373c220202
gcc/config/sh/sh.h original rev. 1.51
...
gcc/config/sh/sh.c original rev. 1.43
> Wed Feb 23 16:42:21 2000 J"orn Rennecke <amylaar@cygnus.co.uk>
>
> * sh.h (LOOP_ALIGN): If aligning loops, call sh_loop_align
> to check for special cases.
> * sh.c (sh_loop_align): Define.
2002-04-05 15:50:14 +00:00
16f8beb6c9
apply gcc original rev. 1.09
...
> Wed Feb 23 16:42:21 2000 J"orn Rennecke <amylaar@cygnus.co.uk>
>
> * final.c (shorten_branches): Make value passed to LOOP_ALIGN
> conform to documentation.
2002-04-05 15:28:07 +00:00
8ed9bfd205
fix incorrect change at rev. 1.2
2002-04-04 20:00:33 +00:00
c701db92ac
Remove .cvsignore file.
...
<URL:http://www.netbsd.org/developers/cvs-repos/notes.html#cvsignore >
2002-04-04 17:08:24 +00:00
306d88f7e2
sync with gnu/dist/gcc/config/sh/sh.md rev. 1.7
...
> apply original gcc rev. 1.29:
>
> * Guard insn splits against illegal regsiters.
2002-04-04 15:39:14 +00:00
dd5f863fa8
sync with gnu/dist/gcc/config/sh/sh.md rev. 1.5
...
> apply gcc original rev. 1.34:
>
> * sh.c (block_lump_real): Add missing clobber of T_REG
2002-04-04 15:33:35 +00:00
a513ce40a3
sync with gnu/dist/gcc/config/sh/sh.md rev. 1.4
...
> gen_call() takes only two arguments (from egcs-current)
2002-04-04 15:30:47 +00:00
e05a3f8f0a
sync with gnu/dist/gcc/config/sh/sh.md rev. 1.2
...
> Sync *{mul,div}*i3 with egcs-current.
> Now we can compile mount_mfs (newfs) with -O2!
2002-04-04 15:27:36 +00:00
4999187555
sync with gnu/dist/gcc/combine.c rev. 1.5
...
> gcc original rev. 1.89:
> * combine.c (if_then_else_cond): If comparing against zero,
> just return thing being compared.
>
> gcc original rev. 1.132:
> * combine.c (combine_simplify_rtx): Don't create an if_then_else
> unless both args are general_operand.
2002-04-04 13:30:38 +00:00
8b3e2136ae
sync with gnu/dist/gcc/combine.c rev. 1.4,1.6
...
> apply gcc original rev. 1.116:
>
> > combine.c (simplify_comparison): Fix typo.
>
> LTU, LEU, GTU, "LEU" -> LTU, LEU, GTU, "GEU"
>
> apply gcc original rev. 1.198 (as rev. 1.116)
>
> > combine.c (simplify_comparison): Fix typo.
>
> GTU, LTU, GEU, "GEU" -> GTU, LTU, GEU, "LEU"
2002-04-04 13:23:37 +00:00
fb1d7cb64f
From Takeshi Shibagaki <ie9t-sbgk@asahi-net.or.jp>:
...
Define OBJECT_FMT_ELF for the benefit of libgcc et al.
2002-04-03 20:31:24 +00:00
eeac1cd146
Make sure the "-G num" option starts with a new paragraph instead of
...
flowwing on to the end of the previous option's description.
2002-04-03 15:14:15 +00:00
7667f320c2
Implement elf32_arm_object_p().
2002-04-03 02:59:13 +00:00
ef861642cb
#define HOST_BITS_PER_LONGLONG 64
2002-04-02 07:28:23 +00:00
c23f2ad0b9
* Use <elfos.h>, not <svr4.h>.
...
* Use <dbxelf.h>.
* Undef CPP_PREDEFINES before defining it.
* Remove the undef of HANDLE_SYSV_PRAGMA; we need this in order for
weak externs to work properly.
2002-04-02 06:22:28 +00:00
002142a2b5
* Don't use NetBSD's ELF headers, use those provided by BFD.
...
* Use bfd_{get,put}_{16,32,64}() rather than NetBSD's htobe32(), etc.
2002-04-01 21:58:09 +00:00
5269e780e1
Update DEFAULT_VECTOR and SELECT_VECS for arm--netbsdelf BFD changes.
2002-04-01 20:25:50 +00:00
fc6514a57b
* Add 2 new BFD targets, "elf32-littlearm-nbsd" and "elf32-bigarm-nbsd".
...
These targets are identical to the non-nbsd versions except that they
mark resulting ELF objects and executables as OSABI_NETBSD version 0.
* Add big-endian ARM BFD vectors to arm-*-netbsdelf. Make the -nbsd
BFD target the default for arm-*-netbsdelf.
2002-04-01 19:53:05 +00:00
2374dc24ce
Add a hack for VAX ELF: compile with -O1
2002-03-30 22:31:13 +00:00
1347ab1b90
Remove check for sparc64 which adds -msoft-quad-float as that is now the
...
default for the compiler on sparc64.
2002-03-28 09:58:09 +00:00
d3debc5e93
Pull changes from gcc3 back to properly emit calls for the _Qp* routines
...
correctly.
Enable -msoft-quad-float as the default moving forward. A complete build with
this compiler produces no errors and passes known regressions problems (awk,
etc).
XXX - -mhard-quad-float no longer will generate code correctly (internal
compiler problems). This is the case all the way to gcc-current so it's not
an easy fix at the moment but no one should be using hard quad math anyways.
2002-03-28 08:14:55 +00:00
4d42b0eae1
Add some EF_VAX flags. (not yet used)
2002-03-28 06:52:57 +00:00
cd2128de3e
Use "pushl $0" in preference to "clrl -(sp)" since the former is clearer
...
in its intent.
2002-03-28 06:51:36 +00:00
bb22d1e732
Instead of removing ARRAY_SIZE just #ifndef around it so patches can be
...
generated against the gcc-2.95.3 tag and this will compile. (as ARRAY_SIZE
isn't provided in libiberty included in gcc-2.95.3 itself
2002-03-27 04:09:03 +00:00
93fcfe16c5
Oops...don't check against &0x1 as the address is already even which is the
...
problem that was being solved in the first place
2002-03-27 00:00:16 +00:00
f3cbb9ef0a
Added ${G_CONFIGDIR} to .PATH. This is needed on m68000 so that
...
libgcc1 gets built.
2002-03-26 23:10:58 +00:00
ef51c9bc83
Back out DEFAULT_SHORT_ENUMS; it's actually causing some real problems,
...
and we need to come up with a better strategy.
Per discussion with Ben Harris.
2002-03-25 23:26:31 +00:00
8c184bd586
This file is long obsolete. Remove it.
2002-03-24 22:56:59 +00:00
41c1f3f785
MACHINE_ARCH can't be "arm26" any more, so don't even try to support it.
2002-03-24 22:24:41 +00:00
e00fcca843
MACHINE_ARCH can't be "arm26" any more.
2002-03-24 22:23:43 +00:00
2198b55512
Add arm/xm-netbsd.h
2002-03-24 18:25:55 +00:00
3c7ffdc2a5
Add arm/xm-netbsd.h which defines ONLY_INT_FIELDS, and use it for
...
NetBSD ARM ELF targets. This is necessary in order for the compiler
to build with packed enums.
2002-03-24 18:05:57 +00:00
371e95a59a
Fix sparc64 so locals and args can be accessed correctly.
...
If one doesn't subtract the stack bias back off %fp before attempting
to access variables stored relative to %fp the wrong address is picked up for
all locals and args. So, calls to the FRAME_LOCALS_ADDRESS or
FRAME_ARGS_ADDRESS macros need a function for sparc64 specific cases.
Provide a function to check for bias's addresses and fix them up.
2002-03-22 03:33:08 +00:00
99db91e2e0
Add m68000-specific `toolchain2netbsd' generated toolchain build/config files.
2002-03-22 00:07:45 +00:00
0e21b8d24d
Added a case for m680[01]0-*-netbsdelf.
2002-03-21 23:54:43 +00:00
7a6b4d460d
Pulled changes from binutils-current necessary for 68010
...
support. Specifically, got the changes to md_prepare_relax_scan
from binutils' revision 1.11 of this file. As a bonus also
now define DWARF2_LINE_MIN_INSN_LENGTH.
2002-03-21 23:53:14 +00:00
addff245d7
Remove the 'admin' category. This didn't turn out the way we
...
had planned, and users are confused what this category was for.
2002-03-21 03:04:00 +00:00
878f329747
Redefine RETURN_IN_MEMORY to match the ATPCS structure-return rules, rather
...
than the APCS ones. As with VFP-format doubles, this should really be done
at run time.
2002-03-19 20:59:22 +00:00
db0c1e532e
Default to the medium any memory model.
2002-03-19 18:12:27 +00:00
f9a48a09ad
Turn on DEFAULT_SHORT_ENUMS, since this will be required by the ARM EABI-2001,
...
which we'd like to comply with.
2002-03-18 20:36:51 +00:00
66bd2c38da
Fix an off-by-one error in elfcore_netbsd_get_lwpid() which caused the
...
returned LWP ID to always be 0, causing problems for creating
note pseudosections in multi-LWP processes.
2002-03-15 19:01:39 +00:00
ad911f7801
Oops, Tracy had already added evbmips and sbmips - delete the double-ups.
2002-03-13 04:11:41 +00:00
acda7e7437
Add port-{acorn32,evbmips,playstation2,sbmips}.
2002-03-13 04:08:54 +00:00
310cee1fde
added sbmips & evbmips
2002-03-13 04:04:22 +00:00
dca4ae94d6
When checking that a potentially-unsigned enum is >= 0, assign it to an int
...
first. This is necessary to avoid warnings with -fshort-enums. Casting
to an int really should be enough, but turns out not to be.
This change will be documented in doc/HACKS.
2002-03-09 13:22:52 +00:00
ee0c88c936
Also supply the FOP_REGNUM value in supply_xmmregs().
...
This fixes problem with getting return value of functions
in debugged programs, as noted in toolchain/15715.
Fix provided by Love <lha@stacken.kth.se> in
http://mail-index.netbsd.org/tech-toolchain/2002/01/17/0000.html
Reviewed by Jason Thorpe.
2002-03-07 19:14:44 +00:00
e44bd33760
version[] 5.0-nb1 --> 5.0nb1, as other archs use
2002-03-06 19:17:09 +00:00
872e77e22e
Remove some unnecessary lines.
2002-03-06 02:40:06 +00:00
60e3a276af
sync with FSF binutils:
...
- also accept "-maltivec"
- enable altivec by default
- add mtvrsave and mfvrsave
the latter two are required for an "altivec" GCC.
2002-03-04 13:35:42 +00:00
b9e5a44a38
Added port-mvmeppc
2002-03-02 22:25:15 +00:00
eff7936f87
include <string.h> as necessary.
2002-03-02 12:21:14 +00:00
32f6aaf031
Use sh to invoke move-if-change and don't depend on it being +x in the
...
repository
2002-03-01 06:06:50 +00:00
f6ce7f3293
Put the infodir-meta rule here (and wrap it) to allow it to be called from the
...
top level Makefile
2002-03-01 06:03:20 +00:00
7ce69c568a
Fix botches in Equivalence Classes paragraph.
2002-03-01 04:20:37 +00:00
a2ba138feb
Fix fallout from collapsing libiberty to MI config.h: libstdc++ includes
...
little bits of libiberty, so -I the non-arch specific directory now.
2002-02-28 02:13:34 +00:00
ac1299bdfb
New toolchain infrastructure for sh3eb, generated with mknative.
2002-02-27 18:45:14 +00:00
8d7b76f87a
Fix OBJS dependency.
2002-02-27 17:58:52 +00:00
6080b95ae4
Make libobjc configury MI.
2002-02-27 17:55:32 +00:00
ab30e43870
Add .MADE to prevent building in $srcdir. (They could have at least
...
referenced $(srcdir) explicitly for the targets, sigh....)
2002-02-27 17:34:48 +00:00
b69d8cc8b2
New toolchain framework for sh3el platforms, generated with mknative.
2002-02-27 17:10:09 +00:00
bfa4c6d7c2
Make new toolchain conf files for libiberty MI.
2002-02-27 17:03:49 +00:00
9991b60633
Remove R_VAX_GOT32O and R_VAX_PLT32O relocs. VAX don't need 'em.
2002-02-27 07:43:34 +00:00
88a892b978
Superseded by src/tools/toolchain/mknative.
2002-02-26 22:32:21 +00:00
3cf9139573
When configured via mknative, G_HOST_* filenames are prefixed with "l-".
...
Strip the prefix.
2002-02-26 22:30:46 +00:00
66c4fd196e
Make hconfig.h common as simply a redirect to config.h; these are
...
identical on a native build. (Simplifies mknative for generating the
arch-dependent bits.)
2002-02-26 20:56:29 +00:00
80c7fa68d5
Use MACHINE_GNU_PLATFORM to identify the machine.
2002-02-26 20:45:34 +00:00
6b83b4af8c
If doing a Canadian cross and the creation of auto-build.h fails, cause
...
the outer "configure" run to fail, too.
2002-02-26 19:17:51 +00:00
a688a18a3f
Tack "nb1" onto VERSION here, too, to sync with "configure".
2002-02-26 18:38:38 +00:00
13dd6cb726
Skip the "misc" directory on recursion. (This contains .el files that
...
will attempt to install into never-never land if emacs isn't installed.
GNU bogon.)
2002-02-26 18:05:20 +00:00
e95d648871
Work with a mkinstalldirs that is not +x.
2002-02-26 18:00:29 +00:00
853b3d3c1a
Workaround some *2netbsd script not preserving the execute bit on various
...
script files (and now that they are in cvs, they can't be fixed).
Future imports of any package should use "-ko" and do away with *2netbsd
altogether, as was the intent of gnu/dist in the first place.
2002-02-26 17:40:52 +00:00
d8d7046773
Switch REAL_ARITHMETIC on (needed for DWARF support).
2002-02-25 17:33:02 +00:00
b5355576f3
Change TARGET_NAME to use ${GNU_ARCH_PLATFORM} instead of
...
${GNU_ARCH_MACHINE}-netbsd. This allows platforms to that were
formerly a.out but ELF to be foo--netbsdelf. It also adds the
missing 2nd "-" which was missing in the former definition.
2002-02-24 22:15:39 +00:00
796a42b608
Only ELF has S_IS_WEAK.
2002-02-24 19:50:14 +00:00
cce4db07ab
Add a missing register prefix.
2002-02-24 01:11:40 +00:00
4ce78967e6
Make some assembler warnings fatal errors. don't allow 4 char register
...
name in OBJ_AOUT mode.
2002-02-24 01:10:31 +00:00
c7ce362189
Teach old (a.out) vax gas about register prefixes.
2002-02-23 21:16:41 +00:00
9fe6d7dc5a
Dump a few comments that don't apply anymore. Remove dead code that is no
...
longer used. Add register prefix support. It's always on for ELF and
optional for a.out (it's recognize register names with or without the %
prefix).
2002-02-23 21:07:29 +00:00
572f2bb969
Teach gcc to emit register prefixes (though unless its ELF, the register
...
prefix is NULL).
2002-02-23 21:02:49 +00:00
bc49a6eede
Fix long time codegen bug. Only the COMPARE, ADD, MINUS operations actually
...
set the C(arry) bit appropriately. All other leave it in an indeterminate
(to GCC) state. Mark that by setting CC_NO_OVERFLOW. Change vax.md emission
of branches that use the C bit to use OUTPUT_JUMP so that if CC_NO_OVERFLOW is
set, branches that would normally use the C bit use opcodes that don't use the
C bit (jgtru -> jneq && jlssu -> jeql). Delete the cmpdi pattern.
2002-02-23 18:36:52 +00:00
4da241dfa0
Whitespace nits.
2002-02-23 17:01:20 +00:00
2365d7ec78
Add a '-s' option which will cause mdsetimage to write back the
...
actual image size copied back into the kernel.
2002-02-23 02:30:37 +00:00
13828b4c70
Implement a cmpdi pattern so gcc won't try to help us. The good news is
...
that's it's smaller. Not sure about faster.
2002-02-22 06:03:36 +00:00
6238aa6c4f
Back out change to NOTICE_CC_UPDATE. It has really bad side effects.
...
I'll fix this another way.
2002-02-22 06:02:33 +00:00
eb2926091c
movl doesn't modify the C bit in the PSL so make sure to communicate that
...
fact to the gcc optimizer so it doesn't emit the wrong branch instructions.
2002-02-21 19:31:36 +00:00
90b4581807
Ahhh! Use tc_fix_adjustable to make gas not resolve symbols into segment
...
based references for weak or external symbols.
2002-02-21 07:30:32 +00:00
1b65728579
Gegneralize the PLT/GOT processing to properly emit relocs even for symbols
...
in the segment/file so that they linker can resolve them for shared libraries.
2002-02-21 06:59:16 +00:00
b32effcb8e
Add some more dependencies so this builds as "make all" also without
...
requiring a make depend first.
2002-02-21 04:13:45 +00:00
71bd5d8f3c
make this build cross by not splitting NATDEPFILES across a line with a
...
line continuation char (\)...
2002-02-21 03:24:50 +00:00
ffefe31a4b
When calling weak or global symbols in the same segment, make sure to use
...
RELAX encoding what can be transmorgified into a PLT reloc.
2002-02-21 00:18:43 +00:00
1bb7ff7e99
Don't emit GOT slots for local symbols since the VAX can always refer to
...
via PC relative addressing.
2002-02-20 18:52:23 +00:00
e9ecffcaf2
When emitting a GOT reloc, use the frag symbol since we want the
...
external name, not where it lives in this object (which would happen
in the symbol was defined locally). All, don't use offset but addnumber.
2002-02-20 18:50:40 +00:00
9fcb3b5a86
If the symbol is weak or external and this is a PCREL reference,
...
generate a fragment instead of fixup. This allows us to convert
to a GOT or PLT reference which can be resolved externally.
2002-02-20 08:12:43 +00:00
1927769e7a
Don't use PCC_STATIC_STRUCT_RETURN.
...
Use movab/pushab when moving LABEL_REFs.
2002-02-20 03:06:37 +00:00
7efea97070
Make sure that non-symbolic GOT references are converted back to
...
PC32 relocs for -Bstatic or -Bsymbolic links.
From Matt Thomas.
2002-02-19 06:16:35 +00:00
b669a91187
Make sure _initialize_vaxnbsd_nat() is called (added manually).
2002-02-19 05:05:38 +00:00
b56d2fc557
Add support for ELF core files.
2002-02-19 05:05:09 +00:00
a962d1ba86
Do not -Dunix or -Darm, as they are namespace-invasive, and cause
...
variants of those CPP symbols to be defined that are contrary to
NetBSD's conventions.
Fixes building Xaw on ARM ELF.
2002-02-12 16:32:39 +00:00
92140a7683
Skip building if the target's .mk file is missing.
2002-02-11 21:45:18 +00:00
1c986fdf46
Skip building anything if the target's .mk files are missing. (Allows the
...
toolchain to be built in segments, or not at all if the in-tree toolchain
is inappropriate for the target.)
2002-02-11 21:36:33 +00:00
msaitoh