]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
21 years agoUse m_length() instead of home-rolled versions.
Poul-Henning Kamp [Wed, 18 Sep 2002 19:44:14 +0000 (19:44 +0000)]
Use m_length() instead of home-rolled versions.

21 years agoUse m_fixhdr() rather than roll our own.
Poul-Henning Kamp [Wed, 18 Sep 2002 19:43:01 +0000 (19:43 +0000)]
Use m_fixhdr() rather than roll our own.

21 years agoMake m_length() and m_fixhdr() return unsigned.
Poul-Henning Kamp [Wed, 18 Sep 2002 19:42:06 +0000 (19:42 +0000)]
Make m_length() and m_fixhdr() return unsigned.

Suggested by: arr

21 years agoUpdate the usage string in the limits(1) manpage to include -b for
Brian Feldman [Wed, 18 Sep 2002 16:25:59 +0000 (16:25 +0000)]
Update the usage string in the limits(1) manpage to include -b for
sbsize. Also, correct the format string in getopt(3) usage to reflect
that -b takes an argument, and correct another case of RLIMIT_SBSIZE
having been forgotten.

21 years agoFix GNU coding violations.
David E. O'Brien [Wed, 18 Sep 2002 16:20:49 +0000 (16:20 +0000)]
Fix GNU coding violations.

21 years agoBasic stubbed-out support for building on sparc64 slipped in w/rev 1.41.
David E. O'Brien [Wed, 18 Sep 2002 16:19:05 +0000 (16:19 +0000)]
Basic stubbed-out support for building on sparc64 slipped in w/rev 1.41.

21 years agoStop using 3 distinct and far varied coding styles in a single file.
David E. O'Brien [Wed, 18 Sep 2002 16:14:36 +0000 (16:14 +0000)]
Stop using 3 distinct and far varied coding styles in a single file.
Pretend to actually adhere to the GNU coding standards.

21 years agoMake the DEBUGF() macro portable by (ugh) adding a Debug() function, which
Juli Mallett [Wed, 18 Sep 2002 16:13:03 +0000 (16:13 +0000)]
Make the DEBUGF() macro portable by (ugh) adding a Debug() function, which
is merely printf() but to stderr.  This takes care of the caveat which lead
to the use of a vararg macro -- getting everything to stderr.

21 years agoIntroduce the m_length() function which will return the accumulated
Poul-Henning Kamp [Wed, 18 Sep 2002 14:57:35 +0000 (14:57 +0000)]
Introduce the m_length() function which will return the accumulated
length of an mbuf-chain and optionally a pointer to the last mbuf.

21 years agoAdd missing #include <sys/mbuf.h>
Poul-Henning Kamp [Wed, 18 Sep 2002 14:21:52 +0000 (14:21 +0000)]
Add missing #include <sys/mbuf.h>

21 years agoThese RFCs and internet-drafts are not really needed in the base
Jacques Vidrine [Wed, 18 Sep 2002 14:17:14 +0000 (14:17 +0000)]
These RFCs and internet-drafts are not really needed in the base
system, and I've not been importing them lately.  cvs rm them now
so they can be cleaned out of the attic later.

Requested by: obrien

21 years agoMove m_fixhdr() from "mbchain" to "mbuf" where it belongs.
Poul-Henning Kamp [Wed, 18 Sep 2002 13:41:37 +0000 (13:41 +0000)]
Move m_fixhdr() from "mbchain" to "mbuf" where it belongs.

21 years agoOptimise the common case where no special encoding is in use (LC_CTYPE is "C"
Tim J. Robbins [Wed, 18 Sep 2002 12:17:28 +0000 (12:17 +0000)]
Optimise the common case where no special encoding is in use (LC_CTYPE is "C"
or "POSIX", other European locales). Use __sgetc() and __sputc() where
possible to avoid a wasteful lock and unlock for each byte and to avoid
function call overhead.

21 years agoLogic error in previous: don't exit the loop when an incomplete multibyte
Tim J. Robbins [Wed, 18 Sep 2002 10:21:41 +0000 (10:21 +0000)]
Logic error in previous: don't exit the loop when an incomplete multibyte
sequence is detected.

21 years agoAlways open file in the DENYNONE mode and let the server to decide what is
Boris Popov [Wed, 18 Sep 2002 09:54:16 +0000 (09:54 +0000)]
Always open file in the DENYNONE mode and let the server to decide what is
good for this file.
This should allow read only access to file which is already opened on server.

21 years agoUpdate supported chipset list.
Søren Schmidt [Wed, 18 Sep 2002 09:40:18 +0000 (09:40 +0000)]
Update supported chipset list.

21 years agoAdd support for the VIA 8235.
Søren Schmidt [Wed, 18 Sep 2002 09:39:37 +0000 (09:39 +0000)]
Add support for the VIA 8235.

Submitted by: Jason Dambrosio <jason@wiz.cx>

21 years agoStyle and markup changes only.
Don Lewis [Wed, 18 Sep 2002 09:34:01 +0000 (09:34 +0000)]
Style and markup changes only.

Capitalize the first letter of the descriptions for the entries in the ERRORS
section if they are complete sentences and end the sentences with periods.

21 years agoImplement additional SMB calls to allow proper update of file size as some
Boris Popov [Wed, 18 Sep 2002 09:27:04 +0000 (09:27 +0000)]
Implement additional SMB calls to allow proper update of file size as some
file servers fail to do it in the right way.

New NFLUSHWIRE flag marks pending flush request(s).

NB: not all cases covered by this commit.

Obtained from: Darwin

21 years agoFix implementation of rc variables $amd_flags and $amd_map_program in
Ralf S. Engelschall [Wed, 18 Sep 2002 08:49:50 +0000 (08:49 +0000)]
Fix implementation of rc variables $amd_flags and $amd_map_program in
rcNG. The $amd_flags variable was already taken over into $rc_flags
by run_rc_command() when amd_precmd() is executed, so changing
$amd_flags there no longer effects the actual execution of amd.
Hence in amd_precmd() the $rc_flags have to be adjusted instead.

21 years ago - Split UMA_ZFLAG_OFFPAGE into UMA_ZFLAG_OFFPAGE and UMA_ZFLAG_HASH.
Jeff Roberson [Wed, 18 Sep 2002 08:26:30 +0000 (08:26 +0000)]
 - Split UMA_ZFLAG_OFFPAGE into UMA_ZFLAG_OFFPAGE and UMA_ZFLAG_HASH.
 - Remove all instances of the mallochash.
 - Stash the slab pointer in the vm page's object pointer when allocating from
   the kmem_obj.
 - Use the overloaded object pointer to find slabs for malloced memory.

21 years agoARIN has annoyingly started using caps when referencing
Mike Barcroft [Wed, 18 Sep 2002 07:51:46 +0000 (07:51 +0000)]
ARIN has annoyingly started using caps when referencing
whois.apnic.net.  To properly receive hints for recursive IP searches,
we convert a buffer to lowercase before searching for magic words.

PR: 42834
Submitted by: Andre Albsmeier <andre.albsmeier@mchp.siemens.de> (mostly)
X-MFC-After: re approval

21 years agoPermit an empty username which is useful for browsing.
Boris Popov [Wed, 18 Sep 2002 07:43:21 +0000 (07:43 +0000)]
Permit an empty username which is useful for browsing.

21 years agoIncrease send/receive queue to accomodate large readx/writex requests.
Boris Popov [Wed, 18 Sep 2002 07:38:10 +0000 (07:38 +0000)]
Increase send/receive queue to accomodate large readx/writex requests.

Receive packets in a small pieces (NB_SORECEIVE_CHUNK), so TCP slowstart will
get its ACKs faster.

Obtained from: Darwin

21 years agoRestore status register A of RTC at resume time.
Mitsuru IWASAKI [Wed, 18 Sep 2002 07:34:04 +0000 (07:34 +0000)]
Restore status register A of RTC at resume time.
This should fix the 'too many RTC interrupts and statclock seems
broken after resume' problem.

MFC after: 1 week

21 years agoImplement C99's va_copy() macro.
Mike Barcroft [Wed, 18 Sep 2002 07:33:16 +0000 (07:33 +0000)]
Implement C99's va_copy() macro.

21 years agoDeprecate the rest of the rune interface.
Tim J. Robbins [Wed, 18 Sep 2002 06:19:12 +0000 (06:19 +0000)]
Deprecate the rest of the rune interface.

21 years agoMark mbmb(), mbrune(), and mbrrune() as deprecated functions. We want to
Tim J. Robbins [Wed, 18 Sep 2002 06:11:21 +0000 (06:11 +0000)]
Mark mbmb(), mbrune(), and mbrrune() as deprecated functions. We want to
get applications to move to the ISO C interfaces as well as have the
freedom to replace the rune interfaces with ones that support stateful
conversions some time in the future.

21 years agoReimplement the functionality of fgetrune(), fputrune(), and fungetrune()
Tim J. Robbins [Wed, 18 Sep 2002 05:58:11 +0000 (05:58 +0000)]
Reimplement the functionality of fgetrune(), fputrune(), and fungetrune()
here in terms of mbrtowc(), wcrtomb(), and the single-byte I/O functions.
The rune I/O functions are about to become deprecated in favour of the
ones provided by ISO C90 Amd. 1 and C99.

21 years agoo Don't include <nl_types.h>, instead provide a typedef for nl_item.
Mike Barcroft [Wed, 18 Sep 2002 05:54:25 +0000 (05:54 +0000)]
o Don't include <nl_types.h>, instead provide a typedef for nl_item.
o Use relatively new visibility primitives to conditionalize some
  constants.

21 years agoMove definition of nl_item type to <sys/_types.h>, so that it can be
Mike Barcroft [Wed, 18 Sep 2002 05:51:23 +0000 (05:51 +0000)]
Move definition of nl_item type to <sys/_types.h>, so that it can be
shared.

21 years agoMostly rewrite the flag handling of uname(1) to be overridable, such as:
Juli Mallett [Wed, 18 Sep 2002 04:33:10 +0000 (04:33 +0000)]
Mostly rewrite the flag handling of uname(1) to be overridable, such as:
  freefall% uname -s
  FreeBSD
  freefall% env UNAME_s=Linnex ./uname -s
  Linnex

MFC after: 1 month

21 years agoMake obrien happy. Add a bad awk script which emulates as much of
Garrett Wollman [Wed, 18 Sep 2002 03:15:25 +0000 (03:15 +0000)]
Make obrien happy.  Add a bad awk script which emulates as much of
gperf's behavior as we ever actually needed here.  This generates
a much-less-efficient keyword recognizer, but it's not like that matters
in this application.  Makefile changes coming once this passes the world
test.

21 years agoUse relatively new visibility primitives for conditionals. Document
Mike Barcroft [Wed, 18 Sep 2002 02:07:08 +0000 (02:07 +0000)]
Use relatively new visibility primitives for conditionals.  Document
unimplemented functions.

21 years agoAdd a toggle to disable VM enforcement.
Robert Watson [Wed, 18 Sep 2002 02:02:08 +0000 (02:02 +0000)]
Add a toggle to disable VM enforcement.

Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs

21 years agoAt the cost of seeming a little gauche, make use of more traditional
Robert Watson [Wed, 18 Sep 2002 02:00:19 +0000 (02:00 +0000)]
At the cost of seeming a little gauche, make use of more traditional
alphabetization for mac_enforce_pipe sysctl.

Obtained from: TrustedBSD Project
Sponsored by: DAPRA, NAI Labs

21 years agoDon't call VOP_LEASE() while holding the accounting mutex.
Robert Watson [Wed, 18 Sep 2002 01:56:13 +0000 (01:56 +0000)]
Don't call VOP_LEASE() while holding the accounting mutex.

21 years agoMFC: Elan support;
Sam Leffler [Tue, 17 Sep 2002 22:39:57 +0000 (22:39 +0000)]
MFC: Elan support;

o Recognize AMD Elan SC520 hostbridge.
o Add initialization code for the AMD Elan sc520 which maps the MMCR
  into KVM and sets the i8254 frequency to the correct value.
o MAKEDEV entry for elan-mmcr device

Reviewed by: phk, jhb
Approved by: re(jhb)

21 years agostyle(9): line up function names
Mike Barcroft [Tue, 17 Sep 2002 22:39:43 +0000 (22:39 +0000)]
style(9): line up function names

21 years agoMove common use of if (DEBUG(FOO)) printf... to DEBUGF(FOO, ...), using
Juli Mallett [Tue, 17 Sep 2002 22:31:26 +0000 (22:31 +0000)]
Move common use of if (DEBUG(FOO)) printf... to DEBUGF(FOO, ...), using
variable length arguments to a macro.  Bump version as this makes DEBUG
statements *always* go to stderr rather than sometimes stdout.  There are
a few stragglers, which I will take care of as soon as I can.  Mostly these
relate to the need-for-death-of some of the remote job code.

Nearby stylistic nits and XXX added/fixed where appropriate.

21 years agoUse the relatively new visibility primitives for conditionals.
Mike Barcroft [Tue, 17 Sep 2002 22:25:40 +0000 (22:25 +0000)]
Use the relatively new visibility primitives for conditionals.

21 years agoo Add typedefs for mode_t, off_t, pid_t rather than including
Mike Barcroft [Tue, 17 Sep 2002 22:22:50 +0000 (22:22 +0000)]
o Add typedefs for mode_t, off_t, pid_t rather than including
  <sys/types.h>.
o Use the relatively new visibility primitives for conditionals.
o Make O_SYNC an alias for O_FSYNC.
o Mark the F* names as deprecated.
o Add some comments to note missing POSIX requirements or options.

21 years agoGuido reported an interesting bug where an FTP connection between a
Matthew Dillon [Tue, 17 Sep 2002 22:21:37 +0000 (22:21 +0000)]
Guido reported an interesting bug where an FTP connection between a
Windows 2000 box and a FreeBSD box could stall.  The problem turned out
to be a timestamp reply bug in the W2K TCP stack.  FreeBSD sends a
timestamp with the SYN, W2K returns a timestamp of 0 in the SYN+ACK
causing FreeBSD to calculate an insane SRTT and RTT, resulting in
a maximal retransmit timeout (60 seconds).  If there is any packet
loss on the connection for the first six or so packets the retransmit
case may be hit (the window will still be too small for fast-retransmit),
causing a 60+ second pause.  The W2K box gives up and closes the
connection.

This commit works around the W2K bug.

15:04:59.374588 FREEBSD.20 > W2K.1036: S 1420807004:1420807004(0) win 65535 <mss 1460,nop,wscale 2,nop,nop,timestamp 188297344 0> (DF) [tos 0x8]
15:04:59.377558 W2K.1036 > FREEBSD.20: S 4134611565:4134611565(0) ack 1420807005 win 17520 <mss 1460,nop,wscale 0,nop,nop,timestamp 0 0> (DF)

Bug reported by: Guido van Rooij <guido@gvr.org>

21 years ago#define<space> -> #define<tab>
Juli Mallett [Tue, 17 Sep 2002 21:29:06 +0000 (21:29 +0000)]
#define<space> -> #define<tab>

21 years agoPrint debugging from DEBUG(VAR) level to stderr. About to macroise this.
Juli Mallett [Tue, 17 Sep 2002 21:26:57 +0000 (21:26 +0000)]
Print debugging from DEBUG(VAR) level to stderr.  About to macroise this.
(Read: More commits to come).

MFC in: 4 days.

21 years agoFixed editing error in previous commit (*blush*).
Bruce Evans [Tue, 17 Sep 2002 19:43:53 +0000 (19:43 +0000)]
Fixed editing error in previous commit (*blush*).

21 years agoUse consistent terminology in the usage message, the man page
John Polstra [Tue, 17 Sep 2002 18:10:39 +0000 (18:10 +0000)]
Use consistent terminology in the usage message, the man page
synopsis, and the man page description ("selector" vs. "sel" and
"addr" vs. "reg").

Fix the usage message and man page synopsis to show that the "value"
argument is not optional.

21 years agoDon't depend on <sys/types.h> pollution in <sys/fcntl.h>.
Mike Barcroft [Tue, 17 Sep 2002 17:59:51 +0000 (17:59 +0000)]
Don't depend on <sys/types.h> pollution in <sys/fcntl.h>.

21 years agoRename new PLONG type to PGTOK as the conversion is more important than the
Juli Mallett [Tue, 17 Sep 2002 17:13:52 +0000 (17:13 +0000)]
Rename new PLONG type to PGTOK as the conversion is more important than the
size (which is mostly undefined anyway).

Submitted by: bde

21 years agoFix standard kse breakge of non-x86 platforms. sigh.
Jake Burkholder [Tue, 17 Sep 2002 16:21:48 +0000 (16:21 +0000)]
Fix standard kse breakge of non-x86 platforms.  sigh.

Pointy hat to: kse

21 years agoCheck for FD_SET overruns.
Jacques Vidrine [Tue, 17 Sep 2002 15:52:35 +0000 (15:52 +0000)]
Check for FD_SET overruns.

Approved by: peter

21 years agoOnly include parent directory Makefile.inc if it exists. Now you can
John Baldwin [Tue, 17 Sep 2002 14:56:23 +0000 (14:56 +0000)]
Only include parent directory Makefile.inc if it exists.  Now you can
cvs co acpi and build acpi utilities w/o needing the usr.sbin Makefile.inc.

21 years agoFixed unsorting of SRCS.
Bruce Evans [Tue, 17 Sep 2002 14:33:23 +0000 (14:33 +0000)]
Fixed unsorting of SRCS.

21 years ago- increment interface output counter. sync w/ netbsd-current
Hajimu UMEMOTO [Tue, 17 Sep 2002 14:25:19 +0000 (14:25 +0000)]
- increment interface output counter.  sync w/ netbsd-current
- increase if_oerrors.  sync w/netbsd

Obtained from: KAME

21 years agoFirewire device support for Apple eMac with PPC kernel.
Katsushi Kobayashi [Tue, 17 Sep 2002 12:50:43 +0000 (12:50 +0000)]
Firewire device support for Apple eMac with PPC kernel.
Contributed by Peter Grehan <grehan@freebsd.org>

21 years agoRegen:
Sheldon Hearn [Tue, 17 Sep 2002 11:47:53 +0000 (11:47 +0000)]
Regen:

* Hart: rev 332 of pcidevs.txt (2002-09-09)

21 years agoAdd /dev/soekris-errled device to control the Error-LED on Soekris cards/boxes.
Poul-Henning Kamp [Tue, 17 Sep 2002 11:47:38 +0000 (11:47 +0000)]
Add /dev/soekris-errled device to control the Error-LED on Soekris cards/boxes.

    # turn LED off
    echo '0' > /dev/soekris-errled

    # turn LED on
    echo '1' > /dev/soekris-errled

    # flash LED (5 hz)
    echo 'f' > /dev/soekris-errled

    # flash LED (4/2 = 2 hz), syntax: "f[1-9]" -> .5 -> 4.5 Hz
    echo 'f4' > /dev/soekris-errled

    # flash digits 1,3 and 7, syntax: "d[1-9]*"
    echo 'd137' > /dev/soekris-errled

Characters not understood are ignored.

21 years agoRemove __RCSID().
Maxim Sobolev [Tue, 17 Sep 2002 11:31:41 +0000 (11:31 +0000)]
Remove __RCSID().

Submitted by: bde

21 years agoFix vsnprintf(3) memory leak for size == 0.
Maxim Konovalov [Tue, 17 Sep 2002 11:28:24 +0000 (11:28 +0000)]
Fix vsnprintf(3) memory leak for size == 0.

PR:             bin/36175
Obtained from:  OpenBSD
Reviewed by:    silence on -audit
MFC after:      5 days

21 years agoExplicitly clear M_FRAG flag on a mbuf with the last fragment to unbreak
Maxim Konovalov [Tue, 17 Sep 2002 11:20:02 +0000 (11:20 +0000)]
Explicitly clear M_FRAG flag on a mbuf with the last fragment to unbreak
ip fragments reassembling for loopback interface.

Discussed with: bde, jlemon
Reviewed by: silence on -net
MFC after: 2 weeks

21 years agoIn rare cases when there is no room for ip options ip_insertoptions()
Maxim Konovalov [Tue, 17 Sep 2002 11:13:04 +0000 (11:13 +0000)]
In rare cases when there is no room for ip options ip_insertoptions()
can fail and corrupt a header length. Initialize len and check what
ip_insertoptions() returns.

Reviewed by: archie, silence on -net
MFC after: 5 days

21 years agoDon't reference cpu_fxsr unless CPU_ENABLE_SSE is defined. This fixes kernel
Maxim Sobolev [Tue, 17 Sep 2002 11:12:10 +0000 (11:12 +0000)]
Don't reference cpu_fxsr unless CPU_ENABLE_SSE is defined. This fixes kernel
in !CPU_ENABLE_SSE case.

21 years agoreject private IPv4 addrs
Hajimu UMEMOTO [Tue, 17 Sep 2002 10:56:52 +0000 (10:56 +0000)]
reject private IPv4 addrs

Obtained from: KAME

21 years ago- reject SIOCSIFADDR if embedded address is in private address range
Hajimu UMEMOTO [Tue, 17 Sep 2002 10:45:51 +0000 (10:45 +0000)]
- reject SIOCSIFADDR if embedded address is in private address range
- reject packets from private address range.  from hitachi

Obtained from: KAME

21 years agoRe-axe.
Bruce Evans [Tue, 17 Sep 2002 10:03:24 +0000 (10:03 +0000)]
Re-axe.

21 years agoMake `as' compile before it is axed. It still uses the archaic BSD
Bruce Evans [Tue, 17 Sep 2002 10:01:02 +0000 (10:01 +0000)]
Make `as' compile before it is axed.  It still uses the archaic BSD
interface setbuffer(), and emulates setbuffer() on USG systems using a
#define of setbuffer() in terms of setvbuf().  The #define is correctly
ifdefed in some places but was not correctly ifdefed here -- i.e., BSD
was essentially configured as USG here.  This became fatal when <stdio.h>
was de-__P(())ified without testing.  This file gets included before
<stdio.h>, so the #define now affects (and breaks)
`setbuffer<left parentheses>' in <stdio.h> where it didn't affect
`setbuffer<whitespace>'.

21 years agoUn-axe this so that it can be fixed before it is axed.
Bruce Evans [Tue, 17 Sep 2002 09:57:07 +0000 (09:57 +0000)]
Un-axe this so that it can be fixed before it is axed.

21 years agoThe previous commit also removed the addition of userland include files to
Peter Wemm [Tue, 17 Sep 2002 09:07:06 +0000 (09:07 +0000)]
The previous commit also removed the addition of userland include files to
the kernel #include path.  While this was not ready at the time (sorry
folks!), it is a good thing.  I think all the loose ends have been tied up
on at least for i386 (LINT compiles) and alpha.

21 years agoMake netatm/spans compile in the kernel without depending on userland
Peter Wemm [Tue, 17 Sep 2002 08:57:52 +0000 (08:57 +0000)]
Make netatm/spans compile in the kernel without depending on userland
include files to provide functions for kernel source (spans_kxdr.c)

21 years agoUpdate man page to reflect current implementation.
Don Lewis [Tue, 17 Sep 2002 08:31:43 +0000 (08:31 +0000)]
Update man page to reflect current implementation.

List all of the error returns shown in the example code in the ERRORS
section.

21 years agoFix a hard sentence break.
Don Lewis [Tue, 17 Sep 2002 08:24:15 +0000 (08:24 +0000)]
Fix a hard sentence break.

21 years agoAdd stub function for cpu_set_upcall_kse()
Peter Wemm [Tue, 17 Sep 2002 07:49:40 +0000 (07:49 +0000)]
Add stub function for  cpu_set_upcall_kse()

21 years ago#if 0 out the following functions:
Peter Wemm [Tue, 17 Sep 2002 07:47:10 +0000 (07:47 +0000)]
#if 0 out the following functions:
  cpu_export_context()
  cpu_set_args()
  cpu_free_kse_mdstorage()
  cpu_export_context()

21 years agoMake this compile after the last kse commit.
Peter Wemm [Tue, 17 Sep 2002 07:44:28 +0000 (07:44 +0000)]
Make this compile after the last kse commit.

21 years agoArgh. I've been reading makefiles for too long. Change comment to a
Peter Wemm [Tue, 17 Sep 2002 07:41:30 +0000 (07:41 +0000)]
Argh.  I've been reading makefiles for too long.  Change comment to a
C-style comment.

21 years agoStub out the calls to get_mcontext and set_mcontext which only exist on
Peter Wemm [Tue, 17 Sep 2002 07:40:15 +0000 (07:40 +0000)]
Stub out the calls to get_mcontext and set_mcontext which only exist on
i386.  This stuff should not be prototyped in MD inludes if the interface
is expected to be MI.

21 years agoAdd i386 to the list of architectures that libc_r is broken on. This
Peter Wemm [Tue, 17 Sep 2002 07:32:36 +0000 (07:32 +0000)]
Add i386 to the list of architectures that libc_r is broken on.  This
effectively removes pppctl from the build for now.  It only compiles on
alpha now (now ironic).

21 years agoIt seems I accidently committed a change that removed the userland includes
Peter Wemm [Tue, 17 Sep 2002 07:22:23 +0000 (07:22 +0000)]
It seems I accidently committed a change that removed the userland includes
from the kernel build.  This broke linux_genassym on the alpha.  For the
kernel, the correct place to get offsetof() is not in /usr/include/stddef.h
but rather <sys/types.h>

21 years agoBandaid to stop failing on non-i386 platforms.
Peter Wemm [Tue, 17 Sep 2002 06:22:51 +0000 (06:22 +0000)]
Bandaid to stop failing on non-i386 platforms.
Add a big ugly #warning as a reminder.

21 years agoMake _fetch_connect() always set the error code.
Bill Fenner [Tue, 17 Sep 2002 05:54:33 +0000 (05:54 +0000)]
Make _fetch_connect() always set the error code.
Tell ftp that _fetch_connect() always sets the error code (http already knew)

21 years agoInclude <sys/types.h> directly rather than depending on <sys/fcntl.h>
Mike Barcroft [Tue, 17 Sep 2002 05:14:09 +0000 (05:14 +0000)]
Include <sys/types.h> directly rather than depending on <sys/fcntl.h>
to include it.  This could be avoided by adding the necessary typedefs
here, or by making users of <sys/file.h> include <sys/types.h> first.

21 years agopid_t will be shared with at least <fcntl.h>.
Mike Barcroft [Tue, 17 Sep 2002 05:05:14 +0000 (05:05 +0000)]
pid_t will be shared with at least <fcntl.h>.

21 years agoBmake bits for GCC 3.2.1-prerelease libstdc++.
Alexander Kabaev [Tue, 17 Sep 2002 04:23:29 +0000 (04:23 +0000)]
Bmake bits for GCC 3.2.1-prerelease libstdc++.
Connect two new source and one header file to the build.

Approved by: obrien

21 years agoMerge FreeBSD modifications into gcc 3.2.1-prerelease:
Alexander Kabaev [Tue, 17 Sep 2002 04:16:58 +0000 (04:16 +0000)]
Merge FreeBSD modifications into gcc 3.2.1-prerelease:
  1.8  printf format error fixes

Approved by: obrien

21 years agoMerge FreeBSD modifications into gcc 3.2.1-prerelease:
Alexander Kabaev [Tue, 17 Sep 2002 04:15:43 +0000 (04:15 +0000)]
Merge FreeBSD modifications into gcc 3.2.1-prerelease:
  1.2  Localize the version number

Approved by: obrien

21 years agoUse the stock 3.2.1-prerelease file.
Alexander Kabaev [Tue, 17 Sep 2002 04:14:38 +0000 (04:14 +0000)]
Use the stock 3.2.1-prerelease file.

Approved by: obrien

21 years agoMerge FreeBSD modifications into gcc 3.2.1-prerelease:
Alexander Kabaev [Tue, 17 Sep 2002 04:10:25 +0000 (04:10 +0000)]
Merge FreeBSD modifications into gcc 3.2.1-prerelease:
  1.2  framework for our kernel printf enhancements
  1.3  suppress prototype warning for K&R main

Approved by: obrien

21 years agoThis commit was generated by cvs2svn to compensate for changes in r103449,
Alexander Kabaev [Tue, 17 Sep 2002 04:04:15 +0000 (04:04 +0000)]
This commit was generated by cvs2svn to compensate for changes in r103449,
which included commits to RCS files with non-trunk default branches.

21 years agoGcc 3.2.1-prerelease libf2c bits from the FSF anoncvs repo gcc-3_2-branch on 16-Sep...
Alexander Kabaev [Tue, 17 Sep 2002 04:04:15 +0000 (04:04 +0000)]
Gcc 3.2.1-prerelease libf2c bits from the FSF anoncvs repo gcc-3_2-branch on 16-Sep-2002 13:23:11 EDT.

21 years agoThis commit was generated by cvs2svn to compensate for changes in r103447,
Alexander Kabaev [Tue, 17 Sep 2002 04:04:04 +0000 (04:04 +0000)]
This commit was generated by cvs2svn to compensate for changes in r103447,
which included commits to RCS files with non-trunk default branches.

21 years agoGcc 3.2.1-prerelease C++ support bits from the FSF anoncvs repo gcc-3_2-branch on...
Alexander Kabaev [Tue, 17 Sep 2002 04:04:04 +0000 (04:04 +0000)]
Gcc 3.2.1-prerelease C++ support bits from the FSF anoncvs repo gcc-3_2-branch on 16-Sep-2002 13:23:11 EDT.

21 years agoThis commit was generated by cvs2svn to compensate for changes in r103445,
Alexander Kabaev [Tue, 17 Sep 2002 04:03:37 +0000 (04:03 +0000)]
This commit was generated by cvs2svn to compensate for changes in r103445,
which included commits to RCS files with non-trunk default branches.

21 years agoGcc 3.2.1-prerelease from the FSF anoncvs repo gcc-3_2-branch on 16-Sep-2002 13:23...
Alexander Kabaev [Tue, 17 Sep 2002 04:03:37 +0000 (04:03 +0000)]
Gcc 3.2.1-prerelease from the FSF anoncvs repo gcc-3_2-branch on 16-Sep-2002 13:23:11 EDT.

21 years agoTempary fix for inet6. The final fix is to change in6_pcbnotify to take pcbinfo instead
Jennifer Yang [Tue, 17 Sep 2002 03:19:43 +0000 (03:19 +0000)]
Tempary fix for inet6. The final fix is to change in6_pcbnotify to take pcbinfo instead
of pcbhead. It is on the way.

21 years agoUse a weak symbol for signalcontext().
Jonathan Mini [Tue, 17 Sep 2002 03:15:47 +0000 (03:15 +0000)]
Use a weak symbol for signalcontext().

21 years agoAdd signalcontext(), which adds a signal frame to a ucontext_t.
Jonathan Mini [Tue, 17 Sep 2002 03:11:07 +0000 (03:11 +0000)]
Add signalcontext(), which adds a signal frame to a ucontext_t.

21 years agoConglomerate printing of ps_pgtok'd data into a PLONG type. I couldn't think
Juli Mallett [Tue, 17 Sep 2002 02:52:44 +0000 (02:52 +0000)]
Conglomerate printing of ps_pgtok'd data into a PLONG type.  I couldn't think
of a better name, except PINT, but I decided to go with assuming LONG to
be safe, rather than assuming INT.

21 years agoBANG! BANG! BANG! Put these bits out of their misery.
David E. O'Brien [Tue, 17 Sep 2002 01:54:56 +0000 (01:54 +0000)]
BANG! BANG! BANG!  Put these bits out of their misery.
Murdered by members of: a.out.die.die.die

ELF is the 1 true path now.  So make good on the src/Makefile threat that
building a.out will not be supported post 4.x.  These bits should either
resurface as a port, or a new port using the latest Binutils bits.  The
later will not support our SunOS-style shared a.out libs; but we shouldn't
need such support by this point in time.

21 years agoInitiate deorbit burn for the i386-only a.out related support. Moves are
Peter Wemm [Tue, 17 Sep 2002 01:49:00 +0000 (01:49 +0000)]
Initiate deorbit burn for the i386-only a.out related support.  Moves are
under way to move the remnants of the a.out toolchain to ports.  As the
comment in src/Makefile said, this stuff is deprecated and one should not
expect this to remain beyond 4.0-REL.  It has already lasted WAY beyond
that.

Notable exceptions:
gcc - I have not touched the a.out generation stuff there.
ldd/ldconfig - still have some code to interface with a.out rtld.
old as/ld/etc - I have not removed these yet, pending their move to ports.
some includes - necessary for ldd/ldconfig for now.

Tested on: i386 (extensively), alpha

21 years agoDisable building of the a.out gas and ld. They will become ports.
David E. O'Brien [Tue, 17 Sep 2002 01:43:21 +0000 (01:43 +0000)]
Disable building of the a.out gas and ld.  They will become ports.