiler; it's compatible with supplied libraries, header files, linker, etc. Or that's what I remember. I couldn't quite bring myself to use a different production compiler from my debugging compiler, even though I'm quite fond of gcc on UNIX. -- Steve Dorner, U of Illinois Computing Services Office Internet: s-dorner@uiuc.edu UUCP: uunet!uiucuxc!uiuc.edu!s-dorner
ml27192@uxa.cso.uiuc.edu (Mark Lanett) (06/26/91)
dorner@pequod.cso.uiuc.edu (Steve Dorner) writes: >In article <1991Jun25.181936.29476@gn.ecn.purdue.edu> jess@gn.ecn.purdue.edu (Jess M Holle) writes: >>In hopes of finding some silverlining in the clouds of MPW, how does the >>speed and size of gcc (the MPW ported version) compare with Think and >>MPW C? >The docs that come with the port indicate several things: >1) The compiler is a pig; 4M to run. >2) The code produced is about 10% faster, around the same size. >3) The compiler is designed as a drop-in replacement for the MPW C compiler; > it's compatible with supplied libraries, header files, linker, etc. Also, it requires Asm (it's not a native compiler). It requires the 3.1 headers because it doesn't understand the #pragma in the 3.2 headers (and apparently isn't going to learn about them), so using it for system 7 stuff would seem difficult. >Or that's what I remember. I couldn't quite bring myself to use a different >production compiler from my debugging compiler, even though I'm quite fond >of gcc on UNIX. >-- >Steve Dorner, U of Illinois Computing Services Office >Internet: s-dorner@uiuc.edu UUCP: uunet!uiucuxc!uiuc.edu!s-dorner -- //----------------------------------------------------------------------------- Mark Lanett mlanett@uiuc.edu Software Tools Group, NCSA
neeri@iis.ethz.ch (Matthias Ulrich Neeracher) (06/26/91)
In article <1991Jun26.004523.23230@ux1.cso.uiuc.edu> dorner@pequod.cso.uiuc.edu (Steve Dorner) writes: >In article <1991Jun25.181936.29476@gn.ecn.purdue.edu> jess@gn.ecn.purdue.edu (Jess M Holle) writes: >>In hopes of finding some silverlining in the clouds of MPW, how does the >>speed and size of gcc (the MPW ported version) compare with Think and >>MPW C? > >The docs that come with the port indicate several things: > >1) The compiler is a pig; 4M to run. Actually, I believe it runs with somewhat less. The interesting thing is that memory usage doesn't seem to grow as quickly as for MPW C. gcc was able to compile a particularly nasty file that MPW C couldn't do in the available memory (Imagine the C code generated from a flex file with 450 looong rules). >2) The code produced is about 10% faster, around the same size. Except for multiplications with constants, for which MPW C generates code that is both faster and smaller. I found the speed difference not that significant, but then I use a lot of pointers and not very tight loops. The similiarities of MPW and GCC code led me to conclude that MPW C code quality is generally underestimated, fiven gcc's reputation. Matthias ---- Matthias Neeracher neeri@iis.ethz.ch "These days, though, you have to be pretty technical before you can even aspire to crudeness." -- William Gibson, _Johnny Mnemonic_
evensen@husc9.harvard.edu (Erik Evensen) (06/26/91)
maybe this is getting ambitious but does anyone know if a) g++ has been ported to work with MPW and b) if it has been ported, will it work with MacApp? If not: something to think about? --Erik (evensen@husc.harvard.edu)