This also aligns the functionality:
- in all cases, the onehot attribute is used to create appropriate
constraints (previously, opt_dff didn't do it at all, and share
created one-hot constraints based on $pmux presence alone, which
is unsound)
- in all cases, shift and mul/div/pow cells are now skipped when
importing the SAT problem (previously only memory_share did this)
— this avoids creating clauses for hard cells that are unlikely
to help with proving the UNSATness needed for optimization
This essentially adds wide port support for free in passes that don't
have a usefully better way of handling wide ports than just breaking
them up to narrow ports, avoiding "please run memory_narrow" annoyance.
When converting a sync transparent read port with const address to async
read port, nothing at all needs to be done other than clk_enable change,
and thus we have no FF cell to return. Handle this case correctly in
the helper and in its users.
When the register being merged into the EN signal happens to be a $sdff,
the current code creates a new $mux for every bit, even if they happen
to be identical (as is usually the case), preventing proper grouping
further down the flow. Fix this by adding a simple cache.
Fixes#2409.
The $div and $mod cells use truncating division semantics (rounding
towards 0), as defined by e.g. Verilog. Another rounding mode, flooring
(rounding towards negative infinity), can be used in e.g. VHDL. The
new $divfloor cell provides this flooring division.
This commit also fixes the handling of $div in opt_expr, which was
previously optimized as if it was $divfloor.
The $div and $mod cells use truncating division semantics (rounding
towards 0), as defined by e.g. Verilog. Another rounding mode, flooring
(rounding towards negative infinity), can be used in e.g. VHDL. The
new $modfloor cell provides this flooring modulo (also known as "remainder"
in several languages, but this name is ambiguous).
This commit also fixes the handling of $mod in opt_expr, which was
previously optimized as if it was $modfloor.
Before this commit, memory_map (which is always a part of a synth
script) would always pick up any $mem cell that was not processed
by a preceding pass and lower it down to $dff/$mux cells.
This is undesirable for two reasons:
* If there is an explicit inference attribute set on a $mem cell,
e.g. (* ram_block *), then it is arguably incorrect to map such
a memory to $dff/$mux cells.
* If memory_map tries to lower a memory that was intended to
be mapped to a large BRAM, it often takes extraordinarily long
time to finish, produces an extremely large log file, and outputs
an unusable design.
After this commit, properly invoked memory_map will not map any
memory that has an explicit inference attribute specified, solving
the first issue, and alleviating the second. The default behavior
is not changed.
Marcelina Kościelnicka
Claire Xenia Wolf
whitequark
Xiretza
David Shah
N. Engelhardt
Alberto Gonzalez
Eddie Hung