]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
23 years agoAdded the linux_sysinfo function to implement sysinfo(2).
Jim Pirzyk [Mon, 23 Jul 2001 06:22:10 +0000 (06:22 +0000)]
Added the linux_sysinfo function to implement sysinfo(2).

PR: kern/27759
Reviewed by: marcel
Approved by: marcel
MFC after: 1 week

23 years agoAdded the proper arguments the sysinfo system call
Jim Pirzyk [Mon, 23 Jul 2001 06:17:34 +0000 (06:17 +0000)]
Added the proper arguments the sysinfo system call

PR: kern/27759
Reviewed by: marcel
Approved by: marcel
Obtained from: Linux man page sysinfo(2)
MFC after: 1 week

23 years agoIf the user sets SENDMAIL_MC to /etc/mail/sendmail.mc, install is told to
Gregory Neil Shapiro [Mon, 23 Jul 2001 02:19:02 +0000 (02:19 +0000)]
If the user sets SENDMAIL_MC to /etc/mail/sendmail.mc, install is told to
install /etc/mail/sendmail.cf to /etc/mail/sendmail.cf and exits with an
error:

===> etc/sendmail
install -c -o root -g wheel -m 644 /etc/mail/sendmail.cf /etc/mail/sendmail.cf
install: /etc/mail/sendmail.cf and /etc/mail/sendmail.cf are the same file
*** Error code 64

Catch this in the Makefile and don't call install if the source and target
are the same file.

Reported by: Alexandr Listopad <laa@reis.zp.ua>
MFC after: 1 week

23 years agoNew make knob, SENDMAIL_M4_FLAGS, modifies the flags passed to m4 when
Gregory Neil Shapiro [Mon, 23 Jul 2001 01:33:54 +0000 (01:33 +0000)]
New make knob, SENDMAIL_M4_FLAGS, modifies the flags passed to m4 when
building a .cf file from a .mc file.

Include -D_FFR_TLS_O_T to enable tls policy control since the sendmail binary
build enables that FFR as well.

PR: conf/28361
MFC after: 1 week

23 years agoGet rid of a compile-time warning by casting to (size_t).
Garance A Drosehn [Mon, 23 Jul 2001 01:19:59 +0000 (01:19 +0000)]
Get rid of a compile-time warning by casting to (size_t).

MFC after: 1 week

23 years agoGet rid of a compile-time warning by casting to (size_t).
Garance A Drosehn [Mon, 23 Jul 2001 01:13:20 +0000 (01:13 +0000)]
Get rid of a compile-time warning by casting to (size_t).

MFC after: 1 week

23 years agoFix a few more minor compile-time warnings, mainly by using size_t where
Garance A Drosehn [Mon, 23 Jul 2001 01:08:15 +0000 (01:08 +0000)]
Fix a few more minor compile-time warnings, mainly by using size_t where
appropriate, and using '("%lu", (unsigned long)asizeval)' when printing
something of type size_t.

MFC after: 1 week

23 years agoGet rid of some compile-time warnings by defining (or casting) variables
Garance A Drosehn [Mon, 23 Jul 2001 00:15:37 +0000 (00:15 +0000)]
Get rid of some compile-time warnings by defining (or casting) variables
as size_t where appropriate.

MFC after: 1 week

23 years agoFix two compile-time warnings by defining fromlen as socklen_t.
Garance A Drosehn [Mon, 23 Jul 2001 00:13:02 +0000 (00:13 +0000)]
Fix two compile-time warnings by defining fromlen as socklen_t.

MFC after: 1 week

23 years agoadd %option nounput
Assar Westerlund [Sun, 22 Jul 2001 23:15:14 +0000 (23:15 +0000)]
add %option nounput

23 years agorevert last commit. using %option nounput is better
Assar Westerlund [Sun, 22 Jul 2001 23:14:33 +0000 (23:14 +0000)]
revert last commit.  using %option nounput is better

noted by: Seth Kingsley <seth.kingsley@windriver.com>

23 years agoadd ecalloc, emalloc, erealloc, estrdup - versions of the e-less
Assar Westerlund [Sun, 22 Jul 2001 22:26:37 +0000 (22:26 +0000)]
add ecalloc, emalloc, erealloc, estrdup - versions of the e-less
functions that exit instead of failing

23 years agoAdd a few missing spl calls in preparation for an MFC of the miibus
Ian Dowse [Sun, 22 Jul 2001 21:39:54 +0000 (21:39 +0000)]
Add a few missing spl calls in preparation for an MFC of the miibus
support.

23 years agoNote that the umass device requires scbus and da
Kris Kennaway [Sun, 22 Jul 2001 21:35:02 +0000 (21:35 +0000)]
Note that the umass device requires scbus and da

23 years agoRemove an old hack that remembered if the card type was Linksys by
Ian Dowse [Sun, 22 Jul 2001 21:31:44 +0000 (21:31 +0000)]
Remove an old hack that remembered if the card type was Linksys by
storing a flag in the global variable 'linksys' during the probe
routine and reading it during the attach routine. We now have the
ED_VENDOR_LINKSYS vendor code stored in sc->vendor, so check that
instead.

23 years agomake driver compile under 4.3-STABLE
Hellmuth Michaelis [Sun, 22 Jul 2001 19:55:05 +0000 (19:55 +0000)]
make driver compile under 4.3-STABLE

23 years agoDon't do sleep state transition if specified sleep state is not
Mitsuru IWASAKI [Sun, 22 Jul 2001 19:13:54 +0000 (19:13 +0000)]
Don't do sleep state transition if specified sleep state is not
supported by the system.

23 years agoAttach the miibus for Linksys/Dlink cards from the attach routine,
Ian Dowse [Sun, 22 Jul 2001 18:58:44 +0000 (18:58 +0000)]
Attach the miibus for Linksys/Dlink cards from the attach routine,
not from the probe routine. This was an oversight when I originally
ported the miibus support to -current, though it was mostly harmless.

We now set the vendor code to the new value ED_VENDOR_LINKSYS in
ed_pccard_Linksys() at probe time. Then ed_pccard_attach() checks
the vendor code, and sets up the miibus if appropriate.

Reviewed by: imp

23 years agoAs per further discussions on hackers redo the SIGCHLD patch to not generate
Matthew Dillon [Sun, 22 Jul 2001 18:47:31 +0000 (18:47 +0000)]
As per further discussions on hackers redo the SIGCHLD patch to not generate
an unexpected user-visible side effect with the sigaction flags.  Also cleanup
a minor union issue.

Submitted by: Rudolf Cejka <cejkar@dcse.fee.vutbr.cz>
MFC addendum: MFC will be combined w/ original commit
MFC after: 3 days

23 years agoFix a bug I introduced in Rev. 1.21 that prevented 'whois test' from
Mike Barcroft [Sun, 22 Jul 2001 18:23:24 +0000 (18:23 +0000)]
Fix a bug I introduced in Rev. 1.21 that prevented 'whois test' from
working.

Reviewed by: des
Approved by: des

23 years agoAdd entry for SMC 2632W card.
Warner Losh [Sun, 22 Jul 2001 18:13:19 +0000 (18:13 +0000)]
Add entry for SMC 2632W card.

Submitted by: lots of people, most recently by Mike Buchanon

Also, Tried to clean up the comments about IRQs to match the new world
order.

23 years agoRemove namespace pollution.
Mike Barcroft [Sun, 22 Jul 2001 18:10:08 +0000 (18:10 +0000)]
Remove namespace pollution.

PR: 14327
Reviewed by: des
Approved by: des
MFC after: 7 days

23 years ago- Fix "off by one" error in VT_WAITACTIVATE. Correctly accept
Kazutaka YOKOTA [Sun, 22 Jul 2001 13:58:23 +0000 (13:58 +0000)]
- Fix "off by one" error in VT_WAITACTIVATE. Correctly accept
  0 as meaning the requesting vty.
- Accept 0 as the requesting vty in VT_ACTIVATE as in VT_WAITACTIVE.

PR: 24423
MFC after: 10 days

23 years agoUse the correct ioctl command to set a VESA mode.
Kazutaka YOKOTA [Sun, 22 Jul 2001 13:30:32 +0000 (13:30 +0000)]
Use the correct ioctl command to set a VESA mode.

PR: 24423
MFC after: 10 days

23 years agoFix some bugs and general brain damage in mounttab:
Ian Dowse [Sun, 22 Jul 2001 12:17:51 +0000 (12:17 +0000)]
Fix some bugs and general brain damage in mounttab:
- Declare mtabhead as an extern in mounttab.h and define it only in
  mounttab.c.
- Remove shared global `verbose' and instead pass it as a parameter.
- Remove the `mtabp' argument to read_mtab(). It served no purpose
  whatsoever, although read_mtab() did use it as a temporary local
  variable.
- Don't check for impossible conditions when parsing mounttab, and
  do detect zero-length fields.
- Correctly test for strtoul() failures - just testing ERANGE is wrong.
- Include a field name in syslog errors, and avoid passing NULL to
  a syslog %s field.
- Don't test if arrays are NULL.
- If there are duplicates when writing out mounttab, keep the last
  entry instead of the first, as it will have a later timestamp.
- Fix a few formatting issues.

Update rpc.umntall and umount to match the mounttab interface changes.

23 years agoFix some man page xrefs.
Mike Pritchard [Sun, 22 Jul 2001 11:51:11 +0000 (11:51 +0000)]
Fix some man page xrefs.

PR: docs/26065
MFC after: 1 week

23 years agoChange the errx() example. Error messages passed to the err(3)
Dima Dorfman [Sun, 22 Jul 2001 11:47:15 +0000 (11:47 +0000)]
Change the errx() example.  Error messages passed to the err(3)
functions shouldn't have the first word capitalized, and shouldn't
have a period at the end.  This is how most of our programs, and most
(all?) of the 4.4BSD programs, are.  In the past, we've even done
sweeps to change things to comply to this.

23 years agoAdd: ``If you have to wrap a long statement, put the operator at the
Dima Dorfman [Sun, 22 Jul 2001 10:35:02 +0000 (10:35 +0000)]
Add: ``If you have to wrap a long statement, put the operator at the
end of the line.''

Reviewed by: alfred, bde

23 years agoDon't put a shell prompt in the example since none of the other man
Dima Dorfman [Sun, 22 Jul 2001 10:33:31 +0000 (10:33 +0000)]
Don't put a shell prompt in the example since none of the other man
pages do it.  Also, don't use brackets to indicate optional options in
the example; instead, add a paragraph below the example explaining it.

23 years agorevert previous commit (bad style and not needed)
Assar Westerlund [Sun, 22 Jul 2001 10:24:31 +0000 (10:24 +0000)]
revert previous commit (bad style and not needed)

Noticed: bde

23 years agoremove -*- nroff -*- thingy
Alfred Perlstein [Sun, 22 Jul 2001 10:18:53 +0000 (10:18 +0000)]
remove -*- nroff -*- thingy

23 years agoadd directory i4brunppp.
Hellmuth Michaelis [Sun, 22 Jul 2001 09:40:58 +0000 (09:40 +0000)]
add directory i4brunppp.
Pointed out by: David H. Wolfskill david@catwhisker.org and
                Crist J. Clark <cristjc@earthlink.net>

23 years agoReplace calls to strncpy with calls to strlcpy, and remove the extra step
Garance A Drosehn [Sun, 22 Jul 2001 07:25:27 +0000 (07:25 +0000)]
Replace calls to strncpy with calls to strlcpy, and remove the extra step
needed to ensure that the result is null-terminated when using strncpy().

MFC after: 8 days

23 years agoError messaging in ipfw(8) was out of hand, almost 50 lines of usage
Crist J. Clark [Sun, 22 Jul 2001 06:40:11 +0000 (06:40 +0000)]
Error messaging in ipfw(8) was out of hand, almost 50 lines of usage
information for any command line error, the actual error message
almost always (and sometimes irretrievably) lost scrolling off the top
of the screen. Now just print the error. Give ipfw(8) no arguments for
the old usage summary.

Thanks to Lyndon Nerenberg <lyndon@orthanc.ab.ca> for the patch and
PR, but I had already done this when ru pointed out the PR.

PR: bin/28729
Approved by: ru
MFC after: 1 week

23 years agoAdd another Dell 3/Di PCI ID. You'd think they had enough of these
Mike Smith [Sun, 22 Jul 2001 04:06:36 +0000 (04:06 +0000)]
Add another Dell 3/Di PCI ID.  You'd think they had enough of these
already...

23 years agoFix a selection of bugs, and improve code layout:
Ian Dowse [Sun, 22 Jul 2001 01:25:25 +0000 (01:25 +0000)]
Fix a selection of bugs, and improve code layout:
- Remove unnecessary and unused local variables.
- Include useful information in error and warning messages.
- Fix the logic for expiring mounttab entries.
- Remove calls to getaddrinfo - the results were not used.
- Simplify some string handling by using snprintf.
- Fix usage.

23 years agoRename the -w option to be -W. The recently-added -w option is different
Garance A Drosehn [Sun, 22 Jul 2001 01:05:20 +0000 (01:05 +0000)]
Rename the -w option to be -W.  The recently-added -w option is different
than the long-standing -w option in NetBSD, so change it before anyone in
FreeBSD gets used to it.  For now, -w is still accepted, but prints out
some warnings via syslog.

MFC after: 1 week

23 years agoCorrect typo.
Yoshihiro Takahashi [Sun, 22 Jul 2001 00:58:15 +0000 (00:58 +0000)]
Correct typo.

23 years agoInclude the remote hostname in RPC-related warning messages. Exit
Ian Dowse [Sun, 22 Jul 2001 00:45:29 +0000 (00:45 +0000)]
Include the remote hostname in RPC-related warning messages. Exit
immediately if a host specified by the -h flag cannot be parsed
instead of attempting to unmount all NFS filesystems, which was
bad.

Add a missing return statement at the end of checkname(); this
could result in a non-zero exit status in some cases even if the
unmount succeeded.

Group two separate NFS-related operations into one block to make
it more obvious that a variable (hostp) is not dereferenced when
uninitialised. Initialise it to NULL anyway to avoid a warning.

Pass in the read_mtab()'s bogus argument as NULL instead of messing
with a local variable to achieve the same effect. A later commit
will clean up this mounttab interface.

23 years agoadd prototype for dosetrlimit
Assar Westerlund [Sun, 22 Jul 2001 00:21:19 +0000 (00:21 +0000)]
add prototype for dosetrlimit

23 years agorevert previous accidental commit
Assar Westerlund [Sun, 22 Jul 2001 00:12:23 +0000 (00:12 +0000)]
revert previous accidental commit

23 years agoget rid of some printf and pointer type warnings
Assar Westerlund [Sun, 22 Jul 2001 00:12:22 +0000 (00:12 +0000)]
get rid of some printf and pointer type warnings

23 years agoChange a few read & write calls to use 'STDOUT_FILENO' instead of '1'.
Garance A Drosehn [Sun, 22 Jul 2001 00:03:21 +0000 (00:03 +0000)]
Change a few read & write calls to use 'STDOUT_FILENO' instead of '1'.

Submitted by: David Hill <david@phobia.ms>
Reviewed by: freebsd-audit  (a little)
MFC after: 1 week

23 years agoThe MA401 is a Prism II
Nick Sayer [Sat, 21 Jul 2001 23:11:13 +0000 (23:11 +0000)]
The MA401 is a Prism II

Approved by: imp

23 years agoACPI CA doesn't recognis S4BIOS as a seperate sleep type anymore, so
Mike Smith [Sat, 21 Jul 2001 21:51:44 +0000 (21:51 +0000)]
ACPI CA doesn't recognis S4BIOS as a seperate sleep type anymore, so
drop support for it here until we work out a better way of handling it.

23 years agoadd <sys/sytm.h> (for cpufunc.h -> rdtsc)
Assar Westerlund [Sat, 21 Jul 2001 20:53:42 +0000 (20:53 +0000)]
add <sys/sytm.h> (for cpufunc.h -> rdtsc)

23 years ago- Allow link-local multicast traffic for client.
Hajimu UMEMOTO [Sat, 21 Jul 2001 19:59:35 +0000 (19:59 +0000)]
- Allow link-local multicast traffic for client.
- Allow ICMPv6 destination unreach, packet too big and NS/NA.
- RIPng also uses link-local to link-local.

MFC after: 1 week

23 years agoadd an unused attribute to yyunput to avoid warnings (errors) when
Assar Westerlund [Sat, 21 Jul 2001 19:47:17 +0000 (19:47 +0000)]
add an unused attribute to yyunput to avoid warnings (errors) when
it's not being used

23 years agoThis was repo copied to src/release/alpha/fixit_crunch.conf.
David E. O'Brien [Sat, 21 Jul 2001 19:01:59 +0000 (19:01 +0000)]
This was repo copied to src/release/alpha/fixit_crunch.conf.
The i386 and PC98 already had their own platform-specific version.

23 years agoI think this is OK for PC98, someone please check.
David E. O'Brien [Sat, 21 Jul 2001 17:55:38 +0000 (17:55 +0000)]
I think this is OK for PC98, someone please check.

As was done to i386/boot_crunch.conf:
+ Build with `lint' defined to cut out some of the `rcsid's (binaries' code
  only, the libs will still have the `rcsid's in them).
+ mount_mfs is OBE.
+ Do not need libipx as I have conditionally diked that functional out of
  ifconfig(8).

23 years agoFix the install kernel on a 1.44MB floppy again.
David E. O'Brien [Sat, 21 Jul 2001 17:38:46 +0000 (17:38 +0000)]
Fix the install kernel on a 1.44MB floppy again.
I had to be aggressively Draconian to succeed.
I diked out:
+ Multia, NoName, PC/EB 64, Aspen Alpine support.
+ SCSI tape support
+ AMI MegaRAID controller support
+ All parallel bus support (includes PLIP)
+ vx (3c590, 3c595), pcn (AMD Am79C97x PCI 10/100), sf (Adaptec AIC-6915),
  sis (SiS 900/SiS 7016), ste (Sundance ST201 (D-Link DFE-550TX)),
  wb (Winbond W89C840F) support.

If the removal of any of this support causes heartburn, please let me know.

23 years agoUnconditionalize these now that they live in ARCH-specific directories.
David E. O'Brien [Sat, 21 Jul 2001 17:32:40 +0000 (17:32 +0000)]
Unconditionalize these now that they live in ARCH-specific directories.

23 years ago+ Build with `lint' defined to cut out some of the `rcsid's (binaries' code
David E. O'Brien [Sat, 21 Jul 2001 17:28:21 +0000 (17:28 +0000)]
+ Build with `lint' defined to cut out some of the `rcsid's (binaries' code
  only, the libs will still have the `rcsid's in them).
+ mount_mfs is OBE.
+ Do not need libipx as I have conditionally diked that functional out of
  ifconfig(8).

23 years ago+ Build with `lint' defined to cut out some of the `rcsid's (binaries' code
David E. O'Brien [Sat, 21 Jul 2001 17:26:55 +0000 (17:26 +0000)]
+ Build with `lint' defined to cut out some of the `rcsid's (binaries' code
  only, the libs will still have the `rcsid's in them).
+ mount_mfs is OBE.
+ The Alpha install does not support SLIP, PCCARD or USB installs; so we do
  not need the associated userland utils.
+ Do not need libipx as I have conditionally diked that functional out of
  ifconfig(8).

23 years agodokern.sh was repo copied from src/release/scripts to
David E. O'Brien [Sat, 21 Jul 2001 17:21:52 +0000 (17:21 +0000)]
dokern.sh was repo copied from src/release/scripts to
src/release/${MACHINE_ARCH}

23 years agoThis has been repo copied to src/release/${MACHINE_ARCH}/dokern.sh
David E. O'Brien [Sat, 21 Jul 2001 17:21:02 +0000 (17:21 +0000)]
This has been repo copied to src/release/${MACHINE_ARCH}/dokern.sh

23 years agoThis has been repo copied to src/release/${MACHINE_ARCH}/boot_crunch.conf
David E. O'Brien [Sat, 21 Jul 2001 17:19:59 +0000 (17:19 +0000)]
This has been repo copied to src/release/${MACHINE_ARCH}/boot_crunch.conf

23 years agoadd <sys/cdefs.h> (for __unused and such)
Assar Westerlund [Sat, 21 Jul 2001 17:12:44 +0000 (17:12 +0000)]
add <sys/cdefs.h> (for __unused and such)

23 years agonull_do_probe: mark as unused
Assar Westerlund [Sat, 21 Jul 2001 17:10:30 +0000 (17:10 +0000)]
null_do_probe: mark as unused

23 years ago(vm_page_select_cache): add prototype
Assar Westerlund [Sat, 21 Jul 2001 17:08:15 +0000 (17:08 +0000)]
(vm_page_select_cache): add prototype

23 years agoAnother hangup bugfix from Juha:
Hellmuth Michaelis [Sat, 21 Jul 2001 16:31:39 +0000 (16:31 +0000)]
Another hangup bugfix from Juha:
 Problem 1 is that the config entry hangup flag is zeroed only at
 CONNECT_ACTIVE_IND in msghdl.c. If any (other) call is disconnected
 after EV_MDO and before CONNECT_ACTIVE_IND, the cleanup routine will
 disconnect the in-progress dialout as well, if its hangup flag is
 nonzero (which it is likely to be) after the previous incarnation of the
 cfg entry. Patch-1 fixes this by clearing the hangup flag as soon as a
 cfg entry is reserved for the call.

Submitted by: Juha-Matti Liukkonen <jml@cubical.fi>

23 years agoFix hangup bug; Juha writes:
Hellmuth Michaelis [Sat, 21 Jul 2001 16:28:28 +0000 (16:28 +0000)]
Fix hangup bug; Juha writes:
 Problem 2 is that doing a local hangup (eg. by writing "H" to the
 dialout device) to a call which is already disconnected results in isdnd
 moving the cfg entry to an illegal state, from which there is no
 recovery. This is tricky because there is no way to synchronize local
 hangup with the remote end (ie. the callee can always hang up at an
 inconvenient time)! Hence, patch-2 alters fsm.c's EV_DRQ state table
 such that the local hangup request is processed or ignored in most
 states, even for disconnected calls.

Submitted by: Juha-Matti Liukkonen <jml@cubical.fi>

23 years agoChange the foreground mount behaviour so that we keep retrying
Ian Dowse [Sat, 21 Jul 2001 15:38:37 +0000 (15:38 +0000)]
Change the foreground mount behaviour so that we keep retrying
forever by default. This matches what mount_nfs did before revision
1.40, and it is the generally expected behaviour for NFS mounts.

Document the current defaults near the start of the man page and
mention the options that can be used to change them.

Discussed on: -hackers

23 years agoUnset MAINTAINER.
Robert Nordier [Sat, 21 Jul 2001 14:32:47 +0000 (14:32 +0000)]
Unset MAINTAINER.

23 years agoAdd a patch from the isdn mailinglist, Juha writes:
Hellmuth Michaelis [Sat, 21 Jul 2001 11:52:31 +0000 (11:52 +0000)]
Add a patch from the isdn mailinglist, Juha writes:

 ppp in 4.x apparently does a close(2) after opening the tun device;
 i4brunppp starts up with only file descriptors 0 and 1 open (to the rbch
 device) -> tun gets opened as 2 -> tun gets closed -> later use results
 in EBADF. A quick fix to i4brunppp.c makes the thing work (I know, this
 is ugly, but I needed it up quick...):

Submitted by: Juha-Matti Liukkonen <jml@cubical.fi>

23 years agoAdd the i4brunppp utility to the example section. This utility is one way
Hellmuth Michaelis [Sat, 21 Jul 2001 11:49:28 +0000 (11:49 +0000)]
Add the i4brunppp utility to the example section. This utility is one way
of taking incoming calls with i4b in conjunction with userland ppp(8).

23 years agoupdate step.
Hellmuth Michaelis [Sat, 21 Jul 2001 11:31:27 +0000 (11:31 +0000)]
update step.

23 years agoupdate ReleaseNotes for isdnd keyword maxconnecttime
Hellmuth Michaelis [Sat, 21 Jul 2001 11:30:38 +0000 (11:30 +0000)]
update ReleaseNotes for isdnd keyword maxconnecttime

23 years agoAfter being port- and addressscanned continously from several sources where
Hellmuth Michaelis [Sat, 21 Jul 2001 11:29:40 +0000 (11:29 +0000)]
After being port- and addressscanned continously from several sources where
my line stayed open for more than eight hours i decided to implement a
maximum connection time; see keyword maxconnecttime in the manual page.

23 years agoConvert from acpi_strerror() to AcpiFormatException()
Mike Smith [Sat, 21 Jul 2001 10:24:37 +0000 (10:24 +0000)]
Convert from acpi_strerror() to AcpiFormatException()

Fix dangling include of the dear departed acpi_ecreg.h

23 years agoAdd missing closing parenthesis.
Ben Smithurst [Sat, 21 Jul 2001 05:33:33 +0000 (05:33 +0000)]
Add missing closing parenthesis.

23 years agoShow an example of how to sort a file listing by size. Hopefully this
David E. O'Brien [Sat, 21 Jul 2001 05:01:50 +0000 (05:01 +0000)]
Show an example of how to sort a file listing by size.  Hopefully this
will reduce the number of people calling for this functionality to be
added to ls(1).

23 years agoUpdate the OSD module to match the ACPI CA 20010717 import.
Mike Smith [Sat, 21 Jul 2001 04:10:01 +0000 (04:10 +0000)]
Update the OSD module to match the ACPI CA 20010717 import.

Submitted by: "Grover, Andrew" <andrew.grover@intel.com> (OsdHardware.c)

23 years agoImplement a "proper" timecounter hung off the ACPI timer device.
Mike Smith [Sat, 21 Jul 2001 04:08:42 +0000 (04:08 +0000)]
Implement a "proper" timecounter hung off the ACPI timer device.

This code is based on the mp_clock code by phk.  It attempts to
detect the PIIX4 (see comments for details) and use a workaround
for its problems.

This code is experimental, and could use some testing and review by a
timekeeping enthusiast.

23 years agoMerge the EC register definitions into the EC module proper, they're not
Mike Smith [Sat, 21 Jul 2001 04:06:44 +0000 (04:06 +0000)]
Merge the EC register definitions into the EC module proper, they're not
used anywhere else.

AcpiOsSleepUsec() -> AcpiOsStall()

23 years agoUse our saved copy of the FADT rather than fetching it again.
Mike Smith [Sat, 21 Jul 2001 04:05:32 +0000 (04:05 +0000)]
Use our saved copy of the FADT rather than fetching it again.

23 years agoDon't call ourselves a "control method" anything, that's not useful.
Mike Smith [Sat, 21 Jul 2001 04:05:08 +0000 (04:05 +0000)]
Don't call ourselves a "control method" anything, that's not useful.

Move the "button pressed" diagnostics to the point where we can print
out which button was actually pressed.

23 years agoThe API for loading tables changed (we no longer explicitly search for the
Mike Smith [Sat, 21 Jul 2001 04:04:03 +0000 (04:04 +0000)]
The API for loading tables changed (we no longer explicitly search for the
RSDP, it's now found via a callback).

AcpiOsSleepUsec() went away, use AcpiOsSleep() instead (we could use
AcpiOsStall() too)

AcpiFormatException() was changed to make more sense (it behaves like
our old acpi_strerror() did), so throw acpi_strerror() away (still
#defined in acpivar.h though, we need to sweep these seperately).

23 years agoMerge local changes with the ACPI CA 20010717 import.
Mike Smith [Sat, 21 Jul 2001 03:57:37 +0000 (03:57 +0000)]
Merge local changes with the ACPI CA 20010717 import.

23 years agoThis commit was generated by cvs2svn to compensate for changes in r80062,
Mike Smith [Sat, 21 Jul 2001 03:55:17 +0000 (03:55 +0000)]
This commit was generated by cvs2svn to compensate for changes in r80062,
which included commits to RCS files with non-trunk default branches.

23 years agoVendor import of the Intel ACPI CA 20010717 snapshot.
Mike Smith [Sat, 21 Jul 2001 03:55:17 +0000 (03:55 +0000)]
Vendor import of the Intel ACPI CA 20010717 snapshot.

23 years agoThis commit was generated by cvs2svn to compensate for changes in r80060,
Mike Smith [Sat, 21 Jul 2001 03:55:15 +0000 (03:55 +0000)]
This commit was generated by cvs2svn to compensate for changes in r80060,
which included commits to RCS files with non-trunk default branches.

23 years agoVendor import of the Intel ACPI CA 20010717 snapshot.
Mike Smith [Sat, 21 Jul 2001 03:55:15 +0000 (03:55 +0000)]
Vendor import of the Intel ACPI CA 20010717 snapshot.

23 years agoHide in the Attic. This isn't built any more (it wont) and it can better
David E. O'Brien [Sat, 21 Jul 2001 02:18:38 +0000 (02:18 +0000)]
Hide in the Attic.  This isn't built any more (it wont) and it can better
collect dust up there.

23 years agoDike out the IPX bits if RELEASE_CRUNCH is defined.
David E. O'Brien [Sat, 21 Jul 2001 00:35:11 +0000 (00:35 +0000)]
Dike out the IPX bits if RELEASE_CRUNCH is defined.

23 years agoHint getaddrinfo() correctly if we're looking up a name that we got from
Brian Somers [Sat, 21 Jul 2001 00:18:54 +0000 (00:18 +0000)]
Hint getaddrinfo() correctly if we're looking up a name that we got from
an AF_INET6 address.

MFC after: 1 week

23 years agoAdd a missing ~ so that the LO_INITIALIZED flag actually gets turned off
John Baldwin [Fri, 20 Jul 2001 23:29:25 +0000 (23:29 +0000)]
Add a missing ~ so that the LO_INITIALIZED flag actually gets turned off
in witness_destroy().

23 years agoPut a knob in the kernel config files to tweak the user max stack size.
Jim Pirzyk [Fri, 20 Jul 2001 20:26:34 +0000 (20:26 +0000)]
Put a knob in the kernel config files to tweak the user max stack size.

PR: kern/28925
Reviewed by: bakul@bitblocks.com and tlambert2@mindspring.com on -arch.
MFC after: 1 week

23 years agogif isn't a count device anymore so don't put a number after it.
Brooks Davis [Fri, 20 Jul 2001 19:53:08 +0000 (19:53 +0000)]
gif isn't a count device anymore so don't put a number after it.

Pointed out by: brian

23 years agoDo more strict checking for an interface.
Hajimu UMEMOTO [Fri, 20 Jul 2001 18:08:17 +0000 (18:08 +0000)]
Do more strict checking for an interface.

Suggested by: matusita@jp.FreeBSD.org
MFC after: 1 week

23 years agoo Add phantom's gethostinfo() function to reduce duplicated code.
Mike Barcroft [Fri, 20 Jul 2001 16:44:44 +0000 (16:44 +0000)]
o Add phantom's gethostinfo() function to reduce duplicated code.
o Increase readability by creating a wrapper for asprintf(3).
o Cast isdigit(3) argument to unsigned char for future locale
  changes.

Submitted by: phantom
Reviewed by: -audit, des
Approved by: des

23 years agoFix some nits I missed in my style cleanup.
David E. O'Brien [Fri, 20 Jul 2001 15:33:04 +0000 (15:33 +0000)]
Fix some nits I missed in my style cleanup.

23 years agoMFCrypto: fixed the remote buffer overflow.
Ruslan Ermilov [Fri, 20 Jul 2001 15:14:03 +0000 (15:14 +0000)]
MFCrypto: fixed the remote buffer overflow.

23 years agoRemoved CSRG ids. CSRG never had anything to do with dpt.
Bruce Evans [Fri, 20 Jul 2001 14:59:15 +0000 (14:59 +0000)]
Removed CSRG ids.  CSRG never had anything to do with dpt.

Don't set BINMODE to 500.  This is not a setuid program.

Note: the dpt utilities have never been attached to the world and
haven't been compilable for a year or two.

23 years agoFix typo; the meaning of the 3rd parameter to genkbd_keyaction() is
Kazutaka YOKOTA [Fri, 20 Jul 2001 14:49:54 +0000 (14:49 +0000)]
Fix typo; the meaning of the 3rd parameter to genkbd_keyaction() is
'up' rather than 'down'.

PR: 22466
MFC after: 10 days

23 years agoFix off by one error for ESC[1J.
Kazutaka YOKOTA [Fri, 20 Jul 2001 14:08:12 +0000 (14:08 +0000)]
Fix off by one error for ESC[1J.

PR: kern/18381
MFC after: 10 days

23 years agoForgot to mention that previous commit (1.4 --> 1.5) would be:
Maxim Sobolev [Fri, 20 Jul 2001 13:25:14 +0000 (13:25 +0000)]
Forgot to mention that previous commit (1.4 --> 1.5) would be:

MFC after: 2 weeks

23 years agoUse reallocf(3) to increase size of malloc'ed buffer instead of
Maxim Sobolev [Fri, 20 Jul 2001 13:20:32 +0000 (13:20 +0000)]
Use reallocf(3) to increase size of malloc'ed buffer instead of
malloc()+memcpy()+free().

23 years agoDefine a couple of subroutines to replace duplicated common code. You
Kazutaka YOKOTA [Fri, 20 Jul 2001 13:10:59 +0000 (13:10 +0000)]
Define a couple of subroutines to replace duplicated common code.  You
shoulnd't see any functional difference.

23 years agoReturn consistent key action codes at key press and release
Kazutaka YOKOTA [Fri, 20 Jul 2001 13:05:57 +0000 (13:05 +0000)]
Return consistent key action codes at key press and release
events. Otherwise you would see unexpected results if shift or
locking keys are defined to give different actions depending
on other shift/locking keys' state.

Please keep the ukbd module and the kernel in sync, otherwise
the USB keyboard won't work after this change.
MFC after: 10 days