Commit Graph

1395 Commits

Author SHA1 Message Date
Miodrag Milanović c6caadfed4
Merge pull request #3902 from YosysHQ/krys/yw_join
yosys-witness concat yw trace files
2023-08-25 15:21:44 +02:00
Asherah Connor 4a475fa7a2 cxxrtl: include iostream when prints are used 2023-08-17 07:08:22 +02:00
Charlotte 2829cd9caa cxxrtl_backend: move sync $print grouping out of dump into analyze 2023-08-11 04:46:52 +02:00
Charlotte ce245b5105 cxxrtl_backend: respect sync `$print` priority
We add a new flow graph node type, PRINT_SYNC, as they don't get handled
with regular CELL_EVALs.  We could probably move this grouping out of
the dump method.
2023-08-11 04:46:52 +02:00
Charlotte 04582f2fb7 verilog_backend: emit sync `$print` cells with same triggers together
Sort by PRIORITY, ensuring output order.
2023-08-11 04:46:52 +02:00
Charlotte 4ffdee65e0 cxxrtl: store comb $print cell last EN/ARGS in module
statics were obviously wrong -- may be multiple instantiations of any
given module.  Extend test to cover this.
2023-08-11 04:46:52 +02:00
Charlotte 843ad9331b cxxrtl: WIP: adjust comb display cells to only fire on change
Naming and use of statics to be possibly revised.
2023-08-11 04:46:52 +02:00
Charlotte 7f7c61c9f0 fmt: remove lzero by lowering during Verilog parse
See https://github.com/YosysHQ/yosys/pull/3721#issuecomment-1502037466
-- this reduces logic within the cell, and makes the rules that apply
much more clear.
2023-08-11 04:46:52 +02:00
Charlotte fc0acd0ad1 cxxrtl: restrict -print-output to cout, cerr 2023-08-11 04:46:52 +02:00
Charlotte f9b149fa7b cxxrtl: add "-print-output" option, test in fmt 2023-08-11 04:46:52 +02:00
Charlotte bfa8b631bf cxxrtl: remove unused signedDivideWithRemainder 2023-08-11 04:46:52 +02:00
Charlotte 3571bf2c2d fmt: fuzz, remove some unnecessary busywork
Removing some signed checks and logic where we've already guaranteed the
values to be positive.  Indeed, in these cases, if a negative value got
through (per my realisation in the signed fuzz harness), it would cause
an infinite loop due to flooring division.
2023-08-11 04:46:52 +02:00
Charlotte 2ae551c0af fmt: fuzz, fix (remove extraneous + incorrect fill)
"blk + chunks" is often an overrun, plus the fill is unnecessary; we
throw blk away immediately.
2023-08-11 04:46:52 +02:00
Charlotte c382d7d3ac fmt: %t/$time support 2023-08-11 04:46:52 +02:00
Charlotte 52dc397a50 cxxrtl: don't use signed divide with unsigned/pos values
Incorrect for unsigned, wasted effort for positive signed.
2023-08-11 04:46:52 +02:00
Charlotte 095b093f4a cxxrtl: first pass of $print impl 2023-08-11 04:46:52 +02:00
whitequark 3f8eab15bb write_verilog: translate $print cells to $write tasks in always blocks. 2023-08-11 04:46:52 +02:00
Jannis Harder 77c7355d53 smtbmc: Avoid quadratic behavior when scanning s-exprs
The previous implementation for finding the end of a top-level s-expr
exhibited quadratic behavior as it would re-scan the complete input for
the current expression for every new line. For large designs with
trivial properties this could easily take seconds and dominate the
runtime over the actual solving.

This change remembers the current nesting level between lines, avoiding
the re-scanning.
2023-08-01 17:19:29 +02:00
N. Engelhardt 43780c9812
Merge pull request #3838 from povik/various-cleanup 2023-07-24 16:24:23 +02:00
Martin Povišer 51ef942547 verilog_backend: Use hashlib dict for `auto_name_map`
This is most likely faster.
2023-07-20 21:00:33 +01:00
Martin Povišer 596743a6b6 verilog_backend: Make the keywords pool static
Do not recreate the keywords pool on every lookup of an identifier.
2023-07-20 21:00:33 +01:00
Martin Povišer 78d13d1956 Mention 'bwmuxmap' in 'write_firrtl' help
The FIRRTL backend does call into the 'bwmuxmap' pass but omits it in
the help message.
2023-07-10 12:45:03 +02:00
Martin Povišer c0b1a7daa4 Drop stray 'cellaigs.h' include from backend passes
This include seems to have been copied over from the JSON backend where
AIG models are sometimes inserted into the JSON output, but these other
backends don't do anything with AIG.
2023-07-10 12:45:03 +02:00
Martin Povišer 06256c0c00 Slightly adjust the wording of "write_blif" help 2023-07-10 12:41:43 +02:00
Charlotte eb397592f0 cxxrtl: add `$divfloor`. 2023-06-28 15:27:06 +01:00
Jannis Harder 596da3f2a6
Merge pull request #3815 from charlottia/py312-syntax 2023-06-26 16:36:58 +02:00
Jannis Harder f9744fdfcd smtbmc: Make cover mode respect --keep-going
As cover mode by default stops looking for further traces when an
assertion fails, it should respect --keep-going.
2023-06-23 10:27:38 +02:00
Charlotte 3f29bdbbc5 smt2: py3.12+: avoid SyntaxWarning.
Python 3.12 emits a SyntaxWarning when encountering invalid escape
sequences.  They still parse as expected.  Marking these raw produces
the same result without the warnings.
2023-06-23 14:40:02 +10:00
Clifford Wolf cff3195caa Improve EDIF lib_cell_ports scan 2023-06-20 10:42:05 +02:00
Clifford Wolf fb9e12761b Add "write_edif -lsbidx" 2023-06-20 10:40:15 +02:00
Charlotte Connor c9d31c3c87 smt2: abits needs to be at least 1 for BitVec
BitVecs need a minimum length of 1; we zero-fill any extra bits in the
extend_u0() calls which works perfectly.
2023-06-13 15:01:45 +10:00
Jannis Harder 06f06c7be2
Merge pull request #3801 from jix/witness-aiw2yw-xbits 2023-06-12 16:12:39 +02:00
Jannis Harder dcc4d6e90b yosys-witness: Don't treat aiw x-bits as don't change
While treating initialization only bits as don't change during later
cycles is correct, actual x-bits should be kept as x-bits.
2023-06-09 15:21:22 +02:00
N. Engelhardt 6f5d984bdb
Merge pull request #3778 from jix/yw_clk2fflogic 2023-06-05 16:15:04 +02:00
Claire Xenia Wolf d7f25165a5 Add ninitff line to aiger .aim files
Signed-off-by: Claire Xenia Wolf <claire@clairexen.net>
2023-06-03 14:38:22 +02:00
Jannis Harder e36c71b5b7 Use clk2fflogic attr on cells to track original FF names in witnesses
This makes clk2fflogic add an attr to $ff cells that carry the state of
the emulated async FF. The $ff output doesn't have any async updates
that happened in the current cycle, but the $ff input does, so the $ff
input corresponds to the async FF's output in the original design.

Hence this patch also makes the following changes to passes besides
clk2fflogic (but only for FFs with the clk2fflogic attr set):

  * opt_clean treats the input as a register name (instead of the
    output)

  * rename -witness ensures that the input has a public name

  * the formal backends (smt2, btor, aiger) will use the input's
    name for the initial state of the FF in witness files

  * when sim reads a yw witness that assigns an initial value to the
    input signal, the state update is redirected to the output

This ensures that yosys witness files for clk2fflogic designs have
useful and stable public signal names. It also makes it possible to
simulate a clk2fflogic witness on the original design (with some
limitations when the original design is already using $ff cells).

It might seem like setting the output of a clk2fflogic FF to update the
input's initial value might not work in general, but it works fine for
these reasons:

  * Witnesses for FFs are only present in the initial cycle, so we do
    not care about any later cycles.

  * The logic that clk2fflogic generates loops the output of the
    genreated FF back to the input, with muxes in between to apply any
    edge or level sensitive updates. So when there are no active updates
    in the current gclk cycle, there is a combinational path from the
    output back to the input.

  * The logic clk2fflogic generates makes sure that an edge sensitive
    update cannot be active in the first cycle (i.e. the past initial
    value is assumed to be whatever it needs to be to avoid an edge).

  * When a level sensitive update is active in the first gclk cycle, it
    is actively driving the output for the whole gclk cycle, so ignoring
    any witness initialization is the correct behavior.
2023-05-25 12:48:02 +02:00
Jannis Harder 4f3d1be96a
Merge pull request #3767 from YosysHQ/krys/yw_fix
Assign wires an smtoffset
2023-05-22 16:10:55 +02:00
Jannis Harder 5fb1223861
Merge pull request #3733 from AdamHillier/aiger-inputs
Add outputs before inputs to the sigmap in the AIGER backend.
2023-05-22 16:09:15 +02:00
Krystine Sherwin 18b44a1e84
yosys-witness: add append option to yw2yw
Can now append a user defined number of steps to input traces when joining.
If the number of steps is +ve, inputs are all set to 0.
If -ve then steps are skipped.
If all of steps are skipped (including init step) then the input trace will not be copied.
If more than one input trace is provided, the append option will need to be provided the same number of times as there are input traces.
2023-05-22 11:44:19 +12:00
Krystine Sherwin 8c9a0b51d4
yosys-witness concat yw trace files
Intended for use with SCY to combine sequential cover traces.
Arbitrary number of inputs, will load all and attempt to join them.  Each trace will be replayed one after the other, in the same order as the files are provided.
Mismatch in init_only fields seems to work fine, with values in subsequent traces being assigned in the initial only if they weren't previously defined.
Uncertain if a mismatch in non init_only fields will cause problems.
Fixes WitnessSig.__eq__().
Adds helper functions to WitnessValues and ReadWitness classes.
2023-05-22 10:13:17 +12:00
Jannis Harder e6f3914800 smt2: Use smt bv offset for `$any*`'s smtoffset
While not setting the smtoffset here was clearly a bug, I think using
`chunk.offset` only worked incidentally. The `smtoffset` is an offset
into the `smtname, smtid` pair (here `"", idcounter`) which corresponds
to the smt bitvector `stringf("%s#%d", get_id(module), idcounter)` which
contains all the chunks this loop is iterating over.

Thus using an incrementing `smtoffset` (like the `$ff`/`$dff` case above
already does) should be the correct fix.
2023-05-18 11:58:09 +02:00
Krystine Sherwin 52ad7a47f3
Assign wires an smtoffset
Wires weren't being assigned an smtoffset value so when generating a yosys witness trace it would also use an offset of 0.
Not sure if this has any other effects, but it fixes the bug I was having.
@jix could you take a look at this?
2023-05-18 10:37:55 +12:00
Benjamin Barzen 8611429237
ABC9: Cell Port Bug Patch (#3670)
* ABC9: RAMB36E1 Bug Patch

* Add simplified testcase

* Also fix xaiger writer for under-width output ports

* Remove old testcase

* Missing top-level input port

* Fix tabs

---------

Co-authored-by: Eddie Hung <eddie@fpgeh.com>
2023-04-22 16:24:36 -07:00
AdamHillier 3861cc31f0 Add outputs before inputs to the sigmap in the AIGER backend. 2023-04-19 11:00:51 +00:00
Catherine 4bb173e256 yosys-smtbmc: support -h/--help (and exit with code 0). 2023-02-27 20:31:00 +00:00
Jannis Harder 160eeab2bb verilog_backend: Do not run bwmuxmap even if in expr mode
While bwmuxmap generates equivalent logic, it doesn't propagate x bits
in the same way, which can be relevant when writing verilog.
2023-02-13 14:00:38 +01:00
Aki Van Ness a90f940615 backends/firrtl: Ensure `modInstance` is valid
This should fix #3648 where when calling `emit_elaborated_extmodules` it
checks to see if a module is a black-box, however there was no
validation that the cell type was actually known, and it just always
assumed that we would get a valid instance, causing a segfault.
2023-02-03 08:27:52 -05:00
Jannis Harder 0f2cb80a26
Merge pull request #3655 from jix/smt2_fix_b_op_width
smt2: Fix operation width computation for boolean producing cells
2023-02-01 18:06:59 +01:00
Jannis Harder 5e82638408 smt2: Fix operation width computation for boolean producing cells
The output width for the boolean value should not influence the
operation width. The previous incorrect width extension would still
produce correct results, but could produce invalid smt2 output for
reduction operators when the output width was larger than the width of
the vector to which the reduction was applied.

This fixes #3654
2023-02-01 12:34:35 +01:00
Jannis Harder b08a880704 backends/rtlil: Do not shorten a value with z bits to 'x 2023-01-29 14:02:25 +01:00