-n has been replaced by the "announce-interval" config option.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

This is equivalent to the -n command line option, which will be removed.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

-i has been replaced by the "domain-number" config option.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

This is equivalent to the -i command line option, which will be removed.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

-l has been replaced by the "latency" config option.
This patch also removes cmd_line_parse_two(), which is now useless.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

This is equivalent to -l NUMBER,NUMBER command line option, which will be
removed.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

-a has been replaced by servo-pi in the config file.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

This is equivalent to -a NUMBER,NUMBER command line option, which will be
removed.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

This get things ready for introducing the servo-pi config option, which
takes two comma separated integers.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

If interface name is empty, look for a network interface with the same
name of the port.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

-b didn't actually work. Even with a single interface and no config
file, the following error message was printed [arch-unix]:

Can not use -b option in multi-link conf

The bug was caused by a problem with the initialization value
of ppg->nlinks.

Removing this option looks wiser than trying to fix it, because
most users just work with a configuration file or -C on the command line,
so fixing -b would just increase the risk of breaking non-unix architectures
with no actual gain.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

For long liks (like 180km) "(signed)s->picos_mu" overflows causing wrong
result of the comparison:
(signed)s->picos_mu < (signed)big_delta_fix)
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

I need to be more careful before pushing to master...
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

comments used to extend to the next terminator (including the
semicolon), but editors paint comment-color the whole line, so we'd
better fix it rather than documenting an uncommon behaviour.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>

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

Even if we count failure events for SNMP (in the switch), we actually
never generate them.  The code has always been tracking phase for
any offset smaller than one clock tick (8ns or 16ns), but we'd better
jump in phase (and leave TRACK_PHASE servo state) and this is a failure.

With this commit, to loos track_phase in wrpc you can just force
the setpoint far enough from the calculated one: e.g.

  [..] cko:2 setp:929 [...]

  wrc# pll sps 0 500

  [...]  ss:'SYNC_PHASE' [...] cko:431 setp:1367 [...]

then it takes 15 iterations to converge, but that's because the
calculations the first time (as shown) are wrong. A real glitch
loosing track_phase would be recovered in 3 iterations.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

This is needed for profiling in wrpc-sw.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

when using VLANs, reception of each VLAN-tagged ptp frame resulted
in printf(), which was syslog-ed to the server... changed printf
into pp_dial(ppsi, frame, 1,...)

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

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 commit has to be preserved as it is included in WRS release 4.2

for tx stamps, when we added the check about the frame content
(to avoid problems with high data load, see 4179fcdc), it got broken
for UDP. That's because the cmesg we get back includes a raw frame
anyways, even if we transmitted UDP or other protocols.

So now check for the trailing part of the frame we get back, to
avoid any extra leading header whatever its lenght.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

During track-phase we used to find the setpoint and then track changes in
the master-to-slave delay, thus not zeroing any initial error that may
happen (e.g. the initial sample was more jittery than average).

Now we use the clock offset as correction source. Setpoint is
adjusted by 1/4th of the current offset, to smooth a little the jitter.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is likely the result of my cleanup of wr-servo, where I forgot
some pieces.

Greg collected some interesting logs in wrpc where the setpoint
was calculated wrongly. This fixes the thing and removes some redundancy.

I "git diff" order:

- avoid WR_UNINITIALIZED, really unused. Do what's needed at
  wr_servo_init() time

- adjust phase to 0 at init time, where s->cur_setpoint is set (we
  really should set and use in a single unified place)

- set delta_ms_prev before entering TRACK_PHASE, not earlier where it
  doesn't fit

- avoid extra checks of non-zero offset.seconds and
offset.nanoseconds, because they are checked at the beginning anyways.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This adds -Wstrict-prototypes -Wmissing-prototypes and takes actions
accordingly.

Some of the solutions are ugly, especially in the bare architectures,
but I prefer to be warned if I take dangerous paths wrt prototypes while
writing further code.

I'm upset by those two not being included in -Wall, as they used to.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The existing rule to check whether WRPCSW_ROOT is properly set (by
environment or default in Kconfig) forced a dependency on all object
file.  Thus, all files were rebuilt every time.

We now force the check only as a dependency for the first object file,
which means ony fsm.o (and then ppsi.o) is built if you run "make" on
a just-built source tree.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This reverts commit 1d5a461d.

actually, the error count must be zeroed every time, not only after
printing: we want to print if the error persists for 5 iterations, not
every 5th time.

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

This is a temporary workaround. We are going to fix the tx/rx functions
to received the header and payload together, but currently they are
separate so the offset must be 0 for raw ethernet.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

It's done to prevent PPSi leaving Slave state on the WRS when high load of
background traffic is present. It's just a temporary workaround. In the future
the HDL of the switch has to be fixed so that PPSi frames have higher priority
than the normal traffic and are not dropped when WRS is overloaded.