Commit Graph

152 Commits

Author SHA1 Message Date
obrien
1dc4a82010 Move -I of cc_tool/ before the GCCDIR ones.
When I imported EGCS into contrib/egcs/ I failed to prune out
egcs/gcc/cp/hash.h which is generated from gxx.gperf.  Thus `cc1plus' wasn't
using the hash.h we generated by cc/cc_tools/Makefile, but rather the one in
egcs/gcc/cp/.

When I imported contrib/gcc/ I did prune gcc/cp/hash.h.  Unfortunately the
GCC maintainers weren't smart on their file nameing and there is also a
egcs/gcc/hash.h (name overloading does NOT work as well on the filesystem
as in C++...).  Due to the -I ordering we are were then picking up gcc/hash.h
when compiling `cc1plus'.
1999-10-12 20:22:38 +00:00
peter
289c0d262f $Id$ -> $FreeBSD$ 1999-08-27 23:37:10 +00:00
peter
0b6fbc2a5b Make gcc work on the Alpha again. alpha/alpha.h is used instead of svr4.h 1999-08-19 21:19:36 +00:00
obrien
07419aec71 Rev 1.36 also includes functional chagnes I intended to commit seperately.
Configuration header inclusion has been moved around to reduce diffs from
the offical GCC distribution.  We now generate the same ``tm.h'' produced by
gcc's `configure' script [minus all the "#ifdef IN_GCC"'s].

Jeff Law of EGCS/Cygus says the new "approved" way of doing configure-related
includes is to list them all in ``tm.h'' rather than having the architure
config headers include large numbers of other configure headers.
1999-08-16 09:55:16 +00:00
obrien
b3af8b5f4b Purely cosmetic cleanups.
- Id tag layout wrong.
 - "FOO =" -> "FOO=".
 - 4 space indention on continuation lines rather than a <tab>

Basic idea checked by:	bde
1999-08-16 04:10:48 +00:00
obrien
e873ea78b5 Move "LIBGCC_SPEC" to freebsd-native.h as it is not approapiate for
submitting to the GCC mainainers for folding into the stock GCC.
1999-08-15 21:59:46 +00:00
obrien
8470008abd Use Bison with "--broken-undeftoken-init". 1999-08-14 22:44:36 +00:00
obrien
033405c257 Rather than deleting the grammer files during building, move them out of the
way.  This way they are available for inspection when debugging problems.
1999-08-14 22:40:01 +00:00
bde
f0457b9b57 Forgotten in previous log message:
PR:		misc/12920
1999-08-08 07:31:47 +00:00
bde
a4f669aaaf Install c++filt in /usr/libexec/${OBJFORMAT}. The version that
was installed in /usr/bin normally got clobbered when objformat
was installed.  Indirection through objformat is correct although
underscore handling is the only thing that differs for aout and
elf -- going through objformat is the easiest way to set c++filt's
underscore handling flag correctly.
1999-08-08 07:30:02 +00:00
obrien
3766f18249 Only use Bison to generate cc1plus's parse code. 1999-08-01 15:44:51 +00:00
obrien
f979029f4c Go back to using Bison for now, due to `make world' breakage.
BTW, if our src/contrib/bison is upgraded from 1.25 to the current 1.27
offering, the build breaks exactly the same way....
1999-07-29 09:49:29 +00:00
obrien
ebe74fdfd5 Use Yacc rather than Bison. 1999-07-28 07:22:08 +00:00
obrien
a4b8ba086c The Unix way is arguments come before an input file name, not after. 1999-07-28 07:19:38 +00:00
bde
1f3f20e6c0 Don't override the (correct) default for BINDIR. 1999-07-04 15:51:39 +00:00
obrien
522e272a7b Rather than hack config/freebsd.h to contain our hacks for a native compiler,
just include them in tm.h (as built from cc/cc_tools/Makefile).

This will reduce the diffs from the vendor sources.

Excellent idea by:	jdp
1999-04-28 18:48:08 +00:00
obrien
d2b4c4a138 use `target' macro crafted in parent Makefile rather than making our own
Submitted by:	Warner Losh
1999-04-25 17:23:28 +00:00
obrien
d0bc3cab3a Hookup FORTRAN info docs.
Submitted by:	Mark Murry
1999-04-25 17:09:48 +00:00
obrien
a099f8cc73 Sorry didn't get this in the last commit: MACHINE_ARCH changes courtesy
of Warner Losh <imp@village.org>.  (he's got plans for FreeBSD mips :-))
1999-04-22 20:12:21 +00:00
obrien
c506ddebe9 Use GNU_ARCH' in most places where MACHINE_ARCH' was used. Building for
mips has some particularlies.
1999-04-22 19:35:16 +00:00
obrien
6b5c8c5794 Have config/i386/freebsd.h directly include att.h. I have no idea why stock
EGCS doesn't just do this.  None of the rest of our other platforms will need
this file.
1999-04-22 00:20:36 +00:00
obrien
b710768bd3 Add FORTRAN support. 1999-04-18 09:44:36 +00:00
obrien
025141faad Move HAVE_CONFIG_H define to global cc/ Makefile so all can share the benifits. 1999-04-18 09:42:56 +00:00
obrien
97db6d0b10 *** empty log message *** 1999-04-18 09:41:49 +00:00
obrien
7a7b31aa6c Use inherited BINDIR defintion. 1999-04-18 09:39:38 +00:00
obrien
c00cfee64b Include the parent Makefile.inc so we get a proper BINDIR definition.
As explained by:	bde
1999-04-18 09:36:11 +00:00
obrien
b89c50491d Hookup the new FORTRAN compiler. Define `NO_FORTRAN' to not compile it. 1999-04-18 09:34:01 +00:00
obrien
f5af675abe Add the EGCS g77 FORTRAN compiler. 1999-04-18 09:33:25 +00:00
obrien
5a2a7a621f Minimum change to convert from home-grown f77 to EGCS's g77.
(Pre-bdelinted)
1999-04-18 09:22:56 +00:00
obrien
d26d808af6 Make room for g77. 1999-04-18 09:17:31 +00:00
markm
3f5095a2fa As the INFO stuff un doc/ works, we may as well build/install it.
OK'ed by:	obrien
1999-04-13 17:56:28 +00:00
bde
33795a33e0 Cleaned up egcs changes (mainly to minimise diffs with the gcc version
and restore order to sorted lists).
1999-04-11 04:32:42 +00:00
bde
b4d0b18a23 Fixed unusable dependencies in `.depend'. The following comment in the
code still applies, but the code attached to it had rotted:

# ../Makefile.inc will put an absolute path to our objdir in CFLAGS.
# Prevent mkdep from using it, so that we don't have to give rules for
# aliases of generated headers.

Many other makefiles seem to have the same bug (i.e., spelling "." as
"${.OBJDIR}" or as an even more complicated alias in -I directives).
1999-04-11 04:11:27 +00:00
bde
20977f06d8 Fixed world breakage. Fixed (?) building with -jN (N > 1).
The previous commit made things worse by breaking the -j0 case and
moving around the breakage for the -jN case.  The fix involves
restoring .ORDER statements that were deleted in the previous commit,
removing wrong special handling of tree-check.h, and rewriting the
generation of fudged dependencies based on an idea I got from the
previous commit (filter out problematic objects first).
1999-04-09 18:46:48 +00:00
obrien
2034f40c32 Another fix for -jN building -- the last fix didn't fix it for everybody.
Submitted by:	Luoqi Chen <luoqi@watermarkgroup.com>
1999-04-08 23:37:46 +00:00
peter
217c68f1e1 Clean up some stuff that's no longer used. 1999-04-08 15:43:36 +00:00
peter
272c0ffe18 Clean up the multilib stuff so that it actually works..
cc -aout now works again.
1999-04-08 15:11:44 +00:00
obrien
5810f4ab82 Remove some of the dependencies. This removes the cycles I was getting on
`make -j 8' builds.
1999-04-08 12:26:07 +00:00
obrien
7b5f3ba39b Due to other explicit dependancies, don't treat tree-check.h as a SRCS. 1999-04-08 09:57:51 +00:00
obrien
36332ab857 Fix a circular dependancy. 1999-04-08 09:21:52 +00:00
obrien
760f9d518d alpha/freebsd-elf.h is being folded into alpha/freebsd.h
Submitted by:	Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp>
1999-04-08 07:38:51 +00:00
obrien
4413bdb033 Use Haifa scheduler on the Alpha. 1999-04-06 20:08:01 +00:00
peter
cc69e950c0 Fix type mismatch.. 1999-04-05 10:18:50 +00:00
obrien
eae521d2fd Correct contents for the EGCS compiler documentation.
Currently it is unbuildable, but an info-aware developer should be able to get
this working w/o too much trouble.
1999-04-05 07:10:13 +00:00
obrien
9c9c4fe46a Conditionalize one more i386'ism. 1999-04-04 20:47:33 +00:00
obrien
1a7add86bb Attempt at creating the right ``tm.h'' file for the Alpha. 1999-04-04 20:40:45 +00:00
imp
e346b32961 Remove mkstemp.c from build. This is now in libc.
Tested by: make buildworld
1999-04-04 20:29:43 +00:00
obrien
dd605303bb didn't get latest version:
Don't conditionalize -I${.CURDIR}/../cc_tools
1999-04-04 17:11:00 +00:00
obrien
582be693c1 Minimum set of changes to switch from Gcc 2.7.2 (in contrib/gcc) to Egcs 1.1.2
(in contrib/egcs)
1999-04-04 16:36:35 +00:00
obrien
3ab362d927 `make world' works fine -- no reason to revert the move to gnu/lib/ 1999-04-02 17:35:14 +00:00