]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
26 years agoWell, nobody objected, so here's my -u (unlink) flag to restore.
Jordan K. Hubbard [Sat, 9 May 1998 05:23:02 +0000 (05:23 +0000)]
Well, nobody objected, so here's my -u (unlink) flag to restore.

26 years agoFix the problem people are having building -current on a -stable system.
John Birrell [Fri, 8 May 1998 21:29:35 +0000 (21:29 +0000)]
Fix the problem people are having building -current on a -stable system.
The headers that are installed in WORLDTMP are part of the interface
that includes libraries like libc, so they must be installed together.
This means that lib-tools and build-tools should be merged. The FreeBSD
build only works in hosted form where it is assumed that the installed
version contains adequate tools to build the latest release.

26 years agoAdd jetdirect 9100/tcp
Poul-Henning Kamp [Fri, 8 May 1998 20:40:15 +0000 (20:40 +0000)]
Add jetdirect 9100/tcp

PR: 5195
Reviewed by: phk
Submitted by: Cy Schubert <cschuber@uumail.gov.bc.ca>

26 years agoAdd a crucial missing _ in NO_SENDMAIL
Poul-Henning Kamp [Fri, 8 May 1998 20:15:39 +0000 (20:15 +0000)]
Add a crucial missing _ in NO_SENDMAIL

PR: 6545
Reviewed by: phk
Submitted by: Jos Backus <jbackus@plex.nl>

26 years agoUnblock SIGPIPE before execv(); convert from sigvec to sigaction
Guido van Rooij [Fri, 8 May 1998 19:15:44 +0000 (19:15 +0000)]
Unblock SIGPIPE before execv(); convert from sigvec to sigaction
Obtained from: OpenBSD

26 years agoIn the SEE ALSO section, {get,set}rlimit are said to be in the section
Masafumi Max NAKANE [Fri, 8 May 1998 14:46:23 +0000 (14:46 +0000)]
In the SEE ALSO section, {get,set}rlimit are said to be in the section
3 which are actually in the section 2.
Pointed out by: Brian Reichert <reichert@numachi.com>
PR: 6480

26 years agoTry experimenting with the split size a little in an effort to
Jordan K. Hubbard [Fri, 8 May 1998 13:49:16 +0000 (13:49 +0000)]
Try experimenting with the split size a little in an effort to
move our installation structure a little further into the 20th century
(just before it's over).
Instigated by: Dag-Erling Coidan <dag-erli@ifi.uio.no>

26 years agoTranslated to C (parameters in a function definition have the same scope
Bruce Evans [Fri, 8 May 1998 07:56:48 +0000 (07:56 +0000)]
Translated to C (parameters in a function definition have the same scope
as variables declared in the main block in the function, so shadowing
of parameters by variables declared in the main block is not just an
obfuscation).

Found by: lint

26 years agoSimplified by using new yacc rules.
Bruce Evans [Fri, 8 May 1998 06:43:07 +0000 (06:43 +0000)]
Simplified by using new yacc rules.

26 years agoSimplified by using new yacc rules and by not generating y.tab.h.
Bruce Evans [Fri, 8 May 1998 06:41:06 +0000 (06:41 +0000)]
Simplified by using new yacc rules and by not generating y.tab.h.

26 years agoMoved lex and yacc source handling from bsd.prog.mk to bsd.dep.mk
Bruce Evans [Fri, 8 May 1998 06:31:05 +0000 (06:31 +0000)]
Moved lex and yacc source handling from bsd.prog.mk to bsd.dep.mk
so that it applies to libraries (and lkms) as well as programs.

26 years agoThis is a hack to workaround source that is coded to use long variables
John Birrell [Fri, 8 May 1998 05:41:57 +0000 (05:41 +0000)]
This is a hack to workaround source that is coded to use long variables
but also assumes that they are 32-bits. This is one place where I don't
think it is appropriate to change 'long' to 'int'. I don't see why the
code couldn't be fixed so that using natural long variables does the
right thing. It's spaggetti code so it'll take some effort. Obviously
NetBSD thought so too because they change 'long' to 'int32_t' etc
and left it at that. As a temporary measure FreeBSD/Alpha can use the
NetBSD code and put this on the list of things to fix.

26 years agoRemote the NetBSD kludge for vfprintf.c
John Birrell [Fri, 8 May 1998 05:17:11 +0000 (05:17 +0000)]
Remote the NetBSD kludge for vfprintf.c

26 years agoDon't assign the va_list variable 'ap' directly to the argtable because
John Birrell [Fri, 8 May 1998 05:10:32 +0000 (05:10 +0000)]
Don't assign the va_list variable 'ap' directly to the argtable because
va_list is not a pointer on alpha. Instead, use the va_arg() macro
to return the address that is stored in the argtable.

26 years agoAdd new named configuration template and remove old template
Andrey A. Chernov [Thu, 7 May 1998 23:42:33 +0000 (23:42 +0000)]
Add new named configuration template and remove old template

26 years agoRedo tcpmux stuff. tcpmux handling is now performed after inetd
Guido van Rooij [Thu, 7 May 1998 18:32:00 +0000 (18:32 +0000)]
Redo tcpmux stuff. tcpmux handling is now performed after inetd
forks. Furthermore, invalid input for tcpmux does not lead to
an exiting inetd.
This patch is recommended for people running tcpmux (which is NOT
enabled by default)

26 years agofix a core dump problem when you don't give -f a parameter... also compare
John-Mark Gurney [Thu, 7 May 1998 14:13:16 +0000 (14:13 +0000)]
fix a core dump problem when you don't give -f a parameter... also compare
results of strcmp against 0 instead of !'ing results...

Closes PR: 6538

26 years agoAn effort to make SLICE/DEVFS play with SCSI.
Julian Elischer [Thu, 7 May 1998 12:13:47 +0000 (12:13 +0000)]
An effort to make SLICE/DEVFS play with SCSI.

Apparently I didn't make my plans to make dev_t and devsw[] go away
under DEVFS quite clear enough to Peter Dufault as he stitched the SCSI
system together using them when he redid the configuration side of things.
This made is rather an effort to remove all vestiges of dev_t and
devsw[] entries from sd.c in DEVFS/SLICE mode.

26 years agoSync with sys/i386/isa/wd.c revision 1.166.
KATO Takenori [Thu, 7 May 1998 08:37:27 +0000 (08:37 +0000)]
Sync with sys/i386/isa/wd.c revision 1.166.

26 years agoSync with sys/i386/isa/fd.c revision 1.112.
KATO Takenori [Thu, 7 May 1998 08:36:48 +0000 (08:36 +0000)]
Sync with sys/i386/isa/fd.c revision 1.112.

26 years agoIn the words of the submitter:
Mike Smith [Thu, 7 May 1998 04:58:58 +0000 (04:58 +0000)]
In the words of the submitter:

---------
Make callers of namei() responsible for releasing references or locks
instead of having the underlying filesystems do it.  This eliminates
redundancy in all terminal filesystems and makes it possible for stacked
transport layers such as umapfs or nullfs to operate correctly.

Quality testing was done with testvn, and lat_fs from the lmbench suite.

Some NFS client testing courtesy of Patrik Kudo.

vop_mknod and vop_symlink still release the returned vpp.  vop_rename
still releases 4 vnode arguments before it returns.  These remaining cases
will be corrected in the next set of patches.
---------

Submitted by: Michael Hancock <michaelh@cet.co.jp>

26 years agocleanup: remove duplicated dump code.
Julian Elischer [Thu, 7 May 1998 02:05:21 +0000 (02:05 +0000)]
cleanup: remove duplicated dump code.
fix oversight on SLICE ioctl code.

26 years agocleanup:
Julian Elischer [Thu, 7 May 1998 01:15:23 +0000 (01:15 +0000)]
cleanup:
take out duplicated dump code

26 years agoUnbotch whitespace and change switch with two cases to if.
Eivind Eklund [Thu, 7 May 1998 00:42:25 +0000 (00:42 +0000)]
Unbotch whitespace and change switch with two cases to if.

Noticed by: bde

26 years agoMake syslogd function in non-secure mode.
Brian Somers [Thu, 7 May 1998 00:39:56 +0000 (00:39 +0000)]
Make syslogd function in non-secure mode.
Log the correct ^2 packet as per the last commit comment.

26 years agoThere is no dump routine for fd so give it an explicit NULL entry
Julian Elischer [Wed, 6 May 1998 23:35:36 +0000 (23:35 +0000)]
There is no dump routine for fd so give it an explicit NULL entry
in the SLICE table.

26 years agoDon't call a dump routine if there isn't one. (e.g. in floppy driver)
Julian Elischer [Wed, 6 May 1998 23:32:48 +0000 (23:32 +0000)]
Don't call a dump routine if there isn't one. (e.g. in floppy driver)

26 years agoAdd dump support to the DEVFS/slice code.
Julian Elischer [Wed, 6 May 1998 22:14:48 +0000 (22:14 +0000)]
Add dump support to the DEVFS/slice code.
now we can actually catch our crashes :-)

Submitted by: Luoqi Chen <luoqi@chen.ml.org> (the man who's everywhere)

26 years agoOnly include bsd.dep.mk in `mk'files that handle C sources. Abuse
Bruce Evans [Wed, 6 May 1998 18:44:04 +0000 (18:44 +0000)]
Only include bsd.dep.mk in `mk'files that handle C sources.  Abuse
bsd.obj.mk instead of bsd.dep.mk for defining the _SUBDIR target
and a default tags target.  Abuse bsd.obj.mk for defining default
cleandepend and depend targets.

26 years agoOverlooked, that newer naming convention is xxx_program instead of xxx_prog.
Andreas Klemm [Wed, 6 May 1998 17:36:16 +0000 (17:36 +0000)]
Overlooked, that newer naming convention is xxx_program instead of xxx_prog.
So changed it to ntpdate_program and xntpd_program.
Backout last change, now we have again named_program, sorry.

26 years agodocument ntpdate_prog and xntpd_prog
Andreas Klemm [Wed, 6 May 1998 17:26:48 +0000 (17:26 +0000)]
document ntpdate_prog and xntpd_prog
fix: named_program -> named_prog to reflect "real life"
(I suggest keeping this shorter name convention as it is
actually used in the config file)

26 years agoDelete stale comment.
Wolfram Schneider [Wed, 6 May 1998 16:53:53 +0000 (16:53 +0000)]
Delete stale comment.
Pointed out by: bde

26 years agoBacked out previous backout. Put y.tab.h back in SRCS.
Bruce Evans [Wed, 6 May 1998 15:33:59 +0000 (15:33 +0000)]
Backed out previous backout.  Put y.tab.h back in SRCS.

26 years agoBacked out previous backout. Put y.tab.h back in SRCS.
Bruce Evans [Wed, 6 May 1998 15:27:57 +0000 (15:27 +0000)]
Backed out previous backout.  Put y.tab.h back in SRCS.

26 years agoBacked out previous backout.
Bruce Evans [Wed, 6 May 1998 15:23:53 +0000 (15:23 +0000)]
Backed out previous backout.

Added y.tab.h to SRCS.  Dependencies on it were missing until recently.

26 years agoRegress to generating foo.c from foo.y via y.tab.c for crufty applications
Bruce Evans [Wed, 6 May 1998 15:01:18 +0000 (15:01 +0000)]
Regress to generating foo.c from foo.y via y.tab.c for crufty applications
that want a y.tab.h file.  This want must be specified by putting y.tab.h
in SRCS (and defaulting to or putting -d in YFLAGS).  This only works if
there is only one yacc parser, of course.  One improvement:  copy y.tab.c
to foo.c instead of renaming it, so that `#line...y.tab.c' statements in
it refer to an existing file.

Regress to not generating explicit rules for .l and .y sources containing
slashes.  This case is unusual and hard to handle properly.

Don't generate an unused dependency when -d is not in YFLAGS.

26 years agoUse `foo/bar.a' instead of `-Lfoo -lbar' for linking to static internal
Bruce Evans [Wed, 6 May 1998 13:23:39 +0000 (13:23 +0000)]
Use `foo/bar.a' instead of `-Lfoo -lbar' for linking to static internal
libraries, so that `ld -f' in can create correct dependencies for
yet-to-be-built libraries.

Use `DIR!= cd ...libbind; make -V .OBJDIR' to find libbind's object dir
if it doesn't seem to be in its usual place relative to ${.OBJDIR}.
This fixes `cd /usr/src/usr.sbin/nslookup; mkdir obj; make'.

26 years agoXFree86 3.3.1 -> 3.3.2
Jordan K. Hubbard [Wed, 6 May 1998 12:01:20 +0000 (12:01 +0000)]
XFree86 3.3.1 -> 3.3.2

26 years agodevfs_add_devswf(9) and devfs_link(9) don't tell you how to
Poul-Henning Kamp [Wed, 6 May 1998 09:13:53 +0000 (09:13 +0000)]
devfs_add_devswf(9) and devfs_link(9) don't tell you how to
remove the device once it's added or linked.

PR: 6524
Reviewed by: phk
Submitted by: Brian Cully <shmit@kublai.com>

26 years agoThe CHIOGSTATUS ioctl of the SCSI media changer driver (sys/scsi/ch.c)
Poul-Henning Kamp [Wed, 6 May 1998 09:11:20 +0000 (09:11 +0000)]
The CHIOGSTATUS ioctl of the SCSI media changer driver (sys/scsi/ch.c)
is broken.  It omits the SCSI_DATA_IN flag in the SCSI READ ELEMENT
STATUS command, which makes the 'chio status' command fail.

PR: 6528
Reviewed by: phk
Submitted by: Hans Huebner <hans@artcom.de>

26 years agoSync with sys/i386/isa/wd.c revision 1.164.
KATO Takenori [Wed, 6 May 1998 08:25:58 +0000 (08:25 +0000)]
Sync with sys/i386/isa/wd.c revision 1.164.

26 years agoCorrect use of .Nm and other cosmetics. Add rcsid. Remove unused #inc.
Philippe Charnier [Wed, 6 May 1998 06:51:42 +0000 (06:51 +0000)]
Correct use of .Nm and other cosmetics. Add rcsid. Remove unused #inc.

26 years agoCorrect use of .Nm, .An and other cosmetics. Add rcsid. Remove unused #inc.
Philippe Charnier [Wed, 6 May 1998 06:49:16 +0000 (06:49 +0000)]
Correct use of .Nm, .An and other cosmetics. Add rcsid. Remove unused #inc.

26 years agoUpdates to support adding driver files outside the i386 area.
Mike Smith [Wed, 6 May 1998 05:45:17 +0000 (05:45 +0000)]
Updates to support adding driver files outside the i386 area.
Submitted by: Daniel O'Connor <doconnor@gsoft.com.au>

26 years agoAs described by the submitter:
Mike Smith [Wed, 6 May 1998 05:29:41 +0000 (05:29 +0000)]
As described by the submitter:

Reverse the VFS_VRELE patch.  Reference counting of vnodes does not need
to be done per-fs.  I noticed this while fixing vfs layering violations.
Doing reference counting in generic code is also the preference cited by
John Heidemann in recent discussions with him.

The implementation of alternative vnode management per-fs is still a valid
requirement for some filesystems but will be revisited sometime later,
most likely using a different framework.

Submitted by: Michael Hancock <michaelh@cet.co.jp>

26 years agoImplement bus_dmamem_* functions and correct a few nits reported by Peter Wemm.
Justin T. Gibbs [Wed, 6 May 1998 01:45:55 +0000 (01:45 +0000)]
Implement bus_dmamem_* functions and correct a few nits reported by Peter Wemm.

26 years agoNow that we have a TAILQ_PREV() that returns the previous object, simplify
Justin T. Gibbs [Wed, 6 May 1998 01:44:12 +0000 (01:44 +0000)]
Now that we have a TAILQ_PREV() that returns the previous object, simplify
some of the buf_queue inline functions.

26 years agoUse the thread-aware errno definition all the time.
John Birrell [Tue, 5 May 1998 22:07:02 +0000 (22:07 +0000)]
Use the thread-aware errno definition all the time.

26 years agoBuild the syscalls (in libc, not libc_r) with weak symbols so that
John Birrell [Tue, 5 May 1998 22:06:16 +0000 (22:06 +0000)]
Build the syscalls (in libc, not libc_r) with weak symbols so that
libpthread can override them as required.

26 years agoRemove extern int errno and #include <errno.h> to get the proper definition.
John Birrell [Tue, 5 May 1998 22:04:13 +0000 (22:04 +0000)]
Remove extern int errno and #include <errno.h> to get the proper definition.

26 years agoRemove leading underscores from the FILE lock functions that POSIX
John Birrell [Tue, 5 May 1998 22:02:29 +0000 (22:02 +0000)]
Remove leading underscores from the FILE lock functions that POSIX
specifies.

26 years agoI have a long list of issues to address here, mostly related to
John Birrell [Tue, 5 May 1998 21:59:19 +0000 (21:59 +0000)]
I have a long list of issues to address here, mostly related to
namespace and revision levels of ANSI and POSIX. This change only
removes the leading underscrore from the FILE locking functions
that POSIX defines.

26 years agoRemove leading underscores for the functions (weak symbols here) that
John Birrell [Tue, 5 May 1998 21:56:42 +0000 (21:56 +0000)]
Remove leading underscores for the functions (weak symbols here) that
POSIX defines.

26 years agoThe __set_ospeed() function is coded against the speed_t type declared
John Birrell [Tue, 5 May 1998 21:54:26 +0000 (21:54 +0000)]
The __set_ospeed() function is coded against the speed_t type declared
in termios.h, but it's prototype in termcap.h and the main file use
the underlying definition (which is now an int, not a long for
compatibility with NetBSD). Really termcap.h should use speed_t too,
but I guess that this might break sources that don't include termios.h
first.

26 years agoChange a couple of long types to int to match NetBSD so that termios
John Birrell [Tue, 5 May 1998 21:49:37 +0000 (21:49 +0000)]
Change a couple of long types to int to match NetBSD so that termios
works on alpha without any modifications.

26 years agoTreat the lock value as volatile.
John Birrell [Tue, 5 May 1998 21:47:58 +0000 (21:47 +0000)]
Treat the lock value as volatile.

26 years agoIn a threaded library, expect the lock field to be declared volatile,
John Birrell [Tue, 5 May 1998 21:46:30 +0000 (21:46 +0000)]
In a threaded library, expect the lock field to be declared volatile,
so provide function prototypes that respect that, avoiding a gcc
warning that `volatile' is being thrown away.

26 years agoAdd variables for the ntpdate and xntpd program, you might want
Andreas Klemm [Tue, 5 May 1998 21:14:27 +0000 (21:14 +0000)]
Add variables for the ntpdate and xntpd program, you might want
to run the binaries from the new ntp v4 port.

26 years agoFix the search sequence for keymaps.
Dag-Erling Smørgrav [Tue, 5 May 1998 19:02:01 +0000 (19:02 +0000)]
Fix the search sequence for keymaps.

PR: bin/6522
Submitted by: Rudolf Cejka

26 years agoFix the 4-8G LBA geometry it was wrong.
Søren Schmidt [Tue, 5 May 1998 14:27:26 +0000 (14:27 +0000)]
Fix the 4-8G LBA geometry it was wrong.

Support >8G drives in CHS mode. This is done by guesstimating the
cylinder count from the LBA size reported. It works on my shiny
new Maxtor 11.5G drive, YMMV.

Reports from users of other big drives (read Quantum bigfoot's)
are welcome...

26 years agoInclude pc98.h instead of isa.h when PC98 is defined.
KATO Takenori [Tue, 5 May 1998 13:21:08 +0000 (13:21 +0000)]
Include pc98.h instead of isa.h when PC98 is defined.

26 years agoDeleted unused item.
KATO Takenori [Tue, 5 May 1998 13:19:34 +0000 (13:19 +0000)]
Deleted unused item.

26 years agoDon't generate y.tab.h, since it is not used.
Bruce Evans [Tue, 5 May 1998 13:19:13 +0000 (13:19 +0000)]
Don't generate y.tab.h, since it is not used.

26 years agoAdded y.tab.h to SRCS. This should have been used to get dependencies
Bruce Evans [Tue, 5 May 1998 13:16:35 +0000 (13:16 +0000)]
Added y.tab.h to SRCS.  This should have been used to get dependencies
on y.tab.h generated.  It will soon be used to tell bsd.dep.mk that
y.tab.h itself needs to be generated.

26 years agoPut y.tab.h back in SRCS. This used to be the best way to get
Bruce Evans [Tue, 5 May 1998 13:00:26 +0000 (13:00 +0000)]
Put y.tab.h back in SRCS.  This used to be the best way to get
dependencies on y.tab.h generated.  It will soon be used to tell
bsd.dep.mk that y.tab.h itself needs to be generated.

26 years agoFix the path.
John Birrell [Tue, 5 May 1998 12:33:21 +0000 (12:33 +0000)]
Fix the path.

26 years agoSpecify an absolute include path.
John Birrell [Tue, 5 May 1998 10:32:47 +0000 (10:32 +0000)]
Specify an absolute include path.

Backout the previous commit in favour of the Makefile.inc added to
the parent directory.

26 years agoOne of those silently included files that just about guarantee that
John Birrell [Tue, 5 May 1998 10:27:39 +0000 (10:27 +0000)]
One of those silently included files that just about guarantee that
nobody has any idea how something like BINDIR gets defined.

26 years agoThe BINDIR definition is an extra directory level away, so include it
John Birrell [Tue, 5 May 1998 08:57:11 +0000 (08:57 +0000)]
The BINDIR definition is an extra directory level away, so include it
explicitly.

26 years agoOops, missed the common part of gas.
John Birrell [Tue, 5 May 1998 08:53:27 +0000 (08:53 +0000)]
Oops, missed the common part of gas.

26 years agoGas build for i386 requires env as a build tool.
John Birrell [Tue, 5 May 1998 08:40:38 +0000 (08:40 +0000)]
Gas build for i386 requires env as a build tool.

26 years agoBacked out previous commit. It was tested, but not for the case where
Bruce Evans [Tue, 5 May 1998 07:36:55 +0000 (07:36 +0000)]
Backed out previous commit.  It was tested, but not for the case where
a separate object tree doesn't exist.  Crufty makefiles will have to
put y.tab.h in SRCS so that we know not to create foo.h from foo.y.

26 years agoAdd a message about editing /etc/group to the beginning of the build.
John Birrell [Tue, 5 May 1998 07:01:40 +0000 (07:01 +0000)]
Add a message about editing /etc/group to the beginning of the build.
FreeBSD needs a few groups that NetBSD doesn't provide by default. This
is annoying because it is the only thing that you need to do to a
NetBSD/Alpha system to bootstrap build FreeBSD/Alpha. Oh well, it
wouldn't be Unix without gratuitous differences, would it?

26 years agoBacked out previous change. Generating vgaio.h from vgaio.y clobbered
Bruce Evans [Tue, 5 May 1998 06:52:09 +0000 (06:52 +0000)]
Backed out previous change.  Generating vgaio.h from vgaio.y clobbered
the source file vgaio.h, and y.tab.h is used so we can't simply avoid
generating the yacc header.

26 years agoAdd usage() and rcsid. Getopt() returns -1 not EOF. Remove unused #includes.
Philippe Charnier [Tue, 5 May 1998 06:24:15 +0000 (06:24 +0000)]
Add usage() and rcsid. Getopt() returns -1 not EOF. Remove unused #includes.
Use .An/.Aq in man page.

26 years agoAdd usage() and rcsid. Remove unused #include. -Wall.
Philippe Charnier [Tue, 5 May 1998 06:13:47 +0000 (06:13 +0000)]
Add usage() and rcsid. Remove unused #include. -Wall.

26 years agoDon't generate a2p.h or y.tab.h from a2p.y. Neither is actually used,
Bruce Evans [Tue, 5 May 1998 05:25:21 +0000 (05:25 +0000)]
Don't generate a2p.h or y.tab.h from a2p.y.  Neither is actually used,
and building a2p.h clobbered the unrelated source file a2p.h if
${.OBJDIR} == ${.CURDIR}.

26 years agoOops, the symlink from y.tab.h doesn't work until its target exists.
Bruce Evans [Tue, 5 May 1998 03:59:27 +0000 (03:59 +0000)]
Oops, the symlink from y.tab.h doesn't work until its target exists.

26 years agoPermit double digit year values to be used in the next millenium.
Alexander Langer [Tue, 5 May 1998 01:53:15 +0000 (01:53 +0000)]
Permit double digit year values to be used in the next millenium.

26 years agoReplace _exit() with exit()
Robert Nordier [Tue, 5 May 1998 00:28:51 +0000 (00:28 +0000)]
Replace _exit() with exit()
Pointed out by: Nathan Torkington <gnat@prometheus.frii.com> PR 5585

26 years agoFix typo.
Robert Nordier [Tue, 5 May 1998 00:12:01 +0000 (00:12 +0000)]
Fix typo.
Pointed out by: Junji SAKAI <sakai.jp.freebsd.org> PR 5621

26 years agoFix a few nits in quoted code fragments and elsewhere.
Robert Nordier [Mon, 4 May 1998 23:16:50 +0000 (23:16 +0000)]
Fix a few nits in quoted code fragments and elsewhere.

26 years agoMake compatible with standard what command:
Robert Nordier [Mon, 4 May 1998 22:26:56 +0000 (22:26 +0000)]
Make compatible with standard what command:
    Implement -s option
    Keep going if a file is not found
    Append ':' to printed filenames
    Include backslash in terminal characters
    Make exit status dependent on a match having been found
    Revise manpage accordingly

26 years agoSupport cross architectures by using specific directories instead of
John Birrell [Mon, 4 May 1998 22:09:10 +0000 (22:09 +0000)]
Support cross architectures by using specific directories instead of
the ones that match the host.

26 years agoi386 gas config files.
John Birrell [Mon, 4 May 1998 22:05:54 +0000 (22:05 +0000)]
i386 gas config files.

26 years agoAlpha gas config files.
John Birrell [Mon, 4 May 1998 22:04:55 +0000 (22:04 +0000)]
Alpha gas config files.

26 years agoSupport cross-architectures using sub-directories.
John Birrell [Mon, 4 May 1998 22:03:00 +0000 (22:03 +0000)]
Support cross-architectures using sub-directories.

26 years agoAdd support for cross-architecture gases. Since the GNU code doesn't
John Birrell [Mon, 4 May 1998 22:01:27 +0000 (22:01 +0000)]
Add support for cross-architecture gases. Since the GNU code doesn't
support more than one architecture at a time, build as from the
default for the host and if CROSS_TOOLS defines other architectures,
build them as as_${arch}

26 years agoBINDIR is now specified in a single place. There are no cross-architecture
John Birrell [Mon, 4 May 1998 21:51:32 +0000 (21:51 +0000)]
BINDIR is now specified in a single place. There are no cross-architecture
issues with this utility.

26 years agoWork out the list of emulations based on the cross-architectures
John Birrell [Mon, 4 May 1998 21:38:46 +0000 (21:38 +0000)]
Work out the list of emulations based on the cross-architectures
defined. The TARGET is now set from an included makefile.

26 years agoAdd cross-architecture support.
John Birrell [Mon, 4 May 1998 21:29:54 +0000 (21:29 +0000)]
Add cross-architecture support.

26 years agoBINDIR is now specified in a single place. There are no cross-architecture
John Birrell [Mon, 4 May 1998 21:27:17 +0000 (21:27 +0000)]
BINDIR is now specified in a single place. There are no cross-architecture
issues with this utility.

26 years agoBIDIR is now specified in a single place. There are no cross-architecture
John Birrell [Mon, 4 May 1998 21:23:23 +0000 (21:23 +0000)]
BIDIR is now specified in a single place. There are no cross-architecture
issues with this utility.

26 years agoEvaluate the TARGET from the architecture. We'll assume that FREEBSD
John Birrell [Mon, 4 May 1998 21:21:17 +0000 (21:21 +0000)]
Evaluate the TARGET from the architecture. We'll assume that FREEBSD
is ELF from here on.

26 years agoRemote i386 cross-architecture support. It's now configurable, and off
John Birrell [Mon, 4 May 1998 21:18:46 +0000 (21:18 +0000)]
Remote i386 cross-architecture support. It's now configurable, and off
by default.

26 years agoAdd cross-architecture support.
John Birrell [Mon, 4 May 1998 21:16:46 +0000 (21:16 +0000)]
Add cross-architecture support.

26 years agoSince the makefiles below binutils with the architecture extension
John Birrell [Mon, 4 May 1998 21:13:50 +0000 (21:13 +0000)]
Since the makefiles below binutils with the architecture extension
are now included according to the cross-architecture support required,
default the BINDIR for i386 to /usr/libexec/elf here instead of in
all the i386 specific makefiles. For all other architectures, BINDIR
is just /usr/bin.

26 years agoAdd makefile support for cross-architectures. Allow CROSS_TOOLS to
John Birrell [Mon, 4 May 1998 21:10:56 +0000 (21:10 +0000)]
Add makefile support for cross-architectures. Allow CROSS_TOOLS to
be defined (in /etc/make.conf, say) and set to the additional architectures
that need to be compiled in. So on alpha I set CROSS_TOOLS = i386.
On i386 you can't build alpha due to lack of 64-bit support on 32-bit
architectures, but that's a GNU problem.

This change relies on makefiles in the binutils sub-directories having
the extension defined in the CROSS_TOOLS, instead of those makefiles
being selected based on the host architecture.

26 years agoFixed races in `make -jN' using new yacc rules.
Bruce Evans [Mon, 4 May 1998 20:09:06 +0000 (20:09 +0000)]
Fixed races in `make -jN' using new yacc rules.

26 years agoSimplified by using new lex rules.
Bruce Evans [Mon, 4 May 1998 20:00:14 +0000 (20:00 +0000)]
Simplified by using new lex rules.