]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
25 years agoClean up some of the documentation at the top.
jkh [Sun, 20 Jun 1999 03:36:46 +0000 (03:36 +0000)]
Clean up some of the documentation at the top.

25 years agoFor consistency with other implementations, check for the existence
alc [Sat, 19 Jun 1999 23:53:13 +0000 (23:53 +0000)]
For consistency with other implementations, check for the existence
of the segment before checking its permissions.

PR: kern/11999
Submitted by: Brooks Davis <brooks@one-eyed-alien.net>

25 years agoImplement -m option to enable/disable slices.
rnordier [Sat, 19 Jun 1999 21:44:43 +0000 (21:44 +0000)]
Implement -m option to enable/disable slices.

25 years agoProvide for enabling/disabling each slice individually.
rnordier [Sat, 19 Jun 1999 21:40:55 +0000 (21:40 +0000)]
Provide for enabling/disabling each slice individually.

25 years agoFix use of e_phoff in place of e_phnum.
rnordier [Sat, 19 Jun 1999 20:50:38 +0000 (20:50 +0000)]
Fix use of e_phoff in place of e_phnum.

25 years agoCorrect and clarify some comments.
rnordier [Sat, 19 Jun 1999 20:40:32 +0000 (20:40 +0000)]
Correct and clarify some comments.

25 years agoAdd device rl0 to GENERIC. Also add al0 (ADMtek PCI ethernet) which I
wpaul [Sat, 19 Jun 1999 20:25:25 +0000 (20:25 +0000)]
Add device rl0 to GENERIC. Also add al0 (ADMtek PCI ethernet) which I
forgot to do previously.

25 years agoGoodbye to vaules, becasue, similiar, backgroud, aquired, freelisat, etc.
rnordier [Sat, 19 Jun 1999 20:20:52 +0000 (20:20 +0000)]
Goodbye to vaules, becasue, similiar, backgroud, aquired, freelisat, etc.

25 years agoFix up the RealTek 8139 driver to work on FreeBSD/alpha. This involves a
wpaul [Sat, 19 Jun 1999 20:17:38 +0000 (20:17 +0000)]
Fix up the RealTek 8139 driver to work on FreeBSD/alpha. This involves a
few changes:

- there was a bug in rl_list_tx_init(): it was calculating the registers
  to initialize incorrectly. Not a problem on the x86 where unaligned
  access are allowed, but a problem on the alpha.

- set rl_btag accordingly depending on the machine type

- rl_rxeof() needs to be sure to longword-align the packet data. This
  is a little tricky since we copy the data out of the receive buffer
  using m_devget(), however there's no way to tell m_devget() to fill
  in the mbufs starting at a particular offset. To get around this,
  we tell m_devget to copy bytes+2 bytes starting at offset offset-2. This
  results in the proper alignment, and we can trim off the two leading
  bytes afterwards with m_adj(). We also allocate some extra space before
  the start of the receive buffer so that we don't get into trouble in
  the case where offset == 0.

- redefine vtophys() in if_rlreg.h for the alpha.

Making this chipset work on the alpha is sort of the inverse of putting
a jet engine on a rowboat (putting a propeller on a 747?) but when
you can get these things for $5 a pop, it's hard to stop people from
buying them.

25 years agoMove the IFSTATUS stuff after the address listing.
phk [Sat, 19 Jun 1999 20:07:24 +0000 (20:07 +0000)]
Move the IFSTATUS stuff after the address listing.

25 years agoMiscellaneous dd(1) changes: mainly fixing variable types (size_t, ssize_t,
green [Sat, 19 Jun 1999 19:49:35 +0000 (19:49 +0000)]
Miscellaneous dd(1) changes: mainly fixing variable types (size_t, ssize_t,
off_t, int, u_int64_t, etc.). dd(1) should now work properly with REALLY
big amounts of data.

25 years agoThanks to Bruce for noticing this.... compare against the *new* nfsnode's
mjacob [Sat, 19 Jun 1999 19:33:44 +0000 (19:33 +0000)]
Thanks to Bruce for noticing this.... compare against the *new* nfsnode's
mount point for seeing whether or not the new nfsnode is already in the
hash queue. We're pretty much guaranteed that the old nfsnode is already
in the hash queue. Wank! Infinite Loop! Looks like just a minor typo....
(ah the influence of fortran ... np && np2... why not nfsnode_the_first &&
nfsnode_the_second???)...

25 years agoThis is the much-awaited cleaned up version of IPFW [ug]id support.
green [Sat, 19 Jun 1999 18:43:33 +0000 (18:43 +0000)]
This is the much-awaited cleaned up version of IPFW [ug]id support.
All relevant changes have been made (including ipfw.8).

25 years agoRemove some unused function and variable declarations.
alc [Sat, 19 Jun 1999 18:42:53 +0000 (18:42 +0000)]
Remove some unused function and variable declarations.

25 years agoAdd a new interface ioctl, to return "aux status".
phk [Sat, 19 Jun 1999 18:42:31 +0000 (18:42 +0000)]
Add a new interface ioctl, to return "aux status".

This is inteded for to allow ifconfig to print various unstructured
information from an interface.

The data is returned from the kernel in ASCII form, see the comment in
if.h for some technicalities.

Canonical cut&paste example to be found in if_tun.c

Initial use:
        Now tun* interfaces tell the PID of the process which opened them.

Future uses could be (volounteers welcome!):
        Have ppp/slip interfaces tell which tty they use.
        Make sync interfaces return their media state: red/yellow/blue
        alarm, timeslot assignment and so on.
        Make ethernets warn about missing heartbeats and/or cables

25 years agoAdd a fast version of bcmp which compares longwords at a time.
dfr [Sat, 19 Jun 1999 16:30:28 +0000 (16:30 +0000)]
Add a fast version of bcmp which compares longwords at a time.

Submitted by: Peter Jeremy <jeremyp@gsmx07.alcatel.com.au>

25 years agoAdd again the ':' after the x option in th eargument list to getopt.
n_hibma [Sat, 19 Jun 1999 13:32:27 +0000 (13:32 +0000)]
Add again the ':' after the x option in th eargument list to getopt.

It disappeared in rev. 1.23 newfs.c

PR: 12292
Submitted by: Cy Schubert <cy@cschuber.net.gov.bc.ca>

25 years agoAdjust mcclock lines to avoid config(8) warning.
dfr [Sat, 19 Jun 1999 13:12:51 +0000 (13:12 +0000)]
Adjust mcclock lines to avoid config(8) warning.

25 years agoConvert signalling NaNs to quiet NaNs as specified by the Alpha Architecture
dfr [Sat, 19 Jun 1999 11:54:35 +0000 (11:54 +0000)]
Convert signalling NaNs to quiet NaNs as specified by the Alpha Architecture
Reference Manual.

Submitted by: Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp>

25 years agoAdd RCS strings to kernel ipfilter files.
green [Sat, 19 Jun 1999 11:35:41 +0000 (11:35 +0000)]
Add RCS strings to kernel ipfilter files.

25 years agosioprobe: Don't try to initialize disabled ports. In particular,
grog [Sat, 19 Jun 1999 08:14:56 +0000 (08:14 +0000)]
sioprobe: Don't try to initialize disabled ports.  In particular,
attempting to initialize sio3 caused problems with ATI video boards,
which cleverly share the I/O space.

PR: kern/12295

25 years agoThis should fix ipfilter for everyone it was broken for. CDEV_MAJOR is _not_
green [Sat, 19 Jun 1999 02:54:04 +0000 (02:54 +0000)]
This should fix ipfilter for everyone it was broken for. CDEV_MAJOR is _not_
-1.

Noticed by: users on freebsd-current

25 years agoAdd a transmit descriptor usage counter and use it to absolutely,
wpaul [Sat, 19 Jun 1999 00:36:56 +0000 (00:36 +0000)]
Add a transmit descriptor usage counter and use it to absolutely,
positively not let ti_encap() fill up the TX ring all the way and wrap
around. This fixes a potential transmit lockup where a really fast
machine (or particular TX traffic pattern) can overrun the end of the
ring.

Reported by: John Plevyak <jplevyak@inktomi.com>

25 years agoChange active flag handling; sanity check passed drive number; pass
rnordier [Fri, 18 Jun 1999 21:49:18 +0000 (21:49 +0000)]
Change active flag handling; sanity check passed drive number; pass
partition pointer to bootstrap; optimise away 10 bytes; cosmetics.

25 years agoK6-family MTRR support
green [Fri, 18 Jun 1999 19:56:02 +0000 (19:56 +0000)]
K6-family MTRR support

This is tested, but I really can't say whether it works entirely. I
don't know exactly what to look for when testing it. So let's say this
is open for testing. Send any results to green@FreeBSD.org

Reviewed by: msmith (long ago)

25 years agoHarmless change to prevent possible problems in the future. I made
green [Fri, 18 Jun 1999 19:24:40 +0000 (19:24 +0000)]
Harmless change to prevent possible problems in the future. I made
sure that i686_mem was only used when
1. CPUID had MTRR set (this was there before)
2. the CPU was GenuineIntel (not there)
3. the CPU is a 686 (also not there)

This should prevent any problems with CPUs that set MTRR but aren't
compatibile with Intel's interface (none that I know of yet.)

25 years agoClarified info on X release 3.3.1.
n_hibma [Fri, 18 Jun 1999 16:53:33 +0000 (16:53 +0000)]
Clarified info on X release 3.3.1.

Submitted/Reviewed by: Joe Touch <touch@ISI.EDU>

25 years agoNew parallel port support for PC98. Old PC98s which have
kato [Fri, 18 Jun 1999 14:48:28 +0000 (14:48 +0000)]
New parallel port support for PC98.  Old PC98s which have
uni-directional parallel port should use olpt driver instead of lpt
driver.

Files ppc.c and ppcreg.h are copied form i386/isa directory with PC98
change.

Submitted by: Akio Morita <amorita@meadow.scphys.kyoto-u.ac.jp>

25 years agoChanged the global `idt' from an array to a pointer so that npx.c
bde [Fri, 18 Jun 1999 14:32:21 +0000 (14:32 +0000)]
Changed the global `idt' from an array to a pointer so that npx.c
automatically hacks on the active copy of the IDT if f00f_hack()
has changed it.  This also allows simplifications in setidt().
This fixes breakage of FP exception handling by rev.1.55 of
sys/kernel.h.  FP exceptions were sent to npx.c's probe handlers
because npx.c "restored" the old handlers to the wrong copy of the
IDT.  The SYSINIT for f00f_hack() was purposely run quite late to
avoid problems like this, but it is bogusly associated with the
SYSINIT for proc0 so it was moved with the latter.

Problem reported and fix tested by:  Martin Cracauer <cracauer@cons.org>

25 years agoFix a coredump with negative (due to overflow) offset values.
ru [Fri, 18 Jun 1999 14:26:07 +0000 (14:26 +0000)]
Fix a coredump with negative (due to overflow) offset values.

25 years agoDon't use the ``next'' redial timer if we have no phone number at all.
brian [Fri, 18 Jun 1999 13:49:01 +0000 (13:49 +0000)]
Don't use the ``next'' redial timer if we have no phone number at all.

25 years agoThis lets you specify "gemdosfs", "shortnames", "longnames", and
jkh [Fri, 18 Jun 1999 12:05:02 +0000 (12:05 +0000)]
This lets you specify "gemdosfs", "shortnames", "longnames", and
"nowin95" as arguments to the "-o" flag, as alternatives to "-G", "-s",
"-l", and "-9"; when running "mount_msdos" by hand, that doesn't let you
do anything you couldn't already do, but if you're letting "mount" run
it, it lets you specify those options, which is especially useful if,
for example, you have an entry in "/etc/fstab" for some file system,
with "noauto" set, so you can conveniently mount a DOS partition from a
removable drive and force it to treat the file system as VFAT rather
than boring old FAT.

Submitted by: Guy Harris <guy@netapp.com>

25 years agoTypo in previous commit.
ru [Fri, 18 Jun 1999 10:41:13 +0000 (10:41 +0000)]
Typo in previous commit.

25 years agoMention that pid 1 is not affected by kill(-1, SIGXXX) too.
ru [Fri, 18 Jun 1999 10:24:12 +0000 (10:24 +0000)]
Mention that pid 1 is not affected by kill(-1, SIGXXX) too.

25 years agoBring in System V run-level patches (turned off by default).
ru [Fri, 18 Jun 1999 09:08:09 +0000 (09:08 +0000)]
Bring in System V run-level patches (turned off by default).
While I'm here, fix some typos in the manpage.

Requested by: des

25 years agoPostpone call to usage() until after all options are processed so that
sheldonh [Fri, 18 Jun 1999 08:52:13 +0000 (08:52 +0000)]
Postpone call to usage() until after all options are processed so that
the -d debugging option produces consistent behaviour.

25 years agoSync with etc.i386/MAKEDEV version 1.199.
dfr [Fri, 18 Jun 1999 06:20:08 +0000 (06:20 +0000)]
Sync with etc.i386/MAKEDEV version 1.199.

25 years agoOn our final pass through ffs_fsync, do all I/O synchronously so that
mckusick [Fri, 18 Jun 1999 05:49:46 +0000 (05:49 +0000)]
On our final pass through ffs_fsync, do all I/O synchronously so that
we can find out if our flush is failing because of write errors. This
change avoids a "flush failed" panic during unrecoverable disk errors.

25 years agoAdd a MAINTAINER line to modules that I maintain.
ken [Fri, 18 Jun 1999 00:41:47 +0000 (00:41 +0000)]
Add a MAINTAINER line to modules that I maintain.

Suggested by: brian, markm

25 years agoReviewed by: the cast of thousands
green [Thu, 17 Jun 1999 23:54:50 +0000 (23:54 +0000)]
Reviewed by: the cast of thousands

This is the change to struct sockets that gets rid of so_uid and replaces
it with a much more useful struct pcred *so_cred. This is here to be able
to do socket-level credential checks (i.e. IPFW uid/gid support, to be added
to HEAD soon). Along with this comes an update to pidentd which greatly
simplifies the code necessary to get a uid from a socket. Soon to come:
a sysctl() interface to finding individual sockets' credentials.

25 years agoAdd a kernel config for PC-Card machines. This is part of the
markm [Thu, 17 Jun 1999 23:53:20 +0000 (23:53 +0000)]
Add a kernel config for PC-Card machines. This is part of the
(safe) PAO integration.

Submitted by: Tatsumi HOSOKAWA

25 years agoMinor English corrections were made; SEE ALSO was also fixed (no commas).
green [Thu, 17 Jun 1999 23:43:35 +0000 (23:43 +0000)]
Minor English corrections were made; SEE ALSO was also fixed (no commas).

25 years agoAdd Id strings
gpalmer [Thu, 17 Jun 1999 23:42:45 +0000 (23:42 +0000)]
Add Id strings

25 years agoReset our start time when the link comes up, correcting
brian [Thu, 17 Jun 1999 22:12:51 +0000 (22:12 +0000)]
Reset our start time when the link comes up, correcting
the reported connection time.

PR: 11943
Submitted by: Peter Mutsaers <plm@xs4all.nl>

25 years agoMore PAO-ish stuff. Wildcard the IRQ bits.
markm [Thu, 17 Jun 1999 21:39:57 +0000 (21:39 +0000)]
More PAO-ish stuff. Wildcard the IRQ bits.

25 years agovm_map_growstack uses vmspace::vm_ssize as though it contained
alc [Thu, 17 Jun 1999 21:29:38 +0000 (21:29 +0000)]
vm_map_growstack uses vmspace::vm_ssize as though it contained
the stack size in bytes when in fact it is the stack size in pages.

25 years agoMore uncontroversial PAO updates. Make the IRQ wildcarding work a lot better,
markm [Thu, 17 Jun 1999 21:07:59 +0000 (21:07 +0000)]
More uncontroversial PAO updates. Make the IRQ wildcarding work a lot better,
and improve the daemonizing code.

Submitted by: Tatsumi HOSOKAWA

25 years agoUpdate the comments on values than can be returned by DEVICE_PROBE.
n_hibma [Thu, 17 Jun 1999 19:22:12 +0000 (19:22 +0000)]
Update the comments on values than can be returned by DEVICE_PROBE.
DEVICE_PROBE can return priorities.

Reviewed by: Doug Rabson <dfr@nlsystems.com>

25 years agoAdd bits of PAO that are non-controversial.
markm [Thu, 17 Jun 1999 19:04:56 +0000 (19:04 +0000)]
Add bits of PAO that are non-controversial.

Submitted by: Tatsumi HOSOKAWA

25 years agoPlease forgive me, it's my first day on the block. Forgot to mention:
sheldonh [Thu, 17 Jun 1999 14:49:19 +0000 (14:49 +0000)]
Please forgive me, it's my first day on the block. Forgot to mention:

PR: 11987

25 years agoNull commit to reflect that the previous commit added warnings for
sheldonh [Thu, 17 Jun 1999 14:48:02 +0000 (14:48 +0000)]
Null commit to reflect that the previous commit added warnings for
previously unreported malloc() failures, and changed the exit status
for those failures to encourage MTA retries.

25 years ago* Add a command-line option to enable debugging mode, which sends error
sheldonh [Thu, 17 Jun 1999 14:44:42 +0000 (14:44 +0000)]
* Add a command-line option to enable debugging mode, which sends error
  messages to stderr instead of syslog.

* Fix manpage with respect to diagnostics.

Reported by: rfg@monkeys.com (Ronald F. Guilmette)

25 years agoSync with sys/i386/i386/machdep.c revision up to 1.341.
kato [Thu, 17 Jun 1999 11:09:39 +0000 (11:09 +0000)]
Sync with sys/i386/i386/machdep.c revision up to 1.341.

25 years agoSync with sys/i386/conf/options.i386 revision 1.117.
kato [Thu, 17 Jun 1999 11:07:52 +0000 (11:07 +0000)]
Sync with sys/i386/conf/options.i386 revision 1.117.

25 years agoMerge from non-crypto version:
ru [Thu, 17 Jun 1999 09:24:37 +0000 (09:24 +0000)]
Merge from non-crypto version:
- "-N" option
- "-E" security fix
- "-s src_addr" option

Requested by: markm

25 years agoVarious fixes for inetd's TCP Wrappers support:
sheldonh [Thu, 17 Jun 1999 09:16:08 +0000 (09:16 +0000)]
Various fixes for inetd's TCP Wrappers support:

1) Handle forking and non-forking internal services correctly.
   Turn on wrapping for internal services because it works now.
2) Preserve server names for each service on HUP.
3) Honour hosts_options(5) severity option.
4) Add IMPLEMENTATION NOTES section to clarify TCP Wrappers
   usage and limitations.

This change may cause previously allowed builtin services (e.g. daytime)
to be denied in existing configurations.

PR: 12097
Reviewed by: markm
1)
Reported by: Pierre Beyssac <pb@fasterix.freenix.org>
2)
Submitted by: Masachika ISHIZUKA <ishizuka@ish.org>
3)
Submitted by: David Malone <dwmalone@maths.tcd.ie>

25 years agoAdd "-N" option to the SYNOPSIS and usage().
ru [Thu, 17 Jun 1999 07:12:39 +0000 (07:12 +0000)]
Add "-N" option to the SYNOPSIS and usage().

25 years agoFixed a missing userland dev_t to kernel dev_t conversion.
bde [Thu, 17 Jun 1999 07:07:55 +0000 (07:07 +0000)]
Fixed a missing userland dev_t to kernel dev_t conversion.

25 years agovm_map_insert sometimes extends an existing vm_map entry, rather than
alc [Thu, 17 Jun 1999 05:49:00 +0000 (05:49 +0000)]
vm_map_insert sometimes extends an existing vm_map entry, rather than
creating a new entry.  vm_map_stack and vm_map_growstack can panic when
a new entry isn't created.  Fixed vm_map_stack and vm_map_growstack.

Also, when extending the stack, always set the protection to VM_PROT_ALL.

25 years agoReformat comment to match indentation of code around it.
julian [Thu, 17 Jun 1999 01:25:25 +0000 (01:25 +0000)]
Reformat comment to match indentation of code around it.

25 years agoMove vm_map_stack and vm_map_growstack after the definition
alc [Thu, 17 Jun 1999 00:39:26 +0000 (00:39 +0000)]
Move vm_map_stack and vm_map_growstack after the definition
of the vm_map_clip_end macro.  (The next commit will modify
vm_map_stack and vm_map_growstack to use vm_map_clip_end.)

25 years agoRemove some unused declarations and duplicate initialization.
alc [Thu, 17 Jun 1999 00:27:39 +0000 (00:27 +0000)]
Remove some unused declarations and duplicate initialization.

25 years agoAdd a vnode argument to VOP_BWRITE to get rid of the last vnode
mckusick [Wed, 16 Jun 1999 23:27:55 +0000 (23:27 +0000)]
Add a vnode argument to VOP_BWRITE to get rid of the last vnode
operator special case. Delete special case code from vnode_if.sh,
vnode_if.src, umap_vnops.c, and null_vnops.c.

25 years agoAdd the -p switch - tells rwhod to ignore POINTOPOINT interfaces.
brian [Wed, 16 Jun 1999 21:05:21 +0000 (21:05 +0000)]
Add the -p switch - tells rwhod to ignore POINTOPOINT interfaces.

Mostly submitted by: Stefan Zehl <sec@42.org>
PR: 12216

25 years agoInit(8) will halt the system if sent USR1 signal,
ru [Wed, 16 Jun 1999 20:01:19 +0000 (20:01 +0000)]
Init(8) will halt the system if sent USR1 signal,
or halt and turn the power off if sent SIGUSR2.

PR: 5451
Submitted by: Leif Neland <leifn@image.dk>
Reworked by: ru
Reviewed by: -hackers

25 years agoClose a race window where a tcp socket is closed while tcp_pcblist is
tegge [Wed, 16 Jun 1999 19:05:17 +0000 (19:05 +0000)]
Close a race window where a tcp socket is closed while tcp_pcblist is
copying out tcp socket info, causing a NULL pointer to be dereferenced.

25 years agoMake syscall tracing work: include "opt_ktrace.h" and make it compile.
dt [Wed, 16 Jun 1999 18:55:35 +0000 (18:55 +0000)]
Make syscall tracing work: include "opt_ktrace.h" and make it compile.

25 years agoSyscall arguments are now properly aligned. Print them, and syscall return
dt [Wed, 16 Jun 1999 18:46:40 +0000 (18:46 +0000)]
Syscall arguments are now properly aligned. Print them, and syscall return
values, as longs, instead of int.

25 years agoMake sure syscall arguments properly aligned in ktrace records.
dt [Wed, 16 Jun 1999 18:37:01 +0000 (18:37 +0000)]
Make sure syscall arguments properly aligned in ktrace records.

Make syscall return value a register_t.

Based on a patch from Hidetoshi Shimokawa.
Mostly reviewed by: Hidetoshi Shimokawa and Bruce Evans.

25 years agoUse vput instead of vrele.
mjacob [Wed, 16 Jun 1999 18:35:58 +0000 (18:35 +0000)]
Use vput instead of vrele.
Reviewed by: Matthew Dillon <dillon@apollo.backplane.com>
Submitted by: Ville-Pertti Keinonen <will@iki.fi>
Obtained from: Matthew Dillon <dillon@apollo.backplane.com>

25 years agoRemove option USBVERBOSE
n_hibma [Wed, 16 Jun 1999 17:34:36 +0000 (17:34 +0000)]
Remove option USBVERBOSE
Add the options for debugging used in dev/usb/*.c

Submitted by: Kazu

25 years agoGrrr! The PNIC II's multicast hash table is only 128 bits wide, not 512
wpaul [Wed, 16 Jun 1999 16:27:30 +0000 (16:27 +0000)]
Grrr! The PNIC II's multicast hash table is only 128 bits wide, not 512
like the original PNIC and the MX98715A (from which the PNIC II is derived).
This requires special handling. Save the card type, and in mx_calchash(),
if we see that the card is a PNIC, return only the low 7 bits of the
hash instead of the low 9 bits.

25 years agoChanged trypbuf to a getpbuf to work around a problem where redundant writes
dg [Wed, 16 Jun 1999 15:54:30 +0000 (15:54 +0000)]
Changed trypbuf to a getpbuf to work around a problem where redundant writes
would occur when clustering them - caused by running out of buffers
and taking a degenerate code path as a result. It appears that waiting
instead for buffers to become available is okay.

Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
Discovered by: Craig A Soules <soules+@andrew.cmu.edu>

25 years agoAdd "-s src_addr" option to allow setting of the source IP address.
ru [Wed, 16 Jun 1999 15:26:22 +0000 (15:26 +0000)]
Add "-s src_addr" option to allow setting of the source IP address.

25 years agoPrint verbose messages when there is error.
yokota [Wed, 16 Jun 1999 14:04:45 +0000 (14:04 +0000)]
Print verbose messages when there is error.
Oked by: des

25 years agoFix non-benign typo.
billf [Wed, 16 Jun 1999 12:26:40 +0000 (12:26 +0000)]
Fix non-benign typo.

Submitted by: Daniel Baker <dbaker@cuckoo.com>

25 years agoClean up bitrot in interrupt tracing code.
tegge [Wed, 16 Jun 1999 03:53:52 +0000 (03:53 +0000)]
Clean up bitrot in interrupt tracing code.

25 years agoIf we still haven't got a sufficient number of free buffers after the
tegge [Wed, 16 Jun 1999 03:19:04 +0000 (03:19 +0000)]
If we still haven't got a sufficient number of free buffers after the
call to flushdirtybuffers() then sleep in waitfreebuffers().
PR: 11697
Reviewed by: David Greenman, Matt Dillon

25 years agoGet rid of the global variable rushjob and replace it with a function in
mckusick [Tue, 15 Jun 1999 23:37:29 +0000 (23:37 +0000)]
Get rid of the global variable rushjob and replace it with a function in
kern/vfs_subr.c named speedup_syncer() which handles the speedup request.
Change the various clients of rushjob to use the new function.

25 years agoIf we retry this operation from the top of this routine, we need to
mjacob [Tue, 15 Jun 1999 23:24:14 +0000 (23:24 +0000)]
If we retry this operation from the top of this routine, we need to
make sure we've freed any allocated resources (to avoid a memory leak)
and and do the right thing with respect to the nfs node hash lock we'd
acquired.

25 years agoAdd a couple of command definitions.
julian [Tue, 15 Jun 1999 23:13:09 +0000 (23:13 +0000)]
Add a couple of command definitions.
Also add offsets into the IDE parameter block so that it is humanly
possible to match the structure to the manufacturer's documentation.
(basically this is just changes to comments)

25 years agoCOM_LLCONSOLE should over-ride COM_DEBUGGER request.
mckusick [Tue, 15 Jun 1999 22:15:10 +0000 (22:15 +0000)]
COM_LLCONSOLE should over-ride COM_DEBUGGER request.

25 years agocam_get_device() was returning 0 on failure, and 1 on success, while
mpp [Tue, 15 Jun 1999 20:03:01 +0000 (20:03 +0000)]
cam_get_device() was returning 0 on failure, and 1 on success, while
camcontrol(8) and the documentation in camlib.c and cam(3) all expect
-1 on failure and 0 on success.  Updated camlib.c to return the values
specified by the documentation.

PR: 12023

25 years agoCorrectly document the two different object file formats that can be
mpp [Tue, 15 Jun 1999 19:19:51 +0000 (19:19 +0000)]
Correctly document the two different object file formats that can be
used with objformat.

PR: 10877

25 years agoDocument the IMPLEMENTATION NOTES section since we have a number of
mpp [Tue, 15 Jun 1999 19:03:54 +0000 (19:03 +0000)]
Document the IMPLEMENTATION NOTES section since we have a number of
man pages that use it.

PR: 10659

25 years agoThis really belongs to section 8.
ru [Tue, 15 Jun 1999 14:04:19 +0000 (14:04 +0000)]
This really belongs to section 8.

25 years agoKill option FAILSAFE.
des [Tue, 15 Jun 1999 13:14:56 +0000 (13:14 +0000)]
Kill option FAILSAFE.

PR: i386/12187
Approved by: bde

25 years agoDocument the usage of escape character in a service name.
ru [Tue, 15 Jun 1999 12:56:38 +0000 (12:56 +0000)]
Document the usage of escape character in a service name.

PR: 7101
Reminded by: jhs

25 years agoMFS: No longer needed with PicoBSD 0.44
roger [Tue, 15 Jun 1999 12:51:57 +0000 (12:51 +0000)]
MFS: No longer needed with PicoBSD 0.44

25 years agoMFS: No longer needed in PicoBSD 0.44
roger [Tue, 15 Jun 1999 12:40:32 +0000 (12:40 +0000)]
MFS: No longer needed in PicoBSD 0.44

25 years agoAdd VIA Apollo Master VT82c570 chipset. (I have 2 of these)
roger [Tue, 15 Jun 1999 12:01:26 +0000 (12:01 +0000)]
Add VIA Apollo Master VT82c570 chipset. (I have 2 of these)
Also, Apollo IDE controller has device ID of 1571 and 0571

25 years agoMFS: No longer needed in PicoBSD 0.44
roger [Tue, 15 Jun 1999 11:45:25 +0000 (11:45 +0000)]
MFS: No longer needed in PicoBSD 0.44

25 years agoMFS: No longer needed in PicoBSD 0.44
roger [Tue, 15 Jun 1999 11:40:13 +0000 (11:40 +0000)]
MFS: No longer needed in PicoBSD 0.44

25 years agoEnsoniq AudioPCI sound cards haven't worked since the newbus integration
newton [Tue, 15 Jun 1999 00:00:10 +0000 (00:00 +0000)]
Ensoniq AudioPCI sound cards haven't worked since the newbus integration
because they haven't been able to attach.  Now fixed.

Submitted by: Christian Weisgerber <naddy@mips.rhein-neckar.de>

25 years agoAdd the System V message queue and semaphore options for better 3rd party
jb [Mon, 14 Jun 1999 22:59:55 +0000 (22:59 +0000)]
Add the System V message queue and semaphore options for better 3rd party
software support out-of-the-box (like i386 does).

25 years agoSupport Alpha and -CURRENT in the "-r" option case.
obrien [Mon, 14 Jun 1999 21:58:59 +0000 (21:58 +0000)]
Support Alpha and -CURRENT in the "-r" option case.

25 years agoInclude <sys/time.h> for correctness. BDE has a better version of this,
nik [Mon, 14 Jun 1999 19:03:46 +0000 (19:03 +0000)]
Include <sys/time.h> for correctness.  BDE has a better version of this,
but it's more complex, and in his words

    Commit your version, since it is the only one that is clearly permitted
    (if not best), and I'll untangle it later.

PR: docs/11589
Reviewed by: Bruce "he kicks ass" Evans

25 years agoIt was possible to circumvent telnet's security ``-E'' (no escape at all)
ru [Mon, 14 Jun 1999 18:46:15 +0000 (18:46 +0000)]
It was possible to circumvent telnet's security ``-E'' (no escape at all)
option by running it with ``-8 -E'' and passing it 0xff character.

PR: 6317
Reviewed by: des, David Borman <dab@bsdi.com> (the author)

25 years agoLast commit was incorrect. Remove the 'rp' PCI device
roger [Mon, 14 Jun 1999 18:16:51 +0000 (18:16 +0000)]
Last commit was incorrect. Remove the 'rp' PCI device

25 years agoComment out the PCI Comtrol Rocketport entries.
roger [Mon, 14 Jun 1999 18:08:17 +0000 (18:08 +0000)]
Comment out the PCI Comtrol Rocketport entries.
The PicoBSD 'isp' build fails when both the PCI and ISA versions of the
'rp' driver are included at the same time