At Fri, 4 Jun 2010 10:16:08 +0100 (BST), Iain Hibbert
<plunky%rya-online.net@localhost> wrote:
Subject: Re: PCC (pcc-libs/csu/netbsd/crtbegin.c) doesn't build in NetBSD-5
>
> that would be nice but manpower does not seem to be available to mirror
> development work onto a branch (I don't have a -5 system and moreover am
> about to go offline for 3 months.. I don't know if anybody else wants to
> work on that in the meantime)
Understood!
> I have imported the latest version to -current yesterday; if you checkout
> the -current src/external/bsd/pcc directory and merge the following
> changes (previously made) to the set lists
>
> src/distrib/sets/lists/base/mi -r1.858
> src/distrib/sets/lists/comp/mi -r1.1385
> src/distrib/sets/lists/comp/shl.mi -r1.103
> src/distrib/sets/lists/man/mi -r1.1187
> src/distrib/sets/lists/comp/mi -r1.1404
>
> then defining the make variables MKPCC=yes and MKPCCCMDS=yes should enable
> you to create a release or distribution build with pcc(1), pcpp(1) and
> ccom(1) binaries for the target architecture.
Awesome! That's even better than I'd hoped for! I'll try pulling the
whole works over to my own netbsd-5 tree and see how it goes, though
I'll probably stick with the current toolchain for building the release.
If I have time or some other excuse I may try -current as well.
> There have also been miscellaneous fixes applied to the NetBSD-current
> source tree to correct GNUC dependencies that you might come across. If
> you want to build parts of NetBSD-5 you will have to discover them..
For the moment I'm most interested in trying my own code with pcc and so
I guess I'll only be worried about header files and such.
I do look forward to a day when the whole tree compiles with pcc and maybe
other third-party compilers as well!
> Please be aware though that pcc is still in development (Ragge is hard at
> work and this morning I recieved notifications that two more issues from
> the bug tracker were fixed :) so you may find other problems. Please
> consider to report them to the pcc-list or bugtracker, I found that Ragge
> is very responsive when the problem can be reduced to show the issue
> directly.
Will do! Thanks again!
--
Greg A. Woods
Planix, Inc.
<woods%planix.com@localhost> +1 416 218 0099 http://www.planix.com/
Attachment:
pgp4DCDwwCwVN.pgp
Description: PGP signature