]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
24 years agoMerge conflicts. This fixes the i386 config bits, merge in the Alpha bits,
obrien [Tue, 20 Jun 2000 08:13:26 +0000 (08:13 +0000)]
Merge conflicts.  This fixes the i386 config bits, merge in the Alpha bits,
and adds arm32 & sparc bits.

24 years agoBinutils now properly supports our ELFness with this file.
obrien [Tue, 20 Jun 2000 08:10:58 +0000 (08:10 +0000)]
Binutils now properly supports our ELFness with this file.
(use the stock file now that it is offically part of Binutils)

24 years agoUse the stock file now. The Binutils maintainers accepted our rev 1.2
obrien [Tue, 20 Jun 2000 07:00:04 +0000 (07:00 +0000)]
Use the stock file now.  The Binutils maintainers accepted our rev 1.2
changes (don't use "/" to start a comment).

24 years agoMerge in our rev 1.2 changes (but do it in a way the Binutils maintainers
obrien [Tue, 20 Jun 2000 06:55:50 +0000 (06:55 +0000)]
Merge in our rev 1.2 changes (but do it in a way the Binutils maintainers
accepted): Restore traditional behavior by acknowledging tabs in strings.

24 years agoMerge in our ELF branding.
obrien [Tue, 20 Jun 2000 06:46:06 +0000 (06:46 +0000)]
Merge in our ELF branding.

24 years agoThis commit was generated by cvs2svn to compensate for changes in r61843,
obrien [Tue, 20 Jun 2000 06:20:42 +0000 (06:20 +0000)]
This commit was generated by cvs2svn to compensate for changes in r61843,
which included commits to RCS files with non-trunk default branches.

24 years agoImport a virgin (but trimed) Binutils 2.10 release.
obrien [Tue, 20 Jun 2000 06:20:42 +0000 (06:20 +0000)]
Import a virgin (but trimed) Binutils 2.10 release.

24 years agoWe are now at version 2.10 release.
obrien [Tue, 20 Jun 2000 06:19:15 +0000 (06:19 +0000)]
We are now at version 2.10 release.

24 years agoNone of these files are used in our build, nor do I really believe
obrien [Tue, 20 Jun 2000 06:08:35 +0000 (06:08 +0000)]
None of these files are used in our build, nor do I really believe
anyone is going to read them.  Many were new with the
"sourceware_binutils-2_10-branch_anoncvs_20000512" import.  Others have
been around since 2.8.1.

24 years agoRemove these Binutils 2.8.1 files.
obrien [Tue, 20 Jun 2000 06:03:50 +0000 (06:03 +0000)]
Remove these Binutils 2.8.1 files.

24 years agoAdd implementation of fetchListFTP.
wes [Tue, 20 Jun 2000 05:32:41 +0000 (05:32 +0000)]
Add implementation of fetchListFTP.

Reviewed by: silence

24 years agoreturn of the accept filter part II
alfred [Tue, 20 Jun 2000 01:09:23 +0000 (01:09 +0000)]
return of the accept filter part II

accept filters are now loadable as well as able to be compiled into
the kernel.

two accept filters are provided, one that returns sockets when data
arrives the other when an http request is completed (doesn't work
with 0.9 requests)

Reviewed by: jmg

24 years agoComment this bad boy. Hopefully the next person that comes along won't
jhb [Mon, 19 Jun 2000 22:50:46 +0000 (22:50 +0000)]
Comment this bad boy.  Hopefully the next person that comes along won't
have to spend a few hours reading the code to figure all this out.

24 years agoTerminate aic_ids[]
brian [Mon, 19 Jun 2000 22:16:14 +0000 (22:16 +0000)]
Terminate aic_ids[]

24 years agoAdd UP1000 to GENERIC
gallatin [Mon, 19 Jun 2000 21:55:42 +0000 (21:55 +0000)]
Add UP1000 to GENERIC

24 years agounbreak kernels without UP1000 support
gallatin [Mon, 19 Jun 2000 21:49:47 +0000 (21:49 +0000)]
unbreak kernels without UP1000 support

24 years agoDon't mis-match interface names in iface_Create()
brian [Mon, 19 Jun 2000 21:43:48 +0000 (21:43 +0000)]
Don't mis-match interface names in iface_Create()

Submitted by: Adrian Penisoara <ady@freebsd.ady.ro>

24 years agoAdd support for the Alpha Processor, Inc. UP1000 system.
gallatin [Mon, 19 Jun 2000 21:15:45 +0000 (21:15 +0000)]
Add support for the Alpha Processor, Inc. UP1000 system.

Reviewed by: dfr
Thanks to:  Alpha Processor Inc. for supplying the hardware.

24 years agoGrrr. I hate CVS. These were supposed to be committed when I did the
markm [Mon, 19 Jun 2000 21:09:27 +0000 (21:09 +0000)]
Grrr. I hate CVS. These were supposed to be committed when I did the
IDEA fix earlier today.

Bring back IDEA from the dead (but not compiled by default).

24 years agomake mixer reads return the value written instead of the value set
cg [Mon, 19 Jun 2000 20:31:58 +0000 (20:31 +0000)]
make mixer reads return the value written instead of the value set

people seem to want this even though it breaks oss spec compliance

24 years agoSupport bounce buffers for ISA DMA on the alpha. This is required for the
gallatin [Mon, 19 Jun 2000 18:41:27 +0000 (18:41 +0000)]
Support bounce buffers for ISA DMA on the alpha.  This is required for the
irongate chipset (used in the UP1000) which does not support scatter/gather
DMA.  We'll still use scatter gather if the core logic chipset supports it.

Reviewed by: dfr

24 years agoRe-commit DNS IPv6 transport support with fixes for IPv4 only
ume [Mon, 19 Jun 2000 18:25:06 +0000 (18:25 +0000)]
Re-commit DNS IPv6 transport support with fixes for IPv4 only
kernel and compatibility issue.

Obtained from: KAME Project

24 years ago- Add some more details to the 'lsdev' output for FAT file systems.
jhb [Mon, 19 Jun 2000 17:44:40 +0000 (17:44 +0000)]
- Add some more details to the 'lsdev' output for FAT file systems.
- Add in support for the EDD (Enhanced Disk Drive) BIOS extensions to
  use LBA mode for accessing drives past cylinder 1024.  This should allow
  us to load a kernel from anywhere on a newer drive up to 2 TB.  Part
  of this came from the PR below.

PR: i386/13847
Submitted by: Tor Egge <Tor.Egge@fast.no>

24 years agoFix a comment botch which broke buildworld.
green [Mon, 19 Jun 2000 15:27:39 +0000 (15:27 +0000)]
Fix a comment botch which broke buildworld.

24 years agoRe-add IDEA. This is not actually built unless asked for by the user.
markm [Mon, 19 Jun 2000 13:59:34 +0000 (13:59 +0000)]
Re-add IDEA. This is not actually built unless asked for by the user.
(To avoid patent hassles).

24 years agoImplemented some optimizations which result in 14 fewer instructions in the
dg [Mon, 19 Jun 2000 00:58:34 +0000 (00:58 +0000)]
Implemented some optimizations which result in 14 fewer instructions in the
receive path.

24 years agoBring the an(4) fixes to wi(4):
roberto [Mon, 19 Jun 2000 00:17:13 +0000 (00:17 +0000)]
Bring the an(4) fixes to wi(4):
- suser check
- splx() fix.

Reminded by: Aaron Campbell <aaron@openbsd.org>

24 years agoBring the ancontrol buffer oflows fixes here too.
roberto [Sun, 18 Jun 2000 23:45:17 +0000 (23:45 +0000)]
Bring the ancontrol buffer oflows fixes here too.

Submitted by: Aaron Campbell <aaron@openbsd.org>

24 years ago- Add suser check before SIOCSAIRONET.
roberto [Sun, 18 Jun 2000 23:40:09 +0000 (23:40 +0000)]
- Add suser check before SIOCSAIRONET.
- Fix a splimp() w/o splx bug in the ioctl routine while I'm here.

Submitted by: Aaron Campbell <aaron@openbsd.org>

24 years agoFix potential buffer overflows (even if ancontrol is not setuid).
roberto [Sun, 18 Jun 2000 23:10:20 +0000 (23:10 +0000)]
Fix potential buffer overflows (even if ancontrol is not setuid).

Submitted by: Aaron Campbell <aaron@cs.dal.ca> from OpenBSD

24 years agoDon't assume that the output of strftime for "%c" ("national
joe [Sun, 18 Jun 2000 22:18:04 +0000 (22:18 +0000)]
Don't assume that the output of strftime for "%c" ("national
representation of time and date") won't change in time.  Instead
of hard coding the locations of the time elements and hoping that
they don't move use strftime to generate the desired formats in
the first place.

PR: bin/7826

24 years agoWhen running with quotas enabled on a filesystem using soft updates,
mckusick [Sun, 18 Jun 2000 22:14:28 +0000 (22:14 +0000)]
When running with quotas enabled on a filesystem using soft updates,
the system would panic when a user's inode quota was exceeded (see
PR 18959 for details). This fixes that problem.

PR: 18959
Submitted by: Jason Godsey <jason@unixguy.fidalgo.net>

24 years agoSome additional performance improvements. When freeing an inode
mckusick [Sun, 18 Jun 2000 22:05:57 +0000 (22:05 +0000)]
Some additional performance improvements. When freeing an inode
check to see if it has been committed to disk. If it has never
been written, it can be freed immediately. For short lived files
this change allows the same inode to be reused repeatedly.
Similarly, when upgrading a fragment to a larger size, if it
has never been claimed by an inode on disk, it too can be freed
immediately making it available for reuse often in the next slowly
growing block of the same file.

24 years agoRemove RECERR from RAY_DEBUG
dmlb [Sun, 18 Jun 2000 21:41:24 +0000 (21:41 +0000)]
Remove RECERR from RAY_DEBUG

24 years agoGeneral tidy up and plough through TODO list.
dmlb [Sun, 18 Jun 2000 21:40:46 +0000 (21:40 +0000)]
General tidy up and plough through TODO list.

Rewrote intro at top of file to reflect my better understanding of how it
the memory mapping works.

Clear the DONE list and move some thoughts into the TODO list.

Remove RECERR from RAY_DEBUG

Start to use a desired network parameter structure, only used in download
code as I've realised that there are some problems with the idea.

Break up ray_rx, and move the data packet handler into a seperate function. This meant some knock on changes in ray_rx_mgt/ray_rx_ctl to do with
mbuf freeing.

Remove some debug code/XXX comments that are out of date.

24 years agoUpgrade instructions for the sourceware_binutils-2_10-branch_anoncvs_20000512
obrien [Sun, 18 Jun 2000 21:32:11 +0000 (21:32 +0000)]
Upgrade instructions for the sourceware_binutils-2_10-branch_anoncvs_20000512
Binutils import.

24 years agoList of files to delete from the Sourceware Binutils' CVS 2_10-branch
obrien [Sun, 18 Jun 2000 21:17:40 +0000 (21:17 +0000)]
List of files to delete from the Sourceware Binutils' CVS 2_10-branch
on 12-May-2000.

24 years agoList of distfile contents to delete using `rm' and `find -delete'.
obrien [Sun, 18 Jun 2000 21:15:52 +0000 (21:15 +0000)]
List of distfile contents to delete using `rm' and `find -delete'.
Simular to the FREEBSD-Xlist used with `tar -X ... -xf'.
This file is typically used when one starts with the Binutils anoncvs repo
rather than a release tarball.

24 years agoAdd a desired network parameter structure to runq entries.
dmlb [Sun, 18 Jun 2000 21:10:58 +0000 (21:10 +0000)]
Add a desired network parameter structure to runq entries.

24 years agoFix building with -DDEBUG.
roberto [Sun, 18 Jun 2000 20:22:11 +0000 (20:22 +0000)]
Fix building with -DDEBUG.

24 years agoRemove the setflags/getflags routines. Their functionality has
joe [Sun, 18 Jun 2000 20:10:41 +0000 (20:10 +0000)]
Remove the setflags/getflags routines.  Their functionality has
been replaced with the library calls fflagstostr and strtofflags.

24 years ago/152x/s/sound/SCSI/
phk [Sun, 18 Jun 2000 15:01:22 +0000 (15:01 +0000)]
/152x/s/sound/SCSI/

24 years agoAdded support for the i82559ER (10/100Mbps NIC for embedded applications).
dg [Sun, 18 Jun 2000 10:26:09 +0000 (10:26 +0000)]
Added support for the i82559ER (10/100Mbps NIC for embedded applications).

Product device ID provided by: Les Biffle <les@ns3.safety.net>

24 years agoFix a printf-style format error
brian [Sun, 18 Jun 2000 10:07:53 +0000 (10:07 +0000)]
Fix a printf-style format error

24 years agobackout accept optimizations.
alfred [Sun, 18 Jun 2000 08:49:13 +0000 (08:49 +0000)]
backout accept optimizations.

Requested by: jmg, dcs, jdp, nate

24 years agoMF4: add support for the Am79C973.
gj [Sun, 18 Jun 2000 08:12:54 +0000 (08:12 +0000)]
MF4: add support for the Am79C973.

24 years agoRemove all but ISP_TARGET_MODE options for isp and ispfw pseudo device.
mjacob [Sun, 18 Jun 2000 06:59:13 +0000 (06:59 +0000)]
Remove all but ISP_TARGET_MODE options for isp and ispfw pseudo device.

24 years agoAllow newer Linksys 10/100 PCMCIA cards to work.
ps [Sun, 18 Jun 2000 05:50:16 +0000 (05:50 +0000)]
Allow newer Linksys 10/100 PCMCIA cards to work.

Reviewed by: imp

24 years agoAdd ispfw as an optional 'device' (for static linking)
mjacob [Sun, 18 Jun 2000 05:35:07 +0000 (05:35 +0000)]
Add ispfw as an optional 'device' (for static linking)

24 years agoMatching commits to pccard for last pcic changes. We now at least to
imp [Sun, 18 Jun 2000 05:28:59 +0000 (05:28 +0000)]
Matching commits to pccard for last pcic changes.  We now at least to
probe/attach.  This is a checkpoint.

24 years agoAlmost make loading work. This is a checkpoint. With these change we
imp [Sun, 18 Jun 2000 05:25:30 +0000 (05:25 +0000)]
Almost make loading work.  This is a checkpoint.  With these change we
can almost kldload this.  More work is ncessary, but I wanted to
checkpoint this now.

24 years agoAdd oldcard, pccard, pcic and sn modules to the build.
imp [Sun, 18 Jun 2000 05:20:26 +0000 (05:20 +0000)]
Add oldcard, pccard, pcic and sn modules to the build.

24 years agoClean up firmware load issues and remove darn near all config options.
mjacob [Sun, 18 Jun 2000 05:18:55 +0000 (05:18 +0000)]
Clean up firmware load issues and remove darn near all config options.
Force alphas to prefer mem mapping as the default.

Basically, we have a pointer to a function which we can call which will
return us a pointer to firmware for the card we have. We call this function
(if it's non-NULL) with the address of our mdvec f/w pointer.

The way this works is that if ispfw (as a module or a static) is loaded,
it initializes the pointer in isp_pci, so we can call into to it to fetch
a pointer to a f/w set.

If ispfw is MOD_UNLOADed, it's retained a pointer to our mdvec f/w pointers,
which then get zeroed out so we don't have any references to data that's
now gone from kernel memory. Removing the f/w saves ~360KBytes.

Alas, there is no autounload mechanism that works for is here.

24 years agoRemove all but one (ISP_TARGET_MODE) option for isp.
mjacob [Sun, 18 Jun 2000 05:15:01 +0000 (05:15 +0000)]
Remove all but one (ISP_TARGET_MODE) option for isp.

24 years agoPreliminary commit of oldcard module. This might not work, but will.
imp [Sun, 18 Jun 2000 05:08:57 +0000 (05:08 +0000)]
Preliminary commit of oldcard module.  This might not work, but will.
This should allow one to load oldcard or newcard for testing.  Please
let me know if this doesn't work.  Don't load this and either of pcic
or pccard.  I've not tried it, but I suspect bad things will happen.

24 years agoadd ispfw module
mjacob [Sun, 18 Jun 2000 05:08:36 +0000 (05:08 +0000)]
add ispfw module

24 years agoadd if_wx module
mjacob [Sun, 18 Jun 2000 05:08:28 +0000 (05:08 +0000)]
add if_wx module

24 years agoAdd if_wx && ispfw modules
mjacob [Sun, 18 Jun 2000 05:08:11 +0000 (05:08 +0000)]
Add if_wx && ispfw modules

24 years agoAdd new functions. Also add comments to existing functions. These
imp [Sun, 18 Jun 2000 05:02:09 +0000 (05:02 +0000)]
Add new functions.  Also add comments to existing functions.  These
are needed for the pccard changes I've just committed.

24 years agoRemoving this bulky one large f/w file. This f/w is now in dev/ispfw.
mjacob [Sun, 18 Jun 2000 04:59:47 +0000 (04:59 +0000)]
Removing this bulky one large f/w file. This f/w is now in dev/ispfw.

24 years agoo Implement some compatibility functions for NEWCARD compat. This is
imp [Sun, 18 Jun 2000 04:59:39 +0000 (04:59 +0000)]
o Implement some compatibility functions for NEWCARD compat.  This is
  incomplete, but will eventually allow the same drivers to function
  with both oldcard and newcard.
o Remove include of opt_bus.h.  It isn't needed and gets in the way of
  module building.

24 years agoOnce we have firmware running (if isp_reset) and this is the first time
mjacob [Sun, 18 Jun 2000 04:56:17 +0000 (04:56 +0000)]
Once we have firmware running (if isp_reset) and this is the first time
through, establish what our LUN width is. Unfortunately, we can't ask
the f/w. If we loaded the f/w, we'll now assume we have expanded LUNs
(SCCLUN for fibre channel, just plain 32 LUN for SCSI). If we didn't
load firmware, assume 8 LUNs for SCSI and 1 LUN for Fibre Channel. We
have to assume only one LUN for Fibre Channel because the LUN setting
in Request Queue entries is in different places whether we have SCCLUN
firmware or not, so the only LUN guaranteed to work for both is LUN 0.

Clean up the rest of isp.c so that ISP2100_SCCLUN defines aren't used-
instead use run time determinants based upon isp->isp_maxluns.

After starting firmware, delay 500us to give it a chance to get rolling.

Fix the interrupt service routine to check for both isr && sema being zero
before thinking this was a spurious interrupt.  Following the manuals,
allow for both Mailbox as well as Queue Reponse type interrupts for regular
SCSI.

24 years agoRemove all ISP2100_SCCLUN define protected code and replace it with
mjacob [Sun, 18 Jun 2000 04:50:26 +0000 (04:50 +0000)]
Remove all ISP2100_SCCLUN define protected code and replace it with
runtime checks.

24 years agoRemove all ISP2100_SCCLUN define based code and replace it with runtime
mjacob [Sun, 18 Jun 2000 04:48:28 +0000 (04:48 +0000)]
Remove all ISP2100_SCCLUN define based code and replace it with runtime
comparisons against the tag isp_maxluns- if > 16, we're SCCLUN based.

On initial regular SCSI startup, disable auto-disconnect.

24 years agoRoll platform minor number. Force definition of SCSI_ISP_FABRIC
mjacob [Sun, 18 Jun 2000 04:47:12 +0000 (04:47 +0000)]
Roll platform minor number. Force definition of SCSI_ISP_FABRIC
(we always support fabric now). Remove SCCLUN definition (we always
support SCCLUN now, if we load the f/w). Add typedef definition of an
external firmware fetch function.

24 years agoRoll core minor version. Set ISP_MAX_LUNS to be off of new isp_maxluns
mjacob [Sun, 18 Jun 2000 04:45:51 +0000 (04:45 +0000)]
Roll core minor version. Set ISP_MAX_LUNS to be off of new isp_maxluns
tag in softc.

24 years agoadd "disable autodisconnect" flags
mjacob [Sun, 18 Jun 2000 04:44:41 +0000 (04:44 +0000)]
add "disable autodisconnect" flags

24 years agoMake pcic compile again with newcard
imp [Sun, 18 Jun 2000 04:44:40 +0000 (04:44 +0000)]
Make pcic compile again with newcard

24 years agocleanup i_int_X vs. uint_X definitions
mjacob [Sun, 18 Jun 2000 04:43:55 +0000 (04:43 +0000)]
cleanup i_int_X vs. uint_X definitions

24 years agoBuild sn as a module
imp [Sun, 18 Jun 2000 04:43:29 +0000 (04:43 +0000)]
Build sn as a module

24 years agoadd MBOX_GET_RESOURCE_COUNT command
mjacob [Sun, 18 Jun 2000 04:41:14 +0000 (04:41 +0000)]
add MBOX_GET_RESOURCE_COUNT command

24 years agoAdd in (separate files for different board's firmware) new files for ispfw
mjacob [Sun, 18 Jun 2000 04:37:44 +0000 (04:37 +0000)]
Add in (separate files for different board's firmware) new files for ispfw
loadable module.

24 years agoFix breakage where we never were attaching SES devices because inq_len
mjacob [Sun, 18 Jun 2000 04:19:08 +0000 (04:19 +0000)]
Fix breakage where we never were attaching SES devices because inq_len
was not being set > 0.

24 years agoFix obvious braino in assignment statement. usermod -g should work again.
davidn [Sun, 18 Jun 2000 02:48:56 +0000 (02:48 +0000)]
Fix obvious braino in assignment statement. usermod -g should work again.
PR: 17877
Submitted by: pius@zyan.com

24 years agoReorder the "prob" section in the output of list/show so it can be copy/pasted
billf [Sun, 18 Jun 2000 02:48:19 +0000 (02:48 +0000)]
Reorder the "prob" section in the output of list/show so it can be copy/pasted
into add without problems.

The previous commit had the other half of this original patch which handled
tcpflags/tcpflgs confusion in output/input.

24 years agoBe a little more strict about handling command line args. This allows user and
davidn [Sun, 18 Jun 2000 02:39:17 +0000 (02:39 +0000)]
Be a little more strict about handling command line args. This allows user and
group names like 'help' 'mod' 'user' 'group' etc. to work correctly without
requiring the -n style invocation.
PR: misc/17069

24 years agoFix a long-standing bug in the rename case previously masked by another bug
davidn [Sun, 18 Jun 2000 02:18:15 +0000 (02:18 +0000)]
Fix a long-standing bug in the rename case previously masked by another bug
just fixed in .db file updating.

24 years agoSecond and hopefully final fix to .db sync when renaming a user; we must run
davidn [Sun, 18 Jun 2000 02:16:07 +0000 (02:16 +0000)]
Second and hopefully final fix to .db sync when renaming a user; we must run
a full reindex in this case to remove the old record. #ifdef -u capability
since this is available on FreeBSD only.
PR: bin/16418
Problem pointed out by: Masachika ISHIZUKA <ishizuka@ish.org>

24 years agoAdd wx and ispfw loadable module defaults.
mjacob [Sat, 17 Jun 2000 23:09:51 +0000 (23:09 +0000)]
Add wx and ispfw loadable module defaults.

24 years agoDeal with quoted arguments. This hack parser uses whitespace to delimit
peter [Sat, 17 Jun 2000 20:10:55 +0000 (20:10 +0000)]
Deal with quoted arguments.  This hack parser uses whitespace to delimit
fields, not lex/yacc grammar so it is not an exact match but should be
close enough for most cases.
Deal with 'port?', 'irq?' style specifications.  These are parsed as
seperate values in lex/yacc in config(8) but tripped up this helper tool.

24 years agoClarify (I hope..) what a hose is. One gets used to DEC (nee, CPQ) speak
wilko [Sat, 17 Jun 2000 20:04:03 +0000 (20:04 +0000)]
Clarify (I hope..) what a hose is. One gets used to DEC (nee, CPQ) speak
all too easily.

Submitted by: Alec Wolman

24 years agoreplace a line erroneously removed in 1.28, should fix xmms
cg [Sat, 17 Jun 2000 19:29:40 +0000 (19:29 +0000)]
replace a line erroneously removed in 1.28, should fix xmms
make buf_clear handle unaligned lengths

24 years agoUse while (<>) instead of while(<STDIN>) so that perl will automagically
peter [Sat, 17 Jun 2000 19:06:13 +0000 (19:06 +0000)]
Use while (<>) instead of while(<STDIN>) so that perl will automagically
deal with filename arguments.  It is amazing how much you forget over time.

Thanks to the people that reminded me this.  I knew there was an easy way
that didn't involve messing with $argv, filehandles, etc, but just could
not remember - all of my books are on the opposite side of the planet..

24 years ago- Moved "hint" informations to GENERIC.hints.
nyan [Sat, 17 Jun 2000 14:46:20 +0000 (14:46 +0000)]
- Moved "hint" informations to GENERIC.hints.
- Cosmetic changes.

24 years agoSwitch over to using the new fflagstostr and strtofflags library calls.
joe [Sat, 17 Jun 2000 14:19:33 +0000 (14:19 +0000)]
Switch over to using the new fflagstostr and strtofflags library calls.

24 years agoMake a note of fflagstostr and strtofflags in the 'see also' section.
joe [Sat, 17 Jun 2000 14:03:34 +0000 (14:03 +0000)]
Make a note of fflagstostr and strtofflags in the 'see also' section.

24 years agoAdd strtofflags and fflagstostr to libc.
joe [Sat, 17 Jun 2000 11:55:57 +0000 (11:55 +0000)]
Add strtofflags and fflagstostr to libc.

24 years agoModify strtofflags so that it returns a malloced string instead of a
joe [Sat, 17 Jun 2000 11:09:24 +0000 (11:09 +0000)]
Modify strtofflags so that it returns a malloced string instead of a
pointer to a static buffer.

24 years agoAdd back "NO_MODULES" as using ``make -DMODULES_WITH_WORLD'' in the
obrien [Sat, 17 Jun 2000 10:56:15 +0000 (10:56 +0000)]
Add back "NO_MODULES" as using ``make -DMODULES_WITH_WORLD'' in the
kernel build dir doesn't really say what your intentions are.

24 years agoIf "MODULES_WITH_WORLD" is defined, sys/modules will be built with the
obrien [Sat, 17 Jun 2000 10:51:56 +0000 (10:51 +0000)]
If "MODULES_WITH_WORLD" is defined, sys/modules will be built with the
world as was our old way, rather than when building a kernel.

Some people do not like the new way, and the release building still assumes
modules are built with the world.

24 years agoAdd support for emulating the RDMSR and WRMSR instructions into BTX. In
jhb [Sat, 17 Jun 2000 06:45:40 +0000 (06:45 +0000)]
Add support for emulating the RDMSR and WRMSR instructions into BTX.  In
theory, this should allow the K7V Athlon motherboard to boot ok with boot
virus protection enabled.  However, I have no hardware to test this.  It
shouldn't break anything though. :)

Prodded by: Kelly Yancey <kbyanc@posi.net>

24 years agoMerged from sys/isa/syscons_isa.c revision 1.13.
kato [Sat, 17 Jun 2000 04:54:50 +0000 (04:54 +0000)]
Merged from sys/isa/syscons_isa.c revision 1.13.

24 years agobs, olpt, pckbd are static limited devices.
kato [Sat, 17 Jun 2000 03:58:06 +0000 (03:58 +0000)]
bs, olpt, pckbd are static limited devices.

24 years agoThe "def" arg for fflagstostr is too specialized for ls. The caller
joe [Sat, 17 Jun 2000 01:43:56 +0000 (01:43 +0000)]
The "def" arg for fflagstostr is too specialized for ls.  The caller
can easily translate from "" to whatever it wants to print if no
flags are set.  (ls prints "-" and mtree prints "none".)

Suggested by: bde

24 years agoReturn of the evil file flags! The {s|g}etflags functions were
joe [Sat, 17 Jun 2000 01:28:13 +0000 (01:28 +0000)]
Return of the evil file flags!  The {s|g}etflags functions were
renamed to {s|g}etflagsbyname, which received objections.   They're
now called strtofflags (string to file flags) and fflagstostr (file
flags to string).

Suggested by: bde

24 years ago- Improved passive mode FTP support by aliasing 229 replies.
ru [Fri, 16 Jun 2000 20:36:16 +0000 (20:36 +0000)]
- Improved passive mode FTP support by aliasing 229 replies.
- Stricter checking of PORT/EPRT/227/229 messages format.
- Moved all security checks into one place.

24 years agoImplement SIOCSIFLLADDR, which allows you to change the link-level
wpaul [Fri, 16 Jun 2000 20:14:43 +0000 (20:14 +0000)]
Implement SIOCSIFLLADDR, which allows you to change the link-level
address on an interface. This basically allows you to do what my
little setmac module/utility does via ifconfig. This involves the
following changes:

socket.h: define SIOCSIFLLADDR
if.c: add support for SIOCSIFLLADDR, which resets the values in
      the arpcom struct and sockaddr_dl for the specified interface.
      Note that if the interface is already up, we need to down/up
      it in order to program the underlying hardware's receive filter.
ifconfig.c: add lladdr command
ifconfig.8: document lladdr command

You can now force the MAC address on any ethernet interface to be
whatever you want. (The change is not sticky across reboots of course:
we don't actually reprogram the EEPROM or anything.) Actually, you
can reprogram the MAC address on other kinds of interfaces too; this
shouldn't be ethernet-specific (though at the moment it's limited to
6 bytes of address data).

Nobody ran up to me and said "this is the politically correct way to
do this!" so I don't want to hear any complaints from people who think
I could have done it more elegantly. Consider yourselves lucky I didn't
do it by having ifconfig tread all over /dev/kmem.

24 years agoAdd a few directories that were shared by a lot of ports: section 3
asami [Fri, 16 Jun 2000 19:08:19 +0000 (19:08 +0000)]
Add a few directories that were shared by a lot of ports: section 3
of perl man pages and site_perl directories for alpha and i386 in /usr/local,
and share/aclocal and share/doc/ja in /usr/local and /usr/X11R6.

Reviewed by: the ports list
Approved by: steve (alpha part)

24 years agoMerged from sys/conf/Makefile.i386 revisions 1.192 and 1.193.
kato [Fri, 16 Jun 2000 16:36:50 +0000 (16:36 +0000)]
Merged from sys/conf/Makefile.i386 revisions 1.192 and 1.193.

24 years agoRevert part of my bioops change which implemented panic(8).
phk [Fri, 16 Jun 2000 14:32:13 +0000 (14:32 +0000)]
Revert part of my bioops change which implemented panic(8).