We have three temperatures, called "roof", "core", "case"
(yess rubi like word same lengt, s/he most love four).

There is a "faketemp" command that can set one two or all of them.
This allows to test negative temperatures in printouts (White Rabbit
in Siberia does it, but we southish can't test easily).

All temperatures are initialized to invalid. Example:

   wrc# faketemp -10 -2.5
   wrc# faketemp
   fff60000 fffd8003 80000000
   wrc# temp
   pcb:84.7500 roof:-10.0000 core:-2.4999 case:INVALID
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This costs anoter 220 bytes in the binary, becuase we now loop among
all temperature providers in order to return data to the caller.

The providers are clustered using an ELF section.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This commit adds data structures, methods and a process to manage
temperature.  Currently the only temperature being read is the onewire
thermometer, so there is no difference with what we had before.
The cost is 1.1 kB in binary size.

Well, there is a difference: only the stat engine was reading
temperature, while now the thermometer is read continuously (every 15
seconds, or another Kconfig-urable interval).  This takes quite some
CPU time (18.5ms per polling loop, but allows any user to read
temperature (using the API we now provide) without re-reading the
onewire bus.

The real aim of this is monitoring a board-specific set of sensors,
as implemented and show in the next commit.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Also, this changes how to mac address is retrieved in patching pfilter
rules. I'd better use format_mac(get_mac) for the verbose message,
so get_mac is also the source for patching the rules.

register and get_mac/set_mac are guaranteed to be in-sync.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

These numbers are only available at build time; there is no run-time
command to change them as yet.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

This allows a node built with CONFIG_VLAN to choose at run-time
whether vlans are allowed or not.  We build both pfilter rule-sets,
and according to the active vlan number we write one or the other.

The case "CONFIG_VLAN is not set" is handled by providing default
NULL pointers for beginning and end of the vlan rule-set.

As a side effect, I renamed the files and the functions in
the pfilter-builder.

This is how it works for me, with CONFIG_PFILTER_VERBOSE set,
so you see different rule-set  are used (29 rules vs. 16 rules):

    wrc# vlan set 0
    0 ("0") out of range
    Command "vlan": error -22
    wrc# vlan off
    fixing MAC adress in rule: use 22:33:44:55:66:77
    pfilter rule 00: 4.00000000
    pfilter rule 01: 1.e4466013
    [...]
    pfilter rule 28: 8.00000000
    current vlan: 0 (0x0)
    wrc# ip set 10.0.0.2
    IP-address: 10.0.0.2 (static assignment)

  (and I am now reachable by untagged frames)

    wrc# vlan set 10
    fixing MAC adress in rule: use 22:33:44:55:66:77
    fixing VLAN number in rule: use 10
    pfilter rule 00: 4.00000000
    pfilter rule 01: 1.e4466013
    [...]
    pfilter rule 15: 8.00000000
    current vlan: 10 (0xa)

  (and I am now reachable on vlan 10)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

Conflicts:
	Makefile

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

This allows a run-time change of the active vlan, if CONFIG_VLAN is
active at build time.  The pfilter rules are being rewritten every
time the vlan is changed (and we have one VLAN only for the CPU).

   wrc# vlan set 0
   0 ("0") out of range
   Command "vlan": error -22
   wrc# vlan set pippo
   0 ("pippo") out of range
   Command "vlan": error -22

   wrc# vlan
   current vlan: 10 (0xa)

   wrc# vlan set 20
   current vlan: 20 (0x14)
   wrc# ip set 192.168.20.2
   IP-address: 192.168.20.2 (static assignment)

      (and the node now replies to ping on vlan 20)

   wrc# vlan set 10
   current vlan: 10 (0xa)
   wrc# ip set 192.168.10.2
   IP-address: 192.168.10.2 (static assignment)

      (and the node now replies to ping on vlan 10)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This includes a wrs fix by Maciej and a change that prevents warning
when bulding with CONFIG_HOST_PROCESS (which is not there yet).  I
prefer to merge this now, so the next commit picks only its own little
change into the ppsi submodule.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

This commit adds the possibility to build an init command in the
binary.  It can be useful to install a number of nodes with a special
configuration, without the need to reach the shell and/or eeprom of
each one.

This is useful, for example, to pass syslog configuration
or the (yet to be implemented) definition of multiple vlans for
etherbone and the wr fabric (frame classes 7 and 6 f the packet
filter). Moreover, I personally love to load a master or slave node
for testing without interacting with the node (or changing the flash
over and over).

This commit also makes reading of commands from storage an optional
feature, but only in the advanced/developer options. Thus, you can
build a wrpc that uses the build-time init commands, or the
flash-stored init commands, both or none of them.

Size effects:
     * no effect if you keep the defaults (no build-time init, flash init)
     * +248 bytes (+ the command itself) if you enable build-time init
     * -1696 bytes if you remove flash-based init (discouraged)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

WARNING: wr_switch_defconfig doesn't build, this is fixed in next commit

Unfortunately this commit makes the internal communication between
minic and lan asymmetric: on receive the tag is discarded (and the
longer header falls into the payload) while on transmit it must be
provided by lan.c.

The reason is that on receive we can trim 4 bytes from the payload,
but on send we can't add 4 bytes without a memmove.  The functions
receive pointers to two different header structures, so hopefully
user errors will be signalled by the compiler.

Still, users interact with net.c, which hides vlan completely from
them.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

We used to have "struct ethhdr" local in net.c, passing "void *hdr" to
minic.  We exports the structure to an header so to use it in
communicating with the minic.  Unfortunaltey, ethhdr is defined elsewhere
too, so we'd better rename to wr_ethhdr.

This is a step towards the introduction of vlans, but has no technical
effect by itself.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

the UDP checlsum was miscalculated for odd-sized payload. The last
byte was not used in calculation.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

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

This branch introduces the "ps" command, based on a restructuring of
how main() calls the various activities.

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>

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

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

This commit turns individual calls to polling functions into an array
of tasks.  On this basis we can add profiling and periodic execution
(instead of continuous polling).
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

The compiler is smart enough to turn this into a multiplication by 8:

    cf54:       3d c1 00 03     sli r1,r14,3
    cf58:       59 a1 00 00     sw (r13+0),r1

and any real division by ll would be trapped by ./check-error.c .

Still, I'd better simplify the code.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

All "tasks" must have the same prototype, so we'd better get
void in all of them at the time being (so we only change the nerwork
function).

Meanwhile, use value "0" for "link down", so it can work as a simple
enable flag for the IP polling code.

(note: we query the link status in wrc_ptp_update() too, and we should
rather refer to the global variable)
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This solves http://www.ohwr.org/issues/1218

We used to have a "needIP" variable, but we need three states:
if the address was set by bootp, we run bootp again when the link goes
up, but if it was a static assignment we do not (and keep the static one).

So I renamend needIP to a 3-valued ip_status. The rename ensured the
compiler found all users.  Also, the messages are now 3-headed, so
the user know whether the address is static or it comes from bootp.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

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

Again, it returns 1 if it did some calculation. For profiling.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is a step towards profiling. The functions being polled return
whether or not they did actually work.
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 is needed for the syslog code to know about link up and down
events, to report a link-up with the delay since last link-down.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>