The structure was not updated when the FAULT timer was removed.

After reviewing the code with Maciej, this time-out does not seem to be
used in a very coherent way. We decided then to remove it.

- Make externalPortConfiguration compliant with the PTP specifications
- Optimize code for externalPortConfiguration
- Optimize code for slaveOnly
- Replace "CONFIG_HAS_P2P && ppi->delayMechanism==P2P" by an inline
function

. Better integrate of code optimization with PPSi Kconfig
. PPSi Kconfig changes to simplify the code in PPSi and avoid #ifdef ..
. Improve code optimization
. Move some declaration in ppsi.h in other files to ligthen ppsi.h
. Improvement of timers: global or instance dependent, dyn. allocation,
renaming for better diagnostics, disabling allowed, ...
. New behaviour for the PPS output generation (configurable,...)
. Move control of the timing mode from HAL to PPSi
. Optimization of code when a Protocol extension is not available
. BMCA: Calculate the ebest,... only one time and then trigger all
instances to execute the bmca state descision
. Fix bug in the calculation of the clockIdentity to bring it into
compliance with PTP spec 2008 document
. Fix a bug related to Announce message received in SLAVE/UNCALIB
. WR ext: use dynamic timer and adapt to new PPS generation
. HA ext: Fix bug in fault injection mechanism

- New L1Sync and PTP servo calculation (delayAsym,...)
- Make function calls more uniforms between extensions
- refactoring
- Add a new state machine to enable/disable an extension and provide
only PTP support: New timer, new hooks, ...
- Fix bug in time format conversion
- Code removed when an extension is not required (#ifdef XXX )
- Align the PTP and L1Sync servo calculation (shared fct, same
calculation,...)
- Force only one servo running at a given time for a given instance
- New servo reset hook : called by main loop
- Add function to converte a time into a string
- Force to stay in FAULTY state during 60s
- Optimise function wrs_enable_timing_output()

- Many information were store in a common static structure (to_config).
However depending on the PPSI instance configuration these information
need to be stored per instance. Mainly it concerns the rand value
(which_rand and the counter reset value). Only the timer name stay now
in a common arrea.

- Fix issues introduced in last commit "Timeout management correction"
- Code optimization
- Protect FAULT timeout in case of timeout overload. It can append
because the logMinDelayResponse can be changed on real-time by the
received DELAY_RESP message

- The timeout expressed in ms is stored in an integer value. A Log
interval may have a value in the range -128,127. The timeout cannot
covers all this range. A protection as been added to avoid wrong timeout
calculation which might generate a crash of PPSI.
- A new function is also added pp_timeout_log_to_ms() to verify the log
Interval and calculate the timeout.
- Optimize the execution of pp_timeout_set()

- Move initialization of configurable parameters in L1SYNC init hook
- Set the L1SYNC state to DISABLE in the startup otherwise the state is
not set if the port is not connected
- Remove time printing in diag messages
- restrict call to servo only when the state is UP
- Adjust properly the delay returned by each l1sync state functions
- Set REQUEST timeout value dependent of the delay mechanism (E2E/P2P)
- Bug fix: PP_P2P_MECH used instead of P2P
- Optimize the REQUEST timeout to generate asap a DelayReq message after
the first SYNC message received

This implementation does not include yet the possibility to declare 2
exclusive instances on a same port (WR & HA)
- Some structures are renamed with the suffix _t to make it more
readable.
- Some variables are renamed  to be more clear
- Implementation of masterOnly, slaveOnly and externalPortConfiguration
- All extension specific includes (XX-api.h) is now automatically
included in ppsi.h and do not need to be included in other files.

standard IEEE1588-2018

We bring here a lot of changes to :
- make the implementation more complaiant with the standard
IEEE1588-2018.
  We introduce new structures for the time (TimeInterval, Relative
difference)
  Named used in structures are aligned to the standard, ...
  It is just a first step. More work is needed
- be able to use more than one protocol extension at the same time
  (configurations, structures, ...)
- add the concept of profile
- increase diagnostics by displaying also standard PTP servo
informations

BMC fixed to be called periodically and data sets comparisments fixed
also changed the foreign master table adding and agging and some
minor adpatations in state pre-master

If we stopped sending to the master or the peer (for traffic or
whatever; in my case with "fault drop"), we wouldn't notice the
problem.

This looks like SYNCHRONIZATION_FAULT (9.2.6.12), so this reuses the
almost-unused TO_FAULTY, renaming it to a more generic TO_FAULT.

9.2.6.12 says we should reach uncalibrated, but since uncalibrated doesn't
exits (it is never entered, it's dead and untested code at this point),
I handle the problem just like the timeout receiving announce messages.

For wr, I reset the servo, so the problem can be seen.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

We just reinit timeouts on reception of a delay response This way the
delay request interval is reinitialized and set to the proper value.

previous code failed for logSyncInterval < 0
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

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

This kills the ambiguously-named pp_ms_to_timeout(), and
adds pp_next_delay_{1,2,3}.

This commit meanwhile fixes the timeouts in the pdelay master (who didn't
consider the request timeout), and in the pdelay listener (because
LISTEN mode must send pdelay requests).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

We don't zero timer any more (see preceding commit), no need
to special case anything.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This was only used by timeout_z() (check if expired and disable it)
that is not useful nor used since last commits.

Timeouts are always running (time is always running...), it's just a matter
of whether we check them or not.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

It was not a good idea: there is no need to clear a timeout if
it is being re-initialized in the same function anyways. If, OTOH,
we change state, it won't be referenced any more so there's no need
to disable it.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

We now have the simpler pp_timeout(). Moreover, most uses were
needless, because now we reset all timeouts when we change state.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

Several changes to the timeout API, simplifying the code and allowing
more simplification later.

 * TO_ANN_INTERVAL renamed to TO_ANN_SEND (as opposed to TO_ANN_RECEIPT)

 * TO_SYNC renamed to TO_SYNC_SEND for symmetry with above

 * added timeout_init() to take not of the waiting values

 * timeout_set now receives only the index, not the value (see above)

 * __timeout_set added for those that want a specific value

 * timeout_rand is gone, as timeout_set knows which to randomize

 * all users fixed
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This moves all functions to timeout.c, so the header is more readable.
As a side effect, the binary is smaller (a few hunded bytes for all
archs, almost 1k for wrpc).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

We use the same timeout for delay_req and pdelay_req, so let's call
it "request" alone.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The timeout for sending a requeste can be shareby by delay/pdelay;
same for logMinDelayReqInterval.

Moreover, I took the chance to remove the third extension-specific
timeout slot (WR only uses 1, keep two of them...).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is a long-standing bug introduced when fsm.c was changed to
immediately enter a new state (before that, we had a grace period in
state-faulty).  The problem is exposed by running the suggested
config file, if eth1 and/or eth2 exist but are currently down.

This patch introduces a specific timeout, so we wait 4 seconds before
reinitializing the pp instance; we thus avoid burning 100% CPU Time in
trying again and again.

As a side effect, this reduces timeouts provided to extensions from 4
to 3. I could have reused one of the other timeouts, as faulty state
does not use them, but I'd better have proper timeout names in
diagnostics.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This changes all "Name Surname for CERN" to "Copyright CERN, author
Name Surname".  This is trivial and makes no change, with a few
expections:

- all trivial Makefiles that had "for CERN" lines have now no
attributions, as they are trivial anyways.

- a few unrepresented files have not the copyright notice (especially
ptpdump, that I wrote in 2012).

- a few "for CERN" remain in subprojects, where I didn't touch anyrhing
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This "type" was the only type represented as an array instead of
being a structure. This leads to inconsistent use: if you memcpy
the other ones you need "&", but not here.

So this commit uses an encompassing structure, to have ClockIdentity
behave like other composite fields.  The commit has no effect at all
on the generated binary files for wrpc but it unexpectedly reduces the
size of the output on the PC. Anyways, the commit is trivial, and all
changes can be easily verified to have no effect.

While this increases entropy, the reasoning is that I want to assign
structures instead of running zillions of memcpy functions, with the
hairy PP_CLOCK_IDENTITY_LENGTH.  So later commits, based on this one,
reduce entropy.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is the result, on a master, with "-d 02":

diag-time-2-eth0: posix_time_get:  1362675879.940187923
diag-time-1-eth0: timeout expired: PP_TO_ANN_INTERVAL
diag-time-2-eth0: posix_time_get:  1362675880.081818439
diag-time-2-eth0: posix_time_get:  1362675880.233227916
diag-time-2-eth0: recv stamp: 1362675880.233227916 (user)
diag-time-2-eth0: posix_time_get:  1362675880.233499803
diag-time-1-eth0: timeout expired: PP_TO_SYNC

(time level 1 reports timeouts and set, level 2 reports get-time too).
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 frames should be sent with some dispersion around the timeout
point.  The new function calculates the correct dispersion, but we
need to change the setting of next_delay in order to actually spread
the distribution of announce and other messages.

Also, this commit doesn't spread all message times, some work on the
field is still needed.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

By zeroing the timeout whenever it expires, a number of
state files get simplified a little. So pp_timeout_z() returns
1 or 0 like pp_timeout(), but also clears the timeout when it is
expired.

Moreover, when a timeout expires a message is printed if the
build-time option pp_verbose_time is set (and run-time verbosity is at
level 2). In this way all timeouts are reported, or none of them.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>