]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
22 years agoClarify that is[x]digit() class is the same in any locale
ache [Thu, 29 Nov 2001 15:23:46 +0000 (15:23 +0000)]
Clarify that is[x]digit() class is the same in any locale

22 years agoFix a bug about CIS string comparison. Pccardd should be able to distinguish
iwasaki [Thu, 29 Nov 2001 14:33:57 +0000 (14:33 +0000)]
Fix a bug about CIS string comparison.  Pccardd should be able to distinguish
card "MELCO" "LPC2-T"
and
card "MELCO" "LPC2-TX"
by this fix.

Reported by: Kitagawa Shoichi <sk@xstar.kiu.ac.jp>,
NINOMIYA Hideyuki <nin@shikoku.ne.jp>
MFC after: 1 week

22 years agoDuh. Back out most of my previous commit. It was mostly covered in the
cjc [Thu, 29 Nov 2001 09:44:33 +0000 (09:44 +0000)]
Duh. Back out most of my previous commit. It was mostly covered in the
STANDARDS section of the page. Add one remark there about inet_pton(3)
only understanding decimal values (in contrast to inet_aton(3) and
friends who are happy with 0ac.020.25 for 172.16.0.25).

Caught by: ru
MFC after: 2 days

22 years agoFixed memory leak in dialog_gauge(3).
ru [Thu, 29 Nov 2001 07:42:13 +0000 (07:42 +0000)]
Fixed memory leak in dialog_gauge(3).

PR: gnu/32260
Submitted by: Igor Pokrovsky <tiamat@telegraph.spb.ru>
MFC after: 3 days

22 years agoOops, actually fix the namespace pollution for atoll() as the previous
bde [Thu, 29 Nov 2001 07:08:56 +0000 (07:08 +0000)]
Oops, actually fix the namespace pollution for atoll() as the previous
commit claimed to do.

22 years agoStty is a program, not a function.
green [Thu, 29 Nov 2001 03:26:42 +0000 (03:26 +0000)]
Stty is a program, not a function.

22 years agoBump p1003.1-90 to p1003.1-96.
green [Thu, 29 Nov 2001 03:21:16 +0000 (03:21 +0000)]
Bump p1003.1-90 to p1003.1-96.

22 years agorename() can't guarantee "to" always exists if it didn't exist in the first
gshapiro [Thu, 29 Nov 2001 03:17:12 +0000 (03:17 +0000)]
rename() can't guarantee "to" always exists if it didn't exist in the first
place -- for example, rename("existing", "newfile"); on a read-only file
system.

Reviewed by: green
MFC after: 3 days

22 years agoImplement a more efficient way to assign addresses: read /etc/hosts
luigi [Thu, 29 Nov 2001 03:16:23 +0000 (03:16 +0000)]
Implement a more efficient way to assign addresses: read /etc/hosts
only once into an array of shell variables, and then scan the array
to find entries matching the MAC address.
Associative arrays would really be handy here...

22 years agoBack out national digits support, POSIX explicetely disallows it:
ache [Thu, 29 Nov 2001 03:03:55 +0000 (03:03 +0000)]
Back out national digits support, POSIX explicetely disallows it:

The definition of character class digit requires that only ten characters
-the ones defining digits- can be specified; alternate digits (for
example, Hindi or Kanji) cannot be specified here. However, the encoding
may vary if an implementation supports more than one encoding.

The definition of character class xdigit requires that the characters
included in character class digit are included here also and allows for
different symbols for the hexadecimal digits 10 through 15.

22 years agoAfter the long explanaition of how the inet_* functions interpret
cjc [Thu, 29 Nov 2001 00:56:45 +0000 (00:56 +0000)]
After the long explanaition of how the inet_* functions interpret
Internet addresses, point out that inet_pton(3) only understands
dotted quads with decimal values.

MFC after: 2 days

22 years agoAlphabetize the "Disks and Storage" section of the release notes. No
bmah [Wed, 28 Nov 2001 23:32:24 +0000 (23:32 +0000)]
Alphabetize the "Disks and Storage" section of the release notes.  No
content changes.

22 years agoUpdate release notes: ACPI imported version.
bmah [Wed, 28 Nov 2001 23:27:23 +0000 (23:27 +0000)]
Update release notes:  ACPI imported version.

22 years agoReturn a POSIX-compliant error for write(2) failure.
green [Wed, 28 Nov 2001 23:02:33 +0000 (23:02 +0000)]
Return a POSIX-compliant error for write(2) failure.

22 years agoAdd a BUGS section noting that the basename of a directory containing
cjc [Wed, 28 Nov 2001 21:43:53 +0000 (21:43 +0000)]
Add a BUGS section noting that the basename of a directory containing
periodic(8) scripts must be useable as a sh(1) variable.

MFC after: 2 days

22 years agoo Cache req->td->td_proc->p_ucred->cr_prison in pr to improve
rwatson [Wed, 28 Nov 2001 21:22:05 +0000 (21:22 +0000)]
o Cache req->td->td_proc->p_ucred->cr_prison in pr to improve
  readability.
o Conditionalize only the SYSCTL definitions for the regression
  tree, not the variables itself, decreasing the number of #ifdef
  REGRESSIONs scattered in kern_mib.c, and making the code more
  readable.

Sponsored by: DARPA, NAI Labs

22 years agoFixed namespace pollution and/or breakage of K&R and C90 support related to
bde [Wed, 28 Nov 2001 19:52:25 +0000 (19:52 +0000)]
Fixed namespace pollution and/or breakage of K&R and C90 support related to
the following functions in the following commits:
- atoll() in revs 1.23-1.25
- llabs() and lldiv() in revs 1.22
- strtoq() and strtouq() in revs 1.18
C99 functions must not be declared in C90/POSIX.1-1990 sections, and
"long long" must not be exposed to compilers that don't support it.

Fixed style bugs (mainly misindentation and disorder) related the
following functions in the following commits:
- atoll() in revs 1.23-1.25
- getprogname() in rev.1.21
- sranddev() in revs 1.19-1.20
- strtoq() and strtouq() in rev.1.13
- user_from_uid() in rev.1.1
Breakage of K&R and C90 support used to be avoided by conditializing the
"long long"s for strtoq() and strtouq() on __STRICT_ANSI__, but the
conditionals should have gone away in rev.1.13 when the "long long"s went
away (the problem was moved to the places that declare quad_t and u_quad_t).

22 years agoWhen a socket is not connected, allow the peer "struct sockaddr"
archie [Wed, 28 Nov 2001 19:39:58 +0000 (19:39 +0000)]
When a socket is not connected, allow the peer "struct sockaddr"
to be included in the meta information that is associated with
incoming and outgoing packets.

Reviewed by: julian
MFC after: 1 week

22 years agoFixed namespace pollution in previous commit. The C99 function
bde [Wed, 28 Nov 2001 19:02:01 +0000 (19:02 +0000)]
Fixed namespace pollution in previous commit.  The C99 function
sterror_r() must not be declared in the C90/POSIX.1-1990 section.
Put it in the nonstandard section for now.

22 years agoFix indentation after removing GEMDOS support. Whitespace changes only.
jhb [Wed, 28 Nov 2001 18:29:16 +0000 (18:29 +0000)]
Fix indentation after removing GEMDOS support.  Whitespace changes only.

22 years agoUse suser_td() instead of explicitly checking cr_uid against 0.
jhb [Wed, 28 Nov 2001 18:25:39 +0000 (18:25 +0000)]
Use suser_td() instead of explicitly checking cr_uid against 0.

PR: kern/21809
Submitted by: <mbendiks@eunet.no>
Reviewed by: rwatson

22 years ago- Change parameters of signal handlers in order to be correct (they are
arr [Wed, 28 Nov 2001 17:29:04 +0000 (17:29 +0000)]
- Change parameters of signal handlers in order to be correct (they are
  handed a integer, not void).
- No need to set flags to zero when they already will be.
- It was also noted the manner in which the signal handling has changed
  might possibly generate some problems (hangs possibly) -- these, while
  remaining in the code, will be fixed shortly (within a day).

Submitted by: bde

22 years agoWhitespace fixes so thre 386 versoion of __uint8_swap_uint32 is easier to
jhb [Wed, 28 Nov 2001 17:10:57 +0000 (17:10 +0000)]
Whitespace fixes so thre 386 versoion of __uint8_swap_uint32 is easier to
read.

22 years agoRemove ciss from the modules build until the build issues are resolved.
imp [Wed, 28 Nov 2001 17:08:51 +0000 (17:08 +0000)]
Remove ciss from the modules build until the build issues are resolved.

22 years agoAxe more unused GEMDOS code that was #ifdef atari.
jhb [Wed, 28 Nov 2001 16:56:42 +0000 (16:56 +0000)]
Axe more unused GEMDOS code that was #ifdef atari.

PR: kern/21809
Submitted by: <mbendiks@eunet.no>

22 years agoFix a typo in a comment
brian [Wed, 28 Nov 2001 16:15:52 +0000 (16:15 +0000)]
Fix a typo in a comment

22 years agoRemove the need for an expensive m_devget on the i386, which does not
luigi [Wed, 28 Nov 2001 16:10:37 +0000 (16:10 +0000)]
Remove the need for an expensive m_devget on the i386, which does not
have alignment problems.

On small boxes (e.g. the net4501 from Soekris, featuring a 486/133)
this provides huge performance benefits: the peak forwarding rate
with avg.sized packets goes up by 50-70% because of this change
alone. Faster CPUs might benefit less from this change, but in any
case the CPU has better things to do than waste time on useless
memory-to-memory copies.

Several drivers (for Tulip-like cards) might benefit from a similar
change.

Right now the new behaviour is controlled by a sysctl variable,
hw.sis_quick which defaults to 1 (on), you can set it to 0 to
reintroduce the old behaviour (and compare the results).  The
variable is only there to show how much you can gain with this
change, it will go away soon.

Also, slightly simplify the code to initialize the ring buffers,
and remove a couple of dangerous printf's which could trigger on
any packet in case of mbuf shortage.

MFC-after: 3 days

22 years agoWaits for a keypress before rebooting on panic.
dcs [Wed, 28 Nov 2001 14:00:36 +0000 (14:00 +0000)]
Waits for a keypress before rebooting on panic.

PR: kern/32351
Submitted by: Jonathan Mini <mini@haikugeek.com>
MFC after: 1 week

22 years agoRegen after addition of the Mustek 1200 CU Plus scanner.
sobomax [Wed, 28 Nov 2001 12:38:15 +0000 (12:38 +0000)]
Regen after addition of the Mustek 1200 CU Plus scanner.

Submitted by: "Ilmar S. Habibulin" <ilmar@watson.org>

22 years agoAdd Mustek 1200 CU Plus scanner.
sobomax [Wed, 28 Nov 2001 12:37:04 +0000 (12:37 +0000)]
Add Mustek 1200 CU Plus scanner.

Submitted by: "Ilmar S. Habibulin" <ilmar@watson.org>

22 years agoYet another synch with minor changes in the ACPI CA 20011120 snapshot.
iwasaki [Wed, 28 Nov 2001 12:09:42 +0000 (12:09 +0000)]
Yet another synch with minor changes in the ACPI CA 20011120 snapshot.
We need to call AcpiEnterSleepStatePrep() before AcpiEnterSleepState().

22 years agoDo not write warning messages on stderr - a caller is expected to check
tobez [Wed, 28 Nov 2001 10:55:02 +0000 (10:55 +0000)]
Do not write warning messages on stderr - a caller is expected to check
the return code and errno instead.  Those warnings did not do any good
for daemonized users of initgroups(3), and confused cvs clients that
communicated with non-root cvs pserver.

The committed fix differs from the one suggested in the PR, and was
submitted by ru.

PR: 15421
Approved by: markm
Discussed on: -stable, -current at various times

22 years agoMdoc police.
des [Wed, 28 Nov 2001 10:07:21 +0000 (10:07 +0000)]
Mdoc police.

Submitted by: ru

22 years agoRewrite buffer handling code a bit to handle large values.
ache [Wed, 28 Nov 2001 09:50:24 +0000 (09:50 +0000)]
Rewrite buffer handling code a bit to handle large values.
Add more checks for data overflow.

22 years agoGC cvs-1.8.1 import left-overs.
ru [Wed, 28 Nov 2001 09:31:56 +0000 (09:31 +0000)]
GC cvs-1.8.1 import left-overs.

22 years agoIncrease internal buffer space
ache [Wed, 28 Nov 2001 09:26:57 +0000 (09:26 +0000)]
Increase internal buffer space

22 years agomdoc(7) police: fix one pam_unix(8) left-over, sort xrefs.
ru [Wed, 28 Nov 2001 09:25:03 +0000 (09:25 +0000)]
mdoc(7) police: fix one pam_unix(8) left-over, sort xrefs.

22 years agoWhitespace police.
ru [Wed, 28 Nov 2001 08:52:35 +0000 (08:52 +0000)]
Whitespace police.

Submitted by: cjc, ru

22 years agoAdd Buffalo AirStation 11Mbps CF WLAN card.
imp [Wed, 28 Nov 2001 08:35:08 +0000 (08:35 +0000)]
Add Buffalo AirStation 11Mbps CF WLAN card.

Obtained from: NetBSD (ichiro)

22 years agoSync to 1.24
imp [Wed, 28 Nov 2001 08:33:41 +0000 (08:33 +0000)]
Sync to 1.24

22 years agoMerge from NetBSD:
imp [Wed, 28 Nov 2001 08:33:12 +0000 (08:33 +0000)]
Merge from NetBSD:
1.147 (my changes to sync to FreeBSD's file)
1.148 Buffalo AirStation CFtytpe WLAN Card (ichiro)
1.149 IBM SCSI PCMCIA (yamt)
1.150 <restore 1.148 lost in 1.149> (enami)

22 years agoAdd hi_IN.ISCII-DEV
ache [Wed, 28 Nov 2001 07:23:11 +0000 (07:23 +0000)]
Add hi_IN.ISCII-DEV

22 years agoDon't ever assume that isdigit() is always subset of isxdigit()
ache [Wed, 28 Nov 2001 06:06:27 +0000 (06:06 +0000)]
Don't ever assume that isdigit() is always subset of isxdigit()

22 years agoFix typo
ache [Wed, 28 Nov 2001 05:43:24 +0000 (05:43 +0000)]
Fix typo

22 years agoAdd tip breakage workaround.
imp [Wed, 28 Nov 2001 05:43:10 +0000 (05:43 +0000)]
Add tip breakage workaround.

22 years agoUse stricter tests to disallow national digits > 9
ache [Wed, 28 Nov 2001 05:39:21 +0000 (05:39 +0000)]
Use stricter tests to disallow national digits > 9
Optimize national digits code a bit

22 years agoAllow national (non-ASCII) digits
ache [Wed, 28 Nov 2001 04:57:48 +0000 (04:57 +0000)]
Allow national (non-ASCII) digits

22 years agoSynch with minor changes in the ACPI CA 20011120 snapshot.
msmith [Wed, 28 Nov 2001 04:36:29 +0000 (04:36 +0000)]
Synch with minor changes in the ACPI CA 20011120 snapshot.

22 years agoMerge local changes.
msmith [Wed, 28 Nov 2001 04:34:52 +0000 (04:34 +0000)]
Merge local changes.

Add a fix for a minor error in the PCI routing table creation handler
where the correct size for the buffer is not returned.

22 years agoMy recent changes to add the ctl_renametf routine assumed that print job
gad [Wed, 28 Nov 2001 04:30:47 +0000 (04:30 +0000)]
My recent changes to add the ctl_renametf routine assumed that print job
control-files will always start with 'cfA*'.  It turns out that some
implementations of lpd (such as solaris) may send a control file which
starts with 'cfB*', or really 'cf<anyLetter>*'.  Although such filenames
are very odd, we did used to accept them.  This changes ctl_renametf to
work correctly with them, and fixes up 'lpc clean' to match.

PR: bin/32183
MFC after: 10 days

22 years agoThis commit was generated by cvs2svn to compensate for changes in r87031,
msmith [Wed, 28 Nov 2001 04:29:40 +0000 (04:29 +0000)]
This commit was generated by cvs2svn to compensate for changes in r87031,
which included commits to RCS files with non-trunk default branches.

22 years agoVendor import of the Intel ACPI CA 20011120 snapshot.
msmith [Wed, 28 Nov 2001 04:29:40 +0000 (04:29 +0000)]
Vendor import of the Intel ACPI CA 20011120 snapshot.

22 years agoMake sure the security check output includes a To: line in the
silby [Wed, 28 Nov 2001 04:07:03 +0000 (04:07 +0000)]
Make sure the security check output includes a To: line in the
same way the daily run output does.

22 years agoAllow national (non-ASCII) digits
ache [Wed, 28 Nov 2001 03:57:12 +0000 (03:57 +0000)]
Allow national (non-ASCII) digits

22 years agoImplement strtoimax() and strtoumax()
fenner [Wed, 28 Nov 2001 03:37:06 +0000 (03:37 +0000)]
Implement strtoimax() and strtoumax()

22 years agoAlso mention "long long" in synopsis.
fenner [Wed, 28 Nov 2001 03:36:05 +0000 (03:36 +0000)]
Also mention "long long" in synopsis.

22 years agoReturn a more meaningful errno when the length of the interpreter
jwd [Wed, 28 Nov 2001 03:26:58 +0000 (03:26 +0000)]
Return a more meaningful errno when the length of the interpreter
exceeds MAXSHELLCMDLEN to avoid secondary /bin/sh execution.

Update execve man page to reflect change.

Increase MAXSHELLCMDLEN to a slightly more meaningful value.

PR: kern/32106
Submitted by: b@etek.chalmers.se
Reviewed by: bsd
MFC after: 2 weeks

22 years agoDont print the sysctl node tree unless you're root.
peter [Wed, 28 Nov 2001 03:11:16 +0000 (03:11 +0000)]
Dont print the sysctl node tree unless you're root.

Found by: jkb (Yahoo OS troublemaker)

22 years agoBase 36 is allowed.
fenner [Wed, 28 Nov 2001 02:35:35 +0000 (02:35 +0000)]
Base 36 is allowed.

22 years agoChange to using my custom sparc64 binutils.
obrien [Wed, 28 Nov 2001 02:35:02 +0000 (02:35 +0000)]
Change to using my custom sparc64 binutils.

22 years agoWhitespace formatting
ache [Wed, 28 Nov 2001 02:18:53 +0000 (02:18 +0000)]
Whitespace formatting

22 years agoPut back base > 35 check. If someone dislike it, plese discuss it with
ache [Wed, 28 Nov 2001 02:17:22 +0000 (02:17 +0000)]
Put back base > 35 check. If someone dislike it, plese discuss it with
standards group first.

22 years agoFix just added atoll prototype
ache [Wed, 28 Nov 2001 01:51:00 +0000 (01:51 +0000)]
Fix just added atoll prototype

22 years agoClean up some of the gross macros whitespace wise before I fix the asm
jhb [Wed, 28 Nov 2001 01:42:16 +0000 (01:42 +0000)]
Clean up some of the gross macros whitespace wise before I fix the asm
constraints.

22 years agoAdd atoll(3) to conform POSIX and C99
ache [Wed, 28 Nov 2001 01:22:08 +0000 (01:22 +0000)]
Add atoll(3) to conform POSIX and C99

22 years agoUnderstand national (non-ASCII) digits now
ache [Wed, 28 Nov 2001 00:48:11 +0000 (00:48 +0000)]
Understand national (non-ASCII) digits now
Allow bases >=36 again
Misc cleanup

22 years agoAdd safeguard to prevent charmap symbol length overflowing
ache [Tue, 27 Nov 2001 23:39:14 +0000 (23:39 +0000)]
Add safeguard to prevent charmap symbol length overflowing

22 years agoAdd hi_IN.ISCII-DEV
ache [Tue, 27 Nov 2001 23:24:41 +0000 (23:24 +0000)]
Add hi_IN.ISCII-DEV

22 years agoSupport the 'ciss' device node.
msmith [Tue, 27 Nov 2001 23:12:00 +0000 (23:12 +0000)]
Support the 'ciss' device node.

MFC after: 1 week

22 years agoMove extern definitions to common.h
ache [Tue, 27 Nov 2001 23:11:05 +0000 (23:11 +0000)]
Move extern definitions to common.h
Increase charmap symbol length to 64

22 years agoAdd the 'ciss' driver, which supports the Compaq SmartRAID 5* family of
msmith [Tue, 27 Nov 2001 23:08:37 +0000 (23:08 +0000)]
Add the 'ciss' driver, which supports the Compaq SmartRAID 5* family of
RAID controllers (5300, 532, 5i, etc.)

Thanks to Compaq and Yahoo! for support during the development of this
driver.

MFC after: 1 week

22 years agoRemeber to set the permissions on an md-backed /tmp to 01777, rather than
rwatson [Tue, 27 Nov 2001 22:12:27 +0000 (22:12 +0000)]
Remeber to set the permissions on an md-backed /tmp to 01777, rather than
using the default of 0755.

22 years agoMention entropy_dir="NO" in diskless.8, since /var is an md-backed
rwatson [Tue, 27 Nov 2001 21:35:47 +0000 (21:35 +0000)]
Mention entropy_dir="NO" in diskless.8, since /var is an md-backed
file system and as such will not preserve its contents across boot.

22 years agoRemove GEMDOS support from msdosfs. I don't think anyone is going to
jhb [Tue, 27 Nov 2001 21:00:15 +0000 (21:00 +0000)]
Remove GEMDOS support from msdosfs.  I don't think anyone is going to
port FreeBSD to Atari machines any time soon.

22 years agoUse 'mov' instead of 'lea' for setting the syscall number in %eax as that
jhb [Tue, 27 Nov 2001 20:51:26 +0000 (20:51 +0000)]
Use 'mov' instead of 'lea' for setting the syscall number in %eax as that
is clearer about what we are actually doing.

Requested by: bde

22 years agoUpdate diskless.8 to refer to 'device md' instead of 'options MFS'; in
rwatson [Tue, 27 Nov 2001 20:47:38 +0000 (20:47 +0000)]
Update diskless.8 to refer to 'device md' instead of 'options MFS'; in
-CURRENT, md devices with FFS on top are used in the diskless boot process
instead of MFS.

22 years agoDocument the "console" setting in loader.conf, suggesting vidconsole
rwatson [Tue, 27 Nov 2001 20:37:40 +0000 (20:37 +0000)]
Document the "console" setting in loader.conf, suggesting vidconsole
or comconsole.

22 years agoipfw_modevent(): Don't use an unnatural block to define a variable
dd [Tue, 27 Nov 2001 20:32:47 +0000 (20:32 +0000)]
ipfw_modevent(): Don't use an unnatural block to define a variable
(fcp) that's already defined in the outer block and isn't used
anywhere else.  This silences -Wunused.

Reviewed by: md5(1)

22 years agoRemove debugging printfs that weren't conditional on any debugging
dd [Tue, 27 Nov 2001 20:28:48 +0000 (20:28 +0000)]
Remove debugging printfs that weren't conditional on any debugging
options in handling MOD_{UN,}LOAD (they weren't very useful, anyway).

22 years agoMake the default kernel prefix "kernel:" instead of the boot file,
dd [Tue, 27 Nov 2001 20:02:18 +0000 (20:02 +0000)]
Make the default kernel prefix "kernel:" instead of the boot file,
with the old behavior available via the -o option (it might still be
useful if one has many kernels and cares which messages came from
which).  If the boot file is not used as the prefix, it is still
logged once at startup.

This change is prompted by the fact that the boot file is now much
longer ("/boot/kernel/kernel" vs. "/kernel"), which significanlty
bloats the syslogd output.

Reviewed by: peter

22 years agoIn icmp_reflect(): If the packet was not addressed to us and was
dd [Tue, 27 Nov 2001 19:58:09 +0000 (19:58 +0000)]
In icmp_reflect(): If the packet was not addressed to us and was
received on an interface without an IP address, try to find a
non-loopback AF_INET address to use.  If that fails, drop it.
Previously, we used the address at the top of the in_ifaddrhead list,
which didn't make much sense, and would cause a panic if there were no
AF_INET addresses configured on the system.

PR: 29337, 30524
Reviewed by: ru, jlemon
Obtained from: NetBSD

22 years agoImplement a "kill" DDB command which is an interface to psignal() that
dd [Tue, 27 Nov 2001 19:56:28 +0000 (19:56 +0000)]
Implement a "kill" DDB command which is an interface to psignal() that
respects locks.  Before SMPng, one was able to call psignal()
using the "call" command, but this is no longer possible because it
does not respect locks by itself.  This is very useful when one has
gotten their machine into a state where it is impossible to spawn
ps/kill or su to root.

In this case, respecting locks essentially means trying to aquire the
proc lock before calling psignal().  We can't block in the debugger,
so if trylock fails, the operation fails.  This also means that we
can't use pfind(), since that will attempt to lock the process for us.

Reviewed by: jhb

22 years agoFix the example of suggested default settings. It stated that settings
sheldonh [Tue, 27 Nov 2001 19:39:07 +0000 (19:39 +0000)]
Fix the example of suggested default settings.  It stated that settings
were only of benefit to large filesystems, which recent research
suggests is not the case, and which the original author of the text
no longer endorses.

22 years agoMake extattrctl WARNS?=2-safe:
jedgar [Tue, 27 Nov 2001 18:58:56 +0000 (18:58 +0000)]
Make extattrctl WARNS?=2-safe:
o remove extraneous extern's
o prototype functions
o combine multiple return (0)'s into a single return (0) at the
  end of main()

Approved by: rwatson
Obtained from: TrustedBSD Project

22 years agoThe fxp driver depends on device miibus now.
jdp [Tue, 27 Nov 2001 18:51:08 +0000 (18:51 +0000)]
The fxp driver depends on device miibus now.

22 years agoAdd include of net/route.h, as structures moved around due to the
rwatson [Tue, 27 Nov 2001 17:36:39 +0000 (17:36 +0000)]
Add include of net/route.h, as structures moved around due to the
syncache rely on 'struct route' being defined.  This fixes the
LINT build some.

22 years agos/Kenned y/Kennedy/
phk [Tue, 27 Nov 2001 17:10:54 +0000 (17:10 +0000)]
s/Kenned y/Kennedy/

Pointed out by: $a_lot_of_people

22 years agoCorrect the example introduced in rev 1.29, which suggested a block:frag
sheldonh [Tue, 27 Nov 2001 17:01:17 +0000 (17:01 +0000)]
Correct the example introduced in rev 1.29, which suggested a block:frag
size ratio other than 8:1.  Currently, we only recommend an 8:1
ratio, because the impact of others ratios has not been adequately
investigated.

Also, do not recommend the use of the -c option in the example, since
newfs now automatically calculates the best cyl:cylgrp ratio.

This change was discussed with the author of rev 1.29.

22 years agoThis string of coincidences was too good to let pass.
phk [Tue, 27 Nov 2001 16:53:43 +0000 (16:53 +0000)]
This string of coincidences was too good to let pass.

Obtained from: BBC World Service on 9410kHz

22 years agoFix a bug in the driver -- under load, the receive unit could become
luigi [Tue, 27 Nov 2001 16:29:11 +0000 (16:29 +0000)]
Fix a bug in the driver -- under load, the receive unit could become
idle and the driver would not detect the event, requiring userland
to cycle the interface to bring it up again.
The fix consists in adding SIS_IMR_RX_IDLE to the interrupt mask and
add a command in sis_intr() to restart the receiver when this happens.

While at it, make the test of status bits more efficient.

22 years agoCFLAGS is not suitable for `optimize' thing; it spams Config.pm
ru [Tue, 27 Nov 2001 16:28:26 +0000 (16:28 +0000)]
CFLAGS is not suitable for `optimize' thing; it spams Config.pm
with -nostdinc which breaks Perl ports building.

This change will be re-applied if BDE agrees on having COPTFLAGS
in userland.

PR: bin/31877

22 years agoAdd a pam_set_item(3) man page with an MLINK to pam_get_item(3).
des [Tue, 27 Nov 2001 15:36:35 +0000 (15:36 +0000)]
Add a pam_set_item(3) man page with an MLINK to pam_get_item(3).

PR: docs/32294
Sponsored by: DARPA, NAI Labs
MFC after: 3 days

22 years agoDue to incorrect include ordering, <dlfcn.h> did not get included, so
des [Tue, 27 Nov 2001 15:36:21 +0000 (15:36 +0000)]
Due to incorrect include ordering, <dlfcn.h> did not get included, so
RTLD_NOW got incorrectly defined to 1 (which is RTLD_LAZY in FreeBSD).
In addition, the comment about FreeBSD requiring SHLIB_SYM_PREFIX to
be "_" is incorrect.

Submitted by: tobez (except for the bit about the incorrect comment)

22 years agoIt's idiotic to return PAM_PERM_DENIED when the item argument is NULL;
des [Tue, 27 Nov 2001 15:35:44 +0000 (15:35 +0000)]
It's idiotic to return PAM_PERM_DENIED when the item argument is NULL;
PAM_BUF_ERR is much closer to the truth.

22 years agoAdd .Lb entry for -lpam.
ru [Tue, 27 Nov 2001 15:08:42 +0000 (15:08 +0000)]
Add .Lb entry for -lpam.

Requested by: des

22 years agoDon't automatically unbind/deallocate memory when releasing.
ru [Tue, 27 Nov 2001 14:12:21 +0000 (14:12 +0000)]
Don't automatically unbind/deallocate memory when releasing.
This fixes the VT switching problem with the i810 X driver.

Explained by: David Dawes <dawes@XFree86.Org>
Reviewed by: dfr

22 years agofixed the cast128 calculation with a short cipher key length.
ume [Tue, 27 Nov 2001 14:11:47 +0000 (14:11 +0000)]
fixed the cast128 calculation with a short cipher key length.
the memory was overridden when the key length was less than 16 bytes.

Obtained from: KAME
MFC after: 1 week

22 years agoAdd support for a last-close handler.
des [Tue, 27 Nov 2001 13:26:27 +0000 (13:26 +0000)]
Add support for a last-close handler.
Revert the module version bumps; they're quite pointless as long as the
only pseudofs consumer is linprocfs, which is in the tree.

22 years agoRe-connect NTP docs to build.
ru [Tue, 27 Nov 2001 13:20:07 +0000 (13:20 +0000)]
Re-connect NTP docs to build.

22 years agoRemove check for the DISPLAY environment variable. It's easier to use
tg [Tue, 27 Nov 2001 13:14:49 +0000 (13:14 +0000)]
Remove check for the DISPLAY environment variable. It's easier to use
the `-x' option to get an X window than to temporarily disable DISPLAY.

PR: 32180