Commit Graph

109 Commits

Author SHA1 Message Date
Peter Wemm
fc369c7cd9 Add missing quote after cut/paste error in last commit. :-( 2003-06-26 20:34:08 +00:00
Peter Wemm
01e5a2aede Configure gcc in biarch mode for amd64 so that the -m32 flag is enabled
for things like boot blocks etc.

Reviewed by:  kan
2003-06-26 03:02:32 +00:00
Peter Wemm
a96996a182 When building a shared library, link it against libgcc_pic.a instead of the
non-PIC libgcc.a.  Linking non-pic code into a shared library is not
a good thing.  It happens to break amd64 at compile time, and the ppc
folks want it too.  The problem is mainly with C++ code, unwind-dw2.c
in particular.  Most of the other functions in libgcc.a are self
contained so most of the time it isn't a problem.  The dwarf2 unwinder
is not safe though since it does make global variable references.

Reviewed by:	kan
2003-06-13 22:25:41 +00:00
David E. O'Brien
8f3bf9945d Lock GCC w/o -m32 into 64-bit mode. 2003-05-01 15:00:46 +00:00
David E. O'Brien
a12d8aed9a We're changing the platform name from x86[_-]86 to amd64. 2003-04-26 20:04:38 +00:00
David E. O'Brien
f685377c2e Remove our custom mixed ELF/a.out support. This means the base compiler
now only produce ELF objects.  It also makes us closer to stock GCC, and
simplifies the set of changes we still need from stock GCC on every import.

Applauded by:	peter
Approved by:	re
2002-11-26 18:25:20 +00:00
Ruslan Ermilov
4192cfffe2 Bootstrapping aid for 4.0-RELEASE. 2002-11-13 11:50:40 +00:00
David E. O'Brien
597e16e012 We HAVE_STPCPY now. 2002-10-20 07:50:20 +00:00
Alexander Kabaev
9734ce03a2 Add multilib_options variable definition.
Approved by:	obrien
2002-10-10 04:56:35 +00:00
David E. O'Brien
1ca4970561 We're going to have to use a TARGET_ARCH of "x86_64" rather than "x86-64"
to make building GCC happy.
2002-09-12 15:56:21 +00:00
David E. O'Brien
60f9b09d26 Allow one to profile FORTRAN77 programs. 2002-06-09 00:03:56 +00:00
David E. O'Brien
9291f1b122 Restore revision 1.15 (use profiled C++ libs) which wasn't on the WIP_GCC31
branch and thus was lost in the shuffle on the move to Gcc 3.1.
2002-06-08 19:58:31 +00:00
David E. O'Brien
86f99a57b2 Read specs from /usr/libdata/gcc/specs if it exists. 2002-06-06 03:47:02 +00:00
David E. O'Brien
f09cdb5b44 Search the "backward" C++ compatibility include dir also. 2002-06-05 16:10:59 +00:00
David E. O'Brien
ca9578f3ef I missed the setting of PREFIX_INCLUDE_DIR when I visually santity checked
the output from a autoconfig run.
2002-05-27 17:48:28 +00:00
David E. O'Brien
5a87307b7f Use a more API denoting way to handle what is in libc and what isn't. 2002-05-18 04:49:44 +00:00
Ruslan Ermilov
35abacef2a MD_EXEC_PREFIX doesn't work for the cross-arch compiler.
The change also makes the `cc -print-search-dirs' output
sane (the pre-3.1 way) in the non-cross case.

Draft reviewed by:	obrien
2002-05-16 15:22:58 +00:00
Ruslan Ermilov
2898afe627 Make it possible to build a cross compiler for alpha,
ia64 and sparc64 on systems that do not have atoll(3).
The "cross" here doesn't necessarily mean cross-arch.
2002-05-16 15:18:13 +00:00
David E. O'Brien
fcbdc1f8a0 Add x86-64 bits. 2002-05-15 22:40:50 +00:00
David E. O'Brien
23735e10dd The IA-64 config needs to know that we are using GNU ld & as.
Submitted by:	peter
2002-05-15 21:59:46 +00:00
David E. O'Brien
79c021244b Fixes for building a.out bits.
Submitted by:	bde
2002-05-12 12:01:12 +00:00
David E. O'Brien
d147c3da04 Touching the sjlj setting on IA-64 makes things not build.
Submitted by:	peter
2002-05-10 17:42:19 +00:00
David E. O'Brien
415f2bb46f Gather up the stragglers that depends on genrtl.h. This is -j10 safe now. 2002-05-10 10:21:19 +00:00
David E. O'Brien
01c50f1782 This was *very* -j unsafe. Add a dependency on the common generated
headers to mostly make it -j1 safe.
2002-05-10 10:14:53 +00:00
David E. O'Brien
3cdd876f04 Bmake bits for Gcc 3.1.
Partially made possible by:	Wilko.Bulte@compaq.com
2002-05-10 08:54:50 +00:00
David E. O'Brien
7b4716843d Use MD_EXEC_PREFIX now to get us thru `buildworld'.
The problem is the GCC driver now turns STANDARD_EXEC_PREFIX into a relative
path -- "<basename argv[0]>/../../libexec" for our normal install location.
However, in the middle of `buildworld' we need
"<basename argv[0]>/../../../../libexec" due to the prefix we tell the GCC
driver.  But either the GCC driver is buggy, or we are confusing it, as it
tries to exec "<basename argv[0]>/../../libexec/cpp0" as if it were installed
in the normal place (but isn't).
MD_EXEC_PREFIX is still absolute, so I'll use that for now.  I would like to
later make it so MD_EXEC_PREFIX is set only for `buildworld', as
MD_EXEC_PREFIX is also in the search path for libraries. Don't ask me why!
Another way is to add ${OBJFORMAT_PATH} (as set in CROSSENV) to the PATH
in src/Makefile.inc's WMAKEENV.
2002-05-10 08:41:46 +00:00
David E. O'Brien
66ecbefa3d The HAVE_AS_GOTOFF_IN_DATA definition needs to depend on obj format. 2002-05-07 01:19:56 +00:00
David E. O'Brien
a5cc86f163 Add support for using the profiled versions of the C++ (and related) libs. 2002-05-01 19:19:22 +00:00
David E. O'Brien
9d838dbb47 Remove the #ifdef IN_GCC junk. We *know* we are building GCC with these
bits.  Also remove comment about keeping in sync with other instances in
the source tree -- it was too easy to get out of sync, so the other
instances now use this instance.
2002-04-15 03:41:47 +00:00
David E. O'Brien
d99ab028e5 Note that HAVE_GAS_SHF_MERGE is a new feature, and it can be surprising if
one does not know about it.

Experienced by:	peter
2002-04-15 03:39:20 +00:00
David E. O'Brien
9e3b001017 Bmake bits for GCC 3.1. 2002-04-06 23:18:01 +00:00
David E. O'Brien
bb5545c63e Tell GCC 3.1 our capibilities. 2002-04-05 12:14:36 +00:00
David E. O'Brien
6d57d58120 A little more reorg. 2002-04-05 10:23:19 +00:00
David E. O'Brien
051ab09117 Minor reorg. 2002-04-04 19:36:33 +00:00
David E. O'Brien
fda035bb2c MFC: remove 2.6.3 cc_int shlib cruft and s/GNU_ARCH/TARGET_ARCH/g. 2002-04-04 18:30:57 +00:00
David E. O'Brien
953bfb1637 Remove some local cruft that snuck in yesterday. 2002-04-04 18:24:56 +00:00
David E. O'Brien
5ccf2039e4 Get rid of GCC_ARCH, and just use plain TARGET_ARCH.
We got rid of the MIPS le/be stuff that needed this a long time ago.
2002-04-04 00:11:00 +00:00
David E. O'Brien
e4753251f7 Matching contrib/gcc/gcc.c commit of rev 1.23, remove misleading comment
about 'STANDARD_EXEC_PREFIX'.  It is used now and is needed if one does not
set 'MD_EXEC_PREFIX'.

Do not define a 'MD_EXEC_PREFIX'.  It is not needed, not used in the
cross case, and just ends up causing "/usr/libexec" being added to the
library search path.
2002-03-05 00:39:29 +00:00
David E. O'Brien
a370115d7d Allow for better control over the GCC front-end when building a cross
compiler.

* Undo the diking out of cross compiler logic from gcc.c rev 1.16.
* Add the `CROSS_STARTFILE_PREFIX' knob.
* Add our own definition of `STANDARD_INCLUDE_DIR'.  This should have been
  included in freebsd-native.h rev 1.5.
2002-03-05 00:17:24 +00:00
David E. O'Brien
3d3dea1bbc Move the creation of the insn-*.c files from cc_tools to cc_int.
This gets rid of a cross build problem we have because we build
everything in cc_tools during the `make build-tools' (or `make depend')
stage.
2002-03-02 08:53:36 +00:00
David E. O'Brien
c21c782ba5 Clean up the style a little bit. 2002-03-02 01:06:42 +00:00
David E. O'Brien
4a88cacfed MFC: rev 1.13 (use libgcc_p.a when -pg is in effect) 2002-03-02 00:54:17 +00:00
David E. O'Brien
4cef291124 Use libgcc_p.a when -pg is in effect.
I am committing this here rather than in gcc/config/freebsd.h because the
profiled libgcc only exists with the native system compiler.  It is not
created by a stock FSF build and we will never be able to get these bits
committed to the FSF CVS repo.  Thus this is very much a FreeBSD "native"
issue.
2002-02-24 06:05:22 +00:00
David E. O'Brien
9237b562e5 Allow for better control over the GCC front-end when building a cross
compiler.

* Undo the diking out of cross compiler logic from gcc.c rev 1.16.
* Add the `CROSS_STARTFILE_PREFIX' knob.
* Add our own definition of `STANDARD_INCLUDE_DIR'.  This should have been
  included in freebsd-native.h rev 1.5.
* Minimize a little bit more, things we dike out in the FREEBSD_NATIVE case.

Submitted by:	ru & obrien
2002-02-20 19:41:54 +00:00
David E. O'Brien
11dc790e85 Clean up the style a little bit. 2002-02-20 03:57:21 +00:00
David E. O'Brien
8f625d4e6c Do not define a 'MD_EXEC_PREFIX'. It is not needed, not used in the
cross case, and just ends up causing "/usr/libexec" being added to the
library search path.

Also remove misleading comment about 'STANDARD_EXEC_PREFIX'.  It is needed
if one does not set 'MD_EXEC_PREFIX'.

Submitted by:	ru
2002-02-20 03:10:54 +00:00
David E. O'Brien
8561d1123d Changes in order to compile GCC 3.0.2.
Really irritating changes are the "forced" layering of malloc + friends
in order to use the GNU versions.  Sorry, we have a *very* fine malloc,
and we will use it.  Period.  Even more irritating is that the GNU people
now want to replace ctype also!!  So we partially dike it out here.
2001-12-18 03:19:11 +00:00
David E. O'Brien
e5d7f6b935 Regenerated for GCC 3.0.2. 2001-12-18 03:15:02 +00:00
David E. O'Brien
838bc11f60 We need to build ``.depend'' early in the "build-tools" for the GCC
"build-tools".  If we do not do this, the "depend" stage of
"buildworld" will build ``.depend'' and it will record the wrong
library and header dependencies (DESTDIR=${WORLDTMP}).  Even worse,
the "all" stage may clobber build-architecture-format build tools
built in the "build-tools" stage with target-architecture-format ones.

Submitted by:	ru
2001-09-27 17:14:33 +00:00
David E. O'Brien
e7d01c5d0b Include `dbxelf.h' via tm.h (the approved FSF/GNU way) vs. in our MI header. 2001-03-02 03:00:41 +00:00