- Standard data declarations can now use any integer type
- Parameters and localparams can now use any integer type
- Function returns types can now use any integer type
- Fix `parameter logic`, `localparam reg`, etc. to be 1 bit (previously 32 bits)
- Added longint type (64 bits)
- Unified parser source for integer type widths
- Simplify synthetic localparams for normal calls to update their width
- This step was inadvertently removed alongside `added_mod_children`
- Support redeclaration of constant function arguments
- `eval_const_function` never correctly handled this, but the issue
was not exposed in the existing tests until the recent change to
always attempt constant function evaluation when all-const args
are used
- Check asserts in const_arg_loop and const_func tests
- Add coverage for width mismatch error cases
This would previously complain about an undefined internal macro if the
unapplied macro had not already been used. If it had, it would
incorrectly use the arguments from the previous invocation.
This adds a mechanism for marking certain portions of elaboration as
occurring within unevaluated ternary branches. To enable elaboration of
the overall ternary, this also adds width detection for these
unelaborated function calls.
This is a somewhat obscure edge case I encountered while working on test
cases for earlier changes. Declarations in generate blocks should not be
checked against the list of ports. This change also adds a check
forbidding declarations within generate blocks being tagged as inputs or
outputs.
This fixes binding signed memory reads, signed unary expressions, and
signed complex SigSpecs to ports. This also sets `is_signed` for wires
generated from signed params when -pwires is used. Though not necessary
for any of the current usages, `is_signed` is now appropriately set when
the `extendWidth` helper is used.
Elaboration now attempts constant evaluation of any function call with
only constant arguments, regardless of the context or contents of the
function. This removes the concept of "recommended constant evaluation"
which previously applied to functions with `for` loops or which were
(sometimes erroneously) identified as recursive. Any function call in a
constant context (e.g., `localparam`) or which contains a constant-only
procedural construct (`while` or `repeat`) in its body will fail as
before if constant evaluation does not succeed.
This change set contains a number of bug fixes and improvements related to
scoping and resolution in generate and procedural blocks. While many of the
frontend changes are interdependent, it may be possible bring the techmap
changes in under a separate PR.
Declarations within unnamed generate blocks previously encountered issues
because the data declarations were left un-prefixed, breaking proper scoping.
The LRM outlines behavior for generating names for unnamed generate blocks. The
original goal was to add this implicit labelling, but doing so exposed a number
of issues downstream. Additional testing highlighted other closely related scope
resolution issues, which have been fixed. This change also adds support for
block item declarations within unnamed blocks in SystemVerilog mode.
1. Unlabled generate blocks are now implicitly named according to the LRM in
`label_genblks`, which is invoked at the beginning of module elaboration
2. The Verilog parser no longer wraps explicitly named generate blocks in a
synthetic unnamed generate block to avoid creating extra hierarchy levels
where they should not exist
3. The techmap phase now allows special control identifiers to be used outside
of the topmost scope, which is necessary because such wires and cells often
appear in unlabeled generate blocks, which now prefix the declarations within
4. Some techlibs required modifications because they relied on the previous
invalid scope resolution behavior
5. `expand_genblock` has been simplified, now only expanding the outermost
scope, completely deferring the inspection and elaboration of nested scopes;
names are now resolved by looking in the innermost scope and stepping outward
6. Loop variables now always become localparams during unrolling, allowing them
to be resolved and shadowed like any other identifier
7. Identifiers in synthetic function call scopes are now prefixed and resolved
in largely the same manner as other blocks
before: `$func$\func_01$tests/simple/scopes.blk.v:60$5$\blk\x`
after: `\func_01$func$tests/simple/scopes.v:60$5.blk.x`
8. Support identifiers referencing a local generate scope nested more
than 1 level deep, i.e. `B.C.x` while within generate scope `A`, or using a
prefix of a current or parent scope, i.e. `B.C.D.x` while in `A.B`, `A.B.C`,
or `A.B.C.D`
9. Variables can now be declared within unnamed blocks in SystemVerilog mode
Addresses the following issues: 656, 2423, 2493
The presence of IS_*_INVERTED on FD* cells follows Vivado, which
apparently has been decided by a dice roll. Just assume false if the
parameter doesn't exist.
Fixes#2559.
The changes in #2476 ensured that function inputs like `input x;`
retained their single-bit size when instantiated with a constant
argument and turned into a localparam. That change did not handle the
possibility for an input to be redeclared later on with an explicit
width, such as `integer x;`.
- Signed cell outputs are sign extended when bound to larger wires
- Signed connections are sign extended when bound to larger cell inputs
- Sign extension is performed in hierarchy and flatten phases
- genrtlil indirects signed constants through signed wires
- Other phases producing RTLIL may need to be updated to preserve
signedness information
- Resolves#1418
- Resolves#2265
This test pretty much passes by accident — the `prep` command runs
memory_collect without memory_dff first, which prevents merging read
register into the memory, and thus blocks block RAM inference for a
reason completely unrelated to the attribute.
The attribute setting didn't actually work because it was set on the
containing module instead of the actual memory.
When the register being merged into the EN signal happens to be a $sdff,
the current code creates a new $mux for every bit, even if they happen
to be identical (as is usually the case), preventing proper grouping
further down the flow. Fix this by adding a simple cache.
Fixes#2409.
Previously, `$memwr` and `$meminit` cells were always preserved (along
with the memory itself). With this change, they are instead part of the
main cell mark-and-sweep pass: a memory (and its `$meminit` and `$memwr`
cells) is only preserved iff any associated `$memrd` cell needs to be
preserved.
* xilinx: eliminate SCCs from DSP48E1 model
* xilinx: add SCC test for DSP48E1
* Update techlibs/xilinx/cells_sim.v
* xilinx: Gate DSP48E1 being a whitebox behind ALLOW_WHITEBOX_DSP48E1
Have a test that checks it works through ABC9 when enabled
Quartus assumes unsigned multiplication by default, breaking signed
multiplies, so add an input signedness parameter to the MISTRAL_MUL*
cells to propagate to Quartus' <family>_mac cells.
For connection `assign a = b;`, `sigmap(a)` returns `b`. This is
exactly the opposite of the desired canonicalization for driven bits.
Consider the following code:
module foo(inout a, b);
assign a = b;
endmodule
module bar(output c);
foo f(c, 1'b0);
endmodule
Before this commit, the inout ports would be swapped after flattening
(and cause a crash while attempting to drive a constant value).
This issue was introduced in 9f772eb9.
Fixes#2183.
Our techmap rules for $shift and $shiftx cells contained a special path
that aimed to decompose the shift LSB-first instead of MSB-first in
select cases that come up in pmux lowering. This path was needlessly
overcomplicated and contained bugs.
Instead of doing that, just switch over the main path to iterate
LSB-first (except for the specially-handled MSB for signed shifts
and overflow handling). This also makes the code consistent with
shl/shr/sshl/sshr cells, which are already decomposed LSB-first.
Fixes#2346.
- expand_genblock defers prefixing of items within named sub-blocks
- Allow partially-qualified references to local scopes
- Handle shadowing within generate blocks
- Resolve generate scope references within tasks and functions
- Apply generate scoping to genvars
- Resolves#2214, resolves#1456
The main part is converting ice40_dsp to recognize the new FF types
created in opt_dff instead of trying to recognize the mux patterns on
its own.
The fsm call has been moved upwards because the passes cannot deal with
$dffe/$sdff*, and other optimizations don't help it much anyway.
The main part is converting xilinx_dsp to recognize the new FF types
created in opt_dff instead of trying to recognize the patterns on its
own.
The fsm call has been moved upwards because the passes cannot deal with
$dffe/$sdff*, and other optimizations don't help it much anyway.
Before this fix, equiv_induct only assumed that one of the following is
true:
- defined value of A is equal to defined value of B
- A is undefined
This lets through valuations where A is defined, B is undefined, and
the defined (meaningless) value of B happens to match the defined value
of A. Instead, tighten this up to OR of the following:
- defined value of A is equal to defined value of B, and B is not
undefined
- A is undefined
This parameter will resolve to the name of the cell being mapped. The
first user of this parameter will be synth_intel_alm's Quartus output,
which requires a unique (and preferably descriptive) name passed as
a cell parameter for the memory cells.
Of standard yosys cells, xilinx_srl only works on $_DFF_?_ and
$_DFFE_?P_, which get upgraded to $_SDFFE_?P?P_ by dfflegalize at the
point where xilinx_srl is called for non-abc9. Fix this by running
ff_map.v first, resulting in FDRE cells, which are handled correctly.
This fixes some dfflegalize equivalence checks, and breaks others — and
I strongly suspect the others are due to bad support for multiple
async inputs in `proc` (in particular, lack of proper support for
dlatchsr and sketchy circuits on dffsr control inputs).
Those can be created by `opt_dff` when optimizing `$adff` with const
clock, or with D == Q. Make dfflegalize do the opposite transform
when such dlatches would be otherwise unimplementable.
whitequark
Zachary Snow
Michael Singer
TimRudy
Marcelina Kościelnicka
William D. Jones
Kamil Rakoczy
Claire Xen
David Shah
dalance
Xiretza
Miodrag Milanović
N. Engelhardt
Eddie Hung
Udi Finkelstein
Dan Ravensloft
Lukasz Dalek