4170f63662
Handle arbitrarily long input lines.
...
(Patch taken from gas.new.)
2000-01-24 19:44:37 +00:00
1b4eebafa7
One more .mk bug.
2000-01-24 07:27:15 +00:00
3c08240686
Use mkstemp().
2000-01-24 02:47:51 +00:00
4ec77e0d35
Use mkstemp(3).
2000-01-24 02:03:37 +00:00
fa450d541e
More .mk file fallout. *sigh*
2000-01-23 22:19:27 +00:00
615ac255bd
More .mk file fallout. *sigh*
2000-01-23 22:10:52 +00:00
12b34d4ffe
Only skip some bytes if the associated object is not a dir.
...
Needed for some tar files.
Fixes PR 9274 by Giles Lean <giles@nemeton.com.au>
2000-01-22 14:04:15 +00:00
f6f389921a
#define POSIX
2000-01-19 03:57:17 +00:00
6ed33701a8
fixed typo in example (fixes PR 9023)
2000-01-17 12:44:39 +00:00
493acb76cb
More alloca() -> malloc(). Needed to make some shared libraries link when
...
compiled with -g.
2000-01-13 00:05:32 +00:00
b9db4a8140
Add rudimentry VAX COFF support. enough so objdump, size, nm, etc. are happy
...
but no near enough for gas or ld.
2000-01-11 05:48:27 +00:00
9b25bfd178
Semi-random code cleanup.
...
Also, in the RRS_PARTIAL case, go ahead and allocate a relocation entry if the
bind failed. (Who knows? Maybe some day we'll allow relinking an executable.)
2000-01-07 21:19:36 +00:00
5a63d68e02
Fix two problems:
...
* On other systems, `-Bsymbolic' does not affect undefined (external)
symbols at all. We were generating invalid PLT slots in this case.
* Do not prebind GOT and PLT slots when `-r' is used; only do so if
-Bsymbolic is used {or if we're in RRS_PARTIAL mode, but I'm not
sure that's correct}. Otherwise, we could inadvertantly prebind
symbols when using `-r' with PIC files; when the resulting .o file
is then linked into a shared library, the library would not permit
an executable to override its symbols as it's supposed to.
2000-01-07 20:23:41 +00:00
f59fc24c1e
Made libbfd read and understand our a.out dynamic symbols.
...
Solves PR bin/7576 from Mike Neuman.
2000-01-06 21:50:05 +00:00
d64461f26b
apply some patches from egcs-current (to make gcc more reliable)
2000-01-06 19:20:35 +00:00
5f388312a6
Fix Y2K bug in date parser reported by Trevin Beattie in PR y2k/9123.
2000-01-05 21:34:11 +00:00
47956fff3f
skip gdb if ${MACHINE_ARCH} == "sh3"
2000-01-02 11:20:51 +00:00
6bba0f3f4a
pass arg 4 of the qsort() correctly.
2000-01-02 11:01:02 +00:00
fef1139c8c
BFD for SH ELF
2000-01-02 10:53:52 +00:00
c1d02ec5a3
Made N_TXTADDR do the righ thing with shared libs.
2000-01-01 20:52:57 +00:00
0845ad5a3c
gas.new for NetBSD/sh3
1999-12-26 20:03:56 +00:00
5efb71755d
ld.new for NetBSD/sh3
1999-12-26 19:59:22 +00:00
bec61173c4
libbfd for NetBSD/sh3
1999-12-26 19:48:00 +00:00
bc2c2462e1
one too many backslashes in an nroff expression, fixed per PR 8724
1999-12-20 23:01:20 +00:00
f777a401db
Accept %fp as an alias for %a6 (prefix *MUST* be present for this alias
...
to work).
1999-12-14 17:05:13 +00:00
a5f38cc443
Add news68k
1999-12-10 06:04:01 +00:00
5985528d02
Remove a few redundant splbio() calls (ACQUIRE_LOCK already does splbio())
1999-12-05 20:34:40 +00:00
4a9e81df67
Pass in the appropriate machine directive to the assembler if an -mcpu=...
...
option is given to the compiler. It's silly to have to specify it twice
on a command line (e.g. -mcpu=ev56 -Wa,-mev56), especially considering
that if you don't, and the compiler emits e.g. a BWX instruction, the
assembler will treat it as a macro, open-coding an equivalent, which can
have some serious unwanted side-effects in some situations.
Also, don't treat e.g. 21164 and ev5 as equivalent; the assembler treats
them differently: 21164 enables PALcode-only instructions, while ev5
does not.
1999-12-03 06:37:47 +00:00
aca0169f10
open(2) can legally return 0. Errors only if return value < 0.
...
(Test for good is >= 0.)
1999-12-01 03:45:54 +00:00
b18e309569
Allow building of PIC or profiling libraries to be disabled.
1999-11-27 23:46:09 +00:00
414bf126c6
add TDEPFILES.sh3
...
TODO: add NATDEPFILES.sh3
1999-11-26 14:39:47 +00:00
61dc0599cd
gdb for SH (doesn't work yet)
1999-11-26 09:13:36 +00:00
f0af9ff8e6
add "system.h"
1999-11-26 07:55:14 +00:00
3954eb11cb
s/sparc/sh/
1999-11-26 07:36:48 +00:00
ce8da2329e
Use splbio() in getdirtybuf() to prevent bad things from happening due
...
to queue race conditions. Don't use B_VFLUSH as it's not appropriate for
this case.
1999-11-24 23:13:15 +00:00
d901f6eae0
Be more careful to block bio interrupts for some data structures. There
...
were at least a few missed cases where vp->v_{clean,dirty}blkhd were
unprotected since the softdep/trickle sync merge.
1999-11-23 23:52:40 +00:00
6a4a304da0
Add rs6000coff_vec, which adds aixcoff-rs6000 support. Used with fixcoff.
1999-11-23 02:08:03 +00:00
7e4423b4ea
Set the limit at which flushing is sped up to prevent excessive memory
...
usage a little lower.
1999-11-18 12:20:45 +00:00
0616ece265
add "sh3"
1999-11-17 09:59:09 +00:00
37cdd0b801
simply the conditional for ld.
1999-11-17 02:52:29 +00:00
ea314e139b
Add the parts of the softdep code that fall under Kirk's more GPL-like
...
copyright here. To be optionally included in the kernel using the
prefix and cinclude mechanism.
1999-11-15 18:42:26 +00:00
79db3ed28c
add hpcmips
1999-11-13 06:57:08 +00:00
ef9b07e056
Add register prefixes.
1999-11-10 16:54:10 +00:00
91d9e85847
Add '\n' after error messages.
...
Problem reported in PR 8773 by Johan Danielsson <joda@pdc.kth.se>
1999-11-10 14:53:39 +00:00
e3ff78dac0
Hack in (quite brutally, I might add) a flag to warn about missing
...
register prefixes. Suggested by Ignatios Souvatzis.
1999-11-09 23:38:57 +00:00
fea5b77bc6
don't bother calling fwrite() if count == 0' ld.c
1999-11-08 21:58:00 +00:00
51515f302b
Fix bogus section header. From PR 8735.
1999-11-03 21:45:57 +00:00
377945b324
change the example values of Confidential field from "yes | no" to "no | yes"
...
so that commonly used value is first and it's less probable somebody
would use "yes" accidentaly
1999-10-31 18:19:10 +00:00
cbb06aec98
- document the environment variables used by send-pr to fill in defaults
...
for problem reports. See the ENVIRONMENT section of the manpage.
- document the files used by send-pr to fill in defaults for problem
reports. See the FILES section of the manpage.
This addresses PR misc/1889.
1999-10-30 17:02:43 +00:00
16dcb3a8fd
more radical fix for the last problem, from the egcs/gcc mailing list
...
(http://egcs.cygnus.com/ml/gcc/1999-10/msg00400.html ).
Log message:
* combine.c (get_last_value): If the last set of a register
is after subst_low_cuid, then we can not use it to determine
the register's last value.
1999-10-25 19:30:19 +00:00
mycroft