]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
19 years agoThis commit was generated by cvs2svn to compensate for changes in r147001,
Dag-Erling Smørgrav [Sun, 5 Jun 2005 15:41:57 +0000 (15:41 +0000)]
This commit was generated by cvs2svn to compensate for changes in r147001,
which included commits to RCS files with non-trunk default branches.

19 years agoVendor import of OpenSSH 4.1p1.
Dag-Erling Smørgrav [Sun, 5 Jun 2005 15:41:57 +0000 (15:41 +0000)]
Vendor import of OpenSSH 4.1p1.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r146998,
Dag-Erling Smørgrav [Sun, 5 Jun 2005 15:40:50 +0000 (15:40 +0000)]
This commit was generated by cvs2svn to compensate for changes in r146998,
which included commits to RCS files with non-trunk default branches.

19 years agoVendor import of OpenSSH 4.0p1.
Dag-Erling Smørgrav [Sun, 5 Jun 2005 15:40:50 +0000 (15:40 +0000)]
Vendor import of OpenSSH 4.0p1.

19 years agoHook ichsmb to build.
Takanori Watanabe [Sun, 5 Jun 2005 12:00:03 +0000 (12:00 +0000)]
Hook ichsmb to build.

19 years agoMake ichsmb loadable.
Takanori Watanabe [Sun, 5 Jun 2005 11:55:29 +0000 (11:55 +0000)]
Make ichsmb loadable.

PR: kern/81912
Submitted by:  nork

19 years agoRemove superfluous breaks.
Marius Strobl [Sun, 5 Jun 2005 10:16:27 +0000 (10:16 +0000)]
Remove superfluous breaks.

19 years agoEliminate an unused field from struct aio_liojob.
Alan Cox [Sun, 5 Jun 2005 05:41:48 +0000 (05:41 +0000)]
Eliminate an unused field from struct aio_liojob.

19 years agoLet kmod.mk create an empty .h file.
Ruslan Ermilov [Sun, 5 Jun 2005 05:30:37 +0000 (05:30 +0000)]
Let kmod.mk create an empty .h file.

19 years agoFix deprecated name, and let kmod.mk create an empty .h file.
Ruslan Ermilov [Sun, 5 Jun 2005 05:28:41 +0000 (05:28 +0000)]
Fix deprecated name, and let kmod.mk create an empty .h file.

19 years agoFix indentation of two comment blocks from the last commit.
Andrew Thompson [Sun, 5 Jun 2005 03:49:23 +0000 (03:49 +0000)]
Fix indentation of two comment blocks from the last commit.

Approved by: mlaier (mentor)

19 years agoBump __FreeBSD_version for if_bridge.
Andrew Thompson [Sun, 5 Jun 2005 03:38:03 +0000 (03:38 +0000)]
Bump __FreeBSD_version for if_bridge.

Approved by: mlaier (mentor)

19 years agoConnect if_bridge to the build.
Andrew Thompson [Sun, 5 Jun 2005 03:32:53 +0000 (03:32 +0000)]
Connect if_bridge to the build.

Approved by: mlaier (mentor)

19 years agoAdd hooks into the networking layer to support if_bridge. This changes struct
Andrew Thompson [Sun, 5 Jun 2005 03:13:13 +0000 (03:13 +0000)]
Add hooks into the networking layer to support if_bridge. This changes struct
ifnet so a buildworld is necessary.

Approved by: mlaier (mentor)
Obtained from: NetBSD

19 years agoAdd if_bridge, which provides more advanced Ethernet bridging and 802.1d
Andrew Thompson [Sun, 5 Jun 2005 02:59:26 +0000 (02:59 +0000)]
Add if_bridge, which provides more advanced Ethernet bridging and 802.1d
spanning tree support.

Based on Jason Wright's bridge driver from OpenBSD, and modified by Jason R.
Thorpe in NetBSD.

Reviewed by: mlaier, bms, green
Silence from: -net
Approved by: mlaier (mentor)
Obtained from: NetBSD

19 years agoAvoid casting from (int *) to (size_t *) in order to fix udf_iconv on amd64.
R. Imura [Sun, 5 Jun 2005 02:09:48 +0000 (02:09 +0000)]
Avoid casting from (int *) to (size_t *) in order to fix udf_iconv on amd64.

Reviewed by: scottl
MFC after: 2 weeks

19 years agoUse the ki_udata field that was recently-added to kinfo_proc, instead of
Garance A Drosehn [Sat, 4 Jun 2005 23:40:09 +0000 (23:40 +0000)]
Use the ki_udata field that was recently-added to kinfo_proc, instead of
stealing ki_spare[0] to hold a pointer.

19 years agoAfter some input from bde@ and rereading the datasheet use a MTX_SPIN
Marius Strobl [Sat, 4 Jun 2005 23:24:50 +0000 (23:24 +0000)]
After some input from bde@ and rereading the datasheet use a MTX_SPIN
mutex instead of a MTX_DEF one in order to defer preemption while
reading the date and time registers. If we don't manage to read them
within the time slot where we are guaranteed that no updates occur we
might actually read them during an update in which case the output is
undefined.

19 years agoRewrite some of the regexps so they don't match themselves.
Dag-Erling Smørgrav [Sat, 4 Jun 2005 23:18:33 +0000 (23:18 +0000)]
Rewrite some of the regexps so they don't match themselves.

19 years agoMinor clean up for flags restoration: Use fchflags/lchflags when
Tim Kientzle [Sat, 4 Jun 2005 22:30:36 +0000 (22:30 +0000)]
Minor clean up for flags restoration: Use fchflags/lchflags when
available, stub out flags restore on platforms that don't support it,
update autoconf to probe for fchflags and lchflags support.

19 years agoSort archive_entry_ctime/archive_entry_ctime_nsec correctly.
Tim Kientzle [Sat, 4 Jun 2005 22:26:23 +0000 (22:26 +0000)]
Sort archive_entry_ctime/archive_entry_ctime_nsec correctly.

19 years agoSupport libarchive with C++ by adding appropriate conditional
Tim Kientzle [Sat, 4 Jun 2005 22:24:04 +0000 (22:24 +0000)]
Support libarchive with C++ by adding appropriate conditional
extern "C" declarations.

19 years agoFix one error in the example usage of the archive_write API and fill
Tim Kientzle [Sat, 4 Jun 2005 22:19:25 +0000 (22:19 +0000)]
Fix one error in the example usage of the archive_write API and fill
in a few missing details.  The example code here is now a complete,
functioning example program.

19 years agoFix getsecs(). It was not counting the seconds right. The immediate
Marcel Moolenaar [Sat, 4 Jun 2005 21:55:51 +0000 (21:55 +0000)]
Fix getsecs(). It was not counting the seconds right. The immediate
and visible effect of the bug what that autoboot would boot a kernel
after only a couple of seconds had passed instead of waiting the
full 10 seconds it's supposed to wait by default.
Add my copyright notice, since one was missing and I reimplemented
the one and only function in this file.

MFC after: 1 week

19 years agoReplace the band-aid for allowing to call sunkbd_configure() multiple
Marius Strobl [Sat, 4 Jun 2005 21:54:31 +0000 (21:54 +0000)]
Replace the band-aid for allowing to call sunkbd_configure() multiple
times which was added in the last revision with what should be a proper
solution as long as keyboards that were pluggged in after the kernel
has fully booted aren't supported. I.e. when sunkbd_configure() is
called for the high-level console probe make sure that the keyboard is
both successfully configured (i.e. also probed) and attached. The band-
aid left the possibility to attach the keyboard device to the high-level
console without attaching the keyboard device itself when the keyboard
is plugged in after uart(4) attached but before syscons(4) does.

19 years agoOn AXi and AXmp boards the NS16550 (used to connect keyboard and mouse)
Marius Strobl [Sat, 4 Jun 2005 21:52:56 +0000 (21:52 +0000)]
On AXi and AXmp boards the NS16550 (used to connect keyboard and mouse)
share their IRQ lines with the i8042. Any IRQ activity (typically during
attach) on the NS16550 used to connect the keyboard when actually the
PS/2 keyboard is selected in OFW causes interaction with the OBP i8042
driver resulting in a hang (and vice versa). As RS232 keyboards and mice
obviously aren't meant to be used in parallel with PS/2 ones on these
boards don't attach to these NS16550 in case the RS232 keyboard isn't
selected in order to prevent such hangs.

Ok'ed by: marcel

19 years agoFix delay(). The processor cycle counter is a 32-bit wrapping counter.
Marcel Moolenaar [Sat, 4 Jun 2005 21:50:44 +0000 (21:50 +0000)]
Fix delay(). The processor cycle counter is a 32-bit wrapping counter.
Hence, mask off the upper 32 bits and deal with wrap-arounds.

MFC after: 1 week

19 years agoChange the semantics of uart_cpu_getdev_keyboard() to only match SCCs/
Marius Strobl [Sat, 4 Jun 2005 21:33:18 +0000 (21:33 +0000)]
Change the semantics of uart_cpu_getdev_keyboard() to only match SCCs/
UARTs used to connect keyboards and not also PS/2 keyboards and only
return their package handle in case the keyboard is the preferred one
according to the OFW but otherwise still regardless of whether the
keyboard is used for stdin or not. This is simply achieved by looking
at the 'keyboard' alias and returning the corresponding package handle
in case it refers to a SCC/UART. This is change is done in order to
give the keyboard which the OFW or the user selected in OFW on boards
that support additional types of keyboards besides the RS232 ones also
preference in FreeBSD. It will be also used to determine on Sun AXi and
Sun AXmp boards whether a PS/2 or a RS232 is to be used as these are
sort of mutual exclusive there (see upcoming commit to uart_bus_ebus.c).
Note that Tatung AXi boards have the same issue but the former code
happened to already give the PS/2 keyboard preference by not identifying
the respective UART as keyboard system device there because the PS/2
keyboard node precedes the keyboard UART one in the OFW device tree of
these boards (which isn't the case for the Sun AXi).

Ok'ed by: marcel

19 years ago- In machfb_configure() when probed for the high-level console return
Marius Strobl [Sat, 4 Jun 2005 21:18:30 +0000 (21:18 +0000)]
- In machfb_configure() when probed for the high-level console return
  the number of registered adapters instead of determining again whether
  stdout is a supported card (and which might have failed to attach and
  register).
- Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant
  to return ENODEV for all invocations expect when used to disable the
  cursor and not just when used for enabling the cursor.
- In case the adapter is the OFW stdout move its OFW cursor to the start
  of the last line on halt so OFW output doesn't get intermixed with what
  FreeBSD left on the screen.
- Drop variable names in the prototypes of some functions in order to
  match the style of majority of the prototypes in this file.

19 years ago- In creator_configure() when probed for the high-level console return
Marius Strobl [Sat, 4 Jun 2005 21:15:27 +0000 (21:15 +0000)]
- In creator_configure() when probed for the high-level console return
  the number of registered adapters instead of determining again whether
  stdout is a supported card (and which might have failed to attach and
  register).
- Drop creator_set_mode() and move the relevant parts to creator_fill_rect()
  and creator_putc() respectively. This is a bit cleaner than having to
  make sure that creator_set_mode() was called before creator_fill_rect()
  or creator_putc() are used and matches better what Xorg does.
- Fix a bug in the handling of the FBIOSCURSOR IOCTL; the code was meant
  to return ENODEV for all invocations expect when used to disable the
  cursor and not just when used for enabling the cursor.
- In case the adapter is the OFW stdout move its OFW cursor to the start
  of the last line on halt so OFW output doesn't get intermixed with what
  FreeBSD left on the screen. With hindsight this is what the faking of a
  hardware cursor which was removed in the last revision really was about,
  i.e. to keep the OFW updated about the current cursor position. The new
  approach however is simpler while producing the same result and doesn't
  cause the first letter of the OFW output to be turned into a blank and
  a newline.
- Add variable names to the prototypes of creator_cursor_*() which were
  added in the last revision and list them alphabetically in order to match
  the style of this file.

19 years agoAdd esp(4) to scsi-controller-regex.
Marius Strobl [Sat, 4 Jun 2005 21:05:37 +0000 (21:05 +0000)]
Add esp(4) to scsi-controller-regex.

MFC after: 1 month

19 years agoMFNetBSD: src/libexec/ld.elf_so/arch/sparc64/mdreloc.c 1.26 (partial)
Marius Strobl [Sat, 4 Jun 2005 20:49:15 +0000 (20:49 +0000)]
MFNetBSD: src/libexec/ld.elf_so/arch/sparc64/mdreloc.c 1.26 (partial)

Fix an obvious bug in the 64-bit PLT fixup: the SLLX was by 12 bits, when
it should be 32.

MFC after: 1 month

19 years agoNow that all affected drivers have been changed remove the helpers
Marius Strobl [Sat, 4 Jun 2005 20:31:20 +0000 (20:31 +0000)]
Now that all affected drivers have been changed remove the helpers
for the SYS_RES_IOPORT -> SYS_RES_MEMORY transition again. While it
was helpful to not need to change all of the affected drivers in a
single pass together with ebus(4) we probably shouldn't start into
6.0 with such a hack.
This requires some of the modules of affected drivers to be rebuilt,
namely: auxio(4), snd_audiocs(4) and puc(4).

19 years agoAccount for ebus(4) defaulting to SYS_RES_MEMORY for memory resources
Marius Strobl [Sat, 4 Jun 2005 20:29:28 +0000 (20:29 +0000)]
Account for ebus(4) defaulting to SYS_RES_MEMORY for memory resources
since ebus.c rev. 1.22.

19 years ago- Take advantage of ebus(4) having switched to SYS_RES_MEMORY for memory
Marius Strobl [Sat, 4 Jun 2005 20:27:39 +0000 (20:27 +0000)]
- Take advantage of ebus(4) having switched to SYS_RES_MEMORY for memory
  resources in ebus.c rev. 1.22 and collapse the resource allocation for
  both the EBus and SBus variants into auxio_attach_common().
- For the EBus variant make sure that the resource for controlling the
  LED is actually available; (in theory) we could have ended up using
  the resource without allocating it.

19 years agoUpdate .Dd value.
Alan Cox [Sat, 4 Jun 2005 19:44:09 +0000 (19:44 +0000)]
Update .Dd value.

19 years agoEliminate the original method of requesting notification of aio_read(2) and
Alan Cox [Sat, 4 Jun 2005 19:16:33 +0000 (19:16 +0000)]
Eliminate the original method of requesting notification of aio_read(2) and
aio_write(2) completion through kevent(2).  This method does not work on
64-bit architectures.  It was deprecated in FreeBSD 4.4.  See revisions
1.87 and 1.70.2.7.

Change aio_physwakeup() to call psignal(9) directly rather than indirectly
through a timeout(9).  Discussed with: bde

Correct a bug introduced in revision 1.65 that could result in premature
delivery of a signal if an lio_listio(2) consisted of a mixture of
direct/raw and queued I/O operations.  Observed by: tegge

Eliminate a field from struct kaioinfo that is now unused.

Reviewed by: tegge

19 years agoBetter explain, then actually implement the IPFW ALTQ-rule first-match
Brian Feldman [Sat, 4 Jun 2005 19:04:31 +0000 (19:04 +0000)]
Better explain, then actually implement the IPFW ALTQ-rule first-match
policy.  It may be used to provide more detailed classification of
traffic without actually having to decide its fate at the time of
classification.

MFC after: 1 week

19 years agoMove MAINTAINER documentation to MAINTAINERS
Warner Losh [Sat, 4 Jun 2005 16:57:04 +0000 (16:57 +0000)]
Move MAINTAINER documentation to MAINTAINERS

19 years agoSince there's no emulation person, move the entry to upper section
Warner Losh [Sat, 4 Jun 2005 16:56:35 +0000 (16:56 +0000)]
Since there's no emulation person, move the entry to upper section

19 years agoAdd the proper logic so that we don't try to do SSE stuff unless its
Doug Rabson [Sat, 4 Jun 2005 15:36:48 +0000 (15:36 +0000)]
Add the proper logic so that we don't try to do SSE stuff unless its
enabled.

19 years agoRelease n_hibma's maintainership due to his ENOTIME.
Jens Schweikhardt [Sat, 4 Jun 2005 10:58:39 +0000 (10:58 +0000)]
Release n_hibma's maintainership due to his ENOTIME.

Discussed with, okayed, authorized, requested by, on behalf of: n_hibma (MAINTAINER)

19 years agoFix the source path.
Ruslan Ermilov [Sat, 4 Jun 2005 10:48:21 +0000 (10:48 +0000)]
Fix the source path.

19 years ago - It's 2005 already, I've been working on this for three years.
Jeff Roberson [Sat, 4 Jun 2005 09:24:15 +0000 (09:24 +0000)]
 - It's 2005 already, I've been working on this for three years.

19 years ago - Don't SLOT_USE() in the preempt case, sched_add() has already taken the
Jeff Roberson [Sat, 4 Jun 2005 09:23:28 +0000 (09:23 +0000)]
 - Don't SLOT_USE() in the preempt case, sched_add() has already taken the
   slot for us.  Previously, we would take two slots on every preempt, and
   setrunqueue() would fix it up for us in the non threaded case.  The
   threaded case was simply broken.
 - Clean up flags, prototypes, comments.

19 years agoChanges to tcp_sack_option() that
Paul Saab [Sat, 4 Jun 2005 08:03:28 +0000 (08:03 +0000)]
Changes to tcp_sack_option() that
- Walks the scoreboard backwards from the tail to reduce the number of
  comparisons for each sack option received.
- Introduce functions to add/remove sack scoreboard elements, making
  the code more readable.

Submitted by:   Noritoshi Demizu
Reviewed by:    Raja Mukerji, Mohan Srinivasan

19 years agoFix module build.
Gleb Smirnoff [Sat, 4 Jun 2005 04:43:14 +0000 (04:43 +0000)]
Fix module build.

MFC after: 3 days

19 years agoWrap copyin/copyout for kevent so the 32bit wrapper does not have
Paul Saab [Fri, 3 Jun 2005 23:15:01 +0000 (23:15 +0000)]
Wrap copyin/copyout for kevent so the 32bit wrapper does not have
to malloc nchanges * sizeof(struct kevent) AND/OR nevents *
sizeof(struct kevent) on every syscall.

Glanced at by: peter, jmg
Obtained from: Yahoo!
MFC after: 2 weeks

19 years agoIn newbus land, ivars can only be accessed for direct child, or when
Warner Losh [Fri, 3 Jun 2005 20:12:12 +0000 (20:12 +0000)]
In newbus land, ivars can only be accessed for direct child, or when
the driver has unholy private knowledge of its great-*cgrandchildren.
The ACPI allocation routine lacked such knowledge when it tried to do
a default allocation for all descendants, rather than just its
immeidate children, so would access grandchild's ivar in an unsafe
way.  This could lead to a panic when devices were present which had
no addresses setup by the BIOS, but which were later allocated in a
lazy manner via pci_alloc_map.  As such, only do the default
allocation adjustments for immediate children.  The manner that
acpi_sysres_find accesses the resource list, used later in
acpi_alloc_resource, is safe and proper so no additional test is
needed there.

This fixes a panic when probing an disabled ata controller on some
newer intel blades.

Reported by: dwhite

19 years agoBring in bits I forgot while importing write back support for arm9.
Olivier Houchard [Fri, 3 Jun 2005 19:49:53 +0000 (19:49 +0000)]
Bring in bits I forgot while importing write back support for arm9.

19 years agoMask off the bar's value after the probe test write before testing
Warner Losh [Fri, 3 Jun 2005 19:41:06 +0000 (19:41 +0000)]
Mask off the bar's value after the probe test write before testing
against 0 in pci_alloc_map, just like we do in pci_add_map.  Also,
make sure that we restore the value to the BAR that was there before
if the bar is 0.  Chances are that it was 0 before the write too and
that the restoration is a nop, but better safe than sorry.

Notice by: dwhite

19 years agoRemove my involvement from the regression test stuff, I don't know the modified
Juli Mallett [Fri, 3 Jun 2005 18:42:49 +0000 (18:42 +0000)]
Remove my involvement from the regression test stuff, I don't know the modified
framework well, and nobody's ever sought my help with writing regression tests,
most people just don't bother with them at all, regrettably.

19 years agoRemove trailing whitespaces.
Hiroki Sato [Fri, 3 Jun 2005 18:42:47 +0000 (18:42 +0000)]
Remove trailing whitespaces.

19 years agoNew release notes:
Hiroki Sato [Fri, 3 Jun 2005 18:37:16 +0000 (18:37 +0000)]
New release notes:
FreeBSD/i386 and /amd64 always use the local APIC timer even
on an UP system,
vm.blacklist tunable added,
ipfw now supports IPv4 only rule,
struct netent and getnetbyaddr(3) POSIX conformance and
        the ABI breakage on 64-bit platforms, and
GNU GCC 3.4.2-prerelease --> 3.4.4.

Update release notes:
more machine-specific optimized version of math functions.

19 years agoUse predefined __offsetof__ builtin function when compiling C++ sources.
Alexander Kabaev [Fri, 3 Jun 2005 17:53:36 +0000 (17:53 +0000)]
Use predefined __offsetof__ builtin function when compiling C++ sources.

19 years ago- Avoid a memory leak if realloc(3) fails by using reallocf(3)
Suleiman Souhlal [Fri, 3 Jun 2005 17:38:33 +0000 (17:38 +0000)]
- Avoid a memory leak if realloc(3) fails by using reallocf(3)

Submitted by: Liam J. Foy <liamfoy@dragonflybsd.org>
Approved by: mdodd (in-lieu of mentor who is away)
MFC after: 1 week

19 years agoEnable handling of MS-compatible packing push and pop pragmas. This change was
Alexander Kabaev [Fri, 3 Jun 2005 17:35:13 +0000 (17:35 +0000)]
Enable handling of MS-compatible packing push and pop pragmas. This change was
inadvertedly left out of the last import.

19 years agoRevert last. It was the result of mismerges from cvs not an actual
Warner Losh [Fri, 3 Jun 2005 15:53:37 +0000 (15:53 +0000)]
Revert last.  It was the result of mismerges from cvs not an actual
need.  _sconfig already was there in the correct place.

Noticed by: ru@

19 years agoMissed rwatson's redundancy
Warner Losh [Fri, 3 Jun 2005 15:50:55 +0000 (15:50 +0000)]
Missed rwatson's redundancy

19 years agoBuild sconfig on i386 to match where drivers are made it configures
Warner Losh [Fri, 3 Jun 2005 15:38:39 +0000 (15:38 +0000)]
Build sconfig on i386 to match where drivers are made it configures

19 years agoUpdate to reflect takawata-san's changes to MAINTAINERS
Warner Losh [Fri, 3 Jun 2005 15:35:36 +0000 (15:35 +0000)]
Update to reflect takawata-san's changes to MAINTAINERS

19 years agoFix standalone module build.
Ruslan Ermilov [Fri, 3 Jun 2005 15:31:10 +0000 (15:31 +0000)]
Fix standalone module build.

Reported by: Boris Samorodov <bsam@ipt.ru>

19 years agoRemove libz and pppd redundancy
Warner Losh [Fri, 3 Jun 2005 15:25:13 +0000 (15:25 +0000)]
Remove libz and pppd redundancy

19 years agoRemove a useless printf.
Olivier Houchard [Fri, 3 Jun 2005 15:15:26 +0000 (15:15 +0000)]
Remove a useless printf.

19 years agoTidy up cvs entries (missed one)
Warner Losh [Fri, 3 Jun 2005 15:12:15 +0000 (15:12 +0000)]
Tidy up cvs entries (missed one)

19 years agoDrop maintainership.
Takanori Watanabe [Fri, 3 Jun 2005 15:10:15 +0000 (15:10 +0000)]
Drop maintainership.

19 years agoTidy up cvs entries
Warner Losh [Fri, 3 Jun 2005 15:08:30 +0000 (15:08 +0000)]
Tidy up cvs entries

19 years agoI've not had time to take care of wi lately...
Warner Losh [Fri, 3 Jun 2005 15:04:09 +0000 (15:04 +0000)]
I've not had time to take care of wi lately...

19 years agoRemove myself as maintainer of the OSF/1 compat code. I haven't had
Andrew Gallatin [Fri, 3 Jun 2005 12:37:56 +0000 (12:37 +0000)]
Remove myself as maintainer of the OSF/1 compat code.  I haven't had
an alpha booting -current since my xp1000 died over a year ago, and I
don't think I touched the osf1 code since long before then.

19 years agoRemove the EXAMPLES section that describes how to sort by size using
Dima Dorfman [Fri, 3 Jun 2005 11:38:35 +0000 (11:38 +0000)]
Remove the EXAMPLES section that describes how to sort by size using
sort(1). This functionality is provided by the -S option now, and it
is useful even though a similar effect is achievable with sort(1),
since the latter doesn't work in combination with -h. This option is
also present in NetBSD, OpenBSD, and GNU fileutils, so there's clearly
a demand for it.

Noticed by: asmodai

19 years agoImprove wording: A sort is "in" a particular order, not "by" a
Dima Dorfman [Fri, 3 Jun 2005 11:22:06 +0000 (11:22 +0000)]
Improve wording: A sort is "in" a particular order, not "by" a
particular order.

19 years agoAdd the -S option to sort files by size. NetBSD and OpenBSD already
Dima Dorfman [Fri, 3 Jun 2005 11:05:58 +0000 (11:05 +0000)]
Add the -S option to sort files by size. NetBSD and OpenBSD already
have this option with identical semantics (sorting large files first).
-r can be used to reverse the sort if that is desired.

PR: 81625
Submitted by: Kostas Blekos <mplekos@physics.upatras.gr>, keramida

19 years agomention libpcap.
Hajimu UMEMOTO [Fri, 3 Jun 2005 10:19:26 +0000 (10:19 +0000)]
mention libpcap.

19 years agobump SHLIB_MAJOR as libpcap refers getnetbyname(3).
Hajimu UMEMOTO [Fri, 3 Jun 2005 09:41:26 +0000 (09:41 +0000)]
bump SHLIB_MAJOR as libpcap refers getnetbyname(3).

Noticed by: ru

19 years agoExplain what's so special about share/mk, and be even more friendly.
Ruslan Ermilov [Fri, 3 Jun 2005 09:26:49 +0000 (09:26 +0000)]
Explain what's so special about share/mk, and be even more friendly.

19 years agofix typo in my previous commit.
Hajimu UMEMOTO [Fri, 3 Jun 2005 09:15:17 +0000 (09:15 +0000)]
fix typo in my previous commit.

Submitted by: Jiawei Ye <leafy7382 _at_ gmail.com>

19 years agoMPPC node is not thread safe.
Gleb Smirnoff [Fri, 3 Jun 2005 09:06:00 +0000 (09:06 +0000)]
MPPC node is not thread safe.

PR: kern/79990
Reported by: Arcadiy Ivanov
Reported by: atckoe.zlo @ gmail.com
MFC after: 1 week

19 years agoRevise my status slightly for the bits that I am concerned with.
Mark Murray [Fri, 3 Jun 2005 07:14:54 +0000 (07:14 +0000)]
Revise my status slightly for the bits that I am concerned with.

19 years agoMatch sys/boot in ensure GCC does not use x86 FP registers in integer code.
David E. O'Brien [Fri, 3 Jun 2005 06:55:22 +0000 (06:55 +0000)]
Match sys/boot in ensure GCC does not use x86 FP registers in integer code.

Submitted by: Pawel Worach <pawel.worach@gmail.com>

19 years agoMake diff(1) follow traditional UNIX SOP of later options nullify earlier
David E. O'Brien [Fri, 3 Jun 2005 06:46:56 +0000 (06:46 +0000)]
Make diff(1) follow traditional UNIX SOP of later options nullify earlier
options that are mutually exclusive.

19 years agoFix a typo from previos commit.
Alexander Kabaev [Fri, 3 Jun 2005 05:58:56 +0000 (05:58 +0000)]
Fix a typo from previos commit.

19 years agoSynchronize access to the per process aiocb lists in many of the functions.
Alan Cox [Fri, 3 Jun 2005 05:27:20 +0000 (05:27 +0000)]
Synchronize access to the per process aiocb lists in many of the functions.

19 years agoreflect type change of n_net and getnetbyaddr(3).
Hajimu UMEMOTO [Fri, 3 Jun 2005 04:44:54 +0000 (04:44 +0000)]
reflect type change of n_net and getnetbyaddr(3).

19 years agoUpdate bmake glue for GCC 3.4.4.
Alexander Kabaev [Fri, 3 Jun 2005 04:21:04 +0000 (04:21 +0000)]
Update bmake glue for GCC 3.4.4.

19 years agoCatch up with changes in BDX_OUTPUT_[LR]BRAC callers.
Alexander Kabaev [Fri, 3 Jun 2005 04:19:23 +0000 (04:19 +0000)]
Catch up with changes in BDX_OUTPUT_[LR]BRAC callers.

19 years agoMerge conflicts for GCC 3.4.4.
Alexander Kabaev [Fri, 3 Jun 2005 04:02:20 +0000 (04:02 +0000)]
Merge conflicts for GCC 3.4.4.

19 years agoLocalize GCC version string for FreeBSD.
Alexander Kabaev [Fri, 3 Jun 2005 03:52:25 +0000 (03:52 +0000)]
Localize GCC version string for FreeBSD.

19 years agoStock files.
Alexander Kabaev [Fri, 3 Jun 2005 03:50:42 +0000 (03:50 +0000)]
Stock files.

19 years agomention getnet*(3) ABI breakage.
Hajimu UMEMOTO [Fri, 3 Jun 2005 03:34:21 +0000 (03:34 +0000)]
mention getnet*(3) ABI breakage.

19 years ago- Remove padding for ABI compatibility of n_net member from struct
Hajimu UMEMOTO [Fri, 3 Jun 2005 03:32:06 +0000 (03:32 +0000)]
- Remove padding for ABI compatibility of n_net member from struct
  netent.
- Change 1st argument of getnetbyaddr() to an uint32_t on 64 bit
  arch as well to confirm to POSIX-2001.

These changes break ABI compatibility on 64 bit arch.
There is similar padding issue for ai_addrlen of struct addrinfo.
However, it is leaved as is for now.

Discussed on: arch@, standards@ and current@
X-MFC after: never

19 years agoThis commit was generated by cvs2svn to compensate for changes in r146901,
Alexander Kabaev [Fri, 3 Jun 2005 03:30:16 +0000 (03:30 +0000)]
This commit was generated by cvs2svn to compensate for changes in r146901,
which included commits to RCS files with non-trunk default branches.

19 years agoGcc 3.4.4 F77 runtime support bits.
Alexander Kabaev [Fri, 3 Jun 2005 03:30:16 +0000 (03:30 +0000)]
Gcc 3.4.4 F77 runtime support bits.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r146899,
Alexander Kabaev [Fri, 3 Jun 2005 03:29:59 +0000 (03:29 +0000)]
This commit was generated by cvs2svn to compensate for changes in r146899,
which included commits to RCS files with non-trunk default branches.

19 years agoGcc 3.4.4 Objective C support bits.
Alexander Kabaev [Fri, 3 Jun 2005 03:29:59 +0000 (03:29 +0000)]
Gcc 3.4.4 Objective C support bits.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r146897,
Alexander Kabaev [Fri, 3 Jun 2005 03:29:38 +0000 (03:29 +0000)]
This commit was generated by cvs2svn to compensate for changes in r146897,
which included commits to RCS files with non-trunk default branches.

19 years agoGcc 3.4.4 C++ support bits.
Alexander Kabaev [Fri, 3 Jun 2005 03:29:38 +0000 (03:29 +0000)]
Gcc 3.4.4 C++ support bits.

19 years agoThis commit was generated by cvs2svn to compensate for changes in r146895,
Alexander Kabaev [Fri, 3 Jun 2005 03:28:44 +0000 (03:28 +0000)]
This commit was generated by cvs2svn to compensate for changes in r146895,
which included commits to RCS files with non-trunk default branches.

19 years agoGcc 3.4.4 release.
Alexander Kabaev [Fri, 3 Jun 2005 03:28:44 +0000 (03:28 +0000)]
Gcc 3.4.4 release.

19 years agoAdd support for IPv4 only rules to IPFW2 now that it supports IPv6 as well.
Max Laier [Fri, 3 Jun 2005 01:10:28 +0000 (01:10 +0000)]
Add support for IPv4 only rules to IPFW2 now that it supports IPv6 as well.
This is the last requirement before we can retire ip6fw.

Reviewed by: dwhite, brooks(earlier version)
Submitted by: dwhite (manpage)
Silence from: -ipfw

19 years agoCorrect typo in a comment describing vshiftl().
Hiten Pandya [Thu, 2 Jun 2005 23:56:10 +0000 (23:56 +0000)]
Correct typo in a comment describing vshiftl().