This is being maintained by John Bandhauer <jband@netscape.com>. Feel free to email me with questions or to volunteer to contribute xptcall code for any platform.
Mike Shaver <shaver@mozilla.org> is the best contact regarding 'nix (Unix, Linux, Finux, etc.) ports of xptcall.
Status | Platform | Contributors and ? Possible Contributors | Notes |
---|---|---|---|
Done | Win32 x86 | John Bandhauer <jband@netscape.com> | win32 |
Done | Linux x86 |
John Bandhauer <jband@netscape.com> Ulrich Drepper <drepper@cygnus.com> |
unix |
Done | FreeBSD and NetBSD x86 |
Christoph Toshok <toshok@hungry.com>, John Bandhauer <jband@netscape.com> |
unix (same as Linux 86 code) |
Done | BSD/OS x86 | Bert Driehuis <bert_driehuis@nl.compuware.com> | unix (same as Linux 86 code) Bert contributed patches that *should* do the right thing for all the unixish-x86 versions of this code for GCC 2.7 or 2.8 vs. EGCS 1.1. He notes that the vtbl scheme is different. He is hoping that others will help test the changes using these two compilers on the various platforms where this same code is used. Bert's details |
Done | Mac PPC |
Roger Lawrence <rogerl@netscape.com>, Patrick Beard <beard@netscape.com> |
mac (passing tests and checked in) |
Done | Solaris Sparc |
Roger Lawrence <rogerl@netscape.com>, Chris McAfee <mcafee@netscape.com> |
unix This is checked in and working. |
Done | Solaris Sparc v9 (64bit) |
Stuart Parmenter <pavlov@netscape.com>, Chris Seawood <cls@seawood.org> |
unix This is checked in and (pavlov claims!) working. |
Done | OS/2 | John Fairhurst <mjf35@cam.ac.uk> | I never heard exactly who did what. But mozilla has been working on OS/2 for a long time now. |
Done | OpenVMS Alpha | Colin R. Blake <colin@theblakes.com> | Colin says this is passing all the tests on OpenVMS Alpha! |
Done | NT Alpha | bob meader <bob@guiduck.com> |
bob writes: Enclosed is xptcall for alpha/nt target.. It is a variation of the IRIS port (only targeted for win32). Notice the last 2 files (the change to mozilla\xpcom\build\makefile.win and mozilla\xpcom\build) are needed because I was unable to figure how to do a "declspecexport" from the assembler ASAXP ... if some knows how to do that then those last 2 files won't be needed. I have had someone look over this code at bridge.com (the entry point to compaq/gem compiler team) and this code was given the OK. I consider it "done".
This code lives in the files where the name includes 'alpha' in the win32 directory. |
Done | Linux ARM |
Stefan Hanske<sh990154@mail.uni-greifswald.de> ? Matthew Wilcox <willy@bofh.ai> |
Stefan's code is checked in and he says it is working. |
Done | Linux Sparc |
Anton Blanchard <anton@progsoc.uts.edu.au>,
Roger Lawrence <rogerl@netscape.com>, Brandon Ehle <ehle.3@osu.edu> |
Anton contributed patches to Roger's Sparc code. Anton says it works and passes the tests! (24-Aug-1999) Brandon writes: I've finished testing XPTCALL Sparc Linux on 12 different Sparc machines and it checks out good. |
Done | Linux PPC |
Patrick Beard <beard@netscape.com> Chris Waterson <waterson@netscape.com> Franz Sirl <Franz.Sirl-kernel@lauterbach.com> ? Jason Y. Sproul <jsproul@condor.fddi.wesleyan.edu> ? Sean Chitwood <darkmane@w-link.net> |
waterson said: Mozilla runs on Linux/PPC |
Done | Linux Alpha |
Glen Nakamura <glen.nakamura@usa.net> Dan Morril <morrildl@nycap.rr.com> |
Glen writes:
I am attaching a patch which contains my Linux Alpha xptcall code. It passes TestXPTCInvoke and TestXPC on my machine which runs kernel 2.2.7, glibc 2.1.1, and egcs 1.1.2. I have not tested it with older GNU compilers such as gcc 2.8.x. From looking at the Linux x86 code, I gather that the vtable layout is a little different for those compilers and the code will need minor modifications in order to work properly. I am not sure how much of the code can be used for OpenVMS Alpha and/or Digital UNIX. Currently the code is dependent on the g++ name mangling convention and a few GNU extensions so I'm not sure how useful it will be for the other systems. Hopefully the comments in the code are detailed enough to help people attempting a port.
|
Done | SunOS x86 |
Arthur Jones <aljones@lbl.gov> ? Philip Pokorny <ppokorny@mindspring.com> |
The word I hear is that this is working and done |
Done | AIX PPC | Jim Dunn <jdunn@netscape.com> | Philip K. Warren writes: We have gone through several releases of AIX without any problems. |
Done | Irix |
Jason Heirtzler <jasonh@m7.engr.sgi.com> |
Jason has declared this done. Jason is no longer working at SGI and will not be maintaining this code. There is some doubt as to whether or not this is working for everyone - see bug #10061. Mike Shaver <shaver@mozilla.org> is the interim maintainer until someone more suitable can be found. |
Done | BeOS x86 |
Duncan Wilcox <duncan@be.com> |
unix (yet another reuse of the Linux 86 code!) Duncan says this is all working. He did the code for old cfront style 'this' adjustment for others to use too! |
HELP! | BeOS PPC | - | - |
Done | Compaq Tru64 UNIX (Digital UNIX) |
Steve Streeter <streeter@zk3.dec.com> |
Code passes the tests and is checked in. |
Working | Neutrio x86 |
Jerry L. Kirk <Jerry.Kirk@Nexwarecorp.com> |
Patches for xptc*_unixish_x86.cpp checked in. Waiting for verification that this is really finished. |
Investigating | SCO UW7 and OSR5 |
J. Kean Johnston <jkj@sco.com> Evan Hunt <evanh@sco.com> |
Recent (Feb-2001) email from jkj@sco.com suggests that work will be occuring soon. |
Works | NetBSD/m68k |
Dave Huang <khym@bga.com> |
Dave's changes are in the tree. Waiting for verification that it is really finished. |
Partially Working | NetBSD/arm32 | Mike Pumford <mpumford@black-star.demon.co.uk> | Mike writes: I have started porting to the platform based on the code for Linux ARM. The InvokeByIndex code works correctly when used with TestXPTCInvoke. I am currently working on making TestXPC function correctly. I am doing the porting work with egcs-1.1.2 on NetBSD 1.4P (NetBSD-current snapshot from a couple of days ago). |
Done | Linux ia64 |
HP Ulrich Drepper <drepper@redhat.com> |
bug 40950 comment 15 |
HELP! | All others! | - | - |
Note: I've used the symbol (?) to indicate people who have expressed an interest in possibly contributing code. Just because these people are listed here does not mean that they have commited themselves to do the work. If you would like to contribute then let me know. Feel free to email these folks and offer to help or find out what's going on. We're all in this together.