]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoAssume a finger of regular width when no width value is reported by
philip [Sun, 8 Aug 2004 01:26:00 +0000 (01:26 +0000)]
Assume a finger of regular width when no width value is reported by
the touchpad (which happens when it has no extended capabilities).

Spotted by: dhw
Forgotten by: philip

20 years agoAdd support AOpen VA1000 video capture card
sanpei [Sun, 8 Aug 2004 01:23:39 +0000 (01:23 +0000)]
Add support AOpen VA1000 video capture card

PR: kern/62146
Submitted by: HASHI Hiroaki <hasi@cskk-sv.co.jp>
MFC after: 1 week

20 years agoUse 'int' for certain wide-character conversions instead of wchar_t.
kientzle [Sun, 8 Aug 2004 01:21:10 +0000 (01:21 +0000)]
Use 'int' for certain wide-character conversions instead of wchar_t.
That quiets some compiler warnings on platforms with 16-bit wchar_t.
With this change, libarchive now compiles cleanly on Win32/cygwin.

20 years agoUpdate support for Synaptics Touchpads (Volume IV)
philip [Sun, 8 Aug 2004 01:10:23 +0000 (01:10 +0000)]
Update support for Synaptics Touchpads (Volume IV)

 o Change the motion calculation to result in
   a more reasonable speed of motion

This should fix the 'aiming' problems people have reported.  It also
mitigates (but doesn't completely solve) the 'stalling' problems at
very low speeds.

Tested by: many subscribers to -current
Approved by: njl

20 years agoUpdate support for Synaptics Touchpads (Volume III)
philip [Sun, 8 Aug 2004 01:00:31 +0000 (01:00 +0000)]
Update support for Synaptics Touchpads (Volume III)

 o Catch 'taps' as button presses

 o One finger sends button1, two fingers send button3,
   three fingers send button2 (double-click)

Tested by: many subscribers to -current
Approved by: njl

20 years agoUpdate support for Synaptics Touchpads (Volume II)
philip [Sun, 8 Aug 2004 00:57:07 +0000 (00:57 +0000)]
Update support for Synaptics Touchpads (Volume II)

 o Handle the 'up/down' buttons some touchpads have as
   a z-axis (scrollwheel) as recommended by the specs

 o Report the buttons as button4 and button5 instead
   of button2 and button4, button2 can be emulated by
   pressing button1 and button3 simultaneously.  This
   allows one to use the two extra buttons for other
   purposes if one so desires.

Tested by: many subscribers to -current
Approved by: njl

20 years agoUpdate support for Synaptics Touchpads (Volume I)
philip [Sun, 8 Aug 2004 00:52:11 +0000 (00:52 +0000)]
Update support for Synaptics Touchpads (Volume I)

 o Clean up whitespace and comments in the
   enable_synaptics() probing function

 o Only use (and rely on) the extended capability
   bits when we are told they actually exist

 o Partly ignore the (possibly dated?) part of the
   specification about the mode byte so that we
   can support 'guest devices' too.

Tested by: many subscribers to -current
Approved by: njl

20 years agoThe configure-built Makefile wasn't installing archive.h
kientzle [Sun, 8 Aug 2004 00:48:39 +0000 (00:48 +0000)]
The configure-built Makefile wasn't installing archive.h

20 years agoIf you don't need a header, don't bother including it.
kientzle [Sun, 8 Aug 2004 00:43:21 +0000 (00:43 +0000)]
If you don't need a header, don't bother including it.

20 years agoWith the advent of pmap locking it makes sense for pmap_copy() to be less
alc [Sun, 8 Aug 2004 00:31:58 +0000 (00:31 +0000)]
With the advent of pmap locking it makes sense for pmap_copy() to be less
forgiving about inconsistencies in the source pmap.  Also, remove a new-
line character terminating a nearby panic string.

20 years agoImplement single stepping when we leave the kernel through the EPC syscall
marcel [Sun, 8 Aug 2004 00:28:07 +0000 (00:28 +0000)]
Implement single stepping when we leave the kernel through the EPC syscall
path. The basic problem is that we cannot set the single stepping flag
directly, because we don't leave the kernel via an interrupt return. So,
we need another way to set the single stepping flag.
The way we do this is by enabling the lower-privilege transfer trap, which
gets raised when we drop the privilege level. However, since we're still
running in kernel space (sec), we're not yet done. We clear the lower-
privilege transfer trap, enable the taken-branch trap and continue exiting
the kernel until we branch into user space.
Given the current code, there's a total of two traps this way before
we can raise SIGTRAP.

20 years agoAdd support for Belkin F5U409 serial-usb adapter
sanpei [Sat, 7 Aug 2004 23:45:04 +0000 (23:45 +0000)]
Add support for Belkin F5U409 serial-usb adapter

PR:              kern/69804
Submitted by:    Dan Nelson <dnelson@allantgroup.com>

20 years agoAdd man pages for the cimag(), conj() and creal() functions.
stefanf [Sat, 7 Aug 2004 23:03:36 +0000 (23:03 +0000)]
Add man pages for the cimag(), conj() and creal() functions.

20 years agoOh yeah? 4 born in the Netherlands. Taken relative to the size of
marcel [Sat, 7 Aug 2004 22:09:26 +0000 (22:09 +0000)]
Oh yeah? 4 born in the Netherlands. Taken relative to the size of
the country (i.e. committer density), we're in the top 5 for sure! :-)

20 years agoPush all changes to disk before downgrading a mount from rw to ro.
phk [Sat, 7 Aug 2004 22:05:12 +0000 (22:05 +0000)]
Push all changes to disk before downgrading a mount from rw to ro.

20 years agoSlightly move labels around to make sure we call ast() on our way out
marcel [Sat, 7 Aug 2004 21:55:15 +0000 (21:55 +0000)]
Slightly move labels around to make sure we call ast() on our way out
after a fork(2) in fork_trampoline(). By moving the epc_syscall_return
label immediately before the call to do_ast() in epc_syscall(), we not
only achieve that but also handle the detour through exception_return
when the frame corresponds to an asynchronous kernel entry. Hence, we
simplified fork_trampoline() as a side-effect.

20 years agoDe-inline gdb_cpu_signal() because we need to convert the trap vectors
marcel [Sat, 7 Aug 2004 21:40:52 +0000 (21:40 +0000)]
De-inline gdb_cpu_signal() because we need to convert the trap vectors
related to breakpoints and single stepping into SIGTRAP so gdb(1) knows
why the remote target has stopped. In particular, gdb(1) needs to know
if the reason is something of its own doing.

20 years ago10 born in France now!
nsouch [Sat, 7 Aug 2004 21:14:50 +0000 (21:14 +0000)]
10 born in France now!

20 years agoThe whitespace crusader strikes!
des [Sat, 7 Aug 2004 20:55:53 +0000 (20:55 +0000)]
The whitespace crusader strikes!

20 years agoDon't forget to allocate space for the terminating NUL when converting to
des [Sat, 7 Aug 2004 20:23:50 +0000 (20:23 +0000)]
Don't forget to allocate space for the terminating NUL when converting to
base 64.

PR: misc/70022
Submitted by: Herve Masson <herve-bsdbt@mindstep.com>
MFC after: 3 days

20 years agoAcquire vr lock before entering vr_setcfg() in vr_miibus_statchg(),
rwatson [Sat, 7 Aug 2004 19:40:19 +0000 (19:40 +0000)]
Acquire vr lock before entering vr_setcfg() in vr_miibus_statchg(),
since vr_setcfg() expects it.

Reported by: Mike Bristow <mike@urgle.com>

20 years agoAdd --totals option. Unlike gtar, this reports the bytes actually
kientzle [Sat, 7 Aug 2004 19:25:34 +0000 (19:25 +0000)]
Add --totals option.  Unlike gtar, this reports the bytes actually
written to the archive, not the ones written to the compressor.
For uncompressed archives, these numbers are the same, of course.

20 years agoSplit archive_{read,write}_finish into separate "close" (finish the archive
kientzle [Sat, 7 Aug 2004 19:22:50 +0000 (19:22 +0000)]
Split archive_{read,write}_finish into separate "close" (finish the archive
and close it) and "finish" (destroy the object) functions.   For backwards
compat and simplicity, have "finish" invoke "close" transparently if needed.
This allows clients to close the archive and check end-of-operation
statistics before destroying the object.

20 years agoDon't forget to count the end-of-file padding as part of the finished
kientzle [Sat, 7 Aug 2004 19:21:18 +0000 (19:21 +0000)]
Don't forget to count the end-of-file padding as part of the finished
file size.

20 years agoHaving implemented read support for it, I now know how to document the
kientzle [Sat, 7 Aug 2004 17:24:50 +0000 (17:24 +0000)]
Having implemented read support for it, I now know how to document the
GNU sparse file extension.

20 years agoLinux port corrections:
kientzle [Sat, 7 Aug 2004 17:17:11 +0000 (17:17 +0000)]
Linux port corrections:
   * Add a more reasonable default device for linux
   * Add an autoconf check for the FNM_LEADING_DIR extension
     (which isn't yet obeyed in the code)

20 years agoCorrect an mdoc error; add a sentence about tar's ancestors "tp" and "tap."
kientzle [Sat, 7 Aug 2004 17:15:33 +0000 (17:15 +0000)]
Correct an mdoc error; add a sentence about tar's ancestors "tp" and "tap."

20 years agoRevert unintended commit that came in with 3rd clause removal
imp [Sat, 7 Aug 2004 16:36:46 +0000 (16:36 +0000)]
Revert unintended commit that came in with 3rd clause removal

20 years agoDue to popular demand, hook up geom_vinum to the build.
le [Sat, 7 Aug 2004 16:16:59 +0000 (16:16 +0000)]
Due to popular demand, hook up geom_vinum to the build.

20 years agoAdd a way to force 1:1 mode for libpthread. To do this, define
deischen [Sat, 7 Aug 2004 15:15:38 +0000 (15:15 +0000)]
Add a way to force 1:1 mode for libpthread.  To do this, define
LIBPTHREAD_SYSTEM_SCOPE in the environment.

You can still force libpthread to be built in strictly 1:1 by
adding -DSYSTEM_SCOPE_ONLY to CFLAGS.  This is kept for archs
that don't yet support M:N mode.

Requested by:   rwatson
Reviewed by:    davidxu

20 years agoFix syntax error introduced in the last commit.
le [Sat, 7 Aug 2004 14:37:49 +0000 (14:37 +0000)]
Fix syntax error introduced in the last commit.

Additionally, since this file is written in C and not in Perl,
we need 'continue' instead of 'next'.

20 years agoWe're not yet ready to assert !Giant in kern_fcntl(), as it's called
rwatson [Sat, 7 Aug 2004 14:09:02 +0000 (14:09 +0000)]
We're not yet ready to assert !Giant in kern_fcntl(), as it's called
with Giant from ABI wrappers such as Linux emulation.

Foot shoot off: phk

20 years agoCorrect the last commit so it works in error situations as well.
sos [Sat, 7 Aug 2004 12:49:28 +0000 (12:49 +0000)]
Correct the last commit so it works in error situations as well.

20 years agoAdd altq(9) explaining about the IFQ_* and IFQ_DRV_* macros in if_var.h and
mlaier [Sat, 7 Aug 2004 12:35:56 +0000 (12:35 +0000)]
Add altq(9) explaining about the IFQ_* and IFQ_DRV_* macros in if_var.h and
how to convert drivers.

Obtained from: NetBSD (with changes)
Reviewed by: josef

20 years agoRemove extra "*/".
simokawa [Sat, 7 Aug 2004 10:04:29 +0000 (10:04 +0000)]
Remove extra "*/".

20 years agoAdd another UPDATING entry about recompiling modules implementing network
mlaier [Sat, 7 Aug 2004 09:41:30 +0000 (09:41 +0000)]
Add another UPDATING entry about recompiling modules implementing network
interfaces due to a change in the size of struct ifnet.

20 years agoAdd a "void *if_carp" placeholder to struct ifnet with prospect to bring in
mlaier [Sat, 7 Aug 2004 09:32:04 +0000 (09:32 +0000)]
Add a "void *if_carp" placeholder to struct ifnet with prospect to bring in
the "Common address redundancy protocol" (CARP) during the 5-STABLE cycle.
Hence doing the ABI break now.

Approved by: re (scottl)

20 years agoDocument the remove command.
marcel [Sat, 7 Aug 2004 07:52:31 +0000 (07:52 +0000)]
Document the remove command.

20 years agoImplement a remove command. The remove command iterates the GPT
marcel [Sat, 7 Aug 2004 07:41:37 +0000 (07:41 +0000)]
Implement a remove command. The remove command iterates the GPT
partitions and removes any that matches the pre-conditions. The
options are the same for the add command and are used to select
the partitions to remove.
Currently the remove command without any options deletes all GPT
partitions. This is rather harmful and will need anti-footshooting
measures.

20 years agoAssign the result of getopt() to an int rather than to a char.
stefanf [Sat, 7 Aug 2004 07:05:38 +0000 (07:05 +0000)]
Assign the result of getopt() to an int rather than to a char.

20 years agoFix the calculation of the most negative int64_t value, which
kientzle [Sat, 7 Aug 2004 06:38:40 +0000 (06:38 +0000)]
Fix the calculation of the most negative int64_t value, which
is used on systems that lack C99 headers (such as FreeBSD 4).

20 years agoUse __FBSDID.
marcel [Sat, 7 Aug 2004 06:24:25 +0000 (06:24 +0000)]
Use __FBSDID.

20 years agoMove the definition of M_MEMDESC to a non-optional file. This allows
scottl [Sat, 7 Aug 2004 06:21:37 +0000 (06:21 +0000)]
Move the definition of M_MEMDESC to a non-optional file.  This allows
kernels configurations without the 'mem' device to compile.

20 years agoo Save the partition number (=index) in the internal map. The index
marcel [Sat, 7 Aug 2004 06:10:45 +0000 (06:10 +0000)]
o  Save the partition number (=index) in the internal map. The index
   starts at 1. No index is represented by 0.
o  Change the show command to display the partition number at the expense
   of the partition end columm. We already display the start and size.
o  Enhance the add command to accept the -i option. The -i option allows
   the user to specify which partition number the new partition should
   get.
o  Update the manpage accordingly.

20 years agoRemove dead code. A vm_map's first_free is never NULL (even if the map is
alc [Sat, 7 Aug 2004 05:58:31 +0000 (05:58 +0000)]
Remove dead code.  A vm_map's first_free is never NULL (even if the map is
full).

(This is preparation for an O(log n) implementation of vm_map_findspace().)

Submitted by: Mark W. Krentel

20 years agoIt appears that cdevsw_add/remove were removed quite a while ago, so
imp [Sat, 7 Aug 2004 04:59:16 +0000 (04:59 +0000)]
It appears that cdevsw_add/remove were removed quite a while ago, so
remove them from here.

20 years agoExpand indirect reference to copyright file to the portion of that file
imp [Sat, 7 Aug 2004 04:40:20 +0000 (04:40 +0000)]
Expand indirect reference to copyright file to the portion of that file
that is relevant.

20 years agoEliminate a variable that became unused in the i386 to amd64 conversion.
alc [Sat, 7 Aug 2004 04:31:26 +0000 (04:31 +0000)]
Eliminate a variable that became unused in the i386 to amd64 conversion.

20 years agoPer letter dated July 22, 1999 remove 3rd clause of Berkeley derived software
imp [Sat, 7 Aug 2004 04:28:56 +0000 (04:28 +0000)]
Per letter dated July 22, 1999 remove 3rd clause of Berkeley derived software
(with permission of addtional copyright holders where appropriate)

20 years agoPer letter dated July 22, 1999, delete clause 3 from code directly
imp [Sat, 7 Aug 2004 04:19:37 +0000 (04:19 +0000)]
Per letter dated July 22, 1999, delete clause 3 from code directly
from Berkeley.

20 years agoSince we upgraded compilers, and the kernel build to match, we can't build
imp [Sat, 7 Aug 2004 04:17:03 +0000 (04:17 +0000)]
Since we upgraded compilers, and the kernel build to match, we can't build
new kernels on older userlands.  Document this fact in the entry that talked
about the system upgrade.

20 years agoCreate an EFI partition when the user wants auto defaults. There's
marcel [Sat, 7 Aug 2004 04:03:18 +0000 (04:03 +0000)]
Create an EFI partition when the user wants auto defaults. There's
some confusion as to how large the EFI system partition should be,
but 100MB seems to be either the maximum, the minimum or the default
size, so make the EFI partition 100MB.

20 years agoAdd "make distfile" capabilities to bsdtar, including informational
kientzle [Sat, 7 Aug 2004 03:24:49 +0000 (03:24 +0000)]
Add "make distfile" capabilities to bsdtar, including informational
COPYING file and some conditional compilation cleanups.

20 years agolibarchive now has two complete build systems. The usual "Makefile"
kientzle [Sat, 7 Aug 2004 03:09:28 +0000 (03:09 +0000)]
libarchive now has two complete build systems.  The usual "Makefile"
is present for FreeBSD.  If you "make distfile" on FreeBSD, you will
soon have a tar.gz file suitable for deploying to other systems
(complete with the expected "configure" script, etc).  This latter
relies (at least for now) on the GNU auto??? tools.  (I like autoconf
okay, but someday I hope to write a custom Makefile.in and dispense
with automake, which is somewhat odious.)

As part of this, I've cleaned up some of the conditional
compilation options, added make-foo to construct archive.h dynamically
(it now contains some version constants), and added some useful
informational files.

20 years agoPass the pointy hat, please: Don't blow away the high-order
kientzle [Sat, 7 Aug 2004 02:50:05 +0000 (02:50 +0000)]
Pass the pointy hat, please:  Don't blow away the high-order
mode bits when setting permissions from ACL data.
Thanks to: David Gilbert for first reporting this and
    Jimmy Olgeni for noticing that it only occurred on
    ACL-enabled filesystems.

20 years agoBeef up the "cannot archive this" error message with the
kientzle [Sat, 7 Aug 2004 02:24:20 +0000 (02:24 +0000)]
Beef up the "cannot archive this" error message with the
actual mode that failed, to help track down a bug.

20 years agoo Save pointers to the chunks for root, home, swap, usr, var and tmp in
marcel [Sat, 7 Aug 2004 01:19:54 +0000 (01:19 +0000)]
o  Save pointers to the chunks for root, home, swap, usr, var and tmp in
   global variables. On ia64, save a pointer to the efi chunk as well.
o  At the same time, change checkLabels() to define these globals instead
   of having the caller of checkLabels() pass addresses to variables for
   these. Change the two callers correspondingly.
o  Spent a bit more time adjusting try_auto_label() to prepate for having
   the EFI partition created on ia64.
o  Remove efi_mountpoint(). The EFI chunk is now available without having
   to iterate over the disks and chunks to find it every time we need it.
o  On ia64, now that the root chunk is globally available, set the
   vfs.root.mountfrom tunable in loader.conf. This avoids that one cannot
   boot into FreeBSD after an install. The kernel cannot find the root
   device without a little help...

20 years agoPickup Giant in ath_rx_proc and when handling a beacon miss in order to
sam [Sat, 7 Aug 2004 00:45:05 +0000 (00:45 +0000)]
Pickup Giant in ath_rx_proc and when handling a beacon miss in order to
satisfy the assertion in if_start.

20 years agoAlways isync after a mtmsr. While perhaps not strictly necessary for PSL_EE
grehan [Sat, 7 Aug 2004 00:20:00 +0000 (00:20 +0000)]
Always isync after a mtmsr. While perhaps not strictly necessary for PSL_EE
bit banging according to the OEA, it's better to be conservative than
having to continually audit uses of this inline.

20 years agoAs SLIP directly accesses the tty code from its if_start() routine,
rwatson [Fri, 6 Aug 2004 22:41:13 +0000 (22:41 +0000)]
As SLIP directly accesses the tty code from its if_start() routine,
mark if_sl as IFF_NEEDSGIANT.

20 years agoUse the new prototype for the zone constructor.
cognet [Fri, 6 Aug 2004 22:32:53 +0000 (22:32 +0000)]
Use the new prototype for the zone constructor.

20 years agoFlag a broad range of VFS operations as GIANT_REQUIRED in order to
rwatson [Fri, 6 Aug 2004 22:25:35 +0000 (22:25 +0000)]
Flag a broad range of VFS operations as GIANT_REQUIRED in order to
catch leaking into VFS without Giant.

Inch Giant a little lower in several file descriptor operations on
vnodes to cover only VFS operations that need it, rather than file
flag reading, etc.

20 years agoFix a panic in ata_generic_transaction(). The DMA pointer of the channel
njl [Fri, 6 Aug 2004 22:23:53 +0000 (22:23 +0000)]
Fix a panic in ata_generic_transaction().  The DMA pointer of the channel
was being unconditionally dereferenced but was NULL for PIO requests.
Check the request flags for a DMA transaction before dereferencing.

Reported by: ceri
Tested by: Radek Kozlowski <radek -at- raadradd.com>

20 years agoIn thread_exit(), include more information about the thread/process
rwatson [Fri, 6 Aug 2004 22:06:14 +0000 (22:06 +0000)]
In thread_exit(), include more information about the thread/process
context in the KTR trace record.  In particular, include the same
information as passed for mi_switch() and fork_exit() KTR trace
records.

20 years agoPush UIDINFO_UNLOCK() slightly earlier in chgsbize(), as it's not
rwatson [Fri, 6 Aug 2004 22:04:33 +0000 (22:04 +0000)]
Push UIDINFO_UNLOCK() slightly earlier in chgsbize(), as it's not
needed if we print the local variable version of the limit rather
than the shared version.

20 years agoAvoid acquiring Giant for some common light-weight or already MPSAFE
rwatson [Fri, 6 Aug 2004 22:00:55 +0000 (22:00 +0000)]
Avoid acquiring Giant for some common light-weight or already MPSAFE
fcntl() operations, including:

  F_DUPFD          dup() alias
  F_GETFD          retrieve close-on-exec flag
  F_SETFD          set close-on-exec flag
  F_GETFL          retrieve file descriptor flags

For the remaining fcntl() operations, do acquire Giant, especially
where we call into fo_ioctl() as a result.  We're not yet ready to
push Giant into fo_ioctl().  Once we do, this can all become quite a
bit prettier.

20 years agoGenerate KTR trace records for syscall enter and exit in i386 system
rwatson [Fri, 6 Aug 2004 21:56:26 +0000 (21:56 +0000)]
Generate KTR trace records for syscall enter and exit in i386 system
calls.  Note that the information included is a bit different from the
existing KTR traces generated on powerpc, as I'm primarily interested
in kernel context (thread, syscall #, proc, etc), not the user
arguments to the system call.  Some convergence would be useful here.

20 years agoGenerate KTR trace records for uma_zalloc_arg() and uma_zfree_arg().
rwatson [Fri, 6 Aug 2004 21:52:38 +0000 (21:52 +0000)]
Generate KTR trace records for uma_zalloc_arg() and uma_zfree_arg().
This doesn't trace every event of interest in UMA, but provides
enough basic information to explain lock traces and sleep patterns.

20 years agoCut a KTR record whenever a callout is invoked. Mark whether it runs
rwatson [Fri, 6 Aug 2004 21:49:00 +0000 (21:49 +0000)]
Cut a KTR record whenever a callout is invoked.  Mark whether it runs
with Giant or not, and include the function point so it can be looked
up against the kernel symbol table during trace analysis.

20 years agoAssign KTR flag constants for UMA events and callouts.
rwatson [Fri, 6 Aug 2004 21:46:51 +0000 (21:46 +0000)]
Assign KTR flag constants for UMA events and callouts.

20 years agoWhile we're revisiting old sins, try to clean up the code a little and
des [Fri, 6 Aug 2004 21:35:51 +0000 (21:35 +0000)]
While we're revisiting old sins, try to clean up the code a little and
make it more style(9)ish.

20 years agoMove the source code for these two modules out of src/sys/modules.
des [Fri, 6 Aug 2004 20:37:08 +0000 (20:37 +0000)]
Move the source code for these two modules out of src/sys/modules.

20 years agoI've had 'make release' problems on a large/fast dual processor machine
kensmith [Fri, 6 Aug 2004 20:08:38 +0000 (20:08 +0000)]
I've had 'make release' problems on a large/fast dual processor machine
with doFS.sh consistently dying here because the device didn't exist
in the namespace fast enough after doing the mdconfig.  But the device
did eventually show up.  There have been similar complaints on mailing
lists that might boil down to this being the problem too.

This is obviously a hack, if anyone knows what might cause a delay
between mdconfig running and when the name appears in the /dev namespace
(inside a chroot-ed environment if that matters) I'd be happy to back
this out.

20 years agoNow that make(1) passes variables that were given on the command line
marcel [Fri, 6 Aug 2004 17:07:40 +0000 (17:07 +0000)]
Now that make(1) passes variables that were given on the command line
to subordinate make(1) invocations through MAKEFLAGS, we cannot add
CFLAGS onto the make(1) command line. This will conflict with the
individual makefiles wanting to append to it, which is not respected
when CFLAGS is given on the command line. Hence build breakage.
So, put CFLAGS in the environment instead.

20 years agoFix an off-by-one bug that caused the first character of the buffer to
tjr [Fri, 6 Aug 2004 17:00:09 +0000 (17:00 +0000)]
Fix an off-by-one bug that caused the first character of the buffer to
be uninitialized.

20 years agoDamn, fix the SEE ALSO section.
trhodes [Fri, 6 Aug 2004 16:08:43 +0000 (16:08 +0000)]
Damn, fix the SEE ALSO section.

20 years agoAdd a manual page for the hfa driver.
trhodes [Fri, 6 Aug 2004 16:06:52 +0000 (16:06 +0000)]
Add a manual page for the hfa driver.

Reviewed by: harti

20 years agoDo not use hardware flow control for the moment. There are some issues
marcel [Fri, 6 Aug 2004 15:51:31 +0000 (15:51 +0000)]
Do not use hardware flow control for the moment. There are some issues
with it that need to be understood better before they can be resolved.
This takes time and time is already in short supply.

Reported & tested by: glebius@

20 years agoDon't scare users with a warning about preemption being off when it isn't
jhb [Fri, 6 Aug 2004 15:49:44 +0000 (15:49 +0000)]
Don't scare users with a warning about preemption being off when it isn't
yet safe to have on by default.

20 years agoBump the document date, since the content changed today.
roam [Fri, 6 Aug 2004 15:29:54 +0000 (15:29 +0000)]
Bump the document date, since the content changed today.

Discussed with: ru

20 years agoremove /boot/kernel from the default path.. There is already code that
jmg [Fri, 6 Aug 2004 15:06:06 +0000 (15:06 +0000)]
remove /boot/kernel from the default path..  There is already code that
will prepend the current kernel booting...  This prevents a problem of
loading /boot/kernel's modules when a different kernel has no modules,
but you left your module_load="YES" in loader.conf...

Reviewed by: dcs (minus the help part)

20 years agoFix a case of _SC_CLK_TCK being misspelled as _SC_CLOCK_TCK.
roam [Fri, 6 Aug 2004 14:49:11 +0000 (14:49 +0000)]
Fix a case of _SC_CLK_TCK being misspelled as _SC_CLOCK_TCK.

PR: 69428
Submitted by: Sascha Schneider <suntsu@suntsu.org>
MFC after: 2 weeks

20 years agoAdd the right CFLAGS to show that we have err.h, getaddrinfo and strlcpy.
harti [Fri, 6 Aug 2004 13:43:29 +0000 (13:43 +0000)]
Add the right CFLAGS to show that we have err.h, getaddrinfo and strlcpy.
Add support.c to the list of source files.

20 years agoList some more files to be excluded from the original distribution.
harti [Fri, 6 Aug 2004 13:41:55 +0000 (13:41 +0000)]
List some more files to be excluded from the original distribution.

20 years agoThis commit was generated by cvs2svn to compensate for changes in r133211,
harti [Fri, 6 Aug 2004 13:38:30 +0000 (13:38 +0000)]
This commit was generated by cvs2svn to compensate for changes in r133211,
which included commits to RCS files with non-trunk default branches.

20 years agoVirgin import of bsnmpd 1.7
harti [Fri, 6 Aug 2004 13:38:30 +0000 (13:38 +0000)]
Virgin import of bsnmpd 1.7

20 years agoSpell FTP correctly - in this case, it is used as the name of the protocol,
roam [Fri, 6 Aug 2004 12:56:39 +0000 (12:56 +0000)]
Spell FTP correctly - in this case, it is used as the name of the protocol,
not the program.  Also, bump the document date.

Reminded by: our resident mdoc guard (ru)

20 years agoCross-reference getnameinfo(3), getaddrinfo(3), getipnodebyname(3) and
roam [Fri, 6 Aug 2004 12:02:07 +0000 (12:02 +0000)]
Cross-reference getnameinfo(3), getaddrinfo(3), getipnodebyname(3) and
getipnodebyaddr(3).

PR: 54229
Submitted by: Samuel Tardieu <sam@inf.enst.fr>
MFC after: 2 weeks

20 years agoForced commit to note that the previous fix was actually from:
roam [Fri, 6 Aug 2004 11:39:35 +0000 (11:39 +0000)]
Forced commit to note that the previous fix was actually from:

PR: 68449

20 years agoFix a printf("%b", ..) example.
roam [Fri, 6 Aug 2004 11:24:37 +0000 (11:24 +0000)]
Fix a printf("%b", ..) example.

PR: 68849
Submitted by: Michel Lavondes <fox@vader.aacc.cc.md.us>
MFC after: 1 week

20 years agoAdd Giorgos's description of the ftp-chroot login.conf option.
roam [Fri, 6 Aug 2004 10:43:24 +0000 (10:43 +0000)]
Add Giorgos's description of the ftp-chroot login.conf option.

Reported by: Bill Moran <wmoran@potentialtech.com>
Submitted by: keramida
MFC after: 2 weeks

20 years agoAdd and document kern.geom.stripe.fast_failed sysctl, which shows how
pjd [Fri, 6 Aug 2004 10:19:34 +0000 (10:19 +0000)]
Add and document kern.geom.stripe.fast_failed sysctl, which shows how
many times "fast" mode failed.

20 years agoFields bio_caller[12] should be used by the consumer and fields
pjd [Fri, 6 Aug 2004 10:07:03 +0000 (10:07 +0000)]
Fields bio_caller[12] should be used by the consumer and fields
bio_driver[12] should be used by the provider!

20 years agoBump the document date.
ru [Fri, 6 Aug 2004 10:05:45 +0000 (10:05 +0000)]
Bump the document date.

20 years agoClean up after mechanical renumbering.
ru [Fri, 6 Aug 2004 10:04:06 +0000 (10:04 +0000)]
Clean up after mechanical renumbering.

20 years agoFix I/O leakage. We're cloning bios in g_stripe_start_fast(), but when
pjd [Fri, 6 Aug 2004 09:55:40 +0000 (09:55 +0000)]
Fix I/O leakage. We're cloning bios in g_stripe_start_fast(), but when
something goes wrong while running in "fast" mode, we free all bios and
falling back to "economic" mode. Freeing bios, doesn't mean decrease
bio_children, so bio_inbed couldn't be equal to bio_children and request
was never finished.
Decrease bio_children manually when destroying bios.

Reported by: Sam Lawrance <boris@brooknet.com.au>, simon

20 years agoDo not attempt to clean up data that has not been initialized yet.
roam [Fri, 6 Aug 2004 09:08:33 +0000 (09:08 +0000)]
Do not attempt to clean up data that has not been initialized yet.
This fixes two kernel panics on boot when the xl driver fails to
allocate bus/port/memory resources.

Reviewed by: silence on -net

20 years agoCrypto is now part of the "base" distribution, and the old "release.3"
cperciva [Fri, 6 Aug 2004 08:49:51 +0000 (08:49 +0000)]
Crypto is now part of the "base" distribution, and the old "release.3"
target no longer exists.  Renumber the old release.[4-9] targets to
release.[3-8].

Reminded by: ru

20 years agoCrypto is now part of the base distribution. Remove crypto-install.sh,
cperciva [Fri, 6 Aug 2004 08:42:05 +0000 (08:42 +0000)]
Crypto is now part of the base distribution.  Remove crypto-install.sh,
and add scrypto, skrb5, ssecure into src-install.sh

Reminded by: ru

20 years agoHook up the Hardware Notes to the build again, now that the doc build
simon [Fri, 6 Aug 2004 08:41:22 +0000 (08:41 +0000)]
Hook up the Hardware Notes to the build again, now that the doc build
script has been updated.