d6886078f2
Don't define MAXFLOAT if __NetBSD__ is defined. Mark this XXX. Hopefully,
...
we won't have to worry about it when libg++ is upgraded. This is necessary
for libg++ to compile with -Werror.
1995-12-16 01:03:33 +00:00
f5c9632c68
Sync with change to arch/mips/xm-netbsd.h.
1995-12-15 08:05:11 +00:00
4fec4fc912
Changing "..." to <...> for includes.
1995-12-15 08:03:52 +00:00
b2b08219f3
Vax now uses src/gnu/usr.bin/gcc.
1995-12-07 07:26:32 +00:00
f5d344126b
Have the ns32k architecure use gcc instead of gcc2.
1995-12-03 21:13:18 +00:00
34527405c3
Import gcc-2.7.2. Since it is in the gcc directory instead of the gcc2
...
directory, this is being done now. We will live with two trees until
the "formal" switch over by changing src/gnu/usr.bin/Makefile.
1995-12-01 17:58:53 +00:00
35c8aad7b3
Flush after diddling text, so we get predictable results on machines
...
with separate I & D caches.
1995-11-28 19:29:45 +00:00
9c5e4ba959
extract gecos field from passwd in a more efficent manner (one sed instead of grep, awk, and a sed)
1995-11-23 02:10:43 +00:00
ca4dc54bcd
Handle '&' in gecos field
1995-11-23 02:07:34 +00:00
1f061d8392
Initialize all registers (from John Kohl; PR#1651).
...
The REGISTER_U_ADDR is no longer needed.
1995-11-22 00:25:50 +00:00
908c02bfc8
LIBMATH -> LIBM; (PR #1689 )
1995-10-26 00:58:49 +00:00
2a60e37a76
Link zcmp to zdiff, zcmp.1 to zdiff.1
1995-10-24 22:18:34 +00:00
8c9912db33
Parallel make fixes
1995-10-22 23:20:24 +00:00
a3e3b27287
Parallel make fixes
1995-10-22 23:02:05 +00:00
89c5e30a44
Parallel make fixes
1995-10-22 21:58:21 +00:00
99fc0f40d9
Parallel make fixes
1995-10-22 21:04:42 +00:00
fd3fd7642f
Parallel make Makefile fixes.
1995-10-22 00:21:52 +00:00
6ac54c7a78
Vax support for ld, so we (at least) can link static binaries.
1995-10-19 13:10:17 +00:00
0f25b8485c
Don't try to compile gcc on vax.
1995-10-19 12:55:37 +00:00
9c21bb769d
Kill spaces.
1995-10-15 13:12:18 +00:00
6a91c699b4
We can read %fs and %gs now.
1995-10-11 05:24:49 +00:00
b63cd8a07d
Fix a memory leak in writing numbers.
1995-10-10 00:22:54 +00:00
d04fcadcc2
Fix typo.
1995-10-09 09:24:59 +00:00
ce77d757fd
And initialize those local variables...
1995-10-09 00:14:41 +00:00
9d43c41208
New style RCS IDs.
1995-10-09 00:09:39 +00:00
b74b1d6456
Allow customized trace output, so ldd(1) can be more conveniently used
...
as a filter.
Rearrange cascading stuff: remove two-level limit.
1995-10-08 23:43:28 +00:00
3aa94e17f0
Add -f option for customized output.
1995-10-08 23:39:58 +00:00
9c06f30b7f
Add port-atari as a valid category.
1995-09-30 12:44:39 +00:00
6fbb9d434b
Update version number to match reality
1995-09-27 22:49:28 +00:00
d5a07be95e
Fix typo in example; PR #1534
1995-09-27 19:10:49 +00:00
6ad6202b57
ANSI C++ says that typedefs must specify a type (i.e. can't just
...
default 'int'). One here didn't.
1995-09-27 14:03:23 +00:00
3403044148
consistently declare functions inline or volatile.
1995-09-27 14:02:02 +00:00
70075b5f87
if __NetBSD__ defined, don't prototype strsignal() -- string.h already
...
provides a prototype, and it's different from what was here!
(proto here was const char *..., proto in string.h was just char *...)
1995-09-27 14:01:01 +00:00
6c9bbdc678
add -fno-for-scope to CXXFLAGS on the alpha. Really, this should be
...
conditional on a gcc >= 2.7.0, but there's no good and easy way to do
that short of introducing a new global 'make' variable, which i think
is a bad idea.
1995-09-27 13:55:19 +00:00
6fda0c28d0
alpha can build groff now.
1995-09-27 12:46:07 +00:00
5530d8edb4
Avoid trailing garbage (from Mike Long; PR#1527).
1995-09-27 09:43:21 +00:00
f9d6b3823a
Improvements by Matthias Pfaller (fpu regs ...)
1995-09-26 20:23:36 +00:00
941f24a3cc
Don't make or install cc1 on the pc532 until we upgrade past 2.4.5.
1995-09-24 06:50:41 +00:00
b19a7d87e2
Pass on return value.
1995-09-23 22:48:18 +00:00
d4fe561742
Interface version 4:
...
implement dlexit() (per Doug Rabson)
get actual rtld path from crt0
pass ld_entry back to crt0 in a better place
1995-09-23 22:46:24 +00:00
fa6e06725c
Define `dlexit()' in crt<=>rtld interface (active in v4).
1995-09-23 22:26:15 +00:00
2bb25d0f0f
Make stack-trace in gdb -k work on m68k because the pc-value is now
...
correctly setup. Also fix the loads of warnings about the second arg
of supply_register().
1995-09-16 19:45:03 +00:00
f5c3480dee
Substitute SUBMITTER instead of DEFAULT_SUBMITTER when generating send-pr.el
1995-09-12 19:01:51 +00:00
dc758c27aa
Handle case where mail-default-reply-to == t
1995-09-07 00:21:14 +00:00
1a9159c7ba
Check for zero sized BSS to avoid a big no-op syscall (per Mike Hibler).
1995-08-31 22:07:25 +00:00
9d398c5d3a
Update batch comparison so that pr's can be sent with send-pr.el. PR #1077
1995-08-31 19:06:23 +00:00
e9b2e388aa
Fix typo. Use DEFAULT_CLASS instead of DEFAULT_CONFIDENTIAL for Class
1995-08-31 18:59:58 +00:00
abd78a8514
Use SUBMITTER instead of DEFAULT_SUBMITTER in send-pr.el generation
1995-08-31 18:58:40 +00:00
98150bc36d
Upgraded to send-pr 3.95, with our local modifications
1995-08-31 18:48:23 +00:00
288f1c084f
Changes from Matthias Pfaller to get gdb to work.
1995-08-29 08:03:27 +00:00
thorpej