merge in trunk rev. 3426

merge the rest

copy trunk/freewrt r3264 into a new branch for making the nfo system
because upgrading the old adk branch would be a hell more work

before anyone cries, give them back their colourls for “ls” command
(“l” and “ll” command stay with -F instead)

now that we HAVE mksh, we can use it…

first of a series of commits I'd suggest to not test until I did:

remove busybox ash and the other shells (hush, lash, msh); always
use mksh by default, as /bin/sh, /bin/ash symlink, /bin/mksh native,
and default admin login shell unless bash or zsh are selected

mksh is a superset of ash (I didn't find anything busybox ash can
do that mksh can't), and the upcoming mksh R30 will be even smaller
in -DMKSH_SMALL mode (but that one will lack vi editing mode)

ok wbx@

we planned this for 1.2, but it will greatly help the nfo conversion
and users request shells they can actually use if they don't have a
webinterface handy

breaks rb-2.6/nfsroot

removed wireless-bridge-if, not needed from now

dashver bump

missed dash version

fix selection of STA wlan mode

replace mkpasswd and /usr/bin/encrypt by portable md5crypt on
fwwif and mirmake basis; tested on GNU/Linux for now

(untested) if shutting down, first put off the power led, so we know
ok wbx@

oops

• move bridge configuration to end of /etc/network/interfaces to make it work
• fix echo -e abuse

• fix cat abuse and double-sed
• change lanbridge config to “inet manual”, requested by wbx@

spaetzle: if changing ANYTHING in fwifupdown, PLEASE make sure that the
auto-generated network configs from the base-files package are in sync.
In _all_ possible combinations.

• add zsh to /etc/shells, forgotten by wbx@
• remove lash, hush and msh from the choice for /bin/sh – we require
  either ash or mksh there for FreeWRT 1.1, other shells are not
  compatible enough
• enable the use of zsh as login shell (alongside with ash, hash, msh,
  lush, bash, mksh, /bin/sh)

… if everything were this easy…

wbx@ „warum finden die [unsere Kunden] das und wir nicht?“

deconfigure switch ports correctly on non-wl500gD
agreed wbx@

introduce a new shell function switch_reset() in all “preinit.arch”
files that is a no-op except on the Asus WL-500gD, where, accroding
to wbx@, a special procedure is required to reinitialise the switch
because it is so unlike the switch found on any other router

fixed failsafe cleanup, should fix #322 and #332

unbreak lan<->wlan bridge, WHY WAS THIS COMMENTED OUT?

• /tmp/.booting now exists from preinit to until after rc.local has been run
• pressing the RESET (not EzSetup) button >4 sec will reboot, like yer PC
• bump dashver

• aah, _so_ you're stripping them quotation marks
• fix lladdr wlan (pasto)
• can't be so hard to remove a blank line, eh?

experimental: support changing the lladdr via menuconfig too

• base-files: fix .ipk file name
• build.mk: if FWRT_PACKAGE_BASE_FILES=y also create build_mipsel/root
  (for chroot build_i386/root/)

merge base-files-arch into base-files, experimental, blame me if it breaks

also here reboot after 60s even if the shutdown scripts take longer

fix rc.conf entry generation
• quote properly
• handle case of $2 and $3 omission correctly
• make it possible to define empty default values

MFC to 1.0-stable branch recommended

run shitdown scripts synchronously.
I hope this doesn't break anything, but if the box is going down
faster than /etc/rc.shutdown umounts davfs2 (in my case), heaven
is going down the drain to hell, er I mean, cache inconsistency…

• base-files/files/etc/init.d/rcK, base-files/files/etc/init.d/rcS:
  do correct shebang processing, like the BSD kernel and mksh: get
  first line of script to run if it begins with "#!", strip that and
  any leading space/tab off, then get shell name (terminates with
  space/tab/end of string) and shell argument quoting according to
  unix common proceedings; use /bin/sh as interpreter if none could
  be found
• mk/package.mk: ensure all init scripts are 0755 by default
• busybox: flesh out inetd init script, so that all init scripts in
  FreeWRT either can be found with -path \*/init.d/\* or -name \*.init
• other packages: add #!/bin/sh shebang line to init scripts¹, bump dashver

¹) change this to /bin/ash if you use ash-specific features that mksh
   can also do; never use ash-specific features that mksh cannot do (are
   there any?), change to /bin/mksh if you use mksh-specific features
   and add a run-time dependency on mksh in that case

without /etc/hosts entries for localhost, many apps will behave
strange; this adds the missing IPv6 entries, which won't hurt –
or so I think – if the kernel doesn't support IPv6

fix for two minor error messages

* they occured after issuing 'reboot' on the serial console cli of rb-2.6
* please check base-files/files/etc/hotplug.d/usb/01-mount for correctness
  on a system with usb support

add dependency of base-files ipkg on the svn revision

„mach doch“ wbx@

fix the remainder of the init script fubar

run-time test succeeded after prodding and serial console help by wbx@

by request of wbx@:

make it possible for users to do
| /etc/init.d/<foo> restart
instead of having to do
| /etc/init.d/S*<foo> restart
by moving the order into a magickal™ comment.

Untested.

Also remove the svn:executable properties, because rcS and rcK will
always call them with /bin/sh, and bump dashvers accordingly.

replace get_next_uid() and get_next_gid() with a collision-free algorithm
taken from MirBSD adduser(8) but converted from mksh(1) with internal file
parsing to fork'n'exec'ing grep(1)

untested

may solve #274, #275

support checkouts from older (xml based) svn tools, too
(on a best-effort basis)

requested by wbx@

• optimise: don't read out the “svn status” a hundred times when one time
  is enough (hint: not everything needs to be a gmake variable) – this is
  by me, it annoyed me often enough
• don't use a local call to the svn client programme; this fails to build
  if it's not installed – noticed by Pawel Szubartowski, thanks

nbd (from OpenWrt) says that the /dev/console node on an Asus WL-500g
yields errors on I/O iff the serial console is not attached. This was
failing our failsafe mode; wbx@ pointed me to the OpenWrt code.

Fix this by trying the console (“dd” command taken from OpenWrt) then
redirecting I/O to a pty (idea was from OpenWrt, implementation fixed
by yours truly tg@) if it fails.

Also redirect telnetd I/O to /dev/null in case its stdio fhs are bad,
too. – XXX this might be wrong since this comes after preinit.arch

tested by thengineer (Jakob)

embed adk .config, some dot cleanup

• /etc/preinit.arch is sourced, doesn't need to be executable
  or have a shebang line
• also, whitespace cleanup and simplification
• new: /etc/init.d/rcK calls /etc/rc.shutdown¹ and then /etc/init.d/K* scripts
  just like /etc/init.d/rc.S runs /etc/init.d/S* scripts and then /etc/rc.local
  -> derived from PR#246, thanks for the input, agreed to commit wbx@
• /etc/S* and /etc/K* scripts do not need to be executable any more or
  have a shebang line, they're called from /bin/sh now
• new: /etc/init.d/K20hotplug, from PR#246
• new: /etc/init.d/K00wan, from /etc/profile (“reboot” function), ok wbx@
• remove /etc/version – even wbx@ was surprised it still exists
• /etc/S* and /etc/K* scripts, /etc/init.d/rcS and /etc/init.d/rcK
  now sanitise $PATH
• clean up /etc/profile
• /sbin/hotplug now calls its scripts from /etc/hotplug.d/ with /bin/sh
  instead of sourcing them, thus being more reliable
• simplify /sbin/hotplug

¹) just like /etc/rc.local, /etc/rc.shutdown SHALL NEVER be created by ADK.
   It's for the user to play with ONLY.

make the third parametre of add_rcconf() optional, default NO as per wbx@ policy

• allow the user to choose GNU bash as login shell for the
  “admin” user; requested by <sum> via IRC
• fix bug: bash wasn't in /etc/shells, bad wbx@

add support for /etc/rc.local: _if_ it's executed, it's run
by /bin/sh after all /etc/init.d/S* scripts have finished;
the /etc/rc.local file should _not_ be executable; it _must_
not be created by FreeWRT, it's intended as a means for the
user to add site-local post-startup configuration

* we override LC_ALL (already), no need to do it twice
* LANG is overridden by LC_ALL anyway
* on some systems, GNU dcgettext overrides LC_ALL with
  LANGUAGE, so we force-set that to 'C' as well

merged from branches/common-adk 1310:1685

the following packages are not converted yet:
* alsa
* asterisk
* pmacct
* zaptel
* elinks
* libgd
* lua
* mypackage
* openser
* php*
* snort*
* tmsnc

additionally, freeradius doesn't compile.

happy fixing :P

sync with stable

sync with freewrt_1_0 branch. It was a little bit to early to branch,
but now everybody can work again. New smaller stuff can be committed 
directly in trunk.
Bigger stuff, like a new architecture or a new board should be
developed via a separate branch. Kernel and compiler updates, too.

Any bugfixes or security updates should go into branch and trunk. 

Still send me a patch first and do not commit directly into the branch.

* add PKG_VERSION to every package in default selection
* add broadcom-scripts package as default for 1.0, it adds
  network and switch configuration scripts to base
  suggested by markus to separate these from webif package
* move ed to application and disable by default, sorry tg, but
  not everybody needs it
* be sure that if broadcom-scripts are installed, new ifup/ifdown
  network configuration is not used

switch to FWCF

ok to enable now wbx@
bugreports to tg@

move default to standard dir files

bump package patchlevel for markus@
after revision 641-643's changes

major package Makefile cleanup. remove mostly-clean, add licence header, always use RSTRIP varibale for stripping, rename some init scripts, indent

merge wbx-target-cleanup to trunk.
- mk directory for some top level makefiles (build.mk,vars.mk)
- add a separate configfs partition to all models, 128 kb big
- add model and platform specific startup files and kernel configs
- delete target/linux/package, add all addon kernel packages to packages
- simplify target/linux/brcm-2.4/Makefile, abstraction to mk/ comes later
- add target/image directory, place where the different images are created
- default off for all extra packages, be sure that you enable all packages
  you need to get a dsl-capable router. snapshots will be created by a specific config which contains all needed stuff (iptables, pppoe, pptp, haserl, webif)

generate /etc/shells dynamically, based upon packages
being built (even if marked 'm')

add the ability to choose user's default shell, too
XXX untested

-fixed warnings

* base-files: we'll have variables of the style
  | firewall=NO
  | ssh=YES
  and variables of the style
  | ntpd_flags=-s
  | ntpd_flags=""
  | ntpd_flags=NO
* openntpd: convert to use functions.sh stuff

as discussed with wbx@ some 2-3 weeks ago, change
the BR2_ prefix to FWRT_

tested with a standard build (only the squashfs
image since the jffs2 didn't get built)

as an instant-measure against bad quality of random numbers on
these devices, add a few bytes of entropy from the build system
on the image

after discussion with wbx@ who taught be that flash writes are
not cheap I refrained from ever automatically updating this file
(e.g. via cronjob and on shutdown), so if you want persistent
entropy, you'd do the following command manually on the box:
# dd if=/dev/urandom of=/etc/.rnd bs=512 count=1

once we have our fuse/nvram thingie, entropy ought to be stored
there, and if a new image is flashed, the old and new entropy
ought to be merged. i'll tend to this when js@ is done.

* base-files:
  - BR2_RUNTIME_PACKAGES_URL needed unquoting (my bad for doint
    the quotes, markus' for not removing them ;)
  - asterisks need to be escaped in regexps, see re_format(7) if
    you're on BSD or http://www.mirbsd.org/man7/re_format.htm otherwise
* target/Config.in:
  - this small one-liner (first hunk of r278) by wbx was the cause
    of all these reboots... looks as if the LP module (what's that,
    line printer?) causes them (and my Asus doesn't even have a
    parallel port), revert and unbreak the long-unnoticed thingie

Now it builds again (at least my config, I'm too tired to test
another one, even the default config), and I can login again (with
the regexp fix), but

* ssh logins with public key do not work at all
* the root and admin password are displayed in plain text in the build log
* I cannot choose my default shell for the admin user to be /bin/mksh even
  if I enabled mksh
* the default root shell is /bin/ash should be /bin/sh instead IMHO (just
  in case someone fucks up with the user shell)

At least we got a buildable and runnable system again,
and I learned how much suckage is hidden here... and we
all know we'd better test before commit and develop major
stuff in a branch (note to self and all others: read up
on branching and especially merging in the svn book or
better switch to CVS).

fixed second double quote issue

fixed doubled quotes

* never use && in Makefiles, always use || and reverse test logic
* quote properly
* missing line continuation chars

this is what you people get for using := ...

also fix some indentation while here

without this patch, mkpasswd cannot ever have worked on GNU/Linux even

- due to a hint form wbx I changed the description and removed special
  '-KEY-' string handling from Makefile

- added runtime config for password
- added runtime config for ssh-key
- added mkpasswd for crypt-md5 generation (needed for runtime password config)

instead of setting WRKBUILD, set WRKSRC and let WRKBUILD?=WRKSRC like mirports
(WRKSRC isn't used anywhere though)

* whitespace cleanup, while here
* move WRKDIR and DISTFILES definition into PKG_template
* define WRKDIR via parent BUILD_DIR instead of going up from WRKSRC
* use a sub-directory as WRKDIR, like BSD
* rename PKG_BUILD_DIR into WRKBUILD
* define WRKBUILD relative via WRKDIR
* rename PKG_INSTALL_DIR into WRKINST
* fix a few dependencies
* make use of common cases in the definition of WRKBUILD and
  WRKINST, removing many redundant lines
* use WRKDIR instead of BUILD_DIR in places where the latter
  was used previously
* while here, remove PKG_CAT stuff

this only affects packages, for now

- added "Runtime Configuration" to menuconfig
- added a package URL preselection to "Runtime Configuration" menu

some files don't need our EXTRACT_CMD
empty DISTFILES will do

fix ipkg.conf

_really_ fix svn revision ;)

ok wbx@

do not create /www by default

add OpenWrt trunk revision 3830.