This commit moves the time operations inside the instance.  It
allows to have different time operations for each instance, which will be
useful in the future. For example, one ethernet card may White Rabbit,
and another may not be.

This commit, like moving network operations inside ppi, requires a
change in wrpc-sw.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is needed for the next commit, where time operations become local
to the specific pp instance.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This commit moves the network operations inside the instance.  It
allows to have different operations for each instance, which will be
useful in the future. For example, one ethernet card may have hardware
timestamping and another may not.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This unifies all bare-linux common code. It's an out-of-context commit
(as I'm not adding bare-arm today) but I prefer to avoid doing the same
change twice in things that are going to change these days.

BTW: this commit introduces no changes in the binary files generated,
with the exception of a small difference between i386 and x86-64 in
the receive buffer.  This keeps the x86-64 version but i386 was not
working anyways (and still doesn't work).  That bug fix is in my queue,
but I've other things to do first with ppsi.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

pp_timed_printf is a great resource, but we need to temporarily
silence diagnostic printing in  time_ops.get, in order to be able to
tell stamping-related calls to just printf-related ones.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

I'm going to split diagnostics into groups based on flags. And
with different diagnostics for each pp instance, to be ready to
debug multi-link situations.  This commit adds the flags in the proper
places and the first flag value, used in the next commit.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Too many messages are confusing: let's make a state change more
visible than a re-enter of the same state.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

By moving calc_timeout into the time functions, we simplify
the split-out of ./time-<arch> as a separate subdirectory.

As a side effect, calc_timeout disappears from arch-wrpc, and
wrpc-sw needs to be adapted to this commit, to provide it
inside its own time_ops.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The structure was separately allocated with a size of 16.  This avoids
the alloc and makes the array the right size from the enum of message
types. We thus a remove a number of magic "16" in the code.
wrpc-sw must be updated too, in order to use this commit.

There's still a problem: the code claims the number to be "last sent",
but actually the value is incremented after sending. I'll need to pass
again on this thing.

ps: I changed name from "sent_seq_id" to "sent_seq" to the compiler
complains for previous users, otherwise undetectable.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

We are not even trying to build the complete executable here, so
remove crt0.S and any references to it.

I forgot these pieces in commit
   arch-spec: removed unused source files
which happened 101 commits ago.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

While some parameters are still there, for future use when we'll put
back the pdelay mechanism, the default mechanism is clearly not
pdelay, and incorrectly-used parameters are abetter removed.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Actually, there's no need to build with three pp_printf any more: we
know pp_printf is sound. So let's just print once with "xint"
pp_printf and once with all messages activated, so see the worst size
we get to.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This fixes a bug introduced in 44deab24 (97 commits ago), where I
split WR_DSPOR(ppi) out of DSPOR(ppi), but not all diagnostics has
ever been built. Next commit fixes this, and allowed me to make this
very fix.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

If the state just executed switched to a new state, we should enter
that new stat immediately, and ignore ppi->next_delay.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The various "goto out" in case of error should report the error.
We don't have them at this point, but they are better checked.

And, at exit, there's not need to clear the timeouts.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

Also, remove unused management macros and a wrong "+ 4" in announce
length.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

Thanks Aurelio for reporting. You compiler is bigger than mine.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This patch-set rewrites the timers into much simpler timeouts.  Code
is smaller, binary is smaller, we have less arch-dependent functions
and so on.  Meanwhile I noticed and fixed some buglets that got
unexposed or unnoticed.

The function st_com_restart_annrec_timer is now just a one-liner, and
it really is just a timeout operation.

So call it pp_timeout_restart_annrec instead, and have it inline in
ppsi.h, removing it from common-fun.c
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>

It look like the timeouts were not restarted after they expired (and
the respecti message is sent).  The previous commit, that disables
expired timers automatically, exposed this bug.  Unless I misunderstood,
the 1s timing of announce message was due to the 1s-long delay
between re-invocations of the state-master code.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

commit 2ce296a5 removed this stanza, but it looks like we actually
need to force IP_MULTICAST_LOOP to zero.

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>

This kills all WR-related timeouts because one of them
is enough (they are used in each state as a timeout for
completion of that very state).

Moreover, since ppsi should be extension-independent,
we now define 4 timeouts for the extension in use (and WR
only uses timeout 0).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This bug has no effect, but I'd better commit it separately for
documentation purposes.  Actually, there is not need to clear the
timeouts when they are state-specific and we are changing state.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

This fixes the bug exposed by the previous commit.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

The timer implementatin was incredibly complex.  Now there is no such
thing as a timer in ptp. We only have timeouts.

A timeout is a point in time. We calculate; we can be earlier; we
can be later. The only thing which is arch-specific is calculating.

This replaces every "timer" operation with the equivalent "timeout"
one. All stuff is renamed to avoid confusion between "time" and
"timer".

Unfortunately, the resulting program as of this commit is not working.
the pdelay_request timeout happens continuously, I'm sure I've exposed
a latent bug, which I'm now going to fix.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

We had bare_timeval with both nsec and usec, but this doesn't work.
As a result, our timers were only second-grained and not
millisecond-grained (we got no nsec from clock_gettime).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is a subset of <linux/jiffies.h> with typecheck from
<linux/typecheck.h>.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This small patch-set removes the MSG_SEND_AND_RET function and its
variants.  I find it extremely hard to read multi-line macros,
and also having a function returns implicitly by calling a macro.

I understand why this was born, but now I feel really better with the
code.

This removes the unmanageable MSG_SEND_AND_RET macro, turning it to a
normal function (inline, at the time being, to avoid moving from the
header to a different place).

The function is called __send_and_log because it logs too, but I'd better
move logging to the actual send and recv methods.

BTW: the message "-> FAULTY" (now removed) was wrong, because not all
errors of this function result in FAULTY state in the fsm.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This kills MSG_SEND_AND_RET_VARLEN, turning everything into
MSG_SEND_AND_RET. Such macro now receives the length argument and
doesn't receive the "pdelay" argument any more, because it is zero by
design in this implementation.

This commit has no effect on the compiled files.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

I'm sorry, but I need names better than x,y,z,w . This is my first
step in getting rid of this complex macro.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

No effect: the final program is still called ptpdump.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>