Commit Graph

3971 Commits

Author SHA1 Message Date
Martin Povišer 7179e4f4b8 booth: Improve user interface 2023-09-25 14:50:41 +02:00
Martin Povišer cde2a0b926 booth: Make more use of appropriate helpers
Use the `addFa` helper, do not misuse `new_id` and make other changes
to the transformation code.
2023-09-25 14:50:41 +02:00
Martin Povišer 62302f601d booth: Remove more of unused helpers 2023-09-25 14:50:41 +02:00
Martin Povišer 30f8387b75 booth: Rewrite the main cell selection loop 2023-09-25 14:50:41 +02:00
Martin Povišer 986507f95f booth: Streamline the low-level circuit emission
For the basic single-bit operations, opt for gate cells (`$_AND_` etc.)
instead of the coarse cells (`$and` etc.). For the emission of cells
move to the conventional module methods (`module->addAndGate`) away
from the local helpers. While at it, touch on the surrounding code.
2023-09-25 14:50:41 +02:00
Martin Povišer cb05262fc4 booth: Remove now-unused helpers 2023-09-25 14:50:41 +02:00
Martin Povišer fedd12261f booth: Move away from explicit `Wire` pointers
To represent intermediate signals use the `SigBit`/`SigSpec` classes as
is customary in the Yosys codebase. Do not pass around `Wire` pointers
unless we have special reason to.
2023-09-25 14:50:41 +02:00
Rasmus Munk Larsen 9ed38bf9b6
Speed up the autoname pass by 3x. (#3945)
* Speed up the autoname pass by 2x. This is accomplished by only constructing IdString objects for plain strings that have a higher score.

* Defer creating IdStrings even further. This increases the speedup to 3x.
2023-09-21 09:46:49 +00:00
Rasmus Munk Larsen e0042bdff7 Speed up TopoSort. The main sorting algorithm implementation in TopoSort::sort_worker is 11-12x faster. Overall, the complete sequence of building the graph and sorting is about 2.5-3x faster. The overall impact in e.g. the replace_const_cells optimization pass is a ~25% speedup. End-to-end impact on our synthesis flow is about 3%. 2023-09-20 15:49:05 -07:00
Martin Povišer 54be4aca90
Merge pull request #3924 from andyfox-rushc/master
multpass -- create Booth Encoded multipliers for
2023-09-18 16:46:59 +02:00
Jannis Harder 62b4df4989 dft_tag: Implement `$overwrite_tag` and `$original_tag`
This does not correctly handle an `$overwrite_tag` on a module output,
but since we currently require the user to flatten the design for
cross-module dft, this cannot be observed from within the design, only
by manually inspecting the signals in the design.
2023-09-13 11:32:36 +02:00
Jannis Harder 46a35da28c Add `future` pass to resolve `$future_ff` cells 2023-09-13 11:32:36 +02:00
Jannis Harder 7a0c37b62d Initial dft_tag implementation
This is still missing a mode to rewrite $overwrite_tag and $original_tag
by injecting $set_tag and $get_tag in the right places. It's also
missing bit-precise propagation models for shifts and arithmetic and
requires the design to be flattened.
2023-09-13 11:32:36 +02:00
Miodrag Milanović 88ce47e4f0
Merge pull request #3892 from QuantamHD/dont_use
abc: Exposes dont_use flag in ABC
2023-09-12 14:58:44 +02:00
andyfox-rushc e4fe522767 MultPassWorker -> BoothPassWorker 2023-09-11 13:00:11 -07:00
andyfox-rushc eccc0ae6db Based passes/techmap/Makefile.inc changes on latest in yosys 2023-09-11 12:14:12 -07:00
andyfox-rushc a2c8e47295 multpass.cc -> booth.cc, added author/support contact info 2023-09-11 11:39:13 -07:00
Martin Povišer 5bef9b4e75
Merge pull request #3915 from povik/sim-print
sim: Add print support
2023-09-11 17:03:59 +02:00
andyfox-rushc 1b5287af59 cpa_carry array added to heap 2023-09-10 14:20:30 -07:00
andyfox-rushc 8d4b6c2f69 Switched arrays for signed multiplier construction to heap 2023-09-10 13:31:47 -07:00
andyfox-rushc d77fb81507 2d array -> 1d array in module generator 2023-09-10 12:45:36 -07:00
andyfox-rushc 6d29dc659b renamed passname to booth, replaced connect_sigSpecToWire with connect, updated test script 2023-09-08 15:34:56 -07:00
andyfox-rushc 411acc4a0a Fixed edge size cases for signed/unsigned booth generator 2023-09-08 13:41:31 -07:00
andyfox-rushc fedefa26bc multpass -- create Booth Encoded multipliers for 2023-09-06 16:35:17 -07:00
Jannis Harder e187fc915e xprop: Fix polarity errors and generate hdlnames
* Fixes a non-deterministic polarity error for $eqx/$nex cells
* Fixes a deterministic polarity error for $_NOR_ and $_ORNOT_ cells
* Generates hdlnames when xprop is run after flatten
2023-09-06 19:25:47 +02:00
Martin Povišer d4d951657f sim: Add `-assert` option to fail on failed assertions 2023-09-05 10:46:04 +02:00
Martin Povišer e995dddeaa abc: Warn about replacing undef bits 2023-09-05 10:45:30 +02:00
Martin Povišer c6566b660f memlib.md: Fix typo 2023-09-04 17:38:35 +02:00
Martin Povišer 3de84b959f memory_libmap: Tweak whitespace 2023-09-04 17:38:35 +02:00
Martin Povišer 50d117956c sim: Add print support 2023-09-04 17:12:38 +02:00
Miodrag Milanovic a8809989c4 ecp5_gsr -> lattice_gsr, change opt_lut_ins to accept lattice as tech 2023-08-22 10:50:11 +02:00
Ethan Mahintorabi d525a41497
abc: Exposes dont_use flag in ABC
ABC's read_lib command has a dont_use
cell list that is configurable by the user.

This PR exposes that option to Yosys.

See
5405d4787a/src/map/scl/scl.c (L285)
for documentation on this option.

Signed-off-by: Ethan Mahintorabi <ethanmoon@google.com>
2023-08-15 20:03:46 +00:00
Charlotte 860e3e4056 proc_clean: only consider fully-defined switch operands too. 2023-08-12 02:46:31 +02:00
Charlotte bf84861fc2 proc_clean: only consider fully-defined case operands. 2023-08-12 02:46:31 +02:00
whitequark d51ecde8c2 clean: keep $print cells, since they have unmodelled side effects. 2023-08-11 04:46:52 +02:00
Martin Povišer ff3c7873f5 wreduce: Group reconnections
Group the reconnections, so that instead of producing

  connect $auto$wreduce.cc:455:run$24 [0] 1'0
  connect $auto$wreduce.cc:455:run$23 [31] 1'0
  connect $auto$wreduce.cc:455:run$23 [30] 1'0
  ... (40 more lines)

we produce

  connect $auto$wreduce.cc:461:run$23 [31:11] 21'000000000000000000000
  connect $auto$wreduce.cc:461:run$24 [31:10] 22'0000000000000000000000

.
2023-08-04 14:43:59 +01:00
Martin Povišer f8325f66b7 opt_expr: Fix 'signed X>=0' replacement for wide output ports
If the `$ge` cell we are replacing has wide output port, the upper bits
on the port should be driven to zero. That's not what a `$not` cell with
a single-bit input does. Instead opt for a `$logic_not` cell, which does
zero-pad its output.

Fixes #3867.
2023-08-01 13:50:12 +01:00
N. Engelhardt 43780c9812
Merge pull request #3838 from povik/various-cleanup 2023-07-24 16:24:23 +02:00
Catherine 6965abeefa abc, abc9_exe: fix build on WASI (and others with `const* stdout`).
C does not guarantee that stdout/stderr can be reassigned.
Most platforms do make them assignable, however musl and WASI that
is based on musl do not. WASI does not have `dup2()`; instead it has
its own non-portable version of it that can only assign to previously
allocated fds.

Update the stream redirection code so that it does the right thing
on WASI and other platforms.
2023-07-23 05:13:29 +01:00
Catherine 411b6e98cd abc, abc9_exe: respect `-q` when built with linked ABC.
This is mostly important for YoWASP builds, since those do not have
a way to build with external ABC (I prototyped it but for some reason
ABC always segfaults when built as an independent Wasm binary...)
2023-07-23 02:03:29 +01:00
Martin Povišer f5485b59a9 sim: Bail if there are blackboxes in simulation 2023-07-20 21:01:03 +01:00
Martin Povišer f0ae046c5a opt_share: Fix input confusion with ANDNOT, ORNOT gates
Distinguish between the A, B input ports of `$_ANDNOT_`, `$_ORNOT_`
gates when considering those for sharing. Unlike the input ports of the
other supported single-bit gates, those are not interchangeable.

Fixes #3848.
2023-07-20 20:58:52 +01:00
N. Engelhardt 2be5c0786f
Merge pull request #3826 from nakengelhardt/nak/mem_libmap_print_attr 2023-07-17 16:35:10 +02:00
Martin Povišer eb083c5d4b extract_counter: Update help and comments after UP/DOWN support
Commit fec7dc5c should have added support for up counters so update
the help and comments accordingly.
2023-07-10 12:45:03 +02:00
N. Engelhardt 57de249881 memory_libmap: print additional debug messages when no valid mapping is found 2023-07-06 18:54:32 +02:00
N. Engelhardt 14d50a176d
Merge pull request #3676 from nakengelhardt/dfflegalize_scratchpad_minarg 2023-07-03 17:15:21 +02:00
N. Engelhardt a6be7b4751 memory_libmap: add debug messages for some reasons for rejecting mappings 2023-06-29 14:08:31 +02:00
N. Engelhardt 7542146fc5 memory_libmap: print message about attributes forcing ram kind 2023-06-28 17:48:20 +02:00
Jannis Harder a07f8ac38a check: Also check for conflicts with constant drivers 2023-06-23 18:07:28 +02:00
Claire Xen 51e627686a
Merge pull request #3812 from charlottia/iterator-invalidation
proc_prune: avoid using invalidated iterator
2023-06-21 14:46:25 +02:00
Charlotte 63e4114233 proc_prune: avoid using invalidated iterator
An `std::vector<T>::reverse_iterator` stores the
`std::vector<T>::iterator` which points to the (forwards-ordered)
*following* item.  Thus while `vec.rbegin()` dereferences to the final
item of `vec`, the iterator it wraps (`vec.rbegin().base()`) is equal to
`vec.end()`.

In the remove case here, we advance `it` (backwards), erasing the item
we just advanced past by grabbing its (pre-increment) base
forward-iterator and subtracting 1.

The iterator maths here is obviously all OK, but the forward-iterator
that `it` wraps post-increment actually points to the item we just
removed.  That iterator was invalidated by the `erase()` call.

That this works anyway is (AFAICT) some combination of luck and/or
promises that aren't part of the C++ spec, but MSVC's debug iterator
support picks this up.

`erase()` returns the new iterator that follows the item just erased,
which happens to be the exact one we want our reverse-iterator to wrap
for the next loop; we get a fresh iterator to the same base, now without
the preceding item.
2023-06-21 19:53:08 +10:00
N. Engelhardt 9c7f0e7670 show: truncate very long module names 2023-06-20 12:53:56 +02:00
N. Engelhardt 22c9237716 show: escape angle brackets 2023-06-20 11:17:12 +02:00
N. Engelhardt 7c606bd5a3
Merge pull request #3791 from nakengelhardt/nak/show_attr_wires 2023-06-05 16:18:54 +02:00
N. Engelhardt 6f5d984bdb
Merge pull request #3778 from jix/yw_clk2fflogic 2023-06-05 16:15:04 +02:00
N. Engelhardt 0707b911c7 show: add -viewer none option 2023-06-01 11:00:07 +02:00
N. Engelhardt 4b986c9c65 fix wire color after BUF 2023-05-31 17:38:46 +02:00
N. Engelhardt 26555a998d show -colorattr: extend colors to arrows when wires have attribute 2023-05-26 17:18:21 +02:00
Eddie Hung ec8d7b1c68 abc9_ops -prep_hier to unmap entire module 2023-05-25 18:42:08 +01:00
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 7caeb922a0 sim: Run level triggered async updates to fixpoint during initialization 2023-05-25 12:46:16 +02:00
gatecat 52c8c28d2c Add recover_names pass to recover names post-mapping 2023-05-25 10:55:07 +02:00
Jannis Harder ad2b04d63a sim: Fix cosimulation with nested modules having unconnected inputs
When assigning values to input ports of nested modules in cosimulation,
sim needs to find the actual driver of the signal to perform the
assignment. The existing code didn't handle unconnected inputs in that
scenario.
2023-05-18 16:50:11 +02:00
Muthiah Annamalai (முத்து அண்ணாமலை) 693c609eec
Merge branch 'YosysHQ:master' into main/issue2525 2023-05-16 21:21:32 -07:00
Muthu Annamalai 665e0f6131 remove new line per maintainer request 2023-05-17 04:20:13 +00:00
Miodrag Milanović acfdc5cc42
Merge pull request #3755 from RTLWorks/muthu/issue3498
[YOSYS] Issue #3498 - Fix Synopsys style unquoted Liberty style
2023-05-15 16:34:35 +02:00
Muthiah Annamalai (முத்து அண்ணாமலை) c855502bd5
Update passes/techmap/libparse.cc
Allow Liberty canonical identifier including double quotes in if-body and pass-through for Synopsys-style unquoted identifiers issue#3498

Co-authored-by: Aki <201479+lethalbit@users.noreply.github.com>
2023-05-09 06:40:21 -07:00
Miodrag Milanović 226a224640
Merge pull request #3749 from lethalbit/aki/plugin-stuff
Updated the `plugin` command to better handle paths
2023-05-09 08:46:02 +02:00
N. Engelhardt 266036c6f9
Merge pull request #3756 from YosysHQ/krys/sim_writeback 2023-05-08 16:21:24 +02:00
N. Engelhardt ec56e625f4
Merge pull request #3742 from jix/fix_rename_witness_cell_renames 2023-05-08 16:13:28 +02:00
Krystine Sherwin 5a4e72f57a
Fix sim writeback check for yw_cosim
Writeback of simulation state into initial state was only working for `run()` and `run_cosim_fst()`.
This change moves the writeback into the `write_output_files()` function so that all simulation modes work with the writeback option.
2023-05-08 13:13:09 +12:00
Muthu Annamalai 17cfc969dd [YOSYS] Issue #3498 - Fix Synopsys style unquoted Liberty style function body parsing with unittest 2023-05-06 23:37:47 -07:00
Miodrag Milanović 4251d37f4f
Merge pull request #3610 from YosysHQ/synthprop
Synthesizable properties
2023-05-05 11:03:09 +02:00
Muthu Annamalai 81e089cb60 [YOSYS-2525] fix read_liberty newline handling #2525
- newlines can be allowed in function parsing
2023-05-04 22:30:27 -07:00
Aki Van Ness bb240665b7
plugin: shuffled the `#ifdef WITH_PYTHON`'s around to un-tangle the code and pulled out the check for the `.py` extension so it will complain if you try to load a python extension without python support 2023-05-03 03:35:55 -04:00
Aki Van Ness 572c8df9a8
plugin: Re-vamped how plugin lookup was done to make it more consistent with the rest of yosys, and prevented a case where you could end up with `.so.so` on the end 2023-05-03 02:22:46 -04:00
Jannis Harder 7bff8b63b3 rename: Fix renaming cells in -witness mode
This was renaming cells while iterating over them which would always
cause an assertion failure. Apparently having to rename cells to make
all witness signals public is rarely required, so this slipped through.
2023-04-25 12:39:00 +02:00
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
N. Engelhardt 57897927ff stat: pass down quiet arg 2023-02-28 17:12:55 +01:00
Miodrag Milanović 21e87f7986
Merge pull request #3646 from YosysHQ/lofty/fix-3591
muxcover: do not add decode muxes with x inputs
2023-02-27 16:26:57 +01:00
Jannis Harder 1c667fab2b
Merge pull request #3672 from jix/yw-cosim-hierarchy-fixes
sim: For yw cosim, drive parent module's signals for input ports
2023-02-15 13:45:00 +01:00
N. Engelhardt b562b54c14 dfflegalize: allow setting mince and minsrst args from scratchpad 2023-02-15 12:53:46 +01:00
Jannis Harder ec94703619
Merge pull request #2995 from georgerennie/cover_precond
chformal: Add -coverenable option
2023-02-14 17:46:31 +01:00
Jannis Harder 85f611fb23
Merge pull request #3126 from georgerennie/equiv_make_assertions
equiv_make: Add -make_assert option
2023-02-14 17:15:55 +01:00
Jannis Harder b636af9751 chformal: Note about using -coverenable with the Verific frontend 2023-02-14 17:10:43 +01:00
Miodrag Milanovic 550a5b7b6b Update license 2023-02-13 17:23:26 +01:00
Miodrag Milanovic 713b7d3e26 added support for latched output reset 2023-02-13 17:23:26 +01:00
Miodrag Milanovic 131b557727 Initial implementation of synthesizable assertions 2023-02-13 17:23:26 +01:00
Jannis Harder 1698202ccc sim: For yw cosim, drive parent module's signals for input ports 2023-02-13 12:26:06 +01:00
Miodrag Milanovic 5f33c0e0b2 Updated changelog 2023-02-08 10:11:47 +01:00
N. Engelhardt 417fadbefd
Merge pull request #3625 from povik/show_cleanup 2023-02-06 16:11:26 +01:00
N. Engelhardt 419f91a2b9 add option to fsm_detect to ignore self-resetting 2023-01-30 16:12:53 +01:00
N. Engelhardt ecfa7e9fbc add pmux option to bmuxmap for better fsm detection with verific frontend 2023-01-30 16:12:53 +01:00
Lofty 822c7b0341 muxcover: do not add decode muxes with x inputs 2023-01-26 05:19:45 +00:00
Jannis Harder afac3f2c76 formalff: Fix crash with _NOT_ gates in -hierarchy mode 2023-01-25 12:55:29 +01:00
Miodrag Milanović 8180cc4325
Merge pull request #3624 from jix/sim_yw
Changes to support SBY trace generation with the sim command
2023-01-23 16:55:17 +01:00
Miodrag Milanović 245884a101
Merge pull request #3629 from YosysHQ/micko/clang_fixes
Fixes for some of clang scan-build detected issues
2023-01-23 16:24:22 +01:00
gatecat bfacaddca8 show: Remove left-in debug log_warning
Signed-off-by: gatecat <gatecat@ds0.me>
2023-01-23 13:54:07 +01:00
Claire Xenia Wolf bfc3c20cfb Improve splitcells pass
Signed-off-by: Claire Xenia Wolf <claire@clairexen.net>
2023-01-18 00:31:29 +01:00
Miodrag Milanovic 6574553189 Fixes for some of clang scan-build detected issues 2023-01-17 12:58:08 +01:00