ChangeLog

2011-10-31 23:27 brooks
[r226973] Bring in an updated varsym patch from my P4 tree.
2011-10-31 16:05 brooks
[r226968] Create a branch to merge varsym patches into.
2011-08-11 14:17 brooks
[r224782] Take Kirill's commit bit into safe keeping per his request to core@.

Approved by: core
2011-08-03 20:21 brooks
[r224642] Fix two more $FreeBSD$ keywords.

Reported by: pluknet
Approved by: re (implicit)
2011-08-03 20:00 brooks
[r224640] Enable keyword expansion for $FreeBSD$ on files where it was added it
r224638.

Submitted by: bz
Approved by: re (implicit)
Point hat to: brooks
2011-08-03 19:14 brooks
[r224638] Add support for dynamically adjusted buffers to allow the full use of
the bandwidth of long fat pipes (i.e. 100Mbps+ trans-oceanic or
trans-continental links). Bandwidth-delay products up to 64MB are
supported.

Also add support (not compiled by default) for the None cypher. The
None cypher can only be enabled on non-interactive sessions (those
without a pty where -T was not used) and must be enabled in both
the client and server configuration files and on the client command
line. Additionally, the None cypher will only be activated after
authentication is complete. To enable the None cypher you must add
-DNONE_CIPHER_ENABLED to CFLAGS via the make command line or in
/etc/make.conf.

This code is a style(9) compliant version of these features extracted
from the patches published at:

http://www.psc.edu/networking/projects/hpn-ssh/

Merging this patch has been a collaboration between me and Bjoern.

Reviewed by: bz
Approved by: re (kib), des (maintainer)
2011-06-17 22:01 brooks
[r223213] It looks like the HPN patch didn't track the window size bump in OpenBSD
rev 1.89 back in 2007. Chase the updates to reduce diffs to head
2011-06-17 21:03 brooks
[r223205] Rename HPN-README to the openssh normal README.hpn form and bring in
bz's rework of the document.
2011-06-17 20:55 brooks
[r223204] Wrap all the NONE cipher support from HPN in #ifdef NONE_CIPHER_ENABLED.
Note that this does not include all references to it in the source.
2011-06-17 20:54 brooks
[r223203] Keep a debug message function call to 80 columns.
2011-06-17 20:19 brooks
[r223200] Reduce diffs to openssh-portable.
2011-06-17 20:18 brooks
[r223199] Fix a comment typo.
2011-06-17 18:49 brooks
[r223196] One more bogus directory
2011-06-17 18:48 brooks
[r223195] Remove some bogus directories. Not sure how these got there, but they
make it impossible to build.
2011-06-17 17:55 brooks
[r223193] Fix a couple mismerges
2011-06-17 17:43 brooks
[r223192] Merge lots of style and whitespace improvements from bz@.
2011-06-01 22:56 brooks
[r222587] Remove some of my XXX's. More general whitespace cleanups.
2011-05-12 03:37 brooks
[r221802] Cleaned up, but untested version of openssh5.8-dynwindow_noneswitch.diff
applied.
2011-05-12 02:23 brooks
[r221799] copy the vendor import of openssh 5.8p2 to start an HPN branch
2011-05-12 02:21 brooks
[r221798] make a work directory
2011-02-24 23:51 brooks
[r219018] Enable the check for negative permissions (the group on a file can't do
something "everyone" can) by default.

X-MFC after: never
2011-02-06 13:59 brooks
[r218372] MFC r215213:

Add an (off by default) check for negative permissions (where the
group on a object has less permissions than everyone). These
permissions will not work reliably over NFS if you have more than
14 supplemental groups and are usually not what you mean.
2010-11-13 00:40 brooks
[r215213] Add an (off by default) check for negative permissions (where the
group on a object has less permissions that everyone). These
permissions will not work reliably over NFS if you have more than
14 supplemental groups and are usually not what you mean.

MFC after: 1 week
2010-05-04 15:52 brooks
[r207616] MFC r205073

Regen:
* Hart: rev 671 of pcidevs.txt; 22-01-2008 (D-M-Y).
* Boemler: vendors.txt (2010-03126)

PR: kern/133733
2010-05-03 19:56 brooks
[r207583] MFC r205073

Regen:
* Hart: rev 671 of pcidevs.txt; 22-01-2008 (D-M-Y).
* Boemler: vendors.txt (2010-03126)

PR: kern/133733
2010-04-21 22:34 brooks
[r207026] This work is complete. All relevent changes have been merged to head.
2010-03-12 07:26 brooks
[r205073] Regen:

* Hart: rev 671 of pcidevs.txt; 22-01-2008 (D-M-Y).
* Boemler: vendors.txt (2010-03126)

PR: kern/133733
MFC after: 1 week
2010-03-04 19:35 brooks
[r204724] MFC r201853:

Make options KGSSAPI build and add it to NOTES.

rpcsec_gss_prot.c:
Use kernel printf and headers.

vc_rpcsec_gss.c:
Use a local RPCAUTH_UNIXGIDS definition for 16 instead of using NGROUPS.
2010-03-03 21:47 brooks
[r204671] MFC r201959

Use the correct types to store uids and gids in the credential cache and
eliminate an inappropriate use of NGROUPS.
2010-02-24 22:16 brooks
[r204293] MFC r202143,202163,202341,202342,204278

Replace the static NGROUPS=NGROUPS_MAX+1=1024 with a dynamic
kern.ngroups+1. kern.ngroups can range from NGROUPS_MAX=1023 to
somewhere in the neighborhood of INT_MAX/4 one a system with sufficent
RAM and memory bandwidth. Given that the Windows group limit is
1024, this range should be sufficient for most applications

r202342:
Only allocate the space we need before calling kern_getgroups instead
of allocating what ever the user asks for up to "ngroups_max + 1". On
systems with large values of kern.ngroups this will be more efficient.

The now redundant check that the array is large enough in
kern_getgroups() is deliberate to allow this change to be merged to
stable/8 without breaking potential third party consumers of the API.
2010-02-24 15:52 brooks
[r204278] Don't inforce an upper bound on kern.ngroups. The INT_MAX-1 limit was
too high due to several overflows. The actual limit is somewhere in the
neighborhood of INT_MAX/4 on 64-bit machines, but most systems could not
support such a limit due to a lack of memory and the cost of duplicate
credentials.

Reported by: bde
2010-01-22 20:24 brooks
[r202828] MFC r201261:
Add missing `void' keywords.
2010-01-22 19:51 brooks
[r202823] MFC r201350:

The devices that supported EVFILT_NETDEV kqueue filters were removed in
r195175. Remove all definitions, documentation, and usage.

The change of function signature for vlan_link_state() was not merged to
maintain the ABI.
2010-01-15 07:18 brooks
[r202342] Only allocate the space we need before calling kern_getgroups instead
of allocating what ever the user asks for up to "ngroups_max + 1". On
systems with large values of kern.ngroups this will be more efficient.

The now redundant check that the array is large enough in
kern_getgroups() is deliberate to allow this change to be merged to
stable/8 without breaking potential third party consumers of the API.

Reported by: bde
MFC after: 28 days
2010-01-15 07:05 brooks
[r202341] Since all other comparisons involving ngroups_max use
"ngroups_max + 1", use ">= ngroups_max+1" instead of the equivalent
"> ngroups_max" to reduce confusion.
2010-01-12 18:20 brooks
[r202163] Declare the kern.ngroups sysctl to be read-only, but tunable at boot for
better error reporting.

Submitted by: Matthew Fleming <matthew dot fleming at isilon dot com>
MFC After: 1 month
2010-01-12 07:55 brooks
[r202144] MFH at r202143
2010-01-12 07:49 brooks
[r202143] Replace the static NGROUPS=NGROUPS_MAX+1=1024 with a dynamic
kern.ngroups+1. kern.ngroups can range from NGROUPS_MAX=1023 to
INT_MAX-1. Given that the Windows group limit is 1024, this range
should be sufficient for most applications.

MFC after: 1 month
2010-01-12 07:40 brooks
[r202142] Cap ngroups_max and (INT_MAX-1), any larger and we'll overflow the int
counters.
2010-01-12 07:33 brooks
[r202140] MFC r201444:

The size of credential messages is limited by CMGROUP_MAX rather than
NGROUPS.
2010-01-12 06:14 brooks
[r202138] MFC r201955:
Improve the comment about CMGROUP_MAX.
2010-01-12 06:11 brooks
[r202137] MFC r201954:

Update the comment on printing group membership to reflect that fact
that each group the process is a member of is printed rather than
an entry for each group the user could be a member of.
2010-01-12 06:09 brooks
[r202136] MFC r201953:
Correct the explination text for the kern.ngroups. It reflects the
number of supplemental groups, not the total number of groups.
2010-01-12 06:04 brooks
[r202135] MFC r201443:
Add vlan(4) to all GENERIC kernels.
2010-01-12 06:00 brooks
[r202134] MFC r201443:
Add vlan(4) to all GENERIC kernels.
2010-01-10 00:08 brooks
[r201965] Document the kern.ngroups tunable.
2010-01-09 23:36 brooks
[r201959] Use the correct types to store uids and gids in the credential cache.
2010-01-09 23:24 brooks
[r201955] Improve the comment about CMGROUP_MAX.

MFC after: 3 days
2010-01-09 23:23 brooks
[r201954] Update the comment on printing group membership to reflect that fact
that each groupt the process is a member of is printed rather than an
entry for each group the user could be a member of.

MFC after: 3 days
2010-01-09 23:22 brooks
[r201953] Correct the explination text for the kern.ngroups. It reflects the
number of supplemental groups, not the total number of groups.

MFC after: 3 days
2010-01-09 23:19 brooks
[r201952] Follow HEAD here and just limit the groups to 16 since this is RPC.
2010-01-09 23:18 brooks
[r201950] We do need this assignment to cr_gid here since the group array isn't
copied.
2010-01-09 23:17 brooks
[r201949] Tell the user what limit they exceeded rather than just giving an
unconnected number.
2010-01-09 23:07 brooks
[r201947] Use proper types to store uid's and gid's.
2010-01-09 23:02 brooks
[r201944] Improve a few comments.
2010-01-09 23:02 brooks
[r201943] Tweak the line wrapping of the kern.ngroups decleration to be more
consistent with other lines.
2010-01-09 23:00 brooks
[r201942] Remove some files that should have been removed in a previous botched
merge.
2010-01-08 23:30 brooks
[r201854] MFC at r201853
2010-01-08 23:26 brooks
[r201853] Make options KGSSAPI build and add it to NOTES.

rpcsec_gss_prot.c:
Use kernel printf and headers.

vc_rpcsec_gss.c:
Use a local RPCAUTH_UNIXGIDS definition for 16 instead of using NGROUPS.
2010-01-08 22:34 brooks
[r201843] MFC at r201841
2010-01-08 22:13 brooks
[r201838] MFC r201352
If a filter has already been added, actually return EEXIST when trying
at add it again.
2010-01-08 21:58 brooks
[r201835] MFC r201352

If a filter has already been added, actually return EEXIST when trying
at add it again.
2010-01-08 21:28 brooks
[r201829] MFC r201349:
Ignore .glimpse_* files as generated by the glimpse target. Also ignore
cscope.* files.
2010-01-08 21:28 brooks
[r201828] MFC r201349:
Ignore .glimpse_* files as generated by the glimpse target. Also ignore
cscope.* files.
2010-01-08 21:27 brooks
[r201826] MFC r201349:
Ignore .glimpse_* files as generated by the glimpse target. Also ignore
cscope.* files.
2010-01-03 20:45 brooks
[r201444] The size of credential messages is limited by CMGROUP_MAX rather than
NGROUPS.

MFC after: 1 week
2010-01-03 20:40 brooks
[r201443] Add vlan(4) to all GENERIC kernels.

MFC after: 1 week
2010-01-03 17:19 brooks
[r201437] Sync up to head.
2010-01-03 17:13 brooks
[r201436] Remove NGROUPS definition. Not 100% sure we'll actually want to do
this, but nothing depends on it in the base of this branch anymore and
no properly written program should depend on it being a static value.
2010-01-03 17:11 brooks
[r201435] Use the rpc internal NGRPS instead of NGROUPS which is now the wrong
size.
2010-01-03 17:10 brooks
[r201434] CMGROUP_MAX is the maximum number of groups in the credential here.
2009-12-31 23:52 brooks
[r201358] Remove a couple comment references to NGROUPS/NGROUPS_MAX.
2009-12-31 23:50 brooks
[r201357] s/NGROUPS/ngroups_max+1/ in a KASSERT.
2009-12-31 22:47 brooks
[r201356] Merge up to the latest head
2009-12-31 21:08 brooks
[r201353] Checkpoint the replacement of the static NGROUPS value with a tunable
ngroups_max+1.
2009-12-31 20:56 brooks
[r201352] If a filter has already been added, actually return EEXIST when trying
at add it again.

MFC after: 1 week
2009-12-31 20:29 brooks
[r201350] The devices that supported EVFILT_NETDEV kqueue filters were removed in
r195175. Remove all definitions, documentation, and usage.

fifo_misc.c:
Remove all kqueue tests as fifo_io.c performs all those that
would have remained.

Reviewed by: rwatson
MFC after: 3 weeks
X-MFC note: don't change vlan_link_state() function signature
2009-12-31 19:40 brooks
[r201349] Ignore .glimpse_* files as generated by the glimpse target. Also ignore
cscope.* files.
2009-12-18 06:09 brooks
[r200662] MFC r200366

Add a missing else that negated the truncation of ki_ngroups to NGROUPS.

Submitted by: Dmitry Pryanishnikov <lynx dot ripe at gmail dot com>
2009-12-10 21:16 brooks
[r200366] Add a missing else that negated the truncation of ki_ngroups to
NGROUPS.

Submitted by: Dmitry Pryanishnikov <lynx dot ripe at gmail dot com>
2009-09-24 21:35 brooks
[r197474] MFC r196990:
cr_groups is no longer embedded in struct ucred and is instead stored
in a seperate array. As such we need to use kvm_read rather than bcopy
to populate the ki_groups field.

This fixes a crash when running ps -ax on a coredump.

Reported by: brucec
Tested by: brucec
MFC after: 3 days
Approved by: re@ (kib)
2009-09-24 21:32 brooks
[r197473] MFC r197269:
Allocate space for the group array in a static credential used in
the quota code. One case was correctly handled in r194498, but
this one was missed.

PR: kern/138657
Tested by: PR submitter
MFC after: 3 days
Approved by: re@ (kib)
2009-09-17 12:35 brooks
[r197269] Allocate space for the group array in a static credential used in
the quota code. One case was correctly handled in r194498, but
this one was missed.

PR: kern/138657
Tested by: PR submitter
MFC after: 3 days
2009-09-09 21:51 brooks
[r197050] Change one no to not in r197047

Submitted by: John Nielsen <john at jnielsen net>
2009-09-09 20:28 brooks
[r197047] Mention that a few ed(4) devices don't emit link state change notices
and a workaround for dhclient.

Reviewed by: thierry
MFC after: 3 days
2009-09-08 19:37 brooks
[r196990] cr_groups is no longer embedded in struct ucred and is instead stored
in a seperate array. As such we need to use kvm_read rather than bcopy
to populate the ki_groups field.

This fixes a crash when running ps -ax on a coredump.

Reported by: brucec
Tested by: brucec
MFC after: 3 days
2009-07-24 21:42 brooks
[r195859] MFC r194494 to improve support for 7.x worlds on >=8.0 kernels:

In preparation for raising NGROUPS and NGROUPS_MAX, change base
system callers of getgroups(), getgrouplist(), and setgroups() to
allocate buffers dynamically. Specifically, allocate a buffer of size
sysconf(_SC_NGROUPS_MAX)+1 (+2 in a few cases to allow for overflow).

This (or similar gymnastics) is required for the code to actually follow
the POSIX.1-2008 specification where {NGROUPS_MAX} may differ at runtime
and where getgroups may return {NGROUPS_MAX}+1 results on systems like
FreeBSD which include the primary group.

In id(1), don't pointlessly add the primary group to the list of all
groups, it is always the first result from getgroups(). In principle
the old code was more portable, but this was only done in one of the two
places where getgroups() was called to the overall effect was pointless.

Document the actual POSIX requirements in the getgroups(2) and
setgroups(2) manpages. We do not yet support a dynamic NGROUPS, but we
may in the future.
2009-07-24 19:12 brooks
[r195853] Introduce a new sysctl process mib, kern.proc.groups which adds the
ability to retrieve the group list of each process.

Modify procstat's -s option to query this mib when the kinfo_proc
reports that the field has been truncated. If the mib does not exist,
fall back to the truncated list.

Reviewed by: rwatson
Approved by: re (kib)
MFC after: 2 weeks
2009-07-24 15:03 brooks
[r195843] Revert the changes to struct kinfo_proc in r194498. Instead, fill
in up to 16 (KI_NGROUPS) values and steal a bit from ki_cr_flags
(all bits currently unused) to indicate overflow with the new flag
KI_CRF_GRP_OVERFLOW.

This fixes procstat -s.

Approved by: re (kib)
2009-07-06 01:32 brooks
[r195388] MFC r194493 and r194548

When checking if we can write to a file, use access() instead of a
manual permission check based on stat output. Also, get rid of the
executability check since it is not used.
2009-06-29 19:46 brooks
[r195175] Remove support for the /dev/net/* per-interface devices. They serve
little purpose and are unused in the base system.

The IOCTL functionality is entirely duplicated and routing sockets
provide a richer interface than the kqueue functionality.

Further, it is not practical for these devices to be made sensible in
the face of VIMAGE.

Bump __FreeBSD_version on the off chance that there is any code out
there that actually uses this stuff.

Reviewed by: rwatson
Discussed with: bz, zec
Approved by: re@ (kensmith)
2009-06-20 20:29 brooks
[r194556] Change crsetgroups_locked() (called by crsetgroups()) to sort the
supplemental groups using insertion sort. Use this property in
groupmember() to let us use a binary search instead of the previous
linear search.
2009-06-20 18:52 brooks
[r194552] Use NGROUPS instead of NGROUPS_MAX as the limits on setgroups and
getgroups for ibcs emulation. It seems vanishingly likely any
programs will actually be affected since they probably assume a much
lower value and use a static array size.
2009-06-20 18:24 brooks
[r194548] Restore the check against running as root that I accidentally removed in
r194493.
2009-06-19 21:11 brooks
[r194514] Use insertion sort of sort supplemental groups in crsetgroups(). Take
advantage of this an reimplement groupmember() as a test against
cr_group[0] followed by a binary search of the supplemental groups.
Seems to work in trivial testing.
2009-06-19 19:16 brooks
[r194507] Document crcopysafe() and crsetgroups().

Reminded by: julian
2009-06-19 19:12 brooks
[r194506] Document crcopysafe() and crsetgroups().

Reminded by: julian
2009-06-19 18:42 brooks
[r194503] MFH r194301-194502
2009-06-19 17:10 brooks
[r194498] Rework the credential code to support larger values of NGROUPS and
NGROUPS_MAX, eliminate ABI dependencies on them, and raise the to 1024
and 1023 respectively. (Previously they were equal, but under a close
reading of POSIX, NGROUPS_MAX was defined to be too large by 1 since it
is the number of supplemental groups, not total number of groups.)

The bulk of the change consists of converting the struct ucred member
cr_groups from a static array to a pointer. Do the equivalent in
kinfo_proc.

Introduce new interfaces crcopysafe() and crsetgroups() for duplicating
a process credential before modifying it and for setting group lists
respectively. Both interfaces take care for the details of allocating
groups array. crsetgroups() takes care of truncating the group list
to the current maximum (NGROUPS) if necessary. In the future,
crsetgroups() may be responsible for insuring invariants such as sorting
the supplemental groups to allow groupmember() to be implemented as a
binary search.

Because we can not change struct xucred without breaking application
ABIs, we leave it alone and introduce a new XU_NGROUPS value which is
always 16 and is to be used or NGRPS as appropriate for things such as
NFS which need to use no more than 16 groups. When feasible, truncate
the group list rather than generating an error.

Minor changes:
- Reduce the number of hand rolled versions of groupmember().
- Do not assign to both cr_gid and cr_groups[0].
- Modify ipfw to cache ucreds instead of part of their contents since
they are immutable once referenced by more than one entity.

Submitted by: Isilon Systems (initial implementation)
X-MFC after: never
PR: bin/113398 kern/133867
2009-06-19 15:58 brooks
[r194494] In preparation for raising NGROUPS and NGROUPS_MAX, change base
system callers of getgroups(), getgrouplist(), and setgroups() to
allocate buffers dynamically. Specifically, allocate a buffer of size
sysconf(_SC_NGROUPS_MAX)+1 (+2 in a few cases to allow for overflow).

This (or similar gymnastics) is required for the code to actually follow
the POSIX.1-2008 specification where {NGROUPS_MAX} may differ at runtime
and where getgroups may return {NGROUPS_MAX}+1 results on systems like
FreeBSD which include the primary group.

In id(1), don't pointlessly add the primary group to the list of all
groups, it is always the first result from getgroups(). In principle
the old code was more portable, but this was only done in one of the two
places where getgroups() was called to the overall effect was pointless.

Document the actual POSIX requirements in the getgroups(2) and
setgroups(2) manpages. We do not yet support a dynamic NGROUPS, but we
may in the future.

MFC after: 2 weeks
2009-06-19 15:52 brooks
[r194493] When checking if we can write to a file, use access() instead of a
manual permission check based on stat output. Also, get rid of the
executability check since it is not used.

MFC after: 2 weeks
2009-06-18 02:51 brooks
[r194412] Put parens around the definition of NGROUPS to avoid operator precedence
bugs in code that uses it.

Submitted by: Matthew Fleming <matthew dot fleming at isilon dot com>
2009-06-17 23:34 brooks
[r194410] crextend() is now only used in kern_prot.c so make it static.

Remove the inline from crsetgroups_locked() since the compiler should do
that automagically.
2009-06-17 23:11 brooks
[r194409] s/cr_gid/cr_groups[0]/ so we'll remove the cr_gid reference from head.
2009-06-17 20:56 brooks
[r194398] Use crsetgroups() in crcopy().

When a user has more than a page worth of groups, switch from power of
two allocation to allocating the number of pages actually required. It
seems more likely that a process will need a non-power-of-two number of
pages than that they will use many thousands of groups and add them one
at a time (in which case there would be no difference in number of
allocations until they use more than 3072 groups and the context switch
overhead would likely swamp the extra allocations).
2009-06-17 20:30 brooks
[r194397] Set cr_groups[0] instead of cr_gid here. They are the same and we use
cr_groups[0] almost exclusivly.
2009-06-17 20:26 brooks
[r194394] We're filling in an xucred so use XU_NGROUPS instead of NGRPS.
2009-06-17 19:40 brooks
[r194389] use MIN() instead of min() in one case.
2009-06-17 19:01 brooks
[r194388] Consistently use arrays of {NGROUPS_MAX}+1 with getgroups() and
getgrouplist(). In doing so, use the pattern from the example from the
posix manpages where ngroups_max is a long and set to
sysconf(_SC_NGROUPS_MAX)+1.

In id(1), don't use a combination of getgid() and getgroups() to
retrieve the list of groups. There's no need to account for the
possibility that getgroups does not return the primary gid, espeically
when we only do it in of of the two cases.
2009-06-17 18:42 brooks
[r194382] Include sys/ucred.h to get the definition of XU_NGROUPS.
2009-06-17 15:57 brooks
[r194371] Fix a style(9) bug.

Reported by: jhb
2009-06-16 22:39 brooks
[r194309] Instead of caching an incomplete copy of the interesting bits of struct
ucred, just grab a reference and cache a pointer to the real deal.
2009-06-16 18:22 brooks
[r194302] MFH r194045-194300
2009-06-16 17:08 brooks
[r194299] Use POSIX documentation style {NGROUPS_MAX} notation instead of talking
about NGROUPS or sysconf(_SC_NGROUPS_MAX).

Note the getgroups conforms to POSIX.1.
2009-06-12 06:13 brooks
[r194045] MFH @ r194044
2009-06-09 14:18 brooks
[r193832] Fix some style bugs and return the right error. Also, document the new
out of memory error condition.
2009-06-09 13:50 brooks
[r193830] Back out the new uses of the cr_gid macro introduced in r193537. We're
certaintly not going to split cr_gid out from cr_groups for 8.0.

Also reintroduce a few related comments.
2009-06-09 04:58 brooks
[r193806] Set NGROUPS_MAX to be 1023 since it should be the number of
_supplemental_ groups and thus getgroups() should return up to
NGROUPS_MAX+1 results on FreeBSD (ignoring for the moment the fact
that NGROUPS_MAX isn't actually a maximum value under POSIX). This
limits waste in naive applications to a single page.

Set NGROUPS to NGROUPS_MAX+1 to properly reflect the in-kernel space
requirements.
2009-06-05 21:00 brooks
[r193537] MFC @ r193535
2009-06-05 20:23 brooks
[r193534] Introduce a new crsetgroups() function and use it when copying groups
into ucred structures.

Consistently use the cr_gid macro to reference cr_groups[0].

Increase cr_ngroups and cr_agroups from short to int.
2009-06-05 17:20 brooks
[r193525] Clamp the number of groups ipfw can track to 16 for now.
2009-06-05 16:47 brooks
[r193520] Remove a redundent assignment to cr_groups[0] immediatly after an
assignment to cr_gid (which is currently defined as cr_groups[0]).
2009-06-05 16:45 brooks
[r193519] Clamp the number of groups in the portal credential to XU_NGROUPS (16).
This should be modified along with userspace to be dynamic.
2009-06-05 16:43 brooks
[r193517] Use groupmember() instead of hand rolling our own implementation.
2009-06-01 22:34 brooks
[r193284] Increase NGROUPS_MAX to 32767.
2009-05-29 19:03 brooks
[r193048] Fix a botched merge in r193002 by removing an almost duplicate line.
2009-05-28 23:13 brooks
[r193005] Don't attempt to copy more than XU_NGROUPS gids into the NFS credential
structure.
2009-05-28 22:12 brooks
[r193002] MFH @ r193000
2009-05-28 22:11 brooks
[r193001] it's been too long since I merged and my branch diffs are too big
2009-05-28 21:41 brooks
[r192998] More work toward increasing NGROUPS_MAX and/or allowing kern.ngroups to
be increased at boot time.

* Replace the embedded cr_groups[NGROUPS] member of struct cred with
a pointer to malloc'd storage and a cr_agroups value which tracks
the number of allocated group slots. When more space is required
crextend() is used to add more space. The new crcopysafe() function
calls crextend as needed when saving a copy of a credential from a
process before modification.

* Introduce a new #define, XU_NGROUPS which is always 16 and is used to
set the array size in struct xucred and will be used other places
where the 16 group limit needs to be preserved.

* Replace the embedded ki_groups[KI_NGROUPS] entry in struct kinfo_proc
with a pointer and adapt libkvm accordingly.

Submitted by: Isilon Systems (Matthew Fleming)
2009-05-14 06:50 brooks
[r192087] Use to value returned by sysconf(_SC_NGROUPS_MAX) in favor of
NGROUPS_MAX or NGROUPS since POSIX says that NGROUPS_MAX represents a
lower bound on sysconf(_SC_NGROUPS_MAX).
2009-05-14 06:48 brooks
[r192086] Don't ifndef NGROUPS for a base system program. If it's not defined,
we shouldn't be using it.
2009-05-14 03:54 brooks
[r192084] Only copy as many groups from cr_groups to ki_groups as will fit.
Truncate if there are too many.
2009-05-14 03:51 brooks
[r192083] Switch a couple instances of NGROUPS to NGRPS where the use case is RPC
credentials.
2009-05-08 20:48 brooks
[r191922] Remove a use of NGROUPS by replacing a couple hand rolled checks that
are equivalent to access() calls.
2009-05-07 16:31 brooks
[r191888] Create a complete copy of the source tree to work on increasing
NGROUPS_MAX.
2009-04-15 20:53 brooks
[r191121] The structs ifaliasreq and in_aliasreq have exactly the same layout and
member names, but we really do mean to use in_aliasreq here.

MFC after: 1 week
2009-03-13 07:12 brooks
[r189759] Add support for setting the debug flags on wlan interfaces after the are
created using wlandebug_<ifn> variables.
2008-12-17 17:35 brooks
[r186237] Correct a bug where /etc/rc.d/defaultroute fails to finish by printing a
newline when it fails to obtain an address via DHCP. This made the next
rc script begin its output on the same line.

PR: conf
Submitted by: Bruce Cran <bruce at cran dot org dot uk>
MFC after: 3 days
2008-10-26 19:28 brooks
[r184297] MFC r183974:
Support the remaining options listed in dhcp-options(5) and RFC 2132.

PR: bin/127076
Submitted by: jkim
Approved by: re (kensmith)
2008-10-26 19:22 brooks
[r184296] MFC r183979:
Display usage when pkg_add is called with no arguments.

PR: bin/121093
Submitted by: volker
Approved by: re (kensmith)
2008-10-26 19:14 brooks
[r184294] MFC r183976:
Wireless Mouse device of Sony VGP-WRC1 mouse/keyboard receiver has the
same program interface as Microsoft Wireless Notebook Optical Mouse and
needs a quirk.

PR: usb/122712
Approved by: re (kensmith)
2008-10-17 15:10 brooks
[r183979] Display usage when pkg_add is called with no arguments.

PR: bin/121093
Submitted by: volker
Approved by: portmgr (linimon)
MFC after: 3 days
2008-10-17 14:37 brooks
[r183976] Wireless Mouse device of Sony VGP-WRC1 mouse/keyboard receiver has the
same program interface as Microsoft Wireless Notebook Optical Mouse and
needs a quirk.

PR: usb/122712
MFC after: 3 days
2008-10-17 13:28 brooks
[r183974] Support the remaining options listed in dhcp-options(5) and RFC 2132.

PR: bin/127076
Submitted by: jkim
MFC after: 1 week
2008-10-01 18:46 brooks
[r183517] Remove compat support for vaps_<ifn> and vap_create_<ifn> variables as
promised in r178527. These variables were never in a release version.

Reminded by: sam
2008-09-03 20:33 brooks
[r182742] Replace a line matching /^ $/ with one matching /^$/.

Obtained from: //depot/projects/vimage-commit2/...
2008-09-01 19:25 brooks
[r182633] Wrap an 81 column SYSCTL_NODE decleration.

Obtained from: //depot/projects/vimage-commit2/...
2008-09-01 17:15 brooks
[r182615] Wrap a line that became too long with the addition of V_.

(This file contains many more unwrapped or badly wrapped lines.)
2008-07-25 19:58 brooks
[r180815] Add support for a new login capability, cpumask which allows login
sessions to be pinned to cpus by login class.
2008-07-11 15:17 brooks
[r180448] Put the _cpuset* symbols in FBSDprivate_1.0 instead of trying to put
nonexistant __cpuset* there.
2008-07-10 19:54 brooks
[r180436] style(9): put parentheses around return values.
2008-07-10 17:46 brooks
[r180434] Regen
2008-07-10 17:45 brooks
[r180433] id_t is a 64-bit integer and thus is passed as two arguments like off_t is.
As a result, those arguments must be recombined before calling the real
syscal implementation. This change fixes 32-bit compatibility for
cpuset_getid(), cpuset_setid(), cpuset_getaffinity(), and
cpuset_setaffinity().
2008-06-20 12:47 brooks
[r179891] Please welcome Erik Cederstrand to the ranks of src committers. He will be
working on things related to his automated performance tracking system.
kris and I will mentor.

Approved by: core
2008-06-13 22:42 brooks
[r179782] Take jkh's commit bit into safe keeping per his request.

Approved by: core
2008-06-09 20:03 brooks
[r179689] Use the -n flag to route(8) when calling "route get". Otherwise we hang
for a long time if we get a lease, but DNS isn't working.

MFC after: 1 week
2008-06-05 22:35 brooks
[r179584] MFC r179024: Change a use of u_int32_t to uint32_t.

PR: bin/93172
2008-06-05 22:35 brooks
[r179583] MFC r179024: Change a use of u_int32_t to uint32_t.

PR: bin/93172
2008-06-05 17:26 brooks
[r179566] Fix the wait for default route change I made a few weeks ago by creating
a new defaultroute script that just does the wait. The previous attempt
created a circular dependency through network_ipv6.

Pointy hat to: brooks
2008-05-19 02:49 brooks
[r179111] getopt.c is public domain. Add a comment to that effect.

Remove confusing README.

PR: bin/98911
Submitted by: Jason McIntyre <jmc at kerhand dot co dot uk>
Obtained from: OpenBSD
2008-05-18 02:57 brooks
[r179079] Move the wait for a default route to rc.d/routing. Once we test for
non-dhcp interfaces to negotiate/associate this will make more sense.

This also correctly gets run after both devd and netif are run so it has
a chance of working.
2008-05-17 03:38 brooks
[r179066] The if_check() function performed three actions:
- verified that the ifp->if_snd.ifq_mtx was initalized for
all attached interfaces. This was pointless because it was
initalized for all interfaces in if_attach() so I've removed it.
- Checked that ifp->if_snd.ifq_maxlen is initalized and set it to
ifqmaxlen if unset. This makes more sense in if_attach() so
I moved it there.
- The first call of if_slowtimo(). Delete if_check() and call
if_slowtimo() directly from the SYSINIT().
2008-05-15 20:04 brooks
[r179024] Change a use of u_int32_t to uint32_t.

PR: bin/93172
Submitted by: Robert Millan <rmh at aybabtu dot com>
MFC after: 1 week
2008-05-15 19:27 brooks
[r179020] getopt.c is public domain. Add a comment to that effect.

Remove confusing README.

PR: bin/98911
Submitted by: Jason McIntyre <jmc at kerhand dot co dot uk>
Obtained from: OpenBSD
MFC after: 3 days
2008-05-15 01:59 brooks
[r179007] MFC rev 1.8: When sending packets directly to the DHCP server, use a socket
and send directly rather than bogusly sending it out as a link layer
broadcast (which fails to be received on some networks).

PR: bin/96018
2008-05-15 01:53 brooks
[r179006] MFC: Add very limited support for the isc-dhclient. It will almostly certaintly
only work if there's just one interface doing dhcp. This version implements
the same logic as the version in the PR, but uses pgrep to be less verbose.

PR: conf/95905
2008-05-15 01:06 brooks
[r179003] Change the default value of synchronous_dhclient to NO.

To preserve the existing behavior of etc/rc.d/netif, add code to wait
up to if_up_delay seconds (30 seconds by default) for a default route to
be configured if there are any dhcp interfaces. This should be extended
to test that the interface is actually up.

X-MFC after:
2008-05-15 00:08 brooks
[r179002] Fix last commit and call childif_destroy() correctly.
2008-05-14 23:53 brooks
[r179001] Don't print the interface status if we only create child or destroy
interfaces.

Correctly return status from childif_create().
2008-05-04 16:15 brooks
[r178764] MFC: Emit a warning when the network_interfaces variable is not set to AUTO.
2008-05-03 07:06 brooks
[r178738] Replace a couple mentions of the soon to be removed vaps_<ifn>
variable form with wlans_<ifn>.
2008-04-30 16:29 brooks
[r178695] Emit a warning when the network_interfaces variable is not set to AUTO.

MFC after: 3 days
2008-04-29 21:29 brooks
[r178675] MFC rev 1.8

When sending packets directly to the DHCP server, use a socket and send
directly rather than bogusly sending it out as a link layer broadcast
(which fails to be received on some networks).

PR: bin/96018
MFC after: 2 weeks
2008-04-29 20:43 brooks
[r178672] MFC rev 1.278:

Delay the global registration of the struct ifnet in if_alloc() until after
we're certain the allocation will entierly succeed. This fixes a leak in a
fairly unlikely case.

Reported by: vijay singh <vijjus at rocketmail dot com>
2008-04-29 20:34 brooks
[r178671] MFC per ppp profile variable documentation (revs 1.336 and 1.337)

PR: conf/121452, conf/122127
2008-04-29 20:17 brooks
[r178669] MFC per ppp profile variable documentation (revs 1.336 and 1.337)

PR: conf/121452, conf/122127
2008-04-25 23:50 brooks
[r178527] Replace the prototype vaps_<ifn> and vap_create_<ifn> variables with
more wlans_<ifn> and create_args_<ifn>

Add documentation for these variants and generally update the wireless
device example.

There is are very short lived shim from vaps_<ifn> which produces
a warning and vap_create_<ifn> which does not. Misuse the MFC
notification service to remind me to remove them.

MFC after: 3 weeks
2008-04-23 22:40 brooks
[r178450] Revert rev 1.332 and keep ddb scripts off by default for now. Minidumps
are more flexable and much text-dump like output can be produced from
them so there's a good argument they are a better default.
2008-04-21 18:17 brooks
[r178381] Change the default of ddb_enable to YES so we default to generating textdumps
on panic. This means you get a potentially useful dump even if your system
is running X when you panic.

X-MFC after: never
2008-04-19 22:04 brooks
[r178323] Delay the global registration of the struct ifnet in if_alloc() until after
we're certain the allocation will entierly succeed. This fixes a leak in a
fairly unlikely case.

Reported by: vijay singh <vijjus at rocketmail dot com>
MFC after: 1 week
2008-04-15 23:03 brooks
[r178233] Add very limited support for the isc-dhclient. It will almostly certaintly
only work if there's just one interface doing dhcp. This version implements
the same logic as the version in the PR, but uses pgrep to be less verbose.

PR: conf/95905
MFC after: 1 week
2008-04-15 22:48 brooks
[r178232] When sending packets directly to the DHCP server, use a socket and send
directly rather than bogusly sending it out as a link layer broadcast
(which fails to be received on some networks).

PR: bin/96018
MFC after: 2 weeks
2008-04-15 22:37 brooks
[r178230] MFC revs 1.17-1.20, the functional components of which are:

1.17:
Allow the characters .-+/ to appear in ppp profile names by folding them
to _ when evaluating ppp_<profile>_nat and ppp_<profile>_mode. Document
the per-profile variables.

1.18:
Add support for hardwiring ppp sessions to particular devices with new
per-profile variables of the form ppp_<profile>_unit. No ppp_unit
variable is supported since tying the same unit to more than one profile
won't work.

PR: conf/121452, conf/122127
2008-04-15 22:34 brooks
[r178229] MFC rev 1.20:
Declare _ppp_profile_cleaned, _punct, and _punct_c local in
ppp_start_profile().

Reported by: yar
2008-04-11 00:18 brooks
[r178089] MFC rev 1.19

Add a new function is_default_interface() which determines if this
interface is one with the default route (or there isn't one). Use it to
decide if we should adjust the default route and /etc/resolv.conf.

Fix the delete of the default route. The if statement was totally bogus
and the delete only worked due to a typo. [1]

Reported by: Jordan Coleman <jordan at JordanColeman dot com> [1]
2008-04-10 01:32 brooks
[r178056] Declare _ppp_profile_cleaned, _punct, and _punct_c local in
ppp_start_profile().

Reported by: yar
MFC after: 1 week
2008-04-06 20:40 brooks
[r177960] Fix a stupid typo.

Reported by: bz
2008-04-06 20:39 brooks
[r177959] Fix a stupid typo.

Reviewed by: bz
2008-04-03 21:36 brooks
[r177898] MFC:

Use the new command file feature of ddb(8) to support setting ddb(4)
scripts at boot. This is currently disabled by default. /etc/ddb.conf
contains some potentially reasonable default scripts.

PR: conf/119995
Submitted by: Scot Hetzel <swhetzel at gmail dot com> (Earlier version)
2008-04-03 21:33 brooks
[r177897] NFC the ability to read a file of commands to ddb(8) modeled after the
feature in ipfw(8).
2008-03-30 02:42 brooks
[r177730] Add a new function is_default_interface() which determines if this
interface is one with the default route (or there isn't one). Use it to
decide if we should adjust the default route and /etc/resolv.conf.

Fix the delete of the default route. The if statement was totally bogus
and the delete only worked due to a typo. [1]

Reported by: Jordan Coleman <jordan at JordanColeman dot com> [1]
MFC after: 1 week
2008-03-28 08:19 brooks
[r177684] Use ; instead of : to end a line.

Submitted by: Niclas Zeising <niclas dot zeising at gmail dot com>
2008-03-28 07:57 brooks
[r177683] Add support for hardwiring ppp sessions to particular devices with new
per-profile variables of the form ppp_<profile>_unit. No ppp_unit
variable is supported since tying the same unit to more than one profile
won't work.

PR: conf/122127
MFC after: 1 week
2008-03-28 06:50 brooks
[r177682] Support gif_interface values that don't follow the pattern gif###.
Remove ancient compatablity support for gif_interface="NO".
2008-03-27 03:30 brooks
[r177658] Remove the C flag from slip.log. The current slip userbase does not
justify the presence of a (usually empty) /var/log/slip.log on every
FreeBSD box.
2008-03-26 21:54 brooks
[r177649] Allow the characters .-+/ to appear in ppp profile names by folding them
to _ when evaluating ppp_<profile>_nat and ppp_<profile>_mode. Document
the per-profile variables.

PR: conf/121452, conf/122127 (partial)
MFC after: 1 week
2008-03-12 17:41 brooks
[r177114] Sync with current merging revs 1.27 and 1.31.

1.27:
Use get_if_var() to retrieve interface specific values of dhclient_flags
and background_dhclient. This allows interfaces who's names are not
valid parts of shell variables and shortens the code.

1.31:
When the state of the interface changes rapidly enough there is a race
where dhclient is in the process of exiting due to the link going down
when the link coming up causes devd to try and start a new one. This
results is the link being up, but no dhclient running.

Work around this race by checking a second time after a one second delay
before refusing to start a dhclient instance due to one already being
running.
2008-03-12 17:35 brooks
[r177113] MFC rev 1.31:

When the state of the interface changes rapidly enough there is a race
where dhclient is in the process of exiting due to the link going down
when the link coming up causes devd to try and start a new one. This
results is the link being up, but no dhclient running.

Work around this race by checking a second time after a one second delay
before refusing to start a dhclient instance due to one already being
running.
2008-03-12 17:29 brooks
[r177112] MFC rev 1.18:

During PREINIT, when giving the interface the address 0.0.0.0, do it as an
alias to avoid distrubing other addresses.

PR: bin/119255
Submitted by: Jaakko Heinonen <jh at saunalahti dot fi>
2008-03-05 18:32 brooks
[r176835] Use the new command file feature of ddb(8) to support setting ddb(4)
scripts at boot. This is currently disabled by default. /etc/ddb.conf
contains some potentially reasonable default scripts.

PR: conf/119995
Submitted by: Scot Hetzel <swhetzel at gmail dot com> (Earlier version)
X-MFC after: textdumps
2008-03-05 17:51 brooks
[r176833] Add the ability to read a file of commands to ddb(8) modeled after the
feature in ipfw(8).
2008-02-01 23:43 brooks
[r175879] When the state of the interface changes rapidly enough (usually due to
rapid wireless association changes in my experience), there is a race
where dhclient is in the process of exiting due to the link going down
when the link coming up causes devd to try and start a new one. This
results is the link being up, but no dhclient running.

Work around this race by checking a second time after a one second delay
before refusing to start a dhclient instance due to one already being
running.

MFC after: 1 week
2008-01-21 23:54 brooks
[r175554] During PREINIT, when giving the interface the address 0.0.0.0, do it as an
alias to avoid distrubing other addresses.

PR: bin/119255
Submitted by: Jaakko Heinonen <jh at saunalahti dot fi>
2007-12-06 16:22 brooks
[r174365] MFC:
To reduce the impact of possible removal of the syscalls required by
libkse in FreeBSD 8.0, do not build or install static versions of libkse
(i.e. libkse*.a) in the default case. Static versions will be built and
installed if libthr is not built or if libkse is the default threading
library.

Discussed on: freebsd-arch
Approved by: re (kensmith)
2007-11-29 21:49 brooks
[r174092] To reduce the impact of possible removal of the syscalls required by
libkse in FreeBSD 8.0, do not build or install static versions of libkse
(i.e. libkse*.a) in the default case. Static versions will be built and
installed if libthr is not built or if libkse is the default threading
library.

Discussed on: freebsd-arch
MFC after: 3 days
2007-09-21 01:26 brooks
[r172260] Use the udp protocol in favor of the nonexistant upd protocol in the
sge_execd entry.

Reported by: emaste
Pointy hat to: brooks
Approved by: re (kensmith)
2007-09-06 19:04 brooks
[r172065] Add service entries for Sun Grid Engine's qmaster and execution service
as per IANA assignments to simplify the installation of the sysutils/sge
port.

Approved by: re (bmah)
2007-05-16 19:59 brooks
[r169619] Update the comments on if_alloc(), if_free(), if_free_type(), and
if_attach.

Remove a comment about pre-3.0 network drivers from if_attach().

Be a bit more consistant about whitespace near comments.
2007-05-16 18:37 brooks
[r169614] The struct if_data members ifi_recvquota and ifi_xmitquota have been
unused for ages. Rename them to ifi_spare_char1 and ifi_spare_char2
respectively to indicate this face.
2007-04-02 14:09 brooks
[r168270] jlemon was my mentor and I got my commit bit in 2001.
2007-03-25 23:58 brooks
[r167889] Use "-ne" instead of "!=" for integer comparison.
2007-03-11 10:48 brooks
[r167451] Spell .Xr without a '/'.

Pointy hat: brooks
2007-03-11 10:24 brooks
[r167448] Fix a couple markup problems in the previous commit and bump the
document date.

Reported by: ru
2007-03-11 06:53 brooks
[r167437] Allow background_fsck_delay to be set to a negative value which delays
the background fsck indefinitely. This allows the administrator to run
it at a convenient time. To support running it from cron, the
forcestart argument now causes the fsck to start with no delay and all
output to be suppressed.
2007-03-02 20:48 brooks
[r167181] Use get_if_var() to retrieve interface specific values of dhclient_flags
and background_dhclient. This allows interfaces who's names are not
valid parts of shell variables and shortens the code.

MFC after: 1 week
2007-02-09 19:03 brooks
[r166604] Include GEOM_LABEL in GENERIC. It's very useful and not well publicized
enough.

Approved by: pjd
2007-02-09 03:11 brooks
[r166581] Sync dhclient and supporting code with CURRENT. In particular:
- Introduce per-interface dhclient_flags_<ifn> and
background_dhclient_<ifn> rc.conf variables.
- Default to sending a client identifer.
- When an invalid host name option is sent, delete it and accept the
lease.
- Fix usage, the call to getopt, and a typo in rc.d/dhclient.
2007-01-29 15:55 brooks
[r166330] Actually implement rev 1.12 for host names and NIS domain names. We
were removing the invalid option, but still rejecting the lease.

Reported by: Yoshihiko Sarumaru <mistral at imasy dot or dot jp>
2007-01-15 17:11 brooks
[r166026] MFC: head.c revs 1.19-20, Makefile rev 1.2

Fix head -c ### where ### is greater than 2^31. Unlike the submitted
patch this uses off_t.

WARNSify and add $FreeBSD$ to Makefile.

PR: bin/107824
Submitted by: Brian Cornell <briancornell at earthlink dot net>
2007-01-11 20:23 brooks
[r165955] Fix build on architectures where off_t is signed by casting to uintmax_t
before comparing with a size_t.
2007-01-11 17:03 brooks
[r165940] Fix head -c ### where ### is greater than 2^31. Unlike the submitted
patch this uses off_t.

WARNSify and add $FreeBSD$ to Makefile.

PR: bin/107824
Submitted by: Brian Cornell <briancornell at earthlink dot net>
MFC after: 3 days
2007-01-09 03:29 brooks
[r165908] MFC rev 1.25:
Pull in /etc/rc.conf.d/network so that ifconfig_<if> variables can be
set there. This is required for consistency with /etc/rc.d/netif.

PR: conf/103893
Submitted by: Nick Hibma <nick at anywi.com>
2006-10-02 18:50 brooks
[r162972] Pull in /etc/rc.conf.d/network so that ifconfig_<if> variables can be
set there. This is required for consistency with /etc/rc.d/netif.

PR: conf/103893
Submitted by: Nick Hibma <nick at anywi.com>
MFC after: 3 days
2006-09-29 03:07 brooks
[r162756] MFC rev 1.7

It is possible for bpf to return a length such that:

length != BPF_WORDALIGN(length)

This meeans that it is possible for this to be true:

interface->rbuf_offset > interface->rbuf_len

Handle this case in the test for running out of packets. While
OpenBSD's solution of setting interface->rbuf_len to
BPF_WORDALIGN(length) is safe due to the size of the buffer, I think
this solution results in less hidden assumptions.

This should fix the problem of dhclient running away and consuming 100%
CPU.

PR: bin/102226
Submitted by: Joost Bekkers <joost at jodocus.org>
Approved by: re (ken)
2006-09-26 02:17 brooks
[r162645] MFC

pccard_ether rev 1.51:
Search the list of up interfaces provided by "ifconfig -ul" instead of
greping for UP in "ifconfig $ifn". This eliminates a dependancy on /usr.

pccard_ether rev 1.53 and network.subr rev 1.173:
Introduce a new method ipv6if which attemptes to figure out if an
interface is an IPv6 interface.

Use this method to decide if we should attempt to configure an interface
with an IPv6 address in pccard_ether. The mechanism pccard_ether uses
to do this is unsuited to the task because it assumes the list of
interfaces it is passed is the full list of IPv6 interfaces and makes
decissions based on that. This is at least a step in the right
direction and is probably about as much as we can MFC safely.

PR: conf/103428
Approved by: re (hrs)
2006-09-26 02:15 brooks
[r162644] MFC rev 1.10

network_ipv6 also does some interface configuration so require it to
run before starting devd so they don't trip over each other.

PR: conf/103428
Approved by: re (hrs)
2006-09-26 01:02 brooks
[r162641] It is possible for bpf to return a length such that:

length != BPF_WORDALIGN(length)

This meeans that it is possible for this to be true:

interface->rbuf_offset > interface->rbuf_len

Handle this case in the test for running out of packets. While
OpenBSD's solution of setting interface->rbuf_len to
BPF_WORDALIGN(length) is safe due to the size of the buffer, I think
this solution results in less hidden assumptions.

This should fix the problem of dhclient running away and consuming 100%
CPU.

PR: bin/102226
Submitted by: Joost Bekkers <joost at jodocus.org>
MFC after: 3 days
2006-09-21 14:29 brooks
[r162522] network_ipv6 also does some interface configuration so require it to run
before starting devd so they don't trip over each other.

PR: conf/103428
2006-09-21 01:44 brooks
[r162490] Introduce a new method ipv6if which attemptes to figure out if an
interface is an IPv6 interface.

Use this method to decide if we should attempt to configure an interface
with an IPv6 address in pccard_ether. The mechanism pccard_ether uses
to do this is unsuited to the task because it assumes the list of
interfaces it is passed is the full list of IPv6 interfaces and makes
decissions based on that. This is at least a step in the right
direction and is probably about as much as we can MFC safely.

PR: conf/103428
MFC after: 3 days
2006-09-20 19:48 brooks
[r162481] Flushing all IPv4 routes when an interface is removed or unconfigured
makes no sense. Remove the undocumented removable_route_flush feature
from pccard_ether.

X-MFC after: never
2006-09-20 19:45 brooks
[r162480] Search the list of up interfaces provided by "ifconfig -ul" instead of
greping for UP in "ifconfig $ifn". This eliminates a dependancy on
/usr.
2006-09-07 17:38 brooks
[r162116] MFC: rev 1.185

Add a new kernel environment variable "boot.netif.mtu" which is used to
set the MTU prior to mounting root via NFS. This is required if the
server supports a higher than default MTU because the client will not
see the responses otherwise.
2006-09-03 17:32 brooks
[r161963] MFC: 1.23

/etc/crontab is similar enough to parse as correct if you run
"crontab /etc/crontab", but not the same format due to the who field.
Add some limited anti-foot-shooting support and refuse to load
/etc/crontab as someone's crontab. Users wishing shoot their foot in
this manner may copy /etc/crontab elsewhere. :)
2006-09-01 16:33 brooks
[r161824] - Document /conf/diskless_remount in the list of special files.
- Note that diskless_remount files may use ".." to support mounts above
the root path.
- Copy dot files when populating directories from /conf. [1]

PR: misc/102724 [1]
Submitted by: Attila Nagy <bra at fsn.hu> [1]
2006-08-30 15:14 brooks
[r161739] When generating localhost entries in /etc/hosts, put localhost before
localhost.domain so that naive users of gethost* think 1:: and 127.0.0.1
are "localhost" which is what the user expects. This also matches the
sample hosts file in src/etc/hosts.
2006-08-22 19:41 brooks
[r161535] MFC: sync /etc/network.subr and /etc/pccard_ether and their defaults
with HEAD. Key changes include:
- Support a limited set of punctuation in interface names ".-/+" by
converting them to '_' before looking up interface variables.
- Don't flush routes by default when an interface is removed.
- Don't generate unnecessicary noise on startup when interfaces are
renamed.

PR: conf/92433
2006-08-18 13:19 brooks
[r161437] Don't try to start interfaces that don't exist.

Reported by: Dominique Goncalves <dominique.goncalves at gmail.com>
2006-08-17 03:03 brooks
[r161386] Introduce a new function, ifexists and use it to avoid attempting to
touch interfaces that don't actually exist in the stop case. In the
process move some IPv4 specific code from ifconfig_down to ipv4_down.

This should solve problems with ifconfig: error messages on boot when
interfaces are renamed.
2006-08-16 17:14 brooks
[r161363] Set removable_route_flush to NO be default. It's clearly the wrong
thing to do in most (all?) cases and certainly should not be the default
now that we're running pccard_ether on all interface creates and
destroys.

MFC after: 3 days
2006-08-09 01:56 brooks
[r161109] Add a new kernel environment variable "boot.netif.mtu" which is used to
set the MTU prior to mounting root via NFS. This is required if the
server supports a higher than default MTU because the client will not
see the responses otherwise.

MFC after: 3 weeks
2006-08-04 21:29 brooks
[r160983] The KAME compatability shims in net_osdep.h have been removed.
2006-08-04 21:27 brooks
[r160981] With exception of the if_name() macro, all definitions in net_osdep.h
were unused or already in if_var.h so add if_name() to if_var.h and
remove net_osdep.h along with all references to it.

Longer term we may want to kill off if_name() entierly since all modern
BSDs have if_xname variables rendering it unnecessicary.
2006-08-04 21:15 brooks
[r160980] Use TAILQ_EMPTY instead of checking if TAILQ_FIRST is NULL.
2006-07-15 02:49 brooks
[r160377] Use TAILQ_FOREACH instead of poking around in the guts of the list
macros.
2006-07-15 02:13 brooks
[r160376] Drop a pointless cast of ifp->if_softc to (struct tap_softc *).
2006-07-15 02:07 brooks
[r160375] Cleanup: Use if_initname to set if_dname, if_dunit, and if_xname instead
of doing it by hand.
2006-07-04 02:02 brooks
[r160093] MFC: removal of dhcp entries
2006-07-04 02:01 brooks
[r160092] Remove dhcp entries. They have been incorrect since the import from
OpenBSD.
2006-05-30 02:10 brooks
[r159058] Remove reference to no longer existant /etc/rc.d/pccard.

PR: conf/98055
Submitted by: thierry herbelot <thierry at herbelot dot com>
2006-05-13 03:42 brooks
[r158509] MFC rev 1.14:

Be more like Windows and Linux and send our hostname in the host-name
option if none is given in the config file. Also add #ifdefd out
support for sending a client ID based on our MAC address.

PR: bin/94743, bin/76401
Submitted by: Frank Behrens <frank at pinky dot sax dot de>
2006-05-13 03:24 brooks
[r158508] MFC revs 1.12-14:

1.12,14 - Use /var/run instead of /etc for temporary files
1.13 - Fix the TIMEOUT case

PR: bin/90518, bin/92187
2006-05-07 23:31 brooks
[r158353] Be more like Windows and Linux and send our hostname in the host-name
option if none is given in the config file. Also add #ifdefd out
support for sending a client ID based on our MAC address.

PR: bin/94743, bin/76401
Submitted by: Frank Behrens <frank at pinky dot sax dot de>
X-MFC after: 6.1-RELEASE
2006-04-21 05:16 brooks
[r157914] MFC rev 1.13:

Fix ypwhich -m. This should be MFC'ed back at least as far as RELENG_5.

Submitted by: Gil Kloepfer <gil@arlut.utexas.edu>
PR: 64445
Approved by: re (scottl)
2006-04-21 05:14 brooks
[r157913] MFC rev 1.13:

Fix ypwhich -m. This should be MFC'ed back at least as far as RELENG_5.

Submitted by: Gil Kloepfer <gil@arlut.utexas.edu>
PR: 64445
Approved by: re (scottl)
2006-04-13 18:34 brooks
[r157737] Spell synchronous with required silent 'h'.

Reported by: ru, ceri
Pointy hat: brooks
2006-04-13 18:27 brooks
[r157736] Add missing _ to $_punct.

Submitted by: Dmitry Pryanishnikov <dmitry at atlantis.dp.ua>
2006-04-13 06:50 brooks
[r157706] Commit the various network interface configutation updates I've been
working on.
1) Make it possible to configure interfaces with certain characters in
their names that aren't valid in shell variables. Currently supported
characters are ".-/+". They are converted into '_' characters.
2) Replace nearly all eval statements in network.subr with a new
function get_if_var which substitues an interface name (after the
translations above) for "IF" in a variable name.
3) Fix list_net_interfaces() in the nodhcp case.
4) Allow the administrator to specify if dhclient should be started
when /etc/rc.d/netif configures the interface or only by devd.
This can be set on both a per interface and system wide basis.

PR: conf/88974 [1,2], conf/92433 [1,2]
2006-02-14 16:41 brooks
[r155685] MFC syscons.c rev 1.441 and documentation:

When SC_DISABLE_KDBKEY or SC_DISABLE_REBOOT are not defined allow the
same behavior to be controlled by the sysctls, hw.syscons.kbd_kbdkey
and hw.syscons.kbd_reboot respectively.

PR: kern/72728
Submitted by: Luca Morettoni <morettoni at libero dot it>
Approved by: re (scottl)
2006-01-26 21:05 brooks
[r154869] Fix rev 1.12.

/tmp may not be writeable yet when dhclient is first run via
/etc/rc.d/netif so using it may not work. Also, writing to a
predictable file in /tmp as root is a really bad idea since a malicious
user may be able to win a race and insert a symlink which will allow
them to cause any file to be overwritten. To solve these problems,
create the tempory file in /var/run which will exist this early and is
writable only by root.

Security: Local risk if users can cause dhclient to run on demand
(such as by unplugging and replugging the network cable).
2006-01-24 06:38 brooks
[r154763] s/predefine/predefined/

PR: gnu/86627
Submitted by: Arne Woerner <arne_woerner at yahoo dot com>
MFC After: 6 days
2006-01-24 06:30 brooks
[r154762] Backout ill considered MFC of 1.20. It doesn't compile.

Pointy hat: brooks
2006-01-24 06:28 brooks
[r154761] MFC revs:
1.17: convert from network byte order to host byte order not to reverse.
1.19: Get rid of the bogus IFP2FC() macro and use IFP2FWC().
1.20: Fix ip_fastforward() return handling.

PR: kern/91307 (1.19)
Found by: Coverity Prevent(tm) (1.20)
2006-01-24 06:11 brooks
[r154760] Give the TIMEOUT case a chance to work by using -t # instead of the
OpenBSD -w # when invoking ping.

PR: bin/92187
Submitted by: "Shin'ya Kumabuchi" <kumabu at t3 dot rim dot or dot jp>
MFC After: 6 days
2006-01-24 06:03 brooks
[r154759] MFC rev 1.5: Mention the -b flag in the SYNOPSIS.
2006-01-24 06:01 brooks
[r154758] MFC rev 1.11:
When we give up on an interface, use the arp(8) command to remove all
entries from the interface rather than using ifconfig's delete command.
This preserves non-dhclient configured addresses (though they are wiped
out when dhclient is restarted).
2006-01-24 05:59 brooks
[r154757] MFC rev 1.13: Allow users to add aliases to the interface.

PR: bin/87465
2006-01-24 05:54 brooks
[r154756] MFC etc/rc.initdiskless rev 1.49: Don't create memory backed file
systems with soft-updates enabled.

PR: conf/85558
Submitted by: Ralf Wenk <RZ dash FreeBSD0605 at hs dash karlsruhe dot de>
2006-01-24 05:49 brooks
[r154755] MFC rev 1.49: Don't create memory backed file systems with soft-updates
enabled, it's pointless and potentially dangerous.

PR: conf/85558
Submitted by: Ralf Wenk <RZ dash FreeBSD0605 at hs dash karlsruhe dot de>
2006-01-24 05:39 brooks
[r154754] MFC the ability to restrict deletions to a particular inteface in the -a
case.
2006-01-14 17:57 brooks
[r154369] When SC_DISABLE_KDBKEY or SC_DISABLE_REBOOT are not defined allow the
same behavior to be controlled by the sysctls, hw.syscons.kbd_kbdkey
and hw.syscons.kbd_reboot respectively.

Apologies to the submitter for taking so long to commit this simple
change.

PR: kern/72728
Submitted by: Luca Morettoni <morettoni at libero dot it>
MFC After: 3 days
2006-01-12 20:27 brooks
[r154269] Belated __FreeBSD_version bump for improvements to the Linux ldconfig
support in etc/rc.d/abi.
2006-01-11 21:59 brooks
[r154239] Creating memory file systems with softupdates enabled is pointless,
don't do it.

PR: conf/85558
Submitted by: Ralf Wenk <RZ dash FreeBSD0605 at hs dash karlsruhe dot de>
MFC after: 5 days
2006-01-11 21:30 brooks
[r154236] Be a little more read-only file system friendly when running the Linux
ldconfig. Build the cache in a temporary directory and only install it
if it's actually different that the installed one.

Also, use "cat tmp > real" to install the temporary file in the real
location to allow the real location to be a symlink to a writable
directory such as /var/run (where the file actually belongs).

MFC After: 5 days
2006-01-11 16:39 brooks
[r154230] MFC rev 1.38: attempt to mount /usr if we need to populate /var and we
can't find mtree.

PR: conf/71757
2006-01-11 06:41 brooks
[r154214] MFC rev 1.6:
When removing the local domain, only do so when the result
will be a host name. This is matches the documented behaviro.
The previous behavior would remove the domain name even if
the result retained a dot.

This fixes rsh connections from a.example.com to example.com.

This is also reported to fix rsh connections from host xxx to
yyy.example.com.

PR: 88301
2006-01-11 06:22 brooks
[r154213] MFC revs 1.46 and 1.47 (from src/etc/rc.initdiskless).

Switch from pax to tar for extracting cpio archives. pax requires a
writable /tmp (or TMPDIR) and thus is unsuitable for this job.

The pax command used was also invalid.

PR: conf/88293
2006-01-11 06:06 brooks
[r154211] MFC:
Allow chkprintcap(8) to be run before lpd is started. Disabled by
default. Default flags create missing directories.

Remove comment about doing this in etc/rc.d/var.

PR: conf/71488
2006-01-11 05:56 brooks
[r154210] MFC revs 1.27 and 1.28.
1.27: don't crash when given an invalid nwkey string
1.28: parse SSIDs correctly and bail immediatly when invalid

PR: misc/88159, bin/74509
2006-01-11 05:37 brooks
[r154209] Get rid of the bogus IFP2FC() macro and use IFP2FWC(). IFP2FC()
attempted to cast a struct ifnet to a struct fw_com which resulted in
data corruption.

PR: kern/91307
Submitted by: Alex Semenyaka <alex at semenyaka do ru>
MFC After: 6 days
2006-01-10 16:15 brooks
[r154191] Update usage to reflect the fact that the -d -a now accepts -i <interface>.

Reminded by: ru
Pointy hat: brooks
2006-01-10 05:56 brooks
[r154166] /etc/crontab is similar enough to parse as correct if you run
"crontab /etc/crontab", but not the same format due to the who field.
Add some limited anti-foot-shooting support and refuse to load
/etc/crontab as someone's crontab. Users wishing shoot their foot in
this manner may copy /etc/crontab elsewhere. :)

MFC After: 1 week
2006-01-10 05:44 brooks
[r154165] Mention the -b flag in the SYNOPSIS.

MFC after: 1 week
2006-01-10 05:33 brooks
[r154164] When we give up on an interface, use the arp(8) command to remove all
entries from the interface rather than using ifconfig's delete command.
This preserves non-dhclient configured addresses (though they are wiped
out when dhclient is restarted).

MFC after: 1 week
2006-01-10 05:17 brooks
[r154162] Allow the -i <interface> option with -d -a to allow all arp entries for
a given interface to be removed.
2006-01-10 04:53 brooks
[r154161] Allow users to add aliases to the interface.

PR: bin/87465 (different solution used)
MFC after: 1 week
2005-12-22 03:47 brooks
[r153630] MFC trimdomain(3) regression tests.
2005-12-22 00:43 brooks
[r153625] Note that an MFC of a fix to trimdomain(3) may result in changes to host
names in log files under certain circumstances.
2005-12-22 00:34 brooks
[r153624] MFC rev 1.6:
When removing the local domain, only do so when the result
will be a host name. This is matches the documented behaviro.
The previous behavior would remove the domain name even if
the result retained a dot.

This fixes rsh connections from a.example.com to example.com.

This is also reported to fix rsh connections from host xxx to
yyy.example.com.

PR: 88301
2005-12-20 21:11 brooks
[r153571] MFC 1.5: we now support BPF write filters unconditionally.
2005-12-20 21:02 brooks
[r153567] MFC 1.12:
When we get a bogus hostname in an option, drop the option
rather than refusing the lease. This allow obtaining leases
on misadministered networks that use host names with
underscores in them.

Reminded by: avatar
2005-12-15 02:13 brooks
[r153434] MFC rev 1.258

remove removable_interfaces example. It's been dead since 6.0-RELEASE,
but I forgot to MFC this.

Reminded by: dougb
Pointy hat: brooks
2005-12-10 03:46 brooks
[r153287] When we get a bogus hostname in an option, drop the option rather than
refusing the lease. This allow obtaining leases on misadministered
networks that use host names with underscores in them.

MFC After: 3 days
2005-12-10 02:24 brooks
[r153286] Sync with HEAD.

- bump the date to represent the addition of ipv4_addrs_<ifn>.
- ru's -mdoc sweep
- bring in documentation of NOAUTO
- remove obsolete documentation of dead removable_interfaces
2005-12-10 02:20 brooks
[r153285] MFC the addition of the ipv4_addrs_<ifn> which allows the addition of
multiple aliases using ranges and CIDR notation.

Submitted by: Philipp Wuensche <cryx dash freebsd at h3q dot com>
2005-12-10 02:14 brooks
[r153284] MFC revs 1.46-1.48 (sync with HEAD):
1.46, 1.47: switch from pax to tar for cpio archive handling.
1.48: umount in reverse order to allow nested mounts.
2005-12-10 02:11 brooks
[r153283] MFC 1.47 and 1.48:

1.47: Be less IPv4 centric. When checking if the interface is
already configured, check if the UP flag is set instead of
checking for the netmask keyword.

1.48: Use more rc.subr bits to clean up pccard_ether and implement
new features. Both the presence of a NOAUTO keyword and an
interface being up can be ignored is the forcestart option
is used. Additionally, a restart option has been added.
2005-12-10 02:06 brooks
[r153282] MFC rev 1.11: remove dhclient dependancy
2005-12-03 01:33 brooks
[r153035] Don't bogusly depend on dhclient. It's now run either by
/etc/rc.d/netif or from devd rather than by the startup scripts.
2005-11-14 23:34 brooks
[r152441] Add a new configuration variable, ipv4_addrs_<ifn>, which adds one or
more IPv4 address from a ranged list in CIRD notation:

ipv4_addrs_ed0="192.168.0.1/24 192.168.1.1-5/28"

In the process move alias processing into new ipv4_up/down functions to
more toward a less IPv4 centric world.

Submitted by: Philipp Wuensche <cryx dash freebsd at h3q dot com>
2005-11-01 01:36 brooks
[r151916] Add items to unmount to the front of the list so they are unmounted in
reverse (thus allowing /conf to be unmounted).
2005-10-31 22:00 brooks
[r151914] Switch from pax to tar for extracting cpio archives. pax requires a
writable /tmp (or TMPDIR) and thus is unsuitable for this job.

Tested by: Joerg Pulz <Joerg dot Pulz at frm2 dot tum dot de>
PR: conf/88293
2005-10-31 21:07 brooks
[r151908] The -x <format> option of pax is for creation of archives, not
extraction.

This will allow cpio archive support to work, at least in situations
where /tmp is writable. Because pax requires a writable /tmp it is
unsuitable for this task, but replacing it will come in a later commit.

Submitted by: Joerg Pulz <Joerg dot Pulz at frm2 dot tum dot de>
PR: conf/88293
2005-10-30 16:31 brooks
[r151883] SSIDs are of length 32-bytes and not NUL terminated. Use the correct
length and bail immediatly if the returned length is wrong rather than
attempting to set an correct value.

This commit differs from the patch in the PR in the use of exit instead
of return and the use of a defined value for the array.

Submitted by: Daan Vreeken [PA4DAN] <Danovitsch at Vitsch dot net>
PR: bin/74509
2005-10-28 22:57 brooks
[r151831] MFC rev 1.6:

Don't print anything if we can't do any localpkg shutdown (start already
does this).

Submitted by: Andre Albsmeier <Andre dot Albsmeier at siemens dot com>
PR: conf/86606
2005-10-28 22:47 brooks
[r151829] s/insure/ensure/ in previous commit. My dictionary and m-w.com say they
are synonymous, but ensure seems slightly closer and does not have the
connotation of buying insurance.

Reported by: Jason McIntyre <jmc at kerhand dot co dot uk>
2005-10-28 21:57 brooks
[r151827] Don't crash when given an invalid nwkey string.

PR: misc/88159
2005-10-20 19:23 brooks
[r151523] Mention the possibility of non-interactive scripts for EDITOR and add a
BUGS section mentioning the requirement that such scripts run "sleep 1"
or equivalent to work reliably.
2005-10-19 21:21 brooks
[r151486] Make devd WARNS=4 clean and bump WARNS accordingly. This will insure
that future variable shadowing bugs don't compile.

Reviewed by: imp
Compiled on: alpha i386 sparc64
2005-10-07 17:20 brooks
[r151067] MFC rev 1.6:

Don't print anything if we can't do any localpkg shutdown (start already
does this).

Submitted by: Andre Albsmeier <Andre dot Albsmeier at siemens dot com>
PR: conf/86606
Approved by: re (kensmith)
2005-10-05 04:46 brooks
[r150956] Add regression tests for trimdomain(3).
2005-10-05 04:42 brooks
[r150955] When removing the local domain, only do so when the result will be a
host name. This is matches the documented behaviro. The previous
behavior would remove the domain name even if the result retained a dot.

This fixes rsh connections from a.example.com to example.com.

Reviewed by: ceri (at least the concept)
2005-10-03 18:20 brooks
[r150882] Use more rc.subr bits to clean up pccard_ether and implement new
features. Both the presence of a NOAUTO keyword and an interface being
up can be ignored is the forcestart option is used. Additionally, a
restart option has been added.

Reviewed by: ume
2005-09-28 19:59 brooks
[r150702] Be less IPv4 centric. When checking if the interface is already
configured, check if the UP flag is set instead of checking for the
netmask keyword.
2005-09-27 18:52 brooks
[r150643] s/routnes/routines/ in help message.

This file is already off the vendor branch and there hasn't been a bc
release in more than 4 years so I can't see any harm in fixing this.

Submitted by: Arne Woerner <arne_woerner at yahoo dot com>
PR: gnu/86627
2005-09-27 02:05 brooks
[r150603] Don't print anything if we can't do any localpkg shutdown (start already
does this).

Submitted by: Andre Albsmeier <Andre dot Albsmeier at siemens dot com>
PR: conf/86606
2005-09-10 17:05 brooks
[r149947] Sync network interface startup code with HEAD.

etc/network.subr: 1.165-1.167
- kill removable_interfaces
- add NOAUTO keyword
- fix quoted entries in ifconfig_<ifn> variables
- always up the interface
- if ifconfig_<ifn> is defined, but empty, don't set it to
ifconfig_DEFAULT
- always configure lo0 first if it exists
etc/pccard_ether: 1.46
- kill removable_interfaces
- add NOAUTO keyword
etc/rc.d/netif: 1.15-1.17
- block Ctrl-C in dhclient
- minor cleanup of the interface list generation code

Approved by: re (scottl)
Requested by: many (death to removable_interfaces!)
2005-09-10 17:02 brooks
[r149946] MFC 1.43 and 1.44:
- Add new template sources to keep /conf more organized.

Approved by: re (scottl)
2005-09-10 17:01 brooks
[r149945] Sync dhclient with HEAD with exception of bpf.c:1.5.

sbin/dhclient/*.c:
- add __FBSDID
sbin/dhclient/conflex.c: 1.3-1.4
- fix a minor buffer overflow in config parsing
- actually support backslash escaping in config files
bin/dhclient/dhclient-script: 1.6-1.10
- don't create or remove routes to our address through 127.0.0.1
- improved detection of the interface of the default route
- support quotes in the medium string
- clear interface state more effectively on failure
- don't update resolv.conf when unchanged
bin/dhclient/dhclient.c: 1.10-1.11
- validate domain-name (used as resolv.conf search string)
- handle superseded subnet-mask correctly

Approved by: re (scottl)
2005-09-08 22:49 brooks
[r149898] Avoid updating resolv.conf when no changes have actually occured.

Submitted by: ume
2005-09-08 22:15 brooks
[r149896] When we fail to aquire a lease, our lease expires without a sucessful
renewal, or we lose link, be more forceful about clearing interface
state so another interface that connects to the same network has a
chance of working. This doesn't address attemping to connect to both at
once, but appears to allow unplugging from a wired interface and then
inserting a wireless card that associates with an AP bridged to the same
LAN.
2005-09-02 18:30 brooks
[r149730] Actually block Ctrl-C (SIGINT=2).

Reported by: sam
Pointy hat to: brooks
2005-09-02 17:35 brooks
[r149727] When we supersed the subnet-mask, write the forced value to the lease
file. This is what the ISC client does.

Submitted by: Rostislav Krasny <rosti dot bsd at gmail dot com>
2005-09-02 17:11 brooks
[r149726] - Alwasy explicitly bring the interface up before configuring it.
- If an interface's ifconfig_<ifn> is set, but empty, don't set it to
ifconfig_DEFAULT. This way interfaces can be disabled even in the
presence of ifconfig_DEFAULT.
- When listing interfaces and network_interfaces=auto, place lo0 first
if it's around.
2005-09-02 17:05 brooks
[r149725] Block SIGQUIT (Ctrl-C) while running in startup mode. This should allow
dhclient's to be killed without stopping all boot progress.

Minor cleanup of the interface list generation code.
2005-08-30 18:20 brooks
[r149639] Introduce a new helper function check_search() derived for res_hnok to
check the domain-name parameter according to the rules for "search"
strings as documented in resolv.conf(5). Specifically, the string must
be no more than 256 bytes long and contain no more than six valid domain
names separated by white space.

The previous unchecked values could result in a mangled resolv.conf
file which could effectively deny access to local sites. This is not
a security issue as rogue dhcp servers could already do this without
sending invalid strings.

Reviewed by: cperciva
MFC After: 3 days
2005-08-26 20:31 brooks
[r149519] The $medium string often contains quoted values with spaces in them (ssids,
for example). Follow the example of the ISC script and wrap ifconfig
calls using $medium in eval "..." so this works.

Reported by: iedowse
2005-08-26 04:06 brooks
[r149483] Support ifconfig_<ifn> variables containing quoted variables with spaces
in them by wrapping the ifconfig command with eval "...".

For example, this allows:

ifconfig_iwi0="DHCP ssid 'foo bar baz'"
2005-08-26 01:25 brooks
[r149481] In read_string(), when the last character was a backslash, unincrement
the output index instead of keeping what ever trash was in the buffer.

Reported by: iedowse
2005-08-26 01:07 brooks
[r149480] Use a more robust, grep-free command to get the interface of the current
default route.

Submitted by: Rostislav Krasny <rosti dot bsd at gmail dot com>
2005-08-26 01:02 brooks
[r149479] Don't and/remove a route to our assigned IP through 127.0.0.1. It
serves no apparent purpose (we commented this out ages ago in the ISC
scripts) and cases problems with some ADSL setups.

Reported by: Rostislav Krasny <rosti dot bsd at gmail dot com>
2005-08-24 01:23 brooks
[r149401] - Remove the removable_interfaces variable. /etc/pccard_ether will
now run on any interface.
- Add a new ifconfig_<ifn> keyword, NOAUTO which prevents configuration
of an interface at boot or via /etc/pccard_ether. This allows
/etc/rc.d/netif to be used to start and stop an interface on a purely
manual basis. The decision to affect pccard_ether may be revisited at
a later date.

Requested by: imp, gallatin (removable_interfaces)
Discussed with: sam, Randy Bush (NOAUTO)
2005-08-24 00:05 brooks
[r149400] MFOpenBSD rev 1.9: fix a buffer overflow when processing config file
lines that are exactly 81 characters in length.

Obtained from: OpenBSD
MFC After: 3 days
2005-08-23 23:59 brooks
[r149399] Add __FBSDID to all .c files in dhclient to aid in determining file
versions when dealing with user problems.
2005-08-21 00:43 brooks
[r149344] MFC rev 1.28, removal of pointless -I.. from CFLAGS.

PR: bin/79229
2005-08-20 04:35 brooks
[r149311] MFC rev 1.9, remove an incorrect and unnecessicary bzero of the softc.

Approved by: re (kensmith)
2005-08-20 04:34 brooks
[r149310] MFC rev 1.81, use if_free_type to avoid leaking struct arpcoms.

Approved by: re (kensmith)
2005-08-20 04:32 brooks
[r149309] MFC rev 1.244, removal of if_findindex and correction of two if_alloc
calls before an if_attach.

Partial MFC of 1.243. The change it was a response to has not been
MFCd, but the comment is relevent without it.

Approved by: re (kensmith)
2005-08-20 04:30 brooks
[r149308] MFC revs 1.257 and 1.216.
- interface startup doc updates
- s/ISC dhclient/OpenBSD dhclient/

Approved by: re (kensmith)
2005-08-18 19:39 brooks
[r149252] Refer to the OpenBSD dhclient instead of the ISC dhclient.

Reported by: Bob Johnson <bob89 at eng dot ufl dot edu>
PR: 84963
2005-08-18 18:36 brooks
[r149243] When we started calling if_findindex() from if_alloc() with an empty
struct ifnet most of if_findindex() become a complex no-op. Remove it
and replace it with a corrected version of the four line for loop it
devolved to plus some error handling. This should probably be replaced
with subr_unit at some point.

Switch from checking ifaddr_byindex to ifnet_byindex when looking for
empty indexes. Since we're doing this from if_alloc/if_free, we can
only be sure that ifnet_byindex will be correct. This fixes panics when
loading the ef(4) module. The panics were caused by the fact that
if_alloc was called four time before if_attach was called and thus
ifaddr_byindex was not set and the same unit was allocated again. This
in turn caused the first if_attach to fail because the ifp was not the
one in ifnet_byindex(ifp->if_index).

Reported by: "Wojciech A. Koszek" <dunstan at freebsd dot czest dot pl>
PR: kern/84987
MFC After: 1 day
2005-08-17 18:53 brooks
[r149208] Remove an unnecessicary bzero that was zeroing the softc's struct ifnet
pointer.

Submitted by: brueffer
Reviewed by: dfr
2005-08-17 00:28 brooks
[r149170] Add two new template sources, /conf/bcast/${ipbca} and /conf/ip/${ip}.
These allow large installations to keep their /conf directory down to a
managable number of entries.

Clean up the handling of dhcp_cookie.
2005-08-16 17:02 brooks
[r149141] - Move IF_ADDR_LOCK_DESTROY(ifp) from if_free to if_free_type.
- Add a note that additions should be made to if_free_type and not
if_free to help avoid this in the future.

This apparently fixes a use after free in if_bridge and may fix bugs
in other direct if_free_type consumers.

Reported by: thompsa
2005-08-15 20:27 brooks
[r149110] Vlan interfaces change their type after ether_ifattach() so we needs to
use if_free_type(ifp, IFT_ETHER) to delete them and stop leaking struct
arpcoms.

Reported by: thompsa
MFC After: 3 days
2005-08-02 02:24 brooks
[r148631] - Document network_interfaces=auto (the default!).
- Remove documentation of pccard_ifconfig as it is now gone.
- Document pccard_ifconfig's replacement ifconfig_DEFAULT.
2005-07-29 23:36 brooks
[r148550] Sync dhclient with HEAD:

- Don't complain when debouncing dhclient startup.
- Fix buffer handling in reveive_packet(). This fixes infinite cpu
eating loops and probably some crashes.
- Spell if_defaultroute route correctly in dhclient-script so we
are allowed to change the default route.
- Document dhclient -b.
- Treat reassociation like association.
- Do not force server-name to be a valid domain name.
- Handle servers that send NUL-terminated host-name options.

Approved by: re (scottl)
2005-07-28 15:30 brooks
[r148484] Further fix receive_packet() by using BPF_WORDALIGN to insure the offset
is properly aligned when we move to the next packet.

Obtained from: ISC dhclient via krw at OpenBSD
2005-07-28 00:24 brooks
[r148465] Don't reject packets with server names containing characters that are
not allowed in domain names. RFC 2132 does not list valid or invalid
characters and the ISC client accepts anything here.

Reported by: ps
2005-07-27 19:25 brooks
[r148451] Fix a bug in the handling of cases where we got a short (or zero)
capture. Zero length captures caused an infinte loop and short captures
probably caused memory corruption and a crash.

Reported by: many
MFC After: 3 days
2005-07-26 18:27 brooks
[r148420] s/if_defaulroute/if_defaultroute/

We may now handle route updates better.
2005-07-26 00:37 brooks
[r148415] Silence the de-bouncing of dhclient start up. The previous output
caused significant mental anguish for some portions of the user
population. :)
2005-07-25 22:19 brooks
[r148411] Change host-name from type "X" to type "t". This allows the client to
accept NUL-terminated strings as required by RFC 2132.

This solution is not perfect as it removes the ability to send
NUL-terminated host-name options which may be required by some broken
servers. Given the current lack of an existance proof of such servers
and the fact that servers that send NUL-terminated domain names do
exist, this seems like an acceptable compromise. A discussion of these
issues can be found at:

http://marc.theaimsgroup.com/?l=dhcp-client&m=96837107208382&w=2

PR: bin/83468
Reported by: Sean Winn <sean at gothic dot net dot au>
MFC-after: 3 days
2005-06-30 17:50 brooks
[r147704] Remove REQUIRE and BEFORE lines since this script is not run by rcorder
at startup. Instead it is called by other scripts.

Approved by: re (network interface startup blanket)
2005-06-30 05:50 brooks
[r147689] Don't complain when we receive smtp, pop, nntp, www, finger, and irc
server options.

Reported by: Max Boyarov <max_b at tut dot by>
Approved by: re (dhclient blanket)
2005-06-30 05:32 brooks
[r147686] People like to do RFC violating things with the domain-name option and
include a space seperated list of domains instead of the domain of the
host. This is supported on too many platforms to break for now so,
remove validation of this option for the moment.

The correct solution longer term is to implement RFC 3397 support and
then treat domain-name options containing space seperated lists of
domains as domain-search options for backwards compatability.

Approved by: re (dhclient blanket)
2005-06-30 05:02 brooks
[r147684] - Remove the pccard_ifconfig variable in favor of a new
ifconfig_DEFAULT variable. Unlike pccard_ifconfig, ifconfig_DEFAULT
applies to all interfaces that do not specify an ifconfig_<ifn>
variable rather than just those listed in removable_interfaces.
- Correct the list of interfaces when network_interfaces and
removable_interfaces are both set by including removable_interfaces
in the list of canidates.
- When listing dhcp interfaces, include those with other ifconfig
options so nat works.

Approved by: re (network interface startup blanket)
2005-06-30 04:52 brooks
[r147682] Add support for starting wpa_supplicant by adding the WPA keyword to an
interface's ifconfig_<ifn> entry in /etc/rc.conf.

Approved by: re (network interface startup blanket)
2005-06-30 04:46 brooks
[r147681] When interfaces are given on the command line, don't attempt to filter
them. Just try to run the given command on them. We need to be able to
run stop functions on interfaces that have been deleted to stop
wpa_supplicant.

Approved by: re (interface startup blanket)
2005-06-29 01:46 brooks
[r147651] Remove isc-dhcp files from HEAD now that we're using the OpenBSD
dhclient.

Reminded by: ru
Approved by: re (blanket dhclient)
2005-06-17 23:54 brooks
[r147472] Fix a number of now bogus casts I missed the first time around (I must
have failed to grep this file properly).

Reported by: Hans Petter Selasky <hselasky at c2i dot net>
Approved by: re (ifnet blanket)
2005-06-17 17:19 brooks
[r147470] Spelling/grammer fixes in comment.

Reported by: Hans Petter Selasky <hselasky at c2i dot net>
Approved by: re (ifnet blanked)
2005-06-17 17:14 brooks
[r147469] Refer to the correct file/function in panic messages.

Reported by: Hans Petter Selasky <hselasky at c2i dot net>
Approved by: re (ifnet blanket)
2005-06-16 18:08 brooks
[r147424] Make sure we actually read the config files before testing values from
them.

Reported by: Darren Pilgrim <dmp at bitfreak dot org>
PR: conf/82313
Approved by: re (network interface startup blanket)
2005-06-15 20:23 brooks
[r147405] Fix four casts of the softc to a struct ifnet.

Hopefully this fixes ed(4) under qemu. I'm shocked that real hardware
is apparently working with these bugs.

Approved by: re (ifnet blanket)
Pointy hat: brooks
2005-06-14 17:47 brooks
[r147386] Move if_alloc() higher in the attach function so sc->ifp is populated
early. I've moved it all the way to the top rather than part way up as
the submitter did.

Submitted by: Jung-uk Kim <jkim at niksun dot com>
Reported by: submitter, le, dougb
Approved by: re (ifnet blanket)
2005-06-14 17:38 brooks
[r147382] Remove a duplicate assignment.

Reported by: delphij
Approved by: re (ifnet blanket)
2005-06-13 23:43 brooks
[r147351] Avoid a null pointer dereference by not expiring our lease if we don't
have an active one.

Submitted by: sam
Reported by: James Snow <snow at teardrop dot org>
Approved by: re (dhclient blanket)
2005-06-13 17:17 brooks
[r147346] Initialze ifp->if_softc.

Submitted by: ume
2005-06-12 16:21 brooks
[r147318] Build on pc98.
2005-06-12 15:33 brooks
[r147317] Stop the interface before detaching and freeing it, rather than after.

Reported by: marius
2005-06-12 00:53 brooks
[r147308] Return NULL instead of a bogus pointer from if_alloc when if_com_alloc
fails.

Move detaching the ifnet from the ifindex_table into if_free so we can
both keep the sanity checks and actually delete the ifnets. [0]

Reported by: gallatin [0]
Approved by: re (blanket)
2005-06-12 00:45 brooks
[r147306] Fix IPv6 neighbor discovery by using IF_LLADDR to get the mac address
instead of a particularly ugly cast + pointer math hack.

Reported by: kuriyama, kris
2005-06-11 16:30 brooks
[r147304] s/sc_enaddr/enaddr/ to fix pc98 build.

Approved by: re (blanket)
2005-06-11 01:37 brooks
[r147291] Move if_alloc() up so it's before mii_phy_probe().
2005-06-11 00:47 brooks
[r147289] Move if_alloc() up in fxp_attach() so there's an ifp before
mii_phy_probe() is called.

Committed via: fxp0
2005-06-11 00:20 brooks
[r147287] Add an evil hack to work around a cast from the softc to the ifnet that
I missed. Since I did no rearrange any softcs, casting the result of
device_get_softc() to (struct ifnet **) and derefrencing it yeilds a
pointer to the ifp. This makes at least vr(4) nics work.
2005-06-10 20:42 brooks
[r147269] Revert the unnecessicary addition of some braces in fxp_attach(). Don't
explicitly free the ifp in fxp_detach(), the call to fxp_release() takes
care of it.
2005-06-10 16:51 brooks
[r147257] Mention interface API changes.
2005-06-10 16:49 brooks
[r147256] Stop embedding struct ifnet at the top of driver softcs. Instead the
struct ifnet or the layer 2 common structure it was embedded in have
been replaced with a struct ifnet pointer to be filled by a call to the
new function, if_alloc(). The layer 2 common structure is also allocated
via if_alloc() based on the interface type. It is hung off the new
struct ifnet member, if_l2com.

This change removes the size of these structures from the kernel ABI and
will allow us to better manage them as interfaces come and go.

Other changes of note:
- Struct arpcom is no longer referenced in normal interface code.
Instead the Ethernet address is accessed via the IFP2ENADDR() macro.
To enforce this ac_enaddr has been renamed to _ac_enaddr.
- The second argument to ether_ifattach is now always the mac address
from driver private storage rather than sometimes being ac_enaddr.

Reviewed by: sobomax, sam
2005-06-10 03:41 brooks
[r147218] Add support for /etc/dhclient-enter-hooks to match /etc/dhclient-exit-hooks.
2005-06-08 18:21 brooks
[r147138] Add support for /etc/dhclient-exit-hooks file.

Tested by: Max Boyarov <max_b at tut dot by>
2005-06-08 00:15 brooks
[r147123] Update to the new world of network configuration. The documenation is
slightly ahead of reality as rc support for WPA is only partialy in
place.

Submitted by: Jeremie Le Hen <jeremie at le-hen dot org>
2005-06-08 00:05 brooks
[r147122] Remove default and documenation for pccard_ether_delay since I removed
it from /etc/pccard_ether.

Submitted by: Jeremie Le Hen <jeremie at le-hen dot org>
2005-06-07 23:59 brooks
[r147121] Fix return values of ifconfig_up/down.

Reported by: Andrea Campi
2005-06-07 18:07 brooks
[r147109] Mention the import of the OpenBSD dhclient, particularly the requirement
of running devd and the loss of DNS update functionality.

Reminded by: sobomax
2005-06-07 14:47 brooks
[r147106] Fix build on 64-bit platforms where time_t is 64 bit. Since where
talking about time related to leases, it should be OK to cast these to
(int)s rather than using intmax_t.

Submitted by: ru
Pointy hat: brooks
2005-06-07 04:51 brooks
[r147090] Fix path to dhclient-script and reconnect to build.
2005-06-07 04:49 brooks
[r147089] Now that dhclient builds and /etc/support has been committed, reconnect
dhclient to the build.
2005-06-07 04:49 brooks
[r147088] Support code for the OpenBSD dhclient. This significantly changes the
way interfaces are configured. Some key points:

- At startup, all interfaces are configured through /etc/rc.d/netif.
- ifconfig_<if> variables my now mix real ifconfig commands the with
DHCP and WPA directives. For example, this allows media
configuration prior to running dhclient.
- /etc/rc.d/dhclient is not run at startup except by netif to start
dhclient on specific interfaces.
- /etc/pccard_ether calls "/etc/rc.d/netif start <if>" to do most of
it's work.
- /etc/pccard_ether no longer takes additional arguments to pass to
ifconfig. Instead, ifconfig_<if> variables are now honored in favor
of pccard_ifconfig when available.
- /etc/pccard_ether will only run on interfaces specified in
removable_interfaces, even if pccard_ifconfig is set.
2005-06-07 04:35 brooks
[r147087] Use SCRIPTS to install dhclient-script instead of rolling our own.
2005-06-07 04:32 brooks
[r147086] FreeBSDize the dhclient-script.

Submitted by: sam
2005-06-07 04:26 brooks
[r147085] Define _PATH_VAREMPTY.

Add a -b option to background immediatly.

Add support for 802.11 routing messages to "instantly" renegotiate
at lease when we associate with a new network.

Submitted by: sam
2005-06-07 04:14 brooks
[r147077] NULL is not a valid time_t.

Submitted by: sam
2005-06-07 04:13 brooks
[r147076] We don't support BPF write filters at this time.

Submitted by: sam
2005-06-07 04:08 brooks
[r147075] Take the OpenBSD version of the Makefile to resolve conflicts as the old
one is irrelevent. Add $FreeBSD.
2005-06-07 04:05 brooks
[r147073] This commit was generated by cvs2svn to compensate for changes in r147072,
which included commits to RCS files with non-trunk default branches.
2005-06-07 04:05 brooks
[r147072] Import the OpenBSD dhclient as shipped with OpenBSD-3.7 (the tag
OPENBSD_3_7).
2005-06-07 04:02 brooks
[r147071] Remove the bits to build the isc-dhclient.
2005-06-07 03:59 brooks
[r147070] Temporarily disconnect dhclient from the build while I import the
OpenBSD version.
2005-06-06 20:19 brooks
[r147062] Add _dhcp user/group as required by the OpenBSD dhclient.
2005-06-06 19:08 brooks
[r147059] Send link state change notifications to /dev/devctl. This is needed to
start the OpenBSD dhclient when links come up.
2005-06-06 02:12 brooks
[r147038] Remove some variables the last commit stopped using so the code compiles.
2005-06-06 00:21 brooks
[r147036] Add missing Va in ifi_epoch entry.
2005-05-28 06:11 brooks
[r146702] Update refrenced URL for SNMP list of ifTypes to refer to iana.org
instead of a dead location on ftp.isi.edu.
2005-04-29 23:02 brooks
[r145693] To allow /etc to be as minimal as possible in a diskless setup, we need
to run initdiskless before we run rcorder on /etc/rc.d. To allow this,
move /etc/rc.d/initdiskless to /etc/rc.initdiskless and run it directly
from /etc/rc.

Remove /etc/rc.d/preseedrandom as it is no longer necessicary (we start
with entropy unblocked) and was only used by initdiskless when it
was needed.

Discussed on: freebsd-rc
Repocopy by: peter
2005-04-29 15:07 brooks
[r145682] Remove the Netgear WG311 from the list. The v1 card is supported, but
the v2 card is a TI. Since we're not attempting to keep this list
complete, removing this is best.

Reported by: Brian Candler <B dot Candler at pobox dot com>
Evan Dower <evantd at hotmail dot com>
MFC After: 1 day
2005-04-26 20:22 brooks
[r145567] Fix a the previous commit. I wanted to remove the if and always run the
body not remove both.

Reported by: ceri
Pointy hat: brooks
2005-04-26 20:00 brooks
[r145566] Don't force IPv6 proto to be printed numericaly.

Noticed by: ceri
2005-04-26 18:10 brooks
[r145565] Introduce a struct icmphdr which contains the type, code, and cksum
fields of an ICMP packet.

Use this to allow ipfw to pullup only these values since it does not use
the rest of the packet and it was failed on ICMP packets because they
were not long enough.

struct icmp should probably be modified to use these at some point, but
that will break a fair bit of code so it can wait for another day.

On the off chance that adding this struct breaks something in ports,
bump __FreeBSD_version.

Reported by: Randy Bush <randy at psg dot com>
Tested by: Randy Bush <randy at psg dot com>
2005-04-18 18:35 brooks
[r145246] Add IPv6 support to IPFW and Dummynet.

Submitted by: Mariano Tortoriello and Raffaele De Lorenzo (via luigi)
2005-04-15 00:47 brooks
[r145093] Centralized finding the protocol header in IP packets in preperation for
IPv6 support. The header in IPv6 is more complex then in IPv4 so we
want to handle skipping over it in one location.

Submitted by: Mariano Tortoriello and Raffaele De Lorenzo (via luigi)
2005-04-12 00:48 brooks
[r144919] Back out previous commit (MFC of removal of if_name() function) because
it breaks the ABI for at least two kernel modules. The PR is still
fixed because the real problem is only in 4.x.
2005-04-12 00:05 brooks
[r144916] MFC:
Since net/net_osdep.c contained only one function that could be
trivially implemented as a macro, do that and remove it. NetBSD did
this quite a while ago.

PR: kern/33201
Prodded by: Matteo Riondato <rionda at gufi dot org>
2005-04-07 20:37 brooks
[r144757] When accessing the sysctl vfs.nfs.iodmax, don't report errors as being
from accessing vfs.nfs.iodmin.
2005-04-06 00:26 brooks
[r144693] Use ACTION_PTR(r) instead of (r->cmd + r->act_ofs).

Reviewed by: md5
2005-04-05 23:38 brooks
[r144691] Make dummynet_flush() match its prototype.
2005-04-05 22:26 brooks
[r144687] Be more specific when complaining about bit masks.
2005-04-03 05:21 brooks
[r144566] Don't init ifp->if_addrhead, if_attach() does it for us.
2005-03-25 22:08 brooks
[r144130] Remove bogus (but harmless) -I.. from CFLAGS. It makes no difference to
.depends other then the commant line.

Also remove -g from CFLAGS. The user should add it to CFLAGS if they
desire debug support.

Reviewed by: ru (in concept)
MFC After: 7 days
2005-03-24 19:26 brooks
[r144075] Use the CTASSERT() macro instead of rolling my own, non-portable one
using #error.

Suggested by: jhb
2005-03-24 17:51 brooks
[r144070] Compile errors are way more useful then panics later.

Replace a KASSERT of LINUX_IFNAMSIZ == IFNAMSIZ with a preprocessor
check and #error message. This will prevent nasty suprises if users
change IFNAMSIZ without updating the linux code appropriatly.
2005-03-03 15:47 brooks
[r143106] Add the -N option to not rotate any files. This is to be used in
cojunction with -C and is used by /etc/rc.d/newsyslog.

I forgot that this was in my perforce tree and not my running system and
thus committed a non-working newsyslog script.

Reported by: des
Pointy hat: brooks
2005-03-02 19:03 brooks
[r143048] It is sufficent to require rcconf rather than initdiskless.
2005-03-02 18:44 brooks
[r143047] Document the newsyslog_* and chkprintcap_* variables I added yesterday.
2005-03-02 16:41 brooks
[r143045] Remove stray else.

Reported by: Tai-hwa Liang <avatar at mmlab dot cse dot yzu dot edu dot tw>
Point hat: brooks
2005-03-02 02:46 brooks
[r142969] Allow chkprintcap(8) to be run before lpd is started. Disabled by
default for now. Default flags create missing directories.

Remove comment about doing this in etc/rc.d/var.

Unlike in the PR, I chose to do this in the lpd script where we reliably
have /usr available.

PR: conf/71488
Submitted by: RZ-FreeBSD0904 at fh-karlsruhe dot de
2005-03-02 01:38 brooks
[r142966] MFC 1.22: When ukbd0 arrives, attach to /dev/ukbd0 rather then /dev/kbd1
since kbd1 might be something else.
2005-03-02 00:58 brooks
[r142965] If we don't have /usr/sbin/mtree, try to mount /usr. We're only likely
to hit this case when /usr is remote and thus hasn't been mounted (since
you're supposed to have /var before mounting remote file systems).
Normal machines that don't have a /var for some reason will have /usr
already available because it's local.
2005-03-02 00:44 brooks
[r142963] Remove a number of entries from the distribution target that are
automaticly created at boot. There's no need to maintain a list of
files and permissions in multiple places. This also means binary
updates won't stomp on log files.

For the record, utmp is created in etc/rc.d/cleanvar, wtmp and lastlog
in etc/rc.d/var, and the reset via etc/rc.d/newsyslog.
2005-03-02 00:40 brooks
[r142962] - Update etc/rc.d/newsyslog to FreeBSD standards and install it.
- Enable it by default, running newsyslog with -CN which creates files
that have the C flag specified in /etc/newsyslog.conf.
- Remove the "newsyslog -CC" call from etc/rc.d/var and the check for
newsyslog.
- Add the C flag to entries in /etc/newsyslog.conf that are currently
installed as part of the base system.

There are two effects from this change:
- Users who delete default syslog files to stop logging to them
will need to set newsyslog_enable=NO in rc.conf or remove the C
flag from those file in /etc/newsyslog.conf or they will come back
on the next boot.
- Diskless systems now create the same set of files that ordinary
systems have by default instead of every file in newsyslog.conf.
2005-03-01 22:08 brooks
[r142957] - Remove the dependency of /usr/bin/touch by using "cp /dev/null <target>"
to create /var/log/lastlog.
- Also create /var/log/wtmp if missing.
- Attempt to create these files unless populate_var is NO rather then
only when /var is empty or populate_var=YES.
2005-02-28 22:03 brooks
[r142835] MFC the change of ifi_epoch to be uptime instead of wall clock time.
The origional commit message follows:

Change the definition of struct if_data's member ifi_epoch from
wall clock time to uptime because wall clock time may go
backwards.

This is a change in the API which will impact SNMP agents who
are using ifi_epoch to set RFC2233's ifCounterDiscontinuityTime.
None are known to exist today. This will not impact applications
that are using the <index, epoch> tuple to verify interface
uniqueness except that it eliminates a race which could lead to
a false assumption of uniqueness.

Discussed with: re (jhb, scottl)
Pointed out by: pkh (way back at EuroBSDCon)
Pointy hat: brooks
2005-02-25 19:46 brooks
[r142501] Change the definition of struct if_data's member ifi_epoch from wall
clock time to uptime because wall clock time may go backwards.

This is a change in the API which will impact SNMP agents who are using
ifi_epoch to set RFC2233's ifCounterDiscontinuityTime. None are know to
exist today. This will not impact applications that are using the
<index, epoch> tuple to verify interface uniqueness except that it
eliminates a race which could lead to a false assumption of uniqueness.

Because this is a behavior change, bump __FreeBSD_version.

Discussed with: re (jhb, scottl)
MFC after: 3 days
Pointed out by: pkh (way back at EuroBSDCon)
Pointy hat: brooks
2005-01-15 01:53 brooks
[r140272] Note the deprecation of the abbreviation of a number of ipfw options.
2005-01-15 01:46 brooks
[r140271] Deprecate unmaintainable uses of strncmp to implement abbreviations.
This commit replaces those with two new functions that simplify the code
and produce warnings that the syntax is deprecated. A small number of
sensible abbreviations may be explicitly added based on user feedback.

There were previously three types of strncmp use in ipfw:
- Most commonly, strncmp(av, "string", sizeof(av)) was used to allow av
to match string or any shortened form of it. I have replaced this
with a new function _substrcmp(av, "string") which returns 0 if av
is a substring of "string", but emits a warning if av is not exactly
"string".

- The next type was two instances of strncmp(av, "by", 2) which allowed
the abbreviation of bytes to "by", "byt", etc. Unfortunately, it
also supported "bykHUygh&*g&*7*ui". I added a second new function
_substrcmp2(av, "by", "bytes") which acts like the strncmp did, but
complains if the user doesn't spell out the word "bytes".

- There is also one correct use of strncmp to match "table(" which might
have another token after it without a space.

Since I changed all the lines anyway, I also fixed the treatment of
strncmp's return as a boolean in many cases. I also modified a few
strcmp cases as well to be fully consistent.
2005-01-08 06:02 brooks
[r139898] Change the USB keyboard example to use /dev/ukbd0 instead of /dev/kbd1.

This example caused me to incorrectly believe that you must use the
generic device nodes when you can in fact use either. It's often better
to use the driver specific node.
2005-01-08 06:00 brooks
[r139897] When ukbd0 arrives, attach to /dev/ukbd0 rather then /dev/kbd1 since
kbd1 might be something else.
2005-01-07 01:21 brooks
[r139821] Write some bit mask limits in hex rather than decimal so they look less
magic.
2004-12-16 04:45 brooks
[r138909] Add Cisco AIR-CB21AG and AIR-PI21AG to the supported list.

I'm using an AIR-CB12AG to make this commit.

Approved by: sam
2004-12-12 21:08 brooks
[r138758] s/do_no_use/do_not_use/

Reported by: ceri
Pointy hat to: brooks
2004-12-12 20:12 brooks
[r138756] MFC the addition of ifi_datalen to struct if_data and ifconfig support
for it. The change to struct if_data differs slightly from the one in
5.3 in that a pad is added after ifi_xmitquota rather then adding
ifi_link_state before ifi_recvquota. This is necessicary to maintain
the 4.x ABI.
2004-12-03 23:43 brooks
[r138364] kmod.mk knows how to create empty opt_*.h files so let it deal with the
NOINET6 case.

Reported by: ru
2004-11-26 19:01 brooks
[r138119] Allow the gif module to be built without IPv6 support when NOINET6 is
defined.

Submitted by: mitrohin a.s. <swp at uni-altai dot ru>
2004-11-25 01:09 brooks
[r138072] Remove a duplicate line from an apparent merge error in rev 1.63.
2004-10-30 13:45 brooks
[r137071] Add a loader menu option to set hint.atkbd.0.flags=0x1 which allows USB
keyboards to work if no PS/2 keyboard is attached. The position in the
menu was chosen to avoid moving option 6 (loader prompt). This should
be a no-op on non-i386/amd64 machines.
2004-10-14 14:38 brooks
[r136512] MFC rev 1.33:

A power failure left the temporary /var/.diskless directory
on my system, and since then my /var was always created as
MFS which was very surprising. Fix this for /tmp and /var.

Approved by: re (kensmith)
2004-10-08 00:24 brooks
[r136243] Since net/net_osdep.c contained only one function that could be
trivially implemented as a macro, do that and remove it. NetBSD did
this quite a while ago.
2004-10-06 04:47 brooks
[r136186] Mention "-d directory" in usage().

Pointy hat to: brooks
2004-10-05 22:16 brooks
[r136174] Don't prepend the directory specified by -d when the file is a relative
path. Doing so makes no sense. I'm not sure allowing relative paths
makes sense either, but I'm not going to break that now.
2004-10-05 22:09 brooks
[r136172] Bump __FreeBSD_version for addition of newsyslog -d.
2004-10-04 23:35 brooks
[r136127] Add a new -d argument which is used to specify an alternate root for log
files similar to DESTDIR in the BSD make process. This only affects log
file paths not config file (-f) or archive directory (-a) paths.
2004-10-04 22:57 brooks
[r136126] MFC: 1.207

Log the renaming of an interface. This should make it easier to follow
kernel log files.

Approved by: re (scottl)
2004-09-30 04:33 brooks
[r135949] MFC: refer to bsdlabel, not disklabel.

Approved by: re (hrs)
2004-09-29 05:28 brooks
[r135922] Update the list of files involved in diskless booting.

Reminded by: pjd
2004-09-24 22:36 brooks
[r135773] MFC: 1.208, 1.209

Fix a LOR where ifconf() used copyout while holding a mutex. This LOR
was seen when configuring addresses on interfaces using ifconfig. This
patch has been verified to work with over eight thousand addresses
assigned to an interface.

Approved by: re (scottl)
LOR id: 031
2004-09-23 21:17 brooks
[r135639] Backout the hack to install mtree, newsyslog, and touch in the cloned
root in the previous commit. We'll fix /etc/rc.d/var directly (someday).
2004-09-23 15:52 brooks
[r135620] Hack around the fact that the current diskless scripts require some
binaries in /usr before /usr is mounted by copying those files to the
root.

Stop making devices. devfs is manditory.
2004-09-23 15:26 brooks
[r135618] Sort the SEE ALSO list correctly. Update the document date.

Reported by: ru
2004-09-23 04:57 brooks
[r135593] Refrence bsdlabel.8 and bsdlabel.5 rather then the obsolete disklabel.8
and disklabel.5. Refrencing bsdlabel.5 is somewhat bogus as it is not
connected to the build and is mostly unwritten at this point.

Reported by: Matthias Andree <matthias.andree at web.de>
PR: docs/72020
2004-09-22 08:59 brooks
[r135568] Fix a LOR where ifconf() used copyout while holding a mutex. This LOR
was seen when configuring addresses on interfaces using ifconfig. This
patch has been verified to work with over eight thousand addresses
assigned to an interface.

LOR id: 031
2004-09-22 04:43 brooks
[r135564] MFC: Makefile:1.240 microtime.9:1.8 microuptime.9:1.6 time.9:1.13

Bring kernel time variable and access function docs up to date by
updating the list of variables and mentioning the bin(up)time functions.

Approved by: re (kensmith)
2004-09-22 04:40 brooks
[r135563] MFC: 1.44

Document the struct if_data members ifi_datalen, ifi_epoch, and
ifi_link_state.

Approved by: re (kensmith)
2004-09-20 07:12 brooks
[r135507] MFC 1.3: Fix a LOR where copyout was called while holding a lock when
listing available cloners (running "ifconfig -C").

Approved by: re (scottl)
Reported by: rwatson
LOR id: 039
2004-09-20 07:06 brooks
[r135506] MFC revs 1.860 and 1.861:

Add 'hint.atkbd.0.flags="0x1"' back to /boot/device.hints on the
install media on i386 and amd64. While the current default is very
useful in server environments, the ability to use USB keyboards is
vastly more important then the ability to use PS/2 keyboards the admin
forgot to plug in during the install process.

Approved by: re (scottl)
PR: kern/71443 kern/71568 (probably others)
2004-09-18 05:02 brooks
[r135416] Log the renaming of an interface. This should make it easier to follow
kernel log files.
2004-09-17 07:24 brooks
[r135366] Bring kernel time variable and access function docs up to date by
updating the list of variables and mentioning the bin(up)time functions.
2004-09-17 06:49 brooks
[r135365] Document the struct if_data members ifi_datalen, ifi_epoch, and
ifi_link_state.

Reminded by: ru
2004-09-15 05:14 brooks
[r135257] Use "cp -f" instead of "rm" and "cp" to break the hardlink to
device.hints.

Submitted by: ru
2004-09-15 04:41 brooks
[r135256] Fix a LOR where copyout was called while holding a lock.

Reported by: rwatson
2004-09-14 04:01 brooks
[r135204] Add 'hint.atkbd.0.flags="0x1"' back to /boot/device.hints on the
install media on i386 and amd64. While the current default is very
useful in server environments, the ability to use USB keyboards is
vastly more important then the ability to use PS/2 keyboards the admin
forgot to plug in during the install process.

PR: kern/71443
2004-09-14 03:44 brooks
[r135200] MFC: The disk labels generated by bsdlabel can no address more than
0xffffffff sectors. Document this limit and avoid installing bogus
labels on disks with more sectors (by refusing to install one at all).

Approved by: re (scottl)
2004-09-13 05:11 brooks
[r135133] MFC the addition of ifi_epoch to struct if_data. The ifi_unused
variable has been replaced with ifi_epoch (and additionally, ifi_timepad
on alpha) so no ABI change has occured.

Approved by: re (kensmith)
2004-09-09 07:46 brooks
[r134973] The disk labels generated by bsdlabel can no address more than
0xffffffff sectors. Document this limit and avoid installing bogus
labels on disks with more sectors.

Allowing the installation of labels addressing as much of the disk as
possiable may be a useful addition in some situations, but this was easy
to implement and should reduce confusion.

PR: bin/71408
2004-09-08 04:50 brooks
[r134933] Re-add ifi_epoch, to struct if_data, this time replacing ifi_unused
to avoid ABI changes. It is set to the last time the interface
counters were zeroed, currently the time if_attach() was called. It is
intentended to be a valid value for RFC2233's ifCounterDiscontinuityTime
and to make it easier for applications to verify that the interface they
find at a given index is the one that was there last time they looked.

Due to space constraints ifi_epoch is a time_t rather then a struct
timeval. SNMP would prefer higher precision, but this unlikely to be
useful in practice.
2004-09-08 04:08 brooks
[r134929] MFC rev 1.200
When detaching an interface, don't leave an obsolete pointer to the
soon to be deleted struct ifnet around.

PR: kern/52260
Submitted by: Jinmei Tatuya <jinmei at isl.rdc.toshiba.co.jp>
2004-09-07 23:26 brooks
[r134921] MFC 1.15:
When an USB keyboard is plugged in to a machine with a builtin
keyboard, cause the USB keyboard to take over from the builtin one.

Approved by: re (scottl)
PR: kern/71354 (partial workaround only)
2004-09-07 23:21 brooks
[r134920] MFC: if.c:1.202 if.h:1.90 ifconfig.c:1.107

Use a spare byte in struct if_data to store the structure size without
increasing it. Add code to ifconfig to use this size to find the
sockaddr_dl after the struct if_data in the routing message. This
allows struct if_data to grow (up to 255 bytes) without breaking
ifconfig.

Approved by: re (scottl)
2004-09-03 07:11 brooks
[r134686] If the argument to the -r flag starts with a ':' or a '|', don't try to
make sure it is a device. GDB special cases these prefixes and treats
:#### as a tcp port on localhost and executes what ever follows '|'.

This allows kgdb to debug via dconschat.

Discussed with: marcel
2004-09-02 05:07 brooks
[r134630] Back out ifi_epoch. The ABI breakage is too disruptive this close to
5-STABLE. ifi_epoch will shortly be reintroduced with less precistion
using the space currently allocated to ifi_unused.
2004-09-01 18:22 brooks
[r134609] Use a spare byte in struct if_data to store the structure size without
increasing it. Add code to ifconfig to use this size to find the
sockaddr_dl after the struct if_data in the routing message. This
allows struct if_data to grow (up to 255 bytes) without breaking
ifconfig.

Submitted by: peter
2004-09-01 15:14 brooks
[r134604] Add warnings about ifconfig incompatabilities caused by the addition of
ifi_epoch.
2004-09-01 01:19 brooks
[r134585] General modernization of coda:
- Ditch NVCODA
- Don't use a static major
- Don't declare functions extern

Reviewed by: peter
2004-09-01 00:08 brooks
[r134584] When an USB keyboard is plugged in to a machine with a builtin keyboard,
cause the USB keyboard to take over from the builtin one. This means my
laptop just works when I'm using it as a desktop.

Reviewed by: imp
2004-08-30 22:15 brooks
[r134540] MFC 1.200:
When detaching an interface, don't leave an obsolete pointer to the
soon to be deleted struct ifnet around.

PR: kern/52260
Approved by: re (scottl)
2004-08-30 06:29 brooks
[r134514] Add a new variable, ifi_epoch, to struct if_data. It is set to the last
time the interface counters were zeroed, currently the time if_attach()
was called. It is indentended to be a valid value for RFC2233's
ifCounterDiscontinuityTime and to make it easier for applications to
verify that the interface they find at a given index is the one that was
there last time they looked.

An if_epoch "compatability" macro has not been created as ifi_epoch has
never been a member of struct ifnet.

Approved by: andre, bms, wollman
2004-08-27 19:42 brooks
[r134399] When detaching an interface, don't leave an obsolete pointer to the
soon to be deleted struct ifnet around.

PR: kern/52260
MFC After: 3 days
2004-07-22 22:50 brooks
[r132557] Actually free the unit when destroying the interface.

Reported by: la at delfi.lt
Tested by: la at delfi.lt
PR: 68618
2004-07-02 19:44 brooks
[r131477] Don't announce the ethernet address when it's 00:00:00:00:00:00. It's
not of any interest. This primairly happens when vlan(4) interfaces are
created.
2004-07-01 17:16 brooks
[r131416] Add support to the uftdi driver for Intrepid Control Systems' vehicle
bus interfaces. These interfaces use the FTDI chipset with different
Vendor and Product IDs.

Add two additional baud rate enumerations. The vehicle bus interfaces
use a baud rate of 2000000. Also add 3000000 as it is the other FTDI
baud divisor special case.

I've commited a slightly different patch from that provided in the PR as
I changed the matching code a bit yesterday.

Submitted by: Mike Durian <durian at shadetreesoftware.com>
PR: kern/67357
2004-07-01 05:14 brooks
[r131391] Product and vendor ids for SIIG US2308 USB to Serial adaptor.

Tested against: rwatson's laptop
2004-06-30 21:00 brooks
[r131369] Static device counts will not be supported in 6.x so place the generated
macros (N<driver>) under BURN_BRIDGES.

Discussed with: peter, imp, scottl, ...
2004-06-23 17:33 brooks
[r130982] el(4) stopped needing to me a count device in December 2000.
2004-06-22 21:07 brooks
[r130940] Fix whitespace errors in previous commit.

Reported by: simon
2004-06-22 20:28 brooks
[r130937] Add entry for network interface cloning changes.
2004-06-22 20:15 brooks
[r130934] Bump FreeBSD_version for if_clone overhaul.
2004-06-22 20:13 brooks
[r130933] Major overhaul of pseudo-interface cloning. Highlights include:

- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.

Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-01 19:30 brooks
[r129957] Add Aerospace Corporation copyrights to EUI64 support files.

Suggested by: marcel, imp
2004-05-27 17:23 brooks
[r129782] Fix stupid patch(1) tricks. Apparently patch thinks all files match the
empty file so if you accidently apply a patch created with diff -N
twice, you get files with duplicate contents.

Reported by: Antoine Brodin <antoine.brodin at laposte.net>
2004-05-26 22:59 brooks
[r129760] Use new eui64(3) functions to print EUI-64s and to allow access to nodes
by EUI-64 and name.

Reviewed by: simokawa
2004-05-26 22:58 brooks
[r129759] Add support for an /etc/eui64 file modeled on /etc/ethers. The API is
modeled on ethers(3) except that all functions are thread-safe.

Reviewed by: simokawa
2004-04-20 23:34 brooks
[r128501] Remove the requirement that the "discard" service be defined in the
services database. Now only services that are actually used need to be
defined.

Submitted by: ume
2004-04-19 05:06 brooks
[r128417] Use an tempory struct ifnet *ifp instead of sc->sc_if to access the
ifnet in stf_clone_create. Also use if_printf() instead of printf().
2004-04-17 04:12 brooks
[r128340] Actually install preseedrandom.

Reported by: bde
2004-04-16 18:36 brooks
[r128324] I added preseedrandom yesterday, not last month.

Reported by: Dmitry Morozovsky <marck at rinet.ru>
2004-04-16 17:13 brooks
[r128322] Document changes in /dev/random initalization. Suggest running
/etc/rc.d/preseedrandom to seed the PRNG in the upgrade instructions.
2004-04-15 18:23 brooks
[r128287] Remove bogus checks on the value of ${entropy_file} and hardcode out
entropy source to /entropy. We have to assume there is no rc.conf at
this stage of the boot process.

Reported by: njl
2004-04-15 17:52 brooks
[r128286] Band-aid diskless booting by running a new preseedrandom script before
initdiskless. The output of several commands and if available the
contents of /entropy are feed into /dev/random to kickstart the PRNG.
/etc/rc.d/initrandom is left alone to maintain the previous behavior as
much as possiable.

Further work in this area is probably needed.

Discussed with: markm
2004-04-14 00:57 brooks
[r128209] Staticize <if>_clone_{create,destroy} functions.

Reviewed by: mlaier
2004-04-12 18:11 brooks
[r128163] Catch up with diskless split "diskless" is no longer provided.

Suggested by: cperciva
2004-03-23 23:22 brooks
[r127345] Overhaul the /etc/rc.d/diskless script by splitting it out into
hostname, resolve, tmp, and var scripts. The latter three are new and
were repo copied. These scripts no longer depend on being booted with
and NFS root instead attempt to automaticly create mfs /tmp and /var
volumes if the they are not writable. This behavior can be overridden
in /etc/rc.conf.

Reviewed by: luigi, pjd
2004-03-13 02:35 brooks
[r126901] Don't allow interfaces to be renamed to the empty string.
While I'm here, errors aren't bools.

Pointed out by: hmp
2004-03-13 02:31 brooks
[r126900] Remove if_withname. It came in with the KAME import, but never got
used. Should someone need its functionality, it's a really expensive
implementation of:
ifnet_byindex(sdl->sdl_index)

Reviewed by: bde, ume
2004-03-12 20:37 brooks
[r126888] Allow kernel with the BOOTP option to boot when DHCP/BOOTP sets the root
path to an absolute path without a host name. Previously, there was a
nasty POLA violation where a system would PXE boot until you added the
BOOTP option and then it would panic instead.

Reviewed by: tegge, Dirk-Willem van Gulik <dirkx at webweaving.org>
(a previous version)
Submitted by: tegge (getip function)
2004-03-12 04:40 brooks
[r126868] If /conf/diskless_remount exists, use it to remount the entire /conf
directory. This allows multiple roots (say for different architectures)
to share the same set of /conf files.
2004-02-04 02:55 brooks
[r125412] Add ifconfig support for network interface renaming. In the process,
reorganize the printing of the interface name when using wildcard
cloning so it is not printed if it we either immediately rename or
destroy the interface.

Reviewed by: ru
2004-02-04 02:54 brooks
[r125411] Add the kernel side of network interface renaming support.

The basic process is to send a routing socket announcement that the
interface has departed, change if_xname, update the sockaddr_dl
associated with the interface, and announce the arrival of the interface
on the routing socket.

As part of this change, ifunit() is greatly simplified by testing
if_xname directly. if_clone_destroy() now uses if_dname to look up the
cloner for the interface and if_dunit to identify the unit number.

Reviewed by: ru, sam (concept)
Vincent Jardin <vjardin AT free.fr>
Max Laier <max AT love2party.net>
2004-02-02 21:55 brooks
[r125345] More macro cleanup. Use the system roundup2() macro instead of making
our own ROUNDUP() macro.

Suggested by: bde
2004-01-31 22:59 brooks
[r125280] Use memcpy plus a manual NUL termination when copying the interface name
from the sdl because strlcpy requires that the source string be
NUL-terminated unlike strncpy.

Submitted by: Peter Jeremy <PeterJeremy at optushome dot com dot au>
2004-01-27 19:35 brooks
[r125109] Cleanup malloc() use in if_attach():
- malloc() returns a void* and does not need a cast
- when called with M_WAITOK, malloc() can not return NULL so don't
check for that case. The result of the check was bogus anyway since
it would leave the interface broken.
2004-01-27 03:15 brooks
[r125062] Clean up macro usage in if_attach():
- Use the system offsetof macro rather then making out own.
- undef ROUND after we use it rather then polluting the whole file.
2004-01-27 01:43 brooks
[r125060] Use IFNAMSIZ instead of a magic value for the length of an interface
name.

Prevent the kernel from potentially overflowing the interface name
variable. The size argument of strlcpy is complex because the name is
not null-terminated in sdl_data.
2004-01-21 22:00 brooks
[r124815] Use the length of the interface name, not the length of its address when
printing the name.

Approved by: brian
2004-01-21 18:05 brooks
[r124805] Don't leak softc's when destroying interfaces.

Init the softc list when loaded.

Noticed by: Maurycy Pawlowski-Wieronski <maurycy at fouk dot org>
2004-01-05 19:14 brooks
[r124149] Don't run newaliases if sendmail_enable=NONE.

Submitted by: Matt Peterson <matt at peterson dot org>
Approved by: silence on freebsd-rc
2003-12-02 21:52 brooks
[r123114] Reconnect ipfstat, ipnat, and ipftest to the build now that if_xname
support is enabled.

Approved by: re (scottl)
2003-12-02 18:33 brooks
[r123110] This commit was generated by cvs2svn to compensate for changes in r123109,
which included commits to RCS files with non-trunk default branches.
2003-12-02 18:33 brooks
[r123109] This is a direct commit to the vendor branch with code from the vendor.

Replace the if_name and if_unit members of struct ifnet with new members
if_xname, if_dname, and if_dunit. if_xname is the name of the interface
and if_dname/unit are the driver name and instance.

This change paves the way for interface renaming and enhanced pseudo
device creation and configuration symantics.

Submitted by: darrenr
Approved by: re (scottl)
2003-12-02 18:28 brooks
[r123108] Use parens more reasionably so we match the vendor code. This is part
of the if_xname change.

Submitted by: darrenr
Approved by: re (scottl)
2003-11-01 23:24 brooks
[r121867] Do if_xname conversion. I missed this one because it's pc98 only.
2003-11-01 00:03 brooks
[r121827] We want the length of the string, not the size of its pointer.
2003-10-31 21:58 brooks
[r121825] Use tabs not spaces.

Pointed out by: Xin LI <delphij at frontfree dot net>
2003-10-31 19:48 brooks
[r121818] Temporarily disable ipfstat and ipnat in /rescue to fix world.
2003-10-31 18:54 brooks
[r121817] Temporarily disconnect ipfstat, ipnat, and ipftest to unbreak world.

Pointy hat to: brooks
2003-10-31 18:32 brooks
[r121816] Replace the if_name and if_unit members of struct ifnet with new members
if_xname, if_dname, and if_dunit. if_xname is the name of the interface
and if_dname/unit are the driver name and instance.

This change paves the way for interface renaming and enhanced pseudo
device creation and configuration symantics.

Approved By: re (in principle)
Reviewed By: njl, imp
Tested On: i386, amd64, sparc64
Obtained From: NetBSD (if_xname)
2003-10-31 02:48 brooks
[r121778] Make TUNDEBUG use if_printf instead of printf.
2003-10-31 01:35 brooks
[r121777] Replace a couple printfs with if_printfs.
2003-10-31 00:44 brooks
[r121776] Initialize if_softc and use it in favor of &ipr_softc[ifp->if_unit] to
get the softc.
2003-10-30 23:30 brooks
[r121775] Remove unused lp_unit member from softc.
2003-10-30 23:12 brooks
[r121772] Remove unused FXP_UNIT() macro.
2003-10-30 23:05 brooks
[r121771] ifp->if_name can't be null when we just assigned a static string to it.
2003-10-30 23:00 brooks
[r121769] style: strcmp() does not return a bool.
2003-10-30 19:45 brooks
[r121752] There is no way to enter the attach routine twice with the same softc
without a detach call in between so don't try to deal with that
possiability.

This is a diff-reduction commit for the upcoming if_xname conversion.
2003-10-28 20:58 brooks
[r121637] Use VLANNAME instead of "vlan".
2003-10-23 16:29 brooks
[r121431] Use IF_MAXUNIT instead of rolling our own identical TUN_MAXUNIT.
2003-08-07 18:25 brooks
[r118630] MFC: Allow diskless_remount files to contain path beginning with /
which will be mounted relative to the NFS root mountpoint.
2003-07-07 21:41 brooks
[r117318] Add support for a -n argument which displays user and group IDs
numerically rather than converting to a user or group name.

MFC After: 1 week
2003-06-30 21:47 brooks
[r117087] Allow diskless_remount files to contain path beginning with / which will
be mounted relative to the NFS root mountpoint.

Reviewed by: dillon at backplane.com
MFC After: 3 days
2002-11-05 05:58 brooks
[r106444] Test for the existance of kernel ppp support by seeing if the kernel has
the module compiled in or loaded instead of bogusly checking for ppp0.
Also if and only if the caller is actually root and the kernel does not
have ppp support, try to load the ppp module before giving up.
2002-10-21 14:38 brooks
[r105632] Use if_printf() instead of printf() to print diagnositic messages in the
form "rdp#: blah".
2002-10-21 14:37 brooks
[r105631] Use if_printf(ifp, "blah") instead of
printf("%s%d: blah", ifp->if_name, ifp->if_unit).
2002-10-21 05:09 brooks
[r105609] All bpf.h/NBPF consumers are gone so stop generating bpf.h
2002-10-21 05:05 brooks
[r105608] Don't include the depricated "bpf.h" and always compile in bpf support
as per current practice.
2002-10-21 03:41 brooks
[r105603] Use if_printf(ifp, "blah") instead of printf("ppp%d: blah", ifp->if_unit).
2002-10-21 03:40 brooks
[r105602] Use if_printf(ifp, "blah") instead of printf("vlan%d: blah", ifp->if_unit).
2002-10-21 03:35 brooks
[r105601] Use if_printf(ifp, "blah") instead of printf("sl%d: blah", sc->sc_if.if_unit).
2002-10-21 02:54 brooks
[r105599] Use if_printf(ifp, "blah") and device_printf(dev, "blah") instead of
printf("%s%d: blah", ifp->if_name, ifp->if_xname). This eliminates the
need to store the unit number in the softc.
2002-10-21 02:51 brooks
[r105598] Use if_printf(ifp, "blah") instead of
printf("%s%d: blah", ifp->if_name, ifp->if_xname).
2002-10-01 01:03 brooks
[r104260] Use if_printf(ifp, "foo") instead of printf("xe%d: foo", ifp->it_unit).
2002-10-01 00:59 brooks
[r104259] Use if_printf(ifp, "blah") instead of printf("vx%d: blah", ifp->if_unit).
2002-10-01 00:58 brooks
[r104258] Use if_printf(ifp, "blah") instead of printf("sr%d: blah", ifp->if_unit).
2002-10-01 00:56 brooks
[r104257] Use if_printf(ifp, "blah") instead of printf("sn%d: blah", ifp->if_unit).
2002-10-01 00:55 brooks
[r104256] Use if_printf() instead of printf() to print "sbni#: blah".
2002-10-01 00:52 brooks
[r104254] Use if_printf(ifp, "blah") instead of printf("fe%d: blah", ifp->if_unit).

A number of functions in this driver still use the unit number in their
printouts because they pass the unit directly as a function argument
instead of passing a softc or struct ifnet pointer. This should be
resolved at a future date.
2002-10-01 00:49 brooks
[r104253] Use if_printf() instead of printf() to print "ed#: blah".
2002-10-01 00:46 brooks
[r104252] Use if_printf(ifp, "blah") instead of
printf(CS_NAME"%d: blah", ifp->if_unit);
2002-10-01 00:44 brooks
[r104251] Use if_printf(ifp, "blah") instead of
printf("%s%d: blah", ifp->if_name, ifp->if_unit);
2002-10-01 00:42 brooks
[r104249] Use if_printf(ifp, "blah") instead of printf("ar%d: blah", ifp->if_unit).
2002-09-24 17:37 brooks
[r103901] Convert most printf()s to if_printf()s.
2002-09-24 17:35 brooks
[r103900] Add a new helper function if_printf() modeled on device_printf(). The
function takes a struct ifnet pointer followed by the usual printf
arguments and prints "<interfacename>: " before the results of printf.
Since this is the primary form of printf calls in network device drivers
and accounts for most uses of the ifnet menber if_unit, this
significantly simplifies many printf()s.
2002-09-06 20:07 brooks
[r103040] Don't include "bpf.h" when compiling on FreeBSD. We always compile in
bpf support these days.
2002-09-02 20:16 brooks
[r102863] Hook up libkern/strlcpy.c and libkern/strlcat.c after repocopy.

Obtained from: OpenBSD
Discussed on: -arch
2002-09-02 20:10 brooks
[r102862] Make SCSI_DELAY setable at boot time and runtime via the
kern.cam.scsi_delay tunable/sysctl.

Reviewed by: mdodd, njl
2002-09-02 00:52 brooks
[r102828] Continue de-counting i4b. Devices i4bctl, i4bcapi, iavc, i4bq921,
i4bq931, i4b, isic, iwic, ifpi, ifpi2, ifpnp, ihfc, and itjc are
no longer count devices. Also remove a few other instances of N<DEVICE>
being used to control compilation of whole files.

Reviewed by: hm
2002-08-28 04:46 brooks
[r102504] MFC: 1.9: spelling
1.10: document cloning
2002-08-28 04:44 brooks
[r102503] MFC: document cloning
2002-08-19 19:22 brooks
[r102130] Fix a couple of bogus return values in previous commit.

Submitted by: "Vladimir B. " Grebenschikov <vova@sw.ru>
Pointy hat to: brooks
2002-08-09 15:30 brooks
[r101588] Make ppp(4) devices clonable and unloadable.
2002-06-13 06:04 brooks
[r98155] Remote pci.h/NPCI usage from i4b code.

Approved by: hm
2002-05-31 06:28 brooks
[r97653] The loop back device hasn't been a count device for a while so remove
the number of interfaces.
2002-05-29 16:16 brooks
[r97473] Restore the irq=0 => irq=255 hack to pci_cfgintr_search(). Just having
it in pci_cfgregread() wasn't sufficent on at least the HP Omnibook 500.

Reviewed by: imp
2002-05-29 06:59 brooks
[r97452] Add a simple manpage for the discard interface derived from lo(4).
2002-05-29 06:55 brooks
[r97451] Document the fact that these devices are allocated at runtime using
cloning and point to two common methods of doing so.
2002-05-25 20:20 brooks
[r97290] Make discard devices clonable and unloadable. Also, change the
interface name from ds# to disc#.
2002-05-25 20:17 brooks
[r97289] Move all unit number management cloned interfaces into the cloning
code. The reverts the API change which made the <if>_clone_destory()
functions return an int instead of void bringing us into closer
alignment with NetBSD.

Reviewed by: net (a long time ago)
2002-04-14 23:18 brooks
[r94695] Fix tx-rate setting for Lucent cards.

Submitted by: Eugene Perevyazko <john@pcs.dp.ua>
2002-04-14 22:08 brooks
[r94687] Support the Siemens SpeedStream PCI card.

PR: kern/35988
Submitted by: Stephen Gunn <csg@waterspout.com>
2002-03-25 10:44 brooks
[r93131] MFC: revs 1.17-1.18: document new sysctls net.link.gif.max_nesting and
net.link.gif.parallel_tunnels.
2002-03-25 10:40 brooks
[r93130] MFC: rev 1.18: Add sysctls net.link.gif.max_nesting and
net.link.gif.parallel_tunnels.
2002-03-25 10:19 brooks
[r93129] MFC: rev 1.37

When using hardware decoding, reconstruct the wire form of the ethernet
header and push it up any attached bpf devices on the parent interface.
This makes hardware vlan decoding more like the normal software path.
2002-03-25 02:55 brooks
[r93115] Remote newly unused variable to unbreak LINT.

Pointy hat to: amorita
2002-03-15 04:20 brooks
[r92300] Add some missing spls which stop stop the once a minute wi_cmd timeout
messages and accompanying .5sec hangs on Lucent cards.

Not an MFC because there are no spls in current.

Submitted by: Tod McQuillin <devin@spamcop.net>
2002-03-11 23:29 brooks
[r92117] Add support for 3Com Airconnect PCI cards.

MFC after: 3 days
2002-03-04 21:46 brooks
[r91648] Add cloning support to the loopback interface.

Submitted by: mux
2002-03-04 21:43 brooks
[r91647] Change the network interface cloning API so the destroy function returns
an int errorcode instead of void in preperation for merging cloning of
the loopback device.

Submitted by: mux
MFC after: 2 weeks
2002-03-04 20:03 brooks
[r91642] MFC: sync with current modulo API changes.

The significant changes in this MFC are:

- Prism II detection from NetBSD.
- Prism 2.5 support.
- 128-bit crypto fixes for Prism chipsets.
2002-03-02 01:41 brooks
[r91554] MFC: rev 1.70
put function return types on a line by themselves as per style(9)
2002-03-02 00:06 brooks
[r91548] Make 128-bit WEP (aka 104-bit WEP) work with Prism based cards.

Submitted by: Thomas Skibo <skibo@pacbell.net>
MFC after: 3 days
2002-02-28 01:03 brooks
[r91454] Some style(9) whitespace fixes and a correction to the copyright.
2002-02-26 20:11 brooks
[r91327] Fix warnings in the gif(4) driver so it compiles with -Werror.
2002-02-26 20:09 brooks
[r91326] Set ifp->if_name to "awi" instead of device_get_name(dev) to avoid
casting away a const qualifier so the awi driver compiles.
2002-02-26 18:24 brooks
[r91324] Staticize an extern that no one else used.
2002-02-26 17:35 brooks
[r91318] Add a missing field to the end of array marker in the pci_ids array.

Submitted by: sumikawa
2002-02-26 02:19 brooks
[r91272] When using hardware decoding, reconstruct the wire form of the ethernet
header and push it up any attached bpf devices on the parent interface.
This makes hardware vlan decoding more like the normal software path.

Tested by: cjtt@employees.org
MFC after: 2 weeks
2002-02-26 01:56 brooks
[r91270] Make gif(4) nesting level and parallel tunnel support tunable at runtime
via sysctl's. The old #defines, MAX_GIF_NEST and XBONEHACK are
currently supported for backwards compatability, but will probably be
removed at some point in the future.
2002-02-20 18:23 brooks
[r90990] Add missing m_free() so we actually drain the send buffer in monitor mode.

Submitted by: Kenjiro Cho <kjc@csl.sony.co.jp>
2002-02-12 17:52 brooks
[r90580] Add support for the Linksys WMP-11, Prism 2.5, PCI adaptor.

Submitted by: Thomas Skibo <skibo@pacbell.net>
MFC after: 2 weeks
2002-01-21 01:26 brooks
[r89602] MFC: Add support for Linksys WDT11 PCI adaptor.

Submitted by: Eric Liedtke <eliedtke@apogeetelecom.com>
Approved by: re
2002-01-21 00:59 brooks
[r89600] Add support for Linksys WDT11 PCI adaptors.

Submitted by: Eric Liedtke <eliedtke@apogeetelecom.com>
2001-12-20 05:38 brooks
[r88247] Remove an extra .Xr which caused a syntax error.

Pointed out by: fenner
2001-12-19 22:25 brooks
[r88230] MFC: Better document the creation and configuration of gif(4) devices.
2001-12-19 21:51 brooks
[r88228] MFC: Initalize ifq_maxlen to silience a harmless warning.
2001-12-17 07:49 brooks
[r88034] Initalize ifq_maxlen to prevent a harmless warning message.

MFC After: 1 day
Pointed out by: jacks@sage-american.com, bmah
2001-12-15 01:07 brooks
[r87968] MFC: make glue for faith(4) module.
2001-12-15 01:06 brooks
[r87967] MFC: Make faith(4) interfaces (un)loadable and clonable.
2001-12-14 22:55 brooks
[r87958] Better document gif interface cloning intrastructure.

MFC After: 2 days
2001-12-09 06:02 brooks
[r87547] MFC: New rc.conf variable, cloned_interfaces, for creating cloned
network devices at startup.
2001-12-04 20:09 brooks
[r87368] MFC: Add make file support for the vlan(4) module.
2001-12-04 20:01 brooks
[r87367] MFC: Make vlan(4) devices loadable, unloadable, and clonable.
2001-12-03 17:28 brooks
[r87276] Don't pass an interface pointer to VLAN_INPUT{,_TAG}. Get it from the
mbuf instead.

Suggested by: fenner
2001-09-28 00:22 brooks
[r84063] Mention that gif interfaces are created at runtime and refer the
administrator to the ifconfig create command.

MFC after: 1 week
2001-09-28 00:02 brooks
[r84060] Remove a couple unintentional mentions of Ethernet that crept in from
ng_ether.c.
2001-09-27 22:11 brooks
[r84056] Add the new ng_gif, ng_gif_demux, and ng_ip_input nodes to NOTES so they
get compiled with LINT.
2001-09-27 21:54 brooks
[r84053] Add ng_ip_input. A new netgraph node for queuing IP packets into the
main IP input processing code.
2001-09-27 21:38 brooks
[r84051] Add an apparently working entry for the BayStack 660, 2mbps direct
sequence wireless card.

Approved by: imp (in principle)
2001-09-27 03:14 brooks
[r84004] NULL commit due to misunderstanding "cvs commit -m <blah>". The
previous log message "/home/brooks/ng_gif.message" should have read:

Add a pair of new netgraph nodes. The ng_gif node is like ng_ether
except that it works on gif interfaces. The ng_gif_demux node attaches
to an ng_gif node to allow separate processing or discarding of
different types of encapsulated traffic.
2001-09-27 00:04 brooks
[r84000] The initial commit contained an error in the license, this is the
correct one.
2001-09-26 23:50 brooks
[r83998] /home/brooks/ng_gif.message
2001-09-26 23:37 brooks
[r83997] Use LIST_ macros instead of TAILQ_ macros to be more like NetBSD.

Obtained from: NetBSD
2001-09-26 22:41 brooks
[r83996] The number of ccd(4) devices is no longer set at compile time so stop
trying to do it in the examples and config files.
2001-09-26 22:02 brooks
[r83995] Call securelevel_gt with a credential. Also, s/p/td->td_proc/.

Submitted by: LINT
2001-09-26 21:53 brooks
[r83994] Include sys/proc.h for the definition of securelevel_ge().

Submitted by: LINT
2001-09-26 01:11 brooks
[r83951] Add a standard hack in the spirit of PCI_ENABLE_IO_MODES to allow systems
with weird PCI-PCI bridge configurations to work. Defining
PCI_ALLOW_UNSUPPORTED_IO_RANGE causes the sanity checks to pass even
with out of range values.

Reviewed by: msmith
2001-09-26 01:02 brooks
[r83950] MFC: monitor mode and style(9) cleanups.
2001-09-25 18:56 brooks
[r83936] The faith(4) device is no longer a count device so don't specify a count.
2001-09-25 18:40 brooks
[r83934] Make faith loadable, unloadable, and clonable.
2001-09-19 21:27 brooks
[r83677] Add a new rc.conf variable, cloned_interfaces, to create cloned
interfaces at boot.
2001-09-19 00:13 brooks
[r83655] Make stf a clonable device.

Yes this really is rather silly and the implementation is overkill given
that you are only allowed one of them, but NetBSD implements cloning on
this device and it's a less cluttered example of cloning then most.
2001-09-10 02:36 brooks
[r83270] Lots of style(9) related white space fixes. Mostly missing spaces
after if, for, while, switch, '{', and '}' plus a bunch of bogus
whitespace at the end of lines.
2001-09-10 02:05 brooks
[r83269] Add support for monitor mode. This means that after enabling the
correct mode via ancontrol, you can use bpf to sniff raw 802.11 frames.
Who want's to port AirSnort. ;-)

Submitted by: Doug Ambrisko <ambrisko@ambrisko.com> (author)
David Wolfskill <david@catwhisker.org> (port to current)
2001-09-06 00:27 brooks
[r83128] gif(4) is no longer a count device, reflect this change.
2001-09-05 23:59 brooks
[r83127] Make wicontrol WARNS=2 clean. This is mostly just adding const to things.
2001-09-05 23:47 brooks
[r83125] Attach if_vlan to the build process.
2001-09-05 23:47 brooks
[r83124] vlan.h is obsolete, don't create it anymore.
2001-09-05 23:33 brooks
[r83123] Deconditionalize vlan support.
2001-09-05 23:04 brooks
[r83122] Update tx(4) to always enable vlan(4) support.

Approved by: semenu
2001-09-05 21:19 brooks
[r83117] Don't specify the number of vlan interfaces any more, they are created
at runtime.
2001-09-05 21:10 brooks
[r83115] Make vlan(4) loadable, unloadable, and clonable. As a side effect,
interfaces must now always enable VLAN support.

Reviewed by: jlemon
MFC after: 3 weeks
2001-09-05 01:06 brooks
[r83043] Add cloning support for the tap(4) device similar to that in the tun(4)
device.

Submitted by: Maksim Yevmenkin <myevmenk@digisle.net>
2001-08-20 18:38 brooks
[r81992] MFC: rev 1.67, make plumb work as documented.

Approved by: re
2001-08-17 22:16 brooks
[r81862] Actuall make plumb work in addition to create as per the manpage.

PR: bin/29812
Submitted by: Joao Carlos Mendes Luis <jonny@eng05.embratel.net.br>
2001-08-16 23:52 brooks
[r81805] MFC: rev 1.10, Cisco Aironet 350 Series PCI id.

Approved by: re
2001-08-06 23:14 brooks
[r81221] Add PCI ID for Cisco Aironet 350 Series cards. The patch differs
slightly from the submitted one.

PR: kern/29415
Submitted by: Bob Fleck <rfleck@cigital.com>
2001-07-30 17:44 brooks
[r80661] MFC: Makefile cleanups, typo fixes, WARNS=2 fixes.
2001-07-30 17:01 brooks
[r80654] Make spacing consistant and set WARNS correctly.
2001-07-30 06:38 brooks
[r80598] MFC: an(4) header duplication removal.
2001-07-27 18:26 brooks
[r80454] Make ancontrol WARNS=2 clean.
2001-07-27 16:13 brooks
[r80451] MFC: r1.2: Don't print a blank line if no keys are set.
2001-07-27 16:05 brooks
[r80449] Cleanup duplication in Aironet headers

PR: kern/29210
Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
MFC after: 3 days
2001-07-27 04:40 brooks
[r80435] Remove -traditional from CFLAGS and add unistd.h to header.h so this
actually compiles. Hopefully, this code didn't rely on some weird
side effect of -traditional.
2001-07-25 16:42 brooks
[r80336] MFC: intrastructure mode has been the default for some time now.
2001-07-25 05:52 brooks
[r80315] Fix a stupid bug which resulted in a blank line in the status output if
WEP is supported, but not keys are set.

MFC after: 3 days
2001-07-25 00:19 brooks
[r80307] IFT_8021_VLAN -> IFT_L2VLAN per if_vlan_var.h rev 1.9. With this change
LINT compiles, but doesn't link.
2001-07-25 00:15 brooks
[r80306] Allow ng_split to be compiled in staticly.

MFC after: 7 weeks
2001-07-24 23:42 brooks
[r80305] Make ng_split part of the module building process.

MFC after: 7 weeks
2001-07-24 23:33 brooks
[r80304] General cleanup providing better style(9) conformance and generally
improved readability. The two real functional changes are that
netgraph now sees this as the "split" node type rather then the
"ng_split" node type and that meta data is passed through without
processing rather then being dropped.

Reviewed by: jhb, julian
MFC after: 7 weeks
2001-07-24 19:27 brooks
[r80302] MFC: interface cloning support.
2001-07-24 19:27 brooks
[r80301] merged interface cloning support.
2001-07-24 19:23 brooks
[r80299] Mention gif(4) changes related to cloning support.

Approved by: imp
2001-07-24 19:10 brooks
[r80298] MFC: Interface cloning support. gif modularity and cloning.
stf modularity and removal of gif dependence.

Reviewed by: brian
2001-07-24 05:34 brooks
[r80247] Bruce gave me a little too much credit ;-), stf didn't get the full
treatment along with gif so it doesn't support cloning yet.
2001-07-24 04:27 brooks
[r80246] Fix four </para> tags that were either missing or misspelled as </para>.

Found by: make lint
Reviewed by: bmah
2001-07-20 19:53 brooks
[r80052] gif isn't a count device anymore so don't put a number after it.

Pointed out by: brian
2001-07-20 02:01 brooks
[r80014] Upbreak the kernel by removing bogus '{'.
2001-07-15 20:32 brooks
[r79766] MFC: Put module directories one per line so adding new modules creates
readable diffs and produces less conflicts with local source trees.
2001-07-15 04:55 brooks
[r79745] Put module directories one per line so adding new modules creates readable
diffs and produces less conflicts with local source trees.
2001-07-14 21:17 brooks
[r79731] Mention new support for Cisco Aironet 350 series adaptors.
2001-07-04 20:49 brooks
[r79251] MFC: -m to show supported media types, better formatting of current and
supported media type output.

PR: 27701
Obtained from: NetBSD
2001-07-04 00:12 brooks
[r79171] MFC: ifconfig support for configuration of 802.11 wireless interfaces.
This commit contains support for the an and wi drivers.
2001-07-02 21:08 brooks
[r79107] Create gif devices in the "gifconfig" stage while configuring them.

Reviewed by: ru, ume
Obtained from: NetBSD
MFC after: 1 week
2001-07-02 21:02 brooks
[r79106] gif(4) and stf(4) modernization:

- Remove gif dependencies from stf.
- Make gif and stf into modules
- Make gif cloneable.

PR: kern/27983
Reviewed by: ru, ume
Obtained from: NetBSD
MFC after: 1 week
2001-07-02 20:52 brooks
[r79105] Support network device cloning via create and destroy options.

Reviewed by: ru, ume
Obtained from: NetBSD
MFC after: 1 week
2001-07-02 20:50 brooks
[r79104] Document network device cloning ioctls.

Reviewed by: ru, ume
MFC after: 1 week
2001-07-02 20:49 brooks
[r79103] Add kernel infrastructure for network device cloning.

Reviewed by: ru, ume
Obtained from: NetBSD
MFC after: 1 week
2001-07-02 17:23 brooks
[r79098] MFC: 1.17 (correctly document and implement -a)
2001-07-02 17:20 brooks
[r79097] MFC: 1.8 (default to an0 if no interface) and 1.9 (print tx speed).
2001-07-02 16:56 brooks
[r79095] Add Cisco Aironet 350 Series.

Approved by: bmah
2001-06-26 00:12 brooks
[r78789] Actually implement -a as documented since import (set access point density).
Remove documentation for short-lived usage as "print keys in ASCII."

PR: bin/21803
MFC after: 1 week
2001-06-26 00:00 brooks
[r78788] Output the current TX rate when showing status.

PR: bin/25470
2001-06-25 22:07 brooks
[r78783] Add support for the Cisco Aironet 350 Series of adaptors. Also, make
a minor ocrrection to the Aironet 340 Series comment.

Approved by: imp (in principle)
2001-06-22 23:35 brooks
[r78639] Fix a number of bugs in the implementation of the WEP related status
functions in ifconfig. "ifconfig an0" should output the correct
status now. Also, make the read and write functions both more
robust and more consistant. This should stop most of the incorrect
size complaints and eliminate the possiability of panics from firmware
that increases resource sizes.

PR: kern/27826
Reviewed by: imp, jlemon
Submitted by: Doug Ambrisko <ambrisko@ambrisko.com>
David Wolfskill <dhw@whistle.com>
2001-06-21 20:44 brooks
[r78568] Slightly lower average age.

Generated by svn2cl 0.13