]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
27 years agoConvert tqdisksort to bufqdisksort. Honor the B_ORDERED buffer flag
gibbs [Sun, 21 Sep 1997 22:10:49 +0000 (22:10 +0000)]
Convert tqdisksort to bufqdisksort.  Honor the B_ORDERED buffer flag
so that meta-data writes go out to the device in the right order.

27 years agoUpdate for new buffer queue data structure.
gibbs [Sun, 21 Sep 1997 22:10:02 +0000 (22:10 +0000)]
Update for new buffer queue data structure.

27 years agobuf.h:
gibbs [Sun, 21 Sep 1997 22:09:24 +0000 (22:09 +0000)]
buf.h:
Change the definition of a buffer queue so that bufqdisksort can
properly deal with bordered writes.

Add inline functions for accessing buffer queues.  This should be
considered an opaque data structure by clients.

callout.h:
New callout implementation.

device.h:
Add support for CAM interrupts.

disk.h:
disklabel.h:
tqdisksort->bufqdisksort

kernel.h:
Add new configuration entries for configuration hooks and calling
cpu_rootconf and cpu_dumpconf.

param.h:
Add a priority for sleeping waiting on config hooks.

proc.h:
Update for new callout implementation.

queue.h:
Add TAILQ_HEAD_INITIALIZER from NetBSD.

systm.h:
Add prototypes for cpu_root/dumpconf, splcam, splsoftcam, etc..

27 years agoConvert to use the new bufq* functions for dealing with buffer
gibbs [Sun, 21 Sep 1997 22:03:22 +0000 (22:03 +0000)]
Convert to use the new bufq* functions for dealing with buffer
queues.

27 years agoUpdate for new callout interface.
gibbs [Sun, 21 Sep 1997 22:02:25 +0000 (22:02 +0000)]
Update for new callout interface.

27 years agoinit_main.c subr_autoconf.c:
gibbs [Sun, 21 Sep 1997 22:00:25 +0000 (22:00 +0000)]
init_main.c subr_autoconf.c:
Add support for "interrupt driven configuration hooks".
A component of the kernel can register a hook, most likely
during auto-configuration, and receive a callback once
interrupt services are available.  This callback will occur before
the root and dump devices are configured, so the configuration
task can affect the selection of those two devices or complete
any tasks that need to be performed prior to launching init.
System boot is posponed so long as a hook is registered.  The
hook owner is responsible for removing the hook once their task
is complete or the system boot can continue.

kern_acct.c kern_clock.c kern_exit.c kern_synch.c kern_time.c:
Change the interface and implementation for the kernel callout
service.  The new implemntaion is based on the work of
Adam M. Costello and George Varghese, published in a technical
report entitled "Redesigning the BSD Callout and Timer Facilities".
The interface used in FreeBSD is a little different than the one
outlined in the paper.  The new function prototypes are:

struct callout_handle timeout(void (*func)(void *),
      void *arg, int ticks);

void untimeout(void (*func)(void *), void *arg,
       struct callout_handle handle);

If a client wishes to remove a timeout, it must store the
callout_handle returned by timeout and pass it to untimeout.

The new implementation gives 0(1) insert and removal of callouts
making this interface scale well even for applications that
keep 100s of callouts outstanding.

See the updated timeout.9 man page for more details.

27 years agoUpdate for changes in the callout interface.
gibbs [Sun, 21 Sep 1997 21:43:54 +0000 (21:43 +0000)]
Update for changes in the callout interface.

27 years agosoundcard.c:
gibbs [Sun, 21 Sep 1997 21:43:35 +0000 (21:43 +0000)]
soundcard.c:
Update for changes in the callout interface.

sequencer.c:
Add a paren that seems to have gone missing.

27 years agoaha1542.c aic6360.c cy.c fd.c ft.c
gibbs [Sun, 21 Sep 1997 21:41:49 +0000 (21:41 +0000)]
aha1542.c aic6360.c cy.c fd.c ft.c
if_ie.c if_wl.c if_zp.c isa.c isa_device.h
labpc.c mcd.c ncr5380.c scd.c seagate.c si.c
sio.c tw.c ultra14f.c wcd.c wd.c:

Update for changes in the callout interface.

apic_vector.s icu_vector.s ipl.s ipl_funcs.c:

Add CAM software/hardware interrupt support.

27 years agoipl.h:
gibbs [Sun, 21 Sep 1997 21:38:54 +0000 (21:38 +0000)]
ipl.h:
Add CAM software/hardware interrupt support.

si.h:
Update for changes in the callout interface.

27 years agoautoconf.c:
gibbs [Sun, 21 Sep 1997 21:38:05 +0000 (21:38 +0000)]
autoconf.c:
Add cpu_rootconf and cpu_dumpconf so that configuring these
two devices can be better controlled by the MI configuration
code.

machdep.c:
MD initialization code for the new callout interface.

trap.c:
Add support for printing out whether cam interrupts are masked
during a panic.

27 years agoAdd shared EISA interrupt support.
gibbs [Sun, 21 Sep 1997 21:35:24 +0000 (21:35 +0000)]
Add shared EISA interrupt support.
Clean up the match routines so that they return const char *

27 years agoMove the rules for aicasm to the MI conf file.
gibbs [Sun, 21 Sep 1997 21:34:31 +0000 (21:34 +0000)]
Move the rules for aicasm to the MI conf file.

27 years agoConvert to the new callout interface.
gibbs [Sun, 21 Sep 1997 21:33:10 +0000 (21:33 +0000)]
Convert to the new callout interface.

27 years agoConvert to the new callout interface.
gibbs [Sun, 21 Sep 1997 21:33:01 +0000 (21:33 +0000)]
Convert to the new callout interface.
Guard against scheduling more than one callout.

27 years agoMove the rules for aicasm to the MI conf file.
gibbs [Sun, 21 Sep 1997 21:32:04 +0000 (21:32 +0000)]
Move the rules for aicasm to the MI conf file.

Add kern/subr_autoconf.c as a standard file as it contains the bulk of
the code for performing interrupt driven configuration.

27 years agoIt turns out that the following:
brian [Sun, 21 Sep 1997 20:26:47 +0000 (20:26 +0000)]
It turns out that the following:
  close(1);
  close(2);
  x = open(ctermid(NULL), O_RDWR|O_NONBLOCK);
  close(0)
on a tty causes select() to return an exception for descriptor x !
This is the case in RELENG_2_2, but not in 2.2.2.  I'm not sure why.
Instead of doing the x=open() and close(0), we just do x=0 now.
Problem pointed out by: Greg Lehey <grog@lemis.com>
                        Tomi Vainio <tomppa@fidata.fi>

27 years agoFreeBSD's original passwd helper is needed here.
markm [Sun, 21 Sep 1997 17:37:08 +0000 (17:37 +0000)]
FreeBSD's original passwd helper is needed here.

27 years agoUnspam rev 1.114 that got spammed when someone commited the rc.conf
joerg [Sun, 21 Sep 1997 16:09:48 +0000 (16:09 +0000)]
Unspam rev 1.114 that got spammed when someone commited the rc.conf
changes: drop `savecore_enable' now that our savecore obeyes the
minfree file.

27 years agoImplement the parts needed for VM86 under SMP.
peter [Sun, 21 Sep 1997 15:03:59 +0000 (15:03 +0000)]
Implement the parts needed for VM86 under SMP.

27 years agoTypo police
brian [Sun, 21 Sep 1997 13:08:00 +0000 (13:08 +0000)]
Typo police
Submitted by: Mark Cammidge <mark@gmtunx.ee.uct.ac.za>

27 years agoAdd a pppctl(8) xref.
brian [Sun, 21 Sep 1997 13:06:43 +0000 (13:06 +0000)]
Add a pppctl(8) xref.
Suggested by: joerg

27 years agoAdd ypwhich(1) and yppoll(8) manual pages.
wosch [Sun, 21 Sep 1997 11:49:18 +0000 (11:49 +0000)]
Add ypwhich(1) and yppoll(8) manual pages.

Obtained from: NetBSD, slightly modified

27 years agoFix some style(9) and formatting problems. tabsize 4 formatting doesn't
peter [Sun, 21 Sep 1997 11:41:12 +0000 (11:41 +0000)]
Fix some style(9) and formatting problems.  tabsize 4 formatting doesn't
look too great with 'more' etc.

Approved by: dyson (with a minor grumble :-)

27 years agoChown(8) compiled with -DSUPPORT_DOT (backward compatibility) does
wosch [Sun, 21 Sep 1997 09:13:57 +0000 (09:13 +0000)]
Chown(8) compiled with -DSUPPORT_DOT (backward compatibility) does
first check for a `.' and then for `:' as a delimiter.
Usernames with a dot will fail.

# chown r.r:bin /tmp/bla
chown: r:bin: illegal group name

Fix: first check for a `:' and then for a `.'

27 years agoAdd support for more than 1 page of idle process stack on SMP systems.
dyson [Sun, 21 Sep 1997 05:50:02 +0000 (05:50 +0000)]
Add support for more than 1 page of idle process stack on SMP systems.

27 years agoRe-institute a bugfix in allocation of anonymous buffer memory.
dyson [Sun, 21 Sep 1997 04:49:30 +0000 (04:49 +0000)]
Re-institute a bugfix in allocation of anonymous buffer memory.

27 years agoChange the M_NAMEI allocations to use the zone allocator. This change
dyson [Sun, 21 Sep 1997 04:24:27 +0000 (04:24 +0000)]
Change the M_NAMEI allocations to use the zone allocator.  This change
plus the previous changes to use the zone allocator decrease the useage
of malloc by half.  The Zone allocator will be upgradeable to be able
to use per CPU-pools, and has more intelligent usage of SPLs.  Additionally,
it has reasonable stats gathering capabilities, while making most calls
inline.

27 years agoRecognize the CS4326 on the Intel PR440FX motherboard. (It works just like
peter [Sun, 21 Sep 1997 03:27:51 +0000 (03:27 +0000)]
Recognize the CS4326 on the Intel PR440FX motherboard.  (It works just like
the normal CS4326 except that it's had it's ID's tweaked for some reason)
Also mark the device as alive in the attach routine so that the pnp system
doesn't think the attach failed.

27 years agoFix a comment-within-a-comment
peter [Sun, 21 Sep 1997 03:21:34 +0000 (03:21 +0000)]
Fix a comment-within-a-comment

27 years agoAccept FIOASYNC (like the old sound driver) since it's generated by
peter [Sun, 21 Sep 1997 03:20:36 +0000 (03:20 +0000)]
Accept FIOASYNC (like the old sound driver) since it's generated by
kern_descrip.c when the user does an fcntl(.., F_SETFL, ...).

27 years agoWe were (I think) missing a vrele() on the vnode for the object loaded
peter [Sun, 21 Sep 1997 03:13:21 +0000 (03:13 +0000)]
We were (I think) missing a vrele() on the vnode for the object loaded
via PT_INTERP (usually /usr/libexec/ld-elf.so.1).

27 years agoAdd "delete 0" before "add 0 0..."
brian [Sun, 21 Sep 1997 02:10:41 +0000 (02:10 +0000)]
Add "delete 0" before "add 0 0..."
Suggested by:

27 years agoDo not complain about non existing shell "/nonexistent"
wosch [Sat, 20 Sep 1997 19:59:54 +0000 (19:59 +0000)]
Do not complain about non existing shell "/nonexistent"
Smarter error check if a uid is already in use.

Submitted by: "Eugene Radchenko" <genie@qsar.chem.msu.su>

27 years agoDo not accept empty user names.
wosch [Sat, 20 Sep 1997 18:26:22 +0000 (18:26 +0000)]
Do not accept empty user names.

27 years agoDon't gratuitously drop termcap.db and vgrindefs.db from the installed
joerg [Sat, 20 Sep 1997 17:10:09 +0000 (17:10 +0000)]
Don't gratuitously drop termcap.db and vgrindefs.db from the installed
set of files.

Reviewed by: jkh

27 years agoPerform a `make distrib-dirs distribution' inside the chroot tree
joerg [Sat, 20 Sep 1997 15:40:55 +0000 (15:40 +0000)]
Perform a `make distrib-dirs distribution' inside the chroot tree
before trying to `make world', so to become less dependant from the
correctness of the environment hosting the `make release'.  The recent
addition of a group `network' made this problem apparent.

27 years agoFor AMD chips, pick up the long description from the chip if
phk [Sat, 20 Sep 1997 13:18:48 +0000 (13:18 +0000)]
For AMD chips, pick up the long description from the chip if
possible. (This is not really a typographical improvement in the
case of the K6 it seems, but AMD appearantly want it too look
that way).  Also if bootverbose, dump some more info about the
chip.

27 years ago_G_VTABLE_LABLE_PREFIX is different between ELF and a.out
peter [Sat, 20 Sep 1997 09:58:49 +0000 (09:58 +0000)]
_G_VTABLE_LABLE_PREFIX is different between ELF and a.out
('_vt.' vs '__vt$')

27 years agoAddition of support of the slightly rogue Promise IDE interface(Dyson), support
dyson [Sat, 20 Sep 1997 07:41:58 +0000 (07:41 +0000)]
Addition of support of the slightly rogue Promise IDE interface(Dyson), support
of multiple PCI IDE controllers(Dyson), and some updates and cleanups from
John Hood, who originally made our IDE DMA stuff work :-).

I have run tests with 7 IDE drives connected to my system, all in DMA
mode, with no errors.  Modulo any bugs, this stuff makes IDE look
really good (within it's limitations.)

Submitted by: John Hood <cgull@smoke.marlboro.vt.us>

27 years agoAdjust the #ifdef KERNEL so that pnp_cinfo is available for sbin/dset.c
peter [Sat, 20 Sep 1997 06:26:28 +0000 (06:26 +0000)]
Adjust the #ifdef KERNEL so that pnp_cinfo is available for sbin/dset.c
Other ways around this might be to #define KERNEL in dset.c for the
"i386/isa/pnp.h" include.

27 years agoMore label editor enhancements.
jkh [Sat, 20 Sep 1997 06:24:17 +0000 (06:24 +0000)]
More label editor enhancements.
Submitted by: Ed Gold <vegold01@starbase.spd.louisville.edu>

27 years agoSynchronize with sys/i386/isa/isa.c revision 1.104.
kato [Sat, 20 Sep 1997 05:29:03 +0000 (05:29 +0000)]
Synchronize with sys/i386/isa/isa.c revision 1.104.

27 years agoSynchronize with sys/i386/conf/options.i386 and sys/i386/isa/sio.c
kato [Sat, 20 Sep 1997 05:28:02 +0000 (05:28 +0000)]
Synchronize with sys/i386/conf/options.i386 and sys/i386/isa/sio.c
revisions 1.60 and 1.182, respectively.

27 years agoAdd missing \ from previous rev to end of a line.
imp [Sat, 20 Sep 1997 04:58:08 +0000 (04:58 +0000)]
Add missing \ from previous rev to end of a line.

27 years agoFix conditional bogon in Label Editor [from Ed Gold]
jkh [Sat, 20 Sep 1997 02:48:48 +0000 (02:48 +0000)]
Fix conditional bogon in Label Editor [from Ed Gold]
Print info boxes without titles.

27 years ago``oops''. I cut/pasted the original free()'s based on mark's suggestion
peter [Sat, 20 Sep 1997 02:29:56 +0000 (02:29 +0000)]
``oops''.  I cut/pasted the original free()'s based on mark's suggestion
rather than extracting the diff from Mark's patch, but it turns out that
I was freeing one allocation twice due to a previous cut/paste braino.
My botch, not Mark's.

Pointed out by:  Mark Valentine <mv@pobox.com>

27 years agoDescribe all know capabilities now
ache [Sat, 20 Sep 1997 01:10:28 +0000 (01:10 +0000)]
Describe all know capabilities now

27 years agoUse snprintf() instead of sprintf() most of the time.
brian [Fri, 19 Sep 1997 22:22:04 +0000 (22:22 +0000)]
Use snprintf() instead of sprintf() most of the time.
Obtained from: OpenBSD

Make usage() consistent with man page.

27 years ago.Ox macro needs to support "OpenBSD 2.1"
wosch [Fri, 19 Sep 1997 21:37:05 +0000 (21:37 +0000)]
.Ox macro needs to support "OpenBSD 2.1"

PR: docs/4561
Submitted by: Norihiro Kumagai <kuma@slab.tnr.sharp.co.jp>

27 years agoTypo.
max [Fri, 19 Sep 1997 16:04:19 +0000 (16:04 +0000)]
Typo.

27 years agobmake'd pnpinfo
jmg [Fri, 19 Sep 1997 15:41:57 +0000 (15:41 +0000)]
bmake'd pnpinfo

27 years agoactivate pnpinfo
jmg [Fri, 19 Sep 1997 15:41:43 +0000 (15:41 +0000)]
activate pnpinfo

27 years agoThis commit was generated by cvs2svn to compensate for changes in r29615,
jmg [Fri, 19 Sep 1997 15:36:00 +0000 (15:36 +0000)]
This commit was generated by cvs2svn to compensate for changes in r29615,
which included commits to RCS files with non-trunk default branches.

27 years agothis is an import of Sujal M. Petal's pnpinfo.. it includes modifications
jmg [Fri, 19 Sep 1997 15:36:00 +0000 (15:36 +0000)]
this is an import of Sujal M. Petal's pnpinfo.. it includes modifications
by Luigi Rizzo and myself...

27 years agoteach sio how to attach to isa PnP cards. This is mainly for use with
jmg [Fri, 19 Sep 1997 15:25:49 +0000 (15:25 +0000)]
teach sio how to attach to isa PnP cards.  This is mainly for use with
internal modems.  Currently detects a USR modem, and a couple Supra
modems...  vendor id's for sio capabile cards welcomed...

document new option EXTRA_SIO that will increase sio's internal data
structures to support X more serial ports...  these are used by the
PnP part of sio for attaching...  If you don't have it specified, it
will default to 2...  This is defaulted to 0 if you don't have PnP
compiled into your kernel...

also document that if you set the PnP flags (pnp x flags y) to 0x1 that
the modem will be refused to be recognized by the sio driver... this
is for people that want the traditional isa driver to probe and attach
the modem... (for keeping legacy sio numbering)

27 years agoteach pnp to keep isa_device structs around, and teach isa.c how to scan
jmg [Fri, 19 Sep 1997 15:20:25 +0000 (15:20 +0000)]
teach pnp to keep isa_device structs around, and teach isa.c how to scan
these structs for conflics...

it still exist that two PnP cards can colide, but this is up to the user
to make sure it doesn't happen...

other modifications to pnp.c to format output properly, and hide more
output behind bootverbose flag...

fix some bugons in pnp.h that would of made it difficult for inclusion
in external programs (for import of pnpinfo)

27 years agomajor modifications to the mixer. currently it doesn't report errors
jmg [Fri, 19 Sep 1997 13:38:51 +0000 (13:38 +0000)]
major modifications to the mixer.  currently it doesn't report errors
that it encounters (such as invalid device), but does print usage upon
this error...

but otherwise allows complex commands such as:
mixer -f /dev/mixer4 synth line 50 -rec mic +rec line =rec cd ^rec mix

27 years agoAdd a sample entry specifying numeric tty settings. People sometimes ask
joerg [Fri, 19 Sep 1997 12:25:09 +0000 (12:25 +0000)]
Add a sample entry specifying numeric tty settings.  People sometimes ask
how to force a local tty line into CLOCAL mode and Xon/Xoff flow control.
This is the preferrable way (over rc.serial).

27 years ago[Regarding the previous patch] This is completely wrong.
phk [Fri, 19 Sep 1997 11:13:16 +0000 (11:13 +0000)]
[Regarding the previous patch] This is completely wrong.

1. ffs_alloc() actually allowed writing one block less one frag (normally
   7 frags or 7/8 blocks) beyond the limit.
2. freebufspace() gives the free space in frags, but `size' is in bytes,
   so the change results in approximately `size' fragments too many being
   reserved.
3. ffs_realloccg() has the same bug but wasn't changed.

PR: 3398
Submitted by: bde
Eyeballed by: phk

27 years agoAdd Wnn6 reverse lookup server port.
asami [Fri, 19 Sep 1997 09:38:50 +0000 (09:38 +0000)]
Add Wnn6 reverse lookup server port.

Submitted by:  Satoshi Taoka <taoka@infonets.hiroshima-u.ac.jp>

27 years agoSynchronize with sys/i386/i386/userconfig.c revision 1.93.
kato [Fri, 19 Sep 1997 08:48:02 +0000 (08:48 +0000)]
Synchronize with sys/i386/i386/userconfig.c revision 1.93.

27 years agoUse An macro for author name.
charnier [Fri, 19 Sep 1997 06:29:52 +0000 (06:29 +0000)]
Use An macro for author name.

27 years agoUse err(3). Typo fixes in man page.
charnier [Fri, 19 Sep 1997 06:28:57 +0000 (06:28 +0000)]
Use err(3). Typo fixes in man page.

27 years agoUse err(3).
charnier [Fri, 19 Sep 1997 06:27:30 +0000 (06:27 +0000)]
Use err(3).

27 years agoAdd the direcotry for the japanese shift-JIS locale
julian [Fri, 19 Sep 1997 01:51:01 +0000 (01:51 +0000)]
Add the direcotry for the japanese shift-JIS locale

27 years agoremove bad comment
julian [Fri, 19 Sep 1997 00:56:49 +0000 (00:56 +0000)]
remove bad comment

27 years agooops
julian [Fri, 19 Sep 1997 00:51:52 +0000 (00:51 +0000)]
oops
somehow the file I was supplied with turned out to be EUC encoding..
luckily they can be translated by a simple c program :)

27 years agoFix argument string for getopt.
dima [Thu, 18 Sep 1997 23:16:17 +0000 (23:16 +0000)]
Fix argument string for getopt.

Submitted by: dillon@best.net

27 years agoFix some problems in the rules file loading and need for modload detection.
danny [Thu, 18 Sep 1997 22:43:48 +0000 (22:43 +0000)]
Fix some problems in the rules file loading and need for modload detection.

Found by: "James E. Housley" <housley@pr-comm.com>

27 years agoUse -ansi -pedantic -trigraphs, as suggested by Bruce.
joerg [Thu, 18 Sep 1997 20:55:50 +0000 (20:55 +0000)]
Use -ansi -pedantic -trigraphs, as suggested by Bruce.

Enforce that only Posix-spec'ed options are being used, so people could
use this as a Posixificator of some kind.

Document the Posix usage in the man page, so to make it clear what's
the difference between Posix c89, and gcc.

27 years agomake/install the japanese SJIS date locale file.
julian [Thu, 18 Sep 1997 19:16:51 +0000 (19:16 +0000)]
make/install the japanese SJIS date locale file.

27 years agoShift JIS versions of date stuff.
julian [Thu, 18 Sep 1997 19:09:08 +0000 (19:09 +0000)]
Shift JIS versions of date stuff.
more changes to come.
Submitted by: Nobuhrio Yasutomi <nobu@psrc.isac.co.jp>
 AM/PM changed to 2 letter versions for now, original SJIS versions
 commented out. change/check later..

27 years agoExecuting binaries on a nullfs (or nullfs-based) filesystem results in
phk [Thu, 18 Sep 1997 18:33:23 +0000 (18:33 +0000)]
Executing binaries on a nullfs (or nullfs-based) filesystem results in
a trap.
PR: 3104
Reviewed by: phk
Submitted by: Dan Walters hannibal@cyberstation.net

27 years agoAdd a "I know it points to my foot!" -f option.
phk [Thu, 18 Sep 1997 18:27:34 +0000 (18:27 +0000)]
Add a "I know it points to my foot!" -f option.
PR: 3288
Reviewed by: phk
Submitted by: Martin Kammerhofer <dada@freepass.tu-graz.ac.at>

27 years agoFix bad assumptions about types.
phk [Thu, 18 Sep 1997 18:19:45 +0000 (18:19 +0000)]
Fix bad assumptions about types.
PR: 1649
Reviewed by: phk
Submitted by: Jason Thorpe <thorpej@nas.nasa.gov>

27 years agoFfs_alloc allow users to write one block beyond the limit.
phk [Thu, 18 Sep 1997 18:07:45 +0000 (18:07 +0000)]
Ffs_alloc allow users to write one block beyond the limit.

PR: 3398
Reviewed by: phk
Submitted by: Wolfram Schneider <wosch@apfel.de>

27 years agoWhoops! Missed one on the XF8633 -> XF86331 update.
jkh [Thu, 18 Sep 1997 17:54:36 +0000 (17:54 +0000)]
Whoops!  Missed one on the XF8633 -> XF86331 update.
PR: 4560

27 years agoCorrect typo.
jkh [Thu, 18 Sep 1997 16:24:36 +0000 (16:24 +0000)]
Correct typo.

27 years agoOver 1000 ports now. Note this in the relevant place.
jkh [Thu, 18 Sep 1997 16:19:17 +0000 (16:19 +0000)]
Over 1000 ports now.  Note this in the relevant place.

27 years agoMany places in the code NULL is used in integer context, where
phk [Thu, 18 Sep 1997 14:08:40 +0000 (14:08 +0000)]
Many places in the code NULL is used in integer context, where
plain 0 should be used.  This happens to work because we #define
NULL to 0, but is stylistically wrong and can cause problems
for people trying to port bits of code to other environments.

PR: 2752
Submitted by: Arne Henrik Juul <arnej@imf.unit.no>

27 years agoMany places in the code NULL is used in integer context, where
phk [Thu, 18 Sep 1997 13:45:34 +0000 (13:45 +0000)]
Many places in the code NULL is used in integer context, where
plain 0 should be used.  This happens to work because we #define
NULL to 0, but is stylistically wrong and can cause problems
for people trying to port bits of code to other environments.

PR: 2752
Submitted by: Arne Henrik Juul <arnej@imf.unit.no>

27 years agoAdd canna and Wnn ports numbers.
asami [Thu, 18 Sep 1997 08:56:53 +0000 (08:56 +0000)]
Add canna and Wnn ports numbers.

Submitted by: Satoshi Taoka <taoka@infonets.hiroshima-u.ac.jp>

27 years agoMissed a place where the extra descriptor buffers would need to be
peter [Thu, 18 Sep 1997 08:28:23 +0000 (08:28 +0000)]
Missed a place where the extra descriptor buffers would need to be
freed.

Submitted by:  Mark Valentine <mark@linus.demon.co.uk>

27 years agoSynchronize with sys/i386/isa/fd.c revision 1.102.
kato [Thu, 18 Sep 1997 08:10:45 +0000 (08:10 +0000)]
Synchronize with sys/i386/isa/fd.c revision 1.102.

27 years agowrap kernel interface in #ifdef KERNEL/#endif
jmg [Thu, 18 Sep 1997 08:04:13 +0000 (08:04 +0000)]
wrap kernel interface in #ifdef KERNEL/#endif

hide more verbosity behind bootverbose

make it so that it doesn't emit as many warnings durning compiling...

27 years agoreduce the number of warnings this file emits during compiling
jmg [Thu, 18 Sep 1997 08:00:48 +0000 (08:00 +0000)]
reduce the number of warnings this file emits during compiling

27 years agoupdate Luigi's driver to poll interface (Peter, you might want to check
jmg [Thu, 18 Sep 1997 07:47:03 +0000 (07:47 +0000)]
update Luigi's driver to poll interface (Peter, you might want to check
that I've done this properly, it does work though :) )...

a few minor fixes to code

part-Submitted-by: Luigi

27 years agoenvironmental -> environment.
charnier [Thu, 18 Sep 1997 06:55:21 +0000 (06:55 +0000)]
environmental -> environment.

27 years agoTypo.
charnier [Thu, 18 Sep 1997 06:54:03 +0000 (06:54 +0000)]
Typo.

27 years agoTypo: environmental variable -> environment variable.
charnier [Thu, 18 Sep 1997 06:49:40 +0000 (06:49 +0000)]
Typo: environmental variable -> environment variable.

27 years agoFix sorting of directories when doing '-d'.
sef [Thu, 18 Sep 1997 06:42:27 +0000 (06:42 +0000)]
Fix sorting of directories when doing '-d'.

PR: bin/4558
Submitted by: Keith Bostic

27 years agoUse err(3). Document -y flag. Remove unused -s flag from getopt string.
charnier [Thu, 18 Sep 1997 06:35:51 +0000 (06:35 +0000)]
Use err(3). Document -y flag. Remove unused -s flag from getopt string.

27 years agoRemove line noise from the quirk list (kernel build stopper)
msmith [Thu, 18 Sep 1997 00:15:55 +0000 (00:15 +0000)]
Remove line noise from the quirk list (kernel build stopper)

27 years agoDon't close(1) in direct mode and then proceed to
brian [Thu, 18 Sep 1997 00:15:25 +0000 (00:15 +0000)]
Don't close(1) in direct mode and then proceed to
isatty(1) !  Keep 0 open for this till the modem's
been set up by either dup()ing 0 or by opening
ctermid(NULL) (if isatty(0)).

Discussed problem with: Tomi Vainio <tomppa@fidata.fi>
Made it finally dawn on me: Angelo Turetta <ATuretta@stylo.it>

27 years agoSupport CHAP using MD4
brian [Wed, 17 Sep 1997 23:17:57 +0000 (23:17 +0000)]
Support CHAP using MD4
Suggested by: jordan

27 years agoAdd the C1557A as a known tape changer (thus non-broken multi-LUN tape
joerg [Wed, 17 Sep 1997 21:49:06 +0000 (21:49 +0000)]
Add the C1557A as a known tape changer (thus non-broken multi-LUN tape
device).

Submitted by: Andre.Albsmeier@mchp.siemens.de (Andre Albsmeier)

27 years agoActivate c89(1).
joerg [Wed, 17 Sep 1997 20:46:58 +0000 (20:46 +0000)]
Activate c89(1).

27 years agoThis commit was generated by cvs2svn to compensate for changes in r29544,
joerg [Wed, 17 Sep 1997 20:44:54 +0000 (20:44 +0000)]
This commit was generated by cvs2svn to compensate for changes in r29544,
which included commits to RCS files with non-trunk default branches.

27 years agoc89 is the Posix.2 name for a C compiler that accepts ANSI/ISO C source
joerg [Wed, 17 Sep 1997 20:44:54 +0000 (20:44 +0000)]
c89 is the Posix.2 name for a C compiler that accepts ANSI/ISO C source
code.  Basically a shortcut to cc -ansi -trigraphs.