]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
18 years agoDocument how we synchronize access to the fields in the cryptocap
pjd [Mon, 22 May 2006 07:49:42 +0000 (07:49 +0000)]
Document how we synchronize access to the fields in the cryptocap
structure.

18 years agoWe must synchronize access to cc_qblocked, because there could be a race
pjd [Mon, 22 May 2006 07:48:45 +0000 (07:48 +0000)]
We must synchronize access to cc_qblocked, because there could be a race
where crypto_invoke() returns ERESTART and before we set cc_qblocked to 1,
crypto_unblock() is called and sets it to 0. This way we mark device as
blocked forever.

Fix it by not setting cc_qblocked in the fast path and by protecting
crypto_invoke() in the crypto_proc thread with CRYPTO_Q_LOCK().
This won't slow things down, because there is no contention - we have
only one crypto thread. Actually it can be slightly faster, because we
save two atomic ops per crypto request.
The fast code path remains lock-less.

18 years agos/packages/packets/
joel [Mon, 22 May 2006 07:36:45 +0000 (07:36 +0000)]
s/packages/packets/

Noticed by: maxim

18 years agoAdd manual page for ipfwpcap(8).
joel [Mon, 22 May 2006 07:14:54 +0000 (07:14 +0000)]
Add manual page for ipfwpcap(8).

Reviewed by: phk, brueffer
Submitted by: Niclas Zeising <lothrandil@n00b.apagnu.se>

18 years agoo Correct comments: remove a reference to non-existent gdbinit.9 man page;
maxim [Mon, 22 May 2006 07:12:25 +0000 (07:12 +0000)]
o Correct comments: remove a reference to non-existent gdbinit.9 man page;
correct a path to the kernel build directory; gdb lost -k key.

Reviewed by: grog
MFC after: 1 week

18 years agoAdd missing case for RQSTYPE_CTIO3- neede for 64 bit target mode.
mjacob [Mon, 22 May 2006 07:07:30 +0000 (07:07 +0000)]
Add missing case for RQSTYPE_CTIO3- neede for 64 bit target mode.

18 years agoReflect int -> size_t changes in glob.h
ache [Mon, 22 May 2006 06:53:35 +0000 (06:53 +0000)]
Reflect int -> size_t changes in glob.h

18 years agoRemove bzero/bcopy vestiges
mjacob [Mon, 22 May 2006 06:51:48 +0000 (06:51 +0000)]
Remove bzero/bcopy vestiges

Be cognizant as to whether we're running 2KLogin f/w in target mode and
do the appropriate loopid load based upon that.

Do a first cut (seems to work, at least for amd64) at 64 bit target
mode for fibre channel cards. We could probably also do it for SPI
cards, but that's not supported right now.

18 years agoRemove bzero/bcopy vestiges.
mjacob [Mon, 22 May 2006 06:49:49 +0000 (06:49 +0000)]
Remove bzero/bcopy vestiges.

Be cognizant as to whether we're running 2KLogin f/w in target mode and
do the appropriate loopid load based upon that.

18 years agoremove bzero/bcopy vestiges
mjacob [Mon, 22 May 2006 06:48:40 +0000 (06:48 +0000)]
remove bzero/bcopy vestiges

18 years agoadd TGT_ANY define
mjacob [Mon, 22 May 2006 06:47:42 +0000 (06:47 +0000)]
add TGT_ANY define

18 years agoFix longstanding bug where exec throttle is 16 bits- not 8.
mjacob [Mon, 22 May 2006 06:47:20 +0000 (06:47 +0000)]
Fix longstanding bug where exec throttle is 16 bits- not 8.

18 years agoReflect size_t changes in glob.h
ache [Mon, 22 May 2006 06:33:19 +0000 (06:33 +0000)]
Reflect size_t changes in glob.h

Obtained from: NetBSD (mostly)

18 years agoAdd size_t declaration
ache [Mon, 22 May 2006 05:57:39 +0000 (05:57 +0000)]
Add size_t declaration

18 years agoMake gl_offs size_t too, as required by POSIX
ache [Mon, 22 May 2006 05:39:57 +0000 (05:39 +0000)]
Make gl_offs size_t too, as required by POSIX

Make non-standard gl_matchc in the similar fashion as gl_pathc size_t too,
like done in NetBSD & others

18 years agoRemove pending actions asked in comments for SHLIB_MAJOR bump, done.
ache [Mon, 22 May 2006 05:12:44 +0000 (05:12 +0000)]
Remove pending actions asked in comments for SHLIB_MAJOR bump, done.

Reviewed by:    ume

18 years agoUse size_t for gl_pathc as asked in the libc's Makefile:
ache [Mon, 22 May 2006 05:10:17 +0000 (05:10 +0000)]
Use size_t for gl_pathc as asked in the libc's Makefile:
# If you bump SHLIB_MAJOR, see standards/55112.

PR:             55112
Reviewed by:    ume

18 years agoRemove the kludge, as asked in the Makefile:
ache [Mon, 22 May 2006 05:04:53 +0000 (05:04 +0000)]
Remove the kludge, as asked in the Makefile:
# If you bump SHLIB_MAJOR, remove the kluge from gen/gethostname.c.

Reviewed by:    ume

18 years agoNew release notes: acpi_dock(4), netcat from OpenBSD 3.9.
bmah [Mon, 22 May 2006 00:59:57 +0000 (00:59 +0000)]
New release notes:  acpi_dock(4), netcat from OpenBSD 3.9.

Modified release notes:  nve(4) update for altq(4) support.

18 years agoWhen allocating a bucket to hold a free'd item in UMA fails, don't
rwatson [Sun, 21 May 2006 23:25:32 +0000 (23:25 +0000)]
When allocating a bucket to hold a free'd item in UMA fails, don't
report this as an allocation failure for the item type.  The failure
will be separately recorded with the bucket type.  This my eliminate
high mbuf allocation failure counts under some circumstances, which
can be alarming in appearance, but not actually a problem in
practice.

MFC after: 2 weeks
Reported by: ps, Peter J. Blok <pblok at bsd4all dot org>,
OxY <oxy at field dot hu>,
Gabor MICSKO <gmicskoa at szintezis dot hu>

18 years agoo Fix a comment: ufs2_dinode.di_blocks counts blocks not bytes actually held.
maxim [Sun, 21 May 2006 21:55:29 +0000 (21:55 +0000)]
o Fix a comment: ufs2_dinode.di_blocks counts blocks not bytes actually held.

18 years agoo Fix a comment: directory whiteout type is DT_WHT not DT_W.
maxim [Sun, 21 May 2006 21:28:34 +0000 (21:28 +0000)]
o Fix a comment: directory whiteout type is DT_WHT not DT_W.

18 years agoo In udp|rip_disconnect() acquire a socket lock before the socket
maxim [Sun, 21 May 2006 19:28:46 +0000 (19:28 +0000)]
o In udp|rip_disconnect() acquire a socket lock before the socket
state modification.  To prevent races do that while holding inpcb
lock.

Reviewed by: rwatson

18 years agoo Add missed error check: in ip_ctloutput() sooptcopyin() returns a
maxim [Sun, 21 May 2006 17:52:08 +0000 (17:52 +0000)]
o Add missed error check: in ip_ctloutput() sooptcopyin() returns a
result but we never examine it.

Reviewed by: rwatson
MFC after: 2 weeks

18 years agoResolve conflicts.
delphij [Sun, 21 May 2006 15:52:24 +0000 (15:52 +0000)]
Resolve conflicts.

18 years agoThis commit was generated by cvs2svn to compensate for changes in r158795,
delphij [Sun, 21 May 2006 15:49:27 +0000 (15:49 +0000)]
This commit was generated by cvs2svn to compensate for changes in r158795,
which included commits to RCS files with non-trunk default branches.

18 years agoImport netcat from OpenBSD 3.9-RELEASE.
delphij [Sun, 21 May 2006 15:49:27 +0000 (15:49 +0000)]
Import netcat from OpenBSD 3.9-RELEASE.

18 years agoBump library majro version for gethostbyaddr(3).
ume [Sun, 21 May 2006 15:15:21 +0000 (15:15 +0000)]
Bump library majro version for gethostbyaddr(3).

18 years agoNuke some compatibility crufts of resolver for 6.X and earlier.
ume [Sun, 21 May 2006 11:29:26 +0000 (11:29 +0000)]
Nuke some compatibility crufts of resolver for 6.X and earlier.

X-MFC after: never

18 years agoFix gethostbyaddr() prototype to conform to IEEE Std 1003.1 on 64 bit
ume [Sun, 21 May 2006 11:27:28 +0000 (11:27 +0000)]
Fix gethostbyaddr() prototype to conform to IEEE Std 1003.1 on 64 bit
arch.

X-MFC after: never

18 years agoReturn EAI_OVERFLOW instead of EAI_MEMORY when the supplied buffer is
ume [Sun, 21 May 2006 11:22:31 +0000 (11:22 +0000)]
Return EAI_OVERFLOW instead of EAI_MEMORY when the supplied buffer is
too short.  This conforms to RFC3493, POSIX and XPG6.

Obtained from: NetBSD

18 years agoUpgrade res_update(3) and the friends to BIND9's one excluding TSIG
ume [Sun, 21 May 2006 11:19:36 +0000 (11:19 +0000)]
Upgrade res_update(3) and the friends to BIND9's one excluding TSIG
support.

X-MFC after: never

18 years agoThis commit was generated by cvs2svn to compensate for changes in r158782,
ume [Sun, 21 May 2006 11:10:07 +0000 (11:10 +0000)]
This commit was generated by cvs2svn to compensate for changes in r158782,
which included commits to RCS files with non-trunk default branches.

18 years agoVendor import of BIND 9.3.2
ume [Sun, 21 May 2006 11:10:07 +0000 (11:10 +0000)]
Vendor import of BIND 9.3.2

18 years agoDon't allow non-root user to set a scheduler policy, otherwise this could
davidxu [Sun, 21 May 2006 00:40:38 +0000 (00:40 +0000)]
Don't allow non-root user to set a scheduler policy, otherwise this could
be a local DOS.

Submitted by: Diane Bruce at db at db.net

18 years ago- Fix path to sysinstall, it hasn't been in /stand for for some time.
simon [Sat, 20 May 2006 22:48:54 +0000 (22:48 +0000)]
- Fix path to sysinstall, it hasn't been in /stand for for some time.
- Put list outside <para> to avoid extra horisontal whitespace in
  output.

MFC after: 1 week

18 years agoRemove reference to "System call restart" in intro(2). intro(2) does not
rodrigc [Sat, 20 May 2006 21:49:03 +0000 (21:49 +0000)]
Remove reference to "System call restart" in intro(2).  intro(2) does not
have such a reference.  Add a reference instead to SA_RESTART in
sigaction(2).

18 years agoRemove stale altq instructions. They don't belong to the driver manpage
mlaier [Sat, 20 May 2006 21:12:06 +0000 (21:12 +0000)]
Remove stale altq instructions.  They don't belong to the driver manpage
anyhow.

18 years agoo .Xr strtonum(3).
maxim [Sat, 20 May 2006 21:11:35 +0000 (21:11 +0000)]
o .Xr strtonum(3).

MFC after: 1 week

18 years agoALTQ-ify nve(4).
mlaier [Sat, 20 May 2006 21:08:09 +0000 (21:08 +0000)]
ALTQ-ify nve(4).

Submitted by: Chris Dionissopoulos
Tested by: Chris Dionissopoulos
MFC after: 4 weeks

18 years agoo Convert to Test::Harness testing infratructure.
maxim [Sat, 20 May 2006 19:18:24 +0000 (19:18 +0000)]
o Convert to Test::Harness testing infratructure.

18 years ago- Add support for filtering the the list of providers by a regular
simon [Sat, 20 May 2006 19:17:47 +0000 (19:17 +0000)]
- Add support for filtering the the list of providers by a regular
  expression, which makes it possible to only see interesting
  providers.  "f" is used inside gstat to set a filter, "F" is used to
  remove current filter.
- Do not print some uninteresting values in the gstat title line.
- Do not print past the end of the screen.
- Read multiple keystrokes per "wait" when gstat is running.
- Remove a redundant != check, right after check of NULL against the
  same variable ("gid").
- Use sysexits.h.
- Do not link against libkvm and libsbuf, they are not actually used.
- Fix a few style(9) issues where I had to touch nearby code anyway.

Approved by: cperciva (mentor)
MFC after: 2 weeks

18 years agoPrevent disappearing SAD entries by implementing MPsafe refcounting.
pjd [Sat, 20 May 2006 15:35:36 +0000 (15:35 +0000)]
Prevent disappearing SAD entries by implementing MPsafe refcounting.
"Why didn't he use SECASVAR_LOCK()/SECASVAR_UNLOCK() macros to
 synchronize access to the secasvar structure's fields?" one may ask.
There were two reasons:
1. refcount(9) is faster then mutex(9) synchronization (one atomic
   operation instead of two).
2. Those macros are not used now at all, so at some point we may decide
   to remove them entirely.

OK'ed by: gnn
MFC after: 2 weeks

18 years agoChange kdump to print more useful information, i.e. it changes from
netchild [Sat, 20 May 2006 14:27:22 +0000 (14:27 +0000)]
Change kdump to print more useful information, i.e. it changes from
 32229 telnet   CALL  mmap(0,0x8000,0x3,0x1002,0xffffffff,0,0,0)
 32229 telnet   CALL  open(0x2807bc28,0,0x1b6)
 32229 telnet   CALL  socket(0x2,0x2,0)
to
 32229 telnet   CALL mmap(0,0x8000,PROT_READ|PROT_WRITE,MAP_PRIVATE|MAP_ANON,0xffffffff,0,0,0)
 32229 telnet   CALL  open(0x2807bc28,O_RDONLY,<unused>0x1b6)
 32229 telnet   CALL  socket(PF_INET,SOCK_DGRAM,0)

David wanted to implement the suggestions which came up at the review from
arch@ too, but real life rejected this proposal. So I commit what we already
got and let another volunteer pick the remaining work from the ideas list.

Submitted by: "David Kirchner" <dpk@dpk.net>
Suggested by: FreeBSD ideas list page
Reviewed by: arch

18 years agoIn IN6_IS_ADDR_V4MAPPED case instead of returning directly set error and
bz [Sat, 20 May 2006 13:26:08 +0000 (13:26 +0000)]
In IN6_IS_ADDR_V4MAPPED case instead of returning directly set error and
goto out so that locks will be dropped.

Reviewed by: rwatson, gnn

18 years agoFix a formatting issue.
brueffer [Sat, 20 May 2006 10:19:53 +0000 (10:19 +0000)]
Fix a formatting issue.

18 years agoMake this example more real world usable: When the manpage first appeared
brueffer [Sat, 20 May 2006 09:49:05 +0000 (09:49 +0000)]
Make this example more real world usable:  When the manpage first appeared
is not interesting, when the driver appeared is.  Most people who use this
example leave the manpage appearance date in and the driver date out.

MFC after: 3 days

18 years agoConvert to use a SYNPOSIS section that mentions kernel modules.
brueffer [Sat, 20 May 2006 09:39:28 +0000 (09:39 +0000)]
Convert to use a SYNPOSIS section that mentions kernel modules.
Also add "device scbus" where needed.

MFC after: 3 days

18 years agoo Sort .Xrs, touch .Dd.
maxim [Sat, 20 May 2006 06:58:46 +0000 (06:58 +0000)]
o Sort .Xrs, touch .Dd.

18 years agoMinimal manpage for the acpi_dock driver. This needs to be fleshed out
brueffer [Sat, 20 May 2006 06:21:42 +0000 (06:21 +0000)]
Minimal manpage for the acpi_dock driver.  This needs to be fleshed out
by someone in the know about this driver.

18 years agodo_file() is called recursively from several places and cannot
delphij [Sat, 20 May 2006 05:47:18 +0000 (05:47 +0000)]
do_file() is called recursively from several places and cannot
safely free() anything related to items that may be recursed on.

Obtained from: DragonFly (rev 1.7, dillon)
PR: bin/94767
MFC After: 1 month

18 years agoSwitch value back to values.
trhodes [Sat, 20 May 2006 04:32:10 +0000 (04:32 +0000)]
Switch value back to values.

Requested by: ru

18 years agoAdd description how to use caching.
ume [Sat, 20 May 2006 03:29:09 +0000 (03:29 +0000)]
Add description how to use caching.

Submitted by: Michael Bushkov <bushman__at__rsu.ru>

18 years agoBelatedly add usr/share/man/man4.sab.4.gz. The sab(4) driver was
marcel [Sat, 20 May 2006 00:30:06 +0000 (00:30 +0000)]
Belatedly add usr/share/man/man4.sab.4.gz. The sab(4) driver was
removed on 2006-04-24.

18 years agoNew release notes: Removed hard-coded limit on IPv4 multicast group
bmah [Fri, 19 May 2006 23:35:26 +0000 (23:35 +0000)]
New release notes:  Removed hard-coded limit on IPv4 multicast group
memberships, linsysfs(5), cached(8), nsswitch.conf now statically
loaded.

Modified release notes:  Note that XFS support is RO, libarchive(3)
extended attributes, don't use entity for usbd(8) removal note.

New manpages noted:  mca(8), sysctl(9).

18 years agoWe don't have d_maj field in cdevsw structure anymore.
sobomax [Fri, 19 May 2006 20:02:44 +0000 (20:02 +0000)]
We don't have d_maj field in cdevsw structure anymore.

18 years agoCorrect decoding a string containing '/'.
jkim [Fri, 19 May 2006 19:06:38 +0000 (19:06 +0000)]
Correct decoding a string containing '/'.

PR: 97485
Submitted by: Mikko Tyolajarvi < mbsd at pacbell dot net >

18 years agoMove clock_lock prototype into <machine/clock.h>, where it is more
sobomax [Fri, 19 May 2006 18:53:50 +0000 (18:53 +0000)]
Move clock_lock prototype into <machine/clock.h>, where it is more
appropriate.

Discussed with: jhb

18 years agoWe have an implementation of generic_bs_rr_1, so use it, as some drivers use
cognet [Fri, 19 May 2006 11:27:02 +0000 (11:27 +0000)]
We have an implementation of generic_bs_rr_1, so use it, as some drivers use
it.

Submitted by: kevlo

18 years agoStyle fixes.
davidxu [Fri, 19 May 2006 06:37:24 +0000 (06:37 +0000)]
Style fixes.

Submitted by: Diane Bruce < db at db dot net >

18 years agoIf the PHY has 1000BASE-T capability, check to see if a 1000BASE-T speed
yongari [Fri, 19 May 2006 03:51:42 +0000 (03:51 +0000)]
If the PHY has 1000BASE-T capability, check to see if a 1000BASE-T speed
was negotiated.

Obtained from:  NetBSD

18 years agoRemove reference to mount_procfs, which is no longer used by mount(8).
rodrigc [Fri, 19 May 2006 00:23:29 +0000 (00:23 +0000)]
Remove reference to mount_procfs, which is no longer used by mount(8).
Add an example using "mount -t procfs".

18 years agoRemove reference to mount_linprocfs, which is no longer used by mount(8).
rodrigc [Fri, 19 May 2006 00:22:45 +0000 (00:22 +0000)]
Remove reference to mount_linprocfs, which is no longer used by mount(8).
Add an example using "mount -t linprocfs".

18 years agoComment out SYSCTL_OMIT_DESCR until it's committed.
cognet [Fri, 19 May 2006 00:11:21 +0000 (00:11 +0000)]
Comment out SYSCTL_OMIT_DESCR until it's committed.

18 years agoChanges to make the NFS client MP safe.
mohans [Fri, 19 May 2006 00:04:24 +0000 (00:04 +0000)]
Changes to make the NFS client MP safe.

Thanks to Kris Kennaway for testing and sending lots of bugs my way.

18 years agoRemove reference to mount_procfs(8), add reference to procfs(5).
rodrigc [Fri, 19 May 2006 00:04:18 +0000 (00:04 +0000)]
Remove reference to mount_procfs(8), add reference to procfs(5).
mount(8) doesn't use mount_procfs(8), and instead passes an fstype
of "procfs" directly to nmount().

18 years agoAdd in a bunch of things to the mfi driver:
ambrisko [Thu, 18 May 2006 23:30:48 +0000 (23:30 +0000)]
Add in a bunch of things to the mfi driver:
      - Linux ioctl support, with the other Linux changes MegaCli
will run if you mount linprocfs & linsysfs then set
sysctl compat.linux.osrelease=2.6.12 or similar.  This works
on i386.  It should work on amd64 but not well tested yet.
StoreLib may or may not work.  Remember to kldload mfi_linux.
      - Add in AEN (Async Event Notification) support so we can
get messages from the firmware when something happens.
Not all messages are in defined in event detail.  Use
event_log to try to figure out what happened.
      - Try to implement something like SIGIO for StoreLib.  Since
mrmonitor doesn't work right I can't fully test it.  StoreLib
works best with the rh9 base.  In theory mrmonitor isn't
needed due to native driver support of AEN :-)
Now we can configure and monitor the RAID better.

Submitted by: IronPort Systems.

18 years ago- When setting up a packet for transmit, if we the tx ring is over half
jhb [Thu, 18 May 2006 23:19:44 +0000 (23:19 +0000)]
- When setting up a packet for transmit, if we the tx ring is over half
  full, kick the binary blob to force it to complete any pending tx
  completions.
- In the watchdog routine, only reset the chip if the blob doesn't complete
  any pending tx completions rather than requiring it to complete all of
  the pending tx completions.

Submitted by: Nathan Whitehorn <nathanw@uchicago.edu>
MFC after: 2 weeks

18 years agoImplement sa11x0_bs_unmap.
cognet [Thu, 18 May 2006 22:03:47 +0000 (22:03 +0000)]
Implement sa11x0_bs_unmap.
Submitted by: kevlo

18 years agoMake this compile (UART_IPEND_* => SER_INT_*).
cognet [Thu, 18 May 2006 22:02:33 +0000 (22:02 +0000)]
Make this compile (UART_IPEND_* => SER_INT_*).

18 years agoInitialize the new members of struct ip_moptions as
bms [Thu, 18 May 2006 19:51:08 +0000 (19:51 +0000)]
Initialize the new members of struct ip_moptions as
a defensive programming measure.

Note that whilst these members are not used by the ip_output()
path, we are passing an instance of struct ip_moptions here
which is declared on the stack (which could be considered a
bad thing).

ip_output() does not consume struct ip_moptions, but in case it
does in future, declare an in_multi vector on the stack too to
behave more like ip_findmoptions() does.

18 years agoMFCs noted: csup(1) (+ use new manpage entity), dhclient(8)
bmah [Thu, 18 May 2006 19:14:07 +0000 (19:14 +0000)]
MFCs noted:  csup(1) (+ use new manpage entity), dhclient(8)
sends hostname, find(1) birthtime primitives, pkg_add(1) -F.

18 years agoCorrect BD_ADDR entry for "Dummy" device in the default hcsecd.conf.
emax [Thu, 18 May 2006 17:53:49 +0000 (17:53 +0000)]
Correct BD_ADDR entry for "Dummy" device in the default hcsecd.conf.
Each byte in BD_ADDR should exactly two nibbles, i.e
"1:2:3:4:5:6" is NOT valid and "01:02:03:04:05:06" is valid BD_ADDR.

MFC after: 1 day

18 years agoNew release notes: pcvt(4) gone, pc98 bootable CDROM support
bmah [Thu, 18 May 2006 17:45:46 +0000 (17:45 +0000)]
New release notes:  pcvt(4) gone, pc98 bootable CDROM support
(and bootable CDROM releases), updated NTP manpages.

Modified release notes:  Munged ipfw(4) release note to remove
now-redundant remark about ip6fw going away "in future releases",
noted addition of new ext2fs(5) manpage.

18 years agoRemove the rest of the alpha-specific MD release documentation
bmah [Thu, 18 May 2006 16:27:53 +0000 (16:27 +0000)]
Remove the rest of the alpha-specific MD release documentation
(release notes and installation guide).

18 years agoRemove the require_kld function I just committed and replace with a nicer solution.
flz [Thu, 18 May 2006 16:04:56 +0000 (16:04 +0000)]
Remove the require_kld function I just committed and replace with a nicer solution.

Reported by: pjd
Approved by: cperciva

18 years agoAdd two new scripts (mdconfig/mdconfig2) to replace old ramdisk{,-own}
flz [Thu, 18 May 2006 15:29:27 +0000 (15:29 +0000)]
Add two new scripts (mdconfig/mdconfig2) to replace old ramdisk{,-own}
scripts. These scripts handle vnode backed md(4) devices.

Old ramdisk{,-own} scripts will stay a bit in CVS to allow some time for
migration since variable names have changed (ramdisk_* -> mdconfig_*).

Two new variables have been introduced to be able to populate the md(4)
device once it has been mounted (mdconfig_*_files and mdconfig_*_cmd).

Use should be as easy as:

mdconfig_md0="-t malloc -s 10m"
mdconfig_md1="-t vnode -f /var/foo.img"

See rc.conf(5) for more information and description of the additional
variables.

Approved by: cperciva

18 years agoSlight grammar fix.
ceri [Thu, 18 May 2006 15:26:27 +0000 (15:26 +0000)]
Slight grammar fix.

18 years agoRemove reference to mount_devfs(8), since mount(8) no longer
rodrigc [Thu, 18 May 2006 13:37:13 +0000 (13:37 +0000)]
Remove reference to mount_devfs(8), since mount(8) no longer
invokes this mount_devfs when mounting devfs file systems.

18 years agoAdd missing stats.hst_nodesc statistics.
pjd [Thu, 18 May 2006 13:22:44 +0000 (13:22 +0000)]
Add missing stats.hst_nodesc statistics.

18 years agoMove flag TDF_UMTXQ into structure umtxq, this eliminates the requirement
davidxu [Thu, 18 May 2006 08:43:46 +0000 (08:43 +0000)]
Move flag TDF_UMTXQ into structure umtxq, this eliminates the requirement
of scheduler lock in some umtx code.

18 years agoSilent Coverity Prevent report by asserting that cap != NULL.
pjd [Thu, 18 May 2006 06:28:39 +0000 (06:28 +0000)]
Silent Coverity Prevent report by asserting that cap != NULL.

Coverity ID: 1414

18 years agoRemove reference to mount_ext2fs(8), add reference to ext2fs(5).
rodrigc [Wed, 17 May 2006 23:29:11 +0000 (23:29 +0000)]
Remove reference to mount_ext2fs(8), add reference to ext2fs(5).
mount no longer invokes mount_ext2fs, it calls nmount() directly
with fstype "ext2fs".

18 years agoAdd a man page for ext2fs.
rodrigc [Wed, 17 May 2006 23:25:52 +0000 (23:25 +0000)]
Add a man page for ext2fs.

18 years agoIn change 1.15, an arbitrary setting of the max RPC record size to
rwatson [Wed, 17 May 2006 23:18:17 +0000 (23:18 +0000)]
In change 1.15, an arbitrary setting of the max RPC record size to
RPC_MAXDATASIZE was introduced.  This is believed to have been debugging
code committed accidentally, although I've been unable to reach the
committer to confirm this.  The effect was to limit the size of RPCs on
TCP and UDP to 9k, well below the default protocol limits in the libc
rpc code.  This change simply removes these introduced limits, falling
back on the libc definitions.

PR: 88856
Reported by: Keith Bostic <bostic at sleepycat dot com>
Testing by: Susan LoVerso <sue at loverso dot southborough dot ma dot us>
Reveiwed by: cel, rees
Review timeout: alfred, mbr
MFC after: 2 weeks

18 years ago- Add C-bus and ISA front-ends for le(4) so it can actually replace
marius [Wed, 17 May 2006 21:25:23 +0000 (21:25 +0000)]
- Add C-bus and ISA front-ends for le(4) so it can actually replace
  lnc(4) on PC98 and i386. The ISA front-end supports the same non-PNP
  network cards as lnc(4) did and additionally a couple of PNP ones.
  Like lnc(4), the C-bus front-end of le(4) only supports C-NET(98)S
  and is untested due to lack of such hardware, but given that's it's
  based on the respective lnc(4) and not too different from the ISA
  front-end it should be highly likely to work.
- Remove the descriptions of le(4), which where converted from lnc(4),
  from sys/i386/conf/NOTES and sys/pc98/conf/NOTES as there's a common
  one in sys/conf/NOTES.

18 years agoAdd le(4). I could actually only test it on alpha, i386 and sparc64 but
marius [Wed, 17 May 2006 20:45:45 +0000 (20:45 +0000)]
Add le(4). I could actually only test it on alpha, i386 and sparc64 but
given that this includes the more problematic platforms I see no reason
why it shouldn't also work on amd64 and ia64.

18 years ago- As only the PCI front-end of le(4) is common to all platforms move its
marius [Wed, 17 May 2006 20:44:01 +0000 (20:44 +0000)]
- As only the PCI front-end of le(4) is common to all platforms move its
  entry to the PCI NICs section so it's in the same spot in all GENERIC
  config files.
- Add a note to the description of pcn(4) informing that is has precedence
  over le(4).

18 years agoUpdate HMAC algorithms.
pjd [Wed, 17 May 2006 18:58:25 +0000 (18:58 +0000)]
Update HMAC algorithms.

18 years agoUpdate manual page: cri_rnd field is gone, cri_mlen field was added.
pjd [Wed, 17 May 2006 18:44:37 +0000 (18:44 +0000)]
Update manual page: cri_rnd field is gone, cri_mlen field was added.

18 years agoo SIOCGIFCONF -> getifaddrs(3) conversion. As a side effect fix
maxim [Wed, 17 May 2006 18:41:05 +0000 (18:41 +0000)]
o SIOCGIFCONF -> getifaddrs(3) conversion.  As a side effect fix
bin/95041: subnet mask mismatch.

PR: bin/95041
Obtained from: NetBSD
Tested by: Hans Lambermont
MFC after: 2 months

18 years agoHonor cri_mlen value.
pjd [Wed, 17 May 2006 18:34:26 +0000 (18:34 +0000)]
Honor cri_mlen value.

Reviewed by: sam
Tested on: hifn(4), ubsec(4)
Compile-tested: safe(4)

18 years ago- The authsize field from auth_hash structure was removed.
pjd [Wed, 17 May 2006 18:30:28 +0000 (18:30 +0000)]
- The authsize field from auth_hash structure was removed.
- Define that we want to receive only 96 bits of HMAC.
- Names of the structues have no longer _96 suffix.

Reviewed by: sam

18 years ago- Fix a very old bug in HMAC/SHA{384,512}. When HMAC is using SHA384
pjd [Wed, 17 May 2006 18:24:17 +0000 (18:24 +0000)]
- Fix a very old bug in HMAC/SHA{384,512}. When HMAC is using SHA384
  or SHA512, the blocksize is 128 bytes, not 64 bytes as anywhere else.
  The bug also exists in NetBSD, OpenBSD and various other independed
  implementations I look at.
- We cannot decide which hash function to use for HMAC based on the key
  length, because any HMAC function can use any key length.
  To fix it split CRYPTO_SHA2_HMAC into three algorithm:
  CRYPTO_SHA2_256_HMAC, CRYPTO_SHA2_384_HMAC and CRYPTO_SHA2_512_HMAC.
  Those names are consistent with OpenBSD's naming.
- Remove authsize field from auth_hash structure.
- Allow consumer to define size of hash he wants to receive.
  This allows to use HMAC not only for IPsec, where 96 bits MAC is requested.
  The size of requested MAC is defined at newsession time in the cri_mlen
  field - when 0, entire MAC will be returned.
- Add swcr_authprepare() function which prepares authentication key.
- Allow to provide key for every authentication operation, not only at
  newsession time by honoring CRD_F_KEY_EXPLICIT flag.
- Make giving key at newsession time optional - don't try to operate on it
  if its NULL.
- Extend COPYBACK()/COPYDATA() macros to handle CRYPTO_BUF_CONTIG buffer
  type as well.
- Accept CRYPTO_BUF_IOV buffer type in swcr_authcompute() as we have
  cuio_apply() now.
- 16 bits for key length (SW_klen) is more than enough.

Reviewed by: sam

18 years ago- Make opencrypto more SMP friendly by dropping the queue lock around
pjd [Wed, 17 May 2006 18:12:44 +0000 (18:12 +0000)]
- Make opencrypto more SMP friendly by dropping the queue lock around
  crypto_invoke(). This allows to serve multiple crypto requests in
  parallel and not bached requests are served lock-less.
  Drivers should not depend on the queue lock beeing held around
  crypto_invoke() and if they do, that's an error in the driver - it
  should do its own synchronization.
- Don't forget to wakeup the crypto thread when new requests is
  queued and only if both symmetric and asymmetric queues are empty.
- Symmetric requests use sessions and there is no way driver can
  disappear when there is an active session, so we don't need to check
  this, but assert this. This is also safe to not use the driver lock
  in this case.
- Assymetric requests don't use sessions, so don't check the driver
  in crypto_kinvoke().
- Protect assymetric operation with the driver lock, because if there
  is no symmetric session, driver can disappear.
- Don't send assymetric request to the driver if it is marked as
  blocked.
- Add an XXX comment, because I don't think migration to another driver
  is safe when there are pending requests using freed session.
- Remove 'hint' argument from crypto_kinvoke(), as it serves no purpose.
- Don't hold the driver lock around kprocess method call, instead use
  cc_koperations to track number of in-progress requests.
- Cleanup register/unregister code a bit.
- Other small simplifications and cleanups.

Reviewed by: sam

18 years agoRemove cri_rnd. It is not used.
pjd [Wed, 17 May 2006 18:04:51 +0000 (18:04 +0000)]
Remove cri_rnd. It is not used.

Reviewed by: sam

18 years agoIf kern.cryptodevallowsoft is TRUE allow also for symmetric software crypto
pjd [Wed, 17 May 2006 18:01:51 +0000 (18:01 +0000)]
If kern.cryptodevallowsoft is TRUE allow also for symmetric software crypto
in kernel. Useful for testing.

Reviewed by: sam

18 years agoForgot about adding cuio_apply() here.
pjd [Wed, 17 May 2006 17:58:05 +0000 (17:58 +0000)]
Forgot about adding cuio_apply() here.

Reviewed by: sam

18 years ago- Implement cuio_apply(), an equivalent to m_apply(9).
pjd [Wed, 17 May 2006 17:56:00 +0000 (17:56 +0000)]
- Implement cuio_apply(), an equivalent to m_apply(9).
- Implement CUIO_SKIP() macro which is only responsible for skipping the given
  number of bytes from iovec list. This allows to avoid duplicating the same
  code in three functions.

Reviewed by: sam

18 years agoDo not call knlist_destroy() in tapclose(). Instead call it when device is
emax [Wed, 17 May 2006 17:05:02 +0000 (17:05 +0000)]
Do not call knlist_destroy() in tapclose(). Instead call it when device is
actually destroyed. Also move call to knlist_init() into tapcreate(). This
should fix panic described in kern/95357.

PR: kern/95357
No response from: freebsd-current@
MFC after: 3 days