]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
26 years agoFixed ownership of share/man/man1aout.
Bruce Evans [Thu, 10 Sep 1998 15:46:41 +0000 (15:46 +0000)]
Fixed ownership of share/man/man1aout.

Didn't fix misformatting of all the perl entries.

26 years agoFix an inefficiency I introduced in my last commit.
Martin Cracauer [Thu, 10 Sep 1998 14:51:06 +0000 (14:51 +0000)]
Fix an inefficiency I introduced in my last commit.
Include "expand.h" vom memalloc.c to pull function declartion into scope

26 years agoThe fix in the previous commit was not sufficient; the upper 24 bits
Kazutaka YOKOTA [Thu, 10 Sep 1998 12:20:09 +0000 (12:20 +0000)]
The fix in the previous commit was not sufficient; the upper 24 bits
of an int argument still contained garbage.
Pointed out by: bde
PR: bin/7799

26 years agoPass the correct argument to npxsave(), otherwise vm86pcb will be
Kazutaka YOKOTA [Thu, 10 Sep 1998 12:16:06 +0000 (12:16 +0000)]
Pass the correct argument to npxsave(), otherwise vm86pcb will be
overwritten.

26 years agoOops missed a line in the previous commit
Søren Schmidt [Thu, 10 Sep 1998 11:23:08 +0000 (11:23 +0000)]
Oops missed a line in the previous commit

26 years agoDorkus sum. I managed to commit an earlier version of this. This fixes the
Mark Murray [Thu, 10 Sep 1998 10:51:11 +0000 (10:51 +0000)]
Dorkus sum. I managed to commit an earlier version of this. This fixes the
broken perl build.

* NOTE * * NOTE * * NOTE *

PERL5 RELIES IN A MAKE WORLD FOR THE BOOTSTRAP!!

* NOTE * * NOTE * * NOTE *

26 years agoEnsure that m_nextpkt is set to NULL after reassembling fragments.
Doug Rabson [Thu, 10 Sep 1998 08:56:40 +0000 (08:56 +0000)]
Ensure that m_nextpkt is set to NULL after reassembling fragments.

26 years agoUpdate info on the bt848 driver.
Søren Schmidt [Thu, 10 Sep 1998 08:20:46 +0000 (08:20 +0000)]
Update info on the bt848 driver.

Submitted by: Roger Hardiman <roger@cs.strath.ac.uk>

26 years agoChanged METEORSINPUT for Hauppauge cards with bt878.
Søren Schmidt [Thu, 10 Sep 1998 08:01:10 +0000 (08:01 +0000)]
Changed METEORSINPUT for Hauppauge cards with bt878.
Also fixed video_open defines and 878 support.

Submitted by: Fred Templin <templin@erg.sri.com>

26 years agoCorrect SECAM B-Delay and add XUSSR channel set.
Søren Schmidt [Thu, 10 Sep 1998 07:52:30 +0000 (07:52 +0000)]
Correct SECAM B-Delay and add XUSSR channel set.

Submitted by: Vsevolod Lobko <seva@alex-ua.com>

26 years agoNote ELF, Perl5 and Atapi tape support.
Jordan K. Hubbard [Thu, 10 Sep 1998 07:03:16 +0000 (07:03 +0000)]
Note ELF, Perl5 and Atapi tape support.

26 years ago(1) Add MLINKS support.
Satoshi Asami [Thu, 10 Sep 1998 06:38:02 +0000 (06:38 +0000)]
(1) Add MLINKS support.
Submitted by: hoek

(2) ELF support.
 2a. Variable PORTOBJFORMAT specifies the object format of the system.
     It is passed down to configure/make via CONFIGURE_ENV/MAKE_ENV,
     and is given to generate-plist via PLIST_SUB.
 2b. In PLIST, substitute lines that end with "/libFOO.so.X" with
     "/libFOO.so.X.0".  (This means PLISTs should only list ELF
     libraries.)
Reviewed by: jb, jdp, hoek, jseger, steve

(3) Perl5-in-system support.  Basically turns USE_PERL5 into a no-op
    if there exists a "/usr/bin/perl5".  Also fix prior breakage by
    dima (${PREFIX} => ${LOCALBASE} in perl5 path).
Reviewed by: markm (sort of)

(4) Install requirement file as "+REQUIRE" so it will be executed
    correctly by pkg_delete.
Reported by: Shigeyuki FUKUSHIMA <shige@kuis.kyoto-u.ac.jp>

(5) Do not disable checksum and makesum when NO_EXTRACT is set.
Pointed out by: hoek and kiri, among other people

26 years agoDon't keep the underlying directory locked while performing the file
Tor Egge [Thu, 10 Sep 1998 02:27:52 +0000 (02:27 +0000)]
Don't keep the underlying directory locked while performing the file
system specific VFS_MOUNT operation.
PR: 1067

26 years agoDon't allow longjmp into a no longer valid stack frame.
Tor Egge [Thu, 10 Sep 1998 02:01:06 +0000 (02:01 +0000)]
Don't allow longjmp into a no longer valid stack frame.

26 years agoI don't want to be the only one to use ``cut''. Use ``sed'' instead.
David E. O'Brien [Thu, 10 Sep 1998 00:27:57 +0000 (00:27 +0000)]
I don't want to be the only one to use ``cut''.  Use ``sed'' instead.

26 years agoAdd ``usr/bin/cut'' to "build-tools" target.
David E. O'Brien [Thu, 10 Sep 1998 00:13:48 +0000 (00:13 +0000)]
Add ``usr/bin/cut'' to "build-tools" target.
Used in usr.sbin/amd/include/newvers.sh.

26 years agoReflect reality
Brian Somers [Wed, 9 Sep 1998 22:51:49 +0000 (22:51 +0000)]
Reflect reality

26 years agoMention some of the major changes to ppp for 3.0
Brian Somers [Wed, 9 Sep 1998 22:50:28 +0000 (22:50 +0000)]
Mention some of the major changes to ppp for 3.0
Fix a typo.

26 years agoDisable kernel_secure_level unless explicitly set in rc.conf. Previously,
James Raynard [Wed, 9 Sep 1998 22:27:21 +0000 (22:27 +0000)]
Disable kernel_secure_level unless explicitly set in rc.conf.  Previously,
it was enabled unless explicitly unset, creating a pitfall for people
like me who upgraded /etc/rc without upgrading /etc/rc.conf.

26 years agoMention which system interface functions are signal-safe.
Brian Somers [Wed, 9 Sep 1998 20:44:51 +0000 (20:44 +0000)]
Mention which system interface functions are signal-safe.
Suggested on -current by: Terry Lambert <tlambert@primenet.com>

26 years agoFixed the usual missing permissions checks in mount(). As for cd9660,
Bruce Evans [Wed, 9 Sep 1998 20:21:18 +0000 (20:21 +0000)]
Fixed the usual missing permissions checks in mount().  As for cd9660,
the damage was limited by the default of 0 for vfs.usermount.

Obtained from: Lite2 via the -current ffs_vfsops.c

26 years agoRemove the NOPERL block to building Perl5.
Mark Murray [Wed, 9 Sep 1998 19:30:45 +0000 (19:30 +0000)]
Remove the NOPERL block to building Perl5.

Perl 5 will rebuild/bootstrap itself next time a make world is done
with this in effect.

26 years agoAdd the wst device (ATAPI tape)
Søren Schmidt [Wed, 9 Sep 1998 18:59:46 +0000 (18:59 +0000)]
Add the wst device (ATAPI tape)
Allow 8 wcd devices instead of 4.

26 years agoMajor numbers allocated for generic SMB/I2C i/o
Nicolas Souchu [Wed, 9 Sep 1998 18:57:59 +0000 (18:57 +0000)]
Major numbers allocated for generic SMB/I2C i/o

26 years agoMajor number allocated for generic SMB i/o -> 106
Nicolas Souchu [Wed, 9 Sep 1998 18:57:38 +0000 (18:57 +0000)]
Major number allocated for generic SMB i/o -> 106

26 years agoMajor number allocated for generic I2C i/o -> 105
Nicolas Souchu [Wed, 9 Sep 1998 18:57:24 +0000 (18:57 +0000)]
Major number allocated for generic I2C i/o -> 105

26 years agoDocument a number of VM sysctl variables with help from old emails
Guy Helmer [Wed, 9 Sep 1998 18:36:14 +0000 (18:36 +0000)]
Document a number of VM sysctl variables with help from old emails
written by John Dyson.

26 years agoAdd the dirs for Perl5.
Mark Murray [Wed, 9 Sep 1998 17:54:47 +0000 (17:54 +0000)]
Add the dirs for Perl5.

26 years agoAdd libperl for Perl5.
Mark Murray [Wed, 9 Sep 1998 17:51:15 +0000 (17:51 +0000)]
Add libperl for Perl5.

26 years agooptions changed with the new AMD.
David E. O'Brien [Wed, 9 Sep 1998 17:37:14 +0000 (17:37 +0000)]
options changed with the new AMD.
Rather than ``nfsv2'', we now have ``vers=3,proto=tcp''

26 years agoAdd the Perl5 bootstrap tools.
Mark Murray [Wed, 9 Sep 1998 17:30:30 +0000 (17:30 +0000)]
Add the Perl5 bootstrap tools.

26 years agogetopt returns -1 not EOF
Warner Losh [Wed, 9 Sep 1998 17:24:31 +0000 (17:24 +0000)]
getopt returns -1 not EOF

26 years agoAdd dependency for subr_bus.c on bus_if.h and device_if.h so that a make
Warner Losh [Wed, 9 Sep 1998 17:23:25 +0000 (17:23 +0000)]
Add dependency for subr_bus.c on bus_if.h and device_if.h so that a make
depend is not required to just build the kernel.

Reviewed by: Doug Rabson

26 years agoPrint warning about block size not being optimal once per open rather than
Warner Losh [Wed, 9 Sep 1998 17:20:32 +0000 (17:20 +0000)]
Print warning about block size not being optimal once per open rather than
once per offending write.

26 years agoFix fencepost error in readlink; inspired by OpenBSD
Warner Losh [Wed, 9 Sep 1998 17:14:17 +0000 (17:14 +0000)]
Fix fencepost error in readlink; inspired by OpenBSD

26 years agoUse size of structure rather than hard coded 256 in readlink. This change
Warner Losh [Wed, 9 Sep 1998 17:12:58 +0000 (17:12 +0000)]
Use size of structure rather than hard coded 256 in readlink.  This change
either came from OpenBSD, or was inspired by work done there.

26 years agosprintf -> snprintf
Warner Losh [Wed, 9 Sep 1998 17:08:46 +0000 (17:08 +0000)]
sprintf -> snprintf

26 years agoAdd alpha-*=freebsd* to configure
Warner Losh [Wed, 9 Sep 1998 17:07:36 +0000 (17:07 +0000)]
Add alpha-*=freebsd* to configure

26 years agoUse readlink correctly
Warner Losh [Wed, 9 Sep 1998 17:05:54 +0000 (17:05 +0000)]
Use readlink correctly

26 years agoFixed bug, caused threads to wait for locks which nobody hold.
Dmitrij Tejblum [Wed, 9 Sep 1998 16:50:33 +0000 (16:50 +0000)]
Fixed bug, caused threads to wait for locks which nobody hold.
Fixed sign extension bug, caused undefined behavior if FILE live in stack.

26 years agoDon't refuse to pass the force and sync options to the kernel. The
Bruce Evans [Wed, 9 Sep 1998 16:26:51 +0000 (16:26 +0000)]
Don't refuse to pass the force and sync options to the kernel.  The
kernel supports them, and the force option will be more important
when mounting of unclean ext2fs file systems is not permitted unless
it is forced.

26 years agoFully use our own generated files for NFSv3.
David E. O'Brien [Wed, 9 Sep 1998 16:11:01 +0000 (16:11 +0000)]
Fully use our own generated files for NFSv3.

26 years agoSet the `pc98' string to the MACHINE if machdep.ispc98 != 0.
KATO Takenori [Wed, 9 Sep 1998 14:58:30 +0000 (14:58 +0000)]
Set the `pc98' string to the MACHINE if machdep.ispc98 != 0.

26 years agoSync with sys/i386/conf/majors.i386 revision 1.47.
KATO Takenori [Wed, 9 Sep 1998 14:40:27 +0000 (14:40 +0000)]
Sync with sys/i386/conf/majors.i386 revision 1.47.

26 years agoSync with sys/i386/conf/files.i386 revision 1.202.
KATO Takenori [Wed, 9 Sep 1998 14:39:56 +0000 (14:39 +0000)]
Sync with sys/i386/conf/files.i386 revision 1.202.

26 years agoSync with sys/i386/conf/options.i386 revision 1.85.
KATO Takenori [Wed, 9 Sep 1998 14:39:20 +0000 (14:39 +0000)]
Sync with sys/i386/conf/options.i386 revision 1.85.

26 years agoRemove perl5 from list of ports to build now that it's in the base system.
Jordan K. Hubbard [Wed, 9 Sep 1998 14:20:22 +0000 (14:20 +0000)]
Remove perl5 from list of ports to build now that it's in the base system.

26 years agoMinimalist fixes to make BMaked perl build.
Mark Murray [Wed, 9 Sep 1998 13:10:47 +0000 (13:10 +0000)]
Minimalist fixes to make BMaked perl build.

26 years agoDon't forget to initialize the inode lock. This bug caused
Bruce Evans [Wed, 9 Sep 1998 13:09:24 +0000 (13:09 +0000)]
Don't forget to initialize the inode lock.  This bug caused
surprisingly few problems.  Most fields were initialized to the
correct values by bzero(), but lk_prio was 0 instead of PINOD (=8),
the lk_wmsg was NULL instead of "ext2in", and lk_lockholder was 0
instead of -1.

Obtained from: Lite2 via the -current ffs_vfsops.c

26 years agoSupport compiling with `gcc -pedantic' (don't use hard newlines in
Bruce Evans [Wed, 9 Sep 1998 12:22:17 +0000 (12:22 +0000)]
Support compiling with `gcc -pedantic' (don't use hard newlines in
(asm) string constants).

26 years agoChange ${MACHINE} into ${MACHINE_ARCH} to support MACHINE=pc98.
KATO Takenori [Wed, 9 Sep 1998 11:27:33 +0000 (11:27 +0000)]
Change ${MACHINE} into ${MACHINE_ARCH} to support MACHINE=pc98.

26 years agoChange i386 in a few paths to ${MACHINE} to support MACHINE=pc98.
KATO Takenori [Wed, 9 Sep 1998 11:22:28 +0000 (11:22 +0000)]
Change i386 in a few paths to ${MACHINE} to support MACHINE=pc98.

26 years agoChange i386 in a few paths to ${MACHINE} to support MACHINE=pc98.
KATO Takenori [Wed, 9 Sep 1998 11:13:55 +0000 (11:13 +0000)]
Change i386 in a few paths to ${MACHINE} to support MACHINE=pc98.

26 years agoDocument why the library is still not being build as shared.
Martin Cracauer [Wed, 9 Sep 1998 10:21:12 +0000 (10:21 +0000)]
Document why the library is still not being build as shared.
PR: gnu/7811

26 years agoFixed disordering and misformatting in previous commit.
Bruce Evans [Wed, 9 Sep 1998 10:17:34 +0000 (10:17 +0000)]
Fixed disordering and misformatting in previous commit.

Don't override correct default targets.

Don't use afterinstall for installing things; it is for fixups after
installing things.

26 years agoBacked out rev.1.28. It was bogus because bsd.man.mk is only used
Bruce Evans [Wed, 9 Sep 1998 10:12:07 +0000 (10:12 +0000)]
Backed out rev.1.28.  It was bogus because bsd.man.mk is only used
(and only usable) by other .mk files that have their own multiple
inclusion protection.

26 years agoTurned on -Wformat -fformat-extensions. -fformat-extensions requires a
Bruce Evans [Wed, 9 Sep 1998 10:04:58 +0000 (10:04 +0000)]
Turned on -Wformat -fformat-extensions.  -fformat-extensions requires a
-current version of gcc.  Without it, -Wformat would complain about all
the nonstandard %[Dbrz] formats in the kernel.

26 years agoDon't define RANLIBMAG if building an elf version of make. The RANLIBMAG
John Birrell [Wed, 9 Sep 1998 08:29:15 +0000 (08:29 +0000)]
Don't define RANLIBMAG if building an elf version of make. The RANLIBMAG
option lets make look inside archive libraries when determining if a
library is out-of-date. I don't think that make should look inside
files, so this fix effectively tells it not to. If the decision is
made (by someone else) that make _should_ be doing this, it will need
to be taught how to recognise elf archive libraries.

Problem noticed by: Joseph Koshy <jkoshy@freebsd.org>

26 years agoDon't use CTL_VFS at the wrong level. This caused loops in the sysctl
Bruce Evans [Wed, 9 Sep 1998 07:41:41 +0000 (07:41 +0000)]
Don't use CTL_VFS at the wrong level.  This caused loops in the sysctl
tree if CTL_VFS happened to get assigned as a type number to a vfs that
has some vfs sysctls.

26 years agoBring in the Perl5 BMake files. This ends the easy stuff.
Mark Murray [Wed, 9 Sep 1998 07:20:33 +0000 (07:20 +0000)]
Bring in the Perl5 BMake files. This ends the easy stuff.

I'll be doiung the rest in stages.

26 years agoThis commit was generated by cvs2svn to compensate for changes in r38980,
Mark Murray [Wed, 9 Sep 1998 07:00:04 +0000 (07:00 +0000)]
This commit was generated by cvs2svn to compensate for changes in r38980,
which included commits to RCS files with non-trunk default branches.

26 years agoInitial import of Perl5. The king is dead; long live the king!
Mark Murray [Wed, 9 Sep 1998 07:00:04 +0000 (07:00 +0000)]
Initial import of Perl5. The king is dead; long live the king!

26 years agoOld Perl is leaving us. Goodbye, faithful friend.
Mark Murray [Wed, 9 Sep 1998 06:49:33 +0000 (06:49 +0000)]
Old Perl is leaving us. Goodbye, faithful friend.

26 years agoAdd the missing rerelease target back.
John Birrell [Wed, 9 Sep 1998 06:07:32 +0000 (06:07 +0000)]
Add the missing rerelease target back.

Reported by: Justin Gibbs.

Add the move-aout-libs upgrade target so that people who have already
gone elf can put their libraries through the mincer. Anyone who hasn't
deleted aout libraries from /usr/lib (but has done a make world putting
the new aout libs in /usr/lib/aout) will be asked for confirmation
to delete them one by one.

26 years agoAdd the missing rerelease target back.
John Birrell [Wed, 9 Sep 1998 06:03:51 +0000 (06:03 +0000)]
Add the missing rerelease target back.

Reported by: Justin Gibbs.

26 years agoAdd a move-aout-libs target to do just that. It takes the well known
John Birrell [Wed, 9 Sep 1998 06:02:31 +0000 (06:02 +0000)]
Add a move-aout-libs target to do just that. It takes the well known
library directories: /usr/lib /usr/lib/compat /usr/local/lib and
/usr/X11R6/lib and adds any other directories that ldconfig knows
about, then removes any that are already aout directories (that is,
with aout as a component of the path) and passes this directory
list to the move_aout_libs.sh script to do the hack (believe it,
it moves libraries that are in use) work.

26 years agoAnother place to set NOPERL.
Mark Murray [Wed, 9 Sep 1998 06:01:37 +0000 (06:01 +0000)]
Another place to set NOPERL.

26 years agoSet NOPERL while I spend the rest of the day bringing in Perl5.
Mark Murray [Wed, 9 Sep 1998 06:01:11 +0000 (06:01 +0000)]
Set NOPERL while I spend the rest of the day bringing in Perl5.
This will doubtless confuse someone, but I'd rather have this than
breakage.

26 years agoAdd a note about the upgrade tool directory.
John Birrell [Wed, 9 Sep 1998 05:48:51 +0000 (05:48 +0000)]
Add a note about the upgrade tool directory.

26 years agoA script for use when upgrading an installed system from aout to
John Birrell [Wed, 9 Sep 1998 05:28:18 +0000 (05:28 +0000)]
A script for use when upgrading an installed system from aout to
elf.

26 years agoFix spurious `free()' when -d g2 is enabled.
Joseph Koshy [Wed, 9 Sep 1998 05:15:38 +0000 (05:15 +0000)]
Fix spurious `free()' when -d g2 is enabled.

PR: 6790
Submitted by: nsmart <njs3@doc.ic.ac.uk>

26 years agoFix typos.
Daniel O'Callaghan [Wed, 9 Sep 1998 05:09:55 +0000 (05:09 +0000)]
Fix typos.

26 years agoSome preliminary updates to the release notes, just as I think of
Jordan K. Hubbard [Wed, 9 Sep 1998 03:37:00 +0000 (03:37 +0000)]
Some preliminary updates to the release notes, just as I think of
them.  By no means a comprehensive list of all the things which should
eventually go in here before 3.0-RELEASE, nor are the CAM issues
reflected here yet.

26 years agoAdd missing comma.
Joseph Koshy [Wed, 9 Sep 1998 03:07:03 +0000 (03:07 +0000)]
Add missing comma.

PR: 7841
Submitted by: NOKUBI Hirotaka <hnokubi@yyy.or.jp>

26 years agoReserve character major 104 for the CAM Transport Layer Services device.
Justin T. Gibbs [Wed, 9 Sep 1998 03:07:02 +0000 (03:07 +0000)]
Reserve character major 104 for the CAM Transport Layer Services device.

26 years agoFix typo.
Joseph Koshy [Wed, 9 Sep 1998 03:04:53 +0000 (03:04 +0000)]
Fix typo.

PR: 7809
Submitted by: Amakawa Shuhei <amakawa@sf.t.u-tokyo.ac.jp>

26 years agoTerminate the defdrives array correctly. (Shame, Rod!)
Mike Smith [Wed, 9 Sep 1998 02:22:28 +0000 (02:22 +0000)]
Terminate the defdrives array correctly. (Shame, Rod!)
Submitted by: Scott Michel <scottm@cs.ucla.edu>

26 years agoAdd $Id$, remove quoting for `...'.
Joseph Koshy [Wed, 9 Sep 1998 01:30:25 +0000 (01:30 +0000)]
Add $Id$, remove quoting for `...'.

Requested by: bde

26 years agoAdd a new library function getobjformat(). It checks all the
John Polstra [Wed, 9 Sep 1998 01:21:25 +0000 (01:21 +0000)]
Add a new library function getobjformat().  It checks all the
standard places ("/etc/objformat", ${OBJFORMAT}, argv) for an
indication of the user's preferred object file format.  This
consolidates some code that was starting to be duplicated in more
and more places.

Use the new function in ldconfig.

Note: I don't think that gcc should use getobjformat(), even though
it could.  The compiler should limit itself to functions that are
widespread, to ease porting and cross-compilation.

26 years agoInitialise lcp::his_mru to the ``set mtu'' value if it's less
Brian Somers [Wed, 9 Sep 1998 00:03:09 +0000 (00:03 +0000)]
Initialise lcp::his_mru to the ``set mtu'' value if it's less
than DEF_MRU, allowing our interface mtu to be decreased
despite negotiation with the peer.

26 years ago- If the OACTIVE flag is set on entry to xl_start(), check to see if the
Bill Paul [Tue, 8 Sep 1998 23:42:10 +0000 (23:42 +0000)]
- If the OACTIVE flag is set on entry to xl_start(), check to see if the
  transmitter is wedged. If so, try to unwedge it, process any descriptors
  that might need to be free()d, then proceed.

- Disable the 'background' autonegotiation performed during bootstrap.
  What happens currently is that the driver starts an autoneg session,
  the sets a timeout in the ifnet structure and returns. Later, when the
  timer expires, the watchdog routine calls the autoneg handler to check
  the results of the session. The problem with this is that the session
  may not complete until some point after we have started to mount NFS
  filesystems, which can cause the mounts to fail. This is especially
  troublesome if booting with an NFS rootfs: we need the interface up
  and running before reaching the mountroot() code.

  The default behavior now is to do the autoneg synchronously, i.e. wait
  5 seconds for the autoneg to complete before exiting the driver attach
  routine. People who want the old behavior can compile the driver with
  XL_BACKGROUND_AUTONEG #defined. This has no effect on autoneg sessions
  initiated by 'ifconfig xl0 media autoselect.'

  This slows the probe down a little, but it's either that or botching
  NFS mounts at bootup.

- If xl_setmode_mii() is called and there's an autoneg session in progress,
  cancel it, _then_ set the modes.

26 years agoAdd new atapi-cd driver that supports atapi CD-R/RW drives.
Søren Schmidt [Tue, 8 Sep 1998 20:57:47 +0000 (20:57 +0000)]
Add new atapi-cd driver that supports atapi CD-R/RW drives.
This is only a stop-gab solution to get atapi burner support into 3.0.

26 years agoAdd 'blank' and 'nextwriteable' commands for atapi CD-R/RW support.
Søren Schmidt [Tue, 8 Sep 1998 20:51:24 +0000 (20:51 +0000)]
Add 'blank' and 'nextwriteable' commands for atapi CD-R/RW support.

26 years agoAdd definitions for the ELF section attribute flags, SHF_*.
John Polstra [Tue, 8 Sep 1998 20:38:06 +0000 (20:38 +0000)]
Add definitions for the ELF section attribute flags, SHF_*.

Submitted by: Robert Nordier <rnordier@nordier.com>

26 years agoMake PCIC_RESUME_RESET an proper option.
Brian Somers [Tue, 8 Sep 1998 18:09:51 +0000 (18:09 +0000)]
Make PCIC_RESUME_RESET an proper option.
My laptop (a CTX Cybernote) needs this.  It claims to have a

  PC-Card VLSI 82C146 (5 mem & 2 I/O windows)

26 years agoAdd support for the RPC 64-bit integer type ``hyper''.
David E. O'Brien [Tue, 8 Sep 1998 17:33:12 +0000 (17:33 +0000)]
Add support for the RPC 64-bit integer type ``hyper''.

26 years agoOops, don't assume that the environment is normal in devfs_mount().
Bruce Evans [Tue, 8 Sep 1998 16:59:37 +0000 (16:59 +0000)]
Oops, don't assume that the environment is normal in devfs_mount().
It isn't for the hidden mountpoint.  The static vfs's haven't been
attached then, so mp->mnt_vfc can't be valid.

26 years agoIf traps are set, they are now executed even when a signal-blocking
Martin Cracauer [Tue, 8 Sep 1998 13:16:52 +0000 (13:16 +0000)]
If traps are set, they are now executed even when a signal-blocking
foreground child is running. Formerly, traps were exceuted after the
next child exit.

The enables the user to put a breaking wrapper around a blocking
application:
  (trap 'echo trap ; exit 1' 2; ./pestyblocker; echo -n)

The "echo -n" after the child call is needed to prevent sh from
optimizing the trap-executing shell away. I'm working on this.

26 years agoUpdate the diffs against current source + some other cleanup.
Andrzej Bialecki [Tue, 8 Sep 1998 12:47:35 +0000 (12:47 +0000)]
Update the diffs against current source + some other cleanup.

26 years agoWhoops, update the makefile too while I'm at it.
Jordan K. Hubbard [Tue, 8 Sep 1998 11:44:37 +0000 (11:44 +0000)]
Whoops, update the makefile too while I'm at it.

26 years agoGo back to using MSDOSFS instead of Robert Nordier's stand-alone DOS
Jordan K. Hubbard [Tue, 8 Sep 1998 11:44:08 +0000 (11:44 +0000)]
Go back to using MSDOSFS instead of Robert Nordier's stand-alone DOS
library.  I have to include MSDOSFS on the boot floppy anyway (so I can
read DOS floppies in a few other contexts) and since both -stable and
-current MSDOSFS handle long filenames, I'm OK on that now.  Should
save about 10K of space.

26 years agoMake totally sure that we have the latest 1.7 version of booteasy
Jordan K. Hubbard [Tue, 8 Sep 1998 10:46:40 +0000 (10:46 +0000)]
Make totally sure that we have the latest 1.7 version of booteasy
(it looked like it already to me, but the PR gives me cause to wonder).
PR: 2514

26 years agoo Set paths internally for pkg_add for tools we might want to invoke.
Jordan K. Hubbard [Tue, 8 Sep 1998 10:42:26 +0000 (10:42 +0000)]
o Set paths internally for pkg_add for tools we might want to invoke.
o Fix bogus suffix handling.
o Tell user when an FTP url is being xferred rather than being silent.
  This sort of violates "the unix way" but it stops people from whacking
  ^C when they think it's hung, too.  Sometimes visual indication of
  success is important.  Doesn't spit out anything if not on a TTY.

26 years agoSync with sys/i386/isa/clock.c revision 1.125.
KATO Takenori [Tue, 8 Sep 1998 09:47:46 +0000 (09:47 +0000)]
Sync with sys/i386/isa/clock.c revision 1.125.

26 years agoFix a cut&paste error which prevented LD_BIND_NOW from working.
Doug Rabson [Tue, 8 Sep 1998 09:47:35 +0000 (09:47 +0000)]
Fix a cut&paste error which prevented LD_BIND_NOW from working.

26 years agoSync with sys/i386/i386/userconfig.c revision 1.107.
KATO Takenori [Tue, 8 Sep 1998 09:47:09 +0000 (09:47 +0000)]
Sync with sys/i386/i386/userconfig.c revision 1.107.

26 years agorestore rev 1.23 to give clear SIGALRM handling example, 68 years
Andrey A. Chernov [Tue, 8 Sep 1998 09:44:46 +0000 (09:44 +0000)]
restore rev 1.23 to give clear SIGALRM handling example, 68 years
are too long to affect real code

26 years agofix a problem with our definition of SNDCTL_DSP_GETBLKSIZE, to let
Luigi Rizzo [Tue, 8 Sep 1998 07:21:00 +0000 (07:21 +0000)]
fix a problem with our definition of SNDCTL_DSP_GETBLKSIZE, to let
the FreeBSD x11amp work (same as in -stable)

26 years agoAdd info on PicoBSD being a part of FreeBSD source tree.
Andrzej Bialecki [Tue, 8 Sep 1998 06:35:43 +0000 (06:35 +0000)]
Add info on PicoBSD being a part of FreeBSD source tree.

26 years agoSync help files with actual contents of "dial" floppy.
Andrzej Bialecki [Tue, 8 Sep 1998 06:04:41 +0000 (06:04 +0000)]
Sync help files with actual contents of "dial" floppy.