18 years agoRaise WARNS to 6:
Chris Pressey [Wed, 2 Feb 2005 06:39:09 +0000 (06:39 +0000)]
Raise WARNS to 6:

- Constify one function argument, mark others as unused.

Submitted-by: Larry Lansing <lansil_at_fuzzynerd_dot_com>
18 years agoRaise WARNS to 6:
Chris Pressey [Wed, 2 Feb 2005 06:22:30 +0000 (06:22 +0000)]
Raise WARNS to 6:

- Add function prototypes.

Submitted-by: Larry Lansing <lansil_at_fuzzynerd_dot_com>
18 years agoMinor adustments to avoid a signed/unsigned comparison compiler warning
Matthew Dillon [Wed, 2 Feb 2005 06:14:51 +0000 (06:14 +0000)]
Minor adustments to avoid a signed/unsigned comparison compiler warning
plus a small cleanup.  No operational changes.

18 years agoThe last commit broke ttyname(), which broke ssh -1. Fix that, plus
Matthew Dillon [Wed, 2 Feb 2005 05:42:01 +0000 (05:42 +0000)]
The last commit broke ttyname(), which broke ssh -1.  Fix that, plus
cleanup a static variable overloading of the name 'buf' to avoid confusion.

Reported-by: Steve Mynott, Peter Avalos
18 years agoFix type-o in last commit.
Matthew Dillon [Wed, 2 Feb 2005 04:57:15 +0000 (04:57 +0000)]
Fix type-o in last commit.

18 years agoChange the ALTQ macro stubs to actually work. For the current ALTQ less
Joerg Sonnenberger [Tue, 1 Feb 2005 23:14:54 +0000 (23:14 +0000)]
Change the ALTQ macro stubs to actually work. For the current ALTQ less
kernel, drv queue and interface queue are identical.

18 years agoAdd a intrmask_t pointer to register_int() and register_swi(), and make
Matthew Dillon [Tue, 1 Feb 2005 22:41:31 +0000 (22:41 +0000)]
Add a intrmask_t pointer to register_int() and register_swi(), and make
the interrupt thread loop set the spl for the duration of the call to the

Allow interrupt threads to be run with or without a critical section based
on the kern.int_use_crit_section sysctl.  The default is to conservatively
run with a critical section for now.

Turning this off will cause cpu usage in interrupts to be properly accounted
for by top, systat, and ps.

18 years agoFix an invariant test that tries to catch locked tokens being left on the
Matthew Dillon [Tue, 1 Feb 2005 22:36:26 +0000 (22:36 +0000)]
Fix an invariant test that tries to catch locked tokens being left on the
stack after a stack pop.

Submitted-by: Aggelos Economopoulos <aoiko@cc.ece.ntua.gr>
18 years agoGCC 3.4 doesn't include a #pragma weak reference, if the symbol
Joerg Sonnenberger [Tue, 1 Feb 2005 22:35:19 +0000 (22:35 +0000)]
GCC 3.4 doesn't include a #pragma weak reference, if the symbol
doesn't exist anyway. This broke the build obviously.
Change the places in libc which used the pragma to use __weak_reference.

18 years agoCount time spent in interrupts in the overall runtime calculation so ps
Matthew Dillon [Tue, 1 Feb 2005 22:33:43 +0000 (22:33 +0000)]
Count time spent in interrupts in the overall runtime calculation so ps
properly reports the cpu usage of interrupt threads.

18 years ago- Fix some duplicate commenting for struct Job, struct GNode and
Max Okumoto [Tue, 1 Feb 2005 22:05:36 +0000 (22:05 +0000)]
- Fix some duplicate commenting for struct Job, struct GNode and
  struct Shell by removing one version and inlining the comments
  into the structure declaration (the comments have been somewhat

- Remove the 3rd clause (from 4) of the BSD license because those
  files have only the UCB copyright

Taken-from: FreeBSD

18 years agoDon't double assign -W -Wall and some more warning flags.
Joerg Sonnenberger [Tue, 1 Feb 2005 21:55:22 +0000 (21:55 +0000)]
Don't double assign -W -Wall and some more warning flags.
WARNS=1 is -Wunknown-pragmas and for gcc 3.4 -Wsystem-headers.
WARNS=2 adds -Wall
WARNS=3 equals to what was set before for WARNS>0.
For WARNS >= 5 don't specify -Wno-uninitialized anymore.

18 years agoFix bug in last commit that broke 'df'. 'sfsp' is now a structural pointer
Matthew Dillon [Tue, 1 Feb 2005 21:52:11 +0000 (21:52 +0000)]
Fix bug in last commit that broke 'df'.  'sfsp' is now a structural pointer
so we increment it by one, not by sizeof(*sp).

18 years ago- Improve how we handle the APM device. Much cleaner code.
Liam J. Foy [Tue, 1 Feb 2005 19:51:28 +0000 (19:51 +0000)]
- Improve how we handle the APM device. Much cleaner code.

18 years agoMove to ND_IFINFO().
Hiroki Sato [Tue, 1 Feb 2005 19:39:07 +0000 (19:39 +0000)]
Move to ND_IFINFO().

18 years ago- Large cleanup/changes to man page.
Liam J. Foy [Tue, 1 Feb 2005 18:43:15 +0000 (18:43 +0000)]
- Large cleanup/changes to man page.

Submitted by: Maxime Labelle <mlabelle@vt220.com>

18 years ago- Do not depend on stat_warn_cont when executing commands.
Liam J. Foy [Tue, 1 Feb 2005 18:32:01 +0000 (18:32 +0000)]
- Do not depend on stat_warn_cont when executing commands.

18 years ago- Tie battd into the build
Liam J. Foy [Tue, 1 Feb 2005 18:20:32 +0000 (18:20 +0000)]
- Tie battd into the build

18 years agostyle(9)
Eirik Nygaard [Tue, 1 Feb 2005 17:53:01 +0000 (17:53 +0000)]

18 years ago- Initial import of the battd utility.
Liam J. Foy [Tue, 1 Feb 2005 17:16:36 +0000 (17:16 +0000)]
- Initial import of the battd utility.

The battd daemon monitors the status of the battery or batteries present
in the system.

18 years ago*_load_pallete's second argument should be const.
Joerg Sonnenberger [Tue, 1 Feb 2005 16:30:02 +0000 (16:30 +0000)]
*_load_pallete's second argument should be const.

18 years ago- Nuke #ifdef SCOPEDROUTING. It was never enabled and is useless now[1].
Hiroki Sato [Tue, 1 Feb 2005 16:09:37 +0000 (16:09 +0000)]
- Nuke #ifdef SCOPEDROUTING.  It was never enabled and is useless now[1].
- Add in6_domifattach(), in6_domifdetach(), and ifp->if_afdata[2].
- ANSIfy.

Obtained from: KAME via FreeBSD
References (FreeBSD):

if.c 1.168[2]
if_var.h 1.59[2]

icmp6.h 1.12[2]
tcp_input.c 1.209[2]

icmp6.c 1.38[1], 1.39[2]
in6.c 1.30[1], 1.31[2]
ip6_forward.c 1.20[1]
in6_gif.c 1.15[1]
in6_proto.c 1.25[2]
in6_ifattach.c 1.17[1], 1.18[2]
ip6_input.c 1.53[1], 1.57[2]
ip6_output.c 1.58[1], 1.59[2]
in6_src.c 1.14[2]
in6_var.h 1.14[2]
nd6.c 1.30[1], 1.32[2]
nd6.h 1.13[2]
nd6_nbr.c 1.19[2]
nd6_rtr.c 1.20[2]
scope6.c 1.8[2]
scope6_var.h 1.3[2]

domain.h 1.18[2]
kernel.h 1.113[2]

18 years agoAlways honor fdp->fd_nrdir as root. Once the loop reached NCF_ROOT,
Joerg Sonnenberger [Tue, 1 Feb 2005 14:43:44 +0000 (14:43 +0000)]
Always honor fdp->fd_nrdir as root. Once the loop reached NCF_ROOT,
fdp->fd_nrdir has to be equal to the current ncp or the original
ncp is not below fdp->fd_nrdir. In the latter case, just return "/"
as fullpath.

18 years agoRemove SCARG junk.
Joerg Sonnenberger [Tue, 1 Feb 2005 13:55:49 +0000 (13:55 +0000)]
Remove SCARG junk.

Suggested-by: drhodus
18 years ago- Add comments to clarify code
Max Okumoto [Tue, 1 Feb 2005 13:05:48 +0000 (13:05 +0000)]
- Add comments to clarify code
- Rewrite compliated for loop as a while loop.

Tested by comparing object files.

18 years agoImport initial version of 1:1 pthread library.
David Xu [Tue, 1 Feb 2005 12:38:27 +0000 (12:38 +0000)]
Import initial version of 1:1 pthread library.

18 years agoAdded code to handle DragonFly tags in expected results files.
Max Okumoto [Tue, 1 Feb 2005 11:19:38 +0000 (11:19 +0000)]
Added code to handle DragonFly tags in expected results files.

18 years agoAdd regression test infrastructure.
Max Okumoto [Tue, 1 Feb 2005 07:07:15 +0000 (07:07 +0000)]
Add regression test infrastructure.

18 years agoencap_getarg() was not properly loading the pointer argument associated
Matthew Dillon [Tue, 1 Feb 2005 05:34:33 +0000 (05:34 +0000)]
encap_getarg() was not properly loading the pointer argument associated
with the m_tag, returning garbage which would then panic the box.  The
code path is primarily used by the GIF interface.

Crash-Reported-by: Peter Avalos <pavalos@theshell.com>
Reviewed-by: Jeffrey Hsu
18 years agoHide pure kernel threads from jailed processes.
Joerg Sonnenberger [Tue, 1 Feb 2005 02:25:45 +0000 (02:25 +0000)]
Hide pure kernel threads from jailed processes.

18 years agoMake const correct.
Joerg Sonnenberger [Tue, 1 Feb 2005 02:19:43 +0000 (02:19 +0000)]
Make const correct.

18 years agoAdd prototype for ciss_print0 in place.
Joerg Sonnenberger [Tue, 1 Feb 2005 02:16:58 +0000 (02:16 +0000)]
Add prototype for ciss_print0 in place.

18 years agoRemove compat code for anything, but DragonFly.
Joerg Sonnenberger [Tue, 1 Feb 2005 00:51:50 +0000 (00:51 +0000)]
Remove compat code for anything, but DragonFly.

18 years agoRemove unused variable.
Joerg Sonnenberger [Tue, 1 Feb 2005 00:05:28 +0000 (00:05 +0000)]
Remove unused variable.

18 years agoRemove GPLed fpemulation, old rp, old awe and pcic.
Joerg Sonnenberger [Mon, 31 Jan 2005 23:44:35 +0000 (23:44 +0000)]
Remove GPLed fpemulation, old rp, old awe and pcic.
dgb is still in, until the persons having the hardware decide that
digi(4) works for them.

Correct spelling of deprecation.

18 years agoMy first commit.
David Xu [Mon, 31 Jan 2005 23:44:17 +0000 (23:44 +0000)]
My first commit.

18 years agoAdd jail_attach support.
Joerg Sonnenberger [Mon, 31 Jan 2005 22:29:59 +0000 (22:29 +0000)]
Add jail_attach support.

Inspired-by: FreeBSD
Manpage-by: Paul Herman
Add sysctl jail.list and jailed() helper functions.

Obtained-from: FreeBSD

Add jexec and jls.

Obtained-from: FreeBSD

Move reference counting into prison_hold / prison_free functions.
Fix an incorrect increment in fork1, the prison structure is not
directly copied and the cr reference counting applies to prison here
too. Begin moving explicit pointer checks for cr_prison != NULL
to the new jailed() function. Move the sys/varsym.h include in
sys/jail.h to the kernel part. Prepare for jail-local securelevel.

18 years agoRemove _THREAD_SAFE depenendancies. Create weakly associated stubs for
Matthew Dillon [Mon, 31 Jan 2005 22:29:48 +0000 (22:29 +0000)]
Remove _THREAD_SAFE depenendancies.  Create weakly associated stubs for
posix threading calls used by libc and weak associations for procedures that
a threading library would otherwise have to override.  This allows a program
to be linked with libc, or linked with libc + a threading library.  libc_r
will be deprecated over time.

Traditionally libc_r had to be linked against a threading library rather then
libc.  Linux, FreeBSD, and other UNIXes have moved either t oa thread-safe
libc or a thread-capable libc.  The threading capability is accomplished by
making the appropriate thread locking calls and providing weakly referenced
dummy stub functions for the case where a threading library is not linked in.
Certain third party software, namely XFree/XOrg, also assume threading
stubs in libc.

This is precursor work to a port of David Xu's 1:1 threading library and is
also intended as a basis to allow DragonFly to support multiple threading

Ported-by: David Rhodus <sdrhodus@gmail.com>
Primarily-ported-from: FreeBSD
Testing and bug fixes by David Xu and Matt Dillon

18 years agoInclude jail.h after proc.h to get struct ucred.
Joerg Sonnenberger [Mon, 31 Jan 2005 22:03:48 +0000 (22:03 +0000)]
Include jail.h after proc.h to get struct ucred.
Fix the printfs for cp_time, it's uint64_t now.

18 years agoAdd missing "return(error)".
Joerg Sonnenberger [Mon, 31 Jan 2005 21:55:18 +0000 (21:55 +0000)]
Add missing "return(error)".

18 years agoDefine in_inithead in net/route.h, it's purpose is the initialisation
Joerg Sonnenberger [Mon, 31 Jan 2005 21:44:10 +0000 (21:44 +0000)]
Define in_inithead in net/route.h, it's purpose is the initialisation
of the AF_INET routing table. This fixes a warning about a missing
prototype in netinet/in_rmx.c and removes the local extern in in_proto.c

18 years agoRemove unused local variable.
Joerg Sonnenberger [Mon, 31 Jan 2005 21:41:00 +0000 (21:41 +0000)]
Remove unused local variable.

18 years agoBuglet in last commit, the first argument of bpf_ptap is the actual bpf_if.
Joerg Sonnenberger [Mon, 31 Jan 2005 21:40:04 +0000 (21:40 +0000)]
Buglet in last commit, the first argument of bpf_ptap is the actual bpf_if.

18 years agoUse PRIx32 to fix compiler warning correctly.
Joerg Sonnenberger [Mon, 31 Jan 2005 21:39:32 +0000 (21:39 +0000)]
Use PRIx32 to fix compiler warning correctly.

18 years agoFix warning by adding explicit () around addition.
Joerg Sonnenberger [Mon, 31 Jan 2005 21:37:52 +0000 (21:37 +0000)]
Fix warning by adding explicit () around addition.
cmp proves that this is identical to before.

18 years agosys/types.h ==> sys/param.h
Joerg Sonnenberger [Mon, 31 Jan 2005 21:27:56 +0000 (21:27 +0000)]
sys/types.h ==> sys/param.h

18 years agosys/types.h ==> sys/param.h
Joerg Sonnenberger [Mon, 31 Jan 2005 21:20:58 +0000 (21:20 +0000)]
sys/types.h ==> sys/param.h

18 years ago- Move deallocation of Buffer object to end of routine.
Max Okumoto [Mon, 31 Jan 2005 21:13:07 +0000 (21:13 +0000)]
- Move deallocation of Buffer object to end of routine.

18 years agoFixes in main.c
Max Okumoto [Mon, 31 Jan 2005 21:11:26 +0000 (21:11 +0000)]
Fixes in main.c
- Cleanup VPATH variable expansion and parsing.
- Cleanup -V flag handling.
- Added comments.
- Fixed negative logic in if-statements.

18 years agoThe last commit couldn't handle cases when $GZIP contained
YONETANI Tomokazu [Mon, 31 Jan 2005 19:28:57 +0000 (19:28 +0000)]
The last commit couldn't handle cases when $GZIP contained
more than one argument and still crashed.
NetBSD CVS keyword is left unchanged as we don't have changes
1.68 and 1.69 yet.

Obtained-from: NetBSD gzip.c revision 1.70

18 years agoRegen.
Joerg Sonnenberger [Mon, 31 Jan 2005 18:18:21 +0000 (18:18 +0000)]

18 years agoCleanup last commit. Remove local ncp, it shadows the parameter, add
Joerg Sonnenberger [Mon, 31 Jan 2005 18:11:06 +0000 (18:11 +0000)]
Cleanup last commit. Remove local ncp, it shadows the parameter, add
return(0) and remove unused locals in vn_fullpath.

18 years agoTake advantage of our new namecache topology to generate the actual paths
Matthew Dillon [Mon, 31 Jan 2005 18:05:09 +0000 (18:05 +0000)]
Take advantage of our new namecache topology to generate the actual paths
associated with open file descriptors, root, and working directories,
and the exec binary's file path.

Adjust the width and alignment of some of the columns.

18 years agoDon't copy blindly MAXPATHLEN byte from fullpath. vn_fullpath
Joerg Sonnenberger [Mon, 31 Jan 2005 17:53:57 +0000 (17:53 +0000)]
Don't copy blindly MAXPATHLEN byte from fullpath. vn_fullpath
allocates MAXPATHLEN, but the buffer is utilized from the end.

18 years agoRegen.
Joerg Sonnenberger [Mon, 31 Jan 2005 17:38:30 +0000 (17:38 +0000)]

18 years agoAdd jail_attach syscall.
Joerg Sonnenberger [Mon, 31 Jan 2005 17:35:35 +0000 (17:35 +0000)]
Add jail_attach syscall.

18 years agoUncomment the entry for kern_chrot in kern_syscall.h and change the
Joerg Sonnenberger [Mon, 31 Jan 2005 17:20:48 +0000 (17:20 +0000)]
Uncomment the entry for kern_chrot in kern_syscall.h and change the
implementation to take the namecache entry directly.

18 years agoSplit vn_fullpath into cache_fullpath and vn_fullpath. The former just
Joerg Sonnenberger [Mon, 31 Jan 2005 17:17:58 +0000 (17:17 +0000)]
Split vn_fullpath into cache_fullpath and vn_fullpath. The former just
takes a namespace entry directly.

18 years agosys/types.h ==> sys/param.h
Joerg Sonnenberger [Mon, 31 Jan 2005 16:51:09 +0000 (16:51 +0000)]
sys/types.h ==> sys/param.h

18 years agoIn preparation for the jail commit, include sys/param.h instead of sys/types.h.
Joerg Sonnenberger [Mon, 31 Jan 2005 16:50:33 +0000 (16:50 +0000)]
In preparation for the jail commit, include sys/param.h instead of sys/types.h.

18 years agoRemove unused compat code like FXP_LOCK, a fake struct mtx.
Joerg Sonnenberger [Mon, 31 Jan 2005 15:39:12 +0000 (15:39 +0000)]
Remove unused compat code like FXP_LOCK, a fake struct mtx.
Remove FXP_UNIT. Don't define sc_if, use the arpcom entry
directly. Use PCIR_BAR instead of PCIR_MAPS.

18 years ago- Add missing disclamer.
Max Okumoto [Mon, 31 Jan 2005 09:49:14 +0000 (09:49 +0000)]
- Add missing disclamer.

18 years ago- Convert negative logic, so code is easyer to read.
Max Okumoto [Mon, 31 Jan 2005 08:30:51 +0000 (08:30 +0000)]
- Convert negative logic, so code is easyer to read.

- Start using Buffer objects, instead of c strings.  That way we
  can pass back one object, instead of passing back seperate length
  and character arrays.

18 years agoAdd a sysctl to control 8254 bios overwrite warnings. Default is on.
Matthew Dillon [Mon, 31 Jan 2005 04:35:17 +0000 (04:35 +0000)]
Add a sysctl to control 8254 bios overwrite warnings.  Default is on.

Submitted-by: "E.Girkantas" <eg@zemaitija.net>
18 years ago- Hide the following characters '(', ')', '{', '}' inside of #defines
Max Okumoto [Sun, 30 Jan 2005 07:20:59 +0000 (07:20 +0000)]
- Hide the following characters '(', ')', '{', '}' inside of #defines
  to make it easyer to read. OPEN_PARN, CLOSE_PAREN, OPEN_BRACKET,

- Re-write for loop into a while loop to clearify operation.

Tested by comparing object files.

18 years agoRedo argv processing to better conform to standards. A NULL argv is no
Matthew Dillon [Sat, 29 Jan 2005 20:54:20 +0000 (20:54 +0000)]
Redo argv processing to better conform to standards.  A NULL argv is no
longer allowed.  If argv[0] is NULL, we still pass a filename for argv[0]
to the underlying program but we no longer attempt to process any further
arguments.  Also rewrite the loop and get rid of the goto to make the code
more readable.

Suggested-by: Maxim Sobolev <sobomax@portaone.com>
18 years agogetblk() has an old crufty API in which the logical block size is not a
Matthew Dillon [Sat, 29 Jan 2005 19:17:06 +0000 (19:17 +0000)]
getblk() has an old crufty API in which the logical block size is not a
well known quantity.  Device drivers standardize on using DEV_BSIZE (512),
while file ops are supposed to use mount->mnt_stat.f_iosize.

The existing code was testing for a non-NULL vnode->v_mountedhere field
but this field is part of a union and only valid for VDIR types.  It was being
improperly tested on non-VDIR vnode types.  In particular, if vn_isdisk()
fails due to the disk device being ripped out from under a filesystem,
the code would fall through and try to use v_mountedhere, leading to a
crash.  It also makes no sense to use the target mount to calculate the
block size for the underlying mount point's vnode, so this test has been
removed entirely.  The vn_isdisk() test has been replaced with an explicit

Finally, note that filesystems like UFS use varying buffer cache buffer
sizes for different areas of the same block device (e.g. bitmap areas,
inode area, file data areas, superblock), which is why DEV_BSIZE is being
used here.  What really needs to happen is for b_blkno to be entirely
removed in favor of a 64 bit offset.

Crash-Reported-by: Vyacheslav Bocharov <list@smz.com.ua>
18 years agoDon't build sysctl as bootstrap tool. This helps us avoiding FreeBSD 4
Joerg Sonnenberger [Sat, 29 Jan 2005 13:51:57 +0000 (13:51 +0000)]
Don't build sysctl as bootstrap tool. This helps us avoiding FreeBSD 4
build problems.

Suggested-by: Chuck Tuffli <chuck_tuffli@agilent.com>
18 years agoWe don't currently build ld-elf as binary for i386 and even if
Joerg Sonnenberger [Sat, 29 Jan 2005 13:21:07 +0000 (13:21 +0000)]
We don't currently build ld-elf as binary for i386 and even if
we want to do that, we wouldn't care about FreeBSD 4.0 support.
Therefore no need to query kernel version here.

18 years agoFix the virtual 'status' file for procfs. The wrong length was being used,
Matthew Dillon [Sat, 29 Jan 2005 08:34:00 +0000 (08:34 +0000)]
Fix the virtual 'status' file for procfs.  The wrong length was being used,
returning a 0-length result every time.

Reported-by: "Simon 'corecode' Schubert" <corecode@fs.ei.tum.de>
18 years agoDo not leave VCTTYISOPEN set if our attempt to open /dev/tty fails, otherwise
Matthew Dillon [Sat, 29 Jan 2005 05:48:17 +0000 (05:48 +0000)]
Do not leave VCTTYISOPEN set if our attempt to open /dev/tty fails, otherwise
we will close the controlling terminal on exit even though we didn't open it.

This case occurs when a jailed shell is created because /dev/tty cannot be
opened in that case.

Reported-by: "Simon 'corecode' Schubert" <corecode@fs.ei.tum.de>
18 years ago- Replace confusing for loop with a do {} while (). Tested by making
Max Okumoto [Sat, 29 Jan 2005 00:30:42 +0000 (00:30 +0000)]
- Replace confusing for loop with a do {} while ().  Tested by making
  sure the object files were the same.

18 years ago- Cleanup Var_Subst() by removing duplicated code.
Max Okumoto [Fri, 28 Jan 2005 23:26:41 +0000 (23:26 +0000)]
- Cleanup Var_Subst() by removing duplicated code.

18 years ago- Factor out the duplicate code and move it into VarCreate() and VarDestroy().
Max Okumoto [Fri, 28 Jan 2005 23:25:18 +0000 (23:25 +0000)]
- Factor out the duplicate code and move it into VarCreate() and VarDestroy().

18 years agoScrap DEC Alpha support.
Sascha Wildner [Fri, 28 Jan 2005 21:08:38 +0000 (21:08 +0000)]
Scrap DEC Alpha support.

18 years agoRemove some old ioctls and structs which were there for backwards
Sascha Wildner [Fri, 28 Jan 2005 20:17:18 +0000 (20:17 +0000)]
Remove some old ioctls and structs which were there for backwards
compatibility with FreeBSD 3.x.

18 years agoAdd IOCTLS section and do some minor cleanup.
Sascha Wildner [Fri, 28 Jan 2005 19:12:11 +0000 (19:12 +0000)]
Add IOCTLS section and do some minor cleanup.

18 years agoMinor Patch
Liam J. Foy [Fri, 28 Jan 2005 16:53:38 +0000 (16:53 +0000)]
Minor Patch

- Fix data type
- Remove unnecessary header
- Instead of if..else.., just set as default behaviour

18 years agoClarify what new code should not cast unused
Max Okumoto [Fri, 28 Jan 2005 09:46:31 +0000 (09:46 +0000)]
Clarify what new code should not cast unused
return values of functions to void.

18 years ago- Fix indentation in VarFind()
Max Okumoto [Fri, 28 Jan 2005 06:00:34 +0000 (06:00 +0000)]
- Fix indentation in VarFind()

18 years agoremove bad semicolon
Joerg Sonnenberger [Thu, 27 Jan 2005 20:49:46 +0000 (20:49 +0000)]
remove bad semicolon

18 years agoANSIfication and minor style cleanups.
Joerg Sonnenberger [Thu, 27 Jan 2005 20:00:00 +0000 (20:00 +0000)]
ANSIfication and minor style cleanups.

18 years agoMount points use a special empty namecache entry to transition from one
Matthew Dillon [Thu, 27 Jan 2005 19:46:48 +0000 (19:46 +0000)]
Mount points use a special empty namecache entry to transition from one
filesystem to another.  It is possible for a stale entry to remain intact
from a prior mount so be sure the new entry is set to an unresolved state
so it is properly re-resolved on later access.

18 years agoTell the user more explicitly what port needs to be installed to get the
Matthew Dillon [Thu, 27 Jan 2005 19:42:32 +0000 (19:42 +0000)]
Tell the user more explicitly what port needs to be installed to get the
mkisofs utility required by the release build.

18 years agoAdjust FreeBSD tag to reflect sychronization.
Sascha Wildner [Thu, 27 Jan 2005 19:17:13 +0000 (19:17 +0000)]
Adjust FreeBSD tag to reflect sychronization.

18 years ago- Sync with FreeBSD.
Sascha Wildner [Thu, 27 Jan 2005 18:06:42 +0000 (18:06 +0000)]
- Sync with FreeBSD.
- Mention DragonFly where appropriate.
- Fix some punctuation.

Noticed-by: Heinrich Rebehn <rebehn@ant.uni-bremen.de>
            Freddie Cash <fcash-ml@sd73.bc.ca>

18 years ago- Add Buf_StripNewlines().
Max Okumoto [Thu, 27 Jan 2005 10:25:19 +0000 (10:25 +0000)]
- Add Buf_StripNewlines().
- Modified Cmd_Exec() to return a Buffer object instead of a 'char *', and
  updated all the callers.

18 years agoUse M_INTWAIT, not M_NOWAIT. We don't really support fast interrupt
Joerg Sonnenberger [Thu, 27 Jan 2005 02:43:12 +0000 (02:43 +0000)]
Use M_INTWAIT, not M_NOWAIT. We don't really support fast interrupt
block devices anyway.

18 years agoAdd a default for Perl 5.8.5 to override the default for
Joerg Sonnenberger [Thu, 27 Jan 2005 02:38:31 +0000 (02:38 +0000)]
Add a default for Perl 5.8.5 to override the default for
system perl.

18 years ago- New function Buf_AppendRange(), which is given a pointer to a string and
Max Okumoto [Thu, 27 Jan 2005 02:30:19 +0000 (02:30 +0000)]
- New function Buf_AppendRange(), which is given a pointer to a string and
  a pointer to the end of the range to append to the Buffer object.

tested by full build world.

18 years ago- New function Buf_Append(), which is given a pointer to a string to
Max Okumoto [Thu, 27 Jan 2005 02:28:48 +0000 (02:28 +0000)]
- New function Buf_Append(), which is given a pointer to a string to
  append to the Buffer object.

18 years ago- Inline silly ADDWORD() macro.
Max Okumoto [Thu, 27 Jan 2005 02:20:45 +0000 (02:20 +0000)]
- Inline silly ADDWORD() macro.

18 years ago- Rewrite Buffer object to maintain only write pointer. There is no
Max Okumoto [Thu, 27 Jan 2005 02:19:45 +0000 (02:19 +0000)]
- Rewrite Buffer object to maintain only write pointer.  There is no
  need to maintain both a write pointer and a length.

test by a full build world.

18 years agoUse and define size_t as specified by SUSv3.
Joerg Sonnenberger [Thu, 27 Jan 2005 01:50:01 +0000 (01:50 +0000)]
Use and define size_t as specified by SUSv3.

18 years agoCode cleanup. Refactor some functions. Push some globals into local scope.
Jeffrey Hsu [Wed, 26 Jan 2005 23:09:58 +0000 (23:09 +0000)]
Code cleanup.  Refactor some functions.  Push some globals into local scope.

18 years ago- Bring in patch to make.1 related to previous patch.
Max Okumoto [Wed, 26 Jan 2005 22:47:25 +0000 (22:47 +0000)]
- Bring in patch to make.1 related to previous patch.

taken-from: FreeBSD
author: harti

18 years agoExplicitly mention in RETURN VALUES that socket() sets
Sascha Wildner [Wed, 26 Jan 2005 21:52:36 +0000 (21:52 +0000)]
Explicitly mention in RETURN VALUES that socket() sets
errno upon failure.

Noticed-by: Liam J. Foy <liamfoy@sepulcrum.org>
18 years agoDon't build a new mbuf and bpf_mtap the old one.
Joerg Sonnenberger [Wed, 26 Jan 2005 16:15:06 +0000 (16:15 +0000)]
Don't build a new mbuf and bpf_mtap the old one.

18 years agoFix the quoting of MAKEFLAGS as noted in rev. 1.47.
Joerg Sonnenberger [Wed, 26 Jan 2005 09:44:21 +0000 (09:44 +0000)]
Fix the quoting of MAKEFLAGS as noted in rev. 1.47.
This still has a slight problem with trailing backslashes
on the comment line, but that will be documented in the man page.

Written-by: Harti Brandt, harti AT freebsd DOT org
18 years agoAdd some spaces for typographical correctness.
Joerg Sonnenberger [Wed, 26 Jan 2005 00:55:44 +0000 (00:55 +0000)]
Add some spaces for typographical correctness.