Partially reverts commit 9c5bffcf93.
The reasoning behind this is that setup.py is intended to strictly consume the Makefile and not be consumed by it. The attempt at using them recursively has caused a number of issues and has rendered Pyosys unusable to some users: See https://github.com/YosysHQ/yosys/issues/5012
Additionally, unlike the previous pyosys installation target, the wheel installation does not respect PREFIX=, only venvs.
For installation inside a venv, the intended method should remain a user manually executing `pip3 install .` instead of relying on the Makefile.
This adds optional in-memory caching of parsed liberty files to speed up
flows that repeatedly parse the same liberty files. To avoid increasing
the memory overhead by default, the caching is disabled by default. The
caching can be controlled globally or on a per path basis using the new
`libcache` command, which also allows purging cached data.
Although the two passes started out very similar, they diverged over
time and now have little in common. Moreover, `techmap` is extremely
complex while `flatten` does not have to be, and this complexity
interferes with improving `flatten`.
This includes the following significant changes:
* Patching ezsat and minisat to disable resource limiting code
on WASM/WASI, since the POSIX functions they use are unavailable.
* Adding a new definition, YOSYS_DISABLE_SPAWN, present if platform
does not support spawning subprocesses (i.e. Emscripten or WASI).
This definition hides the definition of `run_command()`.
* Adding a new Makefile flag, DISABLE_SPAWN, present in the same
condition. This flag disables all passes that require spawning
subprocesses for their function.
This pass is a proper subset of opt_rmdff, which is called by opt, which
is called by every synth flow in the coarse part. Thus, it never
actually does anything and can be safely removed.
A few new attributes are defined for use in cell libraries:
- iopad_external_pin: marks PAD cell's external-facing pin. Pad
insertion will be skipped for ports that are already connected
to such a pin.
- clkbuf_sink: marks an input pin as a clock pin, requesting clock
buffer insertion.
- clkbuf_driver: marks an output pin as a clock buffer output pin.
Clock buffer insertion will be skipped for nets that are already
driven by such a pin.
All three are module attributes that should be set to a comma-separeted
list of pin names.
Clock buffer insertion itself works as follows:
1. All cell ports, starting from bottom up, can be marked as clock sinks
(requesting clock buffer insertion) or as clock buffer outputs.
2. If a wire in a given module is driven by a cell port that is a clock
buffer output, it is in turn also considered a clock buffer output.
3. If an input port in a non-top module is connected to a clock sink in a
contained cell, it is also in turn considered a clock sink.
4. If a wire in a module is driven by a non-clock-buffer cell, and is
also connected to a clock sink port in a contained cell, a clock
buffer is inserted in this module.
5. For the top module, a clock buffer is also inserted on input ports
connected to clock sinks, optionally with a special kind of input
PAD (such as IBUFG for Xilinx).
6. Clock buffer insertion on a given wire is skipped if the clkbuf_inhibit
attribute is set on it.
Clifford has commented on PR #387 stating that he does not like the
driver script and would prefer to only have the core script with
appropriate notes in the documentation.
Also rename to .cc (rather than .cpp) for consistency.
