]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years ago- complete rewrite of the version number parsing code, restoring compatibiliy of...
eik [Tue, 29 Jun 2004 18:52:13 +0000 (18:52 +0000)]
- complete rewrite of the version number parsing code, restoring compatibiliy of 5.x with 4.x and portupgrade

- parse version numbers of ports containing an underscore followed by a number correctly

- handle special strings pl, alpha, beta, pre and rc

PR: 56961

20 years agoReduce the size of pv entries by 15%. This saves 1MB of KVA for mapping
peter [Tue, 29 Jun 2004 15:57:05 +0000 (15:57 +0000)]
Reduce the size of pv entries by 15%.  This saves 1MB of KVA for mapping
pv entries per 1GB of user virtual memory.  (eg: if we had 1GB file was
mmaped into 30 processes, that would theoretically reduce the KVA demand by
30MB for pv entries.  In reality though, we limit pv entries so we don't
have that many at once.)

We used to store the vm_page_t for the page table page.  But we recently
had the pa of the ptp, or can calculate it fairly quickly.  If we wanted
to avoid the shift/mask operation in pmap_pde(), we could recover the
pa but that means we have to store it for a while.

This does not measurably change performance.

Suggested by:  alc
Tested by:  alc

20 years agoRemove 3rd clause from the licence.
marks [Tue, 29 Jun 2004 15:46:12 +0000 (15:46 +0000)]
Remove 3rd clause from the licence.

Approved by: njl

20 years agoUse a default of the FADT for matching CreatorId quirks. Use XSDT to mean
njl [Tue, 29 Jun 2004 14:32:06 +0000 (14:32 +0000)]
Use a default of the FADT for matching CreatorId quirks.  Use XSDT to mean
"match either RSDT or XSDT."

20 years agoFix regression in last commit.
phk [Tue, 29 Jun 2004 08:33:58 +0000 (08:33 +0000)]
Fix regression in last commit.

20 years agoMention that ``set mtu max'' is necessary for PPPoE.
brian [Tue, 29 Jun 2004 07:48:43 +0000 (07:48 +0000)]
Mention that ``set mtu max'' is necessary for PPPoE.

PR: 32040
MFC after: 2 weeks

20 years agoReduce MAXMSS limit by 12 bytes to allow for rfc 1323.
brian [Tue, 29 Jun 2004 07:40:38 +0000 (07:40 +0000)]
Reduce MAXMSS limit by 12 bytes to allow for rfc 1323.

PR: 32717
Submitted by: MORI Kouji <moriko@hh.iij4u.or.jp>
MFC after: 2 weeks

20 years agoMFen 1.732
den [Tue, 29 Jun 2004 07:21:59 +0000 (07:21 +0000)]
MFen 1.732

Content fix *

Submitted by: Sergey Matveychuk <sem@ciam.ru> [*]
Obtained from: The FreeBSD Russian Documentation Project

20 years agoOops, this didn't make it into my submit before I committed: Defer
jhb [Tue, 29 Jun 2004 03:48:49 +0000 (03:48 +0000)]
Oops, this didn't make it into my submit before I committed:  Defer
creation of the sysctl tree for the turnstile profiling stats until a
SI_SUB_LOCK sysinit.  Doing it in init_turnstiles() is too early as it is
called before mi_startup().

20 years agoMess from update & merge - don't release the ifnet lock twice, just once
darrenr [Tue, 29 Jun 2004 03:39:06 +0000 (03:39 +0000)]
Mess from update & merge - don't release the ifnet lock twice, just once
and after we're finished with it.

20 years agoWrap long line.
peter [Tue, 29 Jun 2004 03:13:54 +0000 (03:13 +0000)]
Wrap long line.

20 years agoAdd two new kernel options to allow rudimentary profiling of the internal
jhb [Tue, 29 Jun 2004 02:30:12 +0000 (02:30 +0000)]
Add two new kernel options to allow rudimentary profiling of the internal
hash tables used in the sleep queue and turnstile code.  Each option adds
a sysctl tree under debug containing the maximum depth of any bucket in
the hash table as well as a separate node for each bucket (or chain)
containing the current depth and maximum depth for that bucket.

20 years agoInclude isa/pnpvar.h and remove a duplicate copy of PNP_EISAID.
njl [Tue, 29 Jun 2004 01:33:35 +0000 (01:33 +0000)]
Include isa/pnpvar.h and remove a duplicate copy of PNP_EISAID.

20 years agoWhile booting, use DELAY() for waiting for the EC to respond instead of
njl [Mon, 28 Jun 2004 20:41:30 +0000 (20:41 +0000)]
While booting, use DELAY() for waiting for the EC to respond instead of
msleep().  Until we're finally up, msleep is a no-op.

Explained by: peter

20 years agoInitialize result->backing_object_offset before linking result onto the list of
tegge [Mon, 28 Jun 2004 20:26:35 +0000 (20:26 +0000)]
Initialize result->backing_object_offset before linking result onto the list of
vm objects shadowing source in vm_object_shadow().  This closes a race where
vm_object_collapse() could be called with a partially uninitialized object
argument causing symptoms that looked like hardware problems, e.g.  signal 6,
10, 11 or a /bin/sh busy-waiting for a nonexistant child process.

20 years agoRemove the setting of the pci config variables on power state changes.
imp [Mon, 28 Jun 2004 20:26:21 +0000 (20:26 +0000)]
Remove the setting of the pci config variables on power state changes.
The bus does this now.

20 years agoRemove code to slam the config space on transition to d0.
imp [Mon, 28 Jun 2004 20:09:02 +0000 (20:09 +0000)]
Remove code to slam the config space on transition to d0.

20 years agoRemove burn bridges code that saved/restored the pci config registers
imp [Mon, 28 Jun 2004 20:07:03 +0000 (20:07 +0000)]
Remove burn bridges code that saved/restored the pci config registers
that are now handled in the pci bus layer.  They are no longer
necessary.

20 years agoUse MIN() macro rather than ulmin() inline, and fix stray tab
gallatin [Mon, 28 Jun 2004 19:58:39 +0000 (19:58 +0000)]
Use MIN() macro rather than ulmin() inline, and fix stray tab
that snuck in with my last commit.

Submitted by: green

20 years agoFix alpha - the use of min() on longs was loosing the high bits and
gallatin [Mon, 28 Jun 2004 19:15:40 +0000 (19:15 +0000)]
Fix alpha - the use of min() on longs was loosing the high bits and
returning wrong answers, leading to strange values vm2->vm_{s,t,d}size.

20 years agoBuild usbdevs.h
imp [Mon, 28 Jun 2004 19:06:30 +0000 (19:06 +0000)]
Build usbdevs.h

20 years agoRemove the signal_caught argument from sleepq_timedwait() as it was
jhb [Mon, 28 Jun 2004 18:57:06 +0000 (18:57 +0000)]
Remove the signal_caught argument from sleepq_timedwait() as it was
effectively always zero.

20 years agoanother format nit
imp [Mon, 28 Jun 2004 16:38:53 +0000 (16:38 +0000)]
another format nit

20 years agocasting for printf happiness
imp [Mon, 28 Jun 2004 16:30:49 +0000 (16:30 +0000)]
casting for printf happiness

20 years ago- Execute all of the tasks on the taskqueue during taskqueue_free() after
jhb [Mon, 28 Jun 2004 16:28:23 +0000 (16:28 +0000)]
- Execute all of the tasks on the taskqueue during taskqueue_free() after
  the queue has been removed from the global taskqueue_queues list.  This
  removes the need for the draining queue hack.
- Allow taskqueue_run() to be called with the taskqueue mutex held.  It
  can still be called without the lock for API compatiblity.  In that case
  it will acquire the lock internally.
- Don't lock the individual queue mutex in taskqueue_find() until after the
  strcmp as the global queues mutex is sufficient for the strcmp.
- Simplify taskqueue_thread_loop() now that it can hold the lock across
  taskqueue_run().

Submitted by: bde (mostly)

20 years agoWe need to build miidevs.h when we have miibus, not mii.
imp [Mon, 28 Jun 2004 16:22:20 +0000 (16:22 +0000)]
We need to build miidevs.h when we have miibus, not mii.

20 years agoAdjust the priorities of software interrupts. Bump SWI_CLOCK up to 5
jhb [Mon, 28 Jun 2004 16:21:51 +0000 (16:21 +0000)]
Adjust the priorities of software interrupts.  Bump SWI_CLOCK up to 5
and bump all of the taskqueue swi's to 6.  This gives callouts higher
priority than taskqueue tasks and gives all taskqueue tasks the same
priority.

Discussed with: bde

20 years agoAdjust the priority of the idle threads to be the lowest possible
jhb [Mon, 28 Jun 2004 16:19:50 +0000 (16:19 +0000)]
Adjust the priority of the idle threads to be the lowest possible
priority.  This is just a comestic nit as the idle thread priorities aren't
used by the schedulers.

Reported by: bde

20 years ago- Shorten the names for the TTY related swi interrupt handlers as the
jhb [Mon, 28 Jun 2004 16:17:11 +0000 (16:17 +0000)]
- Shorten the names for the TTY related swi interrupt handlers as the
  'tty:' prefix is largely redundant.
- Fix the priority of the low-priority TTY SWIs that are hung off of the
  softclock thread.

Submitted by: bde (2)

20 years agoDo not m_free packet since IF_HANDOFF (called from netisr_queue) will
rik [Mon, 28 Jun 2004 15:32:24 +0000 (15:32 +0000)]
Do not m_free packet since IF_HANDOFF (called from netisr_queue) will
do it for us, just count it.

20 years agoSome implicit dirs were not having the umask correctly
kientzle [Mon, 28 Jun 2004 14:12:09 +0000 (14:12 +0000)]
Some implicit dirs were not having the umask correctly
applied to their permissions.  Just calculate the
default dir mode once and use it consistently, rather than
trying to remember to calculate it everywhere it's needed.

20 years agoAdd yet another synonym for -n.
kientzle [Mon, 28 Jun 2004 14:02:22 +0000 (14:02 +0000)]
Add yet another synonym for -n.

20 years agoWhitespace fix
den [Mon, 28 Jun 2004 13:36:00 +0000 (13:36 +0000)]
Whitespace fix

Obtained from: The FreeBSD Russian Documentation Project

20 years agoMerged from sys/dev/sio/sio.c revision 1.444.
nyan [Mon, 28 Jun 2004 13:28:20 +0000 (13:28 +0000)]
Merged from sys/dev/sio/sio.c revision 1.444.

20 years agoAdd translation for early-adopter
den [Mon, 28 Jun 2004 13:28:15 +0000 (13:28 +0000)]
Add translation for early-adopter

20 years agoRemove libpthread_dbg, soon it will be replaced by libthread_db which
davidxu [Mon, 28 Jun 2004 12:13:36 +0000 (12:13 +0000)]
Remove libpthread_dbg, soon it will be replaced by libthread_db which
is a defacto standard on other UNIX systems.

20 years agoRegenerate this file so that it gets correct comments for the PCI
mux [Mon, 28 Jun 2004 11:54:30 +0000 (11:54 +0000)]
Regenerate this file so that it gets correct comments for the PCI
databases locations.

20 years agoUpdate the location of the Hart's database as well.
mux [Mon, 28 Jun 2004 11:46:48 +0000 (11:46 +0000)]
Update the location of the Hart's database as well.

20 years agoUpdate to the latest versions of the Boemler and Hart PCI databases.
mux [Mon, 28 Jun 2004 11:36:32 +0000 (11:36 +0000)]
Update to the latest versions of the Boemler and Hart PCI databases.

20 years agoRemove unused includes.
cognet [Mon, 28 Jun 2004 11:35:05 +0000 (11:35 +0000)]
Remove unused includes.

Spotted out by: pjd

20 years agoThe location of the Boemler's PCI database has moved. Update comments
mux [Mon, 28 Jun 2004 11:32:58 +0000 (11:32 +0000)]
The location of the Boemler's PCI database has moved.  Update comments
to reflect reality.

20 years agoFix spelling error in my own paragraph.
dd [Mon, 28 Jun 2004 11:18:47 +0000 (11:18 +0000)]
Fix spelling error in my own paragraph.

20 years agoDocument incorrect handling of multibyte characters in input files
tjr [Mon, 28 Jun 2004 07:19:11 +0000 (07:19 +0000)]
Document incorrect handling of multibyte characters in input files
and character string arguments.

20 years agoDescribe the algorithm used by the -n option.
tjr [Mon, 28 Jun 2004 06:20:28 +0000 (06:20 +0000)]
Describe the algorithm used by the -n option.

20 years agoMove some variable declarations to the top of the file.
tjr [Mon, 28 Jun 2004 06:05:36 +0000 (06:05 +0000)]
Move some variable declarations to the top of the file.

20 years agobde points out that this can't do anything useful. The full patch has
imp [Mon, 28 Jun 2004 05:37:22 +0000 (05:37 +0000)]
bde points out that this can't do anything useful.  The full patch has
other parts that I can't locat at the moment, so back it out until I can.

20 years agoRetire BUS_DMAMAP_NSEGS for sparc64
scottl [Mon, 28 Jun 2004 04:04:43 +0000 (04:04 +0000)]
Retire BUS_DMAMAP_NSEGS for sparc64

20 years agoSwitch sparc64 busdma to use a dynamically allocated segment list rather
scottl [Mon, 28 Jun 2004 03:49:13 +0000 (03:49 +0000)]
Switch sparc64 busdma to use a dynamically allocated segment list rather
than a a stack-limited list.  This removes the artifical limit on s/g list
size.
cvs: ----------------------------------------------------------------------

20 years agoTurns out that jhb didn't really like this. And nate pointed out that
imp [Mon, 28 Jun 2004 03:40:23 +0000 (03:40 +0000)]
Turns out that jhb didn't really like this.  And nate pointed out that
it wasn't a good idea to have the test for NULL on only a limited
subset.  Go back because I'm not sure adding NULL to all the others is
a good idea.

20 years agorman_reserve_resource doesn't set the bushandle, so we have to do it here.
imp [Mon, 28 Jun 2004 03:36:37 +0000 (03:36 +0000)]
rman_reserve_resource doesn't set the bushandle, so we have to do it here.

Badness noted by: njl
Perforce reply not read by: imp

20 years agoWhen opening /dev/io, preserve iopl properly. Otherwise, if you open
imp [Mon, 28 Jun 2004 03:16:32 +0000 (03:16 +0000)]
When opening /dev/io, preserve iopl properly.  Otherwise, if you open
/dev/io multiple times, the first close remove the privs.

20 years agoremove needless blankline
imp [Mon, 28 Jun 2004 03:08:17 +0000 (03:08 +0000)]
remove needless blankline

20 years agoUse APM_UNKNOWN instead of 0xff. Use ~0U instead 0 for the number of
imp [Mon, 28 Jun 2004 02:38:20 +0000 (02:38 +0000)]
Use APM_UNKNOWN instead of 0xff.  Use ~0U instead 0 for the number of
batteries.  This should be -1 to match other code, but since the api
is unsigned ~0U is the same thing.

20 years agoAllow dev to be NULL and assume that a device is not alive or not
imp [Mon, 28 Jun 2004 02:24:04 +0000 (02:24 +0000)]
Allow dev to be NULL and assume that a device is not alive or not
attached.

Reviewed by: njl(?) and jhb

20 years agoFix cut-n-paste bugs in hccontrol(8). Its very strange that no one noticed
emax [Mon, 28 Jun 2004 00:58:53 +0000 (00:58 +0000)]
Fix cut-n-paste bugs in hccontrol(8). Its very strange that no one noticed
them earlier. Thanks to Mike Crosland (mcrosland at xs4all.nl) for pointing
them out.

20 years agoChange the "rtprio" format so it prints an informative string for
gad [Sun, 27 Jun 2004 23:59:38 +0000 (23:59 +0000)]
Change the "rtprio" format so it prints an informative string for
the PRI_ITHD case (instead of just printing the digit '1').

Submitted by: Cyrille Lefevre

20 years agoRefuse to extract an entry from an archive on top of the archive being read.
kientzle [Sun, 27 Jun 2004 23:36:39 +0000 (23:36 +0000)]
Refuse to extract an entry from an archive on top of the archive being read.
(This is the converse of the "don't add an archive to itself".)

20 years agoUpdate -p:
kientzle [Sun, 27 Jun 2004 23:29:44 +0000 (23:29 +0000)]
Update -p:
  * Don't change the umask; the library now ignores the umask if
    you set EXTRACT_PERM
  * Set the EXTRACT_ACL and EXTRACT_FFLAGS bits (used to be
    controlled by EXTRACT_PERM).

20 years agoMore cleanup work on permissions restore:
kientzle [Sun, 27 Jun 2004 23:27:28 +0000 (23:27 +0000)]
More cleanup work on permissions restore:
  * Rename some variables/functions/etc to try to make things clearer.
  * Add separate flags to control fflag/acl restore
  * Collect metadata restore into a single function for clarity
  * Propagate errors in metadata restore back out to the client
  * Fix some places where errors were being returned when they
    shouldn't and vice-versa
  * Modes are now always restored; ARCHIVE_EXTRACT_PERM just controls
    whether or not umask is obeyed.
  * Restore suid/sgid bits only if user/group matches archive
  * Cache the last stat results to try to reduce the number of stat calls

20 years agoAdd two new convenience functions to query the uid/gid from an
kientzle [Sun, 27 Jun 2004 23:16:42 +0000 (23:16 +0000)]
Add two new convenience functions to query the uid/gid from an
archive_entry.

Update the Makefile MLINKS and manpage to bring it up-to-date with
the current status of archive_entry.  At least the manpage actually
lists all of the functions now, even if it doesn't really yet explain
them all.

20 years agoWe don't need GEN_M_CFILES. Fold it into GEN_CFILES definition and
imp [Sun, 27 Jun 2004 23:03:43 +0000 (23:03 +0000)]
We don't need GEN_M_CFILES.  Fold it into GEN_CFILES definition and
adjust as necessary.

Suggested by: bde

20 years agoImprove checking for `ps -t <dev>', and give better error messages when
gad [Sun, 27 Jun 2004 22:56:58 +0000 (22:56 +0000)]
Improve checking for `ps -t <dev>', and give better error messages when
an invalid <dev> is specified.  Aside:  It turns out that the S_ISCHR()
check is true for almost every device that we have (not just tty's).

20 years agoRwatson, write 100 times for tomorrow:
phk [Sun, 27 Jun 2004 21:54:34 +0000 (21:54 +0000)]
Rwatson, write 100 times for tomorrow:

First unlock, then assign NULL to pointer.

20 years agoMake sure to kill the devstat entry for disappearing disks.
phk [Sun, 27 Jun 2004 20:53:20 +0000 (20:53 +0000)]
Make sure to kill the devstat entry for disappearing disks.

PR: 68074
Submitted by: Hendrik Scholz <hscholz@raisdorf.net>

20 years agoOnly build miidevs.h when we have mii in the kernel
imp [Sun, 27 Jun 2004 20:16:56 +0000 (20:16 +0000)]
Only build miidevs.h when we have mii in the kernel
Only build pccarddevs.h when we have pccard in the kernel
Only build usbdevs.h when we have usb in the kernel

Suggested by: bde

20 years agoClean up some consistent confusion between "dev" and "rdev."
kientzle [Sun, 27 Jun 2004 18:38:13 +0000 (18:38 +0000)]
Clean up some consistent confusion between "dev" and "rdev."
Mostly, these were being used correctly even though a lot of
variables and function names were mis-named.

In the process, I found and fixed a couple of latent bugs and
added a guard against adding an archive to itself.

20 years agoWarn about stripping leading '/' when creating archives.
kientzle [Sun, 27 Jun 2004 18:32:14 +0000 (18:32 +0000)]
Warn about stripping leading '/' when creating archives.

20 years ago- Fix style bug in Dell Bluetooth adapter (don't repeat Vendor in
alex [Sun, 27 Jun 2004 17:27:04 +0000 (17:27 +0000)]
- Fix style bug in Dell Bluetooth adapter (don't repeat Vendor in
  device listing).
- Add generic "Port Replicator" entry for the USB Hubs used in
  Dell port replicators.

20 years agoNow that all the in-tree referrences to these old files are gone, go
imp [Sun, 27 Jun 2004 16:59:15 +0000 (16:59 +0000)]
Now that all the in-tree referrences to these old files are gone, go
ahead and remove them.

20 years agoSilence a warning about an unused argument.
tjr [Sun, 27 Jun 2004 16:58:56 +0000 (16:58 +0000)]
Silence a warning about an unused argument.

20 years agoCorrect the description of the -f option: input and output fields are
tjr [Sun, 27 Jun 2004 16:55:25 +0000 (16:55 +0000)]
Correct the description of the -f option: input and output fields are
separated by the field delimiter character, which is not necessarily
the tab character.

20 years agoMFp4: Last references to dev/usb/usbdevs.h converted.
imp [Sun, 27 Jun 2004 16:51:01 +0000 (16:51 +0000)]
MFp4: Last references to dev/usb/usbdevs.h converted.

20 years agoMore tweaks to implicit directory creation. Even such abominations as
kientzle [Sun, 27 Jun 2004 16:44:51 +0000 (16:44 +0000)]
More tweaks to implicit directory creation.  Even such abominations as
a/././b/../b/../c/./../d/e/f now work correctly.  And yes, a/b and a/c
both get created in this example; if you want, you can create an
entire dir heirarchy from a tar archive with only one entry.

More tweaks to umask support: umasks are now obeyed for all objects,
not just directories; the umask used is now the one in effect at the
corresponding call to archive_read_extract(), so clients that want to
tinker with umask during extract should get the expected behavior.

20 years agoHandle multibyte characters when cutting out fields (-f and -d options.)
tjr [Sun, 27 Jun 2004 16:42:33 +0000 (16:42 +0000)]
Handle multibyte characters when cutting out fields (-f and -d options.)

20 years agoUpdate a comment that compared the parser with an obsolete draft of
tjr [Sun, 27 Jun 2004 15:27:15 +0000 (15:27 +0000)]
Update a comment that compared the parser with an obsolete draft of
IEEE Std 1003.2-1992. Most of the "extensions" here were actually
required by the final version of the standard.

20 years agoAdd cross-reference to colrm(1).
tjr [Sun, 27 Jun 2004 14:57:42 +0000 (14:57 +0000)]
Add cross-reference to colrm(1).

20 years agoImplement the -c option correctly in locales with multibyte characters
tjr [Sun, 27 Jun 2004 14:55:07 +0000 (14:55 +0000)]
Implement the -c option correctly in locales with multibyte characters
instead of treating it as a synonym for -b.

20 years agoMFp4:
imp [Sun, 27 Jun 2004 13:25:28 +0000 (13:25 +0000)]
MFp4:

Add Intel Pro100Lan56 card.

Also integrate changes from Carlos Velasco.  Only attch if we're a
network device (to filter out the serial devices).  Also, increment
vpmatch if we match to conform to the pccard match function api.

20 years agoMFp4:
imp [Sun, 27 Jun 2004 13:10:20 +0000 (13:10 +0000)]
MFp4:

Use bus space rather than direct inb/outb.  Minor style changes while
I'm here.  Extremely preliminary support for siliconix ethernet cards
(but more work is required).

20 years agoMFp4:
imp [Sun, 27 Jun 2004 13:07:02 +0000 (13:07 +0000)]
MFp4:

The hack for setting the bus has been moved down into the cbb driver.
I've been running without this hack in my tree for so long I had
forgotten that I'd removed it :-).  Please let me know if this causes
difficulty for your laptop.

20 years agoMFp4: Set the bus handle to the bus handle of the resource, not the
imp [Sun, 27 Jun 2004 13:04:19 +0000 (13:04 +0000)]
MFp4: Set the bus handle to the bus handle of the resource, not the
starting value.  This is more pedantically correct (since the handle
isn't always identical to the start of the resource) and also doesn't
access the innards of struct resource direct (which I forbid in my
tree).  We need to do this for all resource types, not just ioport.

Reviewed by: njl

20 years agoIntroduce a hack that will make geom_gate to work with read-only mounts.
pjd [Sun, 27 Jun 2004 12:56:11 +0000 (12:56 +0000)]
Introduce a hack that will make geom_gate to work with read-only mounts.
Now, when trying to mount file system in read-only mode it tries to
opened a device for writting to be able to update to read-write mode
latter. Ehh.

Discussed with: phk

20 years agoMFp4: First batch of dev/usb/usbdevs.h -> usbdevs.h changes.
imp [Sun, 27 Jun 2004 12:41:44 +0000 (12:41 +0000)]
MFp4: First batch of dev/usb/usbdevs.h -> usbdevs.h changes.

20 years agoAdd usbdevs.h and pccarddevs.h to the build list, as appropriate. This is needed...
imp [Sun, 27 Jun 2004 12:33:20 +0000 (12:33 +0000)]
Add usbdevs.h and pccarddevs.h to the build list, as appropriate.  This is needed to build as part of world or in src/sys/modules.

20 years agoGuard against us having no tty pointer.
phk [Sun, 27 Jun 2004 11:29:54 +0000 (11:29 +0000)]
Guard against us having no tty pointer.

20 years agoFix problems with non-8 space tabs. New options for function
schweikh [Sun, 27 Jun 2004 10:58:37 +0000 (10:58 +0000)]
Fix problems with non-8 space tabs. New options for function
declarations with the opening brace on the same line as the declaration
of arguments all spaces and no tabs (a feature which exists in GNU's
indent). Man page update to follow RSN.

PR: bin/67983
Submitted by: Chip Norkus <wd@teleri.net>
Style guidance and bug for bug compatibility by: bde
MFC after: 2 weeks

20 years agoMake the handling of invalid multibyte sequences more robust by using
tjr [Sun, 27 Jun 2004 10:35:28 +0000 (10:35 +0000)]
Make the handling of invalid multibyte sequences more robust by using
mbrlen() instead of mblen().

20 years agoPrepare jail(8) utility for new functionality which will limit
pjd [Sun, 27 Jun 2004 10:10:16 +0000 (10:10 +0000)]
Prepare jail(8) utility for new functionality which will limit
seeing status of mounted file system for jailed processes.
Pass full path of jail's root directory to the kernel. mount(8) utility is
doing the same thing already.

20 years agoImplement pthread_atfork in libthr. This is mostly from deichen's
mtm [Sun, 27 Jun 2004 10:01:35 +0000 (10:01 +0000)]
Implement pthread_atfork in libthr. This is mostly from deichen's
work in libpthread.

Submitted by: Dan Nelson <dnelson@allantgroup.com>

20 years agoFix a tuner problem.
schweikh [Sun, 27 Jun 2004 09:59:02 +0000 (09:59 +0000)]
Fix a tuner problem.

PR: kern/66006
Submitted by: Josh Elsasser <jre@vineyard.net>
MFC after: 1 week

20 years agoIn the case that the global thread list is being re-initialized after
mtm [Sun, 27 Jun 2004 09:53:06 +0000 (09:53 +0000)]
In the case that the global thread list is being re-initialized after
a fork, make sure that the current thread isn't detached and freed. As
a consequence the thread should be inserted into the head of the
active list only once (in the beginning).

20 years agoThose are unneeded too.
pjd [Sun, 27 Jun 2004 09:06:10 +0000 (09:06 +0000)]
Those are unneeded too.

20 years agoAdd two missing includes and remove two uneeded.
pjd [Sun, 27 Jun 2004 09:03:22 +0000 (09:03 +0000)]
Add two missing includes and remove two uneeded.
This is quite serious fix, because even with MAC framework compiled in,
MAC entry points in those two files were simply ignored.

20 years agoAugment the -T handling:
kientzle [Sun, 27 Jun 2004 06:29:03 +0000 (06:29 +0000)]
Augment the -T handling:
 * Add --null option (sort #defines here)
 * Add process_lines function to util.c that reads newline-terminated
   or null-terminated lines (with self-sizing buffers, etc) and iteratively
   invokes a provided function.  Use this to dramatically simplify:
   -T handling for -c, --exclude-from-file, and --include-from-file.
 * Add -T handling to -x (via include_from_file)

Hopefully, this will fix the openoffice port and a couple of
others that rely on -T and --null.

20 years ago* The mode for implicitly-created dirs is now 0777 modified by the
kientzle [Sun, 27 Jun 2004 05:59:15 +0000 (05:59 +0000)]
* The mode for implicitly-created dirs is now 0777 modified by the
  umask in effect when the archive is closed
* Correct a typo that broke implicit dir creation for non-directories.

Thanks to: Garret A Wollman for pointing out my umask oversight

20 years agoRemoved. It was used for building gdb docs, which we don't do here
marcel [Sun, 27 Jun 2004 03:34:54 +0000 (03:34 +0000)]
Removed. It was used for building gdb docs, which we don't do here
anymore.

20 years agoWe don't have to reach into binutils/doc anymore for inc-hist.diff
marcel [Sun, 27 Jun 2004 03:33:17 +0000 (03:33 +0000)]
We don't have to reach into binutils/doc anymore for inc-hist.diff
The file was repo copied into our doc directory.

20 years agoRepo copied from: src/gnu/usr.bin/binutils/doc/inc-hist.diff
marcel [Sun, 27 Jun 2004 03:31:49 +0000 (03:31 +0000)]
Repo copied from: src/gnu/usr.bin/binutils/doc/inc-hist.diff

20 years agoAcquire the socket buffer lock when calling unp_scan() on
rwatson [Sun, 27 Jun 2004 03:29:25 +0000 (03:29 +0000)]
Acquire the socket buffer lock when calling unp_scan() on
so->so_rcv.sb_mb to prevent the mbuf chain from changing during the
scan.

20 years agoFix "@-" which has apparently been broken for some time. <sigh>
kientzle [Sun, 27 Jun 2004 03:28:13 +0000 (03:28 +0000)]
Fix "@-" which has apparently been broken for some time.  <sigh>
While I'm here, add in a lot more error-checking around append_archive.