The new Kconfig option allows dowloading ppsi.conf at boot time, from
TFTP or otherwise.  The filename in the URL can include IPADDR and/or
MACADDR, so the same configuration string can be used to set up a
batch of switches with different configurations.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This factorized configuration stuff out of wrap_rootfs into a build
target.  It also adds building and installing "conf" so to allow the
following commands in the target (in /wr/etc):

	  make config

	  make oldconfig     (likely after editing dot-config)

	  make defconfig

I'm unable to build menuconfig at this time, because of missing
symbols in the target's libncurses. I hope to fix this when we
bump buildroot up (thanks Adam for that!)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Otherwise "Received SNMP packet(s) from UDP" is written to file everytime the
agent is asked for the information - if we write to ramdisk, RAM becomes full
over the weekend.

Built from commit e38409d

We need it for some copper SFPs, e.g. Cisco MGBT1

We still need them if somebody want to synchronize a regular (non-WR) PTP
device.

For security reasons, php user sessions are destroyed after 10 minutes
of inactivity. After this period, users must login again.

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

wr_mon must be seriously audited, and turned back to be ptp-based
instead of hw-based (we may have two ptp ports on the same hw port).

Meanwhile, this adds the peer's MAC address and the PTP state for each
port in "up" state, if ptp is running there. It's not perfect, but
I find it  useful.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The link lenght is most often wrong, because of missing calibration,
coarse approximation of the diffraction ration and a huge integer
division before multiplication.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
Acked-by: Grzegorz Daniluk <grzegorz.daniluk@cern.ch>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is the sequence of commits, in --reverse order (oldest on top):

   6cdecd8 wr-servo: bugfix for unlikely whole-second offset
   f47f565 wrs-socket: retry getting a tx_timestamp on failure
   d8718e5 wr: trivial: add tracing of the hooks
   de0f017 proto-starndard: invalidate T1 when entering slave mode
   6be7500 proto-standard: call hook->handle_announce for all announce received
   029fbf3 wr: trivial: remove a never-used IPC command
   832bd0d wr-servo: move generation of the state name earlier
   11c199b WR: set system time from WR time
   6ecdf97 wr-servo: run twice as fast
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The "pahse tracking: ON" line was confusing.  It's only role is saying
if tracking has been disabled (by wr_mon itlsef, because a user typed
"t").  So the message is now absent unless tracking is forcibly disabled.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

And force the compiler to complain. It used to be on by default, bah....
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Meanwhile, this cleans up declaration of functions that were not prototypes
in wrsw_hal.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

If you re-build after stuff has been added to Kconfig, the .config you
already have in place is wrong. This means the dot-config at runtime
will have missing items.

We thus force a silent "make oldconfig" at each build step, to ensure
user preverences are saved, but new options get the default value.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Endpoint mode was not correctly displayed in endpointmode.php and
endpointcalibration.php when dotconfig CONFIG_PORTXX_PARAMS did not
include the protocol column.
Now, the forms are correctly displayed with and without the protocol
column in dotconfig.
NOTE: we assume that the protocol column is there or not for all ports or for none of them. (not mixture)

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Text was not fitting the box. It has been reduced by 1.

Adding new functions to get method used, salt and counter for the
different type of encryptation methods.
By this moment, the web interface encrypts using md5 with a random salt
by default.
Old method removed.

Updating password from the web interface now modifies dotconfig instead
of phpusers file.
For security reasons the web interface only saves the password
encrypted.

Login now uses the user and password from dotconfig instead of phpusers
in filesystem.
Now it is suposed to use same username and password for the shell and
the web interface.
Changes are also applicable for both of them.
Nowadays there is only one user: "root"

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The apply_dot-config script was wrong: it wouldn't force on the
empty root password. Thus, if it is set to something non-empty
and then changed (by a network dot-config) to empty, it won't work.

This bug only revealed on nfs-root setups.

Now, it's stupid to keep two symbols (ROOT_PWD_CLEAR and ROOT_PWD_CYPHER),
and I could simplify the script. And re-test. And change docs. No thanks.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>