When the switch is loaded, some PTP outgoing frames get discarded, so we
never get back the timestamp interrupt.  By round-robin'ing the slots,
we get back to the "busy" one, which holds an skb structure that must be
released.

Without this freeing, the skb remains in the output queue of the
socket, and after a while we get EAGAIN in send() because the buffer
is full (400 frames, more or less). And the ptp port is stuck until we
reopen the socket.

This diagnosis exposed the fact that it's not an overflow of pending
descriptors, so I change the message from

  descriptor overflow: tx timestamp pending

to

  discarding tx frame that got no timestamp

Finally, this is pr_warn, not pr_err as we now know it's expected.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

When process is restarted more than 5 times within 10*10 seconds (10 cycles,
each 10 seconds), then monit triggers system restart. Save information, which
process caused restart. Store it in flash to survive over restart, then copy
to /tmp.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Shorten timeout for opening shmem (wrs_shm_get). Clients use their own sleep
between tries.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

SNMPd will try to connect to HAL's and PPSI's shmem at startup. If HAL or PPSI
is not started SNMPd would continue instead of exiting. Later when data from
shmem is needed SNMPd will try to open shmem again. By this it would be
possible to figure out what is happening on switch when HAL, PPSI or RTUd
failed to start.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Report when wrs_watchdog is not present in process' list.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Retry next time when there is an error in "run_once" sections.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

When parameter -g is used there is no need to check other parameters.
Change to avoid waiting for HAL's shmem to be available.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Make possible to start snmpd without successful open of RTUd's shmem
(in other words having RTUd running). SNMPd shall start even on faulty system
since it is a way to report if and what is wrong.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

in file wrsTimingStatusGroup.c
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Limit toc subsections depth to 2. Means don't include listing of failures into
toc.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Remove details of WR-SWITCH-MIB, moved to the wrs_failures document.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

There is no need to expose PTP daemon name in MIB.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Warnings like below are still not solved.
warning: node `XXX' must be contained in at least one conformance group

Command used for smilint:
smilint WR-SWITCH-MIB.txt
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Move wrs_failures from white-rabbit repo to doc/
copy wrlogo

It looks like it is a leftover from the past
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

latex creates such file during run
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Don't remove all pdf files when make clean
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Change document's name to "White Rabbit Switch: Failures and Diagnostics"
Add myself as an author
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

DONE "Temperature inside the box too high"
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

As agreed with Greg portSfpCopper is not needed because portSfpError will
notify about not valid SFP configuration.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

PTP related objects should be called WR-SWITCH-MIB::ptp* instead of
WR-SWITCH-MIB::ppsi*. People need to know it's PTP related, not necessary that
the running PTP engine is PPSi.
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

DONE: "Not supported SFP plugged into the cage (especially non 1-Gb SFP)"
UDATED: "Detected SFP not supported for WR timing"
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

Signed-off-by: Adam Wujek <adam.wujek@cern.ch>

implemented WR-SWITCH-MIB::ppsiClockOffsetPsHR
Signed-off-by: Adam Wujek <adam.wujek@cern.ch>