3
0
Fork 0
mirror of https://github.com/YosysHQ/yosys synced 2025-11-05 13:56:04 +00:00
Commit graph

1908 commits

Author SHA1 Message Date
Robert O'Callahan
45017e19ec Make SigSpec::unpack() non-const 2025-10-31 11:53:39 +00:00
Robert O'Callahan
9a2fd4c31b Switch the SigSpec packed representation to allow just a single chunk that's inline in the SigSpec.
Single-chunk SigSpecs are very common and this avoids a heap allocation. It also simplifies
some algorithms.
2025-10-31 11:53:39 +00:00
Robert O'Callahan
1af5d4f2db Use chunks iterator for SigSpec::extract() 2025-10-31 11:53:39 +00:00
Robert O'Callahan
ab525643a7 Don't reset the hash when unpacking, instead clear the hash whenever bits are modified 2025-10-31 11:53:39 +00:00
Robert O'Callahan
a1f7d6c9bf Use size() instead of direct access to width_ in many places 2025-10-31 11:53:39 +00:00
Robert O'Callahan
dbb8354996 Remove unnecessary pack() from SigSpec::extend_u0() 2025-10-31 11:53:39 +00:00
Robert O'Callahan
d314c47a55 Simplify SigSpec::as_bit() 2025-10-31 11:53:39 +00:00
Robert O'Callahan
be530bef73 Instead of using packing and hashing to compute SigSpec ordering and equality, just use the width and chunkwise comparisons
This avoids having to pack and compute hashes, and generally results in a
simpler ordering.
2025-10-31 11:53:39 +00:00
Robert O'Callahan
973e8a3928 Build a temporary SigChunk list in the iterator in the cases where that's needed 2025-10-31 11:53:39 +00:00
Robert O'Callahan
8c9dd3209a Make SigSpec conversion to vector of SigChunk use chunks iterator 2025-10-31 11:53:39 +00:00
Robert O'Callahan
4672127610 Make SigSpec::parse_rhs use is_chunk to avoid direct access to chunks_ 2025-10-31 11:53:39 +00:00
Robert O'Callahan
11a91af920 Make SigSpec::to_sigbit_pool use chunk iterator 2025-10-31 11:53:39 +00:00
Robert O'Callahan
8cb7cd7ac1 Make SigSpec::to_sigbit_set use chunk iterator 2025-10-31 11:53:39 +00:00
Robert O'Callahan
82f86164d3 Use SigSpec::try_as_const in some places 2025-10-31 11:53:39 +00:00
Robert O'Callahan
0d45d9cc6e Make SigSpec::is_one_hot use try_as_const 2025-10-31 11:53:39 +00:00
Robert O'Callahan
a0e9e2d364 Fix try_as_const/as_wire/as_chunk 2025-10-31 11:53:39 +00:00
Robert O'Callahan
000c081965 Add try_as_const and use the const iterator a bit more 2025-10-31 11:53:38 +00:00
Robert O'Callahan
04a6dbc562 Make SigSpec::has_marked_bits use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
7bd6b4f287 Make SigSpec::has_const(State) use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
213d665ae1 Make SigSpec::has_const use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
5c8f9f14ca Make SigSpec::is_fully_undef use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
b2de56cae2 Make SigSpec::is_fully_def use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
8a88acd9b8 Make SigSpec::is_fully_ones use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
58dbf75885 Make SigSpec::is_fully_zero use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
20e64ee17b Make is_fully_const use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
bf4cfbd72d Make SigSpec::is_wire/is_chunk/is_fully_const use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
fd7b4f4a8b Make SigSpec::updhash() use chunk iterator 2025-10-31 11:53:38 +00:00
Robert O'Callahan
37e4c2e8f8 Make SigSpec::chunks() return an object that can be iterated over without packing the SigSpec 2025-10-31 11:53:38 +00:00
Robert O'Callahan
c4f3e61339 Make Module stop accessing internals of SigSpec 2025-10-29 18:32:10 +00:00
Robert O'Callahan
a27b1a83ae Don't recompute hash using moved-out-of value 2025-10-28 07:41:10 +00:00
Emil J
9a12d92551
Merge pull request #5386 from YosysHQ/emil/liberty-glob-all
Expand wildcards in Liberty file consumers
2025-10-09 20:21:48 +02:00
Emil J
a80462f27f
Merge pull request #5339 from rocallahan/fast-rtlil-parser
Rewrite the RTLIL parser for efficiency
2025-10-08 14:52:37 +02:00
Miodrag Milanović
b7a72811cc
Merge pull request #5408 from rocallahan/atomic-mfp
Make `mfp` const methods thread-safe.
2025-10-08 13:08:16 +02:00
Robert O'Callahan
2f81c55389 Make mfp const methods thread-safe.
In particular, we make the parent links relaxed atomics so concurrent
`ifind()` calls are safe.

This may appear to cause a tiny performance regression but as discussed
in https://yosyshq.discourse.group/t/parallel-optmergepass-implementation/87/16
this is probably just noise.
2025-10-06 22:44:01 +00:00
Mohamed Gaber
80fcce64da
pyosys: fix ref-only classes, implicit conversions
+ cleanup
2025-10-03 11:54:44 +03:00
Mohamed Gaber
447a6cb3f0
misc: WITH_PYTHON -> YOSYS_ENABLE_PYTHON
For consistency.

Also trying a new thing: only rebuilding objects that use the pybind11 library. The idea is these are the only objects that include the Python/pybind headers and thus the only ones that depend on the Python ABI in any capacity, so other objects can be reused across wheel builds. This has the potential to cut down build times.
2025-10-03 11:54:44 +03:00
Mohamed Gaber
54799bb8be
pyosys: globals, set operators for opaque types
There is so much templating going on that compiling wrappers.cc now takes 1m1.668s on an Apple M4…
2025-10-03 11:54:44 +03:00
Mohamed Gaber
88be728353
pyosys: rewrite using pybind11
- Rewrite all Python features to use the pybind11 library instead of boost::python.
  Unlike boost::python, pybind11 is a header-only library that is just included by Pyosys code, saving a lot of compile time on wheels.
- Factor out as much "translation" code from the generator into proper C++ files
- Fix running the embedded interpreter not supporting "from pyosys import libyosys as ys" like wheels
- Move Python-related elements to `pyosys` directory at the root of the repo
- Slight shift in bridging semantics:
  - Containers are declared as "opaque types" and are passed by reference to Python - many methods have been implemented to make them feel right at home without the overhead/ambiguity of copying to Python and then copying back after mutation
  - Monitor/Pass use "trampoline" pattern to support virual methods overridable in Python: virtual methods no longer require `py_` prefix
- Create really short test set for pyosys that just exercises basic functionality
2025-10-03 11:54:44 +03:00
Robert O'Callahan
d45223976a Optimize IdString::empty()
I actually saw this take 0.6% of the time in an `opt_clean` pass (under Module::check()).
Trivial issue, but the fix is also trivial and simple.
2025-10-01 02:49:03 +00:00
Robert O'Callahan
db3d9a1baf When adding named elements to an RTLIL::Module, std::move the name to avoid refcount churn 2025-09-30 21:39:18 +00:00
Robert O'Callahan
86d8dd6224 Add a moving assignment operator for IdString to avoid refcount churn 2025-09-30 21:39:18 +00:00
Robert O'Callahan
39c6e06e3d Add an IdString(std::string_view) constructor for efficiency when we already know the string length 2025-09-30 21:39:18 +00:00
Robert O'Callahan
6b43fca8df Make the Const string constructor take the string by value and move it into the const 2025-09-30 21:39:18 +00:00
Emil J
7719beb4ae
Merge pull request #5349 from rocallahan/cleanup-hashops
Reduce hashops verbiage in `OptMergePass`
2025-09-30 19:34:44 +02:00
Jannis Harder
47639f8a98
Merge pull request #5388 from jix/bufnorm-followup
Refactor and fixes to incremental bufNormalize + related changes
2025-09-29 15:15:29 +02:00
Jannis Harder
86fb2f16f7 bufnorm: Refactor and fix incremental bufNormalize
This fixes some edge cases the previous version didn't handle properly
by simplifying the logic of determining directly driven wires and
representatives to use as buffer inputs.
2025-09-29 08:21:28 +02:00
Akash Levy
4b6b254e31 pyosys cannot parse header with omitted function args 2025-09-28 01:37:30 -07:00
Robert O'Callahan
1c73011e7e Swap SigSpecs using std::swap with moves 2025-09-25 03:04:17 +00:00
Jannis Harder
2dce50516b log: Print static message as fatal error for YOSYS_ABORT 2025-09-24 13:20:27 +02:00
Jannis Harder
86a46b9e5c log: Flush stdout before printing a fatal error to stderr
This hasn't been an issue when using -l to redirect or when stdout is
line buffered, explaining how we didn't notice this earlier, but for
`yosys ... > log` that extra flush is required to ensure all messages
preceding the fatal error are flushed.
2025-09-24 13:20:27 +02:00