2017-07-15 10:35:02 -05:00
|
|
|
.. -*- Mode: rst -*-
|
|
|
|
|
|
|
|
.. include:: ../etc/definitions.rst
|
|
|
|
|
|
|
|
.. URLs that changes between the various backends.
|
|
|
|
.. _Stratus Documentation: file:///usr/share/doc/coriolis2/en/html/stratus/index.html
|
|
|
|
|
|
|
|
.. |ChipStructure-1| image:: ./images/chip-structure-1.png
|
|
|
|
:alt: Chip Top Structure
|
|
|
|
:align: middle
|
|
|
|
:width: 90%
|
|
|
|
|
|
|
|
|
|
|
|
.. _Python Interface to Coriolis:
|
|
|
|
|
2019-05-27 11:49:51 -05:00
|
|
|
|newpage|
|
|
|
|
|
|
|
|
|
2017-07-15 10:35:02 -05:00
|
|
|
Python Interface for |Hurricane| / |Coriolis|
|
|
|
|
=============================================
|
|
|
|
|
|
|
|
The (almost) complete interface of |Hurricane| is exported as a |Python| module
|
2019-06-04 07:32:23 -05:00
|
|
|
and some parts of the other components of |Coriolis| (each one in a separate
|
2017-07-15 10:35:02 -05:00
|
|
|
module). The interface has been made to mirror as closely as possible the
|
|
|
|
C++ one, so the C++ doxygen documentation could be used to write code with
|
|
|
|
either languages.
|
|
|
|
|
|
|
|
`Summary of the C++ Documentation <file:../../../index.html>`_
|
|
|
|
|
|
|
|
A script could be run directly in text mode from the command line or through
|
|
|
|
the graphical interface (see :ref:`Python Scripts in Cgt`).
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
Aside for this requirement, the python script can contain anything valid
|
2019-06-04 07:32:23 -05:00
|
|
|
in |Python|, so don't hesitate to use any package or extension.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
Small example of Python/Stratus script: ::
|
|
|
|
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
import symbolic.cmos
|
|
|
|
from Hurricane import *
|
|
|
|
from Stratus import *
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
def doSomething ():
|
|
|
|
# ...
|
|
|
|
return
|
|
|
|
|
|
|
|
def ScriptMain ( **kw ):
|
|
|
|
editor = None
|
|
|
|
if kw.has_key('editor') and kw['editor']:
|
|
|
|
editor = kw['editor']
|
|
|
|
stratus.setEditor( editor )
|
|
|
|
|
|
|
|
doSomething()
|
|
|
|
return
|
|
|
|
|
|
|
|
if __name__ == "__main__" :
|
|
|
|
kw = {}
|
|
|
|
success = ScriptMain( **kw )
|
|
|
|
shellSuccess = 0
|
|
|
|
if not success: shellSuccess = 1
|
|
|
|
|
|
|
|
sys.exit( shellSuccess )
|
|
|
|
ScriptMain ()
|
|
|
|
|
|
|
|
This typical script can be executed in two ways:
|
|
|
|
|
|
|
|
#. Run directly as a |Python| script, thanks to the ::
|
|
|
|
|
|
|
|
if __name__ == "__main__" :
|
|
|
|
|
|
|
|
part (this is standart |Python|). It is a simple adapter that will
|
2019-06-04 07:32:23 -05:00
|
|
|
call :cb:`ScriptMain()`.
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|
|
|
|
In this case, the ``import symbolic.cmos`` statement at the begining
|
|
|
|
is mandatory.
|
|
|
|
|
2017-07-15 10:35:02 -05:00
|
|
|
#. Through |cgt|, either in text or graphical mode. In that case, the
|
|
|
|
:cb:`ScriptMain()` is directly called trough a sub-interpreter.
|
|
|
|
The arguments of the script are passed through the ``**kw`` dictionnary.
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|
|
|
|
In this case, the ``import symbolic.cmos`` statement at the begining
|
|
|
|
may be omitted.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
+----------------------+-----------------------------------------------+
|
|
|
|
| \*\*kw Dictionnary |
|
|
|
|
+----------------------+-----------------------------------------------+
|
|
|
|
| Parameter Key/Name | Contents type |
|
|
|
|
+======================+===============================================+
|
|
|
|
| ``'cell'`` | A Hurricane cell on which to work. Depending |
|
|
|
|
| | on the context, it may be ``None``. |
|
2019-06-04 07:32:23 -05:00
|
|
|
| | For example, when run from |cgt|, the cell |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | currently loaded in the viewer, if any. |
|
|
|
|
+----------------------+-----------------------------------------------+
|
|
|
|
| ``'editor'`` | The viewer from which the script is run, when |
|
|
|
|
| | lauched through |cgt|. |
|
|
|
|
+----------------------+-----------------------------------------------+
|
|
|
|
|
|
|
|
|
|
|
|
Plugins
|
|
|
|
~~~~~~~
|
|
|
|
|
|
|
|
Plugins are |Python| scripts specially crafted to integrate with |cgt|.
|
|
|
|
Their entry point is a :cb:`ScriptMain()` method as described in
|
|
|
|
`Python Interface to Coriolis`_. They can be called by user scripts
|
|
|
|
through this method.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Chip Placement
|
|
|
|
--------------
|
|
|
|
|
2019-06-04 07:32:23 -05:00
|
|
|
Automatically performs the placement of a complete chip. This plugin, as well
|
2017-07-15 10:35:02 -05:00
|
|
|
as the other P&R tools expect a specific top-level hierarchy for the design.
|
2019-06-04 07:32:23 -05:00
|
|
|
The top-level hierarchy must contain the instances of all the I/O pads and
|
2019-05-27 11:49:51 -05:00
|
|
|
**exactly one** instance named ``corona`` of an eponym cell ``corona``.
|
2019-06-04 07:32:23 -05:00
|
|
|
The ``corona`` cell in turn contains the instance of the chip's core model.
|
2019-05-27 11:49:51 -05:00
|
|
|
|
|
|
|
The intermediate ``corona`` hierarchical level has been introduced to handle
|
2019-06-04 07:32:23 -05:00
|
|
|
the possible decoupling between real I/O pads supplied by a foundry and a
|
2019-05-27 11:49:51 -05:00
|
|
|
symbolic core. So the *chip* level contains only real layout and the corona
|
|
|
|
and below only symbolic layer.
|
|
|
|
|
2019-06-04 07:32:23 -05:00
|
|
|
.. note:: This does not prevent having a design either fully symbolic (pads and core)
|
2019-05-27 11:49:51 -05:00
|
|
|
or fully real.
|
|
|
|
|
2019-06-04 07:32:23 -05:00
|
|
|
.. note:: The ``corona`` also avoids the router to actually have to manage directly
|
|
|
|
the pads which simplify its configuration and besides avoid
|
2019-05-27 11:49:51 -05:00
|
|
|
to have the pads stuffed with blockages.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
|bcenter| |ChipStructure-1| |ecenter|
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
The designer must provide a configuration file that defines the rules for the
|
2017-07-15 10:35:02 -05:00
|
|
|
placement of the top-level hierarchy (that is, the pads and the core).
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
This file must be names ``ioring.py`` and put into the user's configuration
|
|
|
|
directory ``./coriolis2/``
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
Example of chip placement configuration file (for ``AM2901``): ::
|
|
|
|
|
2019-05-27 11:49:51 -05:00
|
|
|
|
|
|
|
from helpers import l, u, n
|
|
|
|
|
2017-07-15 10:35:02 -05:00
|
|
|
chip = \
|
2019-05-27 11:49:51 -05:00
|
|
|
{ 'pads.ioPadGauge' : 'pxlib'
|
|
|
|
, 'pads.south' : [ 'p_a3' , 'p_a2' , 'p_a1' , 'p_r0'
|
|
|
|
, 'p_vddick0', 'p_vssick0', 'p_a0' , 'p_i6'
|
|
|
|
, 'p_i8' , 'p_i7' , 'p_r3' ]
|
|
|
|
, 'pads.east' : [ 'p_zero' , 'p_i0' , 'p_i1' , 'p_i2'
|
|
|
|
, 'p_vddeck0', 'p_vsseck0', 'p_q3' , 'p_b0'
|
|
|
|
, 'p_b1' , 'p_b2' , 'p_b3' ]
|
|
|
|
, 'pads.north' : [ 'p_noe' , 'p_y3' , 'p_y2' , 'p_y1'
|
|
|
|
, 'p_y0' , 'p_vddeck1', 'p_vsseck1', 'p_np'
|
|
|
|
, 'p_ovr' , 'p_cout' , 'p_ng' ]
|
|
|
|
, 'pads.west' : [ 'p_cin' , 'p_i4' , 'p_i5' , 'p_i3'
|
|
|
|
, 'p_ck' , 'p_d0' , 'p_d1' , 'p_d2'
|
|
|
|
, 'p_d3' , 'p_q0' , 'p_f3' ]
|
|
|
|
, 'core.size' : ( l(1500), l(1500) )
|
|
|
|
, 'chip.size' : ( l(3000), l(3000) )
|
|
|
|
, 'chip.clockTree' : True
|
2017-07-15 10:35:02 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
The file must contain *one dictionnary* named ``chip``.
|
|
|
|
|
|
|
|
+----------------------+-------------------------------------------------------+
|
|
|
|
| Chip Dictionnary |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
|
|
|
| Parameter Key/Name | Value/Contents type |
|
|
|
|
+======================+=======================================================+
|
2019-05-27 11:49:51 -05:00
|
|
|
| ``'pad.ioPadGauge'`` | The routing gauge to use for the pad. Must be given |
|
2019-06-04 07:32:23 -05:00
|
|
|
| | as it differs from the one used to route |
|
2019-05-27 11:49:51 -05:00
|
|
|
| | inside the core |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| ``'pad.south'`` | Ordered list (left to right) of pad instance names |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | to put on the south side of the chip |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| ``'pad.east'`` | Ordered list (down to up) of pad instance names |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | to put on the east side of the chip |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| ``'pad.north'`` | Ordered list (left to right) of pad instance names |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | to put on the north side of the chip |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| ``'pad.west'`` | Ordered list (down to up) of pad instance names |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | to put on the west side of the chip |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
|
|
|
| ``'core.size'`` | The size of the core (to be used by the placer) |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| ``'chip.size'`` | The size of the whole chip. The sides must be large |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | enough to accomodate all the pads |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
|
|
|
| ``'chip.clockTree'`` | Whether to generate a clock tree or not. This calls |
|
|
|
|
| | the ClockTree plugin |
|
|
|
|
+----------------------+-------------------------------------------------------+
|
|
|
|
|
|
|
|
Configuration parameters, defaults are defined in ``etc/coriolis2/<STECHNO>/plugins.conf``.
|
|
|
|
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
|
|
|
| Parameter Identifier | Type | Default |
|
|
|
|
+===================================+==================+============================+
|
|
|
|
| **Chip Plugin Parameters** |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``chip.block.rails.count`` | ``Int`` | :cb:`5` |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
|
|
|
| | The minimum number of rails around the core |
|
2019-06-04 07:32:23 -05:00
|
|
|
| | block. Must be odd and above 5. |
|
2017-07-15 10:35:02 -05:00
|
|
|
| | One rail for the clock and at least two pairs |
|
|
|
|
| | of power/grounds |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``chip.block.rails.hWidth`` | ``Int`` | :cb:`12` |lambda| |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| | The horizontal width of the rails |
|
2017-07-15 10:35:02 -05:00
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``chip.block.rails.vWidth`` | ``Int`` | :cb:`12` |lambda| |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
2019-06-04 07:32:23 -05:00
|
|
|
| | The vertical width of the rails |
|
2017-07-15 10:35:02 -05:00
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``chip.block.rails.hSpacing`` | ``Int`` | :cb:`6` |lambda| |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
|
|
|
| | The spacing, *edge to edge* of two adjacent |
|
|
|
|
| | horizontal rails |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``chip.block.rails.vSpacing`` | ``Int`` | :cb:`6` |lambda| |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
|
|
|
| | The spacing, *edge to edge* of two adjacent |
|
|
|
|
| | vertical rails |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
|
|
|
|
|
|
|
.. note::
|
|
|
|
If no clock tree is generated, then the clock rail is *not* created.
|
|
|
|
So even if the requested number of rails ``chip.block.rails.count`` is, say 5,
|
2019-06-04 07:32:23 -05:00
|
|
|
only four rails (2* ``power``, 2* ``ground``) will be generated.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
|
|
|
|
Clock Tree
|
|
|
|
----------
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
Inserts a clock tree into a block. The clock tree uses the H strategy.
|
2017-07-15 10:35:02 -05:00
|
|
|
The clock net is splitted into sub-nets, one for each branch of the
|
|
|
|
tree.
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
* On **chip** design, the sub-nets are created in the model of the
|
2017-07-15 10:35:02 -05:00
|
|
|
core block (then trans-hierarchically flattened to be shown at
|
|
|
|
chip level).
|
|
|
|
* On **blocks**, the sub nets are created directly in the top block.
|
|
|
|
* The sub-nets are named according to a simple geometrical scheme.
|
|
|
|
A common prefix ``ck_htree``, then one postfix by level telling
|
|
|
|
on which quarter of plane the sub-clock is located:
|
|
|
|
|
|
|
|
#. ``_bl``: bottom left plane quarter.
|
|
|
|
#. ``_br``: bottom right plane quarter.
|
|
|
|
#. ``_tl``: top left plane quarter.
|
|
|
|
#. ``_tr``: top right plane quarter.
|
|
|
|
|
|
|
|
We can have ``ck_htree_bl``, ``ck_htree_bl_bl``, ``ch_htree_bl_tl`` and so on.
|
|
|
|
|
|
|
|
The clock tree plugin works in four steps:
|
|
|
|
|
2019-06-04 07:32:23 -05:00
|
|
|
#. Builds the clock tree: creates the top-block abutment box, computes the
|
2019-05-24 06:24:11 -05:00
|
|
|
required levels of H tree and places the clock buffers.
|
2017-07-15 10:35:02 -05:00
|
|
|
#. Once the clock buffers are placed, calls the placer (|etesian|) to place
|
2019-05-24 06:24:11 -05:00
|
|
|
the ordinary standard cells, whithout disturbing clock H-tree buffers.
|
2017-07-15 10:35:02 -05:00
|
|
|
#. At this point we know the exact positions of all the DFFs, so we can
|
|
|
|
connect them to the nearest H-tree leaf clock signal.
|
2019-05-24 06:24:11 -05:00
|
|
|
#. Leaf clock signals that are not connected to any DFFs are removed.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
Netlist reorganisation:
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
* Obviously the top block or chip core model netlist is modified to
|
|
|
|
contain all the clock sub-nets. The interface is *not* changed.
|
2017-07-15 10:35:02 -05:00
|
|
|
* If the top block contains instances of other models *and* those models
|
2019-05-24 06:24:11 -05:00
|
|
|
contain DFFs that get re-connected to the clock sub-nets (from the
|
2019-06-04 07:32:23 -05:00
|
|
|
top level): Changes both the model netlist and interface to propagate
|
2017-07-15 10:35:02 -05:00
|
|
|
the relevant clock sub-nets to the instanciated model. The new model
|
2019-05-27 11:49:51 -05:00
|
|
|
with the added clock signal is renamed with a ``_cts`` suffix.
|
|
|
|
For example, the sub-block model ``ram.vst`` will become ``ram_cts.vst``.
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
.. note::
|
|
|
|
If you are to re-run the clock tree plugin on a netlist, be careful
|
2019-05-27 11:49:51 -05:00
|
|
|
to erase any previously generated ``_cts`` file (both netlist and
|
|
|
|
layout: ``rm *_cts.{ap,vst}``). And restart |cgt| to clear its
|
2017-07-15 10:35:02 -05:00
|
|
|
memory cache.
|
|
|
|
|
|
|
|
Configuration parameters, defaults are defined in ``etc/coriolis2/<STECHNO>/plugins.conf``.
|
|
|
|
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
|
|
|
| Parameter Identifier | Type | Default |
|
|
|
|
+===================================+==================+============================+
|
|
|
|
| **ClockTree Plugin Parameters** |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``clockTree.minimumSide`` | ``Int`` | :cb:`300` |lambda| |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
|
|
|
| | The minimum size below which the clock tree |
|
|
|
|
| | will stop to perform quadri-partitions |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
Migrating the initialisation system to be completely Python-like.
* New: In bootstrap/coriolisEnv.py, add the "etc" directory to the
PYTHONPATH as initialization are now Python modules.
* New: In Hurricane/analogic, first groundwork for the integration of
PIP/MIM/MOM multi-capacitors. Add C++ and Python interface for the
allocation matrix and the list of capacities values.
* Change: In Hurricane::RegularLayer, add a layer parameter to the
constructor so the association between the RegularLayer and it's
BasicLayer can readily be done.
* Change: In Hurricane::Layer, add a new getCut() accessor to get the
cut layer in ViaLayer.
* Change: In Hurricane::DataBase::get(), the Python wrapper should no
longer consider an error if the data-base has not been created yet.
Just return None.
* Bug: In Isobar::PyLayer::getEnclosure() wrapper, if the overall
enclosure is requested, pass the right parameter to the C++ function.
* Change: In AllianceFramework, make public _bindLibraries() and export
it to the Python interface.
* Change: In AllianceFramework::create(), do not longer call bindLibraries().
This now must be done explicitely and afterwards.
* Change: In AllianceFramework::createLibrary() and
Environement::addSYSTEM_LIBRARY(), minor bug corrections that I don't
recall.
* Change: In SearchPath::prepend(), set the selected index to zero and
return it.
* Change: In CRL::System CTOR, add "etc" to the PYTHONPATH as the
configuration files are now organized as Python modules.
* New: In PyCRL, export the CRL::System singleton, it's creation is no
longer triggered by the one of AllianceFramework.
* New: In CRL/etc/, convert most of the configuration files into the
Python module format. For now, keep the old ".conf", but that are no
longer used.
For the real technologies, we cannot keep the directory name as
"180" or "45" as it not allowed by Python syntax, so we create "node180"
or "node45" instead.
Most of the helpers and coriolisInit.py are no longer used now.
To be removed in future commits after being sure that everything
works...
* Bug: In AutoSegment::makeDogleg(AutoContact*), the layer of the contacts
where badly computed when one end of the original segment was attached
to a non-preferred direction segment (mostly on terminal contacts).
Now use the new AutoContact::updateLayer() method.
* Bug: In Dijkstra::load(), limit symetric search area only if the net
is a symmetric one !
* Change: In Katana/python/katanaInit.py, comply with the new initialisation
scheme.
* Change: In Unicorn/cgt.py, comply to the new inititalization scheme.
* Change: In cumulus various Python scripts remove the call to
helpers.staticInitialization() as they are not needed now (we run in
only *one* interpreter, so we correctly share all init).
In plugins/__init__.py, read the new NDA directory variable.
* Bug: In cumulus/plugins/Chip.doCoronafloorplan(), self.railsNb was not
correctly managed when there was no clock.
* Change: In cumulus/plugins/Configuration.coronaContactArray(), compute
the viaPitch from the technology instead of the hard-coded 4.0 lambdas.
In Configuration.loadConfiguration(), read the "ioring.py" from
the new user's settings module.
* Bug: In stratus.dpgen_ADSB2F, gives coordinates translated into DbU to
the XY functions.
In st_model.Save(), use the VstUseConcat flag to get correct VST files.
In st_net.hur_net(), when a net is POWER/GROUND or CLOCK also make it
global.
* Change: In Oroshi/python/WIP_Transistor.py, encapsulate the generator
inside a try/except block to get prettier error (and stop at the first).
2019-10-28 12:09:14 -05:00
|
|
|
|``clockTree.buffer`` | ``String`` | :cb:`buf_x2` |
|
2017-07-15 10:35:02 -05:00
|
|
|
| +------------------+----------------------------+
|
|
|
|
| | The buffer model to use to drive sub-nets |
|
|
|
|
+-----------------------------------+------------------+----------------------------+
|
|
|
|
|
|
|
|
|
|
|
|
Recursive-Save (RSave)
|
|
|
|
----------------------
|
|
|
|
|
2019-05-24 06:24:11 -05:00
|
|
|
Performs a recursive top down save of all the models from the top cell
|
|
|
|
loaded in |cgt|. Forces a write of any non-terminal model. This plugin is used
|
2017-07-15 10:35:02 -05:00
|
|
|
by the clock tree plugin after the netlist clock sub-nets creation.
|
|
|
|
|
|
|
|
|
|
|
|
A Simple Example: AM2901
|
|
|
|
~~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
|
|
|
|
To illustrate the capabilities of |Coriolis| tools and |Python| scripting, a small
|
|
|
|
example, derived from the |Alliance| :cb:`AM2901` is supplied.
|
|
|
|
|
|
|
|
This example contains only the synthetized netlists and the :cb:`doChip.py` script
|
|
|
|
which perform the whole P&R of the design.
|
|
|
|
|
2019-06-04 07:32:23 -05:00
|
|
|
You can generate the chip using one of the following methods:
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
#. **Command line mode:** directly run the script: ::
|
|
|
|
|
2017-10-30 09:33:37 -05:00
|
|
|
dummy@lepka:AM2901> ./doChip -V --cell=amd2901
|
2017-07-15 10:35:02 -05:00
|
|
|
|
|
|
|
#. **Graphic mode:** launch |cgt|, load chip netlist ``amd2901`` (the top cell)
|
|
|
|
then run the |Python| script :cb:`doChip.py`.
|
|
|
|
|
|
|
|
.. note::
|
2019-06-04 07:32:23 -05:00
|
|
|
Between two consecutive run, be sure to erase the netlist/layout generated: ::
|
2017-07-15 10:35:02 -05:00
|
|
|
|
2019-05-27 11:49:51 -05:00
|
|
|
dummy@lepka:AM2901> rm *_cts*.vst *.ap
|