]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
23 years agoBetter handling for the return of snprintf().
brian [Sat, 18 Aug 2001 22:43:11 +0000 (22:43 +0000)]
Better handling for the return of snprintf().

23 years agoIf AF_INET6 isn't available in the kernel, revert to creating AF_INET
brian [Sat, 18 Aug 2001 19:16:56 +0000 (19:16 +0000)]
If AF_INET6 isn't available in the kernel, revert to creating AF_INET
listening sockets in ``set server''.

23 years agoDon't needlessly duplicate what's basically the same copyright.
jake [Sat, 18 Aug 2001 19:09:55 +0000 (19:09 +0000)]
Don't needlessly duplicate what's basically the same copyright.

23 years agoBack out the previous fix to deal with kernels that don't support IPv6,
brian [Sat, 18 Aug 2001 19:07:13 +0000 (19:07 +0000)]
Back out the previous fix to deal with kernels that don't support IPv6,
and implement a far more subtle and correct fix.

The reason behind the infinite loop was that ppp was trying to make up
initial IPv6 numbers and wasn't giving up when it failed unexpectedly to
assign the addresses it just fabricated to it's interface (thinking that
the reason was because another interface was using the same address).
It now attempts this up to 100 times before just failing and trying to
muddle along (in reality, this should never happen more than a couple
of times unless our random number generator doesn't work).

Also, when IPv6 is not available, don't even try to assign the IPv6
interface address in the first place...

23 years agoImplement cpu_wait().
jake [Sat, 18 Aug 2001 18:11:48 +0000 (18:11 +0000)]
Implement cpu_wait().

23 years agoIncrease the size of the phys_avail memory map. Implement pmap_dispose_proc.
jake [Sat, 18 Aug 2001 18:11:13 +0000 (18:11 +0000)]
Increase the size of the phys_avail memory map. Implement pmap_dispose_proc.
Turn some more potentially import functions into nops so we can do stuff
until they matter.

23 years agoSpell ta 1 correctly as ta %xcc, 1. Use %pil for critical enter/exit
jake [Sat, 18 Aug 2001 18:07:37 +0000 (18:07 +0000)]
Spell ta 1 correctly as ta %xcc, 1.  Use %pil for critical enter/exit
instead of pstate.ie.  Note that popc is not implemented in hardware
on certain ultras, so we can't use it for inline ffs (suck).

23 years agoGcc 3.0 requires a .register pseudo-op for certain global registers when
jake [Sat, 18 Aug 2001 18:05:24 +0000 (18:05 +0000)]
Gcc 3.0 requires a .register pseudo-op for certain global registers when
used in assembly language.  Tell it to ignore the registers for now.

23 years agofix typo
julian [Sat, 18 Aug 2001 17:43:29 +0000 (17:43 +0000)]
fix typo

Submitted by: Ian Dowse <iedowse@maths.tcd.ie>

23 years agoExtend comments documenting various supported audio formats and mixer devices.
sobomax [Sat, 18 Aug 2001 15:07:51 +0000 (15:07 +0000)]
Extend comments documenting various supported audio formats and mixer devices.

Reviewed by:    ru, cg
Approved by:    cg
Obtained from:  4Front's OSS Programming Guide
MFC after: 2 weeks

23 years agoClear up what the line "ALL : PARANOID : RFC931 20 : deny" means
dwmalone [Sat, 18 Aug 2001 14:22:52 +0000 (14:22 +0000)]
Clear up what the line "ALL : PARANOID : RFC931 20 : deny" means
to tcp wrappers. The description is a little long, but hopefully
accurate.

23 years agoRun correctly on a machine built without AF_INET6 support
brian [Sat, 18 Aug 2001 13:04:52 +0000 (13:04 +0000)]
Run correctly on a machine built without AF_INET6 support

23 years agoBuild properly with -DNOINET6
brian [Sat, 18 Aug 2001 12:31:54 +0000 (12:31 +0000)]
Build properly with -DNOINET6

23 years agoBuild properly with -DNOIPV6
brian [Sat, 18 Aug 2001 12:30:41 +0000 (12:30 +0000)]
Build properly with -DNOIPV6

23 years agoRemove a blank line that snuck in with the previous commit.
des [Sat, 18 Aug 2001 10:48:55 +0000 (10:48 +0000)]
Remove a blank line that snuck in with the previous commit.
Set the default timeout to 120 seconds instead of 0 (no timeout).

23 years agoImplement compute_stats() in terms of devstat_compute_statistics(). This
ken [Sat, 18 Aug 2001 05:46:59 +0000 (05:46 +0000)]
Implement compute_stats() in terms of devstat_compute_statistics().  This
gets rid of the duplicated code in compute_stats().

Add a new DSM_SKIP statistic type for devstat_compute_statistics() that
causes the subsequent variable argument to be skipped.

Thanks to Sergey Osokin for coding up my idea/code fragment.

Submitted by: "Sergey A. Osokin" <osa@freebsd.org.ru>

23 years agoUnbreak linux compatibility by providing the correct length of the buffer.
mp [Sat, 18 Aug 2001 04:24:30 +0000 (04:24 +0000)]
Unbreak linux compatibility by providing the correct length of the buffer.

Reported by: "Pierre Y. Dampure" <pierre.dampure@westmarsh.com>,
"Niels Chr. Bank-Pedersen" <ncbp@bank-pedersen.dk>
Pointy hat to: mp

23 years agoThere is nothing special that requires SSE to be only on 686 class cpus.
peter [Sat, 18 Aug 2001 04:04:45 +0000 (04:04 +0000)]
There is nothing special that requires SSE to be only on 686 class cpus.
This enables 586-only SMP kernels to compile again.

Problem reported by:  Jacek Jedrzejczak <jacol@ids.gda.pl>

23 years agoCorrect path (../crypto, not ../../crypto)
peter [Sat, 18 Aug 2001 03:36:26 +0000 (03:36 +0000)]
Correct path (../crypto, not ../../crypto)

23 years agoSigh. ufs_lookup() calls ffs_snapgone(), meaning that 'options EXT2FS'
peter [Sat, 18 Aug 2001 03:08:48 +0000 (03:08 +0000)]
Sigh.  ufs_lookup() calls ffs_snapgone(), meaning that 'options EXT2FS'
without 'options FFS' would fail to link.

23 years agoOnly compile the ufs components if FFS || IFS || EXT2FS are present, not
peter [Sat, 18 Aug 2001 02:57:25 +0000 (02:57 +0000)]
Only compile the ufs components if FFS || IFS || EXT2FS are present, not
unconditionally.

23 years agoDon't alocate a 400 byte buffer on the stack,
julian [Sat, 18 Aug 2001 02:53:50 +0000 (02:53 +0000)]
Don't alocate a 400 byte buffer on the stack,
Nor 800 bytes of structures..

MFC after: 2 weeks

23 years agoTurn off SMP on Alpha for the time being. There are some memory ordering
jhb [Sat, 18 Aug 2001 01:12:43 +0000 (01:12 +0000)]
Turn off SMP on Alpha for the time being.  There are some memory ordering
issues that can result in panics and probably data corruption on SMP
systems right now.  When they are fixed this can be turned back on.

23 years agoGrab Giant around trapsignal() at the bottom of trap().
jhb [Sat, 18 Aug 2001 01:08:45 +0000 (01:08 +0000)]
Grab Giant around trapsignal() at the bottom of trap().

Reported by: obrien

23 years agoRemove redundant declarations of functions that are declared in
jhb [Sat, 18 Aug 2001 00:57:32 +0000 (00:57 +0000)]
Remove redundant declarations of functions that are declared in
sys/ptrace.h.

23 years agoNew release note: CDROM booting on IBMs.
bmah [Sat, 18 Aug 2001 00:12:23 +0000 (00:12 +0000)]
New release note:  CDROM booting on IBMs.

Reminded by: mp

23 years agoRe-initialize gp after a jsr.
obrien [Fri, 17 Aug 2001 22:54:26 +0000 (22:54 +0000)]
Re-initialize gp after a jsr.

When rtld runs the .fini section in a shared lib (C++), the code in
question from .../contrib/gdb/config/alpha/crtbegin.asm first calls
__do_globals_dtors_aux and then __do_frame_takedown.  Unfortunately, the
value of gp after a jsr is undefined and in this case had changed from before
the call, probably as a result of calling code in some other shared library.
The normal calling convention for alpha is to re-initialize gp using
'ldgp gp,0(ra)' after a jsr instruction but in this case no such
re-initialization is done. This leads to a bogus value being read for the
address of __do_frame_takedown and a quick segfault.

Submitted by:   dfr
Obtained from:  GCC 3.0

23 years agoAdd some comments.
des [Fri, 17 Aug 2001 22:22:45 +0000 (22:22 +0000)]
Add some comments.

23 years agoActuall make plumb work in addition to create as per the manpage.
brooks [Fri, 17 Aug 2001 22:16:11 +0000 (22:16 +0000)]
Actuall make plumb work in addition to create as per the manpage.

PR: bin/29812
Submitted by: Joao Carlos Mendes Luis <jonny@eng05.embratel.net.br>

23 years agoImplement getpeereid(3), a front-end to the LOCAL_PEERCRED
dd [Fri, 17 Aug 2001 22:09:15 +0000 (22:09 +0000)]
Implement getpeereid(3), a front-end to the LOCAL_PEERCRED
socket option for the Unix domain.  It's weaker than the
socket option (this only returns the uid and gid, while the
socket opt. can return the entire group list), and is
implemented mostly for compatibility with OpenBSD.

23 years agoImplement a LOCAL_PEERCRED socket option which returns a
dd [Fri, 17 Aug 2001 22:01:18 +0000 (22:01 +0000)]
Implement a LOCAL_PEERCRED socket option which returns a
`struct xucred` with the credentials of the connected peer.
Obviously this only works (and makes sense) on SOCK_STREAM
sockets.  This works for both the connect(2) and listen(2)
callers.

There is precise documentation of the semantics in unix(4).

Reviewed by: dwmalone (eyeballed)

23 years agoNew release note: SA-01:53.
bmah [Fri, 17 Aug 2001 21:59:13 +0000 (21:59 +0000)]
New release note:  SA-01:53.

23 years agoWhen discussing optimization levels, don't put the compiler options in
dd [Fri, 17 Aug 2001 21:42:15 +0000 (21:42 +0000)]
When discussing optimization levels, don't put the compiler options in
.Op.  None of the other manual pages do it when discussing options in
the main text, so this one shouldn't, either.  Also, use .Pq instead
of another odd constructhyphenation isn't an issue here, since it's
desireable to have the contents of that digression appear together,
and it's already in another macro, anyway).

23 years agoXref raise(3).
dd [Fri, 17 Aug 2001 21:04:42 +0000 (21:04 +0000)]
Xref raise(3).

23 years agoClarify the description string for one of the StrongARM based cards.
scottl [Fri, 17 Aug 2001 20:59:49 +0000 (20:59 +0000)]
Clarify the description string for one of the StrongARM based cards.

23 years agoTroll for MFCs that haven't been marked as such: KERNCONF, an(4)
bmah [Fri, 17 Aug 2001 19:59:25 +0000 (19:59 +0000)]
Troll for MFCs that haven't been marked as such:  KERNCONF, an(4)
support for Aironet 350, ed(4) support for Linksys cards, aac(4),
ssh(1) not SUID, passwd(1)/pw(8) support for passwd_format,
rc(8) deletes non-directories in /var/run and /var/spool/lock,
fmtcheck(3), sshd(8) X11Forwarding, sshd_config MaxStartups
deprecates ConnectionsPerPeriod.

Some of these MFCs are new, others were reflected in 4-STABLE's
release notes already, and still others were MFC-ed before prior
releases.

Relocate dgm->digi release note item to live next to digi.

Relocate an(4)/AIR350 note to be next to another an(4) entry.

Remove a duplicate entry documenting xargs(1) -J.

Port markup fixes:  sysutils/cs9660_unicode, net/pim6dd, net/pim6sd.

23 years agoMarkup nit: Mark some things with <envar> that used be done with
bmah [Fri, 17 Aug 2001 17:36:50 +0000 (17:36 +0000)]
Markup nit:  Mark some things with <envar> that used be done with
<literal> or <varname>.

23 years agoUse a build-tool to create the .mgc files.
obrien [Fri, 17 Aug 2001 17:21:38 +0000 (17:21 +0000)]
Use a build-tool to create the .mgc files.

Submitted by: ru (partial)
Obtained from: NetBSD (basic idea)
Reviewed by: bde, ru

23 years agoNew/updated release notes: maxusers tunable, vidcontrol(1) -h,
bmah [Fri, 17 Aug 2001 17:18:39 +0000 (17:18 +0000)]
New/updated release notes:  maxusers tunable, vidcontrol(1) -h,
tmpname(3) use of TMPDIR.

23 years agoFix style nits.
obrien [Fri, 17 Aug 2001 16:54:09 +0000 (16:54 +0000)]
Fix style nits.

23 years agoRemoved the (possible) ambiguity in /etc/rc.conf comment.
ru [Fri, 17 Aug 2001 16:05:54 +0000 (16:05 +0000)]
Removed the (possible) ambiguity in /etc/rc.conf comment.

PR: bin/29736

23 years agoNew/modified release notes: Update isdn4bsd entry (users must
bmah [Fri, 17 Aug 2001 16:04:01 +0000 (16:04 +0000)]
New/modified release notes:  Update isdn4bsd entry (users must
use ispppcontrol when doing kernel PPP over ISDN), s/Ukranian/Ukrainian/.

MFCs noted:  cdcontrol(1) next/prev, sysctl(8) -N, sysctl(8) -ao/-ax and
-w deprecated, col(1) -p.

23 years agoRemoved comments that simply describe what is seen by ``cvs di -rRELENG_4''
ru [Fri, 17 Aug 2001 15:51:26 +0000 (15:51 +0000)]
Removed comments that simply describe what is seen by ``cvs di -rRELENG_4''
of that file.  (xtp(4) has already been MFC'ed.)

23 years agomdoc(7) police: collapse multiple spaces.
ru [Fri, 17 Aug 2001 15:25:55 +0000 (15:25 +0000)]
mdoc(7) police: collapse multiple spaces.

23 years agomdoc(7) police: fixed the fatal.
ru [Fri, 17 Aug 2001 15:18:49 +0000 (15:18 +0000)]
mdoc(7) police: fixed the fatal.

23 years agoNote that cdrecord is a part of the sysutils/cdrtools port.
bmah [Fri, 17 Aug 2001 15:16:12 +0000 (15:16 +0000)]
Note that cdrecord is a part of the sysutils/cdrtools port.

(While I'm here, change markup for a reference to the net/smbfs
port to use <port></port> tags.)

Submitted by: kris

23 years agoMFS:
bmah [Fri, 17 Aug 2001 15:01:39 +0000 (15:01 +0000)]
MFS:
s/compatability/compatibility/
s/compatable/compatible/
s/behaviour/behavior/

23 years agomdoc(7) police: markup nit.
ru [Fri, 17 Aug 2001 12:26:00 +0000 (12:26 +0000)]
mdoc(7) police: markup nit.

23 years agoThe ports that convert screendumps into text or png called scr2*, not
sobomax [Fri, 17 Aug 2001 11:32:25 +0000 (11:32 +0000)]
The ports that convert screendumps into text or png called scr2*, not
shot2*.

PR:
Submitted by:   Olexander Kunytsa

23 years agoMake source crunches work as well as object crunches; broken in the
joe [Fri, 17 Aug 2001 11:11:10 +0000 (11:11 +0000)]
Make source crunches work as well as object crunches; broken in the
last commit that fixed object crunches.

Repeat after me, "no excuses for not testing".

23 years agoSimplify overflow calculations a bit
ache [Fri, 17 Aug 2001 11:08:56 +0000 (11:08 +0000)]
Simplify overflow calculations a bit

23 years agoRemove extra check, already done in upper level caller, i.e. in
ache [Fri, 17 Aug 2001 10:43:03 +0000 (10:43 +0000)]
Remove extra check, already done in upper level caller, i.e. in
_fseeko()

23 years agoMention ftell & ftello in EOVERFLOW section too.
ache [Fri, 17 Aug 2001 10:29:09 +0000 (10:29 +0000)]
Mention ftell & ftello in EOVERFLOW section too.

23 years agoAdd more overflow checks in case of fseek()
ache [Fri, 17 Aug 2001 10:22:03 +0000 (10:22 +0000)]
Add more overflow checks in case of fseek()

23 years agoDon't clear "we have offset" flag even if long is overflow for fseek(),
ache [Fri, 17 Aug 2001 10:06:46 +0000 (10:06 +0000)]
Don't clear "we have offset" flag even if long is overflow for fseek(),
there is no harm to have it, it will reduce next call efforts.

23 years agofseek.c:
ache [Fri, 17 Aug 2001 09:57:11 +0000 (09:57 +0000)]
fseek.c:
Resulting fseek() offset must fit in long, required by POSIX (pointed by bde),
so add LONG_MAX and final tests for it.

rewind.c:
1) add missing __sinit() as in fseek() it pretends to be.
2) use clearerr_unlocked() since we already lock stream before _fseeko()
3) don't zero errno at the end, it explicitely required by POSIX as the
only one method to test rewind() error condition.
4) don't clearerr() if error happens in _fseeko()

23 years agocdrecord is now called cdrtools
kris [Fri, 17 Aug 2001 06:42:31 +0000 (06:42 +0000)]
cdrecord is now called cdrtools

Submitted by: stuart nichols <unstable@stac.state.tx.us>

23 years agoadd an entry about myself
fjoe [Fri, 17 Aug 2001 05:59:39 +0000 (05:59 +0000)]
add an entry about myself

Reviewed by: jdp

23 years agoRestore initialization of ahc->unpause lost in rev 1.76. ahc->unpause
gibbs [Fri, 17 Aug 2001 04:22:50 +0000 (04:22 +0000)]
Restore initialization of ahc->unpause lost in rev 1.76.  ahc->unpause
should be initialized to 0 for PCI adapters which was a side-effect of
the memset in ahc_alloc().  EISA/VL adapters require our unpause value to
include the "levelness" of our interrupt.

This change allows EISA adapters to work again.

PR: 29689, 29471

23 years agoFix part of another problem that bde pointed out. This is different
peter [Thu, 16 Aug 2001 23:43:24 +0000 (23:43 +0000)]
Fix part of another problem that bde pointed out.  This is different
to what bde suggested though.

23 years agoDefine _KERNEL as this grubs around where no userland should go.
obrien [Thu, 16 Aug 2001 20:47:01 +0000 (20:47 +0000)]
Define _KERNEL as this grubs around where no userland should go.

23 years agoRemove redundant null-termination. The buffer is already explicitly
peter [Thu, 16 Aug 2001 20:18:43 +0000 (20:18 +0000)]
Remove redundant null-termination.  The buffer is already explicitly
zeroed, and we intentionally leave -1 on the strncpy length to leave
the original \0.

Submitted by: bde

23 years agoDon't explicitly null-terminate. The buffer we are copying into is
peter [Thu, 16 Aug 2001 20:16:20 +0000 (20:16 +0000)]
Don't explicitly null-terminate.  The buffer we are copying into is
already zeroed, and we explicitly leave the last byte untouched.

Submitted by: bde

23 years agoAfter one more day of testing, make what I hope are the final tweaks to
wpaul [Thu, 16 Aug 2001 20:15:16 +0000 (20:15 +0000)]
After one more day of testing, make what I hope are the final tweaks to
prevent/workaround TX lockups in this driver. The secret seems to be to
not let the TX DMA queue become too full. If we have too many packets
in the queue, we should wait for them to drain a bit before trying to
queue more. This should prevent the lockup from occurring, and if it
does occur, there is special code in sf_start() to kick the NIC in the
head and get it going again.

Special thanks to Glen Neff for helping me test this fix.

23 years agoPrototype the ptrace_*()
obrien [Thu, 16 Aug 2001 19:57:36 +0000 (19:57 +0000)]
Prototype the ptrace_*()

23 years agoUpdate the OpenSSH minor-version string.
green [Thu, 16 Aug 2001 19:26:19 +0000 (19:26 +0000)]
Update the OpenSSH minor-version string.

Requested by: obrien
Reviewed by: rwatson

23 years agooops- typo in a previous commit
mjacob [Thu, 16 Aug 2001 17:39:45 +0000 (17:39 +0000)]
oops- typo in a previous commit

23 years agoFix a spelling error in a comment.
mjacob [Thu, 16 Aug 2001 17:31:27 +0000 (17:31 +0000)]
Fix a spelling error in a comment.

23 years agoFix the so-called "half-baked-probe" code that I wrote a long time ago
imp [Thu, 16 Aug 2001 17:30:55 +0000 (17:30 +0000)]
Fix the so-called "half-baked-probe" code that I wrote a long time ago
to properly clear the interrupt register on the no error case.  Also,
set the mcr register to zero when we find we can't support the chip.

This fixes the hang on sio driver attach problem in the new pci pccard
code that some people have reported.  At least on my machine.  I'd
like to get this into 4.4.

Submitted by: bde
PR: kern/29742

MFC after: 1 day

23 years agoAdd more MBOX and ASYNC event defines.
mjacob [Thu, 16 Aug 2001 17:26:03 +0000 (17:26 +0000)]
Add more MBOX and ASYNC event defines.

MFC after: 2 weeks

23 years agoThanks to PHK for spotting: ISPASYNC_UNHANDLED_RESPONSE not
mjacob [Thu, 16 Aug 2001 17:25:41 +0000 (17:25 +0000)]
Thanks to PHK for spotting: ISPASYNC_UNHANDLED_RESPONSE not
handle in isp_async.

23 years agoEnable LIP F8, LIP Reset async events.
mjacob [Thu, 16 Aug 2001 17:25:08 +0000 (17:25 +0000)]
Enable LIP F8, LIP Reset async events.
Be more chatty about SNS failures. Fix
typo for skipped phase mesage. Correct
MBOX_GET_PORT_QUEUE_PARAMS options in
table.

MFC after: 2 weeks

23 years agoMFS: first appeared in FreeBSD 4.4.
ru [Thu, 16 Aug 2001 17:19:30 +0000 (17:19 +0000)]
MFS: first appeared in FreeBSD 4.4.

23 years agoFix typo
julian [Thu, 16 Aug 2001 17:17:54 +0000 (17:17 +0000)]
Fix typo

Submitted by:  BDE
MFC after: 2 weeks

23 years agoOnly allocate teh 1540 byte buffer if we need it..
julian [Thu, 16 Aug 2001 17:16:31 +0000 (17:16 +0000)]
Only allocate teh 1540 byte buffer if we need it..
(lazy allocation)

MFC after: 13 days

23 years agoAdd two missing includes
brian [Thu, 16 Aug 2001 16:58:12 +0000 (16:58 +0000)]
Add two missing includes

23 years agoRemove an unused variable
brian [Thu, 16 Aug 2001 16:58:02 +0000 (16:58 +0000)]
Remove an unused variable

23 years agoPrint a space between the function name and line number.
mikeh [Thu, 16 Aug 2001 16:18:14 +0000 (16:18 +0000)]
Print a space between the function name and line number.

PR: bin/10980
MFC after: 2 weeks

23 years agoAdd $FreeBSD$
mikeh [Thu, 16 Aug 2001 16:16:07 +0000 (16:16 +0000)]
Add $FreeBSD$

MFC after: 2 weeks

23 years agoReduce stack allocation (stack-fast?).
mp [Thu, 16 Aug 2001 16:14:26 +0000 (16:14 +0000)]
Reduce stack allocation (stack-fast?).
    elf_load_file()   =>  352 to 52 bytes
    exec_elf_imgact() => 1072 to 48 bytes
    elf_corehdr()     =>  396 to  8 bytes

Reviewed by: julian

23 years agoFix the @monthly and @weekly shortcuts so that they actually run
mikeh [Thu, 16 Aug 2001 14:23:59 +0000 (14:23 +0000)]
Fix the @monthly and @weekly shortcuts so that they actually run
monthly and weekly, respectively. Also fix the @yearly shortcut so
that it doesn't execute daily during January. OpenBSD and NetBSD also
appear to have this bug.

PR: bin/21152

23 years agomdoc(7) police: restore markup bit that got accidentally lost in rev. 1.44.
ru [Thu, 16 Aug 2001 11:31:18 +0000 (11:31 +0000)]
mdoc(7) police: restore markup bit that got accidentally lost in rev. 1.44.

23 years agomdoc(7) police: replace `\*(Ba' with a simple `|', it's handled specially.
ru [Thu, 16 Aug 2001 11:09:00 +0000 (11:09 +0000)]
mdoc(7) police: replace `\*(Ba' with a simple `|', it's handled specially.

23 years agoFix a dangling {
obrien [Thu, 16 Aug 2001 10:32:42 +0000 (10:32 +0000)]
Fix a dangling {

23 years agoMinor style(9)'ing
obrien [Thu, 16 Aug 2001 10:13:34 +0000 (10:13 +0000)]
Minor style(9)'ing

23 years agostyle(9) and make consistent across platforms
obrien [Thu, 16 Aug 2001 09:29:35 +0000 (09:29 +0000)]
style(9) and make consistent across platforms

23 years agoDocument the sendmail_outbound_enable and sendmail_outbound_flags,
sheldonh [Thu, 16 Aug 2001 09:14:09 +0000 (09:14 +0000)]
Document the sendmail_outbound_enable and sendmail_outbound_flags,
which were introduced 5 months ago.  Looking at the descriptions,
these two look like the stupidest options to have arrived in a while,
but they must be documented now that they've been merged onto the
stable branch.

23 years agoUse the backwards compatability mechanisms so that ps/top etc dont have
peter [Thu, 16 Aug 2001 08:41:15 +0000 (08:41 +0000)]
Use the backwards compatability mechanisms so that ps/top etc dont have
unnecessary breakage.

While here, use explicit sizes for the string fields so that we dont
have unintentional changes again in the future when key tunables change.

This still is not quite right, but a june userland is happy with
a -current kernel with these tweaks.

23 years ago* Update author's e-mail address.
sheldonh [Thu, 16 Aug 2001 08:40:54 +0000 (08:40 +0000)]
* Update author's e-mail address.

* When the author of an utility also wrote its manual page (a rarety, I
  know), it is not necessary to say that the utility "and this manual
  page" were written by the author.

23 years agoUse explicit sizes for the prpsinfo command length string so that
peter [Thu, 16 Aug 2001 08:35:51 +0000 (08:35 +0000)]
Use explicit sizes for the prpsinfo command length string so that
we dont have any more unexpected changes in core dumps.  This gets us
back to the original core dump layout from a few days ago.

23 years agomdoc(7) police: markup nit, removed whitespace at EOL.
ru [Thu, 16 Aug 2001 08:26:04 +0000 (08:26 +0000)]
mdoc(7) police: markup nit, removed whitespace at EOL.

23 years agoUse the .Cm macro to mark up the modifiers of the -s flag's size argument.
sheldonh [Thu, 16 Aug 2001 08:19:09 +0000 (08:19 +0000)]
Use the .Cm macro to mark up the modifiers of the -s flag's size argument.

23 years agomdoc(7) police: markup nits.
ru [Thu, 16 Aug 2001 07:50:54 +0000 (07:50 +0000)]
mdoc(7) police: markup nits.

23 years agomdoc(7) police: Section cross-references are marked with .Sx.
ru [Thu, 16 Aug 2001 07:43:16 +0000 (07:43 +0000)]
mdoc(7) police: Section cross-references are marked with .Sx.
-compat is not a valid keyword.

23 years agoFix a bug in canceling joining threads.
jasone [Thu, 16 Aug 2001 06:31:32 +0000 (06:31 +0000)]
Fix a bug in canceling joining threads.
Do not detach canceled threads.

Reported by: Arno Klaassen <arno@heho.snv.jussieu.fr>
Collaboration with: deischen

23 years agoAdd the `WANT_FORCE_OPTIMIZATION_DOWNGRADE' knob. If set to an integer
obrien [Thu, 16 Aug 2001 06:05:18 +0000 (06:05 +0000)]
Add the `WANT_FORCE_OPTIMIZATION_DOWNGRADE' knob.  If set to an integer
value, it forces GCC to not optimize above this level.  For intance, GCC
made with "WANT_FORCE_OPTIMIZATION_DOWNGRADE=1" is a good setting for the
Alpha platform when building ports.

23 years agoRemove unused variable in example.
dd [Thu, 16 Aug 2001 06:02:20 +0000 (06:02 +0000)]
Remove unused variable in example.

23 years agoDon't allocate a 512 byte buffer on the stack in snplwrite. It's
dd [Thu, 16 Aug 2001 06:00:57 +0000 (06:00 +0000)]
Don't allocate a 512 byte buffer on the stack in snplwrite.  It's
probably harmless in this case, since the latter is called on tty
input, which is usually a result of some system call, so we've got
plenty of stack left.  It's still nice to fix these things, though, in
case somebody ever decides this driver is a good example of something
(perhaps "what you probably shouldn't do").

23 years agoIf the string specifying the allowed options starts with a leading `:',
jkoshy [Thu, 16 Aug 2001 03:27:03 +0000 (03:27 +0000)]
If the string specifying the allowed options starts with a leading `:',
`getopt(3)' should not print a warning for missing argument values.

PR: bin/29625
Reviewed by: mikeh
MFC after: 1 week

23 years agoExplain the relation of getchar() to getc() in less words.
dd [Thu, 16 Aug 2001 03:09:33 +0000 (03:09 +0000)]
Explain the relation of getchar() to getc() in less words.

Submitted by: ru