There was an error in the overflow check of coarse values.
This replaces "> 125*1000*100" with ">= 125*1000*1000"
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

These amounts are *signed*, so we add this kludgy duplication
of code to handle the case of addition of a signed quantity
in fd-irq.c

The arithmetic code was suggested by Alessandro Rubini,
replacing a former, buggy and kludgy implementation by
David Cobas.
Reported-by: Alessandro Rubini <ru@gnudd.com>
Reported-by: Tomasz Wlostowski <twlostow@cern.ch>

This ugly hack (hardcoding the base address of slot 1 fdelay
core) is needed in the SVEC, as we have yet no means to discover
that core's base address via SDB (and will never have, probably).
The slot number of the core would probably better be taken from
a register in the core itself.

Reported-by: Tomasz Wlostowski

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>

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>

lib/fdelay-pulse-tom: dump start/end/delta timestamps. Default pulse width to 250 ns if not specified
Signed-off-by: Luis Fernando Ruiz Gago <lfruiz@cern.ch>

Signed-off-by: Luis Fernando Ruiz Gago <lfruiz@cern.ch>

The new calibration is now there, but some details are still missing,
like using the DAC value for the VCXO and validating against the
installed gateware.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

This is a massive change, spread across several files.
     - the calibration structure is changed
     - it doesn't live at offset 0x1800 any more
     - it must beloaded by SDB structures (not there yet)
     - all calibration users are fixed
     - calibration.c is a new file dealing with it
     - i2c.c has some material less, because of calibration.c
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

The SDB library is an exact copy of the sdbfs/lib directory of repo:
   git://ohwr.org/hdl-core-lib/fpga-config-space.git
at version "sdb-v1.1" (commit de2c98f).

It actually is the same code that is already in fmc-bus/sdb-lib, but
until I finalize eeprom SDB support within fmc-bus I need a copy here.
This directory will disappear when fmc-bus is completed in that respect,
but we need a release with SDB in the eeprom now for fine-delay.
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>

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

This old FIXME is now fixed. Unfortunately the calculation involved is
very boring, given the range of values we deal with.

With this in place, we can now generate a 10MHz square wave and
simulate the inputs for a WR grandmaster.
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: Federico Vaga <federico.vaga@gmail.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>

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

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

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