]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
19 years agoSupport for dynamically loadable and unloadable IP protocols in the ipmux.
andre [Tue, 19 Oct 2004 15:45:57 +0000 (15:45 +0000)]
Support for dynamically loadable and unloadable IP protocols in the ipmux.

With pr_proto_register() it has become possible to dynamically load protocols
within the PF_INET domain.  However the PF_INET domain has a second important
structure called ip_protox[] that is derived from the 'struct protosw inetsw[]'
and takes care of the de-multiplexing of the various protocols that ride on
top of IP packets.

The functions ipproto_[un]register() allow to dynamically adjust the ip_protox[]
array mux in a consistent and easy way.  To register a protocol within
ip_protox[] the existence of a corresponding and matching protocol definition
in inetsw[] is required.  The function does not allow to overwrite an already
registered protocol.  The unregister function simply replaces the mux slot with
the default index pointer to IPPROTO_RAW as it was previously.

19 years agoDetach the Rhine completely on shutdown, rather than merely stopping it
bms [Tue, 19 Oct 2004 15:30:47 +0000 (15:30 +0000)]
Detach the Rhine completely on shutdown, rather than merely stopping it
as the original logic did. This fixes a race with vr_intr() which was
masked on UP systems and manifested on SMP systems.

PR: kern/62889
MFC after: 1 day

19 years agoSupport for dynamically loadable and unloadable protocols within existing protocol
andre [Tue, 19 Oct 2004 15:13:30 +0000 (15:13 +0000)]
Support for dynamically loadable and unloadable protocols within existing protocol
families.

The protosw[] array of any particular protocol family ("domain") is of fixed size
defined at compile time.  This made it impossible to dynamically add or remove any
protocols to or from it.  We work around this by introducing so called SPACER's
which are embedded into the protosw[] array at compile time.  The SPACER's have
a special protocol number (32767) to indicate the fact that they are SPACER's but
are otherwise NULL.  Only as many protocols can be dynamically loaded as SPACER's
are provided in the protosw[] structure.

The pr_usrreqs structure is treated more special and contains pointers to dummy
functions only returning EOPNOTSUPP.  This is needed because the use of those
functions pointers is usually not checked within the kernel because until now it
was assumed to be a valid function pointer.  Instead of fixing all potential
callers we just return a proper error code.

Two new functions provide a clean API to register and unregister a protocol.  The
register function expects a pointer to a valid and complete struct protosw including
a pointer to struct pru_usrreqs provided by the caller.  Upon successful registration
the pr_init() function will be called to finish initialization of the protocol.  The
unregister function restores the SPACER in place of the protocol again.  It is the
responseability of the caller to ensure proper closing of all sockets and freeing
of memory allocation by the unloading protocol.

 sys/protosw.h

  o Define generic PROTO_SPACER to be 32767
  o Prototypes for all pru_*_notsupp() functions
  o Prototypes for pf_proto_[un]register() functions

 kern/uipc_domain.c

  o Global struct pr_usrreqs nousrreqs containing valid pointers to the
    pru_*_notsupp() functions
  o New functions pf_proto_[un]register()

 kern/uipc_socket2.c

  o New functions bodies for all pru_*_notsupp() functions

19 years agoAdd a macro for the destruction of INP_INFO_LOCK's used by loadable modules.
andre [Tue, 19 Oct 2004 14:34:13 +0000 (14:34 +0000)]
Add a macro for the destruction of INP_INFO_LOCK's used by loadable modules.

19 years agoMake comments more clear. Change the order of one if() statement to check the
andre [Tue, 19 Oct 2004 14:31:56 +0000 (14:31 +0000)]
Make comments more clear.  Change the order of one if() statement to check the
more likely variable first.

19 years agoBe more careful to only index valid IP protocols and be more verbose with
andre [Tue, 19 Oct 2004 14:26:44 +0000 (14:26 +0000)]
Be more careful to only index valid IP protocols and be more verbose with
comments.

19 years agoReturn the unit number of a channel instead of a hardcoded '1' from
le [Tue, 19 Oct 2004 10:29:00 +0000 (10:29 +0000)]
Return the unit number of a channel instead of a hardcoded '1' from
the ATA pccard locking function.  This makes pccard devices like
Compact Flash cards work again.

PR:             kern/72805
Submitted by:   James E. Flemer <jflemer@alum.rpi.edu>
MFC in:         2 days

19 years agoForced commit to note that the previous change also elimates calls to
scottl [Tue, 19 Oct 2004 02:44:38 +0000 (02:44 +0000)]
Forced commit to note that the previous change also elimates calls to
bus_dmamap_create|destroy for the rx and tx descriptor buffers.  Since these
buffers are created with bus_dmamem_alloc(), there is no reason to also
create a map, and doing so just wastes memory.

19 years agoUse and alignment of 1 instead of ETHER_ALIGN for rx and tx buffers and jumbo
scottl [Tue, 19 Oct 2004 02:42:49 +0000 (02:42 +0000)]
Use and alignment of 1 instead of ETHER_ALIGN for rx and tx buffers and jumbo
frames.  BGE hardware with the rx alignment bug will still be handled by the
calls to m_adj() that already exist.  m_adj() is probably better suited for
this task anyways.  Just as with if_em, this saves a malloc + several locks
per packet and prevents unneeded data copying within busdma.

19 years agoUse an alignment of 1 instead of PAGE_SIZE for the rx and tx buffer tags.
scottl [Tue, 19 Oct 2004 02:39:27 +0000 (02:39 +0000)]
Use an alignment of 1 instead of PAGE_SIZE for the rx and tx buffer tags.
Since the e1000 DMA engines hava no constraints on the alignment of buffer
transfers, there is no reason to tell busdma that there is.  This save a
minimum of 1 malloc call per packet, which translates to eliminating 4 locks.
It also means that buffers are not needlessly bounced when transfered.  The
end result is a 38% improvement in pps in a 4 way bridging environment.

Obtained from: Sandvine, Inc.

19 years agoWhen dumpdev is set to 'auto', and a suitable swap device is found,
thomas [Mon, 18 Oct 2004 23:40:13 +0000 (23:40 +0000)]
When dumpdev is set to 'auto', and a suitable swap device is found,
create a symbolic link /dev/dumpdev designating that device so
savecore can find and save a previous kernel dump.

19 years agofix (for me) the problems where if_de gets really slow after time
jmg [Mon, 18 Oct 2004 23:06:12 +0000 (23:06 +0000)]
fix (for me) the problems where if_de gets really slow after time
(usually taking 20 seconds to transmit a packet).. no longer fall back
to only transmitting one packet (instead of the entire queue) after we
have processed the entire send queue...  I have no idea why we didn't
start seeing this problem ~6 years ago when this code was introduced...

19 years agoPush acquisition of the accept mutex out of sofree() into the caller
rwatson [Mon, 18 Oct 2004 22:19:43 +0000 (22:19 +0000)]
Push acquisition of the accept mutex out of sofree() into the caller
(sorele()/sotryfree()):

- This permits the caller to acquire the accept mutex before the socket
  mutex, avoiding sofree() having to drop the socket mutex and re-order,
  which could lead to races permitting more than one thread to enter
  sofree() after a socket is ready to be free'd.

- This also covers clearing of the so_pcb weak socket reference from
  the protocol to the socket, preventing races in clearing and
  evaluation of the reference such that sofree() might be called more
  than once on the same socket.

This appears to close a race I was able to easily trigger by repeatedly
opening and resetting TCP connections to a host, in which the
tcp_close() code called as a result of the RST raced with the close()
of the accepted socket in the user process resulting in simultaneous
attempts to de-allocate the same socket.  The new locking increases
the overhead for operations that may potentially free the socket, so we
will want to revise the synchronization strategy here as we normalize
the reference counting model for sockets.  The use of the accept mutex
in freeing of sockets that are not listen sockets is primarily
motivated by the potential need to remove the socket from the
incomplete connection queue on its parent (listen) socket, so cleaning
up the reference model here may allow us to substantially weaken the
synchronization requirements.

RELENG_5_3 candidate.

MFC after: 3 days
Reviewed by: dwhite
Discussed with: gnn, dwhite, green
Reported by: Marc UBM Bocklet <ubm at u-boot-man dot de>
Reported by: Vlad <marchenko at gmail dot com>

19 years agouniq(1) is not an install tool, and using it was causing
ru [Mon, 18 Oct 2004 22:18:23 +0000 (22:18 +0000)]
uniq(1) is not an install tool, and using it was causing
"uniq: not found" during the installworld.

Spotted by: Roman Neuhauser

19 years agoAdd new function ttyinitmode() which sets our systemwide default
phk [Mon, 18 Oct 2004 21:51:27 +0000 (21:51 +0000)]
Add new function ttyinitmode() which sets our systemwide default
modes on a tty structure.

Both the ".init" and the current settings are initialized allowing
the function to be used both at attach and open time.

The function takes an argument to decide if echoing should be enabled.
Echoing should not be enabled for regular physical serial ports
unless they are consoles, in which case they should be configured
by ttyconsolemode() instead.

Use the new function throughout.

19 years agoUtilize SCRIPTS.
ru [Mon, 18 Oct 2004 21:44:17 +0000 (21:44 +0000)]
Utilize SCRIPTS.

19 years agoUtilize FILES, SCRIPTS, and SYMLINKS. While here, fixed a bug in
ru [Mon, 18 Oct 2004 21:42:15 +0000 (21:42 +0000)]
Utilize FILES, SCRIPTS, and SYMLINKS.  While here, fixed a bug in
the implementation of the following feature in revision 1.4:

- Install Makefile.yp as /var/yp/Makefile.dist and link it to
  /var/yp/Makefile only if /var/yp/Makefile doesn't already exist.
  Suggested by Peter Wemm.

The actual code was only symlinking when no /var/yp/Makefile.dist
existed, i.e., never.

19 years agoDrop the NDIS lock before returning from ndis_start().
le [Mon, 18 Oct 2004 21:33:56 +0000 (21:33 +0000)]
Drop the NDIS lock before returning from ndis_start().

PR:             i386/72795
Submitted by:   Frank Mayhar <frank@exit.com>
MFC in:         3 days

19 years agoCatch up with PHK's sio(4) rework [sys/dev/sio/sio.c rev. 1.456].
obrien [Mon, 18 Oct 2004 21:31:05 +0000 (21:31 +0000)]
Catch up with PHK's sio(4) rework [sys/dev/sio/sio.c rev. 1.456].
Also catch up with RCng.

Reviewed by: phk

19 years agoUtilize SCRIPTS.
ru [Mon, 18 Oct 2004 21:29:20 +0000 (21:29 +0000)]
Utilize SCRIPTS.

19 years agoAdd an entry about some tty devices being renamed.
phk [Mon, 18 Oct 2004 21:24:21 +0000 (21:24 +0000)]
Add an entry about some tty devices being renamed.

19 years agoMajor overhaul.
glebius [Mon, 18 Oct 2004 20:13:57 +0000 (20:13 +0000)]
Major overhaul.

 List of functional changes:
   - Make a single device per single node with a single hook.
     This gives us parrallelizm, which can't be achieved on a single
     node with many devices/hooks. This also gives us flexibility - we
     can play with a particular device node, not affecting others.
   - Remove read queue as it is. Use struct ifqueue instead. This change
     removes a lot of extra memcpy()ing, m_devget()ting and m_copymem()ming.
     In ng_device_receivedata() we enqueue an mbuf and wake readers.
     In ngdread() we take one mbuf from qeueue and uiomove() it to
     userspace. If no mbuf is present we optionally block. [1]
   - In ngdwrite() we create an mbuf from uio using m_uiotombuf().
     This is faster then uiomove() into buffer, and then m_copydata(),
     and this is much better than huge m_pullup().
   - Perform locking of device
   - Perform locking of connection list.
   - Clear out _rcvmsg method, since it does nothing good yet.
   - Implement NGM_DEVICE_GET_DEVNAME message.
   - #if 0 ioctl method, while nothing is done here yet.
   - Return immediately from ngdwrite() if uio_resid == 0.

 List of tidyness changes:
   - Introduce device2priv(), to remove cut'n'paste.
   - Use MALLOC/FREE, instead of malloc/free.
   - Use unit2minor().
   - Use UID_ROOT/GID_WHEEL instead of 0/0.
   - Define NGD_DEVICE_DEVNAME, use it.
   - Use more nice macros for debugging. [2]
   - Return Exxx, not -1.

 style(9) changes:
   - No "#endif" after short block.
   - Break long lines.
   - Remove extra spaces, add needed spaces.

[1] Obtained from:      if_tun.c
[2] Obtained from:      ng_pppoe.c
Reviewed by: marks
Approved by: julian (mentor)
MFC after: 1 month

19 years agoAnnotate that get_cyclecount() can be expensive on some platforms,
rwatson [Mon, 18 Oct 2004 19:29:13 +0000 (19:29 +0000)]
Annotate that get_cyclecount() can be expensive on some platforms,
which juxtaposes nicely with the comment just above on how the
harvest function must be cheap.

19 years agoFixed MAKEOBJDIRPREFIX variable type check in case the env(1)
ru [Mon, 18 Oct 2004 17:47:31 +0000 (17:47 +0000)]
Fixed MAKEOBJDIRPREFIX variable type check in case the env(1)
binary is not in PATH, as is the case with installworld, etc.

19 years agoReplaced afterinstall: with FILES.
ru [Mon, 18 Oct 2004 17:20:29 +0000 (17:20 +0000)]
Replaced afterinstall: with FILES.

19 years agoSync with sys/i386/include/endian.h: use the single instruction 'bswap'.
obrien [Mon, 18 Oct 2004 17:19:36 +0000 (17:19 +0000)]
Sync with sys/i386/include/endian.h: use the single instruction 'bswap'.

19 years agoReplaced afterinstall: with equivalent FILES and SYMLINKS.
ru [Mon, 18 Oct 2004 17:16:01 +0000 (17:16 +0000)]
Replaced afterinstall: with equivalent FILES and SYMLINKS.

19 years agoNew directory added for Argentinian timezones.
wollman [Mon, 18 Oct 2004 17:03:53 +0000 (17:03 +0000)]
New directory added for Argentinian timezones.

Submitted by: ache

19 years agoReplaced afterinstall: with FILES.
ru [Mon, 18 Oct 2004 16:39:53 +0000 (16:39 +0000)]
Replaced afterinstall: with FILES.

19 years agoReplaced afterinstall: with FILES.
ru [Mon, 18 Oct 2004 16:26:25 +0000 (16:26 +0000)]
Replaced afterinstall: with FILES.

19 years agoModify behaviour of `xargs -I` in order to:
cperciva [Mon, 18 Oct 2004 15:40:47 +0000 (15:40 +0000)]
Modify behaviour of `xargs -I` in order to:
1. Conform to IEEE Std 1003.1-2004, which state that "Constructed
arguments cannot grow larger than 255 bytes", and
2. Avoid a buffer overflow.

Unfortunately the standard doesn't indicate how xargs is supposed to
handle arguments which (with the appropriate substitutions) would grow
larger than 255 bytes; this solution handles those by making as many
substitutions as possible without overflowing the buffer.

OpenBSD's xargs resolves this in a different direction, by making
all the substitutions and then silently truncating the resulting string.

Since this change may break existing scripts which rely upon the buffer
overflow (255 bytes isn't really all that long...) it will not be MFCed.

19 years agoRemove unused computation of memory size.
thomas [Mon, 18 Oct 2004 14:59:53 +0000 (14:59 +0000)]
Remove unused computation of memory size.

Reviewed by: des

19 years agoCorrect several instances where calls to vfs_getvfs() resulting in
rwatson [Mon, 18 Oct 2004 11:23:11 +0000 (11:23 +0000)]
Correct several instances where calls to vfs_getvfs() resulting in
failure in the NFS server would result in a leaked instance of the NFS
server subsystem lock.  Liberally sprinkle assertions in all target
labels for error unwinding to assert the desired locking state.

RELENG_5_3 candidate.

MFC after: 3 days
Reported by: Wilkinson, Alex <alex dot wilkinson at dsto dot defence dot gov dot au>

19 years agoUpdate to readline 5.0
ache [Mon, 18 Oct 2004 10:13:24 +0000 (10:13 +0000)]
Update to readline 5.0

19 years agoSince sendfile(2) works on regular files only,
yar [Mon, 18 Oct 2004 10:02:04 +0000 (10:02 +0000)]
Since sendfile(2) works on regular files only,
which have no negative offsets, "negative" and
"invalid" are equivalent WRT the offset argument.

Suggested by: bde

19 years agoUtilize FILES and SYMLINKS.
ru [Mon, 18 Oct 2004 08:54:17 +0000 (08:54 +0000)]
Utilize FILES and SYMLINKS.

19 years agoBack out termcap -> ncurses change in LDADD/DPADD
ache [Mon, 18 Oct 2004 08:36:40 +0000 (08:36 +0000)]
Back out termcap -> ncurses change in LDADD/DPADD
It expand functions scope too wide

19 years agoUpgrade to 5.0
ache [Mon, 18 Oct 2004 08:29:47 +0000 (08:29 +0000)]
Upgrade to 5.0

19 years agoCorrect two errors in PG_BUSY management by vm_page_cowfault(). Both
alc [Mon, 18 Oct 2004 08:11:59 +0000 (08:11 +0000)]
Correct two errors in PG_BUSY management by vm_page_cowfault().  Both
errors are in rarely executed paths.
1. Each time the retry_alloc path is taken, the PG_BUSY must be set again.
   Otherwise vm_page_remove() panics.
2. There is no need to set PG_BUSY on the newly allocated page before
   freeing it.  The page already has PG_BUSY set by vm_page_alloc().
   Setting it again could cause an assertion failure.

MFC after: 2 weeks

19 years agoUpgrade to 5.0
ache [Mon, 18 Oct 2004 08:08:03 +0000 (08:08 +0000)]
Upgrade to 5.0

19 years agoUpdate for 5.0
ache [Mon, 18 Oct 2004 07:49:59 +0000 (07:49 +0000)]
Update for 5.0

19 years agoMerge local changes
ache [Mon, 18 Oct 2004 07:25:32 +0000 (07:25 +0000)]
Merge local changes

19 years agoRemove files not in 5.0
ache [Mon, 18 Oct 2004 07:18:06 +0000 (07:18 +0000)]
Remove files not in 5.0

19 years agoUpdate instructions
ache [Mon, 18 Oct 2004 07:15:43 +0000 (07:15 +0000)]
Update instructions

19 years agoThis commit was generated by cvs2svn to compensate for changes in r136647,
ache [Mon, 18 Oct 2004 07:10:33 +0000 (07:10 +0000)]
This commit was generated by cvs2svn to compensate for changes in r136647,
which included commits to RCS files with non-trunk default branches.

19 years agoOfficial patches 001-005
ache [Mon, 18 Oct 2004 07:10:33 +0000 (07:10 +0000)]
Official patches 001-005

19 years agoThis commit was generated by cvs2svn to compensate for changes in r136644,
ache [Mon, 18 Oct 2004 07:02:42 +0000 (07:02 +0000)]
This commit was generated by cvs2svn to compensate for changes in r136644,
which included commits to RCS files with non-trunk default branches.

19 years agoVirgin import of GNU Readline 5.0
ache [Mon, 18 Oct 2004 07:02:42 +0000 (07:02 +0000)]
Virgin import of GNU Readline 5.0

19 years agoAdd a -q flag to pkg_version. It suppresses the output of <,=,>.
obrien [Mon, 18 Oct 2004 05:34:54 +0000 (05:34 +0000)]
Add a -q flag to pkg_version.  It suppresses the output of <,=,>.
This makes it possible to "portupgrade `pkg_version -q -l '<'`".

19 years agoRevert 1.18: It broke Athlon64 builds, which
kientzle [Mon, 18 Oct 2004 05:31:01 +0000 (05:31 +0000)]
Revert 1.18:  It broke Athlon64 builds, which
probably means it also requires a .so version
bump.  Defer it until I finish some related
work on cleaning up error returns throughout
the library.

Thanks to: Conrad J. Sabatier

19 years agoUse STDERR_FILENO instead of 2, as POSIX intended.
kientzle [Mon, 18 Oct 2004 04:34:30 +0000 (04:34 +0000)]
Use STDERR_FILENO instead of 2, as POSIX intended.

Thanks to: Alfred Perlstein

19 years agoRemove setrunelocale()
ache [Mon, 18 Oct 2004 02:06:18 +0000 (02:06 +0000)]
Remove setrunelocale()

19 years agoTypo fix.
keramida [Mon, 18 Oct 2004 00:45:28 +0000 (00:45 +0000)]
Typo fix.

PR: misc/72801
Submitted by: Hywel Mallett <circular@hmallett.co.uk>

19 years agoMFV tzdata2004e
wollman [Mon, 18 Oct 2004 00:03:52 +0000 (00:03 +0000)]
MFV tzdata2004e

Users in Brazil and Argentina should run tzsetup(8).

Obtained from: Arthur David Olson
MFC after re approval

19 years agogetopt(3) prints an "unrecognized option" warning for me, so I
kientzle [Sun, 17 Oct 2004 23:58:17 +0000 (23:58 +0000)]
getopt(3) prints an "unrecognized option" warning for me, so I
shouldn't print another.

MFC after: 15 days

19 years agoCosmetic goof: Use two spaces in "Jan 03 2004" but only one in "Jan 03 12:32"
kientzle [Sun, 17 Oct 2004 23:57:10 +0000 (23:57 +0000)]
Cosmetic goof: Use two spaces in "Jan 03  2004" but only one in "Jan 03 12:32"
MFC after: 15 days

19 years agoThis commit was generated by cvs2svn to compensate for changes in r136632,
wollman [Sun, 17 Oct 2004 23:55:07 +0000 (23:55 +0000)]
This commit was generated by cvs2svn to compensate for changes in r136632,
which included commits to RCS files with non-trunk default branches.

19 years agoLatest timezone database from Arthur Olson. Updates in this round:
wollman [Sun, 17 Oct 2004 23:55:07 +0000 (23:55 +0000)]
Latest timezone database from Arthur Olson.  Updates in this round:

Current or prospective timestamps: Georgia, Uruguay, Argentina, and Brazil

Historial timestamps and/or commentary: Malaysia, Mongolia, Singapore,
Peru, Russua, Israel, United States, Canada

Changes in Argentina and Brazil have created several new time zone regions.
Users are encouraged to rerun tzsetup(8), even if current times appear
correct, to ensure that future times will be interepreted correctly.

Obtained from: Arthur Olson, ftp://elsie.nci.nih.gov/pub/tzdata2004e.tar.gz

19 years agoCorrectly report write errors from the lowest-level
kientzle [Sun, 17 Oct 2004 23:47:30 +0000 (23:47 +0000)]
Correctly report write errors from the lowest-level
output routines back to the compression layer.

19 years agoRefine the error-checking and reporting in the
kientzle [Sun, 17 Oct 2004 23:40:10 +0000 (23:40 +0000)]
Refine the error-checking and reporting in the
"compress" format decompression code.  In particular,
distinguish between EOF and fatal data errors.

19 years agoCorrect the return type of archive_write_data
kientzle [Sun, 17 Oct 2004 23:37:09 +0000 (23:37 +0000)]
Correct the return type of archive_write_data
to match the documentation.

MFC after: 30 days

19 years agoDon't rely on stdio here.
kientzle [Sun, 17 Oct 2004 23:35:16 +0000 (23:35 +0000)]
Don't rely on stdio here.

19 years agoAssert that the containing object is locked in vm_page_io_start() and
alc [Sun, 17 Oct 2004 22:33:40 +0000 (22:33 +0000)]
Assert that the containing object is locked in vm_page_io_start() and
vm_page_io_finish().  The motivation being to transition synchronization of
the vm_page's busy field from the global page queues lock to the per-object
lock.

19 years agoturn off getty(8) on ttyv[0-7]. We don't support syscons on ia64 and
marcel [Sun, 17 Oct 2004 22:20:39 +0000 (22:20 +0000)]
turn off getty(8) on ttyv[0-7]. We don't support syscons on ia64 and
do not have these device special files. Where this previously failed
quietly, it now emits annoying but complete messages at best and
incomprehensible prefixes on average. During all of October, this is
a string of 16 O's, as in:

:
Starting inetd.

Sun Oct 17 15:09:09 PDT 2004
OOOOOOOOOOOOOOOO
FreeBSD/ia64 (itanium.pn.xcllnt.net) (ttyu2)

login:

19 years agoUse cluster if data >= MINCLSIZE.
glebius [Sun, 17 Oct 2004 21:44:11 +0000 (21:44 +0000)]
Use cluster if data >= MINCLSIZE.

Reviewed by: mdodd
Approved by: julian (mentor)
MFC after: 1 month

19 years agoRemove an extra semicolon after a function.
stefanf [Sun, 17 Oct 2004 21:29:53 +0000 (21:29 +0000)]
Remove an extra semicolon after a function.

19 years agoFix typo in rev. 1.58 (missing $) that invalidated the conditional and
marcel [Sun, 17 Oct 2004 21:05:40 +0000 (21:05 +0000)]
Fix typo in rev. 1.58 (missing $) that invalidated the conditional and
caused freebsd-update to be added to the ia64 package list.

19 years agoRemove unnecessary check for curthread == NULL.
alc [Sun, 17 Oct 2004 20:29:28 +0000 (20:29 +0000)]
Remove unnecessary check for curthread == NULL.

19 years agoRemove definition of struct dom_binding, it's non-standard C code and
stefanf [Sun, 17 Oct 2004 19:33:33 +0000 (19:33 +0000)]
Remove definition of struct dom_binding, it's non-standard C code and
unnecessary since src/include/rpcsvc/ypclnt.h's revision 1.10.

19 years agoFix a typo: s/kerncruft/kernelcruft/.
maxim [Sun, 17 Oct 2004 19:02:17 +0000 (19:02 +0000)]
Fix a typo: s/kerncruft/kernelcruft/.

PR: misc/72792
Submitted by: Hywel Mallett

19 years agoRemove the suggestion to use libmap.conf to deal with the version bump.
obrien [Sun, 17 Oct 2004 14:59:18 +0000 (14:59 +0000)]
Remove the suggestion to use libmap.conf to deal with the version bump.
People have a tendency to add things to libmap.conf and forget about them
(as we've seen in the mailing lists before).
The compat4x port should be used instead.

19 years agoSet the real limits on 5-current, add 5-stable, add 6-current.
obrien [Sun, 17 Oct 2004 14:36:30 +0000 (14:36 +0000)]
Set the real limits on 5-current, add 5-stable, add 6-current.

19 years agoFix a bogus variable assignment. You can't expect
schweikh [Sun, 17 Oct 2004 13:39:42 +0000 (13:39 +0000)]
Fix a bogus variable assignment. You can't expect
   _msg="a" \
        " b"
to concat the strings. Use
   _msg="a"
   _msg="$_msg b"
instead (intent is to not exceed 80 chars per line).

MFC after: 1 week

19 years agoThe populate_var variable checks /var for writability, not /tmp.
keramida [Sun, 17 Oct 2004 13:19:10 +0000 (13:19 +0000)]
The populate_var variable checks /var for writability, not /tmp.

Noticed by: Alan Gerber <agerber@ncsu.edu>

19 years agoFixed a typo.
ru [Sun, 17 Oct 2004 11:07:03 +0000 (11:07 +0000)]
Fixed a typo.

Submitted by: Pawel Worach

19 years agoUtilize FILES and SYMLINKS.
ru [Sun, 17 Oct 2004 10:49:30 +0000 (10:49 +0000)]
Utilize FILES and SYMLINKS.

19 years agoUtilize FILES and SYMLINKS.
ru [Sun, 17 Oct 2004 09:48:23 +0000 (09:48 +0000)]
Utilize FILES and SYMLINKS.

19 years agoDocument EMSGSIZE return from recvmsg due to insufficient free files
alfred [Sun, 17 Oct 2004 07:13:34 +0000 (07:13 +0000)]
Document EMSGSIZE return from recvmsg due to insufficient free files
when transfering rights (file descriptors.)

19 years agoRemove the obsolete <rune.h> interface.
tjr [Sun, 17 Oct 2004 06:51:50 +0000 (06:51 +0000)]
Remove the obsolete <rune.h> interface.

19 years agoCatch up with share/mk/bsd.cpu.mk revision 1.43.
obrien [Sun, 17 Oct 2004 05:36:20 +0000 (05:36 +0000)]
Catch up with share/mk/bsd.cpu.mk revision 1.43.

19 years agoEmbellish the AMD64 sections a little. Including supporting 'nocona'.
obrien [Sun, 17 Oct 2004 05:19:45 +0000 (05:19 +0000)]
Embellish the AMD64 sections a little.  Including supporting 'nocona'.

19 years ago+ Simplify by treating the standard x86 CPU names as the CPUTYPE vs.
obrien [Sun, 17 Oct 2004 05:08:29 +0000 (05:08 +0000)]
+ Simplify by treating the standard x86 CPU names as the CPUTYPE vs.
  treating them as an alias.  Treat the shorthand versions as aliases.
+ Separate the x86 GCC CPU CFLAGS from the ICC CFLAGS.  This greatly
  simplifies the GCC section.  It also makes it more clear which CPU's
  have the same ICC CPU CFLAGS.
+ Remove redundancy in the alpha section.
+ Add forgotten ICC CPU CFLAGS for the mobile Intel CPU's added in rev. 1.42.

19 years agoUse <runetype.h> instead of <rune.h> since the latter is going away.
tjr [Sun, 17 Oct 2004 03:02:50 +0000 (03:02 +0000)]
Use <runetype.h> instead of <rune.h> since the latter is going away.

19 years agoRemove support for the obsolete UTF2 encoding.
tjr [Sun, 17 Oct 2004 02:29:15 +0000 (02:29 +0000)]
Remove support for the obsolete UTF2 encoding.

19 years agoBump the libc major version number to 6.
tjr [Sun, 17 Oct 2004 01:52:48 +0000 (01:52 +0000)]
Bump the libc major version number to 6.

19 years agoRemove explicit support for generating LC_CTYPE data files that specify
tjr [Sun, 17 Oct 2004 01:08:11 +0000 (01:08 +0000)]
Remove explicit support for generating LC_CTYPE data files that specify
the obsolete "UTF2" encoding.

19 years agoWhen sf_buf_alloc() replaces a virtual-to-physical mapping, it needn't
alc [Sat, 16 Oct 2004 22:32:50 +0000 (22:32 +0000)]
When sf_buf_alloc() replaces a virtual-to-physical mapping, it needn't
invalidate the TLB(s) if the old mapping wasn't used by the CPU.  With
network interfaces that implement checksum off-loading, the old mapping is
almost never used by the CPU, only by the device driver for setting up the
DMA operation.

Reviewed by: tegge@

19 years agoUpdate manpage for the Vodafone Mobile Connect 3G datacard. Bump
le [Sat, 16 Oct 2004 21:36:19 +0000 (21:36 +0000)]
Update manpage for the Vodafone Mobile Connect 3G datacard.  Bump
document date.

Reminded by:  simon

19 years agoAdd support for the Vodafone Mobile Connect 3G datacard.
le [Sat, 16 Oct 2004 21:27:28 +0000 (21:27 +0000)]
Add support for the Vodafone Mobile Connect 3G datacard.

19 years agoInstall files with mode 444, as God intended.
ru [Sat, 16 Oct 2004 20:52:32 +0000 (20:52 +0000)]
Install files with mode 444, as God intended.

19 years agoUtilize FILES and SYMLINKS.
ru [Sat, 16 Oct 2004 20:42:18 +0000 (20:42 +0000)]
Utilize FILES and SYMLINKS.

19 years agoUtilize FILES and SYMLINKS, other simplifications.
ru [Sat, 16 Oct 2004 19:51:34 +0000 (19:51 +0000)]
Utilize FILES and SYMLINKS, other simplifications.

19 years agoDocument that the length modifier l is ignored for floating point
stefanf [Sat, 16 Oct 2004 16:00:01 +0000 (16:00 +0000)]
Document that the length modifier l is ignored for floating point
conversion specifiers (a, A, e, E, f, F, g and G).

19 years agoExplain it is a negative offset that EINVAL may indicate.
yar [Sat, 16 Oct 2004 09:53:35 +0000 (09:53 +0000)]
Explain it is a negative offset that EINVAL may indicate.
Now readers won't get an impression that pointing to beyond
the current end of file will result in EINVAL.

MFC after: 1 week

19 years agoImprove mdoc(7) markup.
yar [Sat, 16 Oct 2004 09:46:38 +0000 (09:46 +0000)]
Improve mdoc(7) markup.

19 years agoGive details on what will happen if the `offset' argument
yar [Sat, 16 Oct 2004 09:39:25 +0000 (09:39 +0000)]
Give details on what will happen if the `offset' argument
to sendfile(2) falls beyond the end of file.

Touch .Dd.

PR: bin/72649 (in the audit trail)
MFC after: 1 week

19 years agoFix an ancient typo: `sc' usually equals `ns*nt', not `nc*nt'
yar [Sat, 16 Oct 2004 08:36:28 +0000 (08:36 +0000)]
Fix an ancient typo: `sc' usually equals `ns*nt', not `nc*nt'
(the latter is actually the total number of tracks on the disk,
which has no field in disktab(5).)

MFC after: 1 week

19 years agoIf a process needs to be swapped in, wakeup the swapper from within
scottl [Sat, 16 Oct 2004 06:38:22 +0000 (06:38 +0000)]
If a process needs to be swapped in, wakeup the swapper from within
critical_exit as the process is getting scheduled to run.  This is subotimal
but for now avoid the LOR between the scheduler and the sleepq systems.
This is a 5.3 candidate.

Submitted by: davidxu
MFC After: 3 days

19 years agoEnable building with LIBC_SCCS defined.
obrien [Sat, 16 Oct 2004 06:32:43 +0000 (06:32 +0000)]
Enable building with LIBC_SCCS defined.

Bug submitted by: Andrea Campi <andrea+freebsd_current@webcom.it>

19 years agoTry to bring some sanity to the SCM ID's.
obrien [Sat, 16 Oct 2004 06:11:35 +0000 (06:11 +0000)]
Try to bring some sanity to the SCM ID's.
+ spell LIBC_SCCS consistently
+ enable builds with LIBC_SCCS defined to not syntax error
+ minor SCM reformatting to try to have some consistency