Compare commits

...
188 Commits
Author SHA1 Message Date
Wilson Snyder 5b2fbf4f37 Version bump 2022-08-31 19:46:45 -04:00
Wilson Snyder 592dab2bdb Commentary: Changes update 2022-08-31 19:27:43 -04:00
Wilson Snyder 51daa64e9a Fix --hierarchical with order-based pin connections (#3585). 2022-08-31 18:12:21 -04:00
Aleksander Kiryk 2136afde6b Support negated properties (#3572) 2022-08-30 06:33:42 -04:00
Wilson Snyder ea55db7286 Internals: Cleanup some string constructors. No functional change. 2022-08-30 01:02:39 -04:00
Wilson Snyder 6a5f77b278 Internals: Cleanup some string/model constructors. No functional change. 2022-08-29 23:50:32 -04:00
Wilson Snyder 8658a0d7dc Internals: Constructor format update. No functional change. 2022-08-29 23:05:52 -04:00
Wilson Snyder c335aad25f Fix --hierarchical with order-based pin connections (#3583). 2022-08-29 22:49:19 -04:00
Wilson Snyder 9d9d647c1f Fix indentation of --protect import function SV code. 2022-08-29 22:28:02 -04:00
Wilson Snyder d47a37fb76 Internals: Cleanup constructors etc. No functional change. 2022-08-29 22:17:27 -04:00
Arkadiusz Kozdra 0a3a15a66e Support class parameters (#2231) (#3541) 2022-08-28 10:24:55 -04:00
Wilson Snyder 2358ced061 Rename tracing rolloverSize and add test (#3570). 2022-08-28 08:25:02 -04:00
Varun Koyyalagunta 5869fdf7f6 Fix $dump systemtask with --output-split-cfuncs (#3495) (#3497) 2022-08-25 18:29:11 -05:00
Krzysztof Bieganski 1a1d2ecfd9 Enable tracing in generated main (#3578)
Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-08-25 14:55:37 +01:00
Geza Lore 9ac64d0b92 Improve performance of MTask coarsening
Various optimizations to speed up MTasks coarsening (which is the long
pole in the multi-threaded scheduling of very large designs).

The biggest impact ones:
- Use efficient hand written Pairing Heaps for implementing priority
  queues and the scoreboard, instead of the old SortByValueMap. This
  helps us avoid having to sort a lot of merge candidates that we will
  never actually consider and helps a lot in performance.
- Remove unnecessary associative containers and store data structures
  (the heap nodes in particular) directly in the object they relate to.
  This eliminates a huge amount of lookups and helps a lot in
  performance.
- Distribute storage for SiblingMC instances into the LogicMTask
  instances, and combine with the sibling maps. This again eliminates
  hash table lookups and makes storage structures smaller.
- Remove some now bidirectional edge maps, keep only the forward map.

There are also some other smaller optimizations:
- Replaced more unnecessary dynamic_casts with static_casts
- Templated some functions/classes to reduce the number of static
  branches in loops.
- Improves sorting of edges for sibling candidate creation
- Various micro-optimizations here and there

This speeds up MTask coarsening by 3.8x on a large design, which
translates to a 2.5x speedup of the ordering pass in multi-threaded
mode. (Combined with the earlier optimizations, ordering is now 3x
faster.)

Due to the elimination of a lot of the auxiliary data structures, and
ensuring a minimal size for the necessary ones, memory consumption of
the MTask coarsening is also reduced (measured up to 4.4x reduction
though the accuracy of this is low).

The algorithm is identical except for minor alterations of the order
some candidates are added or removed, this can cause perturbation in the
output due to tied scores being broken based on IDs.
2022-08-20 21:18:50 +01:00
Wilson Snyder c6607724cb Fix clang warning. 2022-08-20 14:19:00 -04:00
Wilson Snyder 90dc04cf93 Add --future0 and --future1 options. 2022-08-20 14:01:13 -04:00
Geza Lore 4d81eb021d Revert "Improve performance of MTask coarsening"
This reverts commit 83475008d9.
2022-08-19 18:03:45 +01:00
Geza Lore 83475008d9 Improve performance of MTask coarsening
Various optimizations to speed up MTasks coarsening (which is the long
pole in the multi-threaded scheduling of very large designs).

The biggest impact ones:
- Use efficient hand written Pairing Heaps for implementing priority
  queues and the scoreboard, instead of the old SortByValueMap. This
  helps us avoid having to sort a lot of merge candidates that we will
  never actually consider and helps a lot in performance.
- Remove unnecessary associative containers and store data structures
  (the heap nodes in particular) directly in the object they relate to.
  This eliminates a huge amount of lookups and helps a lot in
  performance.
- Distribute storage for SiblingMC instances into the LogicMTask
  instances, and combine with the sibling maps. This again eliminates
  hash table lookups and makes storage structures smaller.
- Remove some now bidirectional edge maps, keep only the forward map.

There are also some other smaller optimizations:
- Replaced more unnecessary dynamic_casts with static_casts
- Templated some functions/classes to reduce the number of static
  branches in loops.
- Improves sorting of edges for sibling candidate creation
- Various micro-optimizations here and there

This speeds up MTask coarsening by 3.8x on a large design, which
translates to a 2.5x speedup of the ordering pass in multi-threaded
mode. (Combined with the earlier optimizations, ordering is now 3x
faster.)

Due to the elimination of a lot of the auxiliary data structures, and
ensuring a minimal size for the necessary ones, memory consumption of
the MTask coarsening is also reduced (measured up to 4.4x reduction
though the accuracy of this is low).

The algorithm is identical except for minor alterations of the order
some candidates are added or removed, this can cause perturbation in the
output due to tied scores being broken based on IDs.
2022-08-19 16:59:20 +01:00
Geza Lore 03ac7ad730 Make PartPropagateCp specific to the MTask graph
While keeping the client code abstract in PartPropagateCp is nice for
testing, there is performance to be had removing the abstraction. As
this code dominates in scheduling large designs, we eliminate the
abstraction and re-work the testing to use the actual LogicMTask and
MTaskEdge graph types. No functional change intended.
2022-08-19 14:06:11 +01:00
Geza Lore cd50949a7e Reuse MTaskEdge instances in MT scheduling
Instead of deleting then re-allocating MTaskEdge instances when merging
two MTasks, just redirect the edged of the donor MTask to the recipient
MTask. This is both faster as it avoids an allocation and a deletion,
together with one update of the sibling maps, and also makes the
algorithm more stable due to MergeCandidate IDs being stable and
allocated up front for all MTaskEdges, before any SiblingMCs are
allocated.

Perturbations in output are expected as the IDs used to break ties
between merge candidates with equal costs are not updated when
redirecting an edge (on purpose). The relinking of only one end of the
graph edges also perturbs the order in which they are enumerated, which
does change candidate opportunities when the number of edges is larger
than PART_SIBLING_EDGE_LIMIT. Confirmed output is identical when
IDs are updated and edges are updated to appear in their original order.
2022-08-19 14:06:11 +01:00
Geza Lore f0040c7b9a Remove reliance on pointer comparison in MT scheduling
The critical path propagation used to rely on a pointer comparison to
break equal scoring critical path updates. Use the corresponding mtask
ids instead, which is deterministic across invocations.
2022-08-19 14:06:11 +01:00
Geza Lore f8a0389e73 Do not use stepCost when gathering sibling merge candidates
siblingPairFromRelatives gathers neighbours of a vertex, and sorts them.
It then takes the N best nodes, and creates sibling merge candidates
from them. We now use the unadjusted cost instead of the step cost of
the vertices when sorting. This is both faster as we need not do the
log-space rounding to compute stepCost, and will also make similar but
yet cheaper nodes appear closer to the front as we don't lose precision
in rounding, hence they are more likely to be entered as merge
candidates. Note that when creating the merge candidate, we still use
the stepCost, so it's purpose of reducing the propagation of critical
path updates is maintained in full. In summary, this should make both
Verilator and the generated model very slightly faster, at least in
theory, and I have observed minor improvement in places.
2022-08-19 14:06:11 +01:00
Geza Lore b436794773 Add specialized GraphStreamUnordered
GraphStreamUnordered used to be GraphStream<std::less<const
V3GraphVertex*>>, but a lot of performance improvements can be had by a
specialized implementation, so added a highly optimized one. This helps
a lot with --debug-partition.
2022-08-19 14:06:11 +01:00
Geza Lore 90d22cbec6 Fix AstNode::exists return type 2022-08-19 13:22:06 +01:00
Krzysztof Bieganski 33e2acfe61 Fix AstNode::forall return type (#3559)
Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-08-19 12:33:17 +01:00
Ryszard Rozak db5fdfb0ee Fix === with some tristate constants (#3551). 2022-08-18 07:03:05 -04:00
Krzysztof Bieganski 951cd73fe0 Handle MemberSel in V3EmitV.cpp (#3555) 2022-08-18 06:33:45 -04:00
Arkadiusz Kozdra 0eeb40b975 Fix converting subclasses to string (#3552) 2022-08-17 18:08:43 -04:00
Wilson Snyder 93272c13fd Tests: Confirm fixed (#181) 2022-08-15 22:17:36 -04:00
Wilson Snyder 43abaeb055 Tests: Confirm fixed (#485) 2022-08-15 22:17:17 -04:00
Wilson Snyder 18b9e661c9 Tests: Confirm fixed (#446) 2022-08-15 22:17:09 -04:00
Wilson Snyder f435d96241 Fix case statement comparing string literal (#3544). 2022-08-15 21:56:09 -04:00
github action d32e3f042f Apply 'make format' 2022-08-12 10:56:12 +00:00
Mostafa Gamal df5f95a5bd Fix nested default assignment for struct pattern (#3511) (#3524) 2022-08-12 06:55:07 -04:00
Drew Ranck b0c475205b Fix void-cast queue pop_front or pop_back (#3542) (#3364)
Fix compile error for queue method usage, if it is the
first statement in a block of code, and the return
value is not used. Example:

>  if (foo)
>    void'(bar.pop_front());
2022-08-12 06:51:25 -04:00
Wilson Snyder 1e2219347e Internals: Cleanup ifdef, move up not under compilver version ifdef 2022-08-11 17:41:43 -04:00
Wilson Snyder cbe1b8e266 Fix segfault exporting non-existant package (#3535). 2022-08-08 17:53:50 -04:00
Mariusz Glebocki 2b12fe5773 Internals: Construct V3Number with correct type instead of changing it manually. (#3529) 2022-08-08 08:17:02 -04:00
Yutetsu TAKATSUKASA d20f22beb1 Fix tristate logic when reading inout port in a module #3399 (#3523)
* Tests: Add a test to reproduce #3399

* Fix #3399. When reading an inout port in a module, it should refer the
original inout port, not the generated MODTEMP.
2022-08-07 21:12:57 +09:00
Wilson Snyder f4fe10844b Tests: Fix t_flag_help.pl (#3532). 2022-08-07 04:57:59 -04:00
Mariusz Glebocki 122e89ffde Fix V3Number::isMsbXZ(). (#3530) 2022-08-05 19:12:52 +01:00
Geza Lore 96a4b3e5a5 Update clang-format config and apply
- Regroup and sort #include directives (like we used to, but automatic)
- Set AlwaysBreakTemplateDeclarations to true
2022-08-05 12:00:24 +01:00
Geza Lore fac8e76923 Rework SortByValueMap for better performance
Keep a single std::set of key/value pairs, and a single unordered_map
from key to iterators into the set. Also improve some of the accessing
mechanisms using modern C++. This speeds up multi-threaded ordering by
about 10%.
2022-08-03 21:17:02 +01:00
Geza Lore b864f5f5ba V3Partition: use static_cast with LogicMTaskVertex
dynamic_cast is not free, and the mtask graph contains only
LogicMTaskVertex vertices, use static_cast instead for some speedup.
2022-08-03 17:05:01 +01:00
Geza Lore f9f66d787e Fix integer overflow in V3Unroll (#3451) 2022-08-03 09:41:30 +01:00
Geza Lore bd211c87aa astgen: split 'visit' method declarations from definitions
Add definitions to V3Ast.cpp, and use static_cast.
This fixes a lot of clang-tidy noise.
2022-08-02 17:53:19 +01:00
Geza Lore 6c33e6e889 Tell clang-tidy .h files are C++ (not C) headers 2022-08-02 17:53:19 +01:00
Kamil Rakoczy cfb6fd8b34 Reduce max RSS usage (#3483)
By constant folding nodes earlier in V3Expand, we can save some max RSS on large designs.
2022-08-02 13:36:14 +01:00
Geza Lore cb60663d49 V3Gate: Defer substitutions until required as well
Similarly to the earlier patch that defers constant folding on optimized
logic, now we also defer the variable substitutions as well. This again
eliminates a lot of traversals, and yields another ~10x speedup of V3Gate
on a design where V3Gate used to dominate while producing identical
results.
2022-08-01 12:54:41 +01:00
Geza Lore 0d2bf23d82 V3Gate: Defer constant folding until required
Rather than constant folding each logic block after every substitution,
only constant fold updated blocks when re-analysed, or at the end. This
removes a lot of invocations of V3Const on large blocks that can be
optimized well, and should yield the same result.

This speeds up V3Gate by ~4x on a design where V3Gate dominates.
2022-07-31 20:42:04 +01:00
Geza Lore 682a60e325 Cleanup V3Gate, no functional change 2022-07-31 20:07:54 +01:00
Geza Lore 2ab6272cc7 Use AstNode::foreach in V3Gate
This yields a little speedup.
2022-07-31 20:05:25 +01:00
Geza Lore 152a6cd886 Improve AstNode::foreach (also exists and forall)
Speed improvements:
- Use a direct, recursion-free implementation
- Improve pre-fetching

Functionality:
- Support remove/replace of currently iterated node
2022-07-31 19:07:32 +01:00
Wilson Snyder f91793e931 Revert - SC overrides cause non-override clang error. 2022-07-30 13:53:54 -04:00
Wilson Snyder a2d26b45bb Internals: Fix some clang-tidy issues. No functional change intended. 2022-07-30 11:54:28 -04:00
Wilson Snyder dce8f3d25d Internals: Spacing from develop-v5. No functional change. 2022-07-30 11:54:28 -04:00
Geza Lore 38e5b6c1ad Replace __gcov_flush with __gcov_dump
__gcov_flush was a private function and was removed from later GCC
versions (at least from 11.2.0, possibly earlier). Replace with the
documented public __gcov_dump.
2022-07-30 16:02:03 +01:00
Wilson Snyder b9d7819faa Internals: Fix some cppcheck issues. Some dump functions fixed. 2022-07-30 10:01:39 -04:00
Yutetsu TAKATSUKASA 1f9323d086 Set correct dtype in replaceShiftSame() (#3520)
* Tests: Add a test to reproduce bug3399

* Fix3399. Set the correct dtype in replaceShiftSame().

* Tests: update stats.

* Update Changes
2022-07-29 07:05:04 +09:00
Geza Lore 574dbfded1 V3MergeCond: Fix incorrect merge of assignments to the condition 2022-07-28 15:50:02 +01:00
Wilson Snyder 2a87387eb3 Documentation fixes (#3514) 2022-07-28 08:41:01 -04:00
Geza Lore a5ddd10e31 Tests: compare VCD files both ways
vcddiff is a bit broken, and sometimes 'vcddiff a b' fails while the
files are indeed equivalent. There is a chance however that 'vcddif b a'
will succeed in this case, so compare trace files both ways when
checking test results and claim success if vcddiff succeeds in at least
one direction.
2022-07-27 10:48:02 +01:00
github action e871cd8a44 Apply 'make format' 2022-07-25 21:47:29 +00:00
Mostafa Gamal 7b431b37c7 Fix struct pattern assignment (#2328) (#3517). 2022-07-25 17:46:22 -04:00
Gustav Svensk eeef5ab4de Fix sformat string incorrectly cleared (#3515) (#3519). 2022-07-25 17:36:34 +02:00
Geza Lore ac4ec87942 Respect clang's default -fbracket-depth by default
Set default value of --comp-limit-parens to 240, to respect default
 maximum nesting of parentheses in clang (which is controlled by
 -fbracket-depth and defaults to 256). For code generation consistency,
 also use the same default with gcc.
2022-07-25 12:59:26 +01:00
Geza Lore 290c2e0388 Mark FileLine::v3errorEndFatal as noreturn 2022-07-25 12:51:02 +01:00
Geza Lore 89924bda51 Always type '$clog2' as signed 32 2022-07-25 12:48:13 +01:00
Yutetsu TAKATSUKASA 60eab3eb8c Fix wrong result of bit op tree optimization #3509 (#3516)
* Tests: Add a test to reproduce #3509

* Tests: Compile without tautological-compare check because bit op tree optimization is disabled in the test.

* Internals: Dedup code. No functional change is intended.

* Fix #3509.

"2'b10 == (2'b11 & {1'b0, val[0]})"  and "2'b10 != (2'b11 & {1'b0, val[0]})" were
wrongly optimized to "!val[0]" and "val[0]" respectively.
Now properly optimize them to 1'b0 and 1'b1.

* Commentary

* Commentary: Update Changes
2022-07-24 19:54:37 +09:00
Geza Lore e0b61ceabd Remove legacy #ifdef SYSTEMC_64BIT_PATCHES
These days this is always false, see #3505
2022-07-21 15:01:17 +01:00
Geza Lore 30e3edb81d Remove deprecated and unused timescale override defines
These have been 'deprecated' for 2 years and are otherwise unused except
for using a temporary placeholder value, which I have inlined with the
default value.

Also remove the now VL_TIME_STR_CONVERT utility function (and
corresponding unit tests), which have no references in any project on
GitHub.
2022-07-20 14:06:09 +01:00
Arkadiusz Kozdra 542e324869 Wildcard index type support for associative arrays (#3501).
Associative arrays that specify a wildcard index type may be indexed by
integral expressions of any size, with leading zeros removed
automatically.  A natural representation for such expressions is a
string, especially that the standard explicitly specifies automatic
casts from string indices to bit vectors of equivalent size.
The automatic cast part is done implicitly by the existing type system.

A simpler way to just make this work would be to convert wildcard index
type to a string type directly in the parser code, but several new AST
classes are needed to make sure illegal method calls are detected.
The verilated data structure implementation is reused, because there is
no need for differentiating the behavior on C++ side.
2022-07-20 15:01:36 +02:00
Geza Lore 1c5e5704f5 Fix iteration fixup in AstNode::addHereThisAsNext
Previous version broke verialor_ext_tests due to iteration order
mismatch after 3fc8249429
2022-07-20 13:08:51 +01:00
Geza Lore 1d400dd98c Configure tracing at run-time, instead of compile time (#3504)
All remaining use of conditional compilation in the tracing
implementation of the run-time library are replaced with the use of
VerilatedModel::traceConfig, and is now done at run-time.
2022-07-20 11:27:10 +01:00
Geza Lore a4ed3c2086 Make parallel tracing switchable at run-time 2022-07-19 17:13:13 +01:00
Geza Lore efb5caad22 Improve robustness of trace configuration
Always fail if adding a model to a trace file that has already executed
a dump. We used to do this before as well, though in a less robust way.
We will be relying on this property more in the future, so improve the
check.
2022-07-19 14:16:08 +01:00
Geza Lore 3a002b6cf2 Remove VerilatedVcd::m_evcd and related dead code.
The legacy code that was using this was removed earlier, and m_evcd was
constant false, so removed.
2022-07-19 13:58:18 +01:00
Geza Lore f8b7981be4 Make use of FST writer thread switchable at run-time.
Always build the FST libray with -DFST_WRITER_PARALLEL, iff VL_THREADED.
This supports run-time enablement of the FST writer thread, and has no
measurable performance impact on single threaded tracing but simplifies
the library build.

Note: the actual choice of using the fst writer thread is still compile
time, but can now be made run-time easily.
2022-07-19 13:48:03 +01:00
Geza Lore b55ee79d86 Fix typo 2022-07-19 12:36:21 +01:00
Geza Lore af70db88db Remove unused method 2022-07-19 11:32:16 +01:00
Geza Lore 7ef033f876 Ensure generated Makefile for hierarchical build is stable.
Avoid iterating unordered_map. Iterate sorted blocks instead.
2022-07-19 11:32:01 +01:00
Geza Lore db59c07f27 Implement trace offloading with fewer ifdefs
Step towards a proper run-time library. Reduce the amount of ifdefs in
the implementation of offloaded tracing. There are still a very small
number of ifdefs left, which will need more careful changes in order to
keep user API compatibility.
2022-07-19 11:31:35 +01:00
Geza Lore 9085e34d70 Pass VerilatedModel at trace registration time 2022-07-19 11:00:09 +01:00
Arkadiusz Kozdra 0dfa7d3af5 Internals: const-qualify findDType function. No functional change. (#3502) 2022-07-18 18:58:55 +02:00
Todd Strader b0e796ca83 Public combo propagation issues (#2905) 2022-07-15 11:44:32 -04:00
Geza Lore 3bd830eacf Minor clean up of initialization 2022-07-13 18:24:48 +01:00
Geza Lore f4efcbde5c Remove simple use of static data from V3OutFormatter::indentSpaces 2022-07-13 16:15:21 +01:00
Geza Lore 658819bb71 Trivial static const -> constexpr 2022-07-13 16:01:03 +01:00
Geza Lore 3fc8249429 Use AstNode::addHereThisAsNext in a few places 2022-07-13 13:57:00 +01:00
Geza Lore e0a38ce2c2 Remove unnecessary AstNode::clearIter() 2022-07-13 13:57:00 +01:00
Geza Lore 178e1789b5 Make AstNode::addHereThisAsNext always O(1)
Using unlinkFrBackWithNext is O(n) in the size of the list if unlinking
from the middle, so addHereThisAsNext also had this complexity. This
patch implements addHereThisAsNext directly, which is always O(1).
2022-07-13 12:13:40 +01:00
William D. Jones 108c900387 Fix unique_ptr memory header for MinGW64 (#3493). 2022-07-13 06:38:03 -04:00
Wilson Snyder 63507e8e29 Internals: Favor UASSERT_OBJ when have object. 2022-07-12 18:02:57 -04:00
Geza Lore 87f1e06c41 Small algorithmic improvement of PartContraction::siblingPairFromRelatives
Use std::partial_sort for the non-exhaustive case. This is O(n) instead
of O(n*log(n)) in the size of the candidate list being sorted. (It
actually is O(n*log(k)), but k is constant 6 in the non-exhaustive
case).
2022-07-12 19:10:01 +01:00
Geza Lore 7e8bafd217 Remove static data use from PartContraction::siblingPairFromRelatives
Use std::sort with lambda rather than qsort with static function and
static data. Verilation performance neutral.
2022-07-12 19:09:40 +01:00
Geza Lore 457ad07ade Remove unnecessary static state from V3EmitCFunc 2022-07-12 17:51:17 +01:00
Geza Lore 79c901c220 Tighten signatures/implementaion of VerilatedModel abstract methods. 2022-07-12 16:06:08 +01:00
Geza Lore b61d819fcb Move contextp() under VerilatedModel 2022-07-12 16:06:08 +01:00
Geza Lore f4038e3674 Move thread pool and execution profiler into the context. (#3477)
Fixes #3454
2022-07-12 11:41:15 +01:00
Arkadiusz Kozdra 8377514127 Add support for $test$plusargs(expr) (#3489) 2022-07-11 06:21:35 -04:00
Wilson Snyder d8ea989eda Tests/examples: Remove some legacy Verilator:: calls. 2022-07-09 09:50:50 -04:00
Wilson Snyder 5f3316d3dc * Fix empty string arguments to display (#3484). 2022-07-09 08:30:57 -04:00
Wilson Snyder a4fddb3fbe Fix table misoptimizing away display (#3488). 2022-07-09 07:55:46 -04:00
Wilson Snyder 3d71716a8a Internals: Constructor style cleanup. No functional change. 2022-07-09 07:40:07 -04:00
Yutetsu TAKATSUKASA 9f37cef1bb Fix #3470 of incorrect bit op tree optimization (#3476)
* Tests: Add a test to reproduce #3470

* Update LSB during return path of traversal. No functional change is intended.

* Introduce LeafInfo::m_msb

* Update LeafInfo::m_msb when visitin AstCCast

* Internals: Add comment, reorder. No functional change is intended.

* Delete explicit from copy constructor to fix build error.

* Update Changes

* Internals: Remove unused parameter. No functional change is intended.

* Tests: Add explanation to t_const_opt.
2022-07-06 08:33:37 +09:00
Geza Lore 0de1bbc85b Add and use VL_CONSTEXPR_CXX17 2022-07-05 14:21:28 +01:00
Geza Lore 3aa8624658 Set 'threads' in tests via parameter to compile
This is in preparation to #3454.
2022-07-05 12:33:41 +01:00
Geza Lore 42b711b862 Don't use 'assert' in profiler initialization 2022-07-05 12:18:54 +01:00
Mariusz Glebocki 2873dbe154 Optimize file writing by using a memory buffer. (#3461) 2022-07-04 10:23:31 -04:00
Felix Yan 31a83cb0d8 Fix use of C++17 on Arch Linux (#3479) 2022-06-29 11:24:30 -04:00
Geza Lore 383e384739 Remove always true cfg_with_threaded from test driver 2022-06-27 15:23:32 +01:00
Geza Lore 1bb6433649 Improve worker thread shutdown.
Always ensure worker thread task queue is drained before shutting down.
2022-06-27 15:03:36 +01:00
Yutetsu TAKATSUKASA ced39d0982 Internals: preparation for fixing #3470 (#3475)
* Internals: Let LeafInfo class. No functional change is intended.

* Internals: Rename LeafInfo::width -> LeafInfo::varWidth(). No functional change is intende.
2022-06-27 22:41:33 +09:00
Wilson Snyder fc4d6a62af Remove VL_PROFILER ifdef. Partial (#3454). 2022-06-22 20:06:23 -04:00
Wilson Snyder fa99cbbc73 Commentary: Fix mis-sorted option names. No functional change. 2022-06-21 19:28:26 -04:00
Wilson Snyder 49455721a3 Commentary 2022-06-21 19:28:23 -04:00
Unai Martinez-Corral 11032b1936 Fix bisonpre for MSYS2 (#3471) 2022-06-20 11:59:27 -04:00
Wilson Snyder f2fba51fe2 devel release 2022-06-19 15:13:29 -04:00
Wilson Snyder 7c79f0d431 Version bump (Changes update) 2022-06-19 15:10:23 -04:00
Wilson Snyder 41b9b30b9b Version bump 2022-06-19 15:01:17 -04:00
Wilson Snyder 1fa82ffb7b Commentary: Update ChangeLog 2022-06-19 15:00:10 -04:00
Wilson Snyder 4f93ac6477 Internals: Style modernization. No functional change intended. 2022-06-15 18:49:32 -04:00
Krzysztof Bieganski f7533010c6 Internals: Add setNoopt() function to LifeVisitor (#3468) 2022-06-15 18:11:03 -04:00
Todd Strader 47b650d821 Fix public unpacked input ports (#3465) 2022-06-15 07:41:59 -04:00
Geza Lore 0c2c097377 Add -fno-merge-cond-motion option
This disables code motion during V3MergeCond, for debugging.
2022-06-13 14:16:11 +01:00
Geza Lore d721f70690 Commentary 2022-06-13 12:14:37 +01:00
Kevin Kiningham ea8aaa21e8 Fix compile error under strict C++11 mode (#3463) 2022-06-13 12:14:02 +01:00
Kamil Rakoczy 660d1059b0 With --no-decoration, remove output whitespace (#3460)
Signed-off-by: Kamil Rakoczy <[email protected]>
2022-06-10 07:26:33 -04:00
Wilson Snyder e7dc2de14b Fix BLKANDNBLK on $readmem/$writemem (#3379). 2022-06-04 12:43:18 -04:00
Wilson Snyder 32100a3c91 Tests: Fix racy tests for develop-v5. 2022-06-04 12:17:04 -04:00
Wilson Snyder 67f7432dd7 Commentary (#3436). 2022-06-04 08:37:42 -04:00
Wilson Snyder 59dc2853e3 Support concat assignment to packed array (#3446). 2022-06-03 21:32:13 -04:00
Wilson Snyder ada58465b2 Add -f<optimization> options to replace -O<letter> options (#3436). 2022-06-03 20:43:16 -04:00
Wilson Snyder 173f57c636 Changed --no-merge-const-pool to -fno-merge-const-pool (#3436). 2022-06-03 19:41:59 -04:00
Wilson Snyder 1f3e8640f7 Examples: -Os should be a compile flag, not Verilator flag. 2022-06-03 18:45:39 -04:00
Wilson Snyder 6039e9dcc3 Commentary 2022-06-02 21:32:22 -04:00
Yutetsu TAKATSUKASA d64f979f99 Fix BitOpTree optimization to consider polarity of frozen node (#3445) (#3459)
* Tests: add a test to another failing case of #3445

* Consider polarity as lsb in BitOpTree optimization.
2022-06-01 09:26:16 +09:00
Huanghuang Zhou 0c53d19113 Commentary: InstrCountVisitor documentation (#3457)
Signed-off-by: huanghuang.zhou <[email protected]>
2022-05-31 07:10:58 -04:00
Geza Lore c64a07fd09 CI: fix cache keys in test jobs 2022-05-30 18:35:59 +01:00
Geza Lore 694919b9d1 CI: add ubuntu-22.04 regressions 2022-05-30 18:34:41 +01:00
Geza Lore 606b35853b Configure and compile with C++17 on Ubuntu 22.04
The packaged libsystemc on Ubuntu 22.04 uses C++17, so default to that
on that platform. Keep C++14 elsewhere.
2022-05-30 18:34:41 +01:00
Yutetsu TAKATSUKASAandGeza Lore 26b7452178 Fix #3445 of BitOpTreeOpt (#3453)
* Tests: Check BitOpTree statistics in t_const_opt.

* Tests: Add a test to reproduce #3445

* Fix #3445. Don't forget LSB of frozen node in BitOpTreeOpt.

* Apply suggestions from code review

Co-authored-by: Geza Lore <[email protected]>
2022-05-30 19:33:06 +09:00
Geza Lore b51f887567 Perform VCD tracing in parallel when using --threads (#3449)
VCD tracing is now parallelized using the same thread pool as the model.
We achieve this by breaking the top level trace functions into multiple
top level functions (as many as --threads), and after emitting the time
stamp to the VCD file on the main thread, we execute the tracing
functions in parallel on the same thread pool as the model (which we
pass to the trace file during registration), tracing into a secondary
per thread buffer. The main thread will then stitch (memcpy) the buffers
together into the output file.

This makes the `--trace-threads` option redundant with `--trace`, which
now only affects `--trace-fst`. FST tracing uses the previous offloading
scheme.

This obviously helps a lot in VCD tracing performance, and I have seen
better than Amdahl speedup, namely I get 3.9x on XiangShan 4T (2.7x on
OpenTitan 4T).
2022-05-29 19:08:39 +01:00
Geza Lore c4b8675d77 Always inline some small, hot trace routines 2022-05-28 12:47:09 +01:00
Geza Lore a7cd7a1ed9 Initialize VerilatedTrace members in class 2022-05-28 12:47:07 +01:00
Geza Lore a48c779367 Rename verilated_trace_imp.cpp -> verilated_trace_imp.h
Also fix file header to describe purpose of this file.
2022-05-28 12:20:35 +01:00
Geza Lore cf1eccc24f Make local function 'static' in verilated_profiler.h 2022-05-28 12:17:39 +01:00
Geza Lore d45caca011 Remove legacy VCD tracing API
This has not been used by Verilator for a while, but was kept for
compatibility with some external code. Now removed.
2022-05-28 12:07:24 +01:00
Geza Lore 0722f47539 Improve V3MergeCond by reordering statements (#3125)
V3MergeCond merges consecutive conditional `_ = cond ? _ : _` and
`if (cond) ...` statements. This patch adds an analysis and ordering
phase that moves statements with identical conditions closer to each
other, in order to enable more merging opportunities. This in turn
eliminates a lot of repeated conditionals which reduced dynamic branch
count and branch misprediction rate. Observed 6.5% improvement on
multi-threaded large designs, at the cost of less than 2% increase in
Verilation speed.
2022-05-27 16:57:51 +01:00
github action a372e010bd Apply 'make format' 2022-05-25 04:51:51 +00:00
Wilson Snyder 530817191e Support non-ANSI interface port declarations (#3439). 2022-05-25 00:50:50 -04:00
Geza Lore 1282548a1c Only consider definitions in t_flag_csplit 2022-05-20 18:02:00 +01:00
Geza Lore c7610ed044 Fix FST tracing thread in CMake build 2022-05-20 17:04:46 +01:00
Geza Lore ffc1c51526 Commentary 2022-05-20 16:44:53 +01:00
Geza Lore b130a8cfeb Add -DVM_TRACE_VCD in model builds with Make with --trace 2022-05-20 16:44:38 +01:00
Geza Lore 551bd284dd Rename some internals related to multi-threaded tracing
Rename the implementation internals of current multi-threaded tracing to
be "offload mode". No functional change, nor user interface change
intended.
2022-05-20 16:44:35 +01:00
Krzysztof Bieganski 9edccfdffa Initial support for intra-assignment timing controls, net delays (#3427)
This is a pre-PR to #3363.

Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-17 19:19:44 +01:00
Geza Lore 1a056f6db9 Fix invalid conditional merging when starting at 'c = c ? a : b'
Fixes #3409.
2022-05-17 18:36:40 +01:00
Krzysztof Bieganski e018eb7bac Support AstClass::repairCache() after V3Class (#3431)
This is a pre-PR to #3363.

Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-17 09:22:43 -04:00
Krzysztof Bieganski 0a91ddf38a Tests: Better grep check in t_foreach (#3435)
This is a pre-PR to #3363.

Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-17 09:20:59 -04:00
Krzysztof Bieganski 67bb2c640e Tests: Rename t_timing_clkgen to t_timing_clkgen1 (#3430)
This is a pre-PR to #3363, which will introduce more clock gen tests.

Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-17 09:19:51 -04:00
Krzysztof Bieganski 561eaa311d Tests: Enable CI testing with GCC 10 (#3432)
This is a pre-PR to #3363.

Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-17 09:19:06 -04:00
Krzysztof Bieganski 3f7a248ed4 Refactor some of the Begin handling to a separate function (#3426)
Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-16 20:45:33 +01:00
Krzysztof Bieganski ecaa07a72a Rename AstTimingControl to AstEventControl (#3425)
Signed-off-by: Krzysztof Bieganski <[email protected]>
2022-05-16 20:44:41 +01:00
Wilson Snyder 99bdc27be3 Internals: Cleanup some statics, trivial part towards (#3419) 2022-05-15 14:26:55 -04:00
Wilson Snyder c3c46967dc Tests: Appease sanitizer (#3121). 2022-05-15 11:50:52 -04:00
Wilson Snyder 7f1a9239ab Commentary, fix typo (#3121) 2022-05-15 11:14:07 -04:00
Geza Lore c8102c8ffe Fix typo 2022-05-15 16:01:35 +01:00
Wilson Snyder 5aa12e9b51 Add assert when VerilatedContext is mis-deleted (#3121). 2022-05-15 10:51:03 -04:00
Geza Lore 829437b20b Commentary - dependencies 2022-05-15 15:25:46 +01:00
Wilson Snyder 3c4131d45d Fix 'with' operator with type casting (#3387). 2022-05-15 09:53:48 -04:00
Wilson Snyder ae8d8ee1ac Fix crash with misuse of display. 2022-05-15 09:29:45 -04:00
Geza Lore 89ec3d16dc Allow const nodes in VNRef
No functional change.
2022-05-15 13:30:07 +01:00
Wilson Snyder 7602c983c5 Commentary 2022-05-14 18:16:31 -04:00
HungMingWu 560efb2c9e Internals: Fix memory leak in V3FileLine (#3407) (#3408). No functional change intended. 2022-05-14 18:15:38 -04:00
Wilson Snyder c2328ef46a Spelling fixes. 2022-05-14 16:12:57 -04:00
Wilson Snyder f6035447ae Internals: Use mutable for mutexes. No functional change. 2022-05-13 07:21:39 -04:00
Wilson Snyder 38438b3373 Internals: Cleanup some defaults. No functional change. 2022-05-12 23:30:39 -04:00
Wilson Snyder 71dedccbbe Support compile time trace signal selection with tracing_on/off (#3323). 2022-05-12 22:28:08 -04:00
Martin Stadler cddbe4642f Fix cmake rules to support higher-level targest (#3377) (#3386).
Don't add linker flags as generator expression to support linking `TARGET` to higher-level targets in a top-level CMakeLists.txt file.
2022-05-11 21:33:05 -04:00
Wilson Snyder bdfdc737a0 Internals: Cleanup V3Config. No functional change intended. 2022-05-11 00:47:52 -04:00
Wilson Snyder 3d045c3aee Internals: Cleanup some verilog.y formatting. No functional change. 2022-05-09 00:37:51 -04:00
HungMingWu 9583f152ee Fix compile error when enable VL_LEAK_CHECKS (#3411).
Signed-off-by: HungMingWu <[email protected]>
2022-05-08 20:49:13 -04:00
Wilson Snyder 5b2755d28d Untabify verilog.y (#3412). No functional change. 2022-05-08 20:46:18 -04:00
Kamil Rakoczy 9378259779 Fix UNOPTFLAT warning from initial static var (#3406)
Signed-off-by: Kamil Rakoczy <[email protected]>
2022-05-06 10:24:03 +02:00
Wilson Snyder 3d762282b9 Fix hang with large case statement optimization (#3405). 2022-05-05 07:02:52 -04:00
Wilson Snyder 30783e6a79 devel release 2022-05-02 22:23:05 -04:00
564 changed files with 21904 additions and 11342 deletions
+21 -10
View File
@@ -17,7 +17,7 @@ AllowShortLoopsOnASingleLine: true
AlwaysBreakAfterDefinitionReturnType: None
AlwaysBreakAfterReturnType: None
AlwaysBreakBeforeMultilineStrings: false
AlwaysBreakTemplateDeclarations: false
AlwaysBreakTemplateDeclarations: true
BinPackArguments: true
BinPackParameters: true
BraceWrapping:
@@ -59,15 +59,27 @@ ForEachMacros:
- foreach
- Q_FOREACH
- BOOST_FOREACH
IncludeBlocks: Preserve
# Include grouping/sorting
SortIncludes: true
IncludeBlocks: Regroup
IncludeCategories:
- Regex: '^"(llvm|llvm-c|clang|clang-c)/'
Priority: 2
- Regex: '^(<|"(gtest|gmock|isl|json)/)'
Priority: 3
- Regex: '.*'
Priority: 1
IncludeIsMainRegex: '(Test)?$'
- Regex: '"(config_build|verilated_config|verilatedos)\.h"'
Priority: -1 # Sepecials before main header
- Regex: '(<|")verilated.*'
Priority: 1 # Runtime headers
- Regex: '"V3.*__gen.*\.h"'
Priority: 3 # Generated internal headers separately
- Regex: '"V3.*"'
Priority: 2 # Internal header
- Regex: '".*"'
Priority: 4 # Other non-system headers
- Regex: '<[[:alnum:]_.]+>'
Priority: 5 # Simple system headers next
- Regex: '<.*>'
Priority: 6 # Other system headers next
IncludeIsMainRegex: '$'
IndentCaseLabels: false
IndentPPDirectives: None
IndentWidth: 4
@@ -91,7 +103,6 @@ PenaltyExcessCharacter: 1000000
PenaltyReturnTypeOnItsOwnLine: 60
PointerAlignment: Left
ReflowComments: true
SortIncludes: false
SortUsingDeclarations: true
SpaceAfterCStyleCast: false
SpaceAfterTemplateKeyword: true
+19 -3
View File
@@ -29,7 +29,7 @@ jobs:
strategy:
fail-fast: false
matrix:
os: [ubuntu-20.04, ubuntu-18.04]
os: [ubuntu-22.04, ubuntu-20.04, ubuntu-18.04]
compiler:
- { cc: clang, cxx: clang++ }
- { cc: gcc, cxx: g++ }
@@ -37,9 +37,16 @@ jobs:
exclude:
# Build pull requests only with ubuntu-20.04 and without m32
- os: ${{ github.event_name == 'pull_request' && 'ubuntu-18.04' || 'do-not-exclude' }}
- os: ${{ github.event_name == 'pull_request' && 'ubuntu-22.04' || 'do-not-exclude' }}
- m32: ${{ github.event_name == 'pull_request' && 1 || 'do-not-exclude' }}
# Build -m32 only on ubuntu-20.04
- {os: ubuntu-18.04, m32: 1}
- {os: ubuntu-22.04, m32: 1}
include:
# Build GCC 10 on ubuntu-20.04
- os: ubuntu-20.04
compiler: { cc: gcc-10, cxx: g++-10 }
m32: 0
runs-on: ${{ matrix.os }}
name: Build | ${{ matrix.os }} | ${{ matrix.compiler.cc }} ${{ matrix.m32 && '| -m32' || '' }}
env:
@@ -90,7 +97,7 @@ jobs:
strategy:
fail-fast: false
matrix:
os: [ubuntu-20.04, ubuntu-18.04]
os: [ubuntu-22.04, ubuntu-20.04, ubuntu-18.04]
compiler:
- { cc: clang, cxx: clang++ }
- { cc: gcc, cxx: g++ }
@@ -99,9 +106,18 @@ jobs:
exclude:
# Build pull requests only with ubuntu-20.04 and without m32
- os: ${{ github.event_name == 'pull_request' && 'ubuntu-18.04' || 'do-not-exclude' }}
- os: ${{ github.event_name == 'pull_request' && 'ubuntu-22.04' || 'do-not-exclude' }}
- m32: ${{ github.event_name == 'pull_request' && 1 || 'do-not-exclude' }}
# Build -m32 only on ubuntu-20.04
- {os: ubuntu-18.04, m32: 1}
- {os: ubuntu-22.04, m32: 1}
include:
# Test with GCC 10 on ubuntu-20.04 without m32
- {os: ubuntu-20.04, compiler: { cc: gcc-10, cxx: g++-10 }, m32: 0, suite: dist-vlt-0}
- {os: ubuntu-20.04, compiler: { cc: gcc-10, cxx: g++-10 }, m32: 0, suite: dist-vlt-1}
- {os: ubuntu-20.04, compiler: { cc: gcc-10, cxx: g++-10 }, m32: 0, suite: dist-vlt-2}
- {os: ubuntu-20.04, compiler: { cc: gcc-10, cxx: g++-10 }, m32: 0, suite: vltmt-0}
- {os: ubuntu-20.04, compiler: { cc: gcc-10, cxx: g++-10 }, m32: 0, suite: vltmt-1}
runs-on: ${{ matrix.os }}
name: Test | ${{ matrix.os }} | ${{ matrix.compiler.cc }} | ${{ matrix.suite }} ${{ matrix.m32 && '| -m32' || '' }}
env:
@@ -110,7 +126,7 @@ jobs:
CI_M32: ${{ matrix.m32 }}
CC: ${{ matrix.compiler.cc }}
CXX: ${{ matrix.compiler.cxx }}
CACHE_BASE_KEY: test-${{ matrix.os }}-${{ matrix.compiler.cc }}-m32=${{ matrix.m32 }}-${ matrix.suite }}
CACHE_BASE_KEY: test-${{ matrix.os }}-${{ matrix.compiler.cc }}-m32=${{ matrix.m32 }}-${{ matrix.suite }}
CCACHE_MAXSIZE: 64M # Per build matrix entry (2160M in total)
VERILATOR_ARCHIVE: verilator-${{ github.sha }}-${{ matrix.os }}-${{ matrix.compiler.cc }}${{ matrix.m32 && '-m32' || '' }}.tar.gz
steps:
+77 -13
View File
@@ -8,6 +8,70 @@ The changes in each Verilator version are described below. The
contributors that suggested a given feature are shown in []. Thanks!
Verilator 4.226 2022-08-31
==========================
**Minor:**
* Add --future0 and --future1 options.
* Support class parameters (#2231) (#3541). [Arkadiusz Kozdra/Antmicro]
* Support wildcard index associative arrays (#3501). [Arkadiusz Kozdra/Antmicro]
* Support negated properties (#3572). [Aleksander Kiryk]
* Support $test$plusargs(expr) (#3489).
* Rename trace rolloverSize() (#3570).
* Improve Verilation speed with --threads on large designs. [Geza Lore]
* Fix struct pattern assignment (#2328) (#3517). [Mostafa Gamal]
* Fix public combo propagation issues (#2905). [Todd Strader]
* Fix incorrect tristate logic (#3399) [shareefj, Vighnesh Iyer]
* Fix incorrect bit op tree optimization (#3470). [algrobman]
* Fix bisonpre for MSYS2 (#3471).
* Fix max memory usage (#3483). [Kamil Rakoczy]
* Fix empty string arguments to display (#3484). [Grulfen]
* Fix table misoptimizing away display (#3488). [Stefan Post]
* Fix unique_ptr memory header for MinGW64 (#3493).
* Fix $dump systemtask with --output-split-cfuncs (#3495) (#3497). [Varun Koyyalagunta]
* Fix wrong bit op tree optimization (#3509). [Nathan Graybeal]
* Fix nested default assignment for struct pattern (#3511) (#3524). [Mostafa Gamal]
* Fix sformat string incorrectly cleared (#3515) (#3519). [Gustav Svensk]
* Fix segfault exporting non-existant package (#3535).
* Fix void-cast queue pop_front or pop_back (#3542) (#3364). [Drew Ranck]
* Fix case statement comparing string literal (#3544). [Gustav Svensk]
* Fix === with some tristate constants (#3551). [Ryszard Rozak]
* Fix converting subclasses to string (#3552). [Arkadiusz Kozdra/Antmicro]
* Fix --hierarchical with order-based pin connections (#3583) (#3585). [Kelin9298]
Verilator 4.224 2022-06-19
==========================
**Major:**
* VCD tracing is now parallelized with --threads (#3449). [Geza Lore, Shunyao CAD]
**Minor:**
* Add -f<optimization> options to replace -O<letter> options (#3436).
* Changed --no-merge-const-pool to -fno-merge-const-pool (#3436).
* Changed --no-decoration to remove output whitespace (#3460). [Kamil Rakoczy]
* Support compile time trace signal selection with tracing_on/off (#3323). [Shunyao CAD]
* Support non-ANSI interface port declarations (#3439). [Geza Lore, Shunyao CAD]
* Support concat assignment to packed array (#3446).
* Improve conditional merging optimization (#3125). [Geza Lore, Shunyao CAD]
* Define VM_TRACE_VCD when tracing in VCD format. [Geza Lore, Shunyao CAD]
* Add assert when VerilatedContext is mis-deleted (#3121). [Rupert Swarbrick]
* Internal prep work towards timing control. [Krzysztof Bieganski/Antmicro]
* Fix hang with large case statement optimization (#3405). [Mike Urbach]
* Fix UNOPTFLAT warning from initial static var (#3406). [Kamil Rakoczy]
* Fix compile error when enable VL_LEAK_CHECKS (#3411). [HungMingWu]
* Fix cmake rules to support higher-level targets (#3377) (#3386). [Martin Stadler]
* Fix BLKANDNBLK on $readmem/$writemem (#3379). [Alex Solomatnikov]
* Fix 'with' operator with type casting (#3387). [xiak95]
* Fix incorrect conditional merging (#3409). [Raynard Qiao]
* Fix passing VL_TRACE_FST_WRITER_THREAD in CMake build. [Geza Lore, Shunyao CAD]
* Fix compile error under strict C++11 mode (#3463). [Kevin Kiningham]
* Fix public unpacked input ports (#3465). [Todd Strader]
Verilator 4.222 2022-05-02
==========================
@@ -43,7 +107,7 @@ Verilator 4.220 2022-03-12
* Add VERILATOR_VERSION_INTEGER for determining API (#3343). [Larry Doolittle]
* Improve various V3Combine algorithm details (#3328). [Yutetsu TAKATSUKASA]
* Improve various V3Order algorithm details. [Geza Lore]
* Fix macOS arm64 build (#3285) (#3291). [Guokai Chen]
* Fix MacOS arm64 build (#3285) (#3291). [Guokai Chen]
* Fix signed number operation (#3294) (#3308). [Raynard Qiao]
* Fix FST traces to include vector range (#3296) (#3297). [Jamie Iles]
* Fix skipping public enum values with four-state values (#3303).
@@ -108,7 +172,7 @@ Verilator 4.216 2021-12-05
**Minor:**
* Internal code cleanups and improvements. [Geza Lore]
* Improve --thread verilation-time performance.
* Improve --thread Verilation-time performance.
* Support task name in $display %m (#3211). [Julie Schwartz]
* Make 'bit', 'logic' and 'time' types unsigned by default. [Geza Lore]
* Optimize $random concatenates/selects (#3114).
@@ -132,7 +196,7 @@ Verilator 4.214 2021-10-17
**Major:**
* Add profile-guided optmization of mtasks (#3150).
* Add profile-guided optimization of mtasks (#3150).
**Minor:**
@@ -162,7 +226,7 @@ Verilator 4.212 2021-09-01
* Fix re-evaluation of logic dependent on state set in DPI exports (#3091). [Geza Lore]
* Support unpacked array localparams in tasks/functions (#3078). [Geza Lore]
* Support timeunit/timeprecision in $unit.
* Support assignment patterns as children of pins (#3041). [Krzysztof Bieganski]
* Support assignment patterns as children of pins (#3041). [Krzysztof Bieganski/Antmicro]
* Add --instr-count-dpi to tune assumed DPI import cost for multithreaded
model scheduling. Default value changed to 200 (#3068). [Yinan Xu]
* Output files are split based on the set of headers required
@@ -198,7 +262,7 @@ Verilator 4.210 2021-07-07
* Add --prof-c to pass profiling to compiler (#3059). [Alexander Grobman]
* Optimize a lot more model variables into function locals (#3027). [Geza Lore]
* Support middle-of-design nested topmodules (#3026). [Dan Petrisko]
* Support middle-of-design nested top modules (#3026). [Dan Petrisko]
* Remove deprecated --no-relative-cfuncs option (#3024). [Geza Lore]
* Remove deprecated --inhibit-sim option (#3035). [Geza Lore]
* Merge const static data globally into a new constant pool (#3013). [Geza Lore]
@@ -227,7 +291,7 @@ Verilator 4.204 2021-06-12
* Prep work towards better ccache hashing/performance. [Geza Lore]
* Fix assertion failure in bitOpTree optimization (#2891) (#2899). [Raynard Qiao]
* Fix DPI functions not seen as vpiModule (#2893). [Todd Strader]
* Fix bounds check in VL_SEL_IWII (#2910). [Krzysztof Bieganski]
* Fix bounds check in VL_SEL_IWII (#2910). [Krzysztof Bieganski/Antmicro]
* Fix slowdown in elaboration (#2911). [Nathan Graybeal]
* Fix initialization of assoc in assoc array (#2914). [myftptoyman]
* Fix make support for gmake 3.x (#2920) (#2921). [Philipp Wagner]
@@ -237,10 +301,10 @@ Verilator 4.204 2021-06-12
* Fix to emit 'else if' without nesting (#2944). [Geza Lore]
* Fix part select issues in LATCH warning (#2948) (#2938). [Julien Margetts]
* Fix to not emit empty files with low split limits (#2961). [Geza Lore]
* Fix merging of assignments in C++ code (#2970). [Ruper Swarbrick]
* Fix merging of assignments in C++ code (#2970). [Rupert Swarbrick]
* Fix unused variable warnings (#2991). [Pieter Kapsenberg]
* Fix --protect-ids when using SV classes (#2994). [Geza Lore]
* Fix constant function calls with uninit value (#2995). [yanx21]
* Fix constant function calls with uninitialized value (#2995). [yanx21]
* Fix Makefiles to support Windows EXEEXT usage (#3008). [Miodrag Milanovic]
@@ -278,7 +342,7 @@ Verilator 4.202 2021-04-24
* Fix Cygwin example compile issues (#2856). [Mark Shaw]
* Fix select of with index variable (#2880). [Alexander Grobman]
* Fix cmake version number to be numeric (#2881). [Yuri Victorovich]
* Fix MinGW not supportting 'localtime_r' (#2882). [HyungKi Jeong]
* Fix MinGW not supporting 'localtime_r' (#2882). [HyungKi Jeong]
* Fix cast from packed, typedef'ed interface signal (#2884). [Todd Strader]
* Fix VPI package reported as vpiModule (#2885). [Todd Strader]
* Fix dumping waveforms to multiple FST files (#2889). [David Metz]
@@ -342,7 +406,7 @@ Verilator 4.108 2021-01-10
**Major:**
* Many VPI changes for IEEE compatibility, which may alter behavior from previous releases.
* Support randomize() class method and rand (#2607). [Krzysztof Bieganski]
* Support randomize() class method and rand (#2607). [Krzysztof Bieganski/Antmicro]
**Minor:**
@@ -402,7 +466,7 @@ Verilator 4.104 2020-11-14
* Support queue and associative array 'with' statements (#2616).
* Support queue slicing (#2326).
* Support associative array pattern assignments and defaults.
* Support static methods and typedefs in classes (#2615). [Krzysztof Bieganski]
* Support static methods and typedefs in classes (#2615). [Krzysztof Bieganski/Antmicro]
* Add error on typedef referencing self (#2539). [Cody Piersall]
* With --debug, turn off address space layout randomization.
* Fix iteration over mutating list bug in VPI (#2588). [Kaleb Barrett]
@@ -425,7 +489,7 @@ Verilator 4.102 2020-10-15
* Support # as a comment in -f files (#2497). [phantom-killua]
* Support 'this' (#2585). [Rafal Kapuscik]
* Support defines for FST tracing (#2592). [Markus Krause]
* Support non-overapping implication inside properties (#1292). [Peter Monsson]
* Support non-overlapping implication inside properties (#1292). [Peter Monsson]
* Fix timescale with --hierarchical (#2554). [Yutetsu TAKATSUKASA]
* Fix cmake build with --hierarchical (#2560). [Yutetsu TAKATSUKASA]
* Fix -G dropping public indication (#2561). [Andrew Goessling]
@@ -451,7 +515,7 @@ Verilator 4.100 2020-09-07
* Support (with limitations) class extern, class extends, virtual class.
* Support $urandom, $urandom_range without stability.
* Support assume property. [Peter Monsson]
* Support non-overapping implication inside properties (#1292). [Peter Monsson]
* Support non-overlapping implication inside properties (#1292). [Peter Monsson]
* Fix false DECLFILENAME on black-boxed modules (#2430). [Philipp Wagner]
* Fix naming of "id : begin" blocks.
* Fix class constructor error on assignments to const.
+5 -2
View File
@@ -322,7 +322,10 @@ cppcheck: $(CPPCHECK_DEP)
$(CPPCHECK) $(CPPCHECK_FLAGS) -DVL_DEBUG=1 -DVL_CPPCHECK=1 -DVL_THREADED=1 $(CPPCHECK_INC) $<
CLANGTIDY = clang-tidy
CLANGTIDY_FLAGS = -config='' -checks='-fuchsia-*,-cppcoreguidelines-avoid-c-arrays,-cppcoreguidelines-init-variables'
CLANGTIDY_FLAGS = -config='' \
-header-filter='.*' \
-checks='-fuchsia-*,-cppcoreguidelines-avoid-c-arrays,-cppcoreguidelines-init-variables,-cppcoreguidelines-avoid-goto,-modernize-avoid-c-arrays,-readability-magic-numbers,-readability-simplify-boolean-expr,-cppcoreguidelines-macro-usage' \
CLANGTIDY_DEP = $(subst .h,.h.tidy,$(CPPCHECK_H)) \
$(subst .cpp,.cpp.tidy,$(CPPCHECK_CPP))
CLANGTIDY_DEFS = -DVL_DEBUG=1 -DVL_THREADED=1 -DVL_CPPCHECK=1
@@ -331,7 +334,7 @@ clang-tidy: $(CLANGTIDY_DEP)
%.cpp.tidy: %.cpp
$(CLANGTIDY) $(CLANGTIDY_FLAGS) $< -- $(CLANGTIDY_DEFS) $(CPPCHECK_INC) | 2>&1 tee $@
%.h.tidy: %.h
$(CLANGTIDY) $(CLANGTIDY_FLAGS) $< -- $(CLANGTIDY_DEFS) $(CPPCHECK_INC) | 2>&1 tee $@
$(CLANGTIDY) $(CLANGTIDY_FLAGS) $< -- $(CLANGTIDY_DEFS) $(CPPCHECK_INC) -x c++-header | 2>&1 tee $@
analyzer-src:
-rm -rf src/obj_dbg
+11 -10
View File
@@ -21,7 +21,7 @@ Welcome to Verilator
.. list-table::
* - **Welcome to Verilator, the fastest Verilog/SystemVerilog simulator.**
* Accepts synthesizable Verilog or SystemVerilog
* Accepts Verilog or SystemVerilog
* Performs lint code-quality checks
* Compiles into multithreaded C++, or SystemC
* Creates XML to front-end your own tools
@@ -57,17 +57,18 @@ files, the "Verilated" code.
The user writes a little C++/SystemC wrapper file, which instantiates the
"Verilated" model of the user's top level module. These C++/SystemC files
are then compiled by a C++ compiler (gcc/clang/MSVC++). The resulting
executable performs the design simulation. Verilator also supports linking
its generated libraries, optionally encrypted, into other simulators.
are then compiled by a C++ compiler (gcc/clang/MSVC++). Executing the
resulting executable performs the design simulation. Verilator also
supports linking Verilated generated libraries, optionally encrypted, into
other simulators.
Verilator may not be the best choice if you are expecting a full featured
replacement for NC-Verilog, VCS or another commercial Verilog simulator, or
if you are looking for a behavioral Verilog simulator e.g. for a quick
class project (we recommend `Icarus Verilog`_ for this.) However, if you
are looking for a path to migrate SystemVerilog to C++ or SystemC, or your
team is comfortable writing just a touch of C++ code, Verilator is the tool
for you.
replacement for Incisive, ModelSim/Questa, VCS or another commercial
Verilog simulator, or if you are looking for a behavioral Verilog simulator
e.g. for a quick class project (we recommend `Icarus Verilog`_ for this.)
However, if you are looking for a path to migrate SystemVerilog to C++ or
SystemC, or your team is comfortable writing just a touch of C++ code,
Verilator is the tool for you.
Performance
+22 -19
View File
@@ -284,33 +284,34 @@ detailed descriptions of these arguments.
--bbox-unsup Blackbox unsupported language features
--bin <filename> Override Verilator binary
--build Build model executable/library after Verilation
-CFLAGS <flags> C++ compiler arguments for makefile
--cc Create C++ output
--cdc Clock domain crossing analysis
-CFLAGS <flags> C++ compiler arguments for makefile
--clk <signal-name> Mark specified signal as clock
--make <build-tool> Generate scripts for specified build tool
--no-clk <signal-name> Prevent marking specified signal as clock
--compiler <compiler-name> Tune for specified C++ compiler
--converge-limit <loops> Tune convergence settle time
--coverage Enable all coverage
--coverage-line Enable line coverage
--coverage-max-width <width> Maximum array depth for coverage
--coverage-toggle Enable toggle coverage
--coverage-user Enable SVL user coverage
--coverage-underscore Enable coverage of _signals
--coverage-user Enable SVL user coverage
-D<var>[=<value>] Set preprocessor define
--debug Enable debugging
--debug-check Enable debugging assertions
--no-debug-leak Disable leaking memory in --debug mode
--debugi <level> Enable debugging at a specified level
--debugi-<srcfile> <level> Enable debugging a source file at a level
--no-decoration Disable comments and symbol decorations
--default-language <lang> Default language to parse
+define+<var>=<value> Set preprocessor define
--dpi-hdr-only Only produce the DPI header file
--dump-defines Show preprocessor defines with -E
--dump-tree Enable dumping .tree files
--dump-tree-addrids Use short identifiers instead of addresses
--dump-treei <level> Enable dumping .tree files at a level
--dump-treei-<srcfile> <level> Enable dumping .tree file at a source file at a level
--dump-tree-addrids Use short identifiers instead of addresses
-E Preprocess, but do not compile
--error-limit <value> Abort after this number of errors
--exe Link to create executable
@@ -319,7 +320,9 @@ detailed descriptions of these arguments.
-f <file> Parse arguments from a file
-FI <file> Force include of a file
--flatten Force inlining of all modules, tasks and functions
-fno-<optimization> Disable internal optimization stage
-G<name>=<value> Overwrite top-level parameter
--gate-stmts <value> Tune gate optimizer depth
--gdb Run Verilator under GDB interactively
--gdbbt Run Verilator under GDB for backtrace
--generate-key Create random key for --protect-key
@@ -327,55 +330,50 @@ detailed descriptions of these arguments.
--help Display this help
--hierarchical Enable hierarchical Verilation
-I<dir> Directory to search for includes
-j <jobs> Parallelism for --build
--gate-stmts <value> Tune gate optimizer depth
--if-depth <value> Tune IFDEPTH warning
+incdir+<dir> Directory to search for includes
--inline-mult <value> Tune module inlining
--instr-count-dpi <value> Assumed dynamic instruction count of DPI imports
-LDFLAGS <flags> Linker pre-object arguments for makefile
-j <jobs> Parallelism for --build
--l2-name <value> Verilog scope name of the top module
--language <lang> Default language standard to parse
-LDFLAGS <flags> Linker pre-object arguments for makefile
--lib-create <name> Create a DPI library
+libext+<ext>+[ext]... Extensions for finding modules
--lint-only Lint, but do not make output
--make <build-tool> Generate scripts for specified build tool
-MAKEFLAGS <flags> Arguments to pass to make during --build
--max-num-width <value> Maximum number width (default: 64K)
--MMD Create .d dependency files
--MP Create phony dependency targets
--Mdir <directory> Name of output object directory
--no-merge-const-pool Disable merging of different types in const pool
--MMD Create .d dependency files
--mod-prefix <topname> Name to prepend to lower classes
--no-clk <signal-name> Prevent marking specified signal as clock
--no-decoration Disable comments and symbol decorations
--no-pins64 Don't use uint64_t's for 33-64 bit sigs
--no-skip-identical Disable skipping identical output
--MP Create phony dependency targets
+notimingchecks Ignored
-O0 Disable optimizations
-O3 High performance optimizations
-O<optimization-letter> Selectable optimizations
-o <executable> Name of final executable
--no-order-clock-delay Disable ordering clock enable assignments
--no-verilate Skip verilation and just compile previously Verilated code.
--output-split <statements> Split .cpp files into pieces
--output-split-cfuncs <statements> Split model functions
--output-split-ctrace <statements> Split tracing functions
-P Disable line numbers and blanks with -E
--pins-bv <bits> Specify types for top level ports
--pins-sc-uint Specify types for top level ports
--pins-sc-biguint Specify types for top level ports
--pins-sc-uint Specify types for top level ports
--pins-uint8 Specify types for top level ports
--no-pins64 Don't use uint64_t's for 33-64 bit sigs
--pipe-filter <command> Filter all input through a script
--pp-comments Show preprocessor comments with -E
--prefix <topname> Name of top level class
--private Debugging; see docs
--prof-c Compile C++ code with profiling
--prof-cfuncs Name functions for profiling
--prof-exec Enable generating execution profile for gantt chart
--prof-pgo Enable generating profiling data for PGO
--protect-key <key> Key for symbol protection
--protect-ids Hash identifier names for obscurity
--protect-key <key> Key for symbol protection
--protect-lib <name> Create a DPI protected library
--private Debugging; see docs
--public Debugging; see docs
--public-flat-rw Mark all variables, etc as public_flat_rw
-pvalue+<name>=<value> Overwrite toplevel parameter
@@ -386,6 +384,7 @@ detailed descriptions of these arguments.
--rr Run Verilator and record with rr
--savable Enable model save-restore
--sc Create SystemC output
--no-skip-identical Disable skipping identical output
--stats Create statistics file
--stats-vars Provide statistics on variables
-sv Enable SystemVerilog parsing
@@ -405,7 +404,7 @@ detailed descriptions of these arguments.
--trace-max-width <width> Maximum array depth for tracing
--trace-params Enable tracing of parameters
--trace-structs Enable tracing structure names
--trace-threads <threads> Enable waveform creation on separate threads
--trace-threads <threads> Enable FST waveform creation on separate threads
--trace-underscore Enable tracing of _signals
-U<var> Undefine preprocessor define
--unroll-count <loops> Tune maximum loop iterations
@@ -413,6 +412,7 @@ detailed descriptions of these arguments.
--unused-regexp <regexp> Tune UNUSED lint signals
-V Verbose version and config
-v <filename> Verilog library
--no-verilate Skip verilation and just compile previously Verilated code.
+verilog1995ext+<ext> Synonym for +1364-1995ext+<ext>
+verilog2001ext+<ext> Synonym for +1364-2001ext+<ext>
--version Displays program version and exits
@@ -427,6 +427,9 @@ detailed descriptions of these arguments.
-Wno-lint Disable all lint warnings
-Wno-style Disable all style warnings
-Wpedantic Warn on compliance-test issues
-Wwarn-<message> Enable specified warning message
-Wwarn-lint Enable lint warning message
-Wwarn-style Enable style warning message
--x-assign <mode> Assign non-initial Xs to this value
--x-initial <mode> Assign initial Xs to this value
--x-initial-edge Enable initial X->0 and X->1 edge triggers
+7 -3
View File
@@ -54,8 +54,12 @@ if [ "$CI_BUILD_STAGE_NAME" = "build" ]; then
if [ "$CI_OS_NAME" = "linux" ]; then
sudo apt-get update
sudo apt-get install libfl-dev libgoogle-perftools-dev ccache
if [ "$CI_RUNS_ON" = "ubuntu-20.04" ]; then
sudo apt-get install libfl-dev ccache
if [ "$CI_RUNS_ON" != "ubuntu-22.04" ]; then
# Some conflict of libunwind verison on 22.04, can live without it for now
sudo apt-get install libgoogle-perftools-dev
fi
if [ "$CI_RUNS_ON" = "ubuntu-20.04" ] || [ "$CI_RUNS_ON" = "ubuntu-22.04" ]; then
sudo apt-get install libsystemc libsystemc-dev
fi
if [ "$COVERAGE" = 1 ]; then
@@ -85,7 +89,7 @@ elif [ "$CI_BUILD_STAGE_NAME" = "test" ]; then
sudo apt-get update
# libfl-dev needed for internal coverage's test runs
sudo apt-get install gdb gtkwave lcov libfl-dev ccache
if [ "$CI_RUNS_ON" = "ubuntu-20.04" ]; then
if [ "$CI_RUNS_ON" = "ubuntu-20.04" ] || [ "$CI_RUNS_ON" = "ubuntu-22.04" ]; then
sudo apt-get install libsystemc-dev
fi
if [ "$CI_M32" = 1 ]; then
+13 -9
View File
@@ -5,14 +5,14 @@
# General Public License Version 3 or the Perl Artistic License Version 2.0.
# SPDX-License-Identifier: LGPL-3.0-only OR Artistic-2.0
#AC_INIT([Verilator],[#.### YYYY-MM-DD])
#AC_INIT([Verilator],[#.### devel])
AC_INIT([Verilator],[4.222 2022-05-02],
[https://verilator.org],
[verilator],[https://verilator.org])
# When releasing, also update header of Changes file
# and commit using "devel release" or "Version bump" message
# Then 'make maintainer-dist'
#AC_INIT([Verilator],[#.### YYYY-MM-DD])
#AC_INIT([Verilator],[#.### devel])
AC_INIT([Verilator],[4.226 2022-08-31],
[https://verilator.org],
[verilator],[https://verilator.org])
AC_CONFIG_HEADERS(src/config_build.h)
AC_CONFIG_FILES(Makefile src/Makefile src/Makefile_obj include/verilated.mk include/verilated_config.h verilator.pc verilator-config.cmake verilator-config-version.cmake)
@@ -348,14 +348,18 @@ AC_SUBST(CFG_CXXFLAGS_PROFILE)
# Flag to select newest language standard supported
# Macros work such that first option that passes is the one we take
# Currently enabled c++14 due to packaged SystemC dependency
# c++14 is the newest that Verilator is regressed to support
# Currently enable c++17/c++14 due to packaged SystemC dependency
# c++17 is the newest that Verilator is regularly tested to support
# c++11 is the oldest that Verilator supports
# gnu is requried for Cygwin to compile verilated.h successfully
#_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=gnu++20)
#_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=c++20)
#_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=gnu++17)
#_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=c++17)
case "$(which lsb_release 2>&1 > /dev/null && lsb_release -d)" in
*Arch*Linux* | *Ubuntu*22.04*)
_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=gnu++17)
_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=c++17)
;;
esac
_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=gnu++14)
_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=c++14)
_MY_CXX_CHECK_SET(CFG_CXXFLAGS_STD_NEWEST,-std=gnu++11)
+1
View File
@@ -2,3 +2,4 @@
*.html
*.pdf
_build
guide/spelling.txt
+10
View File
@@ -5,11 +5,13 @@ Please see the Verilator manual for 200+ additional contributors. Thanks to all.
Adrien Le Masle
Ahmed El-Mahmoudy
Aleksander Kiryk
Alex Chadwick
Àlex Torregrosa
Aliaksei Chapyzhenka
Ameya Vikram Singh
Andreas Kuster
Arkadiusz Kozdra
Chris Randall
Chuxuan Wang
Conor McCullough
@@ -32,9 +34,11 @@ Gianfranco Costamagna
Glen Gibb
Graham Rushton
Guokai Chen
Gustav Svensk
Harald Heckmann
Howard Su
Huang Rui
Huanghuang Zhou
HungMingWu
HyungKi Jeong
Iru Cai
@@ -67,16 +71,19 @@ Lukasz Dalek
Maarten De Braekeleer
Maciej Sobkowski
Marco Widmer
Mariusz Glebocki
Markus Krause
Marlon James
Marshal Qiao
Martin Schmidt
Martin Stadler
Matthew Ballance
Michael Killough
Michaël Lefebvre
Mike Popoloski
Miodrag Milanović
Morten Borup Petersen
Mostafa Gamal
Nandu Raj
Nathan Kohagen
Nathan Myers
@@ -93,6 +100,7 @@ Rafal Kapuscik
Raynard Qiao
Richard Myers
Rupert Swarbrick
Ryszard Rozak
Samuel Riedel
Sean Cross
Sebastien Van Cauwenberghe
@@ -109,9 +117,11 @@ Tomasz Gorochowik
Tymoteusz Blazejczyk
Udi Finkelstein
Unai Martinez-Corral
Varun Koyyalagunta
Vassilis Papaefstathiou
Veripool API Bot
Victor Besyakov
William D. Jones
Wilson Snyder
Xi Zhang
Yoda Lee
+2
View File
@@ -35,6 +35,7 @@ default:
# Intermediate rules
vl-extract: ../bin/verilator ../Changes
ln -sf ../spelling.txt guide/spelling.txt
mkdir -p _build/gen
$(PYTHON3) bin/vl_sphinx_extract ../bin/verilator
sed 's/`/\&96;/g' < ../Changes > _build/gen/Changes
@@ -77,6 +78,7 @@ clean mostlyclean distclean maintainer-clean::
rm -f *.pg *.pgs *.toc *.tp *.tps *.vr *.vrs *.idx
rm -f *.ev *.evs *.ov *.ovs *.cv *.cvs *.ma *.mas
rm -f *.tex
rm -f guide/spelling.txt
distclean maintainer-clean::
rm -f *.info* *.1 *.html *.pdf $(INFOS)
+13 -53
View File
@@ -86,58 +86,17 @@ Connecting to C++
In C++ output mode (:vlopt:`--cc`), the Verilator generated model class is a
simple C++ class. The user must write a C++ wrapper and main loop for the
simulation, which instantiates the model class, and link with the Verilated
model. Here is a simple example:
model.
.. code-block:: C++
Refer to ``examples/make_tracing_c`` in the distribution for a detailed
commented example.
#include <verilated.h> // Defines common routines
#include <iostream> // Need std::cout
#include "Vtop.h" // From Verilating "top.v"
Top level IO signals are read and written as members of the model. You
call the model's :code:`eval()` method to evaluate the model. When the
simulation is complete call the model's :code:`final()` method to execute
any SystemVerilog final blocks, and complete any assertions. See
:ref:`Evaluation Loop`.
Vtop *top; // Instantiation of model
uint64_t main_time = 0; // Current simulation time
// This is a 64-bit integer to reduce wrap over issues and
// allow modulus. This is in units of the timeprecision
// used in Verilog (or from --timescale-override)
double sc_time_stamp() { // Called by $time in Verilog
return main_time; // converts to double, to match
// what SystemC does
}
int main(int argc, char** argv) {
Verilated::commandArgs(argc, argv); // Remember args
top = new Vtop; // Create model
top->reset_l = 0; // Set some inputs
while (!Verilated::gotFinish()) {
if (main_time > 10) {
top->reset_l = 1; // Deassert reset
}
if ((main_time % 10) == 1) {
top->clk = 1; // Toggle clock
}
if ((main_time % 10) == 6) {
top->clk = 0;
}
top->eval(); // Evaluate model
cout << top->out << endl; // Read a output
main_time++; // Time passes...
}
top->final(); // Done simulating
// // (Though this example doesn't get here)
delete top;
}
Note top level IO signals are read and written as members of the model. You
call the :code:`eval()` method to evaluate the model. When the simulation is
complete call the :code:`final()` method to execute any SystemVerilog final
blocks, and complete any assertions. See :ref:`Evaluation Loop`.
Connecting to SystemC
@@ -446,14 +405,15 @@ accesses the above signal "readme" would be:
int main(int argc, char** argv, char** env) {
Verilated::commandArgs(argc, argv);
Vour* top = new Vour;
Verilated::internalsDump(); // See scopes to help debug
while (!Verilated::gotFinish()) {
const std::unique_ptr<VerilatedContext> contextp{new VerilatedContext};
const std::unique_ptr<Vour> top{new Vour{contextp.get()}};
contextp->internalsDump(); // See scopes to help debug
while (!contextp->gotFinish()) {
top->eval();
VerilatedVpi::callValueCbs(); // For signal callbacks
read_and_check();
}
delete top;
return 0;
}
EOF
+5
View File
@@ -20,6 +20,11 @@ Option `--cdc`
The experimental `--cdc` option is believed to be generally unused and is
planned for removal no sooner than January 2023.
Option `-O<letter>`
The debug `-O<letter>` options have been replaced with
`-fno-<optimization>` debug options to match GCC. The old options are
planned for removal no sooner than June 2023.
Option `--prof-threads`
The `--prof-threads` option has been superseded by the `--prof-exec` and
`--prof-pgo` options and is planned for removal no sooner than April 2023.
+2 -2
View File
@@ -8,7 +8,7 @@ Simulation Runtime Arguments
The following are the arguments that may be passed to a Verilated
executable, provided that executable calls
:code:`Verilated::commandArgs()`.
:code:`VerilatedContext*->commandArgs(argc, argv)`.
All simulation runtime arguments begin with "+verilator", so that the
user's executable may skip over all "+verilator" arguments when parsing its
@@ -96,7 +96,7 @@ Summary:
.. option:: +verilator+noassert
Disable assert checking per runtime argument. This is the same as
calling :code:`Verilated::assertOn(false)` in the model.
calling :code:`VerilatedContext*->assertOn(false)` in the model.
.. option:: +verilator+V
+202 -112
View File
@@ -129,16 +129,6 @@ Summary:
is also used). Verilator manages the build itself, and for this --build
requires GNU Make to be available on the platform.
.. option:: -CFLAGS <flags>
Add specified C compiler argument to the generated makefiles. For
multiple flags either pass them as a single argument with space
separators quoted in the shell (:command:`-CFLAGS "-a -b"`), or use
multiple -CFLAGS options (:command:`-CFLAGS -a -CFLAGS -b`).
When make is run on the generated makefile these will be passed to the
C++ compiler (g++/clang++/msvc++).
.. option:: --cc
Specifies C++ without SystemC output mode; see also :vlopt:`--sc`
@@ -156,6 +146,16 @@ Summary:
have interest in adding more traditional CDC checks, please contact the
authors.
.. option:: -CFLAGS <flags>
Add specified C compiler argument to the generated makefiles. For
multiple flags either pass them as a single argument with space
separators quoted in the shell (:command:`-CFLAGS "-a -b"`), or use
multiple -CFLAGS options (:command:`-CFLAGS -a -CFLAGS -b`).
When make is run on the generated makefile these will be passed to the
C++ compiler (g++/clang++/msvc++).
.. option:: --clk <signal-name>
With :vlopt:`--clk`, the specified signal-name is taken as a root clock
@@ -179,6 +179,11 @@ Summary:
remove it from the combinatorial logic reevaluation checking code. This
may greatly improve performance.
.. option:: --no-clk <signal-name>
Prevent the specified signal from being marked as clock. See
:vlopt:`--clk`.
.. option:: --compiler <compiler-name>
Enables workarounds for the specified C++ compiler (list below).
@@ -288,6 +293,13 @@ Summary:
<--debugi>`). Higher levels produce more detailed messages. See
:vlopt:`--debug` for other implications of enabling debug.
.. option:: --no-decoration
When creating output Verilated code, minimize comments, white space,
symbol names and other decorative items, at the cost of greatly reduced
readability. This may assist C++ compile times. This will not typically
change the ultimate model's performance, but may in some cases.
.. option:: --default-language <value>
Select the language to be used by default when first processing each
@@ -431,6 +443,78 @@ Summary:
flattening large designs may require significant CPU time, memory and
storage.
.. option:: -fno-acyc-simp
.. option:: -fno-assemble
.. option:: -fno-case
.. option:: -fno-combine
.. option:: -fno-const
.. option:: -fno-const-bit-op-tree
.. option:: -fno-dedup
.. option:: -fno-expand
.. option:: -fno-gate
.. option:: -fno-inline
.. option:: -fno-life
.. option:: -fno-life-post
.. option:: -fno-localize
.. option:: -fno-merge-cond
.. option:: -fno-merge-cond-motion
.. option:: -fno-merge-const-pool
.. option:: -fno-reloop
.. option:: -fno-reorder
.. option:: -fno-split
.. option:: -fno-subst
.. option:: -fno-subst-const
.. option:: -fno-table
Rarely needed. Disables one of the internal optimization steps. These
are typically used only when recommended by a maintainer to help debug
or work around an issue.
.. option:: -future0 <option>
Rarely needed. Suppress an unknown Verilator option for an option that
takes no additional arguments. This is used to allow scripts written
with pragmas for a later version of Verilator to run under a older
version. e.g. :code:`-future0 option --option` would on older versions
that do not understand :code:`--option` or :code:`+option` suppress what
would otherwise be an invalid option error, and on newer versions that
implement :code:`--option`, :code:`-future0 option --option` would have
the :code:`-future0 option` ignored and the :code:`--option` would
function appropriately.
.. option:: -future1 <option>
Rarely needed. Suppress an unknown Verilator option for an option that
takes an additional argument. This is used to allow scripts written
with pragmas for a later version of Verilator to run under a older
version. e.g. :code:`-future1 option --option arg` would on older
versions that do not understand :code:`--option arg` or :code:`+option
arg` suppress what would otherwise be an invalid option error, and on
newer versions that implement :code:`--option arg`, :code:`-future1
option --option arg` would have the :code:`-future1 option` ignored and
the :code:`--option arg` would function appropriately.
.. option:: -G<name>=<value>
Overwrites the given parameter of the toplevel module. The value is
@@ -537,21 +621,6 @@ Summary:
to limit the number of parallel build jobs but attempt to execute all
independent build steps in parallel.
.. option:: -LDFLAGS <flags>
Add specified C linker arguments to the generated makefiles. For multiple
flags either pass them as a single argument with space separators quoted
in the shell (``-LDFLAGS "-a -b"``), or use multiple -LDFLAGS arguments
(``-LDFLAGS -a -LDFLAGS -b``).
When make is run on the generated makefile these will be passed to the
C++ linker (ld) **after** the primary file being linked. This flag is
called :vlopt:`-LDFLAGS` as that's the traditional name in simulators;
it's would have been better called LDLIBS as that's the Makefile
variable it controls. (In Make, LDFLAGS is before the first object,
LDLIBS after. -L libraries need to be in the Make variable LDLIBS, not
LDFLAGS.)
.. option:: --l2-name <value>
Instead of using the module name when showing Verilog scope, use the
@@ -568,12 +637,20 @@ Summary:
A synonym for :vlopt:`--default-language`, for compatibility with other
tools and earlier versions of Verilator.
.. option:: +libext+<ext>[+<ext>][...]
.. option:: -LDFLAGS <flags>
Specify the extensions that should be used for finding modules. If for
example module "my" is referenced, look in :file:`my.<ext>`. Note
"+libext+" is fairly standard across Verilog tools. Defaults to
".v+.sv".
Add specified C linker arguments to the generated makefiles. For multiple
flags either pass them as a single argument with space separators quoted
in the shell (``-LDFLAGS "-a -b"``), or use multiple -LDFLAGS arguments
(``-LDFLAGS -a -LDFLAGS -b``).
When make is run on the generated makefile these will be passed to the
C++ linker (ld) **after** the primary file being linked. This flag is
called :vlopt:`-LDFLAGS` as that's the traditional name in simulators;
it's would have been better called LDLIBS as that's the Makefile
variable it controls. (In Make, LDFLAGS is before the first object,
LDLIBS after. -L libraries need to be in the Make variable LDLIBS, not
LDFLAGS.)
.. option:: --lib-create <name>
@@ -592,6 +669,13 @@ Summary:
See also :vlopt:`--protect-lib`.
.. option:: +libext+<ext>[+<ext>][...]
Specify the extensions that should be used for finding modules. If for
example module "my" is referenced, look in :file:`my.<ext>`. Note
"+libext+" is fairly standard across Verilog tools. Defaults to
".v+.sv".
.. option:: --lint-only
Check the files for lint violations only, do not create any other
@@ -630,17 +714,6 @@ Summary:
Set the maximum number literal width (e.g. in 1024'd22 this it the
1024). Defaults to 64K.
.. option:: --MMD =item --no-MMD
Enable/disable creation of .d dependency files, used for make dependency
detection, similar to gcc -MMD option. By default this option is
enabled for :vlopt:`--cc` or :vlopt:`--sc` modes.
.. option:: --MP
When creating .d dependency files with :vlopt:`--MMD` option, make phony
targets. Similar to :command:`gcc -MP` option.
.. option:: --Mdir <directory>
Specifies the name of the Make object directory. All generated files
@@ -648,40 +721,23 @@ Summary:
The directory is created if it does not exist and the parent directories
exist; otherwise manually create the Mdir before calling Verilator.
.. option:: --no-merge-const-pool
.. option:: --MMD
Rarely needed. In order to minimize cache footprint, values of different
data type, that are yet emitted identically in C++ are merged in the
constant pool. This option disables this and causes every constant pool
entry with a distinct data type to be emitted separately.
.. option:: --no-MMD
Enable/disable creation of .d dependency files, used for make dependency
detection, similar to gcc -MMD option. By default this option is
enabled for :vlopt:`--cc` or :vlopt:`--sc` modes.
.. option:: --mod-prefix <topname>
Specifies the name to prepend to all lower level classes. Defaults to
the same as :vlopt:`--prefix`.
.. option:: --no-clk <signal-name>
.. option:: --MP
Prevent the specified signal from being marked as clock. See
:vlopt:`--clk`.
.. option:: --no-decoration
When creating output Verilated code, minimize comments, white space,
symbol names and other decorative items, at the cost of greatly reduced
readability. This may assist C++ compile times. This will not typically
change the ultimate model's performance, but may in some cases.
.. option:: --no-pins64
Backward compatible alias for :vlopt:`--pins-bv 33 <--pins-bv>`.
.. option:: --no-skip-identical =item --skip-identical
Rarely needed. Disables or enables skipping execution of Verilator if
all source files are identical, and all output files exist with newer
dates. By default this option is enabled for :vlopt:`--cc` or
:vlopt:`--sc` modes only.
When creating .d dependency files with :vlopt:`--MMD` option, make phony
targets. Similar to :command:`gcc -MP` option.
.. option:: +notimingchecks
@@ -703,9 +759,9 @@ Summary:
Rarely needed. Enables or disables a specific optimizations, with the
optimization selected based on the letter passed. A lowercase letter
disables an optimization, an upper case letter enables it. This is
intended for debugging use only; see the source code for
version-dependent mappings of optimizations to -O letters.
disables an optimization, an upper case letter enables it. This option
is deprecated and the various `-f<optimization>` arguments should be
used instead.
.. option:: -o <executable>
@@ -760,11 +816,6 @@ Summary:
With :vlopt:`-E`, disable generation of :code:`&96;line` markers and
blank lines, similar to :command:`gcc -P`.
.. option:: --pins64
Backward compatible alias for :vlopt:`--pins-bv 65 <--pins-bv>`. Note
that's a 65, not a 64.
.. option:: --pins-bv <width>
Specifies SystemC inputs/outputs of greater than or equal to <width>
@@ -797,6 +848,15 @@ Summary:
of uint32_t. Likewise pins of width 9-16 will use uint16_t instead of
uint32_t.
.. option:: --pins64
Backward compatible alias for :vlopt:`--pins-bv 65 <--pins-bv>`. Note
that's a 65, not a 64.
.. option:: --no-pins64
Backward compatible alias for :vlopt:`--pins-bv 33 <--pins-bv>`.
.. option:: --pipe-filter <command>
Rarely needed. Verilator will spawn the specified command as a
@@ -826,6 +886,11 @@ Summary:
prepended to the name of the :vlopt:`--top` option, or V prepended to
the first Verilog filename passed on the command line.
.. option:: --private
Opposite of :vlopt:`--public`. Is the default; this option exists for
backwards compatibility.
.. option:: --prof-c
When compiling the C++ code, enable the compiler's profiling flag
@@ -847,35 +912,18 @@ Summary:
.. option:: --prof-exec
Enable collection of execution trace, that can be convered into a gantt
Enable collection of execution trace, that can be converted into a gantt
chart with verilator_gantt See :ref:`Execution Profiling`.
.. option:: --prof-pgo
Enable collection of profiling data for profile guided verilation. Currently
Enable collection of profiling data for profile guided Verilation. Currently
this is only useful with :vlopt:`--threads`. See :ref:`Thread PGO`.
.. option:: --prof-threads
Deprecated. Same as --prof-exec and --prof-pgo together.
.. option:: --protect-key <key>
Specifies the private key for :vlopt:`--protect-ids`. For best security
this key should be 16 or more random bytes, a reasonable secure choice
is the output of :command:`verilator --generate-key` . Typically, a key
would be created by the user once for a given protected design library,
then every Verilator run for subsequent versions of that library would
be passed the same :vlopt:`--protect-key`. Thus, if the input Verilog is
similar between library versions (Verilator runs), the Verilated code
will likewise be mostly similar.
If :vlopt:`--protect-key` is not specified and a key is needed,
Verilator will generate a new key for every Verilator run. As the key is
not saved, this is best for security, but means every Verilator run will
give vastly different output even for identical input, perhaps harming
compile times (and certainly thrashing any "ccache").
.. option:: --protect-ids
Hash any private identifiers (variable, module, and assertion block
@@ -896,6 +944,23 @@ Summary:
prototypes. Use of the VPI is not recommended as many design details
may be exposed, and an INSECURE warning will be issued.
.. option:: --protect-key <key>
Specifies the private key for :vlopt:`--protect-ids`. For best security
this key should be 16 or more random bytes, a reasonable secure choice
is the output of :command:`verilator --generate-key` . Typically, a key
would be created by the user once for a given protected design library,
then every Verilator run for subsequent versions of that library would
be passed the same :vlopt:`--protect-key`. Thus, if the input Verilog is
similar between library versions (Verilator runs), the Verilated code
will likewise be mostly similar.
If :vlopt:`--protect-key` is not specified and a key is needed,
Verilator will generate a new key for every Verilator run. As the key is
not saved, this is best for security, but means every Verilator run will
give vastly different output even for identical input, perhaps harming
compile times (and certainly thrashing any "ccache").
.. option:: --protect-lib <name>
Produces a DPI library similar to :vlopt:`--lib-create`, but hides
@@ -907,11 +972,6 @@ Summary:
in the distribution for a demonstration of how to build and use the DPI
library.
.. option:: --private
Opposite of :vlopt:`--public`. Is the default; this option exists for
backwards compatibility.
.. option:: --public
This is only for historical debug use. Using it may result in
@@ -1004,6 +1064,15 @@ Summary:
Specifies SystemC output mode; see also :vlopt:`--cc` option.
.. option:: --skip-identical
.. option:: --no-skip-identical
Rarely needed. Disables or enables skipping execution of Verilator if
all source files are identical, and all output files exist with newer
dates. By default this option is enabled for :vlopt:`--cc` or
:vlopt:`--sc` modes only.
.. option:: --stats
Creates a dump file with statistics on the design in
@@ -1041,7 +1110,8 @@ Summary:
is not thread safe. With "--threads 1", the generated model is single
threaded but may run in a multithreaded environment. With "--threads N",
where N >= 2, the model is generated to run multithreaded on up to N
threads. See :ref:`Multithreading`.
threads. See :ref:`Multithreading`. This option also applies to
:vlopt:`--trace` (but not :vlopt:`--trace-fst`).
.. option:: --threads-dpi all
@@ -1119,7 +1189,8 @@ Summary:
Having tracing compiled in may result in some small performance losses,
even when tracing is not turned on during model execution.
See also :vlopt:`--trace-threads` option.
When using :vlopt:`--threads`, VCD tracing is parallelized, using the
same number of threads as passed to :vlopt:`--threads`.
.. option:: --trace-coverage
@@ -1173,12 +1244,12 @@ Summary:
.. option:: --trace-threads *threads*
Enable waveform tracing using separate threads. This is typically faster
in simulation runtime but uses more total compute. This option is
independent of, and works with, both :vlopt:`--trace` and
:vlopt:`--trace-fst`. Different trace formats can take advantage of
more trace threads to varying degrees. Currently VCD tracing can utilize
at most "--trace-threads 1", and FST tracing can utilize at most
"--trace-threads 2". This overrides :vlopt:`--no-threads` .
in simulation runtime but uses more total compute. This option only
applies to :vlopt:`--trace-fst`. FST tracing can utilize at most
"--trace-threads 2". This overrides :vlopt:`--no-threads`.
This option is accepted, but has absolutely no effect with
:vlopt:`--trace`, which respects :vlopt:`--threads` instead.
.. option:: --trace-underscore
@@ -1676,9 +1747,28 @@ The grammar of configuration commands is as follows:
.. option:: tracing_off [-file "<filename>" [-lines <line> [ - <line> ]]]
Enable/disable waveform tracing for all future signals declared in the
specified filename (or wildcard with '\*' or '?', or all files if
omitted) and range of line numbers (or all lines if omitted).
.. option:: tracing_on [-scope "<scopename>" [-levels <levels> ]]
For tracing_off, instances below any module in the files/ranges
specified will also not be traced.
.. option:: tracing_off [-scope "<scopename>" [-levels <levels> ]]
Enable/disable waveform tracing for all future signals declared in
all files.
With -file, enable/disable waveform tracing in the specified
filename (or wildcard with '\*' or '?'), and -line range of line
numbers (or all lines if omitted).
For tracing_off with -file, instances below any module in the
files/ranges specified will also not be traced. To overcome this
feature, use tracing_on on the upper module declaration and on any
cells, or use the -scope flavor of the command.
With -scope enable/disable waveform tracing for the specified scope (or
wildcard with '\*' or '?'), and optional --levels number of levels
below. These controls only take place after other file/line/module
based controls have indicated the signal should be traced.
With -levels (used with -scope), the number of levels below that
scope which the rule is to match, where 0 means all levels below, 1
the exact level as the provided scope, and 2 meaning an additional
level of children below the provided scope, etc.
+1 -1
View File
@@ -217,7 +217,7 @@ or "`ifdef`"'s may break other tools.
:option:`public_flat` signals.
To force a marked signal from C++, set the corresponding `__VforceVal`
variable to the desired value, and the `__VforceEn` signal to the bitmask
variable to the desired value, and the `__VforceEn` signal to the bit-mask
indicating which bits of the signal to force. To force all bits of the
target signal, set `__VforceEn` to all ones. To release the signal (or part
thereof), set appropriate bits of the `__VforceEn` signal to zero.
+29 -14
View File
@@ -72,23 +72,38 @@ a good thing for getting working silicon.
Will Verilator output remain under my own license/copyright?
""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
Yes, it's just like using GCC on your programs; this is why Verilator uses
the "GNU **Lesser** Public License Version 3" instead of the more typical
"GNU Public License". See the licenses for details, but in brief, if you
change Verilator itself or the header files Verilator includes, you must
make the source code available under the GNU Lesser Public License.
However, Verilator output (the Verilated code) only "include"s the licensed
files, and so you are **not** required to open-source release any output
from Verilator.
Your SystemVerilog, VPI/DPI, or main() C++ code remains under your own license.
It's just like how using GCC on your programs does not change the copyright
of your program; this is why Verilator uses the "GNU **Lesser** Public
License Version 3" instead of the more typical "GNU Public License". See
the licenses for details.
Some examples:
* Any SystemVerilog or other input fed into Verilator remain your own.
* Any of your VPI/DPI C++ routines that Verilator calls remain your own.
* Any of your main() C++ code that calls into Verilator remain your own.
* If you change Verilator itself, for example changing or adding a file
under the src/ directory in the repository, you must make the source code
available under the GNU Lesser Public License.
* If you change a header Verilator provides, for example under include/ in
the repository, you must make the source code available under the GNU
Lesser Public License.
You also have the option of using the Perl Artistic License, which again
does not require you to release your Verilog or generated code, and also
allows you to modify Verilator for internal use without distributing the
modified version. But please contribute back to the community!
does not require you to release your Verilog, C++, or generated code. This
license also allows you to modify Verilator for internal use without
distributing the modified version. But please contribute back to the
community!
One limit is that you cannot under either license release a closed-source
Verilog simulation product incorporating Verilator. That is you can have a
commercial product, but must make the source code available.
Under both license you can offer a commercial product that is based on
Verilator either directly or embedded within. However under both licenses,
any changes you make to Verilator for such a product must be open sourced.
As is standard with Open Source, contributions back to Verilator will be
placed under the Verilator copyright and LGPL/Artistic license. Small test
+12 -6
View File
@@ -50,7 +50,7 @@ For --cc/--sc, it creates:
- Make include file with class names (from --make gmake)
* - *{prefix}*\ _hier.mk
- Make file for hierarchy blocks (from --make gmake)
* - *{prefix|*\ _hierMkArgs.f
* - *{prefix}*\ _hierMkArgs.f
- Arguments for hierarchical Verilation (from --make gmake)
* - *{prefix}*\ _hierCMakeArgs.f
- Arguments for hierarchical Verilation (from --make cmake)
@@ -62,13 +62,17 @@ For --cc/--sc, it creates:
- Top level (SystemVerilog $root) internal header file
* - *{prefix}*\ ___024root.cpp
- Top level (SystemVerilog $root) internal C++ file
* - *{prefix}*___024root*{__n}*\ .cpp
- Additional top level internal C++ files (from --output-split)
* - *{prefix}*\ ___024root\ *{__n}*\ .cpp
- Additional top level internal C++ files
* - *{prefix}*\ ___024root\ *{__DepSet_hash__n}*\ .cpp
- Additional top level internal C++ files (hashed to reduce build times)
* - *{prefix}*\ ___024root__Slow\ *{__n}*\ .cpp
- Infrequent cold routines
* - *{prefix}*\ ___024root__Trace{__n}*\ .cpp
* - *{prefix}*\ ___024root\ *{__DepSet_hash__n}*\ .cpp
- Infrequent cold routines (hashed to reduce build times)
* - *{prefix}*\ ___024root__Trace\ *{__n}*\ .cpp
- Wave file generation code (from --trace)
* - *{prefix}*\ ___024root__Trace__Slow{__n}*\ .cpp
* - *{prefix}*\ ___024root__Trace__Slow\ *{__n}*\ .cpp
- Wave file generation code (from --trace)
* - *{prefix}*\ __Dpi.h
- DPI import and export declarations (from --dpi)
@@ -87,7 +91,9 @@ For --cc/--sc, it creates:
* - *{prefix}{each_verilog_module}*\ .cpp
- Lower level internal C++ files
* - *{prefix}{each_verilog_module}{__n}*\ .cpp
- Additional lower C++ files (from --output-split)
- Additional lower C++ files
* - *{prefix}{each_verilog_module}{__DepSet_hash__n}*\ .cpp
- Additional lower C++ files (hased to reduce build times)
For --hierarchy mode, it creates:
+13 -5
View File
@@ -87,7 +87,7 @@ compiles under all the options above, plus using MSVC++.
Install Prerequisites
---------------------
To build or run Verilator you need these standard packages:
To build or run Verilator, you need these standard packages:
::
@@ -98,13 +98,20 @@ To build or run Verilator you need these standard packages:
sudo apt-get install libfl-dev # Ubuntu only (ignore if gives error)
sudo apt-get install zlibc zlib1g zlib1g-dev # Ubuntu only (ignore if gives error)
To build or run the following are optional but should be installed for good
performance:
To build or run Verilator, the following are optional but should be installed
for good performance:
::
sudo apt-get install ccache # If present at build, needed for run
sudo apt-get install libgoogle-perftools-dev numactl perl-doc
sudo apt-get install libgoogle-perftools-dev numactl
The following is optional but is recommended for nicely rendered command line
help when running Verilator:
::
sudo apt-get install perl-doc
To build Verilator you will need to install these packages; these do not
need to be present to run Verilator:
@@ -118,7 +125,8 @@ Those developing Verilator itself may also want these (see internals.rst):
::
sudo apt-get install gdb graphviz cmake clang clang-format-11 gprof lcov
sudo pip3 install sphinx sphinx_rtd_theme breathe
sudo apt-get install yapf3
sudo pip3 install sphinx sphinx_rtd_theme sphinxcontrib-spelling breathe
cpan install Pod::Perldoc
cpan install Parallel::Forker
+2 -2
View File
@@ -343,7 +343,7 @@ Verilator supports the procedural `force` (and corresponding `release`)
statement. The behavior of the `force` statement however does not entirely
comply with the IEEE 1800 SystemVerilog standard. According to the standard,
when a procedural statement of the form `force a = b;` is executed, the
simulation should behave as if from that point onwards, a continuous
simulation should behave as if from that point forwards, a continuous
assignment `assign a = b;` have been added to override the drivers of `a`.
More specifically: the value of `a` should be updated, whenever the value of
`b` changes, all the way until a `release a;` statement is executed.
@@ -489,7 +489,7 @@ $test$plusargs, $value$plusargs
.. code-block:: C++
Verilated::commandArgs(argc, argv);
{VerilatedContext*} ->commandArgs(argc, argv);
to register the command line before calling $test$plusargs or
$value$plusargs.
+1 -1
View File
@@ -405,7 +405,7 @@ others as they prove beneficial.
Thread Profile-Guided Optimization
----------------------------------
Verilator supports profile-guided optimization (verilation) of multi-threaded
Verilator supports profile-guided optimization (Verilation) of multi-threaded
models (Thread PGO) to improve performance.
When using multithreading, Verilator computes how long macro tasks take and
+14 -10
View File
@@ -221,9 +221,13 @@ model, it may be beneficial to performance to adjust the
influences the partitioning of the model by adjusting the assumed execution
time of DPI imports.
The :vlopt:`--trace-threads` options can be used to produce trace dumps
using multiple threads. If :vlopt:`--trace-threads` is set without
:vlopt:`--threads`, then :vlopt:`--trace-threads` will imply
When using :vlopt:`--trace` to perform VCD tracing, the VCD trace
construction is parallelized using the same number of threads as specified
with :vlopt:`--threads`, and is executed on the same thread pool as the model.
The :vlopt:`--trace-threads` options can be used with :vlopt:`--trace-fst`
to offload FST tracing using multiple threads. If :vlopt:`--trace-threads` is
given without :vlopt:`--threads`, then :vlopt:`--trace-threads` will imply
:vlopt:`--threads 1 <--threads>`, i.e.: the support libraries will be
thread safe.
@@ -231,12 +235,12 @@ With :vlopt:`--trace-threads 0 <--trace-threads>`, trace dumps are produced
on the main thread. This again gives the highest single thread performance.
With :vlopt:`--trace-threads {N} <--trace-threads>`, where N is at least 1,
N additional threads will be created and managed by the trace files (e.g.:
VerilatedVcdC or VerilatedFstC), to generate the trace dump. The main
thread will be released to proceed with execution as soon as possible,
though some blocking of the main thread is still necessary while capturing
the trace. Different trace formats can utilize a various number of
threads. See the :vlopt:`--trace-threads` option.
up to N additional threads will be created and managed by the trace files
(e.g.: VerilatedFstC), to offload construction of the trace dump. The main
thread will be released to proceed with execution as soon as possible, though
some blocking of the main thread is still necessary while capturing the
trace. FST tracing can utilize up to 2 offload threads, so there is no use
of setting :vlopt:`--trace-threads` higher than 2 at the moment.
When running a multithreaded model, the default Linux task scheduler often
works against the model, by assuming threads are short lived, and thus
@@ -441,7 +445,7 @@ SystemC include directories and link to the SystemC libraries.
.. describe:: TRACE_THREADS
Optional. Generated multi-threaded trace dumping, same as
Optional. Generated multi-threaded FST trace dumping, same as
"--trace-threads".
.. describe:: TOP_MODULE
+1 -1
View File
@@ -1506,7 +1506,7 @@ List Of Warnings
Error that a construct might be legal according to IEEE but is not
currently supported by Verilator.
A typical workaround is to recode the construct into a simpler and more
A typical workaround is to rewrite the construct into a simpler and more
common alternative language construct.
Alternatively, check if the construct is supported by other tools, and
+1 -1
View File
@@ -274,7 +274,7 @@ path through the graph is the sum of macro-task execution costs. Sarkar
does almost the same thing, except that he has nonzero estimates for
synchronization costs.
Verilator's cost estimates are assigned by ``InstrCountCostVisitor``. This
Verilator's cost estimates are assigned by ``InstrCountVisitor``. This
class is perhaps the most fragile piece of the multithread
implementation. It's easy to have a bug where you count something cheap
(eg. accessing one element of a huge array) as if it were expensive (eg.
+55
View File
@@ -2,9 +2,11 @@ ABCp
Accellera
Affe
Aleksander
Alexandre
Ami
Amir
Anastasiadis
Anikin
Antonin
Antwerpen
Arasanipalai
@@ -70,6 +72,7 @@ Eda
Eddleston
Egbert
Egil
Ehab
Eiler
Eivind
El
@@ -82,6 +85,7 @@ Fabrizio
Fekete
Ferrandi
Flachs
Flavien
Foletto
Forker
Francillon
@@ -104,6 +108,7 @@ Goessling
Gonnen
Goorah
Gossner
Graybeal
Grobman
Gunter
Guo
@@ -121,6 +126,7 @@ Hornung
Hossell
Hsu
Hyperthreading
Ibrahim
Iles
Inlines
Inout
@@ -134,6 +140,7 @@ Jannis
Jens
Jeras
Jiuyang
Joannou
Joly
Jullien
Junji
@@ -147,6 +154,7 @@ Kamendje
Kandadi
Kaplan
Karge
Karlsson
Katz
Katzman
Keren
@@ -161,6 +169,7 @@ Kolecki
Koonce
Korteland
Koszek
Kouping
Kravitz
Krolnik
Kruse
@@ -194,6 +203,7 @@ Mednick
Michiels
Microsystems
Milanovic
Millis
MinW
Mindspeed
Miodrag
@@ -224,12 +234,14 @@ Petr
Piechotka
Piersall
Plunkett
Popolon
Popov
Prabhat
Prabhu
Prateek
Pre
Preprocess
Pretet
Priyadharshini
Pullup
Pulver
@@ -262,12 +274,15 @@ Shi
Shinkarovsky
Shirakawa
Shuba
Shunyao
Slatter
SoC
Sobhan
Sokorac
Solaris
Solomatnikov
Solt
Southwell
Srini
Stamness
Stephane
@@ -286,6 +301,7 @@ SystemVerilog
Takatsukasa
Tambe
Tariq
Tejada
Tengstrand
Terpstra
Thiede
@@ -293,8 +309,10 @@ Thierry
Thyer
Tichelaar
Tomov
Tood
Topa
Tota
Trefor
Tresidder
Tri
Tristate
@@ -304,11 +322,13 @@ Ubixum
Uints
Undefines
Unsized
Urbach
Uselib
Usha
Usuario
VERILATOR
Vasu
Vdeeptemp
Vdly
Vemumtab
Vemuri
@@ -351,7 +371,9 @@ Zhang
abirkmanis
accessor
accessors
adrienlemasle
agrobman
ahouska
al
ala
andit
@@ -372,8 +394,12 @@ backtrace
backtraces
basename
bbox
benchmarking
biguint
biops
bitOpTree
bitOpTree
bitop
bitstoreal
blackbox
bokke
@@ -389,6 +415,7 @@ casez
casted
cb
ccache
ccall
cdc
ceil
celldefine
@@ -411,6 +438,8 @@ concat
concats
config
const
constexpr
constpool
coredump
countbits
countones
@@ -426,6 +455,7 @@ da
dat
datadir
datafiles
david
ddd
deassign
debugi
@@ -434,6 +464,7 @@ defname
defparam
demangling
der
dereference
desassign
destructor
detections
@@ -462,6 +493,7 @@ dumpvars
dut
dx
elab
elike
elsif
endcase
endcelldefine
@@ -480,15 +512,18 @@ enums
env
envvar
eof
errae
erroring
et
eval
evals
exe
executables
expr
extern
fanin
fasttrace
fauto
fbranch
fclose
fdisplay
@@ -503,6 +538,7 @@ filt
flto
flushCall
fopen
forceable
foreach
fprintf
fprofile
@@ -521,6 +557,7 @@ funcs
fwrite
gantt
gcc
gcda
gdb
genblk
genvar
@@ -552,6 +589,7 @@ incdir
includer
inferfaces
inhibitSim
initarray
initializer
initializers
inits
@@ -570,6 +608,7 @@ killua
lang
lcov
ld
leavinel
len
libext
libgoogle
@@ -584,6 +623,7 @@ ln
loc
localparam
localparams
localtime
logicals
longint
lsb
@@ -618,6 +658,7 @@ multithreaded
multithreading
mutexes
mux
myftptoyman
mysignal
namespace
nand
@@ -642,15 +683,19 @@ onehot
ooo
oprofile
oversubscription
parallelized
param
parameterized
params
parens
pawel
pc
pdf
perf
perftools
pgo
picoChip
pinIndex
pinout
plusargs
pmos
@@ -676,6 +721,7 @@ prev
printf
printtimescale
profcfunc
profiler
prototyptes
ps
pthread
@@ -703,12 +749,14 @@ regs
reloop
resetall
respecified
rodata
rr
rst
runtime
runtimes
rw
sVerilator
sawatzke
sc
scalared
sccanf
@@ -724,6 +772,7 @@ specparam
splitme
spp
sqrt
src
srcdir
srcfile
sscanf
@@ -743,6 +792,7 @@ subcells
subexpressions
submodule
submodules
substring
sv
svBitVal
svBitVecVal
@@ -777,6 +827,7 @@ trunc
txt
typ
typedef
typedef'ed
typedefed
typedefs
typename
@@ -790,6 +841,7 @@ uniquified
unistd
unlink
unlinked
unnamedblk
unopt
unoptflat
unoptimizable
@@ -839,11 +891,14 @@ writeme
writemem
writememb
writememh
xiak
xin
xml
xnor
xout
xuejiazidi
yanx
ypq
yurivict
zdave
Øyvind
+1 -1
View File
@@ -33,5 +33,5 @@ add_executable(example ../make_tracing_c/sim_main.cpp)
# Add the Verilated circuit to the target
verilate(example COVERAGE TRACE
INCLUDE_DIRS "../make_tracing_c"
VERILATOR_ARGS -f ../make_tracing_c/input.vc -Os -x-assign 0
VERILATOR_ARGS -f ../make_tracing_c/input.vc -x-assign fast
SOURCES ../make_tracing_c/top.v)
+1 -1
View File
@@ -45,7 +45,7 @@ set_property(
# Add the Verilated circuit to the target
verilate(example SYSTEMC COVERAGE TRACE
INCLUDE_DIRS "../make_tracing_sc"
VERILATOR_ARGS -f ../make_tracing_sc/input.vc -Os -x-assign 0
VERILATOR_ARGS -f ../make_tracing_sc/input.vc -x-assign fast
SOURCES ../make_tracing_sc/top.v)
verilator_link_systemc(example)
+5 -2
View File
@@ -21,11 +21,14 @@ int main(int argc, char** argv, char** env) {
// Prevent unused variable warnings
if (false && argc && argv && env) {}
// Construct a VerilatedContext to hold simulation time, etc.
VerilatedContext* contextp = new VerilatedContext;
// Construct the Verilated model, from Vtop.h generated from Verilating "top.v"
Vtop* top = new Vtop;
Vtop* top = new Vtop{contextp};
// Simulate until $finish
while (!Verilated::gotFinish()) {
while (!contextp->gotFinish()) {
// Evaluate model
top->eval();
+1 -1
View File
@@ -33,7 +33,7 @@ VERILATOR_FLAGS =
# Generate C++
VERILATOR_FLAGS += -cc
# Optimize
VERILATOR_FLAGS += -Os -x-assign 0
VERILATOR_FLAGS += -x-assign fast
# Warn abount lint issues; may not want this on less solid designs
VERILATOR_FLAGS += -Wall
# This example does not use vl_time_stamp but rather
+1 -1
View File
@@ -36,7 +36,7 @@ VERILATOR_FLAGS += -cc --exe
# Generate makefile dependencies (not shown as complicates the Makefile)
#VERILATOR_FLAGS += -MMD
# Optimize
VERILATOR_FLAGS += -Os -x-assign 0
VERILATOR_FLAGS += -x-assign fast
# Warn abount lint issues; may not want this on less solid designs
VERILATOR_FLAGS += -Wall
# Make waveforms
+2
View File
@@ -34,6 +34,8 @@ int main(int argc, char** argv, char** env) {
// Using unique_ptr is similar to
// "VerilatedContext* contextp = new VerilatedContext" then deleting at end.
const std::unique_ptr<VerilatedContext> contextp{new VerilatedContext};
// Do not instead make Vtop as a file-scope static variable, as the
// "C++ static initialization order fiasco" may cause a crash
// Set debug level, 0 is off, 9 is highest presently used
// May be overridden by commandArgs argument parsing
+1 -1
View File
@@ -37,7 +37,7 @@ VERILATOR_FLAGS += -sc --exe
# Generate makefile dependencies (not shown as complicates the Makefile)
#VERILATOR_FLAGS += -MMD
# Optimize
VERILATOR_FLAGS += -Os -x-assign 0
VERILATOR_FLAGS += -x-assign fast
# Warn abount lint issues; may not want this on less solid designs
VERILATOR_FLAGS += -Wall
# Make waveforms
+106 -44
View File
@@ -47,27 +47,34 @@
#define VERILATOR_VERILATED_CPP_
#include "verilatedos.h"
#include "verilated_imp.h"
#include "verilated_config.h"
#include "verilatedos.h"
#include "verilated_imp.h"
#include <algorithm>
#include <cctype>
#include <cerrno>
#include <cstdlib>
#include <sstream>
#include <sys/stat.h> // mkdir
#include <list>
#include <limits>
#include <list>
#include <sstream>
#include <utility>
#include <sys/stat.h> // mkdir
// clang-format off
#if defined(_WIN32) || defined(__MINGW32__)
# include <direct.h> // mkdir
#endif
#ifdef VL_THREADED
# include "verilated_threads.h"
#endif
// clang-format on
#include "verilated_trace.h"
// Max characters in static char string for VL_VALUE_STRING
constexpr unsigned VL_VALUE_STRING_MAX_WIDTH = 8192;
@@ -673,7 +680,10 @@ std::string _vl_vsformat_time(char* tmp, T ld, int timeunit, bool left, size_t w
if (std::numeric_limits<T>::is_integer) {
constexpr int b = 128;
constexpr int w = VL_WORDS_I(b);
VlWide<w> tmp0, tmp1, tmp2, tmp3;
VlWide<w> tmp0;
VlWide<w> tmp1;
VlWide<w> tmp2;
VlWide<w> tmp3;
WDataInP shifted = VL_EXTEND_WQ(b, 0, tmp0, static_cast<QData>(ld));
if (shift < 0) {
@@ -691,7 +701,8 @@ std::string _vl_vsformat_time(char* tmp, T ld, int timeunit, bool left, size_t w
= VL_EXTEND_WQ(b, 0, tmp2, std::numeric_limits<uint64_t>::max()); // breaks shifted
if (VL_GT_W(w, integer, max64Bit)) {
WDataOutP v = VL_ASSIGN_W(b, tmp3, integer); // breaks fracDigitsPow10
VlWide<w> zero, ten;
VlWide<w> zero;
VlWide<w> ten;
VL_ZERO_W(b, zero);
VL_EXTEND_WI(b, 0, ten, 10);
char buf[128]; // 128B is obviously long enough to represent 128bit integer in decimal
@@ -1453,11 +1464,12 @@ void VL_SFORMAT_X(int obits, void* destp, const char* formatp, ...) VL_MT_SAFE {
void VL_SFORMAT_X(int obits_ignored, std::string& output, const char* formatp, ...) VL_MT_SAFE {
if (obits_ignored) {}
output = "";
std::string temp_output;
va_list ap;
va_start(ap, formatp);
_vl_vsformat(output, formatp, ap);
_vl_vsformat(temp_output, formatp, ap);
va_end(ap);
output = temp_output;
}
std::string VL_SFORMATF_NX(const char* formatp, ...) VL_MT_SAFE {
@@ -1606,8 +1618,8 @@ IData VL_SYSTEM_IW(int lhswords, const WDataInP lhsp) VL_MT_SAFE {
return code >> 8; // Want exit status
}
IData VL_TESTPLUSARGS_I(const char* formatp) VL_MT_SAFE {
const std::string& match = Verilated::threadContextp()->impp()->argPlusMatch(formatp);
IData VL_TESTPLUSARGS_I(const std::string& format) VL_MT_SAFE {
const std::string& match = Verilated::threadContextp()->impp()->argPlusMatch(format.c_str());
return match.empty() ? 0 : 1;
}
@@ -1840,8 +1852,7 @@ VlReadMem::VlReadMem(bool hex, int bits, const std::string& filename, QData star
, m_bits{bits}
, m_filename(filename) // Need () or GCC 4.8 false warning
, m_end{end}
, m_addr{start}
, m_linenum{0} {
, m_addr{start} {
m_fp = std::fopen(filename.c_str(), "r");
if (VL_UNLIKELY(!m_fp)) {
// We don't report the Verilog source filename as it slow to have to pass it down
@@ -1975,8 +1986,7 @@ void VlReadMem::setData(void* valuep, const std::string& rhs) {
VlWriteMem::VlWriteMem(bool hex, int bits, const std::string& filename, QData start, QData end)
: m_hex{hex}
, m_bits{bits}
, m_addr{0} {
, m_bits{bits} {
if (VL_UNLIKELY(start > end)) {
VL_FATAL_MT(filename.c_str(), 0, "", "$writemem invalid address range");
return;
@@ -2158,29 +2168,6 @@ void VL_WRITEMEM_N(bool hex, // Hex format, else binary
//===========================================================================
// Timescale conversion
// Helper function for conversion of timescale strings
// Converts (1|10|100)(s|ms|us|ns|ps|fs) to power of then
int VL_TIME_STR_CONVERT(const char* strp) VL_PURE {
int scale = 0;
if (!strp) return 0;
if (*strp++ != '1') return 0;
while (*strp == '0') {
++scale;
++strp;
}
switch (*strp++) {
case 's': break;
case 'm': scale -= 3; break;
case 'u': scale -= 6; break;
case 'n': scale -= 9; break;
case 'p': scale -= 12; break;
case 'f': scale -= 15; break;
default: return 0;
}
if ((scale < 0) && (*strp++ != 's')) return 0;
if (*strp) return 0;
return scale;
}
static const char* vl_time_str(int scale) VL_PURE {
static const char* const names[]
= {"100s", "10s", "1s", "100ms", "10ms", "1ms", "100us", "10us", "1us",
@@ -2289,11 +2276,22 @@ VerilatedContext::VerilatedContext()
}
// Must declare here not in interface, as otherwise forward declarations not known
VerilatedContext::~VerilatedContext() {}
VerilatedContext::~VerilatedContext() {
checkMagic(this);
m_magic = 0x1; // Arbitrary but 0x1 is what Verilator src uses for a deleted pointer
}
void VerilatedContext::checkMagic(const VerilatedContext* contextp) {
if (VL_UNLIKELY(!contextp || contextp->m_magic != MAGIC)) {
VL_FATAL_MT("", 0, "", // LCOV_EXCL_LINE
"Attempt to create model using a bad/deleted VerilatedContext pointer");
}
}
VerilatedContext::Serialized::Serialized() {
m_timeunit = VL_TIME_UNIT; // Initial value until overriden by _Vconfigure
m_timeprecision = VL_TIME_PRECISION; // Initial value until overriden by _Vconfigure
constexpr int8_t picosecond = -12;
m_timeunit = picosecond; // Initial value until overriden by _Vconfigure
m_timeprecision = picosecond; // Initial value until overriden by _Vconfigure
}
void VerilatedContext::assertOn(bool flag) VL_MT_SAFE {
@@ -2313,7 +2311,7 @@ std::string VerilatedContext::dumpfile() const VL_MT_SAFE_EXCLUDES(m_timeDumpMut
return m_dumpfile;
}
std::string VerilatedContext::dumpfileCheck() const VL_MT_SAFE_EXCLUDES(m_timeDumpMutex) {
const std::string out = dumpfile();
std::string out = dumpfile();
if (VL_UNLIKELY(out.empty())) {
VL_PRINTF_MT("%%Warning: $dumpvar ignored as not proceeded by $dumpfile\n");
return "";
@@ -2418,6 +2416,33 @@ const char* VerilatedContext::timeprecisionString() const VL_MT_SAFE {
return vl_time_str(timeprecision());
}
void VerilatedContext::threads(unsigned n) {
if (n == 0) VL_FATAL_MT(__FILE__, __LINE__, "", "%Error: Simulation threads must be >= 1");
if (m_threadPool) {
VL_FATAL_MT(
__FILE__, __LINE__, "",
"%Error: Cannot set simulation threads after the thread pool has been created.");
}
#if VL_THREADED
if (m_threads == n) return; // To avoid unnecessary warnings
m_threads = n;
const unsigned hardwareThreadsAvailable = std::thread::hardware_concurrency();
if (m_threads > hardwareThreadsAvailable) {
VL_PRINTF_MT("%%Warning: System has %u hardware threads but simulation thread count set "
"to %u. This will likely cause significant slowdown.\n",
hardwareThreadsAvailable, m_threads);
}
#else
if (n > 1) {
VL_PRINTF_MT("%%Warning: Verilator run-time library built without VL_THREADS. Ignoring "
"call to 'VerilatedContext::threads' with argument %u.\n",
n);
}
#endif
}
void VerilatedContext::commandArgs(int argc, const char** argv) VL_MT_SAFE_EXCLUDES(m_argMutex) {
const VerilatedLockGuard lock{m_argMutex};
m_args.m_argVec.clear(); // Empty first, then add
@@ -2448,6 +2473,33 @@ void VerilatedContext::internalsDump() const VL_MT_SAFE {
VerilatedImp::userDump();
}
void VerilatedContext::addModel(VerilatedModel* modelp) {
threadPoolp(); // Ensure thread pool is created, so m_threads cannot change any more
if (modelp->threads() > m_threads) {
std::ostringstream msg;
msg << "VerilatedContext has " << m_threads << " threads but model '"
<< modelp->modelName() << "' (instantiated as '" << modelp->hierName()
<< "') was Verilated with --threads " << modelp->threads() << ".\n";
const std::string str = msg.str();
VL_FATAL_MT(__FILE__, __LINE__, modelp->hierName(), str.c_str());
}
}
VerilatedVirtualBase* VerilatedContext::threadPoolp() {
if (m_threads == 1) return nullptr;
#if VL_THREADED
if (!m_threadPool) m_threadPool.reset(new VlThreadPool{this, m_threads - 1});
#endif
return m_threadPool.get();
}
VerilatedVirtualBase*
VerilatedContext::enableExecutionProfiler(VerilatedVirtualBase* (*construct)(VerilatedContext&)) {
if (!m_executionProfiler) m_executionProfiler.reset(construct(*this));
return m_executionProfiler.get();
}
//======================================================================
// VerilatedContextImp:: Methods - command line
@@ -2657,6 +2709,7 @@ const VerilatedScopeNameMap* VerilatedContext::scopeNameMap() VL_MT_SAFE {
VerilatedSyms::VerilatedSyms(VerilatedContext* contextp)
: _vm_contextp__(contextp ? contextp : Verilated::threadContextp()) {
VerilatedContext::checkMagic(_vm_contextp__);
Verilated::threadContextp(_vm_contextp__);
#ifdef VL_THREADED
__Vm_evalMsgQp = new VerilatedEvalMsgQueue;
@@ -2664,6 +2717,7 @@ VerilatedSyms::VerilatedSyms(VerilatedContext* contextp)
}
VerilatedSyms::~VerilatedSyms() {
VerilatedContext::checkMagic(_vm_contextp__);
#ifdef VL_THREADED
delete __Vm_evalMsgQp;
#endif
@@ -2755,7 +2809,7 @@ void Verilated::runFlushCallbacks() VL_MT_SAFE {
// When running internal code coverage (gcc --coverage, as opposed to
// verilator --coverage), dump coverage data to properly cover failing
// tests.
VL_GCOV_FLUSH();
VL_GCOV_DUMP();
}
void Verilated::addExitCb(VoidPCb cb, void* datap) VL_MT_SAFE {
@@ -2838,6 +2892,14 @@ void VerilatedImp::versionDump() VL_MT_SAFE {
VL_PRINTF_MT(" Version: %s %s\n", Verilated::productName(), Verilated::productVersion());
}
//===========================================================================
// VerilatedModel:: Methods
VerilatedModel::VerilatedModel(VerilatedContext& context)
: m_context{context} {}
std::unique_ptr<VerilatedTraceConfig> VerilatedModel::traceConfig() const { return nullptr; }
//===========================================================================
// VerilatedModule:: Methods
+71 -7
View File
@@ -81,6 +81,7 @@
#endif
// clang-format on
class VerilatedContext;
class VerilatedContextImp;
class VerilatedContextImpData;
class VerilatedCovContext;
@@ -90,6 +91,9 @@ class VerilatedFstC;
class VerilatedFstSc;
class VerilatedScope;
class VerilatedScopeNameMap;
template <class, class>
class VerilatedTrace;
class VerilatedTraceConfig;
class VerilatedVar;
class VerilatedVarNameMap;
class VerilatedVcd;
@@ -147,7 +151,7 @@ extern uint32_t VL_THREAD_ID() VL_MT_SAFE;
#if VL_THREADED
#define VL_LOCK_SPINS 50000 /// Number of times to spin for a mutex before relaxing
#define VL_LOCK_SPINS 50000 /// Number of times to spin for a mutex before yielding
/// Mutex, wrapped to allow -fthread_safety checks
class VL_CAPABILITY("mutex") VerilatedMutex final {
@@ -252,6 +256,40 @@ public:
#endif
};
//=========================================================================
/// Base class of a Verilator generated (Verilated) model.
///
/// VerilatedModel is a base class of the user facing primary class generated
/// by Verilator.
class VerilatedModel VL_NOT_FINAL {
VL_UNCOPYABLE(VerilatedModel);
VerilatedContext& m_context; // The VerilatedContext this model is instantiated under
protected:
explicit VerilatedModel(VerilatedContext& context);
virtual ~VerilatedModel() = default;
public:
/// Returns the VerilatedContext this model is instantiated under
/// Used to get to e.g. simulation time via contextp()->time()
inline VerilatedContext* contextp() const { return &m_context; }
/// Returns the hierarchical name of this module instance.
virtual const char* hierName() const = 0;
/// Returns the name of this model (the name of the generated model class).
virtual const char* modelName() const = 0;
/// Returns the thread level parallelism, this model was Verilated with. Always 1 or higher.
virtual unsigned threads() const = 0;
private:
// The following are for use by Verilator internals only
template <class, class>
friend class VerilatedTrace;
// Run-time trace configuration requested by this model
virtual std::unique_ptr<VerilatedTraceConfig> traceConfig() const;
};
//=========================================================================
/// Base class for all Verilated module classes.
@@ -266,10 +304,6 @@ public:
const char* name() const { return m_namep; } ///< Return name of module
};
/// Declare a module, ala SC_MODULE
#define VL_MODULE(modname) class modname VL_NOT_FINAL : public VerilatedModule
// Not class final in VL_MODULE, as users might be abstracting our models (--hierarchical)
//=========================================================================
// Functions overridable by user defines
// (Internals however must use VL_PRINTF_MT, which calls these.)
@@ -362,6 +396,16 @@ protected:
// Implementation details
const std::unique_ptr<VerilatedContextImpData> m_impdatap;
// Number of threads to use for simulation (size of m_threadPool + 1 for main thread)
#ifdef VL_THREADED
unsigned m_threads = std::thread::hardware_concurrency();
#else
const unsigned m_threads = 1;
#endif
// The thread pool shared by all models added to this context
std::unique_ptr<VerilatedVirtualBase> m_threadPool;
// The execution profiler shared by all models added to this context
std::unique_ptr<VerilatedVirtualBase> m_executionProfiler;
// Coverage access
std::unique_ptr<VerilatedVirtualBase> m_coveragep; // Pointer for coveragep()
@@ -374,6 +418,10 @@ protected:
// List of free descriptors in the MCT region [4, 32)
std::vector<IData> m_fdFreeMct VL_GUARDED_BY(m_fdMutex);
// Magic to check for bad construction
static constexpr uint64_t MAGIC = 0xC35F9A6E5298EE6EULL; // SHA256 "VerilatedContext"
uint64_t m_magic = MAGIC;
private:
// CONSTRUCTORS
VL_UNCOPYABLE(VerilatedContext);
@@ -491,6 +539,12 @@ public:
/// Get time precision as IEEE-standard text
const char* timeprecisionString() const VL_MT_SAFE;
/// Get number of threads used for simulation (including the main thread)
unsigned threads() const { return m_threads; }
/// Set number of threads used for simulation (including the main thread)
/// Can only be called before the thread pool is created (before first model is added).
void threads(unsigned n);
/// Allow traces to at some point be enabled (disables some optimizations)
void traceEverOn(bool flag) VL_MT_SAFE {
if (flag) calcUnusedSigs(true);
@@ -506,13 +560,20 @@ public:
/// releases - contact the authors before production use.
void scopesDump() const VL_MT_SAFE;
public: // But for internal use only
// METHODS - public but for internal use only
// Internal: access to implementation class
VerilatedContextImp* impp() { return reinterpret_cast<VerilatedContextImp*>(this); }
const VerilatedContextImp* impp() const {
return reinterpret_cast<const VerilatedContextImp*>(this);
}
void addModel(VerilatedModel*);
VerilatedVirtualBase* threadPoolp();
VerilatedVirtualBase*
enableExecutionProfiler(VerilatedVirtualBase* (*construct)(VerilatedContext&));
// Internal: $dumpfile
void dumpfile(const std::string& flag) VL_MT_SAFE_EXCLUDES(m_timeDumpMutex);
std::string dumpfile() const VL_MT_SAFE_EXCLUDES(m_timeDumpMutex);
@@ -535,6 +596,10 @@ public: // But for internal use only
// Internal: Serialization setup
static constexpr size_t serialized1Size() VL_PURE { return sizeof(m_s); }
void* serialized1Ptr() VL_MT_UNSAFE { return &m_s; }
// Internal: Check magic number
static void checkMagic(const VerilatedContext* contextp);
void selfTestClearMagic() { m_magic = 0x2; }
};
//===========================================================================
@@ -822,7 +887,6 @@ public:
}
#endif
public:
// METHODS - INTERNAL USE ONLY (but public due to what uses it)
// Internal: Create a new module name by concatenating two strings
// Returns pointer to thread-local static data (overwritten on next call)
+2 -21
View File
@@ -56,6 +56,7 @@ VK_CPPFLAGS_ALWAYS += \
-DVM_SC=$(VM_SC) \
-DVM_TRACE=$(VM_TRACE) \
-DVM_TRACE_FST=$(VM_TRACE_FST) \
-DVM_TRACE_VCD=$(VM_TRACE_VCD) \
$(CFG_CXXFLAGS_NO_UNUSED) \
ifeq ($(CFG_WITH_CCWARN),yes) # Local... Else don't burden users
@@ -141,26 +142,6 @@ ifneq ($(VM_THREADS),0)
endif
endif
ifneq ($(VM_TRACE_THREADS),0)
ifneq ($(VM_TRACE_THREADS),)
ifeq ($(findstring -DVL_THREADED,$(CPPFLAGS)),)
$(error VM_TRACE_THREADS requires VM_THREADS)
endif
CPPFLAGS += -DVL_TRACE_THREADED
VK_C11=1
VK_LIBS_THREADED=1
endif
endif
ifneq ($(VM_TRACE_FST_WRITER_THREAD),0)
ifneq ($(VM_TRACE_FST_WRITER_THREAD),)
CPPFLAGS += -DVL_TRACE_FST_WRITER_THREAD
VK_C11=1
VK_LIBS_THREADED=1
endif
endif
ifneq ($(VK_C11),0)
ifneq ($(VK_C11),)
# Need C++11 at least, so always default to newest
@@ -186,7 +167,7 @@ VM_SLOW += $(VM_CLASSES_SLOW) $(VM_SUPPORT_SLOW)
VK_FAST_OBJS = $(addsuffix .o, $(VM_FAST))
VK_SLOW_OBJS = $(addsuffix .o, $(VM_SLOW))
VK_USER_OBJS = $(addsuffix .o, $(VM_USER_CLASSES))
VK_USER_OBJS = $(addsuffix .o, $(VM_USER_CLASSES))
# Note VM_GLOBAL_FAST and VM_GLOBAL_SLOW holds the files required from the
# run-time library. In practice everything is actually in VM_GLOBAL_FAST,
+13 -12
View File
@@ -22,8 +22,10 @@
//=============================================================================
#include "verilatedos.h"
#include "verilated.h"
#include "verilated_cov.h"
#include "verilated.h"
#include "verilated_cov_key.h"
#include <deque>
@@ -69,22 +71,23 @@ public: // But only local to this file
// This isn't in the header file for auto-magic conversion because it
// inlines to too much code and makes compilation too slow.
template <class T> class VerilatedCoverItemSpec final : public VerilatedCovImpItem {
template <class T>
class VerilatedCoverItemSpec final : public VerilatedCovImpItem {
private:
// MEMBERS
T* m_countp; // Count value
public:
// METHODS
// cppcheck-suppress truncLongCastReturn
virtual uint64_t count() const override { return *m_countp; }
virtual void zero() const override { *m_countp = 0; }
uint64_t count() const override { return *m_countp; }
void zero() const override { *m_countp = 0; }
// CONSTRUCTORS
// cppcheck-suppress noExplicitConstructor
explicit VerilatedCoverItemSpec(T* countp)
: m_countp{countp} {
*m_countp = 0;
}
virtual ~VerilatedCoverItemSpec() override = default;
~VerilatedCoverItemSpec() override = default;
};
//=============================================================================
@@ -103,7 +106,7 @@ private:
using ItemList = std::deque<VerilatedCovImpItem*>;
// MEMBERS
VerilatedMutex m_mutex; // Protects all members
mutable VerilatedMutex m_mutex; // Protects all members
ValueIndexMap m_valueIndexes VL_GUARDED_BY(m_mutex); // Unique arbitrary value for values
IndexValueMap m_indexValues VL_GUARDED_BY(m_mutex); // Unique arbitrary value for keys
ItemList m_items VL_GUARDED_BY(m_mutex); // List of all items
@@ -122,11 +125,7 @@ public:
protected:
friend class VerilatedCovContext;
virtual ~VerilatedCovImp() override { clearGuts(); }
static VerilatedCovImp& imp() VL_MT_SAFE {
static VerilatedCovImp s_singleton;
return s_singleton;
}
~VerilatedCovImp() override { clearGuts(); }
private:
// PRIVATE METHODS
@@ -209,7 +208,7 @@ private:
// Forward to . so we have a whole word
const std::string suffix = *bpost ? std::string{bpost + 1} : "";
const std::string out = prefix + "*" + suffix;
std::string out = prefix + "*" + suffix;
// cout << "\nch pre="<<prefix<<" s="<<suffix<<"\nch a="<<old<<"\nch b="<<add
// <<"\ncho="<<out<<endl;
@@ -515,8 +514,10 @@ VerilatedCovContext* VerilatedCov::threadCovp() VL_MT_SAFE {
VerilatedCovContext* VerilatedContext::coveragep() VL_MT_SAFE {
static VerilatedMutex s_mutex;
// cppcheck-suppress identicalInnerCondition
if (VL_UNLIKELY(!m_coveragep)) {
const VerilatedLockGuard lock{s_mutex};
// cppcheck-suppress identicalInnerCondition
if (VL_LIKELY(!m_coveragep)) { // Not redundant, prevents race
m_coveragep.reset(new VerilatedCovImp);
}
+7 -4
View File
@@ -26,6 +26,7 @@
#define VERILATOR_VERILATED_COV_H_
#include "verilatedos.h"
#include "verilated.h"
#include <iostream>
@@ -88,7 +89,8 @@ class VerilatedCovImp;
//=============================================================================
// Convert VL_COVER_INSERT value arguments to strings, is \internal
template <class T> std::string vlCovCvtToStr(const T& t) VL_PURE {
template <class T>
std::string vlCovCvtToStr(const T& t) VL_PURE {
std::ostringstream os;
os << t;
return os.str();
@@ -120,7 +122,8 @@ public:
/// Zero coverage points
void zero() VL_MT_SAFE;
public: // But Internal use only
// METHODS - public but Internal use only
// Insert a coverage item
// We accept from 1-30 key/value pairs, all as strings.
// Call _insert1, followed by _insert2 and _insert3
@@ -157,8 +160,8 @@ protected:
friend class VerilatedCovImp;
// CONSTRUCTORS
// Internal: Only made as part of VerilatedCovImp
VerilatedCovContext() {}
virtual ~VerilatedCovContext() {}
VerilatedCovContext() = default;
~VerilatedCovContext() override = default;
// METHODS
// Internal: access to implementation class
+2
View File
@@ -28,7 +28,9 @@
#define VERILATOR_VERILATED_DPI_CPP_
#include "verilatedos.h"
#include "verilated_dpi.h"
#include "verilated_imp.h"
// On MSVC++ we need svdpi.h to declare exports, not imports
+1
View File
@@ -27,6 +27,7 @@
#define VERILATOR_VERILATED_DPI_H_
#include "verilatedos.h"
#include "verilated.h" // Also presumably included by caller
#include "verilated_sym_props.h"
+57 -21
View File
@@ -28,7 +28,7 @@
#include "verilated_fst_c.h"
// GTKWave configuration
#ifdef VL_TRACE_FST_WRITER_THREAD
#ifdef VL_THREADED
# define HAVE_LIBPTHREAD
# define FST_WRITER_PARALLEL
#endif
@@ -83,15 +83,16 @@ static_assert(static_cast<int>(FST_ST_VCD_PROGRAM) == static_cast<int>(VLT_TRACE
//=============================================================================
// Specialization of the generics for this trace format
#define VL_DERIVED_T VerilatedFst
#include "verilated_trace_imp.cpp"
#undef VL_DERIVED_T
#define VL_SUB_T VerilatedFst
#define VL_BUF_T VerilatedFstBuffer
#include "verilated_trace_imp.h"
#undef VL_SUB_T
#undef VL_BUF_T
//=============================================================================
// VerilatedFst
VerilatedFst::VerilatedFst(void* fst)
: m_fst{fst} {}
VerilatedFst::VerilatedFst(void* /*fst*/) {}
VerilatedFst::~VerilatedFst() {
if (m_fst) fstWriterClose(m_fst);
@@ -104,14 +105,12 @@ void VerilatedFst::open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex) {
m_fst = fstWriterCreate(filename, 1);
fstWriterSetPackType(m_fst, FST_WR_PT_LZ4);
fstWriterSetTimescaleFromString(m_fst, timeResStr().c_str()); // lintok-begin-on-ref
#ifdef VL_TRACE_FST_WRITER_THREAD
fstWriterSetParallelMode(m_fst, 1);
#endif
if (m_useFstWriterThread) fstWriterSetParallelMode(m_fst, 1);
fullDump(true); // First dump must be full for fst
m_curScope.clear();
VerilatedTrace<VerilatedFst>::traceInit();
Super::traceInit();
// Clear the scope stack
auto it = m_curScope.begin();
@@ -133,14 +132,14 @@ void VerilatedFst::open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex) {
void VerilatedFst::close() VL_MT_SAFE_EXCLUDES(m_mutex) {
const VerilatedLockGuard lock{m_mutex};
VerilatedTrace<VerilatedFst>::closeBase();
Super::closeBase();
fstWriterClose(m_fst);
m_fst = nullptr;
}
void VerilatedFst::flush() VL_MT_SAFE_EXCLUDES(m_mutex) {
const VerilatedLockGuard lock{m_mutex};
VerilatedTrace<VerilatedFst>::flushBase();
Super::flushBase();
fstWriterFlushContext(m_fst);
}
@@ -162,7 +161,7 @@ void VerilatedFst::declare(uint32_t code, const char* name, int dtypenum, fstVar
int lsb) {
const int bits = ((msb > lsb) ? (msb - lsb) : (lsb - msb)) + 1;
const bool enabled = VerilatedTrace<VerilatedFst>::declCode(code, name, bits, false);
const bool enabled = Super::declCode(code, name, bits, false);
if (!enabled) return;
std::string nameasstr = namePrefix() + name;
@@ -245,18 +244,55 @@ void VerilatedFst::declDouble(uint32_t code, const char* name, int dtypenum, fst
declare(code, name, dtypenum, vardir, vartype, array, arraynum, false, 63, 0);
}
//=============================================================================
// Get/commit trace buffer
VerilatedFst::Buffer* VerilatedFst::getTraceBuffer() {
#ifdef VL_THREADED
if (offload()) return new OffloadBuffer{*this};
#endif
return new Buffer{*this};
}
void VerilatedFst::commitTraceBuffer(VerilatedFst::Buffer* bufp) {
#ifdef VL_THREADED
if (offload()) {
OffloadBuffer* const offloadBufferp = static_cast<OffloadBuffer*>(bufp);
if (offloadBufferp->m_offloadBufferWritep) {
m_offloadBufferWritep = offloadBufferp->m_offloadBufferWritep;
return; // Buffer will be deleted by the offload thread
}
}
#endif
delete bufp;
}
//=============================================================================
// Configure
void VerilatedFst::configure(const VerilatedTraceConfig& config) {
// If at least one model requests the FST writer thread, then use it
m_useFstWriterThread |= config.m_useFstWriterThread;
}
//=============================================================================
// VerilatedFstBuffer implementation
//=============================================================================
// Trace rendering primitives
// Note: emit* are only ever called from one place (full* in
// verilated_trace_imp.cpp, which is included in this file at the top),
// verilated_trace_imp.h, which is included in this file at the top),
// so always inline them.
VL_ATTR_ALWINLINE
void VerilatedFst::emitBit(uint32_t code, CData newval) {
void VerilatedFstBuffer::emitBit(uint32_t code, CData newval) {
VL_DEBUG_IFDEF(assert(m_symbolp[code]););
fstWriterEmitValueChange(m_fst, m_symbolp[code], newval ? "1" : "0");
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitCData(uint32_t code, CData newval, int bits) {
void VerilatedFstBuffer::emitCData(uint32_t code, CData newval, int bits) {
char buf[VL_BYTESIZE];
VL_DEBUG_IFDEF(assert(m_symbolp[code]););
cvtCDataToStr(buf, newval << (VL_BYTESIZE - bits));
@@ -264,7 +300,7 @@ void VerilatedFst::emitCData(uint32_t code, CData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitSData(uint32_t code, SData newval, int bits) {
void VerilatedFstBuffer::emitSData(uint32_t code, SData newval, int bits) {
char buf[VL_SHORTSIZE];
VL_DEBUG_IFDEF(assert(m_symbolp[code]););
cvtSDataToStr(buf, newval << (VL_SHORTSIZE - bits));
@@ -272,7 +308,7 @@ void VerilatedFst::emitSData(uint32_t code, SData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitIData(uint32_t code, IData newval, int bits) {
void VerilatedFstBuffer::emitIData(uint32_t code, IData newval, int bits) {
char buf[VL_IDATASIZE];
VL_DEBUG_IFDEF(assert(m_symbolp[code]););
cvtIDataToStr(buf, newval << (VL_IDATASIZE - bits));
@@ -280,7 +316,7 @@ void VerilatedFst::emitIData(uint32_t code, IData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitQData(uint32_t code, QData newval, int bits) {
void VerilatedFstBuffer::emitQData(uint32_t code, QData newval, int bits) {
char buf[VL_QUADSIZE];
VL_DEBUG_IFDEF(assert(m_symbolp[code]););
cvtQDataToStr(buf, newval << (VL_QUADSIZE - bits));
@@ -288,7 +324,7 @@ void VerilatedFst::emitQData(uint32_t code, QData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitWData(uint32_t code, const WData* newvalp, int bits) {
void VerilatedFstBuffer::emitWData(uint32_t code, const WData* newvalp, int bits) {
int words = VL_WORDS_I(bits);
char* wp = m_strbuf;
// Convert the most significant word
@@ -304,6 +340,6 @@ void VerilatedFst::emitWData(uint32_t code, const WData* newvalp, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedFst::emitDouble(uint32_t code, double newval) {
void VerilatedFstBuffer::emitDouble(uint32_t code, double newval) {
fstWriterEmitValueChange(m_fst, m_symbolp[code], &newval);
}
+75 -29
View File
@@ -31,26 +31,32 @@
#include <string>
#include <vector>
class VerilatedFstBuffer;
//=============================================================================
// VerilatedFst
// Base class to create a Verilator FST dump
// This is an internally used class - see VerilatedFstC for what to call from applications
class VerilatedFst final : public VerilatedTrace<VerilatedFst> {
class VerilatedFst final : public VerilatedTrace<VerilatedFst, VerilatedFstBuffer> {
public:
using Super = VerilatedTrace<VerilatedFst, VerilatedFstBuffer>;
private:
// Give the superclass access to private bits (to avoid virtual functions)
friend class VerilatedTrace<VerilatedFst>;
friend VerilatedFstBuffer; // Give the buffer access to the private bits
//=========================================================================
// FST specific internals
void* m_fst;
void* m_fst = nullptr;
std::map<uint32_t, fstHandle> m_code2symbol;
std::map<int, fstEnumHandle> m_local2fstdtype;
std::list<std::string> m_curScope;
fstHandle* m_symbolp = nullptr; // same as m_code2symbol, but as an array
char* m_strbuf = nullptr; // String buffer long enough to hold maxBits() chars
bool m_useFstWriterThread = false; // Whether to use the separate FST writer thread
// CONSTRUCTORS
VL_UNCOPYABLE(VerilatedFst);
void declare(uint32_t code, const char* name, int dtypenum, fstVarDir vardir,
@@ -60,31 +66,29 @@ protected:
//=========================================================================
// Implementation of VerilatedTrace interface
// Implementations of protected virtual methods for VerilatedTrace
virtual void emitTimeChange(uint64_t timeui) override;
// Called when the trace moves forward to a new time point
void emitTimeChange(uint64_t timeui) override;
// Hooks called from VerilatedTrace
virtual bool preFullDump() override { return isOpen(); }
virtual bool preChangeDump() override { return isOpen(); }
bool preFullDump() override { return isOpen(); }
bool preChangeDump() override { return isOpen(); }
// Implementations of duck-typed methods for VerilatedTrace. These are
// called from only one place (namely full*) so always inline them.
inline void emitBit(uint32_t code, CData newval);
inline void emitCData(uint32_t code, CData newval, int bits);
inline void emitSData(uint32_t code, SData newval, int bits);
inline void emitIData(uint32_t code, IData newval, int bits);
inline void emitQData(uint32_t code, QData newval, int bits);
inline void emitWData(uint32_t code, const WData* newvalp, int bits);
inline void emitDouble(uint32_t code, double newval);
// Trace buffer management
Buffer* getTraceBuffer() override;
void commitTraceBuffer(Buffer*) override;
// Configure sub-class
void configure(const VerilatedTraceConfig&) override;
public:
//=========================================================================
// External interface to client code
// (All must be threadsafe)
// CONSTRUCTOR
explicit VerilatedFst(void* fst = nullptr);
~VerilatedFst();
// METHODS - All must be thread safe
// Open the file; call isOpen() to see if errors
void open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex);
// Close the file
@@ -97,11 +101,6 @@ public:
//=========================================================================
// Internal interface to Verilator generated code
// Inside dumping routines, declare a data type
void declDTypeEnum(int dtypenum, const char* name, uint32_t elements, unsigned int minValbits,
const char** itemNamesp, const char** itemValuesp);
// Inside dumping routines, declare a signal
void declBit(uint32_t code, const char* name, int dtypenum, fstVarDir vardir,
fstVarType vartype, bool array, int arraynum);
void declBus(uint32_t code, const char* name, int dtypenum, fstVarDir vardir,
@@ -112,18 +111,65 @@ public:
fstVarType vartype, bool array, int arraynum, int msb, int lsb);
void declDouble(uint32_t code, const char* name, int dtypenum, fstVarDir vardir,
fstVarType vartype, bool array, int arraynum);
void declDTypeEnum(int dtypenum, const char* name, uint32_t elements, unsigned int minValbits,
const char** itemNamesp, const char** itemValuesp);
};
#ifndef DOXYGEN
// Declare specialization here as it's used in VerilatedFstC just below
template <> void VerilatedTrace<VerilatedFst>::dump(uint64_t timeui);
template <> void VerilatedTrace<VerilatedFst>::set_time_unit(const char* unitp);
template <> void VerilatedTrace<VerilatedFst>::set_time_unit(const std::string& unit);
template <> void VerilatedTrace<VerilatedFst>::set_time_resolution(const char* unitp);
template <> void VerilatedTrace<VerilatedFst>::set_time_resolution(const std::string& unit);
template <> void VerilatedTrace<VerilatedFst>::dumpvars(int level, const std::string& hier);
template <>
void VerilatedFst::Super::dump(uint64_t time);
template <>
void VerilatedFst::Super::set_time_unit(const char* unitp);
template <>
void VerilatedFst::Super::set_time_unit(const std::string& unit);
template <>
void VerilatedFst::Super::set_time_resolution(const char* unitp);
template <>
void VerilatedFst::Super::set_time_resolution(const std::string& unit);
template <>
void VerilatedFst::Super::dumpvars(int level, const std::string& hier);
#endif
//=============================================================================
// VerilatedFstBuffer
class VerilatedFstBuffer VL_NOT_FINAL {
// Give the trace file access to the private bits
friend VerilatedFst;
friend VerilatedFst::Super;
friend VerilatedFst::Buffer;
friend VerilatedFst::OffloadBuffer;
VerilatedFst& m_owner; // Trace file owning this buffer. Required by subclasses.
// The FST file handle
void* const m_fst = m_owner.m_fst;
// code to fstHande map, as an array
const fstHandle* const m_symbolp = m_owner.m_symbolp;
// String buffer long enough to hold maxBits() chars
char* const m_strbuf = m_owner.m_strbuf;
// CONSTRUCTOR
explicit VerilatedFstBuffer(VerilatedFst& owner)
: m_owner{owner} {}
virtual ~VerilatedFstBuffer() = default;
//=========================================================================
// Implementation of VerilatedTraceBuffer interface
// Implementations of duck-typed methods for VerilatedTraceBuffer. These are
// called from only one place (the full* methods), so always inline them.
VL_ATTR_ALWINLINE inline void emitBit(uint32_t code, CData newval);
VL_ATTR_ALWINLINE inline void emitCData(uint32_t code, CData newval, int bits);
VL_ATTR_ALWINLINE inline void emitSData(uint32_t code, SData newval, int bits);
VL_ATTR_ALWINLINE inline void emitIData(uint32_t code, IData newval, int bits);
VL_ATTR_ALWINLINE inline void emitQData(uint32_t code, QData newval, int bits);
VL_ATTR_ALWINLINE inline void emitWData(uint32_t code, const WData* newvalp, int bits);
VL_ATTR_ALWINLINE inline void emitDouble(uint32_t code, double newval);
};
//=============================================================================
// VerilatedFstC
/// Create a FST dump file in C standalone (no SystemC) simulations.
+2 -4
View File
@@ -22,6 +22,7 @@
//=============================================================================
#include "verilatedos.h"
#include "verilated_fst_sc.h"
//======================================================================
@@ -30,7 +31,7 @@
void VerilatedFstSc::open(const char* filename) {
if (!sc_core::sc_get_curr_simcontext()->elaboration_done()) {
vl_fatal(__FILE__, __LINE__, "VerilatedFstSc",
("%Error: VerilatedFstSc::open(\"" + std::string(filename)
("%Error: VerilatedFstSc::open(\"" + std::string{filename}
+ "\") is called before sc_core::sc_start(). "
"Run sc_core::sc_start(sc_core::SC_ZERO_TIME) before opening a wave file.")
.c_str());
@@ -63,9 +64,6 @@ void VerilatedFstSc::trace(const unsigned int&, const std::string&, const char**
DECL_TRACE_METHOD_B( unsigned short )
DECL_TRACE_METHOD_B( unsigned int )
DECL_TRACE_METHOD_B( unsigned long )
#ifdef SYSTEMC_64BIT_PATCHES
DECL_TRACE_METHOD_B( unsigned long long)
#endif
DECL_TRACE_METHOD_B( char )
DECL_TRACE_METHOD_B( short )
DECL_TRACE_METHOD_B( int )
+4 -5
View File
@@ -23,8 +23,9 @@
#define _VERILATED_FST_SC_H_ 1
#include "verilatedos.h"
#include "verilated_sc.h"
#include "verilated_fst_c.h"
#include "verilated_sc.h"
#include <string>
@@ -51,7 +52,7 @@ public:
}
spTrace()->set_time_resolution(sc_get_time_resolution().to_string());
}
virtual ~VerilatedFstSc() { close(); }
virtual ~VerilatedFstSc() /*override*/ { close(); }
// METHODS
/// Called by SystemC simulate()
@@ -60,6 +61,7 @@ public:
}
// Override VerilatedFstC. Must be called after starting simulation.
// cppcheck-suppress missingOverride // GCC won't accept override
virtual void open(const char* filename) /*override*/ VL_MT_SAFE;
private:
@@ -95,9 +97,6 @@ private:
DECL_TRACE_METHOD_B( unsigned short )
DECL_TRACE_METHOD_B( unsigned int )
DECL_TRACE_METHOD_B( unsigned long )
#ifdef SYSTEMC_64BIT_PATCHES
DECL_TRACE_METHOD_B( unsigned long long)
#endif
DECL_TRACE_METHOD_B( char )
DECL_TRACE_METHOD_B( short )
DECL_TRACE_METHOD_B( int )
+1 -19
View File
@@ -146,7 +146,7 @@ extern IData VL_SYSTEM_IW(int lhswords, WDataInP const lhsp);
extern IData VL_SYSTEM_IQ(QData lhs);
inline IData VL_SYSTEM_II(IData lhs) VL_MT_SAFE { return VL_SYSTEM_IQ(lhs); }
extern IData VL_TESTPLUSARGS_I(const char* formatp);
extern IData VL_TESTPLUSARGS_I(const std::string& format);
extern const char* vl_mc_scan_plusargs(const char* prefixp); // PLIish
//=========================================================================
@@ -256,25 +256,7 @@ extern void _vl_debug_print_w(int lbits, WDataInP const iwp);
//=========================================================================
// Pli macros
extern int VL_TIME_STR_CONVERT(const char* strp) VL_PURE;
// These are deprecated and used only to establish the default precision/units.
// Use Verilator timescale-override for better control.
// clang-format off
#ifndef VL_TIME_PRECISION
# ifdef VL_TIME_PRECISION_STR
# define VL_TIME_PRECISION VL_TIME_STR_CONVERT(VL_STRINGIFY(VL_TIME_PRECISION_STR))
# else
# define VL_TIME_PRECISION (-12) ///< Timescale default units if not in Verilog - picoseconds
# endif
#endif
#ifndef VL_TIME_UNIT
# ifdef VL_TIME_UNIT_STR
# define VL_TIME_UNIT VL_TIME_STR_CONVERT(VL_STRINGIFY(VL_TIME_PRECISION_STR))
# else
# define VL_TIME_UNIT (-12) ///< Timescale default units if not in Verilog - picoseconds
# endif
#endif
#if defined(SYSTEMC_VERSION)
/// Return current simulation time
+17 -22
View File
@@ -91,7 +91,7 @@ class VerilatedEvalMsgQueue final {
std::atomic<uint64_t> m_depth; // Current depth of queue (see comments below)
VerilatedMutex m_mutex; // Mutex protecting queue
mutable VerilatedMutex m_mutex; // Mutex protecting queue
VerilatedThreadQueue m_queue VL_GUARDED_BY(m_mutex); // Message queue
public:
// CONSTRUCTORS
@@ -143,11 +143,10 @@ class VerilatedThreadMsgQueue final {
public:
// CONSTRUCTORS
VerilatedThreadMsgQueue() {}
~VerilatedThreadMsgQueue() {
// The only call of this with a non-empty queue is a fatal error.
// So this does not flush the queue, as the destination queue is not known to this class.
}
VerilatedThreadMsgQueue() = default;
~VerilatedThreadMsgQueue() = default;
// The only call of destructor with a non-empty queue is a fatal error.
// So this does not flush the queue, as the destination queue is not known to this class.
private:
VL_UNCOPYABLE(VerilatedThreadMsgQueue);
@@ -188,11 +187,11 @@ class VerilatedFpList final {
public:
using const_iterator = FILE* const*;
explicit VerilatedFpList() {}
explicit VerilatedFpList() = default;
const_iterator begin() const { return m_fp; }
const_iterator end() const { return m_fp + m_sz; }
std::size_t size() const { return m_sz; }
std::size_t capacity() const { return 31; }
static std::size_t capacity() { return 31; }
void push_back(FILE* fd) {
if (VL_LIKELY(size() < capacity())) m_fp[m_sz++] = fd;
}
@@ -236,12 +235,11 @@ class VerilatedContextImp final : VerilatedContext {
return s_s;
}
private:
public: // But only for verilated*.cpp
// CONSTRUCTORS - no data can live here, use only VerilatedContext
VerilatedContextImp() = delete;
~VerilatedContextImp() = delete;
public: // But only for verilated*.cpp
// METHODS - extending into VerilatedContext, call via impp()->
// Random seed handling
@@ -272,13 +270,12 @@ public: // But only for verilated*.cpp
std::string argPlusMatch(const char* prefixp) VL_MT_SAFE_EXCLUDES(m_argMutex);
std::pair<int, char**> argc_argv() VL_MT_SAFE_EXCLUDES(m_argMutex);
public: // But only for verilated*.cpp
// METHODS - scope name
// METHODS - scope name - INTERNAL only for verilated*.cpp
void scopeInsert(const VerilatedScope* scopep) VL_MT_SAFE;
void scopeErase(const VerilatedScope* scopep) VL_MT_SAFE;
public: // But only for verilated*.cpp
// METHODS - file IO
// METHODS - file IO - INTERNAL only for verilated*.cpp
IData fdNewMcd(const char* filenamep) VL_MT_SAFE_EXCLUDES(m_fdMutex) {
const VerilatedLockGuard lock{m_fdMutex};
if (m_fdFreeMct.empty()) return 0;
@@ -298,7 +295,7 @@ public: // But only for verilated*.cpp
const std::size_t start = std::max<std::size_t>(31UL + 1UL + 3UL, m_fdps.size());
const std::size_t excess = 10;
m_fdps.resize(start + excess);
std::fill(m_fdps.begin() + start, m_fdps.end(), (FILE*)0);
std::fill(m_fdps.begin() + start, m_fdps.end(), static_cast<FILE*>(nullptr));
m_fdFree.resize(excess);
for (std::size_t i = 0, id = start; i < m_fdFree.size(); ++i, ++id) {
m_fdFree[i] = id;
@@ -344,7 +341,7 @@ public: // But only for verilated*.cpp
if (VL_UNLIKELY(idx <= 2)) return; // stdout/stdin/stderr
if (VL_UNLIKELY(!m_fdps[idx])) return; // Already free
std::fclose(m_fdps[idx]);
m_fdps[idx] = (FILE*)0;
m_fdps[idx] = nullptr;
m_fdFree.push_back(idx);
} else {
// MCD case
@@ -464,7 +461,6 @@ public:
// METHODS - debug
static void versionDump() VL_MT_SAFE;
public:
// METHODS - user scope tracking
// We implement this as a single large map instead of one map per scope.
// There's often many more scopes than userdata's and thus having a ~48byte
@@ -486,7 +482,8 @@ public:
return it->second;
}
public: // But only for verilated.cpp
// METHODS - But only for verilated.cpp
// Symbol table destruction cleans up the entries for each scope.
static void userEraseScope(const VerilatedScope* scopep) VL_MT_SAFE {
// Slow ok - called once/scope on destruction, so we only iterate.
@@ -512,8 +509,7 @@ public: // But only for verilated.cpp
}
}
public: // But only for verilated*.cpp
// METHODS - hierarchy
// METHODS - hierarchy - only for verilated*.cpp
static void hierarchyAdd(const VerilatedScope* fromp, const VerilatedScope* top) VL_MT_SAFE {
// Slow ok - called at construction for VPI accessible elements
const VerilatedLockGuard lock{s().m_hierMapMutex};
@@ -534,8 +530,7 @@ public: // But only for verilated*.cpp
return &s().m_hierMap;
}
public: // But only for verilated*.cpp
// METHODS - export names
// METHODS - export names - only for verilated*.cpp
// Each function prototype is converted to a function number which we
// then use to index a 2D table also indexed by scope number, because we
+44 -10
View File
@@ -17,6 +17,7 @@
//=============================================================================
#include "verilatedos.h"
#include "verilated_profiler.h"
#if VL_THREADED
@@ -60,54 +61,87 @@ uint16_t VlExecutionRecord::getcpu() {
//=============================================================================
// VlExecutionProfiler implementation
template <size_t N> size_t roundUptoMultipleOf(size_t value) {
template <size_t N>
static size_t roundUptoMultipleOf(size_t value) {
static_assert((N & (N - 1)) == 0, "'N' must be a power of 2");
size_t mask = N - 1;
return (value + mask) & ~mask;
}
VlExecutionProfiler::VlExecutionProfiler() {
VlExecutionProfiler::VlExecutionProfiler(VerilatedContext& context)
: m_context{context} {
// Setup profiling on main thread
setupThread(0);
}
void VlExecutionProfiler::configure(const VerilatedContext& context) {
void VlExecutionProfiler::configure() {
if (VL_UNLIKELY(m_enabled)) {
--m_windowCount;
if (VL_UNLIKELY(m_windowCount == context.profExecWindow())) {
if (VL_UNLIKELY(m_windowCount == m_context.profExecWindow())) {
VL_DEBUG_IF(VL_DBG_MSGF("+ profile start collection\n"););
clear(); // Clear the profile after the cache warm-up cycles.
m_tickBegin = VL_CPU_TICK();
} else if (VL_UNLIKELY(m_windowCount == 0)) {
const uint64_t tickEnd = VL_CPU_TICK();
VL_DEBUG_IF(VL_DBG_MSGF("+ profile end\n"););
const std::string& fileName = context.profExecFilename();
const std::string& fileName = m_context.profExecFilename();
dump(fileName.c_str(), tickEnd);
m_enabled = false;
}
return;
}
const uint64_t startReq = context.profExecStart() + 1; // + 1, so we can start at time 0
const uint64_t startReq = m_context.profExecStart() + 1; // + 1, so we can start at time 0
if (VL_UNLIKELY(m_lastStartReq < startReq && VL_TIME_Q() >= context.profExecStart())) {
if (VL_UNLIKELY(m_lastStartReq < startReq && VL_TIME_Q() >= m_context.profExecStart())) {
VL_DEBUG_IF(VL_DBG_MSGF("+ profile start warmup\n"););
VL_DEBUG_IF(assert(m_windowCount == 0););
m_enabled = true;
m_windowCount = context.profExecWindow() * 2;
m_windowCount = m_context.profExecWindow() * 2;
m_lastStartReq = startReq;
}
}
VerilatedVirtualBase* VlExecutionProfiler::construct(VerilatedContext& context) {
VlExecutionProfiler* const selfp = new VlExecutionProfiler{context};
#if VL_THREADED
if (VlThreadPool* const threadPoolp = static_cast<VlThreadPool*>(context.threadPoolp())) {
for (int i = 0; i < threadPoolp->numThreads(); ++i) {
// Data to pass to worker thread initialization
struct Data {
VlExecutionProfiler* const selfp;
const uint32_t threadId;
} data{selfp, static_cast<uint32_t>(i + 1)};
// Initialize worker thread
threadPoolp->workerp(i)->addTask(
[](void* userp, bool) {
Data* const datap = static_cast<Data*>(userp);
datap->selfp->setupThread(datap->threadId);
},
&data);
// Wait until initializationis complete
threadPoolp->workerp(i)->wait();
}
}
#endif
return selfp;
}
void VlExecutionProfiler::setupThread(uint32_t threadId) {
// Reserve some space in the thread-local profiling buffer, in order to try to avoid malloc
// while profiling.
t_trace.reserve(RESERVED_TRACE_CAPACITY);
// Register thread-local buffer in list of all buffers
bool exists;
{
const VerilatedLockGuard lock{m_mutex};
bool exists = !m_traceps.emplace(threadId, &t_trace).second;
assert(!exists);
exists = !m_traceps.emplace(threadId, &t_trace).second;
}
if (VL_UNLIKELY(exists)) {
VL_FATAL_MT(__FILE__, __LINE__, "", "multiple initialization of profiler on some thread");
}
}
+16 -13
View File
@@ -24,10 +24,6 @@
#include "verilatedos.h"
#ifndef VL_PROFILER
#error "verilated_profiler.h/cpp expects VL_PROFILER (from --prof-{exec, pgo}"
#endif
#include "verilated.h"
#include <array>
@@ -38,13 +34,14 @@
#include <vector>
class VlExecutionProfiler;
class VlThreadPool;
//=============================================================================
// Macros to simplify generated code
#define VL_EXEC_TRACE_ADD_RECORD(vlSymsp) \
if (VL_UNLIKELY((vlSymsp)->__Vm_executionProfiler.enabled())) \
(vlSymsp)->__Vm_executionProfiler.addRecord()
if (VL_UNLIKELY((vlSymsp)->__Vm_executionProfilerp->enabled())) \
(vlSymsp)->__Vm_executionProfilerp->addRecord()
//=============================================================================
// Return high-precision counter for profiling, or 0x0 if not available
@@ -136,7 +133,7 @@ static_assert(std::is_trivially_destructible<VlExecutionRecord>::value,
//=============================================================================
// VlExecutionProfiler is for collecting profiling data about model execution
class VlExecutionProfiler final {
class VlExecutionProfiler final : public VerilatedVirtualBase {
// CONSTANTS
// In order to try to avoid dynamic memory allocations during the actual profiling phase,
@@ -154,8 +151,9 @@ class VlExecutionProfiler final {
using ExecutionTrace = std::vector<VlExecutionRecord>;
// STATE
VerilatedContext& m_context; // The context this profiler is under
static VL_THREAD_LOCAL ExecutionTrace t_trace; // thread-local trace buffers
VerilatedMutex m_mutex;
mutable VerilatedMutex m_mutex;
// Map from thread id to &t_trace of given thread
std::map<uint32_t, ExecutionTrace*> m_traceps VL_GUARDED_BY(m_mutex);
@@ -167,31 +165,36 @@ class VlExecutionProfiler final {
public:
// CONSTRUCTOR
VlExecutionProfiler();
explicit VlExecutionProfiler(VerilatedContext& context);
~VlExecutionProfiler() override = default;
// METHODS
// Is profiling enabled
inline bool enabled() const { return m_enabled; }
// Append a trace record to the trace buffer of the current thread
inline VlExecutionRecord& addRecord() {
static inline VlExecutionRecord& addRecord() {
t_trace.emplace_back();
return t_trace.back();
}
// Configure profiler (called in beginning of 'eval')
void configure(const VerilatedContext&);
void configure();
// Setup profiling on a particular thread;
void setupThread(uint32_t threadId);
// Clear all profiling data
void clear() VL_MT_SAFE_EXCLUDES(m_mutex);
// Write profiling data into file
void dump(const char* filenamep, uint64_t tickEnd) VL_MT_SAFE_EXCLUDES(m_mutex);
// Passed to VerilatedContext to create the VlExecutionProfiler profiler instance
static VerilatedVirtualBase* construct(VerilatedContext& context);
};
//=============================================================================
// VlPgoProfiler is for collecting profiling data for PGO
template <std::size_t T_Entries> class VlPgoProfiler final {
template <std::size_t T_Entries>
class VlPgoProfiler final {
// TYPES
struct Record final {
const std::string m_name; // Hashed name of mtask/etc
@@ -228,7 +231,7 @@ void VlPgoProfiler<T_Entries>::write(const char* modelp, const std::string& file
// So when we have multiple models in an executable, possibly even
// running on different threads, each will have a different symtab so
// each will collect is own data correctly. However when each is
// destroid we need to get all the data, not keep overwriting and only
// destroyed we need to get all the data, not keep overwriting and only
// get the last model's data.
static bool s_firstCall = true;
+6 -4
View File
@@ -24,8 +24,10 @@
#define VERILATOR_VERILATED_SAVE_CPP_
#include "verilatedos.h"
#include "verilated.h"
#include "verilated_save.h"
#include "verilated.h"
#include "verilated_imp.h"
#include <cerrno>
@@ -38,13 +40,13 @@
# include <unistd.h>
#endif
#ifndef O_LARGEFILE // For example on WIN32
#ifndef O_LARGEFILE // WIN32 headers omit this
# define O_LARGEFILE 0
#endif
#ifndef O_NONBLOCK
#ifndef O_NONBLOCK // WIN32 headers omit this
# define O_NONBLOCK 0
#endif
#ifndef O_CLOEXEC
#ifndef O_CLOEXEC // WIN32 headers omit this
# define O_CLOEXEC 0
#endif
// clang-format on
+11 -9
View File
@@ -23,6 +23,7 @@
#define VERILATOR_VERILATED_SAVE_C_H_
#include "verilatedos.h"
#include "verilated.h"
#include <string>
@@ -77,7 +78,7 @@ public:
virtual void flush() VL_MT_UNSAFE_ONE {}
/// Write data to stream
VerilatedSerialize& write(const void* __restrict datap, size_t size) VL_MT_UNSAFE_ONE {
const uint8_t* __restrict dp = (const uint8_t* __restrict)datap;
const uint8_t* __restrict dp = static_cast<const uint8_t* __restrict>(datap);
while (size) {
bufferCheck();
size_t blk = size;
@@ -194,16 +195,16 @@ public:
/// Construct new object
VerilatedSave() = default;
/// Flush, close and destruct
virtual ~VerilatedSave() override { close(); }
~VerilatedSave() override { close(); }
// METHODS
/// Open the file; call isOpen() to see if errors
void open(const char* filenamep) VL_MT_UNSAFE_ONE;
/// Open the file; call isOpen() to see if errors
void open(const std::string& filename) VL_MT_UNSAFE_ONE { open(filename.c_str()); }
/// Flush and close the file
virtual void close() override VL_MT_UNSAFE_ONE;
void close() override VL_MT_UNSAFE_ONE;
/// Flush data to file
virtual void flush() override VL_MT_UNSAFE_ONE;
void flush() override VL_MT_UNSAFE_ONE;
};
//=============================================================================
@@ -221,7 +222,7 @@ public:
/// Construct new object
VerilatedRestore() = default;
/// Flush, close and destruct
virtual ~VerilatedRestore() override { close(); }
~VerilatedRestore() override { close(); }
// METHODS
/// Open the file; call isOpen() to see if errors
@@ -229,9 +230,9 @@ public:
/// Open the file; call isOpen() to see if errors
void open(const std::string& filename) VL_MT_UNSAFE_ONE { open(filename.c_str()); }
/// Close the file
virtual void close() override VL_MT_UNSAFE_ONE;
virtual void flush() override VL_MT_UNSAFE_ONE {}
virtual void fill() override VL_MT_UNSAFE_ONE;
void close() override VL_MT_UNSAFE_ONE;
void flush() override VL_MT_UNSAFE_ONE {}
void fill() override VL_MT_UNSAFE_ONE;
};
//=============================================================================
@@ -287,7 +288,8 @@ inline VerilatedDeserialize& operator>>(VerilatedDeserialize& os, std::string& r
uint32_t len = 0;
os >> len;
rhs.resize(len);
return os.read((void*)rhs.data(), len);
// C cast is required below
return os.read((void*)(rhs.data()), len);
}
VerilatedSerialize& operator<<(VerilatedSerialize& os, VerilatedContext* rhsp);
VerilatedDeserialize& operator>>(VerilatedDeserialize& os, VerilatedContext* rhsp);
-1
View File
@@ -134,7 +134,6 @@ public:
initUnpacked(ulims);
}
public:
~VerilatedVarProps() = default;
// METHODS
bool magicOk() const { return m_magic == MAGIC; }
+1
View File
@@ -30,6 +30,7 @@
#define VERILATOR_VERILATED_SYMS_H_
#include "verilatedos.h"
#include "verilated.h"
#include "verilated_sym_props.h"
+34 -46
View File
@@ -22,11 +22,8 @@
//=============================================================================
#include "verilatedos.h"
#include "verilated_threads.h"
#ifdef VL_PROFILER
#include "verilated_profiler.h"
#endif
#include "verilated_threads.h"
#include <cstdio>
#include <memory>
@@ -51,68 +48,59 @@ VlMTaskVertex::VlMTaskVertex(uint32_t upstreamDepCount)
//=============================================================================
// VlWorkerThread
VlWorkerThread::VlWorkerThread(uint32_t threadId, VerilatedContext* contextp,
VlExecutionProfiler* profilerp)
VlWorkerThread::VlWorkerThread(VerilatedContext* contextp)
: m_ready_size{0}
, m_exiting{false}
, m_cthread{startWorker, this, threadId, profilerp}
, m_contextp{contextp} {}
, m_cthread{startWorker, this, contextp} {}
VlWorkerThread::~VlWorkerThread() {
m_exiting.store(true, std::memory_order_release);
wakeUp();
shutdown();
// The thread should exit; join it.
m_cthread.join();
}
static void shutdownTask(void*, bool) {
// Deliberately empty, we use the address of this function as a magic number
}
void VlWorkerThread::shutdown() { addTask(shutdownTask, nullptr); }
void VlWorkerThread::wait() {
// Enqueue a task that sets this flag. Execution is in-order so this ensures completion.
std::atomic<bool> flag{false};
addTask([](void* flagp, bool) { static_cast<std::atomic<bool>*>(flagp)->store(true); }, &flag);
// Spin wait
for (unsigned i = 0; i < VL_LOCK_SPINS; ++i) {
if (flag.load()) return;
VL_CPU_RELAX();
}
// Yield wait
while (!flag.load()) std::this_thread::yield();
}
void VlWorkerThread::workerLoop() {
ExecRec work;
work.m_fnp = nullptr;
// Wait for the first task without spinning, in case the thread is never actually used.
dequeWork</* SpinWait: */ false>(&work);
while (true) {
if (VL_LIKELY(!work.m_fnp)) dequeWork(&work);
// Do this here, not above, to avoid a race with the destructor.
if (VL_UNLIKELY(m_exiting.load(std::memory_order_acquire))) break;
if (VL_LIKELY(work.m_fnp)) {
work.m_fnp(work.m_selfp, work.m_evenCycle);
work.m_fnp = nullptr;
}
if (VL_UNLIKELY(work.m_fnp == shutdownTask)) break;
work.m_fnp(work.m_selfp, work.m_evenCycle);
// Wait for next task with spinning.
dequeWork</* SpinWait: */ true>(&work);
}
}
void VlWorkerThread::startWorker(VlWorkerThread* workerp, uint32_t threadId,
VlExecutionProfiler* profilerp) {
Verilated::threadContextp(workerp->m_contextp);
#ifdef VL_PROFILER
// Note: setupThread is not defined without VL_PROFILER, hence the #ifdef. Still, we might
// not be profiling execution (e.g.: PGO only), so profilerp might still be nullptr.
if (profilerp) profilerp->setupThread(threadId);
#endif
void VlWorkerThread::startWorker(VlWorkerThread* workerp, VerilatedContext* contextp) {
Verilated::threadContextp(contextp);
workerp->workerLoop();
}
//=============================================================================
// VlThreadPool
VlThreadPool::VlThreadPool(VerilatedContext* contextp, int nThreads,
VlExecutionProfiler* profiler) {
// --threads N passes nThreads=N-1, as the "main" threads counts as 1
++nThreads;
const unsigned cpus = std::thread::hardware_concurrency();
if (cpus < nThreads) {
static int warnedOnce = 0;
if (!warnedOnce++) {
VL_PRINTF_MT("%%Warning: System has %u CPUs but model Verilated with"
" --threads %d; may run slow.\n",
cpus, nThreads);
}
}
// Create worker threads
for (uint32_t threadId = 1; threadId < nThreads; ++threadId) {
m_workers.push_back(new VlWorkerThread{threadId, contextp, profiler});
}
VlThreadPool::VlThreadPool(VerilatedContext* contextp, unsigned nThreads) {
for (unsigned i = 0; i < nThreads; ++i) m_workers.push_back(new VlWorkerThread{contextp});
}
VlThreadPool::~VlThreadPool() {
+24 -26
View File
@@ -24,6 +24,7 @@
#define VERILATOR_VERILATED_THREADS_H_
#include "verilatedos.h"
#include "verilated.h" // for VerilatedMutex and clang annotations
#ifndef VL_THREADED
@@ -50,6 +51,9 @@
#endif
// clang-format on
class VlExecutionProfiler;
class VlThreadPool;
// VlMTaskVertex and VlThreadpool will work with multiple model class types.
// Since the type is opaque to VlMTaskVertex and VlThreadPool, represent it
// as a void* here.
@@ -129,20 +133,14 @@ public:
}
};
class VlExecutionProfiler;
class VlThreadPool;
class VlWorkerThread final {
private:
// TYPES
struct ExecRec {
VlExecFnp m_fnp; // Function to execute
VlSelfP m_selfp; // Symbol table to execute
bool m_evenCycle; // Even/odd for flag alternation
ExecRec()
: m_fnp{nullptr}
, m_selfp{nullptr}
, m_evenCycle{false} {}
VlExecFnp m_fnp = nullptr; // Function to execute
VlSelfP m_selfp = nullptr; // Symbol table to execute
bool m_evenCycle = false; // Even/odd for flag alternation
ExecRec() = default;
ExecRec(VlExecFnp fnp, VlSelfP selfp, bool evenCycle)
: m_fnp{fnp}
, m_selfp{selfp}
@@ -150,7 +148,7 @@ private:
};
// MEMBERS
VerilatedMutex m_mutex;
mutable VerilatedMutex m_mutex;
std::condition_variable_any m_cv;
// Only notify the condition_variable if the worker is waiting
bool m_waiting VL_GUARDED_BY(m_mutex) = false;
@@ -162,26 +160,24 @@ private:
// Store the size atomically, so we can spin wait
std::atomic<size_t> m_ready_size;
std::atomic<bool> m_exiting; // Worker thread should exit
std::thread m_cthread; // Underlying C++ thread record
VerilatedContext* const m_contextp; // Context for spawned thread
VL_UNCOPYABLE(VlWorkerThread);
public:
// CONSTRUCTORS
explicit VlWorkerThread(uint32_t threadId, VerilatedContext* contextp,
VlExecutionProfiler* profilerp);
explicit VlWorkerThread(VerilatedContext* contextp);
~VlWorkerThread();
// METHODS
template <bool SpinWait>
inline void dequeWork(ExecRec* workp) VL_MT_SAFE_EXCLUDES(m_mutex) {
// Spin for a while, waiting for new data
for (int i = 0; i < VL_LOCK_SPINS; ++i) {
if (VL_LIKELY(m_ready_size.load(std::memory_order_relaxed))) { //
break;
if VL_CONSTEXPR_CXX17 (SpinWait) {
for (unsigned i = 0; i < VL_LOCK_SPINS; ++i) {
if (VL_LIKELY(m_ready_size.load(std::memory_order_relaxed))) break;
VL_CPU_RELAX();
}
VL_CPU_RELAX();
}
VerilatedLockGuard lock{m_mutex};
while (m_ready.empty()) {
@@ -195,8 +191,7 @@ public:
m_ready.erase(m_ready.begin());
m_ready_size.fetch_sub(1, std::memory_order_relaxed);
}
inline void wakeUp() { addTask(nullptr, nullptr, false); }
inline void addTask(VlExecFnp fnp, VlSelfP selfp, bool evenCycle)
inline void addTask(VlExecFnp fnp, VlSelfP selfp, bool evenCycle = false)
VL_MT_SAFE_EXCLUDES(m_mutex) {
bool notify;
{
@@ -207,12 +202,15 @@ public:
}
if (notify) m_cv.notify_one();
}
void shutdown(); // Finish current tasks, then terminate thread
void wait(); // Blocks calling thread until all tasks complete in this thread
void workerLoop();
static void startWorker(VlWorkerThread* workerp, uint32_t threadId,
VlExecutionProfiler* profilerp);
static void startWorker(VlWorkerThread* workerp, VerilatedContext* contextp);
};
class VlThreadPool final {
class VlThreadPool final : public VerilatedVirtualBase {
// MEMBERS
std::vector<VlWorkerThread*> m_workers; // our workers
@@ -221,8 +219,8 @@ public:
// Construct a thread pool with 'nThreads' dedicated threads. The thread
// pool will create these threads and make them available to execute tasks
// via this->workerp(index)->addTask(...)
VlThreadPool(VerilatedContext* contextp, int nThreads, VlExecutionProfiler* profilerp);
~VlThreadPool();
VlThreadPool(VerilatedContext* contextp, unsigned nThreads);
~VlThreadPool() override;
// METHODS
inline int numThreads() const { return m_workers.size(); }
+267 -121
View File
@@ -28,26 +28,35 @@
#include "verilated_trace_defs.h"
#include <bitset>
#include <condition_variable>
#include <memory>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>
#ifdef VL_TRACE_THREADED
# include <condition_variable>
#ifdef VL_THREADED
# include <deque>
# include <thread>
#endif
// clang-format on
#ifdef VL_TRACE_THREADED
class VlThreadPool;
template <class T_Buffer>
class VerilatedTraceBuffer;
template <class T_Buffer>
class VerilatedTraceOffloadBuffer;
#ifdef VL_THREADED
//=============================================================================
// Threaded tracing
// Offloaded tracing
// A simple synchronized first in first out queue
template <class T> class VerilatedThreadQueue final { // LCOV_EXCL_LINE // lcov bug
template <class T>
class VerilatedThreadQueue final { // LCOV_EXCL_LINE // lcov bug
private:
VerilatedMutex m_mutex; // Protects m_queue
mutable VerilatedMutex m_mutex; // Protects m_queue
std::condition_variable_any m_cv;
std::deque<T> m_queue VL_GUARDED_BY(m_mutex);
@@ -88,7 +97,7 @@ public:
// Commands used by thread tracing. Anonymous enum in class, as we want
// it scoped, but we also want the automatic conversion to integer types.
class VerilatedTraceCommand final {
class VerilatedTraceOffloadCommand final {
public:
// These must all fit in 4 bit at the moment, as the tracing routines
// pack parameters in the top bits.
@@ -102,36 +111,65 @@ public:
CHG_WDATA = 0x6,
CHG_DOUBLE = 0x8,
// TODO: full..
TIME_CHANGE = 0xd,
TIME_CHANGE = 0xc,
TRACE_BUFFER = 0xd,
END = 0xe, // End of buffer
SHUTDOWN = 0xf // Shutdown worker thread, also marks end of buffer
};
};
#endif
//=============================================================================
// VerilatedTraceConfig
// Simple data representing trace configuration required by generated models.
class VerilatedTraceConfig final {
public:
const bool m_useParallel; // Use parallel tracing
const bool m_useOffloading; // Offloading trace rendering
const bool m_useFstWriterThread; // Use the separate FST writer thread
VerilatedTraceConfig(bool useParallel, bool useOffloading, bool useFstWriterThread)
: m_useParallel{useParallel}
, m_useOffloading{useOffloading}
, m_useFstWriterThread{useFstWriterThread} {}
};
//=============================================================================
// VerilatedTrace
// VerilatedTrace uses F-bounded polymorphism to access duck-typed
// implementations in the format specific derived class, which must be passed
// as the type parameter T_Derived
template <class T_Derived> class VerilatedTrace VL_NOT_FINAL {
// T_Trace is the format specific subclass of VerilatedTrace.
// T_Buffer is the format specific base class of VerilatedTraceBuffer.
template <class T_Trace, class T_Buffer>
class VerilatedTrace VL_NOT_FINAL {
public:
using Buffer = VerilatedTraceBuffer<T_Buffer>;
using OffloadBuffer = VerilatedTraceOffloadBuffer<T_Buffer>;
//=========================================================================
// Generic tracing internals
using initCb_t = void (*)(void*, T_Derived*, uint32_t); // Type of init callbacks
using dumpCb_t = void (*)(void*, T_Derived*); // Type of all but init callbacks
using initCb_t = void (*)(void*, T_Trace*, uint32_t); // Type of init callbacks
using dumpCb_t = void (*)(void*, Buffer*); // Type of dump callbacks
using dumpOffloadCb_t = void (*)(void*, OffloadBuffer*); // Type of offload dump callbacks
using cleanupCb_t = void (*)(void*, T_Trace*); // Type of cleanup callbacks
private:
// Give the buffer (both base and derived) access to the private bits
friend T_Buffer;
friend Buffer;
friend OffloadBuffer;
struct CallbackRecord {
// Note: would make these fields const, but some old STL implementations
// (the one in Ubuntu 14.04 with GCC 4.8.4 in particular) use the
// assignment operator on inserting into collections, so they don't work
// with const fields...
union {
initCb_t m_initCb; // The callback function
dumpCb_t m_dumpCb; // The callback function
union { // The callback
initCb_t m_initCb;
dumpCb_t m_dumpCb;
dumpOffloadCb_t m_dumpOffloadCb;
cleanupCb_t m_cleanupCb;
};
void* m_userp; // The user pointer to pass to the callback (the symbol table)
CallbackRecord(initCb_t cb, void* userp)
@@ -140,65 +178,110 @@ private:
CallbackRecord(dumpCb_t cb, void* userp)
: m_dumpCb{cb}
, m_userp{userp} {}
CallbackRecord(dumpOffloadCb_t cb, void* userp)
: m_dumpOffloadCb{cb}
, m_userp{userp} {}
CallbackRecord(cleanupCb_t cb, void* userp)
: m_cleanupCb{cb}
, m_userp{userp} {}
};
uint32_t* m_sigs_oldvalp; // Old value store
EData* m_sigs_enabledp; // Bit vector of enabled codes (nullptr = all on)
uint64_t m_timeLastDump; // Last time we did a dump
bool m_offload = false; // Use the offload thread (ignored if !VL_THREADED)
bool m_parallel = false; // Use parallel tracing (ignored if !VL_THREADED)
#ifdef VL_THREADED
struct ParallelWorkerData {
const dumpCb_t m_cb; // The callback
void* const m_userp; // The use pointer to pass to the callback
Buffer* const m_bufp; // The buffer pointer to pass to the callback
std::atomic<bool> m_ready{false}; // The ready flag
mutable VerilatedMutex m_mutex; // Mutex for suspension until ready
std::condition_variable_any m_cv; // Condition variable for suspension
bool m_waiting VL_GUARDED_BY(m_mutex) = false; // Whether a thread is suspended in wait()
void wait();
ParallelWorkerData(dumpCb_t cb, void* userp, Buffer* bufp)
: m_cb{cb}
, m_userp{userp}
, m_bufp{bufp} {}
};
// Passed a ParallelWorkerData*, second argument is ignored
static void parallelWorkerTask(void*, bool);
#endif
protected:
uint32_t* m_sigs_oldvalp = nullptr; // Previous value store
EData* m_sigs_enabledp = nullptr; // Bit vector of enabled codes (nullptr = all on)
private:
std::vector<bool> m_sigs_enabledVec; // Staging for m_sigs_enabledp
std::vector<CallbackRecord> m_initCbs; // Routines to initialize traciong
std::vector<CallbackRecord> m_initCbs; // Routines to initialize tracing
std::vector<CallbackRecord> m_fullCbs; // Routines to perform full dump
std::vector<CallbackRecord> m_fullOffloadCbs; // Routines to perform offloaded full dump
std::vector<CallbackRecord> m_chgCbs; // Routines to perform incremental dump
std::vector<CallbackRecord> m_chgOffloadCbs; // Routines to perform offloaded incremental dump
std::vector<CallbackRecord> m_cleanupCbs; // Routines to call at the end of dump
bool m_fullDump; // Whether a full dump is required on the next call to 'dump'
uint32_t m_nextCode; // Next code number to assign
uint32_t m_numSignals; // Number of distinct signals
uint32_t m_maxBits; // Number of bits in the widest signal
bool m_fullDump = true; // Whether a full dump is required on the next call to 'dump'
uint32_t m_nextCode = 0; // Next code number to assign
uint32_t m_numSignals = 0; // Number of distinct signals
uint32_t m_maxBits = 0; // Number of bits in the widest signal
std::vector<std::string> m_namePrefixStack{""}; // Path prefixes to add to signal names
std::vector<std::pair<int, std::string>> m_dumpvars; // dumpvar() entries
char m_scopeEscape;
double m_timeRes; // Time resolution (ns/ms etc)
double m_timeUnit; // Time units (ns/ms etc)
char m_scopeEscape = '.';
double m_timeRes = 1e-9; // Time resolution (ns/ms etc)
double m_timeUnit = 1e-0; // Time units (ns/ms etc)
uint64_t m_timeLastDump = 0; // Last time we did a dump
bool m_didSomeDump = false; // Did at least one dump (i.e.: m_timeLastDump is valid)
VerilatedContext* m_contextp = nullptr; // The context used by the traced models
std::unordered_set<const VerilatedModel*> m_models; // The collection of models being traced
void addCallbackRecord(std::vector<CallbackRecord>& cbVec, CallbackRecord& cbRec)
void addCallbackRecord(std::vector<CallbackRecord>& cbVec, CallbackRecord&& cbRec)
VL_MT_SAFE_EXCLUDES(m_mutex);
// Equivalent to 'this' but is of the sub-type 'T_Derived*'. Use 'self()->'
// Equivalent to 'this' but is of the sub-type 'T_Trace*'. Use 'self()->'
// to access duck-typed functions to avoid a virtual function call.
T_Derived* self() { return static_cast<T_Derived*>(this); }
T_Trace* self() { return static_cast<T_Trace*>(this); }
void runCallbacks(const std::vector<CallbackRecord>& cbVec);
void runOffloadedCallbacks(const std::vector<CallbackRecord>& cbVec);
// Flush any remaining data for this file
static void onFlush(void* selfp) VL_MT_UNSAFE_ONE;
// Close the file on termination
static void onExit(void* selfp) VL_MT_UNSAFE_ONE;
#ifdef VL_TRACE_THREADED
// Number of total trace buffers that have been allocated
uint32_t m_numTraceBuffers;
// Size of trace buffers
size_t m_traceBufferSize;
#ifdef VL_THREADED
// Number of total offload buffers that have been allocated
uint32_t m_numOffloadBuffers = 0;
// Size of offload buffers
size_t m_offloadBufferSize = 0;
// Buffers handed to worker for processing
VerilatedThreadQueue<uint32_t*> m_buffersToWorker;
VerilatedThreadQueue<uint32_t*> m_offloadBuffersToWorker;
// Buffers returned from worker after processing
VerilatedThreadQueue<uint32_t*> m_buffersFromWorker;
VerilatedThreadQueue<uint32_t*> m_offloadBuffersFromWorker;
protected:
// Write pointer into current buffer
uint32_t* m_traceBufferWritep;
// End of trace buffer
uint32_t* m_traceBufferEndp;
// The worker thread itself
uint32_t* m_offloadBufferWritep = nullptr;
// End of offload buffer
uint32_t* m_offloadBufferEndp = nullptr;
private:
// The offload worker thread itself
std::unique_ptr<std::thread> m_workerThread;
// Get a new trace buffer that can be populated. May block if none available
uint32_t* getTraceBuffer();
// Get a new offload buffer that can be populated. May block if none available
uint32_t* getOffloadBuffer();
// The function executed by the worker thread
void workerThreadMain();
// The function executed by the offload worker thread
void offloadWorkerThreadMain();
// Wait until given buffer is placed in m_buffersFromWorker
void waitForBuffer(const uint32_t* bufferp);
// Wait until given offload buffer is placed in m_offloadBuffersFromWorker
void waitForOffloadBuffer(const uint32_t* bufferp);
// Shut down and join worker, if it's running, otherwise do nothing
void shutdownWorker();
void shutdownOffloadWorker();
#endif
// CONSTRUCTORS
@@ -208,13 +291,12 @@ protected:
//=========================================================================
// Internals available to format specific implementations
VerilatedMutex m_mutex; // Ensure dump() etc only called from single thread
mutable VerilatedMutex m_mutex; // Ensure dump() etc only called from single thread
uint32_t nextCode() const { return m_nextCode; }
uint32_t numSignals() const { return m_numSignals; }
uint32_t maxBits() const { return m_maxBits; }
void fullDump(bool value) { m_fullDump = value; }
uint64_t timeLastDump() { return m_timeLastDump; }
double timeRes() const { return m_timeRes; }
double timeUnit() const { return m_timeUnit; }
@@ -235,6 +317,14 @@ protected:
void closeBase();
void flushBase();
#ifdef VL_THREADED
inline bool offload() const { return m_offload; }
inline bool parallel() const { return m_parallel; }
#else
static constexpr bool offload() { return false; }
static constexpr bool parallel() { return false; }
#endif
//=========================================================================
// Virtual functions to be provided by the format specific implementation
@@ -246,6 +336,13 @@ protected:
virtual bool preFullDump() = 0;
virtual bool preChangeDump() = 0;
// Trace buffer management
virtual Buffer* getTraceBuffer() = 0;
virtual void commitTraceBuffer(Buffer*) = 0;
// Configure sub-class
virtual void configure(const VerilatedTraceConfig&) = 0;
public:
//=========================================================================
// External interface to client code
@@ -266,19 +363,51 @@ public:
// Call
void dump(uint64_t timeui) VL_MT_SAFE_EXCLUDES(m_mutex);
//=========================================================================
// Internal interface to Verilator generated code
//=========================================================================
// Non-hot path internal interface to Verilator generated code
void addModel(VerilatedModel*) VL_MT_SAFE_EXCLUDES(m_mutex);
void addInitCb(initCb_t cb, void* userp) VL_MT_SAFE;
void addFullCb(dumpCb_t cb, void* userp) VL_MT_SAFE;
void addFullCb(dumpOffloadCb_t cb, void* userp) VL_MT_SAFE;
void addChgCb(dumpCb_t cb, void* userp) VL_MT_SAFE;
void addCleanupCb(dumpCb_t cb, void* userp) VL_MT_SAFE;
void addChgCb(dumpOffloadCb_t cb, void* userp) VL_MT_SAFE;
void addCleanupCb(cleanupCb_t cb, void* userp) VL_MT_SAFE;
void scopeEscape(char flag) { m_scopeEscape = flag; }
void pushNamePrefix(const std::string&);
void popNamePrefix(unsigned count = 1);
};
//=============================================================================
// VerilatedTraceBuffer
// T_Buffer is the format specific base class of VerilatedTraceBuffer.
// The format-specific hot-path methods use duck-typing via T_Buffer for performance.
template <class T_Buffer>
class VerilatedTraceBuffer VL_NOT_FINAL : public T_Buffer {
protected:
// Type of the owner trace file
using Trace = typename std::remove_cv<
typename std::remove_reference<decltype(T_Buffer::m_owner)>::type>::type;
static_assert(std::has_virtual_destructor<T_Buffer>::value, "");
static_assert(std::is_base_of<VerilatedTrace<Trace, T_Buffer>, Trace>::value, "");
friend Trace; // Give the trace file access to the private bits
friend std::default_delete<VerilatedTraceBuffer<T_Buffer>>;
uint32_t* const m_sigs_oldvalp; // Previous value store
EData* const m_sigs_enabledp; // Bit vector of enabled codes (nullptr = all on)
explicit VerilatedTraceBuffer(Trace& owner);
~VerilatedTraceBuffer() override = default;
public:
//=========================================================================
// Hot path internal interface to Verilator generated code
@@ -296,7 +425,7 @@ public:
// duck-typed void emitWData(uint32_t code, const WData* newvalp, int bits) = 0;
// duck-typed void emitDouble(uint32_t code, double newval) = 0;
uint32_t* oldp(uint32_t code) { return m_sigs_oldvalp + code; }
VL_ATTR_ALWINLINE inline uint32_t* oldp(uint32_t code) { return m_sigs_oldvalp + code; }
// Write to previous value buffer value and emit trace entry.
void fullBit(uint32_t* oldp, CData newval);
@@ -307,89 +436,32 @@ public:
void fullWData(uint32_t* oldp, const WData* newvalp, int bits);
void fullDouble(uint32_t* oldp, double newval);
#ifdef VL_TRACE_THREADED
// Threaded tracing. Just dump everything in the trace buffer
inline void chgBit(uint32_t code, CData newval) {
m_traceBufferWritep[0] = VerilatedTraceCommand::CHG_BIT_0 | newval;
m_traceBufferWritep[1] = code;
m_traceBufferWritep += 2;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgCData(uint32_t code, CData newval, int bits) {
m_traceBufferWritep[0] = (bits << 4) | VerilatedTraceCommand::CHG_CDATA;
m_traceBufferWritep[1] = code;
m_traceBufferWritep[2] = newval;
m_traceBufferWritep += 3;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgSData(uint32_t code, SData newval, int bits) {
m_traceBufferWritep[0] = (bits << 4) | VerilatedTraceCommand::CHG_SDATA;
m_traceBufferWritep[1] = code;
m_traceBufferWritep[2] = newval;
m_traceBufferWritep += 3;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgIData(uint32_t code, IData newval, int bits) {
m_traceBufferWritep[0] = (bits << 4) | VerilatedTraceCommand::CHG_IDATA;
m_traceBufferWritep[1] = code;
m_traceBufferWritep[2] = newval;
m_traceBufferWritep += 3;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgQData(uint32_t code, QData newval, int bits) {
m_traceBufferWritep[0] = (bits << 4) | VerilatedTraceCommand::CHG_QDATA;
m_traceBufferWritep[1] = code;
*reinterpret_cast<QData*>(m_traceBufferWritep + 2) = newval;
m_traceBufferWritep += 4;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgWData(uint32_t code, const WData* newvalp, int bits) {
m_traceBufferWritep[0] = (bits << 4) | VerilatedTraceCommand::CHG_WDATA;
m_traceBufferWritep[1] = code;
m_traceBufferWritep += 2;
for (int i = 0; i < (bits + 31) / 32; ++i) { *m_traceBufferWritep++ = newvalp[i]; }
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
inline void chgDouble(uint32_t code, double newval) {
m_traceBufferWritep[0] = VerilatedTraceCommand::CHG_DOUBLE;
m_traceBufferWritep[1] = code;
// cppcheck-suppress invalidPointerCast
*reinterpret_cast<double*>(m_traceBufferWritep + 2) = newval;
m_traceBufferWritep += 4;
VL_DEBUG_IF(assert(m_traceBufferWritep <= m_traceBufferEndp););
}
#define CHG(name) chg##name##Impl
#else
#define CHG(name) chg##name
#endif
// In non-threaded mode, these are called directly by the trace callbacks,
// and are called chg*. In threaded mode, they are called by the worker
// In non-offload mode, these are called directly by the trace callbacks,
// and are called chg*. In offload mode, they are called by the worker
// thread and are called chg*Impl
// Check previous dumped value of signal. If changed, then emit trace entry
inline void CHG(Bit)(uint32_t* oldp, CData newval) {
VL_ATTR_ALWINLINE inline void chgBit(uint32_t* oldp, CData newval) {
const uint32_t diff = *oldp ^ newval;
if (VL_UNLIKELY(diff)) fullBit(oldp, newval);
}
inline void CHG(CData)(uint32_t* oldp, CData newval, int bits) {
VL_ATTR_ALWINLINE inline void chgCData(uint32_t* oldp, CData newval, int bits) {
const uint32_t diff = *oldp ^ newval;
if (VL_UNLIKELY(diff)) fullCData(oldp, newval, bits);
}
inline void CHG(SData)(uint32_t* oldp, SData newval, int bits) {
VL_ATTR_ALWINLINE inline void chgSData(uint32_t* oldp, SData newval, int bits) {
const uint32_t diff = *oldp ^ newval;
if (VL_UNLIKELY(diff)) fullSData(oldp, newval, bits);
}
inline void CHG(IData)(uint32_t* oldp, IData newval, int bits) {
VL_ATTR_ALWINLINE inline void chgIData(uint32_t* oldp, IData newval, int bits) {
const uint32_t diff = *oldp ^ newval;
if (VL_UNLIKELY(diff)) fullIData(oldp, newval, bits);
}
inline void CHG(QData)(uint32_t* oldp, QData newval, int bits) {
VL_ATTR_ALWINLINE inline void chgQData(uint32_t* oldp, QData newval, int bits) {
const uint64_t diff = *reinterpret_cast<QData*>(oldp) ^ newval;
if (VL_UNLIKELY(diff)) fullQData(oldp, newval, bits);
}
inline void CHG(WData)(uint32_t* oldp, const WData* newvalp, int bits) {
VL_ATTR_ALWINLINE inline void chgWData(uint32_t* oldp, const WData* newvalp, int bits) {
for (int i = 0; i < (bits + 31) / 32; ++i) {
if (VL_UNLIKELY(oldp[i] ^ newvalp[i])) {
fullWData(oldp, newvalp, bits);
@@ -397,11 +469,85 @@ public:
}
}
}
inline void CHG(Double)(uint32_t* oldp, double newval) {
VL_ATTR_ALWINLINE inline void chgDouble(uint32_t* oldp, double newval) {
// cppcheck-suppress invalidPointerCast
if (VL_UNLIKELY(*reinterpret_cast<double*>(oldp) != newval)) fullDouble(oldp, newval);
}
#undef CHG
};
#ifdef VL_THREADED
//=============================================================================
// VerilatedTraceOffloadBuffer
// T_Buffer is the format specific base class of VerilatedTraceBuffer.
// The format-specific hot-path methods use duck-typing via T_Buffer for performance.
template <class T_Buffer>
class VerilatedTraceOffloadBuffer final : public VerilatedTraceBuffer<T_Buffer> {
using typename VerilatedTraceBuffer<T_Buffer>::Trace;
friend Trace; // Give the trace file access to the private bits
uint32_t* m_offloadBufferWritep; // Write pointer into current buffer
uint32_t* const m_offloadBufferEndp; // End of offload buffer
explicit VerilatedTraceOffloadBuffer(Trace& owner);
~VerilatedTraceOffloadBuffer() override = default;
public:
//=========================================================================
// Hot path internal interface to Verilator generated code
// Offloaded tracing. Just dump everything in the offload buffer
inline void chgBit(uint32_t code, CData newval) {
m_offloadBufferWritep[0] = VerilatedTraceOffloadCommand::CHG_BIT_0 | newval;
m_offloadBufferWritep[1] = code;
m_offloadBufferWritep += 2;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgCData(uint32_t code, CData newval, int bits) {
m_offloadBufferWritep[0] = (bits << 4) | VerilatedTraceOffloadCommand::CHG_CDATA;
m_offloadBufferWritep[1] = code;
m_offloadBufferWritep[2] = newval;
m_offloadBufferWritep += 3;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgSData(uint32_t code, SData newval, int bits) {
m_offloadBufferWritep[0] = (bits << 4) | VerilatedTraceOffloadCommand::CHG_SDATA;
m_offloadBufferWritep[1] = code;
m_offloadBufferWritep[2] = newval;
m_offloadBufferWritep += 3;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgIData(uint32_t code, IData newval, int bits) {
m_offloadBufferWritep[0] = (bits << 4) | VerilatedTraceOffloadCommand::CHG_IDATA;
m_offloadBufferWritep[1] = code;
m_offloadBufferWritep[2] = newval;
m_offloadBufferWritep += 3;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgQData(uint32_t code, QData newval, int bits) {
m_offloadBufferWritep[0] = (bits << 4) | VerilatedTraceOffloadCommand::CHG_QDATA;
m_offloadBufferWritep[1] = code;
*reinterpret_cast<QData*>(m_offloadBufferWritep + 2) = newval;
m_offloadBufferWritep += 4;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgWData(uint32_t code, const WData* newvalp, int bits) {
m_offloadBufferWritep[0] = (bits << 4) | VerilatedTraceOffloadCommand::CHG_WDATA;
m_offloadBufferWritep[1] = code;
m_offloadBufferWritep += 2;
for (int i = 0; i < (bits + 31) / 32; ++i) { *m_offloadBufferWritep++ = newvalp[i]; }
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
inline void chgDouble(uint32_t code, double newval) {
m_offloadBufferWritep[0] = VerilatedTraceOffloadCommand::CHG_DOUBLE;
m_offloadBufferWritep[1] = code;
// cppcheck-suppress invalidPointerCast
*reinterpret_cast<double*>(m_offloadBufferWritep + 2) = newval;
m_offloadBufferWritep += 4;
VL_DEBUG_IF(assert(m_offloadBufferWritep <= m_offloadBufferEndp););
}
};
#endif
#endif // guard
@@ -10,32 +10,32 @@
// SPDX-License-Identifier: LGPL-3.0-only OR Artistic-2.0
//
//=============================================================================
///
/// \file
/// \brief Verilated common-format tracing implementation code
///
/// This file must be compiled and linked against all Verilated objects
/// that use --trace.
///
/// Use "verilator --trace" to add this to the Makefile for the linker.
///
//
// Verilated tracing implementation code template common to all formats.
// This file is included by the format specific implementations and
// should not be used otherwise.
//
//=============================================================================
// clang-format off
#ifndef VL_CPPCHECK
#ifndef VL_DERIVED_T
#if !defined(VL_SUB_T) || !defined(VL_BUF_T)
# error "This file should be included in trace format implementations"
#endif
#include "verilated_intrinsics.h"
#include "verilated_trace.h"
#ifdef VL_THREADED
# include "verilated_threads.h"
# include <list>
#endif
#if 0
# include <iostream>
# define VL_TRACE_THREAD_DEBUG(msg) std::cout << "TRACE THREAD: " << msg << std::endl
# define VL_TRACE_OFFLOAD_DEBUG(msg) std::cout << "TRACE OFFLOAD THREAD: " << msg << std::endl
#else
# define VL_TRACE_THREAD_DEBUG(msg)
# define VL_TRACE_OFFLOAD_DEBUG(msg)
#endif
// clang-format on
@@ -78,37 +78,39 @@ static std::string doubleToTimescale(double value) {
return valuestr; // Gets converted to string, so no ref to stack
}
#ifdef VL_TRACE_THREADED
#ifdef VL_THREADED
//=========================================================================
// Buffer management
template <> uint32_t* VerilatedTrace<VL_DERIVED_T>::getTraceBuffer() {
template <>
uint32_t* VerilatedTrace<VL_SUB_T, VL_BUF_T>::getOffloadBuffer() {
uint32_t* bufferp;
// Some jitter is expected, so some number of alternative trace buffers are
// Some jitter is expected, so some number of alternative offlaod buffers are
// required, but don't allocate more than 8 buffers.
if (m_numTraceBuffers < 8) {
if (m_numOffloadBuffers < 8) {
// Allocate a new buffer if none is available
if (!m_buffersFromWorker.tryGet(bufferp)) {
++m_numTraceBuffers;
if (!m_offloadBuffersFromWorker.tryGet(bufferp)) {
++m_numOffloadBuffers;
// Note: over allocate a bit so pointer comparison is well defined
// if we overflow only by a small amount
bufferp = new uint32_t[m_traceBufferSize + 16];
bufferp = new uint32_t[m_offloadBufferSize + 16];
}
} else {
// Block until a buffer becomes available
bufferp = m_buffersFromWorker.get();
bufferp = m_offloadBuffersFromWorker.get();
}
return bufferp;
}
template <> void VerilatedTrace<VL_DERIVED_T>::waitForBuffer(const uint32_t* buffp) {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::waitForOffloadBuffer(const uint32_t* buffp) {
// Slow path code only called on flush/shutdown, so use a simple algorithm.
// Collect buffers from worker and stash them until we get the one we want.
std::deque<uint32_t*> stash;
do { stash.push_back(m_buffersFromWorker.get()); } while (stash.back() != buffp);
do { stash.push_back(m_offloadBuffersFromWorker.get()); } while (stash.back() != buffp);
// Now put them back in the queue, in the original order.
while (!stash.empty()) {
m_buffersFromWorker.put_front(stash.back());
m_offloadBuffersFromWorker.put_front(stash.back());
stash.pop_back();
}
}
@@ -116,17 +118,20 @@ template <> void VerilatedTrace<VL_DERIVED_T>::waitForBuffer(const uint32_t* buf
//=========================================================================
// Worker thread
template <> void VerilatedTrace<VL_DERIVED_T>::workerThreadMain() {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::offloadWorkerThreadMain() {
bool shutdown = false;
do {
uint32_t* const bufferp = m_buffersToWorker.get();
uint32_t* const bufferp = m_offloadBuffersToWorker.get();
VL_TRACE_THREAD_DEBUG("");
VL_TRACE_THREAD_DEBUG("Got buffer: " << bufferp);
VL_TRACE_OFFLOAD_DEBUG("");
VL_TRACE_OFFLOAD_DEBUG("Got buffer: " << bufferp);
const uint32_t* readp = bufferp;
std::unique_ptr<Buffer> traceBufp; // We own the passed tracebuffer
while (true) {
const uint32_t cmd = readp[0];
const uint32_t top = cmd >> 4;
@@ -139,71 +144,79 @@ template <> void VerilatedTrace<VL_DERIVED_T>::workerThreadMain() {
switch (cmd & 0xF) {
//===
// CHG_* commands
case VerilatedTraceCommand::CHG_BIT_0:
VL_TRACE_THREAD_DEBUG("Command CHG_BIT_0 " << top);
chgBitImpl(oldp, 0);
case VerilatedTraceOffloadCommand::CHG_BIT_0:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_BIT_0 " << top);
traceBufp->chgBit(oldp, 0);
continue;
case VerilatedTraceCommand::CHG_BIT_1:
VL_TRACE_THREAD_DEBUG("Command CHG_BIT_1 " << top);
chgBitImpl(oldp, 1);
case VerilatedTraceOffloadCommand::CHG_BIT_1:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_BIT_1 " << top);
traceBufp->chgBit(oldp, 1);
continue;
case VerilatedTraceCommand::CHG_CDATA:
VL_TRACE_THREAD_DEBUG("Command CHG_CDATA " << top);
case VerilatedTraceOffloadCommand::CHG_CDATA:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_CDATA " << top);
// Bits stored in bottom byte of command
chgCDataImpl(oldp, *readp, top);
traceBufp->chgCData(oldp, *readp, top);
readp += 1;
continue;
case VerilatedTraceCommand::CHG_SDATA:
VL_TRACE_THREAD_DEBUG("Command CHG_SDATA " << top);
case VerilatedTraceOffloadCommand::CHG_SDATA:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_SDATA " << top);
// Bits stored in bottom byte of command
chgSDataImpl(oldp, *readp, top);
traceBufp->chgSData(oldp, *readp, top);
readp += 1;
continue;
case VerilatedTraceCommand::CHG_IDATA:
VL_TRACE_THREAD_DEBUG("Command CHG_IDATA " << top);
case VerilatedTraceOffloadCommand::CHG_IDATA:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_IDATA " << top);
// Bits stored in bottom byte of command
chgIDataImpl(oldp, *readp, top);
traceBufp->chgIData(oldp, *readp, top);
readp += 1;
continue;
case VerilatedTraceCommand::CHG_QDATA:
VL_TRACE_THREAD_DEBUG("Command CHG_QDATA " << top);
case VerilatedTraceOffloadCommand::CHG_QDATA:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_QDATA " << top);
// Bits stored in bottom byte of command
chgQDataImpl(oldp, *reinterpret_cast<const QData*>(readp), top);
traceBufp->chgQData(oldp, *reinterpret_cast<const QData*>(readp), top);
readp += 2;
continue;
case VerilatedTraceCommand::CHG_WDATA:
VL_TRACE_THREAD_DEBUG("Command CHG_WDATA " << top);
chgWDataImpl(oldp, readp, top);
case VerilatedTraceOffloadCommand::CHG_WDATA:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_WDATA " << top);
traceBufp->chgWData(oldp, readp, top);
readp += VL_WORDS_I(top);
continue;
case VerilatedTraceCommand::CHG_DOUBLE:
VL_TRACE_THREAD_DEBUG("Command CHG_DOUBLE " << top);
chgDoubleImpl(oldp, *reinterpret_cast<const double*>(readp));
case VerilatedTraceOffloadCommand::CHG_DOUBLE:
VL_TRACE_OFFLOAD_DEBUG("Command CHG_DOUBLE " << top);
traceBufp->chgDouble(oldp, *reinterpret_cast<const double*>(readp));
readp += 2;
continue;
//===
// Rare commands
case VerilatedTraceCommand::TIME_CHANGE:
VL_TRACE_THREAD_DEBUG("Command TIME_CHANGE " << top);
case VerilatedTraceOffloadCommand::TIME_CHANGE:
VL_TRACE_OFFLOAD_DEBUG("Command TIME_CHANGE " << top);
readp -= 1; // No code in this command, undo increment
emitTimeChange(*reinterpret_cast<const uint64_t*>(readp));
readp += 2;
continue;
case VerilatedTraceOffloadCommand::TRACE_BUFFER:
VL_TRACE_OFFLOAD_DEBUG("Command TRACE_BUFFER " << top);
readp -= 1; // No code in this command, undo increment
traceBufp.reset(*reinterpret_cast<Buffer* const*>(readp));
readp += 2;
continue;
//===
// Commands ending this buffer
case VerilatedTraceCommand::END: VL_TRACE_THREAD_DEBUG("Command END"); break;
case VerilatedTraceCommand::SHUTDOWN:
VL_TRACE_THREAD_DEBUG("Command SHUTDOWN");
case VerilatedTraceOffloadCommand::END: //
VL_TRACE_OFFLOAD_DEBUG("Command END");
break;
case VerilatedTraceOffloadCommand::SHUTDOWN:
VL_TRACE_OFFLOAD_DEBUG("Command SHUTDOWN");
shutdown = true;
break;
//===
// Unknown command
default: { // LCOV_EXCL_START
VL_TRACE_THREAD_DEBUG("Command UNKNOWN");
VL_PRINTF_MT("Trace command: 0x%08x\n", cmd);
VL_TRACE_OFFLOAD_DEBUG("Command UNKNOWN " << cmd);
VL_FATAL_MT(__FILE__, __LINE__, "", "Unknown trace command");
break;
} // LCOV_EXCL_STOP
@@ -214,23 +227,24 @@ template <> void VerilatedTrace<VL_DERIVED_T>::workerThreadMain() {
break;
}
VL_TRACE_THREAD_DEBUG("Returning buffer");
VL_TRACE_OFFLOAD_DEBUG("Returning buffer");
// Return buffer
m_buffersFromWorker.put(bufferp);
m_offloadBuffersFromWorker.put(bufferp);
} while (VL_LIKELY(!shutdown));
}
template <> void VerilatedTrace<VL_DERIVED_T>::shutdownWorker() {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::shutdownOffloadWorker() {
// If the worker thread is not running, done..
if (!m_workerThread) return;
// Hand an buffer with a shutdown command to the worker thread
uint32_t* const bufferp = getTraceBuffer();
bufferp[0] = VerilatedTraceCommand::SHUTDOWN;
m_buffersToWorker.put(bufferp);
uint32_t* const bufferp = getOffloadBuffer();
bufferp[0] = VerilatedTraceOffloadCommand::SHUTDOWN;
m_offloadBuffersToWorker.put(bufferp);
// Wait for it to return
waitForBuffer(bufferp);
waitForOffloadBuffer(bufferp);
// Join the thread and delete it
m_workerThread->join();
m_workerThread.reset(nullptr);
@@ -241,80 +255,72 @@ template <> void VerilatedTrace<VL_DERIVED_T>::shutdownWorker() {
//=============================================================================
// Life cycle
template <> void VerilatedTrace<VL_DERIVED_T>::closeBase() {
#ifdef VL_TRACE_THREADED
shutdownWorker();
while (m_numTraceBuffers) {
delete[] m_buffersFromWorker.get();
--m_numTraceBuffers;
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::closeBase() {
#ifdef VL_THREADED
if (offload()) {
shutdownOffloadWorker();
while (m_numOffloadBuffers) {
delete[] m_offloadBuffersFromWorker.get();
--m_numOffloadBuffers;
}
}
#endif
}
template <> void VerilatedTrace<VL_DERIVED_T>::flushBase() {
#ifdef VL_TRACE_THREADED
// Hand an empty buffer to the worker thread
uint32_t* const bufferp = getTraceBuffer();
*bufferp = VerilatedTraceCommand::END;
m_buffersToWorker.put(bufferp);
// Wait for it to be returned. As the processing is in-order,
// this ensures all previous buffers have been processed.
waitForBuffer(bufferp);
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::flushBase() {
#ifdef VL_THREADED
if (offload()) {
// Hand an empty buffer to the worker thread
uint32_t* const bufferp = getOffloadBuffer();
*bufferp = VerilatedTraceOffloadCommand::END;
m_offloadBuffersToWorker.put(bufferp);
// Wait for it to be returned. As the processing is in-order,
// this ensures all previous buffers have been processed.
waitForOffloadBuffer(bufferp);
}
#endif
}
//=============================================================================
// Callbacks to run on global events
template <> void VerilatedTrace<VL_DERIVED_T>::onFlush(void* selfp) {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::onFlush(void* selfp) {
// This calls 'flush' on the derived class (which must then get any mutex)
reinterpret_cast<VL_DERIVED_T*>(selfp)->flush();
reinterpret_cast<VL_SUB_T*>(selfp)->flush();
}
template <> void VerilatedTrace<VL_DERIVED_T>::onExit(void* selfp) {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::onExit(void* selfp) {
// This calls 'close' on the derived class (which must then get any mutex)
reinterpret_cast<VL_DERIVED_T*>(selfp)->close();
reinterpret_cast<VL_SUB_T*>(selfp)->close();
}
//=============================================================================
// VerilatedTrace
template <>
VerilatedTrace<VL_DERIVED_T>::VerilatedTrace()
: m_sigs_oldvalp{nullptr}
, m_sigs_enabledp{nullptr}
, m_timeLastDump{0}
, m_fullDump{true}
, m_nextCode{0}
, m_numSignals{0}
, m_maxBits{0}
, m_scopeEscape{'.'}
, m_timeRes{1e-9}
, m_timeUnit {
1e-9
}
#ifdef VL_TRACE_THREADED
, m_numTraceBuffers { 0 }
#endif
{
VerilatedTrace<VL_SUB_T, VL_BUF_T>::VerilatedTrace() {
set_time_unit(Verilated::threadContextp()->timeunitString());
set_time_resolution(Verilated::threadContextp()->timeprecisionString());
}
template <> VerilatedTrace<VL_DERIVED_T>::~VerilatedTrace() {
template <>
VerilatedTrace<VL_SUB_T, VL_BUF_T>::~VerilatedTrace() {
if (m_sigs_oldvalp) VL_DO_CLEAR(delete[] m_sigs_oldvalp, m_sigs_oldvalp = nullptr);
if (m_sigs_enabledp) VL_DO_CLEAR(delete[] m_sigs_enabledp, m_sigs_enabledp = nullptr);
Verilated::removeFlushCb(VerilatedTrace<VL_DERIVED_T>::onFlush, this);
Verilated::removeExitCb(VerilatedTrace<VL_DERIVED_T>::onExit, this);
#ifdef VL_TRACE_THREADED
closeBase();
#endif
Verilated::removeFlushCb(VerilatedTrace<VL_SUB_T, VL_BUF_T>::onFlush, this);
Verilated::removeExitCb(VerilatedTrace<VL_SUB_T, VL_BUF_T>::onExit, this);
if (offload()) closeBase();
}
//=========================================================================
// Internals available to format specific implementations
template <> void VerilatedTrace<VL_DERIVED_T>::traceInit() VL_MT_UNSAFE {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::traceInit() VL_MT_UNSAFE {
// Note: It is possible to re-open a trace file (VCD in particular),
// so we must reset the next code here, but it must have the same number
// of codes on re-open
@@ -359,25 +365,28 @@ template <> void VerilatedTrace<VL_DERIVED_T>::traceInit() VL_MT_UNSAFE {
}
// Set callback so flush/abort will flush this file
Verilated::addFlushCb(VerilatedTrace<VL_DERIVED_T>::onFlush, this);
Verilated::addExitCb(VerilatedTrace<VL_DERIVED_T>::onExit, this);
Verilated::addFlushCb(VerilatedTrace<VL_SUB_T, VL_BUF_T>::onFlush, this);
Verilated::addExitCb(VerilatedTrace<VL_SUB_T, VL_BUF_T>::onExit, this);
#ifdef VL_TRACE_THREADED
// Compute trace buffer size. we need to be able to store a new value for
// each signal, which is 'nextCode()' entries after the init callbacks
// above have been run, plus up to 2 more words of metadata per signal,
// plus fixed overhead of 1 for a termination flag and 3 for a time stamp
// update.
m_traceBufferSize = nextCode() + numSignals() * 2 + 4;
#ifdef VL_THREADED
if (offload()) {
// Compute offload buffer size. we need to be able to store a new value for
// each signal, which is 'nextCode()' entries after the init callbacks
// above have been run, plus up to 2 more words of metadata per signal,
// plus fixed overhead of 1 for a termination flag and 3 for a time stamp
// update.
m_offloadBufferSize = nextCode() + numSignals() * 2 + 4;
// Start the worker thread
m_workerThread.reset(new std::thread{&VerilatedTrace<VL_DERIVED_T>::workerThreadMain, this});
// Start the worker thread
m_workerThread.reset(
new std::thread{&VerilatedTrace<VL_SUB_T, VL_BUF_T>::offloadWorkerThreadMain, this});
}
#endif
}
template <>
bool VerilatedTrace<VL_DERIVED_T>::declCode(uint32_t code, const char* namep, uint32_t bits,
bool tri) {
bool VerilatedTrace<VL_SUB_T, VL_BUF_T>::declCode(uint32_t code, const char* namep, uint32_t bits,
bool tri) {
if (VL_UNCOVERABLE(!code)) {
VL_FATAL_MT(__FILE__, __LINE__, "", "Internal: internal trace problem, code 0 is illegal");
}
@@ -421,28 +430,32 @@ bool VerilatedTrace<VL_DERIVED_T>::declCode(uint32_t code, const char* namep, ui
//=========================================================================
// Internals available to format specific implementations
template <> std::string VerilatedTrace<VL_DERIVED_T>::timeResStr() const {
template <>
std::string VerilatedTrace<VL_SUB_T, VL_BUF_T>::timeResStr() const {
return doubleToTimescale(m_timeRes);
}
//=========================================================================
// External interface to client code
template <> void VerilatedTrace<VL_DERIVED_T>::set_time_unit(const char* unitp) VL_MT_SAFE {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::set_time_unit(const char* unitp) VL_MT_SAFE {
m_timeUnit = timescaleToDouble(unitp);
}
template <> void VerilatedTrace<VL_DERIVED_T>::set_time_unit(const std::string& unit) VL_MT_SAFE {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::set_time_unit(const std::string& unit) VL_MT_SAFE {
set_time_unit(unit.c_str());
}
template <> void VerilatedTrace<VL_DERIVED_T>::set_time_resolution(const char* unitp) VL_MT_SAFE {
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::set_time_resolution(const char* unitp) VL_MT_SAFE {
m_timeRes = timescaleToDouble(unitp);
}
template <>
void VerilatedTrace<VL_DERIVED_T>::set_time_resolution(const std::string& unit) VL_MT_SAFE {
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::set_time_resolution(const std::string& unit) VL_MT_SAFE {
set_time_resolution(unit.c_str());
}
template <>
void VerilatedTrace<VL_DERIVED_T>::dumpvars(int level, const std::string& hier) VL_MT_SAFE {
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::dumpvars(int level, const std::string& hier) VL_MT_SAFE {
if (level == 0) {
m_dumpvars.clear(); // empty = everything on
} else {
@@ -455,18 +468,114 @@ void VerilatedTrace<VL_DERIVED_T>::dumpvars(int level, const std::string& hier)
}
}
template <> void VerilatedTrace<VL_DERIVED_T>::dump(uint64_t timeui) VL_MT_SAFE_EXCLUDES(m_mutex) {
#ifdef VL_THREADED
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::parallelWorkerTask(void* datap, bool) {
ParallelWorkerData* const wdp = reinterpret_cast<ParallelWorkerData*>(datap);
// Run the task
wdp->m_cb(wdp->m_userp, wdp->m_bufp);
// Mark buffer as ready
const VerilatedLockGuard lock{wdp->m_mutex};
wdp->m_ready.store(true);
if (wdp->m_waiting) wdp->m_cv.notify_one();
}
template <>
VL_ATTR_NOINLINE void VerilatedTrace<VL_SUB_T, VL_BUF_T>::ParallelWorkerData::wait() {
// Spin for a while, waiting for the buffer to become ready
for (int i = 0; i < VL_LOCK_SPINS; ++i) {
if (VL_LIKELY(m_ready.load(std::memory_order_relaxed))) return;
VL_CPU_RELAX();
}
// We have been spinning for a while, so yield the thread
VerilatedLockGuard lock{m_mutex};
m_waiting = true;
m_cv.wait(lock, [this] { return m_ready.load(std::memory_order_relaxed); });
m_waiting = false;
}
#endif
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::runCallbacks(const std::vector<CallbackRecord>& cbVec) {
#ifdef VL_THREADED
if (parallel()) {
// If tracing in parallel, dispatch to the thread pool
VlThreadPool* threadPoolp = static_cast<VlThreadPool*>(m_contextp->threadPoolp());
// List of work items for thread (std::list, as ParallelWorkerData is not movable)
std::list<ParallelWorkerData> workerData;
// We use the whole pool + the main thread
const unsigned threads = threadPoolp->numThreads() + 1;
// Main thread executes all jobs with index % threads == 0
std::vector<ParallelWorkerData*> mainThreadWorkerData;
// Enuque all the jobs
for (unsigned i = 0; i < cbVec.size(); ++i) {
const CallbackRecord& cbr = cbVec[i];
// Always get the trace buffer on the main thread
Buffer* const bufp = getTraceBuffer();
// Create new work item
workerData.emplace_back(cbr.m_dumpCb, cbr.m_userp, bufp);
// Grab the new work item
ParallelWorkerData* const itemp = &workerData.back();
// Enqueue task to thread pool, or main thread
if (unsigned rem = i % threads) {
threadPoolp->workerp(rem - 1)->addTask(parallelWorkerTask, itemp);
} else {
mainThreadWorkerData.push_back(itemp);
}
}
// Execute main thead jobs
for (ParallelWorkerData* const itemp : mainThreadWorkerData) {
parallelWorkerTask(itemp, false);
}
// Commit all trace buffers in order
for (ParallelWorkerData& item : workerData) {
// Wait until ready
item.wait();
// Commit the buffer
commitTraceBuffer(item.m_bufp);
}
// Done
return;
}
#endif
// Fall back on sequential execution
for (const CallbackRecord& cbr : cbVec) {
Buffer* const traceBufferp = getTraceBuffer();
cbr.m_dumpCb(cbr.m_userp, traceBufferp);
commitTraceBuffer(traceBufferp);
}
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::runOffloadedCallbacks(
const std::vector<CallbackRecord>& cbVec) {
// Fall back on sequential execution
for (const CallbackRecord& cbr : cbVec) {
#ifdef VL_THREADED
Buffer* traceBufferp = getTraceBuffer();
cbr.m_dumpOffloadCb(cbr.m_userp, static_cast<OffloadBuffer*>(traceBufferp));
commitTraceBuffer(traceBufferp);
#else
VL_FATAL_MT(__FILE__, __LINE__, "", "Unreachable");
#endif
}
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::dump(uint64_t timeui) VL_MT_SAFE_EXCLUDES(m_mutex) {
// Not really VL_MT_SAFE but more VL_MT_UNSAFE_ONE.
// This does get the mutex, but if multiple threads are trying to dump
// chances are the data being dumped will have other problems
const VerilatedLockGuard lock{m_mutex};
if (VL_UNCOVERABLE(m_timeLastDump && timeui <= m_timeLastDump)) { // LCOV_EXCL_START
if (VL_UNCOVERABLE(m_didSomeDump && timeui <= m_timeLastDump)) { // LCOV_EXCL_START
VL_PRINTF_MT("%%Warning: previous dump at t=%" PRIu64 ", requesting t=%" PRIu64
", dump call ignored\n",
m_timeLastDump, timeui);
return;
} // LCOV_EXCL_STOP
m_timeLastDump = timeui;
m_didSomeDump = true;
Verilated::quiesce();
@@ -477,58 +586,64 @@ template <> void VerilatedTrace<VL_DERIVED_T>::dump(uint64_t timeui) VL_MT_SAFE_
if (!preChangeDump()) return;
}
#ifdef VL_TRACE_THREADED
// Currently only incremental dumps run on the worker thread
uint32_t* bufferp = nullptr;
if (VL_LIKELY(!m_fullDump)) {
// Get the trace buffer we are about to fill
bufferp = getTraceBuffer();
m_traceBufferWritep = bufferp;
m_traceBufferEndp = bufferp + m_traceBufferSize;
if (offload()) {
#ifdef VL_THREADED
// Currently only incremental dumps run on the worker thread
if (VL_LIKELY(!m_fullDump)) {
// Get the offload buffer we are about to fill
bufferp = getOffloadBuffer();
m_offloadBufferWritep = bufferp;
m_offloadBufferEndp = bufferp + m_offloadBufferSize;
// Tell worker to update time point
m_traceBufferWritep[0] = VerilatedTraceCommand::TIME_CHANGE;
*reinterpret_cast<uint64_t*>(m_traceBufferWritep + 1) = timeui;
m_traceBufferWritep += 3;
// Tell worker to update time point
m_offloadBufferWritep[0] = VerilatedTraceOffloadCommand::TIME_CHANGE;
*reinterpret_cast<uint64_t*>(m_offloadBufferWritep + 1) = timeui;
m_offloadBufferWritep += 3;
} else {
// Update time point
flushBase();
emitTimeChange(timeui);
}
#else
VL_FATAL_MT(__FILE__, __LINE__, "", "Unreachable");
#endif
} else {
// Update time point
flushBase();
emitTimeChange(timeui);
}
#else
// Update time point
emitTimeChange(timeui);
#endif
// Run the callbacks
if (VL_UNLIKELY(m_fullDump)) {
m_fullDump = false; // No more need for next dump to be full
for (uint32_t i = 0; i < m_fullCbs.size(); ++i) {
const CallbackRecord& cbr = m_fullCbs[i];
cbr.m_dumpCb(cbr.m_userp, self());
if (offload()) {
runOffloadedCallbacks(m_fullOffloadCbs);
} else {
runCallbacks(m_fullCbs);
}
} else {
for (uint32_t i = 0; i < m_chgCbs.size(); ++i) {
const CallbackRecord& cbr = m_chgCbs[i];
cbr.m_dumpCb(cbr.m_userp, self());
if (offload()) {
runOffloadedCallbacks(m_chgOffloadCbs);
} else {
runCallbacks(m_chgCbs);
}
}
for (uint32_t i = 0; i < m_cleanupCbs.size(); ++i) {
const CallbackRecord& cbr = m_cleanupCbs[i];
cbr.m_dumpCb(cbr.m_userp, self());
cbr.m_cleanupCb(cbr.m_userp, self());
}
#ifdef VL_TRACE_THREADED
if (VL_LIKELY(bufferp)) {
// Mark end of the trace buffer we just filled
*m_traceBufferWritep++ = VerilatedTraceCommand::END;
#ifdef VL_THREADED
if (offload() && VL_LIKELY(bufferp)) {
// Mark end of the offload buffer we just filled
*m_offloadBufferWritep++ = VerilatedTraceOffloadCommand::END;
// Assert no buffer overflow
assert(m_traceBufferWritep - bufferp <= m_traceBufferSize);
assert(m_offloadBufferWritep - bufferp <= m_offloadBufferSize);
// Pass it to the worker thread
m_buffersToWorker.put(bufferp);
m_offloadBuffersToWorker.put(bufferp);
}
#endif
}
@@ -537,101 +652,100 @@ template <> void VerilatedTrace<VL_DERIVED_T>::dump(uint64_t timeui) VL_MT_SAFE_
// Non-hot path internal interface to Verilator generated code
template <>
void VerilatedTrace<VL_DERIVED_T>::addCallbackRecord(std::vector<CallbackRecord>& cbVec,
CallbackRecord& cbRec)
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addModel(VerilatedModel* modelp)
VL_MT_SAFE_EXCLUDES(m_mutex) {
const VerilatedLockGuard lock{m_mutex};
if (VL_UNCOVERABLE(timeLastDump() != 0)) { // LCOV_EXCL_START
const std::string msg = (std::string{"Internal: "} + __FILE__ + "::" + __FUNCTION__
+ " called with already open file");
VL_FATAL_MT(__FILE__, __LINE__, "", msg.c_str());
const bool firstModel = m_models.empty();
const bool newModel = m_models.insert(modelp).second;
VerilatedContext* const contextp = modelp->contextp();
// Validate
if (!newModel) { // LCOV_EXCL_START
VL_FATAL_MT(__FILE__, __LINE__, "",
"The same model has already been added to this trace file");
}
if (VL_UNCOVERABLE(m_contextp && contextp != m_contextp)) {
VL_FATAL_MT(__FILE__, __LINE__, "",
"A trace file instance can only handle models from the same context");
}
if (VL_UNCOVERABLE(m_didSomeDump)) {
VL_FATAL_MT(__FILE__, __LINE__, "",
"Cannot add models to a trace file if 'dump' has already been called");
} // LCOV_EXCL_STOP
cbVec.push_back(cbRec);
}
template <> void VerilatedTrace<VL_DERIVED_T>::addInitCb(initCb_t cb, void* userp) VL_MT_SAFE {
CallbackRecord cbr{cb, userp};
addCallbackRecord(m_initCbs, cbr);
}
template <> void VerilatedTrace<VL_DERIVED_T>::addFullCb(dumpCb_t cb, void* userp) VL_MT_SAFE {
CallbackRecord cbr{cb, userp};
addCallbackRecord(m_fullCbs, cbr);
}
template <> void VerilatedTrace<VL_DERIVED_T>::addChgCb(dumpCb_t cb, void* userp) VL_MT_SAFE {
CallbackRecord cbr{cb, userp};
addCallbackRecord(m_chgCbs, cbr);
}
template <> void VerilatedTrace<VL_DERIVED_T>::addCleanupCb(dumpCb_t cb, void* userp) VL_MT_SAFE {
CallbackRecord cbr{cb, userp};
addCallbackRecord(m_cleanupCbs, cbr);
}
// Keep hold of the context
m_contextp = contextp;
template <> void VerilatedTrace<VL_DERIVED_T>::pushNamePrefix(const std::string& prefix) {
m_namePrefixStack.push_back(m_namePrefixStack.back() + prefix);
}
// Get the desired trace config from the model
const std::unique_ptr<VerilatedTraceConfig> configp = modelp->traceConfig();
#ifndef VL_THREADED
if (configp->m_useOffloading) {
VL_FATAL_MT(__FILE__, __LINE__, "", "Cannot use trace offloading without VL_THREADED");
}
#endif
template <> void VerilatedTrace<VL_DERIVED_T>::popNamePrefix(unsigned count) {
while (count--) m_namePrefixStack.pop_back();
assert(!m_namePrefixStack.empty());
}
// Configure trace base class
if (!firstModel) {
if (m_offload != configp->m_useOffloading) {
VL_FATAL_MT(__FILE__, __LINE__, "",
"Either all or no models using the same trace file must use offloading");
}
}
m_offload = configp->m_useOffloading;
// If at least one model requests parallel tracing, then use it
m_parallel |= configp->m_useParallel;
//=========================================================================
// Hot path internal interface to Verilator generated code
if (VL_UNCOVERABLE(m_parallel && m_offload)) { // LCOV_EXCL_START
VL_FATAL_MT(__FILE__, __LINE__, "", "Cannot use parallel tracing with offloading");
} // LCOV_EXCL_STOP
// These functions must write the new value back into the old value store,
// and subsequently call the format specific emit* implementations. Note
// that this file must be included in the format specific implementation, so
// the emit* functions can be inlined for performance.
template <> void VerilatedTrace<VL_DERIVED_T>::fullBit(uint32_t* oldp, CData newval) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitBit(code, newval);
}
template <> void VerilatedTrace<VL_DERIVED_T>::fullCData(uint32_t* oldp, CData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitCData(code, newval, bits);
}
template <> void VerilatedTrace<VL_DERIVED_T>::fullSData(uint32_t* oldp, SData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitSData(code, newval, bits);
}
template <> void VerilatedTrace<VL_DERIVED_T>::fullIData(uint32_t* oldp, IData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitIData(code, newval, bits);
}
template <> void VerilatedTrace<VL_DERIVED_T>::fullQData(uint32_t* oldp, QData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*reinterpret_cast<QData*>(oldp) = newval;
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitQData(code, newval, bits);
// Configure format specific sub class
configure(*(configp.get()));
}
template <>
void VerilatedTrace<VL_DERIVED_T>::fullWData(uint32_t* oldp, const WData* newvalp, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
for (int i = 0; i < VL_WORDS_I(bits); ++i) oldp[i] = newvalp[i];
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
self()->emitWData(code, newvalp, bits);
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addCallbackRecord(std::vector<CallbackRecord>& cbVec,
CallbackRecord&& cbRec)
VL_MT_SAFE_EXCLUDES(m_mutex) {
const VerilatedLockGuard lock{m_mutex};
cbVec.push_back(cbRec);
}
template <> void VerilatedTrace<VL_DERIVED_T>::fullDouble(uint32_t* oldp, double newval) {
const uint32_t code = oldp - m_sigs_oldvalp;
*reinterpret_cast<double*>(oldp) = newval;
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
// cppcheck-suppress invalidPointerCast
self()->emitDouble(code, newval);
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addInitCb(initCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_initCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addFullCb(dumpCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_fullCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addFullCb(dumpOffloadCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_fullOffloadCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addChgCb(dumpCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_chgCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addChgCb(dumpOffloadCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_chgOffloadCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::addCleanupCb(cleanupCb_t cb, void* userp) VL_MT_SAFE {
addCallbackRecord(m_cleanupCbs, CallbackRecord{cb, userp});
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::pushNamePrefix(const std::string& prefix) {
m_namePrefixStack.push_back(m_namePrefixStack.back() + prefix);
}
template <>
void VerilatedTrace<VL_SUB_T, VL_BUF_T>::popNamePrefix(unsigned count) {
while (count--) m_namePrefixStack.pop_back();
assert(!m_namePrefixStack.empty());
}
//=========================================================================
@@ -724,41 +838,96 @@ static inline void cvtQDataToStr(char* dstp, QData value) {
#define cvtEDataToStr cvtIDataToStr
//=============================================================================
//=========================================================================
// VerilatedTraceBuffer
#ifdef VERILATED_VCD_TEST
template <>
VerilatedTraceBuffer<VL_BUF_T>::VerilatedTraceBuffer(Trace& owner)
: VL_BUF_T{owner}
, m_sigs_oldvalp{owner.m_sigs_oldvalp}
, m_sigs_enabledp{owner.m_sigs_enabledp} {}
void verilated_trace_imp_selftest() {
#define SELF_CHECK(got, exp) \
do { \
if ((got) != (exp)) VL_FATAL_MT(__FILE__, __LINE__, "", "%Error: selftest"); \
} while (0)
// These functions must write the new value back into the old value store,
// and subsequently call the format specific emit* implementations. Note
// that this file must be included in the format specific implementation, so
// the emit* functions can be inlined for performance.
#define SELF_CHECK_TS(scale) \
SELF_CHECK(doubleToTimescale(timescaleToDouble(scale)), std::string{scale});
SELF_CHECK_TS("100s");
SELF_CHECK_TS("10s");
SELF_CHECK_TS("1s");
SELF_CHECK_TS("100ms");
SELF_CHECK_TS("10ms");
SELF_CHECK_TS("1ms");
SELF_CHECK_TS("100us");
SELF_CHECK_TS("10us");
SELF_CHECK_TS("1us");
SELF_CHECK_TS("100ns");
SELF_CHECK_TS("10ns");
SELF_CHECK_TS("1ns");
SELF_CHECK_TS("100ps");
SELF_CHECK_TS("10ps");
SELF_CHECK_TS("1ps");
SELF_CHECK_TS("100fs");
SELF_CHECK_TS("10fs");
SELF_CHECK_TS("1fs");
SELF_CHECK_TS("100as");
SELF_CHECK_TS("10as");
SELF_CHECK_TS("1as");
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullBit(uint32_t* oldp, CData newval) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitBit(code, newval);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullCData(uint32_t* oldp, CData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitCData(code, newval, bits);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullSData(uint32_t* oldp, SData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitSData(code, newval, bits);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullIData(uint32_t* oldp, IData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*oldp = newval; // Still copy even if not tracing so chg doesn't call full
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitIData(code, newval, bits);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullQData(uint32_t* oldp, QData newval, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
*reinterpret_cast<QData*>(oldp) = newval;
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitQData(code, newval, bits);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullWData(uint32_t* oldp, const WData* newvalp, int bits) {
const uint32_t code = oldp - m_sigs_oldvalp;
for (int i = 0; i < VL_WORDS_I(bits); ++i) oldp[i] = newvalp[i];
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
emitWData(code, newvalp, bits);
}
template <>
void VerilatedTraceBuffer<VL_BUF_T>::fullDouble(uint32_t* oldp, double newval) {
const uint32_t code = oldp - m_sigs_oldvalp;
*reinterpret_cast<double*>(oldp) = newval;
if (VL_UNLIKELY(m_sigs_enabledp && !(VL_BITISSET_W(m_sigs_enabledp, code)))) return;
// cppcheck-suppress invalidPointerCast
emitDouble(code, newval);
}
#ifdef VL_THREADED
//=========================================================================
// VerilatedTraceOffloadBuffer
template <>
VerilatedTraceOffloadBuffer<VL_BUF_T>::VerilatedTraceOffloadBuffer(VL_SUB_T& owner)
: VerilatedTraceBuffer<VL_BUF_T>{owner}
, m_offloadBufferWritep{owner.m_offloadBufferWritep}
, m_offloadBufferEndp{owner.m_offloadBufferEndp} {
if (m_offloadBufferWritep) {
using This = VerilatedTraceBuffer<VL_BUF_T>*;
// Tack on the buffer address
static_assert(2 * sizeof(uint32_t) >= sizeof(This),
"This should be enough on all plafrorms");
*m_offloadBufferWritep++ = VerilatedTraceOffloadCommand::TRACE_BUFFER;
*reinterpret_cast<This*>(m_offloadBufferWritep) = static_cast<This>(this);
m_offloadBufferWritep += 2;
}
}
#endif
#endif // VL_CPPCHECK
+69 -38
View File
@@ -91,9 +91,9 @@ class VlReadMem final {
const int m_bits; // Bit width of values
const std::string& m_filename; // Filename
const QData m_end; // End address (as specified by user)
FILE* m_fp; // File handle for filename
QData m_addr; // Next address to read
int m_linenum; // Line number last read from file
FILE* m_fp = nullptr; // File handle for filename
QData m_addr = 0; // Next address to read
int m_linenum = 0; // Line number last read from file
bool m_anyAddr = false; // Had address directive in the file
public:
VlReadMem(bool hex, int bits, const std::string& filename, QData start, QData end);
@@ -107,8 +107,8 @@ public:
class VlWriteMem final {
const bool m_hex; // Hex format
const int m_bits; // Bit width of values
FILE* m_fp; // File handle for filename
QData m_addr; // Next address to write
FILE* m_fp = nullptr; // File handle for filename
QData m_addr = 0; // Next address to write
public:
VlWriteMem(bool hex, int bits, const std::string& filename, QData start, QData end);
~VlWriteMem();
@@ -131,7 +131,8 @@ public:
static int _vl_cmp_w(int words, WDataInP const lwp, WDataInP const rwp) VL_MT_SAFE;
template <std::size_t T_Words> struct VlWide final {
template <std::size_t T_Words>
struct VlWide final {
// MEMBERS
// This should be the only data member, otherwise generated static initializers need updating
EData m_storage[T_Words]; // Contents of the packed array
@@ -163,7 +164,8 @@ VlWide<T_Words>& VL_CVT_W_A(const WDataInP inp, const VlWide<T_Words>&) {
return *((VlWide<T_Words>*)inp);
}
template <std::size_t T_Words> std::string VL_TO_STRING(const VlWide<T_Words>& obj) {
template <std::size_t T_Words>
std::string VL_TO_STRING(const VlWide<T_Words>& obj) {
return VL_TO_STRING_W(T_Words, obj.data());
}
@@ -174,7 +176,8 @@ template <std::size_t T_Words> std::string VL_TO_STRING(const VlWide<T_Words>& o
//
// Bound here is the maximum size() allowed, e.g. 1 + SystemVerilog bound
// For dynamic arrays it is always zero
template <class T_Value, size_t T_MaxSize = 0> class VlQueue final {
template <class T_Value, size_t T_MaxSize = 0>
class VlQueue final {
private:
// TYPES
using Deque = std::deque<T_Value>;
@@ -199,7 +202,8 @@ public:
// Standard copy constructor works. Verilog: assoca = assocb
// Also must allow conversion from a different T_MaxSize queue
template <size_t U_MaxSize = 0> VlQueue operator=(const VlQueue<T_Value, U_MaxSize>& rhs) {
template <size_t U_MaxSize = 0>
VlQueue operator=(const VlQueue<T_Value, U_MaxSize>& rhs) {
m_deque = rhs.privateDeque();
if (VL_UNLIKELY(T_MaxSize && T_MaxSize < m_deque.size())) m_deque.resize(T_MaxSize - 1);
return *this;
@@ -289,7 +293,7 @@ public:
// Can't just overload operator[] or provide a "at" reference to set,
// because we need to be able to insert only when the value is set
T_Value& at(int32_t index) {
static T_Value s_throwAway;
static VL_THREAD_LOCAL T_Value s_throwAway;
// Needs to work for dynamic arrays, so does not use T_MaxSize
if (VL_UNLIKELY(index < 0 || index >= m_deque.size())) {
s_throwAway = atDefault();
@@ -300,7 +304,7 @@ public:
}
// Accessing. Verilog: v = assoc[index]
const T_Value& at(int32_t index) const {
static T_Value s_throwAway;
static VL_THREAD_LOCAL T_Value s_throwAway;
// Needs to work for dynamic arrays, so does not use T_MaxSize
if (VL_UNLIKELY(index < 0 || index >= m_deque.size())) {
return atDefault();
@@ -330,7 +334,8 @@ public:
// Methods
void sort() { std::sort(m_deque.begin(), m_deque.end()); }
template <typename Func> void sort(Func with_func) {
template <typename Func>
void sort(Func with_func) {
// with_func returns arbitrary type to use for the sort comparison
std::sort(m_deque.begin(), m_deque.end(), [=](const T_Value& a, const T_Value& b) {
// index number is meaninless with sort, as it changes
@@ -338,7 +343,8 @@ public:
});
}
void rsort() { std::sort(m_deque.rbegin(), m_deque.rend()); }
template <typename Func> void rsort(Func with_func) {
template <typename Func>
void rsort(Func with_func) {
// with_func returns arbitrary type to use for the sort comparison
std::sort(m_deque.rbegin(), m_deque.rend(), [=](const T_Value& a, const T_Value& b) {
// index number is meaninless with sort, as it changes
@@ -373,7 +379,8 @@ public:
}
return out;
}
template <typename Func> VlQueue find(Func with_func) const {
template <typename Func>
VlQueue find(Func with_func) const {
VlQueue out;
IData index = 0;
for (const auto& i : m_deque) {
@@ -382,7 +389,8 @@ public:
}
return out;
}
template <typename Func> VlQueue<IData> find_index(Func with_func) const {
template <typename Func>
VlQueue<IData> find_index(Func with_func) const {
VlQueue<IData> out;
IData index = 0;
for (const auto& i : m_deque) {
@@ -391,7 +399,8 @@ public:
}
return out;
}
template <typename Func> VlQueue find_first(Func with_func) const {
template <typename Func>
VlQueue find_first(Func with_func) const {
// Can't use std::find_if as need index number
IData index = 0;
for (const auto& i : m_deque) {
@@ -400,7 +409,8 @@ public:
}
return VlQueue{};
}
template <typename Func> VlQueue<IData> find_first_index(Func with_func) const {
template <typename Func>
VlQueue<IData> find_first_index(Func with_func) const {
IData index = 0;
for (const auto& i : m_deque) {
if (with_func(index, i)) return VlQueue<IData>::cons(index);
@@ -408,7 +418,8 @@ public:
}
return VlQueue<IData>{};
}
template <typename Func> VlQueue find_last(Func with_func) const {
template <typename Func>
VlQueue find_last(Func with_func) const {
IData index = m_deque.size() - 1;
for (auto& item : vlstd::reverse_view(m_deque)) {
if (with_func(index, item)) return VlQueue::cons(item);
@@ -416,7 +427,8 @@ public:
}
return VlQueue{};
}
template <typename Func> VlQueue<IData> find_last_index(Func with_func) const {
template <typename Func>
VlQueue<IData> find_last_index(Func with_func) const {
IData index = m_deque.size() - 1;
for (auto& item : vlstd::reverse_view(m_deque)) {
if (with_func(index, item)) return VlQueue<IData>::cons(index);
@@ -442,7 +454,8 @@ public:
for (const auto& i : m_deque) out += i;
return out;
}
template <typename Func> T_Value r_sum(Func with_func) const {
template <typename Func>
T_Value r_sum(Func with_func) const {
T_Value out(0); // Type must have assignment operator
IData index = 0;
for (const auto& i : m_deque) out += with_func(index++, i);
@@ -456,7 +469,8 @@ public:
for (; it != m_deque.end(); ++it) out *= *it;
return out;
}
template <typename Func> T_Value r_product(Func with_func) const {
template <typename Func>
T_Value r_product(Func with_func) const {
if (m_deque.empty()) return T_Value(0);
auto it = m_deque.begin();
IData index = 0;
@@ -474,7 +488,8 @@ public:
for (; it != m_deque.end(); ++it) out &= *it;
return out;
}
template <typename Func> T_Value r_and(Func with_func) const {
template <typename Func>
T_Value r_and(Func with_func) const {
if (m_deque.empty()) return T_Value(0);
auto it = m_deque.begin();
IData index = 0;
@@ -489,7 +504,8 @@ public:
for (const auto& i : m_deque) out |= i;
return out;
}
template <typename Func> T_Value r_or(Func with_func) const {
template <typename Func>
T_Value r_or(Func with_func) const {
T_Value out(0); // Type must have assignment operator
IData index = 0;
for (const auto& i : m_deque) out |= with_func(index++, i);
@@ -500,7 +516,8 @@ public:
for (const auto& i : m_deque) out ^= i;
return out;
}
template <typename Func> T_Value r_xor(Func with_func) const {
template <typename Func>
T_Value r_xor(Func with_func) const {
T_Value out(0); // Type must have assignment operator
IData index = 0;
for (const auto& i : m_deque) out ^= with_func(index++, i);
@@ -520,7 +537,8 @@ public:
}
};
template <class T_Value> std::string VL_TO_STRING(const VlQueue<T_Value>& obj) {
template <class T_Value>
std::string VL_TO_STRING(const VlQueue<T_Value>& obj) {
return obj.to_string();
}
@@ -529,7 +547,8 @@ template <class T_Value> std::string VL_TO_STRING(const VlQueue<T_Value>& obj) {
// There are no multithreaded locks on this; the base variable must
// be protected by other means
//
template <class T_Key, class T_Value> class VlAssocArray final {
template <class T_Key, class T_Value>
class VlAssocArray final {
private:
// TYPES
using Map = std::map<T_Key, T_Value>;
@@ -654,19 +673,22 @@ public:
}
return out;
}
template <typename Func> VlQueue<T_Value> find(Func with_func) const {
template <typename Func>
VlQueue<T_Value> find(Func with_func) const {
VlQueue<T_Value> out;
for (const auto& i : m_map)
if (with_func(i.first, i.second)) out.push_back(i.second);
return out;
}
template <typename Func> VlQueue<T_Key> find_index(Func with_func) const {
template <typename Func>
VlQueue<T_Key> find_index(Func with_func) const {
VlQueue<T_Key> out;
for (const auto& i : m_map)
if (with_func(i.first, i.second)) out.push_back(i.first);
return out;
}
template <typename Func> VlQueue<T_Value> find_first(Func with_func) const {
template <typename Func>
VlQueue<T_Value> find_first(Func with_func) const {
const auto it
= std::find_if(m_map.begin(), m_map.end(), [=](const std::pair<T_Key, T_Value>& i) {
return with_func(i.first, i.second);
@@ -674,7 +696,8 @@ public:
if (it == m_map.end()) return VlQueue<T_Value>{};
return VlQueue<T_Value>::cons(it->second);
}
template <typename Func> VlQueue<T_Key> find_first_index(Func with_func) const {
template <typename Func>
VlQueue<T_Key> find_first_index(Func with_func) const {
const auto it
= std::find_if(m_map.begin(), m_map.end(), [=](const std::pair<T_Key, T_Value>& i) {
return with_func(i.first, i.second);
@@ -682,7 +705,8 @@ public:
if (it == m_map.end()) return VlQueue<T_Value>{};
return VlQueue<T_Key>::cons(it->first);
}
template <typename Func> VlQueue<T_Value> find_last(Func with_func) const {
template <typename Func>
VlQueue<T_Value> find_last(Func with_func) const {
const auto it
= std::find_if(m_map.rbegin(), m_map.rend(), [=](const std::pair<T_Key, T_Value>& i) {
return with_func(i.first, i.second);
@@ -690,7 +714,8 @@ public:
if (it == m_map.rend()) return VlQueue<T_Value>{};
return VlQueue<T_Value>::cons(it->second);
}
template <typename Func> VlQueue<T_Key> find_last_index(Func with_func) const {
template <typename Func>
VlQueue<T_Key> find_last_index(Func with_func) const {
const auto it
= std::find_if(m_map.rbegin(), m_map.rend(), [=](const std::pair<T_Key, T_Value>& i) {
return with_func(i.first, i.second);
@@ -724,7 +749,8 @@ public:
for (const auto& i : m_map) out += i.second;
return out;
}
template <typename Func> T_Value r_sum(Func with_func) const {
template <typename Func>
T_Value r_sum(Func with_func) const {
T_Value out(0); // Type must have assignment operator
for (const auto& i : m_map) out += with_func(i.first, i.second);
return out;
@@ -737,7 +763,8 @@ public:
for (; it != m_map.end(); ++it) out *= it->second;
return out;
}
template <typename Func> T_Value r_product(Func with_func) const {
template <typename Func>
T_Value r_product(Func with_func) const {
if (m_map.empty()) return T_Value(0);
auto it = m_map.begin();
T_Value out{with_func(it->first, it->second)};
@@ -753,7 +780,8 @@ public:
for (; it != m_map.end(); ++it) out &= it->second;
return out;
}
template <typename Func> T_Value r_and(Func with_func) const {
template <typename Func>
T_Value r_and(Func with_func) const {
if (m_map.empty()) return T_Value(0);
auto it = m_map.begin();
T_Value out{with_func(it->first, it->second)};
@@ -766,7 +794,8 @@ public:
for (const auto& i : m_map) out |= i.second;
return out;
}
template <typename Func> T_Value r_or(Func with_func) const {
template <typename Func>
T_Value r_or(Func with_func) const {
T_Value out(0); // Type must have assignment operator
for (const auto& i : m_map) out |= with_func(i.first, i.second);
return out;
@@ -776,7 +805,8 @@ public:
for (const auto& i : m_map) out ^= i.second;
return out;
}
template <typename Func> T_Value r_xor(Func with_func) const {
template <typename Func>
T_Value r_xor(Func with_func) const {
T_Value out(0); // Type must have assignment operator
for (const auto& i : m_map) out ^= with_func(i.first, i.second);
return out;
@@ -839,7 +869,8 @@ void VL_WRITEMEM_N(bool hex, int bits, const std::string& filename,
/// This class may get exposed to a Verilated Model's top I/O, if the top
/// IO has an unpacked array.
template <class T_Value, std::size_t T_Depth> struct VlUnpacked final {
template <class T_Value, std::size_t T_Depth>
struct VlUnpacked final {
// MEMBERS
// This should be the only data member, otherwise generated static initializers need updating
T_Value m_storage[T_Depth]; // Contents of the unpacked array
+164 -344
View File
@@ -38,13 +38,13 @@
# include <unistd.h>
#endif
#ifndef O_LARGEFILE // For example on WIN32
#ifndef O_LARGEFILE // WIN32 headers omit this
# define O_LARGEFILE 0
#endif
#ifndef O_NONBLOCK
#ifndef O_NONBLOCK // WIN32 headers omit this
# define O_NONBLOCK 0
#endif
#ifndef O_CLOEXEC
#ifndef O_CLOEXEC // WIN32 headers omit this
# define O_CLOEXEC 0
#endif
@@ -62,12 +62,24 @@ constexpr unsigned VL_TRACE_MAX_VCD_CODE_SIZE = 5; // Maximum length of a VCD s
// cache-lines.
constexpr unsigned VL_TRACE_SUFFIX_ENTRY_SIZE = 8; // Size of a suffix entry
//=============================================================================
// Utility functions: TODO: put these in a common place and share them.
template <size_t N>
static size_t roundUpToMultipleOf(size_t value) {
static_assert((N & (N - 1)) == 0, "'N' must be a power of 2");
size_t mask = N - 1;
return (value + mask) & ~mask;
}
//=============================================================================
// Specialization of the generics for this trace format
#define VL_DERIVED_T VerilatedVcd
#include "verilated_trace_imp.cpp"
#undef VL_DERIVED_T
#define VL_SUB_T VerilatedVcd
#define VL_BUF_T VerilatedVcdBuffer
#include "verilated_trace_imp.h"
#undef VL_SUB_T
#undef VL_BUF_T
//=============================================================================
//=============================================================================
@@ -108,13 +120,13 @@ void VerilatedVcd::open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex) {
// Set member variables
m_filename = filename; // "" is ok, as someone may overload open
openNextImp(m_rolloverMB != 0);
openNextImp(m_rolloverSize != 0);
if (!isOpen()) return;
dumpHeader();
// When using rollover, the first chunk contains the header only.
if (m_rolloverMB) openNextImp(true);
if (m_rolloverSize) openNextImp(true);
}
void VerilatedVcd::openNext(bool incFilename) VL_MT_SAFE_EXCLUDES(m_mutex) {
@@ -168,7 +180,7 @@ void VerilatedVcd::openNextImp(bool incFilename) {
}
bool VerilatedVcd::preChangeDump() {
if (VL_UNLIKELY(m_rolloverMB && m_wroteBytes > m_rolloverMB)) openNextImp(true);
if (VL_UNLIKELY(m_rolloverSize && m_wroteBytes > m_rolloverSize)) openNextImp(true);
return isOpen();
}
@@ -183,7 +195,7 @@ void VerilatedVcd::makeNameMap() {
deleteNameMap();
m_namemapp = new NameMap;
VerilatedTrace<VerilatedVcd>::traceInit();
Super::traceInit();
// Though not speced, it's illegal to generate a vcd with signals
// not under any module - it crashes at least two viewers.
@@ -218,13 +230,19 @@ VerilatedVcd::~VerilatedVcd() {
if (m_wrBufp) VL_DO_CLEAR(delete[] m_wrBufp, m_wrBufp = nullptr);
deleteNameMap();
if (m_filep && m_fileNewed) VL_DO_CLEAR(delete m_filep, m_filep = nullptr);
#ifdef VL_THREADED
if (parallel()) {
assert(m_numBuffers == m_freeBuffers.size());
for (auto& pair : m_freeBuffers) VL_DO_CLEAR(delete[] pair.first, pair.first = nullptr);
}
#endif
}
void VerilatedVcd::closePrev() {
// This function is on the flush() call path
if (!isOpen()) return;
VerilatedTrace<VerilatedVcd>::flushBase();
Super::flushBase();
bufferFlush();
m_isOpen = false;
m_filep->close();
@@ -245,20 +263,15 @@ void VerilatedVcd::close() VL_MT_SAFE_EXCLUDES(m_mutex) {
// This function is on the flush() call path
const VerilatedLockGuard lock{m_mutex};
if (!isOpen()) return;
if (m_evcd) {
printStr("$vcdclose ");
printQuad(timeLastDump());
printStr(" $end\n");
}
closePrev();
// closePrev() called VerilatedTrace<VerilatedVcd>::flush(), so we just
// closePrev() called Super::flush(), so we just
// need to shut down the tracing thread here.
VerilatedTrace<VerilatedVcd>::closeBase();
Super::closeBase();
}
void VerilatedVcd::flush() VL_MT_SAFE_EXCLUDES(m_mutex) {
const VerilatedLockGuard lock{m_mutex};
VerilatedTrace<VerilatedVcd>::flushBase();
Super::flushBase();
bufferFlush();
}
@@ -277,12 +290,12 @@ void VerilatedVcd::printQuad(uint64_t n) {
printStr(buf);
}
void VerilatedVcd::bufferResize(uint64_t minsize) {
void VerilatedVcd::bufferResize(size_t minsize) {
// minsize is size of largest write. We buffer at least 8 times as much data,
// writing when we are 3/4 full (with thus 2*minsize remaining free)
if (VL_UNLIKELY(minsize > m_wrChunkSize)) {
const char* oldbufp = m_wrBufp;
m_wrChunkSize = minsize * 2;
m_wrChunkSize = roundUpToMultipleOf<1024>(minsize * 2);
m_wrBufp = new char[m_wrChunkSize * 8];
std::memcpy(m_wrBufp, oldbufp, m_writep - oldbufp);
m_writep = m_wrBufp + (m_writep - oldbufp);
@@ -292,7 +305,7 @@ void VerilatedVcd::bufferResize(uint64_t minsize) {
}
void VerilatedVcd::bufferFlush() VL_MT_UNSAFE_ONE {
// This function can be called from the trace thread
// This function can be called from the trace offload thread
// This function is on the flush() call path
// We add output data to m_writep.
// When it gets nearly full we dump it using this routine which calls write()
@@ -463,14 +476,16 @@ void VerilatedVcd::declare(uint32_t code, const char* name, const char* wirep, b
int arraynum, bool tri, bool bussed, int msb, int lsb) {
const int bits = ((msb > lsb) ? (msb - lsb) : (lsb - msb)) + 1;
const bool enabled = VerilatedTrace<VerilatedVcd>::declCode(code, name, bits, tri);
const bool enabled = Super::declCode(code, name, bits, tri);
if (m_suffixes.size() <= nextCode() * VL_TRACE_SUFFIX_ENTRY_SIZE) {
m_suffixes.resize(nextCode() * VL_TRACE_SUFFIX_ENTRY_SIZE * 2, 0);
}
// Make sure write buffer is large enough (one character per bit), plus header
bufferResize(bits + 1024);
// Keep upper bound on bytes a single signal cna emit into the buffer
m_maxSignalBytes = std::max<size_t>(m_maxSignalBytes, bits + 32);
// Make sure write buffer is large enough, plus header
bufferResize(m_maxSignalBytes + 1024);
if (!enabled) return;
@@ -497,38 +512,29 @@ void VerilatedVcd::declare(uint32_t code, const char* name, const char* wirep, b
// Print reference
std::string decl = "$var ";
if (m_evcd) {
decl += "port";
} else {
decl += wirep; // usually "wire"
}
decl += wirep; // usually "wire"
constexpr size_t bufsize = 1000;
char buf[bufsize];
VL_SNPRINTF(buf, bufsize, " %2d ", bits);
decl += buf;
if (m_evcd) {
VL_SNPRINTF(buf, bufsize, "<%u", code);
decl += buf;
} else {
// Add string code to decl
char* const endp = writeCode(buf, code);
*endp = '\0';
decl += buf;
// Build suffix array entry
char* const entryp = &m_suffixes[code * VL_TRACE_SUFFIX_ENTRY_SIZE];
const size_t length = endp - buf;
assert(length <= VL_TRACE_MAX_VCD_CODE_SIZE);
// 1 bit values don't have a ' ' separator between value and string code
const bool isBit = bits == 1;
entryp[0] = ' '; // Separator
// Use memcpy as we checked size above, and strcpy is flagged unsafe
std::memcpy(entryp + !isBit, buf,
std::strlen(buf)); // Code (overwrite separator if isBit)
entryp[length + !isBit] = '\n'; // Replace '\0' with line termination '\n'
// Set length of suffix (used to increment write pointer)
entryp[VL_TRACE_SUFFIX_ENTRY_SIZE - 1] = !isBit + length + 1;
}
// Add string code to decl
char* const endp = writeCode(buf, code);
*endp = '\0';
decl += buf;
// Build suffix array entry
char* const entryp = &m_suffixes[code * VL_TRACE_SUFFIX_ENTRY_SIZE];
const size_t length = endp - buf;
assert(length <= VL_TRACE_MAX_VCD_CODE_SIZE);
// 1 bit values don't have a ' ' separator between value and string code
const bool isBit = bits == 1;
entryp[0] = ' '; // Separator
// Use memcpy as we checked size above, and strcpy is flagged unsafe
std::memcpy(entryp + !isBit, buf,
std::strlen(buf)); // Code (overwrite separator if isBit)
entryp[length + !isBit] = '\n'; // Replace '\0' with line termination '\n'
// Set length of suffix (used to increment write pointer)
entryp[VL_TRACE_SUFFIX_ENTRY_SIZE - 1] = !isBit + length + 1;
decl += " ";
decl += basename;
if (array) {
@@ -562,26 +568,69 @@ void VerilatedVcd::declArray(uint32_t code, const char* name, bool array, int ar
void VerilatedVcd::declDouble(uint32_t code, const char* name, bool array, int arraynum) {
declare(code, name, "real", array, arraynum, false, false, 63, 0);
}
#ifdef VL_TRACE_VCD_OLD_API
void VerilatedVcd::declTriBit(uint32_t code, const char* name, bool array, int arraynum) {
declare(code, name, "wire", array, arraynum, true, false, 0, 0);
}
void VerilatedVcd::declTriBus(uint32_t code, const char* name, bool array, int arraynum, int msb,
int lsb) {
declare(code, name, "wire", array, arraynum, true, true, msb, lsb);
}
void VerilatedVcd::declTriQuad(uint32_t code, const char* name, bool array, int arraynum, int msb,
int lsb) {
declare(code, name, "wire", array, arraynum, true, true, msb, lsb);
}
void VerilatedVcd::declTriArray(uint32_t code, const char* name, bool array, int arraynum, int msb,
int lsb) {
declare(code, name, "wire", array, arraynum, true, true, msb, lsb);
}
#endif // VL_TRACE_VCD_OLD_API
//=============================================================================
// Trace rendering prinitives
// Get/commit trace buffer
VerilatedVcd::Buffer* VerilatedVcd::getTraceBuffer() {
VerilatedVcd::Buffer* const bufp = new Buffer{*this};
#ifdef VL_THREADED
if (parallel()) {
// Note: This is called from VeriltedVcd::dump, which already holds the lock
// If no buffer available, allocate a new one
if (m_freeBuffers.empty()) {
constexpr size_t pageSize = 4096;
// 4 * m_maxSignalBytes, so we can reserve 2 * m_maxSignalBytes at the end for safety
size_t startingSize = roundUpToMultipleOf<pageSize>(4 * m_maxSignalBytes);
m_freeBuffers.emplace_back(new char[startingSize], startingSize);
++m_numBuffers;
}
// Grab a buffer
const auto pair = m_freeBuffers.back();
m_freeBuffers.pop_back();
// Initialize
bufp->m_writep = bufp->m_bufp = pair.first;
bufp->m_size = pair.second;
bufp->adjustGrowp();
}
#endif
// Return the buffer
return bufp;
}
void VerilatedVcd::commitTraceBuffer(VerilatedVcd::Buffer* bufp) {
if (parallel()) {
#if VL_THREADED
// Note: This is called from VeriltedVcd::dump, which already holds the lock
// Resize output buffer. Note, we use the full size of the trace buffer, as
// this is a lot more stable than the actual occupancy of the trace buffer.
// This helps us to avoid re-allocations due to small size changes.
bufferResize(bufp->m_size);
// Compute occupancy of buffer
const size_t usedSize = bufp->m_writep - bufp->m_bufp;
// Copy to output buffer
std::memcpy(m_writep, bufp->m_bufp, usedSize);
// Adjust write pointer
m_writep += usedSize;
// Flush if necessary
bufferCheck();
// Put buffer back on free list
m_freeBuffers.emplace_back(bufp->m_bufp, bufp->m_size);
#else
VL_FATAL_MT(__FILE__, __LINE__, "", "Unreachable");
#endif
} else {
// Needs adjusting for emitTimeChange
m_writep = bufp->m_writep;
}
delete bufp;
}
//=============================================================================
// VerilatedVcdBuffer implementation
//=============================================================================
// Trace rendering primitives
static inline void
VerilatedVcdCCopyAndAppendNewLine(char* writep, const char* suffixp) VL_ATTR_NO_SANITIZE_ALIGN;
@@ -606,26 +655,59 @@ static inline void VerilatedVcdCCopyAndAppendNewLine(char* writep, const char* s
#endif
}
void VerilatedVcd::finishLine(uint32_t code, char* writep) {
const char* const suffixp = m_suffixes.data() + code * VL_TRACE_SUFFIX_ENTRY_SIZE;
void VerilatedVcdBuffer::finishLine(uint32_t code, char* writep) {
const char* const suffixp = m_suffixes + code * VL_TRACE_SUFFIX_ENTRY_SIZE;
VL_DEBUG_IFDEF(assert(suffixp[0]););
VerilatedVcdCCopyAndAppendNewLine(writep, suffixp);
// Now write back the write pointer incremented by the actual size of the
// suffix, which was stored in the last byte of the suffix buffer entry.
m_writep = writep + suffixp[VL_TRACE_SUFFIX_ENTRY_SIZE - 1];
bufferCheck();
if (m_owner.parallel()) {
#ifdef VL_THREADED
// Double the size of the buffer if necessary
if (VL_UNLIKELY(m_writep >= m_growp)) {
// Compute occupied size of current buffer
const size_t usedSize = m_writep - m_bufp;
// We are always doubling the size
m_size *= 2;
// Allocate the new buffer
char* const newBufp = new char[m_size];
// Copy from current buffer to new buffer
std::memcpy(newBufp, m_bufp, usedSize);
// Delete current buffer
delete[] m_bufp;
// Make new buffer the current buffer
m_bufp = newBufp;
// Adjust write pointer
m_writep = m_bufp + usedSize;
// Adjust resize limit
adjustGrowp();
}
#else
VL_FATAL_MT(__FILE__, __LINE__, "", "Unreachable");
#endif
} else {
// Flush the write buffer if there's not enough space left for new information
// We only call this once per vector, so we need enough slop for a very wide "b###" line
if (VL_UNLIKELY(m_writep > m_wrFlushp)) {
m_owner.m_writep = m_writep;
m_owner.bufferFlush();
m_writep = m_owner.m_writep;
}
}
}
//=============================================================================
// emit* trace routines
// Note: emit* are only ever called from one place (full* in
// verilated_trace_imp.cpp, which is included in this file at the top),
// verilated_trace_imp.h, which is included in this file at the top),
// so always inline them.
VL_ATTR_ALWINLINE
void VerilatedVcd::emitBit(uint32_t code, CData newval) {
void VerilatedVcdBuffer::emitBit(uint32_t code, CData newval) {
// Don't prefetch suffix as it's a bit too late;
char* wp = m_writep;
*wp++ = '0' | static_cast<char>(newval);
@@ -633,7 +715,7 @@ void VerilatedVcd::emitBit(uint32_t code, CData newval) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitCData(uint32_t code, CData newval, int bits) {
void VerilatedVcdBuffer::emitCData(uint32_t code, CData newval, int bits) {
char* wp = m_writep;
*wp++ = 'b';
cvtCDataToStr(wp, newval << (VL_BYTESIZE - bits));
@@ -641,7 +723,7 @@ void VerilatedVcd::emitCData(uint32_t code, CData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitSData(uint32_t code, SData newval, int bits) {
void VerilatedVcdBuffer::emitSData(uint32_t code, SData newval, int bits) {
char* wp = m_writep;
*wp++ = 'b';
cvtSDataToStr(wp, newval << (VL_SHORTSIZE - bits));
@@ -649,7 +731,7 @@ void VerilatedVcd::emitSData(uint32_t code, SData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitIData(uint32_t code, IData newval, int bits) {
void VerilatedVcdBuffer::emitIData(uint32_t code, IData newval, int bits) {
char* wp = m_writep;
*wp++ = 'b';
cvtIDataToStr(wp, newval << (VL_IDATASIZE - bits));
@@ -657,7 +739,7 @@ void VerilatedVcd::emitIData(uint32_t code, IData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitQData(uint32_t code, QData newval, int bits) {
void VerilatedVcdBuffer::emitQData(uint32_t code, QData newval, int bits) {
char* wp = m_writep;
*wp++ = 'b';
cvtQDataToStr(wp, newval << (VL_QUADSIZE - bits));
@@ -665,7 +747,7 @@ void VerilatedVcd::emitQData(uint32_t code, QData newval, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitWData(uint32_t code, const WData* newvalp, int bits) {
void VerilatedVcdBuffer::emitWData(uint32_t code, const WData* newvalp, int bits) {
int words = VL_WORDS_I(bits);
char* wp = m_writep;
*wp++ = 'b';
@@ -682,272 +764,10 @@ void VerilatedVcd::emitWData(uint32_t code, const WData* newvalp, int bits) {
}
VL_ATTR_ALWINLINE
void VerilatedVcd::emitDouble(uint32_t code, double newval) {
void VerilatedVcdBuffer::emitDouble(uint32_t code, double newval) {
char* wp = m_writep;
// Buffer can't overflow before VL_SNPRINTF; we sized during declaration
VL_SNPRINTF(wp, m_wrChunkSize, "r%.16g", newval);
VL_SNPRINTF(wp, m_maxSignalBytes, "r%.16g", newval);
wp += std::strlen(wp);
finishLine(code, wp);
}
#ifdef VL_TRACE_VCD_OLD_API
void VerilatedVcd::fullBit(uint32_t code, const uint32_t newval) {
// Note the &1, so we don't require clean input -- makes more common no change case faster
*oldp(code) = newval;
*m_writep++ = ('0' + static_cast<char>(newval & 1));
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullBus(uint32_t code, const uint32_t newval, int bits) {
*oldp(code) = newval;
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = ((newval & (1L << bit)) ? '1' : '0');
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullQuad(uint32_t code, const uint64_t newval, int bits) {
(*(reinterpret_cast<uint64_t*>(oldp(code)))) = newval;
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = ((newval & (1ULL << bit)) ? '1' : '0');
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullArray(uint32_t code, const uint32_t* newval, int bits) {
for (int word = 0; word < (((bits - 1) / 32) + 1); ++word) { oldp(code)[word] = newval[word]; }
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = ((newval[(bit / 32)] & (1L << (bit & 0x1f))) ? '1' : '0');
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullArray(uint32_t code, const uint64_t* newval, int bits) {
for (int word = 0; word < (((bits - 1) / 64) + 1); ++word) { oldp(code)[word] = newval[word]; }
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = ((newval[(bit / 64)] & (1ULL << (bit & 0x3f))) ? '1' : '0');
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullTriBit(uint32_t code, const uint32_t newval, const uint32_t newtri) {
oldp(code)[0] = newval;
oldp(code)[1] = newtri;
*m_writep++ = "01zz"[newval | (newtri << 1)];
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullTriBus(uint32_t code, const uint32_t newval, const uint32_t newtri,
int bits) {
oldp(code)[0] = newval;
oldp(code)[1] = newtri;
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = "01zz"[((newval >> bit) & 1) | (((newtri >> bit) & 1) << 1)];
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullTriQuad(uint32_t code, const uint64_t newval, const uint64_t newtri,
int bits) {
(*(reinterpret_cast<uint64_t*>(oldp(code)))) = newval;
(*(reinterpret_cast<uint64_t*>(oldp(code + 1)))) = newtri;
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
*m_writep++ = "01zz"[((newval >> bit) & 1ULL) | (((newtri >> bit) & 1ULL) << 1ULL)];
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullTriArray(uint32_t code, const uint32_t* newvalp, const uint32_t* newtrip,
int bits) {
for (int word = 0; word < (((bits - 1) / 32) + 1); ++word) {
oldp(code)[word * 2] = newvalp[word];
oldp(code)[word * 2 + 1] = newtrip[word];
}
*m_writep++ = 'b';
for (int bit = bits - 1; bit >= 0; --bit) {
uint32_t valbit = (newvalp[(bit / 32)] >> (bit & 0x1f)) & 1;
uint32_t tribit = (newtrip[(bit / 32)] >> (bit & 0x1f)) & 1;
*m_writep++ = "01zz"[valbit | (tribit << 1)];
}
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
void VerilatedVcd::fullDouble(uint32_t code, const double newval) {
// cppcheck-suppress invalidPointerCast
(*(reinterpret_cast<double*>(oldp(code)))) = newval;
// Buffer can't overflow before VL_SNPRINTF; we sized during declaration
VL_SNPRINTF(m_writep, m_wrChunkSize, "r%.16g", newval);
m_writep += std::strlen(m_writep);
*m_writep++ = ' ';
m_writep = writeCode(m_writep, code);
*m_writep++ = '\n';
bufferCheck();
}
#endif // VL_TRACE_VCD_OLD_API
//======================================================================
//======================================================================
//======================================================================
#ifdef VERILATED_VCD_TEST
#include <iostream>
extern void verilated_trace_imp_selftest();
uint32_t v1, v2, s1, s2[3];
uint32_t tri96[3];
uint32_t tri96__tri[3];
uint64_t quad96[2];
uint64_t tquad;
uint64_t tquad__tri;
uint8_t ch;
uint64_t timestamp = 1;
double doub = 0.0;
float flo = 0.0f;
void vcdInit(void*, VerilatedVcd* vcdp, uint32_t) {
vcdp->scopeEscape('.');
vcdp->pushNamePrefix("top.");
/**/ vcdp->declBus(0x2, "v1", -1, 0, 5, 1);
/**/ vcdp->declBus(0x3, "v2", -1, 0, 6, 1);
/**/ vcdp->pushNamePrefix("sub1.");
/***/ vcdp->declBit(0x4, "s1", -1, 0);
/***/ vcdp->declBit(0x5, "ch", -1, 0);
/**/ vcdp->popNamePrefix();
/**/ vcdp->pushNamePrefix("sub2.");
/***/ vcdp->declArray(0x6, "s2", -1, 0, 40, 3);
/**/ vcdp->popNamePrefix();
vcdp->popNamePrefix();
// Note need to add 3 for next code.
vcdp->pushNamePrefix("top2.");
/**/ vcdp->declBus(0x2, "t2v1", -1, 0, 4, 1);
/**/ vcdp->declTriBit(0x10, "io1", -1, 0);
/**/ vcdp->declTriBus(0x12, "io5", -1, 0, 4, 0);
/**/ vcdp->declTriArray(0x16, "io96", -1, 0, 95, 0);
/**/ // Note need to add 6 for next code.
/**/ vcdp->declDouble(0x1c, "doub", -1, 0);
/**/ // Note need to add 2 for next code.
/**/ vcdp->declArray(0x20, "q2", -1, 0, 95, 0);
/**/ // Note need to add 4 for next code.
/**/ vcdp->declTriQuad(0x24, "tq", -1, 0, 63, 0);
/**/ // Note need to add 4 for next code.
vcdp->popNamePrefix();
}
void vcdFull(void*, VerilatedVcd* vcdp) {
vcdp->fullBus(0x2, v1, 5);
vcdp->fullBus(0x3, v2, 7);
vcdp->fullBit(0x4, s1);
vcdp->fullBus(0x5, ch, 2);
vcdp->fullArray(0x6, &s2[0], 38);
vcdp->fullTriBit(0x10, tri96[0] & 1, tri96__tri[0] & 1);
vcdp->fullTriBus(0x12, tri96[0] & 0x1f, tri96__tri[0] & 0x1f, 5);
vcdp->fullTriArray(0x16, tri96, tri96__tri, 96);
vcdp->fullDouble(0x1c, doub);
vcdp->fullArray(0x20, &quad96[0], 96);
vcdp->fullTriQuad(0x24, tquad, tquad__tri, 64);
}
void vcdChange(void*, VerilatedVcd* vcdp) {
vcdp->chgBus(0x2, v1, 5);
vcdp->chgBus(0x3, v2, 7);
vcdp->chgBit(0x4, s1);
vcdp->chgBus(0x5, ch, 2);
vcdp->chgArray(0x6, &s2[0], 38);
vcdp->chgTriBit(0x10, tri96[0] & 1, tri96__tri[0] & 1);
vcdp->chgTriBus(0x12, tri96[0] & 0x1f, tri96__tri[0] & 0x1f, 5);
vcdp->chgTriArray(0x16, tri96, tri96__tri, 96);
vcdp->chgDouble(0x1c, doub);
vcdp->chgArray(0x20, &quad96[0], 96);
vcdp->chgTriQuad(0x24, tquad, tquad__tri, 64);
}
// clang-format off
void vcdTestMain(const char* filenamep) {
verilated_trace_imp_selftest();
v1 = v2 = s1 = 0;
s2[0] = s2[1] = s2[2] = 0;
tri96[2] = tri96[1] = tri96[0] = 0;
tri96__tri[2] = tri96__tri[1] = tri96__tri[0] = ~0;
quad96[1] = quad96[0] = 0;
ch = 0;
doub = 0;
tquad = tquad__tri = 0;
{
VerilatedVcdC* vcdp = new VerilatedVcdC;
vcdp->evcd(true);
vcdp->set_time_unit("1ms");
vcdp->set_time_unit(std::string{"1ms"});
vcdp->set_time_resolution("1ns");
vcdp->set_time_resolution(std::string{"1ns"});
vcdp->spTrace()->addInitCb(&vcdInit, 0);
vcdp->spTrace()->addFullCb(&vcdFull, 0);
vcdp->spTrace()->addChgCb(&vcdChange, 0);
vcdp->open(filenamep);
// Dumping
vcdp->dump(++timestamp);
v1 = 0xfff;
tri96[2] = 4; tri96[1] = 2; tri96[0] = 1;
tri96__tri[2] = tri96__tri[1] = tri96__tri[0] = ~0; // Still tri
quad96[1] = 0xffffffff; quad96[0] = 0;
doub = 1.5;
flo = 1.4f;
vcdp->dump(++timestamp);
v2 = 0x1;
s2[1] = 2;
tri96__tri[2] = tri96__tri[1] = tri96__tri[0] = 0; // enable w/o data change
quad96[1] = 0; quad96[0] = ~0;
doub = -1.66e13;
flo = 0.123f;
tquad = 0x00ff00ff00ff00ffULL;
tquad__tri = 0x0000fffff0000ffffULL;
vcdp->dump(++timestamp);
ch = 2;
tri96[2] = ~4; tri96[1] = ~2; tri96[0] = ~1;
doub = -3.33e-13;
vcdp->dump(++timestamp);
vcdp->dump(++timestamp);
# ifdef VERILATED_VCD_TEST_64BIT
const uint64_t bytesPerDump = 15ULL;
for (uint64_t i = 0; i < ((1ULL << 32) / bytesPerDump); i++) {
v1 = i;
vcdp->dump(++timestamp);
}
# endif
vcdp->close();
VL_DO_CLEAR(delete vcdp, vcdp = nullptr);
}
}
#endif
// clang-format on
//********************************************************************
// ;compile-command: "v4make test_regress/t/t_trace_c_api.pl"
//
// Local Variables:
// End:
+196 -288
View File
@@ -28,7 +28,196 @@
#include <string>
#include <vector>
class VerilatedVcd;
class VerilatedVcdBuffer;
class VerilatedVcdFile;
//=============================================================================
// VerilatedVcd
// Base class to create a Verilator VCD dump
// This is an internally used class - see VerilatedVcdC for what to call from applications
class VerilatedVcd VL_NOT_FINAL : public VerilatedTrace<VerilatedVcd, VerilatedVcdBuffer> {
public:
using Super = VerilatedTrace<VerilatedVcd, VerilatedVcdBuffer>;
private:
friend VerilatedVcdBuffer; // Give the buffer access to the private bits
//=========================================================================
// VCD specific internals
VerilatedVcdFile* m_filep; // File we're writing to
bool m_fileNewed; // m_filep needs destruction
bool m_isOpen = false; // True indicates open file
std::string m_filename; // Filename we're writing to (if open)
uint64_t m_rolloverSize = 0; // File size to rollover at
int m_modDepth = 0; // Depth of module hierarchy
char* m_wrBufp; // Output buffer
char* m_wrFlushp; // Output buffer flush trigger location
char* m_writep; // Write pointer into output buffer
size_t m_wrChunkSize; // Output buffer size
size_t m_maxSignalBytes = 0; // Upper bound on number of bytes a single signal can generate
uint64_t m_wroteBytes = 0; // Number of bytes written to this file
std::vector<char> m_suffixes; // VCD line end string codes + metadata
using NameMap = std::map<const std::string, const std::string>;
NameMap* m_namemapp = nullptr; // List of names for the header
#ifdef VL_THREADED
// Vector of free trace buffers as (pointer, size) pairs.
std::vector<std::pair<char*, size_t>> m_freeBuffers;
size_t m_numBuffers = 0; // Number of trace buffers allocated
#endif
void bufferResize(size_t minsize);
void bufferFlush() VL_MT_UNSAFE_ONE;
inline void bufferCheck() {
// Flush the write buffer if there's not enough space left for new information
// We only call this once per vector, so we need enough slop for a very wide "b###" line
if (VL_UNLIKELY(m_writep > m_wrFlushp)) bufferFlush();
}
void openNextImp(bool incFilename);
void closePrev();
void closeErr();
void makeNameMap();
void deleteNameMap();
void printIndent(int level_change);
void printStr(const char* str);
void printQuad(uint64_t n);
void printTime(uint64_t timeui);
void declare(uint32_t code, const char* name, const char* wirep, bool array, int arraynum,
bool tri, bool bussed, int msb, int lsb);
void dumpHeader();
static char* writeCode(char* writep, uint32_t code);
// CONSTRUCTORS
VL_UNCOPYABLE(VerilatedVcd);
protected:
//=========================================================================
// Implementation of VerilatedTrace interface
// Called when the trace moves forward to a new time point
void emitTimeChange(uint64_t timeui) override;
// Hooks called from VerilatedTrace
bool preFullDump() override { return isOpen(); }
bool preChangeDump() override;
// Trace buffer management
Buffer* getTraceBuffer() override;
void commitTraceBuffer(Buffer*) override;
// Configure sub-class
void configure(const VerilatedTraceConfig&) override{};
public:
//=========================================================================
// External interface to client code
// CONSTRUCTOR
explicit VerilatedVcd(VerilatedVcdFile* filep = nullptr);
~VerilatedVcd();
// ACCESSORS
// Set size in bytes after which new file should be created.
void rolloverSize(uint64_t size) { m_rolloverSize = size; }
// METHODS - All must be thread safe
// Open the file; call isOpen() to see if errors
void open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex);
// Open next data-only file
void openNext(bool incFilename) VL_MT_SAFE_EXCLUDES(m_mutex);
// Close the file
void close() VL_MT_SAFE_EXCLUDES(m_mutex);
// Flush any remaining data to this file
void flush() VL_MT_SAFE_EXCLUDES(m_mutex);
// Return if file is open
bool isOpen() const VL_MT_SAFE { return m_isOpen; }
//=========================================================================
// Internal interface to Verilator generated code
void declBit(uint32_t code, const char* name, bool array, int arraynum);
void declBus(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declQuad(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declArray(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declDouble(uint32_t code, const char* name, bool array, int arraynum);
};
#ifndef DOXYGEN
// Declare specialization here as it's used in VerilatedFstC just below
template <>
void VerilatedVcd::Super::dump(uint64_t time);
template <>
void VerilatedVcd::Super::set_time_unit(const char* unitp);
template <>
void VerilatedVcd::Super::set_time_unit(const std::string& unit);
template <>
void VerilatedVcd::Super::set_time_resolution(const char* unitp);
template <>
void VerilatedVcd::Super::set_time_resolution(const std::string& unit);
template <>
void VerilatedVcd::Super::dumpvars(int level, const std::string& hier);
#endif // DOXYGEN
//=============================================================================
// VerilatedVcdBuffer
class VerilatedVcdBuffer VL_NOT_FINAL {
// Give the trace file ans sub-classes access to the private bits
friend VerilatedVcd;
friend VerilatedVcd::Super;
friend VerilatedVcd::Buffer;
friend VerilatedVcd::OffloadBuffer;
VerilatedVcd& m_owner; // Trace file owning this buffer. Required by subclasses.
// Write pointer into output buffer (in parallel mode, this is set up in 'getTraceBuffer')
char* m_writep = m_owner.parallel() ? nullptr : m_owner.m_writep;
// Output buffer flush trigger location (only used when not parallel)
char* const m_wrFlushp = m_owner.parallel() ? nullptr : m_owner.m_wrFlushp;
// VCD line end string codes + metadata
const char* const m_suffixes = m_owner.m_suffixes.data();
// The maximum number of bytes a single signal can emit
const size_t m_maxSignalBytes = m_owner.m_maxSignalBytes;
#ifdef VL_THREADED
// Additional data for parallel tracing only
char* m_bufp = nullptr; // The beginning of the trace buffer
size_t m_size = 0; // The size of the buffer at m_bufp
char* m_growp = nullptr; // Resize limit pointer
void adjustGrowp() {
m_growp = (m_bufp + m_size) - (2 * m_maxSignalBytes);
assert(m_growp >= m_bufp + m_maxSignalBytes);
}
#endif
void finishLine(uint32_t code, char* writep);
// CONSTRUCTOR
explicit VerilatedVcdBuffer(VerilatedVcd& owner)
: m_owner{owner} {}
virtual ~VerilatedVcdBuffer() = default;
//=========================================================================
// Implementation of VerilatedTraceBuffer interface
// Implementations of duck-typed methods for VerilatedTraceBuffer. These are
// called from only one place (the full* methods), so always inline them.
VL_ATTR_ALWINLINE inline void emitBit(uint32_t code, CData newval);
VL_ATTR_ALWINLINE inline void emitCData(uint32_t code, CData newval, int bits);
VL_ATTR_ALWINLINE inline void emitSData(uint32_t code, SData newval, int bits);
VL_ATTR_ALWINLINE inline void emitIData(uint32_t code, IData newval, int bits);
VL_ATTR_ALWINLINE inline void emitQData(uint32_t code, QData newval, int bits);
VL_ATTR_ALWINLINE inline void emitWData(uint32_t code, const WData* newvalp, int bits);
VL_ATTR_ALWINLINE inline void emitDouble(uint32_t code, double newval);
};
//=============================================================================
// VerilatedFile
@@ -52,280 +241,6 @@ public:
virtual ssize_t write(const char* bufp, ssize_t len) VL_MT_UNSAFE;
};
//=============================================================================
// VerilatedVcd
// Base class to create a Verilator VCD dump
// This is an internally used class - see VerilatedVcdC for what to call from applications
class VerilatedVcd VL_NOT_FINAL : public VerilatedTrace<VerilatedVcd> {
private:
// Give the superclass access to private bits (to avoid virtual functions)
friend class VerilatedTrace<VerilatedVcd>;
//=========================================================================
// VCD specific internals
VerilatedVcdFile* m_filep; // File we're writing to
bool m_fileNewed; // m_filep needs destruction
bool m_isOpen = false; // True indicates open file
bool m_evcd = false; // True for evcd format
std::string m_filename; // Filename we're writing to (if open)
uint64_t m_rolloverMB = 0; // MB of file size to rollover at
int m_modDepth = 0; // Depth of module hierarchy
char* m_wrBufp; // Output buffer
const char* m_wrFlushp; // Output buffer flush trigger location
char* m_writep; // Write pointer into output buffer
uint64_t m_wrChunkSize; // Output buffer size
uint64_t m_wroteBytes = 0; // Number of bytes written to this file
std::vector<char> m_suffixes; // VCD line end string codes + metadata
using NameMap = std::map<const std::string, const std::string>;
NameMap* m_namemapp = nullptr; // List of names for the header
void bufferResize(uint64_t minsize);
void bufferFlush() VL_MT_UNSAFE_ONE;
inline void bufferCheck() {
// Flush the write buffer if there's not enough space left for new information
// We only call this once per vector, so we need enough slop for a very wide "b###" line
if (VL_UNLIKELY(m_writep > m_wrFlushp)) bufferFlush();
}
void openNextImp(bool incFilename);
void closePrev();
void closeErr();
void makeNameMap();
void deleteNameMap();
void printIndent(int level_change);
void printStr(const char* str);
void printQuad(uint64_t n);
void printTime(uint64_t timeui);
void declare(uint32_t code, const char* name, const char* wirep, bool array, int arraynum,
bool tri, bool bussed, int msb, int lsb);
void dumpHeader();
static char* writeCode(char* writep, uint32_t code);
void finishLine(uint32_t code, char* writep);
// CONSTRUCTORS
VL_UNCOPYABLE(VerilatedVcd);
protected:
//=========================================================================
// Implementation of VerilatedTrace interface
// Implementations of protected virtual methods for VerilatedTrace
virtual void emitTimeChange(uint64_t timeui) override;
// Hooks called from VerilatedTrace
virtual bool preFullDump() override { return isOpen(); }
virtual bool preChangeDump() override;
// Implementations of duck-typed methods for VerilatedTrace. These are
// called from only one place (namely full*) so always inline them.
inline void emitBit(uint32_t code, CData newval);
inline void emitCData(uint32_t code, CData newval, int bits);
inline void emitSData(uint32_t code, SData newval, int bits);
inline void emitIData(uint32_t code, IData newval, int bits);
inline void emitQData(uint32_t code, QData newval, int bits);
inline void emitWData(uint32_t code, const WData* newvalp, int bits);
inline void emitDouble(uint32_t code, double newval);
public:
//=========================================================================
// External interface to client code
explicit VerilatedVcd(VerilatedVcdFile* filep = nullptr);
~VerilatedVcd();
// ACCESSORS
// Set size in megabytes after which new file should be created
void rolloverMB(uint64_t rolloverMB) { m_rolloverMB = rolloverMB; }
// METHODS
// Open the file; call isOpen() to see if errors
void open(const char* filename) VL_MT_SAFE_EXCLUDES(m_mutex);
// Open next data-only file
void openNext(bool incFilename) VL_MT_SAFE_EXCLUDES(m_mutex);
// Close the file
void close() VL_MT_SAFE_EXCLUDES(m_mutex);
// Flush any remaining data to this file
void flush() VL_MT_SAFE_EXCLUDES(m_mutex);
// Return if file is open
bool isOpen() const VL_MT_SAFE { return m_isOpen; }
//=========================================================================
// Internal interface to Verilator generated code
void declBit(uint32_t code, const char* name, bool array, int arraynum);
void declBus(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declQuad(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declArray(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declDouble(uint32_t code, const char* name, bool array, int arraynum);
#ifdef VL_TRACE_VCD_OLD_API
//=========================================================================
// Note: These are only for testing for backward compatibility with foreign
// code and is not used by Verilator. Do not use these as there is no
// guarantee of functionality.
void declTriBit(uint32_t code, const char* name, bool array, int arraynum);
void declTriBus(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declTriQuad(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void declTriArray(uint32_t code, const char* name, bool array, int arraynum, int msb, int lsb);
void fullBit(uint32_t* oldp, CData newval) { fullBit(oldp - this->oldp(0), newval); }
void fullCData(uint32_t* oldp, CData newval, int bits) {
fullBus(oldp - this->oldp(0), newval, bits);
}
void fullSData(uint32_t* oldp, SData newval, int bits) {
fullBus(oldp - this->oldp(0), newval, bits);
}
void fullIData(uint32_t* oldp, IData newval, int bits) {
fullBus(oldp - this->oldp(0), newval, bits);
}
void fullQData(uint32_t* oldp, QData newval, int bits) {
fullQuad(oldp - this->oldp(0), newval, bits);
}
void fullWData(uint32_t* oldp, const WData* newvalp, int bits) {
fullArray(oldp - this->oldp(0), newvalp, bits);
}
void fullDouble(uint32_t* oldp, double newval) { fullDouble(oldp - this->oldp(0), newval); }
inline void chgBit(uint32_t* oldp, CData newval) { chgBit(oldp - this->oldp(0), newval); }
inline void chgCData(uint32_t* oldp, CData newval, int bits) {
chgBus(oldp - this->oldp(0), newval, bits);
}
inline void chgSData(uint32_t* oldp, SData newval, int bits) {
chgBus(oldp - this->oldp(0), newval, bits);
}
inline void chgIData(uint32_t* oldp, IData newval, int bits) {
chgBus(oldp - this->oldp(0), newval, bits);
}
inline void chgQData(uint32_t* oldp, QData newval, int bits) {
chgQuad(oldp - this->oldp(0), newval, bits);
}
inline void chgWData(uint32_t* oldp, const WData* newvalp, int bits) {
chgArray(oldp - this->oldp(0), newvalp, bits);
}
inline void chgDouble(uint32_t* oldp, double newval) {
chgDouble(oldp - this->oldp(0), newval);
}
// Inside dumping routines, dump one signal, faster when not inlined
// due to code size reduction.
void fullBit(uint32_t code, const uint32_t newval);
void fullBus(uint32_t code, const uint32_t newval, int bits);
void fullQuad(uint32_t code, const uint64_t newval, int bits);
void fullArray(uint32_t code, const uint32_t* newvalp, int bits);
void fullArray(uint32_t code, const uint64_t* newvalp, int bits);
void fullTriBit(uint32_t code, const uint32_t newval, const uint32_t newtri);
void fullTriBus(uint32_t code, const uint32_t newval, const uint32_t newtri, int bits);
void fullTriQuad(uint32_t code, const uint64_t newval, const uint64_t newtri, int bits);
void fullTriArray(uint32_t code, const uint32_t* newvalp, const uint32_t* newtrip, int bits);
void fullDouble(uint32_t code, const double newval);
// Inside dumping routines, dump one signal if it has changed.
// We do want to inline these to avoid calls when the value did not change.
inline void chgBit(uint32_t code, const uint32_t newval) {
const uint32_t diff = oldp(code)[0] ^ newval;
if (VL_UNLIKELY(diff)) fullBit(code, newval);
}
inline void chgBus(uint32_t code, const uint32_t newval, int bits) {
const uint32_t diff = oldp(code)[0] ^ newval;
if (VL_UNLIKELY(diff)) {
if (VL_UNLIKELY(bits == 32 || (diff & ((1U << bits) - 1)))) {
fullBus(code, newval, bits);
}
}
}
inline void chgQuad(uint32_t code, const uint64_t newval, int bits) {
const uint64_t diff = (*(reinterpret_cast<uint64_t*>(oldp(code)))) ^ newval;
if (VL_UNLIKELY(diff)) {
if (VL_UNLIKELY(bits == 64 || (diff & ((1ULL << bits) - 1)))) {
fullQuad(code, newval, bits);
}
}
}
inline void chgArray(uint32_t code, const uint32_t* newvalp, int bits) {
for (int word = 0; word < (((bits - 1) / 32) + 1); ++word) {
if (VL_UNLIKELY(oldp(code)[word] ^ newvalp[word])) {
fullArray(code, newvalp, bits);
return;
}
}
}
inline void chgArray(uint32_t code, const uint64_t* newvalp, int bits) {
for (int word = 0; word < (((bits - 1) / 64) + 1); ++word) {
if (VL_UNLIKELY(*(reinterpret_cast<uint64_t*>(oldp(code + 2 * word)))
^ newvalp[word])) {
fullArray(code, newvalp, bits);
return;
}
}
}
inline void chgTriBit(uint32_t code, const uint32_t newval, const uint32_t newtri) {
const uint32_t diff = ((oldp(code)[0] ^ newval) | (oldp(code)[1] ^ newtri));
if (VL_UNLIKELY(diff)) {
// Verilator 3.510 and newer provide clean input, so the below
// is only for back compatibility
if (VL_UNLIKELY(diff & 1)) { // Change after clean?
fullTriBit(code, newval, newtri);
}
}
}
inline void chgTriBus(uint32_t code, const uint32_t newval, const uint32_t newtri, int bits) {
const uint32_t diff = ((oldp(code)[0] ^ newval) | (oldp(code)[1] ^ newtri));
if (VL_UNLIKELY(diff)) {
if (VL_UNLIKELY(bits == 32 || (diff & ((1U << bits) - 1)))) {
fullTriBus(code, newval, newtri, bits);
}
}
}
inline void chgTriQuad(uint32_t code, const uint64_t newval, const uint64_t newtri, int bits) {
const uint64_t diff = (((*(reinterpret_cast<uint64_t*>(oldp(code)))) ^ newval)
| ((*(reinterpret_cast<uint64_t*>(oldp(code + 1)))) ^ newtri));
if (VL_UNLIKELY(diff)) {
if (VL_UNLIKELY(bits == 64 || (diff & ((1ULL << bits) - 1)))) {
fullTriQuad(code, newval, newtri, bits);
}
}
}
inline void chgTriArray(uint32_t code, const uint32_t* newvalp, const uint32_t* newtrip,
int bits) {
for (int word = 0; word < (((bits - 1) / 32) + 1); ++word) {
if (VL_UNLIKELY((oldp(code)[word * 2] ^ newvalp[word])
| (oldp(code)[word * 2 + 1] ^ newtrip[word]))) {
fullTriArray(code, newvalp, newtrip, bits);
return;
}
}
}
inline void chgDouble(uint32_t code, const double newval) {
// cppcheck-suppress invalidPointerCast
if (VL_UNLIKELY((*(reinterpret_cast<double*>(oldp(code)))) != newval)) {
fullDouble(code, newval);
}
}
// METHODS
// Old/standalone API only
void evcd(bool flag) { m_evcd = flag; }
#endif // VL_TRACE_VCD_OLD_API
};
#ifndef DOXYGEN
// Declare specializations here they are used in VerilatedVcdC just below
template <> void VerilatedTrace<VerilatedVcd>::dump(uint64_t timeui);
template <> void VerilatedTrace<VerilatedVcd>::set_time_unit(const char* unitp);
template <> void VerilatedTrace<VerilatedVcd>::set_time_unit(const std::string& unit);
template <> void VerilatedTrace<VerilatedVcd>::set_time_resolution(const char* unitp);
template <> void VerilatedTrace<VerilatedVcd>::set_time_resolution(const std::string& unit);
template <> void VerilatedTrace<VerilatedVcd>::dumpvars(int level, const std::string& hier);
#endif // DOXYGEN
//=============================================================================
// VerilatedVcdC
/// Class representing a VCD dump file in C standalone (no SystemC)
@@ -344,7 +259,6 @@ public:
/// Destruct, flush, and close the dump
virtual ~VerilatedVcdC() { close(); }
public:
// METHODS - User called
/// Return if file is open
@@ -357,8 +271,12 @@ public:
/// The header is only in the first file created, this allows
/// "cat" to be used to combine the header plus any number of data files.
void openNext(bool incFilename = true) VL_MT_SAFE { m_sptrace.openNext(incFilename); }
/// Set size in megabytes after which new file should be created
void rolloverMB(size_t rolloverMB) VL_MT_SAFE { m_sptrace.rolloverMB(rolloverMB); }
/// Set size in bytes after which new file should be created
/// This will create a header file, followed by each separate file
/// which might be larger than the given size (due to chunking and
/// alignment to a start of a given time's dump). Any file but the
/// first may be removed. Cat files together to create viewable vcd.
void rolloverSize(size_t size) VL_MT_SAFE { m_sptrace.rolloverSize(size); }
/// Close dump
void close() VL_MT_SAFE { m_sptrace.close(); }
/// Flush dump
@@ -396,16 +314,6 @@ public:
// Internal class access
inline VerilatedVcd* spTrace() { return &m_sptrace; }
#ifdef VL_TRACE_VCD_OLD_API
//=========================================================================
// Note: These are only for testing for backward compatibility with foreign
// code and is not used by Verilator. Do not use these as there is no
// guarantee of functionality.
// Use evcd format
void evcd(bool flag) VL_MT_UNSAFE_ONE { m_sptrace.evcd(flag); }
#endif
};
#endif // guard
+2 -4
View File
@@ -22,6 +22,7 @@
//=============================================================================
#include "verilatedos.h"
#include "verilated_vcd_sc.h"
//======================================================================
@@ -30,7 +31,7 @@
void VerilatedVcdSc::open(const char* filename) {
if (!sc_core::sc_get_curr_simcontext()->elaboration_done()) {
vl_fatal(__FILE__, __LINE__, "VerilatedVcdSc",
("%Error: VerilatedVcdSc::open(\"" + std::string(filename)
("%Error: VerilatedVcdSc::open(\"" + std::string{filename}
+ "\") is called before sc_core::sc_start(). "
"Run sc_core::sc_start(sc_core::SC_ZERO_TIME) before opening a wave file.")
.c_str());
@@ -63,9 +64,6 @@ void VerilatedVcdSc::trace(const unsigned int&, const std::string&, const char**
DECL_TRACE_METHOD_B( unsigned short )
DECL_TRACE_METHOD_B( unsigned int )
DECL_TRACE_METHOD_B( unsigned long )
#ifdef SYSTEMC_64BIT_PATCHES
DECL_TRACE_METHOD_B( unsigned long long)
#endif
DECL_TRACE_METHOD_B( char )
DECL_TRACE_METHOD_B( short )
DECL_TRACE_METHOD_B( int )
+3 -4
View File
@@ -23,6 +23,7 @@
#define VERILATOR_VERILATED_VCD_SC_H_
#include "verilatedos.h"
#include "verilated_sc.h"
#include "verilated_vcd_c.h"
@@ -54,7 +55,7 @@ public:
spTrace()->set_time_resolution(sc_get_time_resolution().to_string());
}
/// Destruct, flush, and close the dump
virtual ~VerilatedVcdSc() { close(); }
virtual ~VerilatedVcdSc() /*override*/ { close(); }
// METHODS - for SC kernel
// Called by SystemC simulate()
@@ -63,6 +64,7 @@ public:
}
// Override VerilatedVcdC. Must be called after starting simulation.
// cppcheck-suppress missingOverride // GCC won't accept override
virtual void open(const char* filename) /*override*/ VL_MT_SAFE;
private:
@@ -98,9 +100,6 @@ private:
DECL_TRACE_METHOD_B( unsigned short )
DECL_TRACE_METHOD_B( unsigned int )
DECL_TRACE_METHOD_B( unsigned long )
#ifdef SYSTEMC_64BIT_PATCHES
DECL_TRACE_METHOD_B( unsigned long long)
#endif
DECL_TRACE_METHOD_B( char )
DECL_TRACE_METHOD_B( short )
DECL_TRACE_METHOD_B( int )
+50 -48
View File
@@ -26,8 +26,9 @@
#define VERILATOR_VERILATED_VPI_CPP_
#include "verilated.h"
#include "verilated_vpi.h"
#include "verilated.h"
#include "verilated_imp.h"
#include <list>
@@ -81,8 +82,9 @@ public:
// To simplify our free list, we use a size large enough for all derived types
// We reserve word zero for the next pointer, as that's safer in case a
// dangling reference to the original remains around.
static const size_t chunk = 96;
if (VL_UNCOVERABLE(size > chunk)) VL_FATAL_MT(__FILE__, __LINE__, "", "increase chunk");
static constexpr size_t CHUNK_SIZE = 96;
if (VL_UNCOVERABLE(size > CHUNK_SIZE))
VL_FATAL_MT(__FILE__, __LINE__, "", "increase CHUNK_SIZE");
if (VL_LIKELY(t_freeHead)) {
uint8_t* const newp = t_freeHead;
t_freeHead = *(reinterpret_cast<uint8_t**>(newp));
@@ -90,7 +92,7 @@ public:
return newp + 8;
}
// +8: 8 bytes for next
uint8_t* newp = reinterpret_cast<uint8_t*>(::operator new(chunk + 8));
uint8_t* newp = reinterpret_cast<uint8_t*>(::operator new(CHUNK_SIZE + 8));
*(reinterpret_cast<uint32_t*>(newp)) = activeMagic();
return newp + 8;
}
@@ -134,12 +136,12 @@ public:
VerilatedVpioTimedCb(uint64_t id, QData time)
: m_id{id}
, m_time{time} {}
virtual ~VerilatedVpioTimedCb() override = default;
~VerilatedVpioTimedCb() override = default;
static VerilatedVpioTimedCb* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioTimedCb*>(reinterpret_cast<VerilatedVpioTimedCb*>(h));
}
virtual uint32_t type() const override { return vpiCallback; }
virtual PLI_INT32 dovpi_remove_cb() override;
uint32_t type() const override { return vpiCallback; }
PLI_INT32 dovpi_remove_cb() override;
};
class VerilatedVpioReasonCb final : public VerilatedVpio {
@@ -153,12 +155,12 @@ public:
VerilatedVpioReasonCb(uint64_t id, PLI_INT32 reason)
: m_id{id}
, m_reason{reason} {}
virtual ~VerilatedVpioReasonCb() override = default;
~VerilatedVpioReasonCb() override = default;
static VerilatedVpioReasonCb* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioReasonCb*>(reinterpret_cast<VerilatedVpioReasonCb*>(h));
}
virtual uint32_t type() const override { return vpiCallback; }
virtual PLI_INT32 dovpi_remove_cb() override;
uint32_t type() const override { return vpiCallback; }
PLI_INT32 dovpi_remove_cb() override;
};
class VerilatedVpioConst final : public VerilatedVpio {
@@ -167,11 +169,11 @@ class VerilatedVpioConst final : public VerilatedVpio {
public:
explicit VerilatedVpioConst(int32_t num)
: m_num{num} {}
virtual ~VerilatedVpioConst() override = default;
~VerilatedVpioConst() override = default;
static VerilatedVpioConst* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioConst*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiConstant; }
uint32_t type() const override { return vpiConstant; }
int32_t num() const { return m_num; }
};
@@ -199,10 +201,10 @@ public:
}
const VerilatedVar* varp() const { return m_varp; }
const VerilatedScope* scopep() const { return m_scopep; }
virtual uint32_t size() const override { return get_range().elements(); }
virtual const VerilatedRange* rangep() const override { return &get_range(); }
virtual const char* name() const override { return m_varp->name(); }
virtual const char* fullname() const override {
uint32_t size() const override { return get_range().elements(); }
const VerilatedRange* rangep() const override { return &get_range(); }
const char* name() const override { return m_varp->name(); }
const char* fullname() const override {
static VL_THREAD_LOCAL std::string t_out;
t_out = std::string{m_scopep->name()} + "." + name();
return t_out.c_str();
@@ -213,12 +215,12 @@ class VerilatedVpioParam final : public VerilatedVpioVarBase {
public:
VerilatedVpioParam(const VerilatedVar* varp, const VerilatedScope* scopep)
: VerilatedVpioVarBase{varp, scopep} {}
virtual ~VerilatedVpioParam() override = default;
~VerilatedVpioParam() override = default;
static VerilatedVpioParam* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioParam*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiParameter; }
uint32_t type() const override { return vpiParameter; }
void* varDatap() const { return m_varp->datap(); }
};
@@ -228,13 +230,13 @@ class VerilatedVpioRange final : public VerilatedVpio {
public:
explicit VerilatedVpioRange(const VerilatedRange* range)
: m_range{range} {}
virtual ~VerilatedVpioRange() override = default;
~VerilatedVpioRange() override = default;
static VerilatedVpioRange* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioRange*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiRange; }
virtual uint32_t size() const override { return m_range->elements(); }
virtual const VerilatedRange* rangep() const override { return m_range; }
uint32_t type() const override { return vpiRange; }
uint32_t size() const override { return m_range->elements(); }
const VerilatedRange* rangep() const override { return m_range; }
};
class VerilatedVpioRangeIter final : public VerilatedVpio {
@@ -245,12 +247,12 @@ class VerilatedVpioRangeIter final : public VerilatedVpio {
public:
explicit VerilatedVpioRangeIter(const VerilatedRange* range)
: m_range{range} {}
virtual ~VerilatedVpioRangeIter() override = default;
~VerilatedVpioRangeIter() override = default;
static VerilatedVpioRangeIter* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioRangeIter*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiIterator; }
virtual vpiHandle dovpi_scan() override {
uint32_t type() const override { return vpiIterator; }
vpiHandle dovpi_scan() override {
if (VL_UNLIKELY(m_done)) {
delete this; // IEEE 37.2.2 vpi_scan at end does a vpi_release_handle
return nullptr;
@@ -267,14 +269,14 @@ protected:
public:
explicit VerilatedVpioScope(const VerilatedScope* scopep)
: m_scopep{scopep} {}
virtual ~VerilatedVpioScope() override = default;
~VerilatedVpioScope() override = default;
static VerilatedVpioScope* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioScope*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiScope; }
uint32_t type() const override { return vpiScope; }
const VerilatedScope* scopep() const { return m_scopep; }
virtual const char* name() const override { return m_scopep->name(); }
virtual const char* fullname() const override { return m_scopep->name(); }
const char* name() const override { return m_scopep->name(); }
const char* fullname() const override { return m_scopep->name(); }
};
class VerilatedVpioVar VL_NOT_FINAL : public VerilatedVpioVarBase {
@@ -307,7 +309,7 @@ public:
m_mask.u32 = 0;
}
}
virtual ~VerilatedVpioVar() override {
~VerilatedVpioVar() override {
if (m_prevDatap) VL_DO_CLEAR(delete[] m_prevDatap, m_prevDatap = nullptr);
}
static VerilatedVpioVar* castp(vpiHandle h) {
@@ -317,7 +319,7 @@ public:
uint8_t mask_byte(int idx) const { return m_mask.u8[idx & 3]; }
uint32_t entSize() const { return m_entSize; }
uint32_t index() const { return m_index; }
virtual uint32_t type() const override {
uint32_t type() const override {
return (varp()->dims() > 1) ? vpiMemory : vpiReg; // but might be wire, logic
}
void* prevDatap() const { return m_prevDatap; }
@@ -338,14 +340,14 @@ public:
m_index = index;
m_varDatap = (static_cast<uint8_t*>(varp->datap())) + entSize() * offset;
}
virtual ~VerilatedVpioMemoryWord() override = default;
~VerilatedVpioMemoryWord() override = default;
static VerilatedVpioMemoryWord* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioMemoryWord*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiMemoryWord; }
virtual uint32_t size() const override { return varp()->packed().elements(); }
virtual const VerilatedRange* rangep() const override { return &(varp()->packed()); }
virtual const char* fullname() const override {
uint32_t type() const override { return vpiMemoryWord; }
uint32_t size() const override { return varp()->packed().elements(); }
const VerilatedRange* rangep() const override { return &(varp()->packed()); }
const char* fullname() const override {
static VL_THREAD_LOCAL std::string t_out;
constexpr size_t LEN_MAX_INDEX = 25;
char num[LEN_MAX_INDEX];
@@ -363,12 +365,12 @@ class VerilatedVpioVarIter final : public VerilatedVpio {
public:
explicit VerilatedVpioVarIter(const VerilatedScope* scopep)
: m_scopep{scopep} {}
virtual ~VerilatedVpioVarIter() override = default;
~VerilatedVpioVarIter() override = default;
static VerilatedVpioVarIter* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioVarIter*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiIterator; }
virtual vpiHandle dovpi_scan() override {
uint32_t type() const override { return vpiIterator; }
vpiHandle dovpi_scan() override {
if (VL_LIKELY(m_scopep->varsp())) {
const VerilatedVarNameMap* const varsp = m_scopep->varsp();
if (VL_UNLIKELY(!m_started)) {
@@ -404,15 +406,15 @@ public:
, m_varp{varp}
, m_iteration{varp->unpacked().right()}
, m_direction{VL_LIKELY(varp->unpacked().left() > varp->unpacked().right()) ? 1 : -1} {}
virtual ~VerilatedVpioMemoryWordIter() override = default;
~VerilatedVpioMemoryWordIter() override = default;
static VerilatedVpioMemoryWordIter* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioMemoryWordIter*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiIterator; }
uint32_t type() const override { return vpiIterator; }
void iterationInc() {
if (!(m_done = (m_iteration == m_varp->unpacked().left()))) m_iteration += m_direction;
}
virtual vpiHandle dovpi_scan() override {
vpiHandle dovpi_scan() override {
if (VL_UNLIKELY(m_done)) {
delete this; // IEEE 37.2.2 vpi_scan at end does a vpi_release_handle
return nullptr;
@@ -437,9 +439,9 @@ public:
static VerilatedVpioModule* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioModule*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiModule; }
virtual const char* name() const override { return m_name; }
virtual const char* fullname() const override { return m_fullname; }
uint32_t type() const override { return vpiModule; }
const char* name() const override { return m_name; }
const char* fullname() const override { return m_fullname; }
};
class VerilatedVpioModuleIter final : public VerilatedVpio {
@@ -451,12 +453,12 @@ public:
: m_vec{&vec} {
m_it = m_vec->begin();
}
virtual ~VerilatedVpioModuleIter() override = default;
~VerilatedVpioModuleIter() override = default;
static VerilatedVpioModuleIter* castp(vpiHandle h) {
return dynamic_cast<VerilatedVpioModuleIter*>(reinterpret_cast<VerilatedVpio*>(h));
}
virtual uint32_t type() const override { return vpiIterator; }
virtual vpiHandle dovpi_scan() override {
uint32_t type() const override { return vpiIterator; }
vpiHandle dovpi_scan() override {
if (m_it == m_vec->end()) {
delete this; // IEEE 37.2.2 vpi_scan at end does a vpi_release_handle
return nullptr;
+2 -1
View File
@@ -25,6 +25,7 @@
#define VERILATOR_VERILATED_VPI_H_
#include "verilatedos.h"
#include "verilated.h"
#include "verilated_syms.h"
@@ -47,7 +48,7 @@ public:
static bool callValueCbs() VL_MT_UNSAFE_ONE;
/// Call callbacks of arbitrary types.
/// User wrapper code should call this from their main loops.
static bool callCbs(const uint32_t reason) VL_MT_UNSAFE_ONE;
static bool callCbs(uint32_t reason) VL_MT_UNSAFE_ONE;
/// Returns time of the next registered VPI callback, or
/// ~(0ULL) if none are registered
static QData cbNextDeadline() VL_MT_UNSAFE_ONE;
+50 -23
View File
@@ -40,6 +40,7 @@
#ifdef __GNUC__
# define VL_ATTR_ALIGNED(alignment) __attribute__((aligned(alignment)))
# define VL_ATTR_ALWINLINE __attribute__((always_inline))
# define VL_ATTR_NOINLINE __attribute__((noinline))
# define VL_ATTR_COLD __attribute__((cold))
# define VL_ATTR_HOT __attribute__((hot))
# define VL_ATTR_NORETURN __attribute__((noreturn))
@@ -82,6 +83,9 @@
#ifndef VL_ATTR_ALWINLINE
# define VL_ATTR_ALWINLINE ///< Attribute to inline, even when not optimizing
#endif
#ifndef VL_ATTR_NOINLINE
# define VL_ATTR_NOINLINE ///< Attribute to never inline, even when optimizing
#endif
#ifndef VL_ATTR_COLD
# define VL_ATTR_COLD ///< Attribute that function is rarely executed
#endif
@@ -214,22 +218,33 @@
// C++-2011
#if __cplusplus >= 201103L || defined(__GXX_EXPERIMENTAL_CXX0X__) || defined(VL_CPPCHECK)
# ifndef VL_NO_LEGACY
// These are deprecated historical defines. We leave them in case users referenced them.
# define VL_EQ_DELETE = delete
# define vl_unique_ptr std::unique_ptr
# define vl_unordered_map std::unordered_map
# define vl_unordered_set std::unordered_set
# define VL_INCLUDE_UNORDERED_MAP <unordered_map>
# define VL_INCLUDE_UNORDERED_SET <unordered_set>
# define VL_FINAL final
# define VL_MUTABLE mutable
# define VL_OVERRIDE override
# endif
#else
# error "Verilator requires a C++11 or newer compiler"
#endif
#ifndef VL_NO_LEGACY
// These are deprecated historical defines. We leave them in case users referenced them.
# define VL_EQ_DELETE = delete
# define vl_unique_ptr std::unique_ptr
# define vl_unordered_map std::unordered_map
# define vl_unordered_set std::unordered_set
# define VL_INCLUDE_UNORDERED_MAP <unordered_map>
# define VL_INCLUDE_UNORDERED_SET <unordered_set>
# define VL_FINAL final
# define VL_MUTABLE mutable
# define VL_OVERRIDE override
#endif
//=========================================================================
// C++-2017
#if __cplusplus >= 201703L
# define VL_CONSTEXPR_CXX17 constexpr
#else
# define VL_CONSTEXPR_CXX17
#endif
//=========================================================================
// Optimization
@@ -241,14 +256,11 @@
// Internal coverage
#ifdef VL_GCOV
extern "C" {
void __gcov_flush(); // gcc sources gcc/gcov-io.h has the prototype
}
// Flush internal code coverage data before e.g. std::abort()
# define VL_GCOV_FLUSH() \
__gcov_flush()
extern "C" void __gcov_dump();
// Dump internal code coverage data before e.g. std::abort()
# define VL_GCOV_DUMP() __gcov_dump()
#else
# define VL_GCOV_FLUSH()
# define VL_GCOV_DUMP()
#endif
//=========================================================================
@@ -442,7 +454,8 @@ using ssize_t = uint32_t; ///< signed size_t; returned from read()
// or 0x0 if not implemented on this platform
#define VL_GET_CPU_TICK(val) \
{ \
uint32_t hi, lo; \
uint32_t hi; \
uint32_t lo; \
asm volatile("rdtsc" : "=a"(lo), "=d"(hi)); \
(val) = ((uint64_t)lo) | (((uint64_t)hi) << 32); \
}
@@ -517,12 +530,20 @@ using ssize_t = uint32_t; ///< signed size_t; returned from read()
#define VL_STRINGIFY(x) VL_STRINGIFY2(x)
#define VL_STRINGIFY2(x) #x
//=========================================================================
// Offset of field in type
// Address zero can cause compiler problems
#define VL_OFFSETOF(type, field) \
(reinterpret_cast<size_t>(&(reinterpret_cast<type*>(0x10000000)->field)) - 0x10000000)
//=========================================================================
// Conversions
namespace vlstd {
template <typename T> struct reverse_wrapper {
template <typename T>
struct reverse_wrapper {
const T& m_v;
explicit reverse_wrapper(const T& a_v)
@@ -532,10 +553,16 @@ template <typename T> struct reverse_wrapper {
};
// C++20's std::ranges::reverse_view
template <typename T> reverse_wrapper<T> reverse_view(const T& v) { return reverse_wrapper<T>(v); }
template <typename T>
reverse_wrapper<T> reverse_view(const T& v) {
return reverse_wrapper<T>(v);
}
// C++17's std::as_const
template <class T> T const& as_const(T& v) { return v; }
template <class T>
T const& as_const(T& v) {
return v;
}
}; // namespace vlstd
//=========================================================================
+10 -9
View File
@@ -29,10 +29,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Active.h"
#include "V3Ast.h"
#include "V3Const.h"
#include "V3Global.h"
#include "V3Graph.h"
#include <unordered_map>
@@ -75,7 +76,7 @@ public:
class LatchDetectGraph final : public V3Graph {
protected:
LatchDetectGraphVertex* m_curVertexp; // Current latch detection graph vertex
LatchDetectGraphVertex* m_curVertexp = nullptr; // Current latch detection graph vertex
std::vector<AstVarRef*> m_outputs; // Vector of lvalues encountered on this pass
VL_DEBUG_FUNC; // Declare debug()
@@ -290,13 +291,13 @@ private:
// STATE
LatchDetectGraph m_graph; // Graph used to detect latches in combo always
// VISITORS
virtual void visit(AstVarRef* nodep) {
virtual void visit(AstVarRef* nodep) override {
const AstVar* const varp = nodep->varp();
if (nodep->access().isWriteOrRW() && varp->isSignal() && !varp->isUsedLoopIdx()) {
m_graph.addAssignment(nodep);
}
}
virtual void visit(AstNodeIf* nodep) {
virtual void visit(AstNodeIf* nodep) override {
if (!nodep->isBoundsCheck()) {
LatchDetectGraphVertex* const parentp = m_graph.currentp();
LatchDetectGraphVertex* const branchp = m_graph.addPathVertex(parentp, "BRANCH", true);
@@ -308,7 +309,7 @@ private:
}
}
//--------------------
virtual void visit(AstNode* nodep) { iterateChildren(nodep); }
virtual void visit(AstNode* nodep) override { iterateChildren(nodep); }
public:
// CONSTRUCTORS
@@ -317,7 +318,7 @@ public:
iterate(nodep);
m_graph.latchCheck(nodep, kwd == VAlwaysKwd::ALWAYS_LATCH);
}
virtual ~ActiveLatchCheckVisitor() {}
~ActiveLatchCheckVisitor() override = default;
};
//######################################################################
@@ -387,7 +388,7 @@ public:
: m_check{check} {
iterate(nodep);
}
virtual ~ActiveDlyVisitor() override = default;
~ActiveDlyVisitor() override = default;
};
//######################################################################
@@ -530,7 +531,7 @@ private:
// Warn and/or convert any delayed assignments
if (combo && !sequent) {
ActiveDlyVisitor{nodep, ActiveDlyVisitor::CT_COMB};
{ ActiveDlyVisitor{nodep, ActiveDlyVisitor::CT_COMB}; }
const ActiveLatchCheckVisitor latchvisitor{nodep, kwd};
} else if (!combo && sequent) {
ActiveDlyVisitor{nodep, ActiveDlyVisitor::CT_SEQ};
@@ -598,7 +599,7 @@ private:
public:
// CONSTRUCTORS
explicit ActiveVisitor(AstNetlist* nodep) { iterate(nodep); }
virtual ~ActiveVisitor() override = default;
~ActiveVisitor() override = default;
};
//######################################################################
+3 -2
View File
@@ -26,11 +26,12 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3ActiveTop.h"
#include "V3Ast.h"
#include "V3SenTree.h"
#include "V3Const.h"
#include "V3Global.h"
#include "V3SenTree.h"
//######################################################################
// Active class functions
+2 -1
View File
@@ -17,9 +17,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Assert.h"
#include "V3Ast.h"
#include "V3Global.h"
#include "V3Stats.h"
//######################################################################
+1 -1
View File
@@ -20,8 +20,8 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Error.h"
#include "V3Ast.h"
#include "V3Error.h"
//============================================================================
+2 -1
View File
@@ -20,9 +20,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3AssertPre.h"
#include "V3Ast.h"
#include "V3Global.h"
//######################################################################
// Assert class functions
+65 -14
View File
@@ -18,10 +18,11 @@
#include "verilatedos.h"
#include "V3Ast.h"
#include "V3File.h"
#include "V3Global.h"
#include "V3Broken.h"
#include "V3EmitV.h"
#include "V3File.h"
#include "V3Global.h"
#include "V3String.h"
#include <iomanip>
@@ -205,7 +206,7 @@ string AstNode::prettyName(const string& namein) {
string AstNode::prettyTypeName() const {
if (name() == "") return typeName();
return string(typeName()) + " '" + prettyName() + "'";
return std::string{typeName()} + " '" + prettyName() + "'";
}
//######################################################################
@@ -281,13 +282,13 @@ void AstNode::addNextHere(AstNode* newp) {
// This could be at head, tail, or both (single)
// New could be head of single node, or list
UASSERT(newp, "Null item passed to addNext");
UASSERT(!newp->backp(), "New node (back) already assigned?");
UASSERT_OBJ(!newp->backp(), newp, "New node (back) already assigned?");
debugTreeChange(this, "-addHereThs: ", __LINE__, false);
debugTreeChange(newp, "-addHereNew: ", __LINE__, true);
newp->editCountInc();
AstNode* const addlastp = newp->m_headtailp; // Last node in list to be added
UASSERT(!addlastp->m_nextp, "Headtailp tail isn't at the tail");
UASSERT_OBJ(!addlastp->m_nextp, addlastp, "Headtailp tail isn't at the tail");
// Forward links
AstNode* const oldnextp = this->m_nextp;
@@ -437,7 +438,7 @@ void VNRelinker::dump(std::ostream& str) const {
AstNode* AstNode::unlinkFrBackWithNext(VNRelinker* linkerp) {
debugTreeChange(this, "-unlinkWNextThs: ", __LINE__, true);
AstNode* const oldp = this;
UASSERT(oldp->m_backp, "Node has no back, already unlinked?");
UASSERT_OBJ(oldp->m_backp, oldp, "Node has no back, already unlinked?");
oldp->editCountInc();
AstNode* const backp = oldp->m_backp;
if (linkerp) {
@@ -497,7 +498,7 @@ AstNode* AstNode::unlinkFrBackWithNext(VNRelinker* linkerp) {
AstNode* AstNode::unlinkFrBack(VNRelinker* linkerp) {
debugTreeChange(this, "-unlinkFrBkThs: ", __LINE__, true);
AstNode* const oldp = this;
UASSERT(oldp->m_backp, "Node has no back, already unlinked?");
UASSERT_OBJ(oldp->m_backp, oldp, "Node has no back, already unlinked?");
oldp->editCountInc();
AstNode* const backp = oldp->m_backp;
if (linkerp) {
@@ -565,7 +566,7 @@ void AstNode::relink(VNRelinker* linkerp) {
}
AstNode* const newp = this;
UASSERT(linkerp && linkerp->m_backp, "Need non-empty linker");
UASSERT(!newp->backp(), "New node already linked?");
UASSERT_OBJ(!newp->m_backp, newp, "New node already linked?");
newp->editCountInc();
if (debug() > 8) {
@@ -631,11 +632,56 @@ void AstNode::relinkOneLink(AstNode*& pointpr, // Ref to pointer that gets set
}
void AstNode::addHereThisAsNext(AstNode* newp) {
// {old}->this->{next} becomes {old}->new->this->{next}
VNRelinker handle;
this->unlinkFrBackWithNext(&handle);
newp->addNext(this);
handle.relink(newp);
// {back}->this->{next} becomes {back}->new->this->{next}
UASSERT_OBJ(!newp->backp(), newp, "New node already linked?");
UASSERT_OBJ(this->m_backp, this, "'this' node has no back, already unlinked?");
UASSERT_OBJ(newp->m_headtailp, newp, "m_headtailp not set on new node");
//
AstNode* const backp = this->m_backp;
AstNode* const newLastp = newp->m_headtailp;
//
this->editCountInc();
// Common linkage
newLastp->m_nextp = this;
this->m_backp = newLastp;
newp->m_backp = backp;
// newLastp will not be the last node in the list as 'this' will follow it.
// If newLastp == newp, then below handles newp becoming head
newLastp->m_headtailp = nullptr;
// Linkage dependent on position
if (backp && backp->m_nextp == this) {
// If 'this' is not at the head of a list, then the new node will also not be at the head
// of a list, so we can just link in the new node in the middle.
backp->m_nextp = newp;
newp->m_headtailp = nullptr;
} else {
// If 'this' is at the head of a list, then the new node becomes the head of that list.
if (backp) {
if (backp->m_op1p == this) {
backp->m_op1p = newp;
} else if (backp->m_op2p == this) {
backp->m_op2p = newp;
} else if (backp->m_op3p == this) {
backp->m_op3p = newp;
} else {
UASSERT_OBJ(backp->m_op4p == this, this, "Don't know where newp should go");
backp->m_op4p = newp;
}
}
// We also need to update m_headtailp.
AstNode* const tailp = this->m_headtailp;
this->m_headtailp = nullptr;
newp->m_headtailp = tailp;
tailp->m_headtailp = newp;
}
// Iterator fixup
if (newLastp->m_iterpp) *(newLastp->m_iterpp) = this;
if (this->m_iterpp) {
*(this->m_iterpp) = newp;
this->m_iterpp = nullptr;
}
//
debugTreeChange(this, "-addHereThisAsNext: ", __LINE__, true);
}
void AstNode::swapWith(AstNode* bp) {
@@ -758,7 +804,7 @@ void AstNode::deleteTree() {
#ifdef VL_LEAK_CHECKS
void* AstNode::operator new(size_t size) {
// Optimization note: Aligning to cache line is a loss, due to lost packing
const AstNode* const objp = static_cast<AstNode*>(::operator new(size));
AstNode* const objp = static_cast<AstNode*>(::operator new(size));
V3Broken::addNewed(objp);
return objp;
}
@@ -1248,3 +1294,8 @@ void VNDeleter::doDeletes() {
for (AstNode* const nodep : m_deleteps) nodep->deleteTree();
m_deleteps.clear();
}
//######################################################################
// VNVisitor
#include "V3Ast__gen_visitor_defns.h" // From ./astgen
+337 -174
View File
@@ -20,11 +20,13 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Broken.h"
#include "V3Error.h"
#include "V3FileLine.h"
#include "V3Number.h"
#include "V3Global.h"
#include "V3Broken.h"
#include "V3Number.h"
#include "V3Ast__gen_classes.h" // From ./astgen
#include <cmath>
#include <functional>
@@ -34,8 +36,6 @@
#include <unordered_set>
#include <utility>
#include <vector>
#include "V3Ast__gen_classes.h" // From ./astgen
// Things like:
// class V3AstNode;
@@ -95,7 +95,7 @@ public:
// const char* ascii() const {...};
enum en m_e;
// cppcheck-suppress uninitVar // responsibility of each subclass
inline VNType() {}
inline VNType() = default;
// cppcheck-suppress noExplicitConstructor
inline VNType(en _e)
: m_e{_e} {}
@@ -997,7 +997,7 @@ public:
return false;
}
//
VNumRange() {}
VNumRange() = default;
VNumRange(int hi, int lo, bool littleEndian) { init(hi, lo, littleEndian); }
VNumRange(int left, int right)
: m_left{left}
@@ -1123,7 +1123,7 @@ class VNUser final {
} m_u;
public:
VNUser() {}
VNUser() = default;
// non-explicit:
// cppcheck-suppress noExplicitConstructor
VNUser(int i) {
@@ -1133,7 +1133,7 @@ public:
explicit VNUser(void* p) { m_u.up = p; }
~VNUser() = default;
// Casters
template <class T> //
template <class T>
typename std::enable_if<std::is_pointer<T>::value, T>::type to() const {
return reinterpret_cast<T>(m_u.up);
}
@@ -1297,10 +1297,8 @@ public:
/// Return edited nodep; see comments in V3Ast.cpp
AstNode* iterateSubtreeReturnEdits(AstNode* nodep);
#include "V3Ast__gen_visitor.h" // From ./astgen
// Things like:
// virtual void visit(AstBreak* nodep) { visit((AstNodeStmt*)(nodep)); }
// virtual void visit(AstNodeStmt* nodep) { visit((AstNode*)(nodep)); }
virtual void visit(AstNode* nodep) = 0;
#include "V3Ast__gen_visitor_decls.h" // From ./astgen
};
//######################################################################
@@ -1334,25 +1332,27 @@ inline std::ostream& operator<<(std::ostream& os, const VNRelinker& rhs) {
return os;
}
//######################################################################
// Callback base class to determine if node matches some formula
// ######################################################################
// Callback base class to determine if node matches some formula
class VNodeMatcher VL_NOT_FINAL {
public:
virtual bool nodeMatch(const AstNode* nodep) const { return true; }
};
//######################################################################
// AstNode -- Base type of all Ast types
// ######################################################################
// AstNode -- Base type of all Ast types
// Prefetch a node.
#define ASTNODE_PREFETCH_NON_NULL(nodep) \
do { \
VL_PREFETCH_RD(&((nodep)->m_nextp)); \
VL_PREFETCH_RD(&((nodep)->m_type)); \
} while (false)
// The if() makes it faster, even though prefetch won't fault on null pointers
#define ASTNODE_PREFETCH(nodep) \
do { \
if (nodep) { \
VL_PREFETCH_RD(&((nodep)->m_nextp)); \
VL_PREFETCH_RD(&((nodep)->m_type)); \
} \
if (nodep) ASTNODE_PREFETCH_NON_NULL(nodep); \
} while (false)
class AstNode VL_NOT_FINAL {
@@ -1509,6 +1509,16 @@ public:
AstNode* firstAbovep() const { // Returns nullptr when second or later in list
return ((backp() && backp()->nextp() != this) ? backp() : nullptr);
}
// isFirstInMyListOfStatements(n) -- implemented by child classes:
// AstNodeBlock, AstCaseItem, AstNodeIf, AstNodeFTask, and possibly others.
virtual bool isFirstInMyListOfStatements(AstNode* n) const { return false; }
// isStandaloneBodyStmt == Do we need a ; on generated cpp for this node?
bool isStandaloneBodyStmt() {
return (!firstAbovep() // we're 2nd or later in the list, so yes need ;
// If we're first in the list, check what backp() thinks of us:
|| (backp() && backp()->isFirstInMyListOfStatements(this)));
}
uint8_t brokenState() const { return m_brokenState; }
void brokenState(uint8_t value) { m_brokenState = value; }
@@ -1555,7 +1565,7 @@ public:
static string dedotName(const string& namein); // Name with dots removed
static string prettyName(const string& namein); // Name for printing out to the user
static string prettyNameQ(const string& namein) { // Quoted pretty name (for errors)
return string("'") + prettyName(namein) + "'";
return std::string{"'"} + prettyName(namein) + "'";
}
static string
encodeName(const string& namein); // Encode user name into internal C representation
@@ -1717,13 +1727,13 @@ public:
void dtypeSetVoid() { dtypep(findVoidDType()); }
// Data type locators
AstNodeDType* findBitDType() { return findBasicDType(VBasicDTypeKwd::LOGIC); }
AstNodeDType* findDoubleDType() { return findBasicDType(VBasicDTypeKwd::DOUBLE); }
AstNodeDType* findStringDType() { return findBasicDType(VBasicDTypeKwd::STRING); }
AstNodeDType* findSigned32DType() { return findBasicDType(VBasicDTypeKwd::INTEGER); }
AstNodeDType* findUInt32DType() { return findBasicDType(VBasicDTypeKwd::UINT32); }
AstNodeDType* findUInt64DType() { return findBasicDType(VBasicDTypeKwd::UINT64); }
AstNodeDType* findCHandleDType() { return findBasicDType(VBasicDTypeKwd::CHANDLE); }
AstNodeDType* findBitDType() const { return findBasicDType(VBasicDTypeKwd::LOGIC); }
AstNodeDType* findDoubleDType() const { return findBasicDType(VBasicDTypeKwd::DOUBLE); }
AstNodeDType* findStringDType() const { return findBasicDType(VBasicDTypeKwd::STRING); }
AstNodeDType* findSigned32DType() const { return findBasicDType(VBasicDTypeKwd::INTEGER); }
AstNodeDType* findUInt32DType() const { return findBasicDType(VBasicDTypeKwd::UINT32); }
AstNodeDType* findUInt64DType() const { return findBasicDType(VBasicDTypeKwd::UINT64); }
AstNodeDType* findCHandleDType() const { return findBasicDType(VBasicDTypeKwd::CHANDLE); }
AstNodeDType* findEmptyQueueDType() const;
AstNodeDType* findVoidDType() const;
AstNodeDType* findQueueIndexDType() const;
@@ -1787,7 +1797,6 @@ public:
void deleteTree(); // Always deletes the next link
void checkTree(); // User Interface version
void checkIter() const;
void clearIter() { m_iterpp = nullptr; }
void dumpPtrs(std::ostream& os = std::cout) const;
void dumpTree(std::ostream& os = std::cout, const string& indent = " ",
int maxDepth = 0) const;
@@ -1858,16 +1867,12 @@ private:
// For internal use only.
// Note: specializations for particular node types are provided by 'astgen'
template <typename T> inline static bool privateTypeTest(const AstNode* nodep);
template <typename T>
inline static bool privateTypeTest(const AstNode* nodep);
// For internal use only.
// Note: specializations for particular node types are provided below
template <typename T_Node> inline static bool privateMayBeUnder(const AstNode* nodep) {
return true;
}
// For internal use only.
template <typename TargetType, typename DeclType> constexpr static bool uselessCast() {
template <typename TargetType, typename DeclType>
constexpr static bool uselessCast() {
using NonRef = typename std::remove_reference<DeclType>::type;
using NonPtr = typename std::remove_pointer<NonRef>::type;
using NonCV = typename std::remove_cv<NonPtr>::type;
@@ -1875,7 +1880,8 @@ private:
}
// For internal use only.
template <typename TargetType, typename DeclType> constexpr static bool impossibleCast() {
template <typename TargetType, typename DeclType>
constexpr static bool impossibleCast() {
using NonRef = typename std::remove_reference<DeclType>::type;
using NonPtr = typename std::remove_pointer<NonRef>::type;
using NonCV = typename std::remove_cv<NonPtr>::type;
@@ -1884,20 +1890,23 @@ private:
public:
// For use via the VN_IS macro only
template <typename T, typename E> inline static bool privateIs(const AstNode* nodep) {
template <typename T, typename E>
inline static bool privateIs(const AstNode* nodep) {
static_assert(!uselessCast<T, E>(), "Unnecessary VN_IS, node known to have target type.");
static_assert(!impossibleCast<T, E>(), "Unnecessary VN_IS, node cannot be this type.");
return nodep && privateTypeTest<T>(nodep);
}
// For use via the VN_CAST macro only
template <typename T, typename E> inline static T* privateCast(AstNode* nodep) {
template <typename T, typename E>
inline static T* privateCast(AstNode* nodep) {
static_assert(!uselessCast<T, E>(),
"Unnecessary VN_CAST, node known to have target type.");
static_assert(!impossibleCast<T, E>(), "Unnecessary VN_CAST, node cannot be this type.");
return nodep && privateTypeTest<T>(nodep) ? reinterpret_cast<T*>(nodep) : nullptr;
}
template <typename T, typename E> inline static const T* privateCast(const AstNode* nodep) {
template <typename T, typename E>
inline static const T* privateCast(const AstNode* nodep) {
static_assert(!uselessCast<T, E>(),
"Unnecessary VN_CAST, node known to have target type.");
static_assert(!impossibleCast<T, E>(), "Unnecessary VN_CAST, node cannot be this type.");
@@ -1905,7 +1914,8 @@ public:
}
// For use via the VN_AS macro only
template <typename T, typename E> inline static T* privateAs(AstNode* nodep) {
template <typename T, typename E>
inline static T* privateAs(AstNode* nodep) {
static_assert(!uselessCast<T, E>(), "Unnecessary VN_AS, node known to have target type.");
static_assert(!impossibleCast<T, E>(), "Unnecessary VN_AS, node cannot be this type.");
UASSERT_OBJ(!nodep || privateTypeTest<T>(nodep), nodep,
@@ -1913,7 +1923,8 @@ public:
<< "'");
return reinterpret_cast<T*>(nodep);
}
template <typename T, typename E> inline static const T* privateAs(const AstNode* nodep) {
template <typename T, typename E>
inline static const T* privateAs(const AstNode* nodep) {
static_assert(!uselessCast<T, E>(), "Unnecessary VN_AS, node known to have target type.");
static_assert(!impossibleCast<T, E>(), "Unnecessary VN_AS, node cannot be this type.");
UASSERT_OBJ(!nodep || privateTypeTest<T>(nodep), nodep,
@@ -1924,167 +1935,117 @@ public:
// Predicate that returns true if the given 'nodep' might have a descendant of type 'T_Node'.
// This is conservative and is used to speed up traversals.
template <typename T_Node> inline static bool mayBeUnder(const AstNode* nodep) {
// Note: specializations for particular node types are provided below
template <typename T_Node>
static bool mayBeUnder(const AstNode* nodep) {
static_assert(!std::is_const<T_Node>::value,
"Type parameter 'T_Node' should not be const qualified");
static_assert(std::is_base_of<AstNode, T_Node>::value,
"Type parameter 'T_Node' must be a subtype of AstNode");
return privateMayBeUnder<T_Node>(nodep);
return true;
}
// Predicate that is true for node subtypes 'T_Node' that do not have any children
// This is conservative and is used to speed up traversals.
// Note: specializations for particular node types are provided below
template <typename T_Node>
static constexpr bool isLeaf() {
static_assert(!std::is_const<T_Node>::value,
"Type parameter 'T_Node' should not be const qualified");
static_assert(std::is_base_of<AstNode, T_Node>::value,
"Type parameter 'T_Node' must be a subtype of AstNode");
return false;
}
private:
template <typename T_Arg, bool VisitNext>
static void foreachImpl(
// Using std::conditional for const correctness in the public 'foreach' functions
typename std::conditional<std::is_const<T_Arg>::value, const AstNode*, AstNode*>::type
nodep,
std::function<void(T_Arg*)> f) {
// Using std::conditional for const correctness in the public 'foreach' functions
template <typename T_Arg>
using ConstCorrectAstNode =
typename std::conditional<std::is_const<T_Arg>::value, const AstNode, AstNode>::type;
// Note: Using a loop to iterate the nextp() chain, instead of tail recursion, because
// debug builds don't eliminate tail calls, causing stack overflow on long lists of nodes.
do {
// Prefetch children and next
ASTNODE_PREFETCH(nodep->op1p());
ASTNODE_PREFETCH(nodep->op2p());
ASTNODE_PREFETCH(nodep->op3p());
ASTNODE_PREFETCH(nodep->op4p());
if /* TODO: 'constexpr' in C++17 */ (VisitNext) ASTNODE_PREFETCH(nodep->nextp());
template <typename T_Arg>
inline static void foreachImpl(ConstCorrectAstNode<T_Arg>* nodep,
const std::function<void(T_Arg*)>& f, bool visitNext);
// Apply function in pre-order
if (privateTypeTest<typename std::remove_const<T_Arg>::type>(nodep)) {
f(static_cast<T_Arg*>(nodep));
}
template <typename T_Arg, bool Default>
inline static bool predicateImpl(ConstCorrectAstNode<T_Arg>* nodep,
const std::function<bool(T_Arg*)>& p);
// Traverse children (including their 'nextp()' chains), unless futile
if (mayBeUnder<typename std::remove_const<T_Arg>::type>(nodep)) {
if (AstNode* const op1p = nodep->op1p()) foreachImpl<T_Arg, true>(op1p, f);
if (AstNode* const op2p = nodep->op2p()) foreachImpl<T_Arg, true>(op2p, f);
if (AstNode* const op3p = nodep->op3p()) foreachImpl<T_Arg, true>(op3p, f);
if (AstNode* const op4p = nodep->op4p()) foreachImpl<T_Arg, true>(op4p, f);
}
// Traverse 'nextp()' chain if requested
if /* TODO: 'constexpr' in C++17 */ (VisitNext) {
nodep = nodep->nextp();
} else {
break;
}
} while (nodep);
}
template <typename T_Arg, bool Default, bool VisitNext>
static bool predicateImpl(
// Using std::conditional for const correctness in the public 'foreach' functions
typename std::conditional<std::is_const<T_Arg>::value, const AstNode*, AstNode*>::type
nodep,
std::function<bool(T_Arg*)> p) {
// Note: Using a loop to iterate the nextp() chain, instead of tail recursion, because
// debug builds don't eliminate tail calls, causing stack overflow on long lists of nodes.
do {
// Prefetch children and next
ASTNODE_PREFETCH(nodep->op1p());
ASTNODE_PREFETCH(nodep->op2p());
ASTNODE_PREFETCH(nodep->op3p());
ASTNODE_PREFETCH(nodep->op4p());
if /* TODO: 'constexpr' in C++17 */ (VisitNext) ASTNODE_PREFETCH(nodep->nextp());
// Apply function in pre-order
if (privateTypeTest<typename std::remove_const<T_Arg>::type>(nodep)) {
if (p(static_cast<T_Arg*>(nodep)) != Default) return !Default;
}
// Traverse children (including their 'nextp()' chains), unless futile
if (mayBeUnder<typename std::remove_const<T_Arg>::type>(nodep)) {
if (AstNode* const op1p = nodep->op1p()) {
if (predicateImpl<T_Arg, Default, true>(op1p, p) != Default) return !Default;
}
if (AstNode* const op2p = nodep->op2p()) {
if (predicateImpl<T_Arg, Default, true>(op2p, p) != Default) return !Default;
}
if (AstNode* const op3p = nodep->op3p()) {
if (predicateImpl<T_Arg, Default, true>(op3p, p) != Default) return !Default;
}
if (AstNode* const op4p = nodep->op4p()) {
if (predicateImpl<T_Arg, Default, true>(op4p, p) != Default) return !Default;
}
}
// Traverse 'nextp()' chain if requested
if /* TODO: 'constexpr' in C++17 */ (VisitNext) {
nodep = nodep->nextp();
} else {
break;
}
} while (nodep);
return Default;
}
template <typename T_Node> constexpr static void checkTypeParameter() {
template <typename T_Node>
constexpr static bool checkTypeParameter() {
static_assert(!std::is_const<T_Node>::value,
"Type parameter 'T_Node' should not be const qualified");
static_assert(std::is_base_of<AstNode, T_Node>::value,
"Type parameter 'T_Node' must be a subtype of AstNode");
return true;
}
public:
// Traverse subtree and call given function 'f' in pre-order on each node that has type
// 'T_Node'. Prefer 'foreach' over simple VNVisitor that only needs to handle a single (or a
// few) node types, as it's easier to write, but more importantly, the dispatch to the
// operation function in 'foreach' should be completely predictable by branch target caches in
// modern CPUs, while it is basically unpredictable for VNVisitor.
template <typename T_Node> void foreach (std::function<void(T_Node*)> f) {
checkTypeParameter<T_Node>();
foreachImpl<T_Node, /* VisitNext: */ false>(this, f);
// 'T_Node'. The node passd to the function 'f' can be removed or replaced, but other editing
// of the iterated tree is not safe. Prefer 'foreach' over simple VNVisitor that only needs to
// handle a single (or a few) node types, as it's easier to write, but more importantly, the
// dispatch to the operation function in 'foreach' should be completely predictable by branch
// target caches in modern CPUs, while it is basically unpredictable for VNVisitor.
template <typename T_Node>
void foreach (std::function<void(T_Node*)> f) {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
foreachImpl<T_Node>(this, f, /* visitNext: */ false);
}
// Same as above, but for 'const' nodes
template <typename T_Node> void foreach (std::function<void(const T_Node*)> f) const {
checkTypeParameter<T_Node>();
foreachImpl<const T_Node, /* VisitNext: */ false>(this, f);
template <typename T_Node>
void foreach (std::function<void(const T_Node*)> f) const {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
foreachImpl<const T_Node>(this, f, /* visitNext: */ false);
}
// Same as 'foreach' but also follows 'this->nextp()'
template <typename T_Node> void foreachAndNext(std::function<void(T_Node*)> f) {
checkTypeParameter<T_Node>();
foreachImpl<T_Node, /* VisitNext: */ true>(this, f);
template <typename T_Node>
void foreachAndNext(std::function<void(T_Node*)> f) {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
foreachImpl<T_Node>(this, f, /* visitNext: */ true);
}
// Same as 'foreach' but also follows 'this->nextp()'
template <typename T_Node> void foreachAndNext(std::function<void(const T_Node*)> f) const {
checkTypeParameter<T_Node>();
foreachImpl<const T_Node, /* VisitNext: */ true>(this, f);
template <typename T_Node>
void foreachAndNext(std::function<void(const T_Node*)> f) const {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
foreachImpl<const T_Node>(this, f, /* visitNext: */ true);
}
// Given a predicate function 'p' return true if and only if there exists a node of type
// 'T_Node' that satisfies the predicate 'p'. Returns false if no node of type 'T_Node' is
// present. Traversal is performed in some arbitrary order and is terminated as soon as the
// result can be determined.
template <typename T_Node> bool exists(std::function<bool(T_Node*)> p) {
checkTypeParameter<T_Node>();
return predicateImpl<T_Node, /* Default: */ false, /* VisitNext: */ false>(this, p);
template <typename T_Node>
bool exists(std::function<bool(T_Node*)> p) {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
return predicateImpl<T_Node, /* Default: */ false>(this, p);
}
// Same as above, but for 'const' nodes
template <typename T_Node> void exists(std::function<bool(const T_Node*)> p) const {
checkTypeParameter<T_Node>();
return predicateImpl<const T_Node, /* Default: */ false, /* VisitNext: */ false>(this, p);
template <typename T_Node>
bool exists(std::function<bool(const T_Node*)> p) const {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
return predicateImpl<const T_Node, /* Default: */ false>(this, p);
}
// Given a predicate function 'p' return true if and only if all nodes of type
// 'T_Node' satisfy the predicate 'p'. Returns true if no node of type 'T_Node' is
// present. Traversal is performed in some arbitrary order and is terminated as soon as the
// result can be determined.
template <typename T_Node> bool forall(std::function<bool(T_Node*)> p) {
checkTypeParameter<T_Node>();
return predicateImpl<T_Node, /* Default: */ true, /* VisitNext: */ false>(this, p);
template <typename T_Node>
bool forall(std::function<bool(T_Node*)> p) {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
return predicateImpl<T_Node, /* Default: */ true>(this, p);
}
// Same as above, but for 'const' nodes
template <typename T_Node> void forall(std::function<bool(const T_Node*)> p) const {
checkTypeParameter<T_Node>();
return predicateImpl<const T_Node, /* Default: */ true, /* VisitNext: */ false>(this, p);
template <typename T_Node>
bool forall(std::function<bool(const T_Node*)> p) const {
static_assert(checkTypeParameter<T_Node>(), "Invalid type parameter 'T_Node'");
return predicateImpl<const T_Node, /* Default: */ true>(this, p);
}
int nodeCount() const {
@@ -2098,22 +2059,214 @@ public:
// Specialisations of privateTypeTest
#include "V3Ast__gen_impl.h" // From ./astgen
// Specializations of privateMayBeUnder
template <> inline bool AstNode::privateMayBeUnder<AstCell>(const AstNode* nodep) {
// Specializations of AstNode::mayBeUnder
template <>
inline bool AstNode::mayBeUnder<AstCell>(const AstNode* nodep) {
return !VN_IS(nodep, NodeStmt) && !VN_IS(nodep, NodeMath);
}
template <> inline bool AstNode::privateMayBeUnder<AstNodeAssign>(const AstNode* nodep) {
template <>
inline bool AstNode::mayBeUnder<AstNodeAssign>(const AstNode* nodep) {
return !VN_IS(nodep, NodeMath);
}
template <> inline bool AstNode::privateMayBeUnder<AstVarScope>(const AstNode* nodep) {
return !VN_IS(nodep, NodeStmt) && !VN_IS(nodep, NodeMath);
template <>
inline bool AstNode::mayBeUnder<AstVarScope>(const AstNode* nodep) {
if (VN_IS(nodep, VarScope)) return false; // Should not nest
if (VN_IS(nodep, Var)) return false;
if (VN_IS(nodep, Active)) return false;
if (VN_IS(nodep, NodeStmt)) return false;
if (VN_IS(nodep, NodeMath)) return false;
return true;
}
template <> inline bool AstNode::privateMayBeUnder<AstExecGraph>(const AstNode* nodep) {
template <>
inline bool AstNode::mayBeUnder<AstExecGraph>(const AstNode* nodep) {
if (VN_IS(nodep, ExecGraph)) return false; // Should not nest
if (VN_IS(nodep, NodeStmt)) return false; // Should be directly under CFunc
return true;
}
// Specializations of AstNode::isLeaf
template <>
constexpr bool AstNode::isLeaf<AstNodeVarRef>() {
return true;
}
template <>
constexpr bool AstNode::isLeaf<AstVarRef>() {
return true;
}
template <>
constexpr bool AstNode::isLeaf<AstVarXRef>() {
return true;
}
// foreach implementation
template <typename T_Arg>
void AstNode::foreachImpl(ConstCorrectAstNode<T_Arg>* nodep, const std::function<void(T_Arg*)>& f,
bool visitNext) {
// Checking the function is bound up front eliminates this check from the loop at invocation
if (!f) {
nodep->v3fatal("AstNode::foreach called with unbound function"); // LCOV_EXCL_LINE
} else {
// Pre-order traversal implemented directly (without recursion) for speed reasons. The very
// first iteration (the one that operates on the input nodep) is special, as we might or
// might not need to enqueue nodep->nextp() depending on VisitNext, while in all other
// iterations, we do want to enqueue nodep->nextp(). Duplicating code (via
// 'foreachImplVisit') for the initial iteration here to avoid an extra branch in the loop
using T_Arg_NonConst = typename std::remove_const<T_Arg>::type;
using Node = ConstCorrectAstNode<T_Arg>;
// Traversal stack
std::vector<Node*> stack; // Kept as a vector for easy resizing
Node** basep = nullptr; // Pointer to base of stack
Node** topp = nullptr; // Pointer to top of stack
Node** limp = nullptr; // Pointer to stack limit (when need growing)
// We prefetch this far into the stack
constexpr int prefetchDistance = 2;
// Grow stack to given size
const auto grow = [&](size_t size) VL_ATTR_ALWINLINE {
const ptrdiff_t occupancy = topp - basep;
stack.resize(size);
basep = stack.data() + prefetchDistance;
topp = basep + occupancy;
limp = basep + size - 5; // We push max 5 items per iteration
};
// Initial stack size
grow(32);
// We want some non-null pointers at the beginning. These will be prefetched, but not
// visited, so the root node will suffice. This eliminates needing branches in the loop.
for (int i = -prefetchDistance; i; ++i) basep[i] = nodep;
// Visit given node, enqueue children for traversal
const auto visit = [&](Node* currp) VL_ATTR_ALWINLINE {
// Type test this node
if (AstNode::privateTypeTest<T_Arg_NonConst>(currp)) {
// Call the client function
f(static_cast<T_Arg*>(currp));
// Short circuit if iterating leaf nodes
if VL_CONSTEXPR_CXX17 (isLeaf<T_Arg_NonConst>()) return;
}
// Enqueue children for traversal, unless futile
if (mayBeUnder<T_Arg_NonConst>(currp)) {
if (AstNode* const op4p = currp->op4p()) *topp++ = op4p;
if (AstNode* const op3p = currp->op3p()) *topp++ = op3p;
if (AstNode* const op2p = currp->op2p()) *topp++ = op2p;
if (AstNode* const op1p = currp->op1p()) *topp++ = op1p;
}
};
// Enqueue the next of the root node, if required
if (visitNext && nodep->nextp()) *topp++ = nodep->nextp();
// Visit the root node
visit(nodep);
// Visit the rest of the tree
while (VL_LIKELY(topp > basep)) {
// Pop next node in the traversal
Node* const headp = *--topp;
// Prefetch in case we are ascending the tree
ASTNODE_PREFETCH_NON_NULL(topp[-prefetchDistance]);
// Ensure we have stack space for nextp and the 4 children
if (VL_UNLIKELY(topp >= limp)) grow(stack.size() * 2);
// Enqueue the next node
if (headp->nextp()) *topp++ = headp->nextp();
// Visit the head node
visit(headp);
}
}
}
// predicate implementation
template <typename T_Arg, bool Default>
bool AstNode::predicateImpl(ConstCorrectAstNode<T_Arg>* nodep,
const std::function<bool(T_Arg*)>& p) {
// Implementation similar to foreach, but abort traversal as soon as result is determined.
if (!p) {
nodep->v3fatal("AstNode::foreach called with unbound function"); // LCOV_EXCL_LINE
} else {
using T_Arg_NonConst = typename std::remove_const<T_Arg>::type;
using Node = ConstCorrectAstNode<T_Arg>;
// Traversal stack
std::vector<Node*> stack; // Kept as a vector for easy resizing
Node** basep = nullptr; // Pointer to base of stack
Node** topp = nullptr; // Pointer to top of stack
Node** limp = nullptr; // Pointer to stack limit (when need growing)
// We prefetch this far into the stack
constexpr int prefetchDistance = 2;
// Grow stack to given size
const auto grow = [&](size_t size) VL_ATTR_ALWINLINE {
const ptrdiff_t occupancy = topp - basep;
stack.resize(size);
basep = stack.data() + prefetchDistance;
topp = basep + occupancy;
limp = basep + size - 5; // We push max 5 items per iteration
};
// Initial stack size
grow(32);
// We want some non-null pointers at the beginning. These will be prefetched, but not
// visited, so the root node will suffice. This eliminates needing branches in the loop.
for (int i = -prefetchDistance; i; ++i) basep[i] = nodep;
// Visit given node, enqueue children for traversal, return true if result determined.
const auto visit = [&](Node* currp) VL_ATTR_ALWINLINE {
// Type test this node
if (AstNode::privateTypeTest<T_Arg_NonConst>(currp)) {
// Call the client function
if (p(static_cast<T_Arg*>(currp)) != Default) return true;
// Short circuit if iterating leaf nodes
if VL_CONSTEXPR_CXX17 (isLeaf<T_Arg_NonConst>()) return false;
}
// Enqueue children for traversal, unless futile
if (mayBeUnder<T_Arg_NonConst>(currp)) {
if (AstNode* const op4p = currp->op4p()) *topp++ = op4p;
if (AstNode* const op3p = currp->op3p()) *topp++ = op3p;
if (AstNode* const op2p = currp->op2p()) *topp++ = op2p;
if (AstNode* const op1p = currp->op1p()) *topp++ = op1p;
}
return false;
};
// Visit the root node
if (visit(nodep)) return !Default;
// Visit the rest of the tree
while (VL_LIKELY(topp > basep)) {
// Pop next node in the traversal
Node* const headp = *--topp;
// Prefetch in case we are ascending the tree
ASTNODE_PREFETCH_NON_NULL(topp[-prefetchDistance]);
// Ensure we have stack space for nextp and the 4 children
if (VL_UNLIKELY(topp >= limp)) grow(stack.size() * 2);
// Enqueue the next node
if (headp->nextp()) *topp++ = headp->nextp();
// Visit the head node
if (visit(headp)) return !Default;
}
return Default;
}
}
inline std::ostream& operator<<(std::ostream& os, const AstNode* rhs) {
if (!rhs) {
os << "nullptr";
@@ -2127,7 +2280,7 @@ inline void VNRelinker::relink(AstNode* newp) { newp->AstNode::relink(this); }
//######################################################################
// VNRef is std::reference_wrapper that can only hold AstNode subtypes
template <typename T_Node> //
template <typename T_Node>
class VNRef final : public std::reference_wrapper<T_Node> {
static_assert(std::is_base_of<AstNode, T_Node>::value,
"Type parameter 'T_Node' must be a subtype of AstNode");
@@ -2150,16 +2303,16 @@ static_assert(sizeof(VNRef<AstNode>) == sizeof(std::reference_wrapper<AstNode>),
// without having to copy nodes into the collections.
// Forward declaration to avoid including V3Hasher.h which needs V3Ast.h (this file).
size_t V3HasherUncachedHash(AstNode&);
size_t V3HasherUncachedHash(const AstNode&);
// Specialization of std::hash for VNRef
template <typename T_Node> //
template <typename T_Node>
struct std::hash<VNRef<T_Node>> final {
size_t operator()(VNRef<T_Node> r) const { return V3HasherUncachedHash(r); }
};
// Specialization of std::equal_to for VNRef
template <typename T_Node> //
template <typename T_Node>
struct std::equal_to<VNRef<T_Node>> final {
size_t operator()(VNRef<T_Node> ra, VNRef<T_Node> rb) const {
return ra.get().sameTree(&(rb.get()));
@@ -2423,6 +2576,7 @@ public:
AstNode* stmtsp() const { return op1p(); } // op1 = List of statements
void addStmtsp(AstNode* nodep) { addNOp1p(nodep); }
bool unnamed() const { return m_unnamed; }
bool isFirstInMyListOfStatements(AstNode* nodep) const override { return nodep == stmtsp(); }
};
class AstNodePreSel VL_NOT_FINAL : public AstNode {
@@ -2444,7 +2598,7 @@ public:
void fromp(AstNode* nodep) { return setOp1p(nodep); }
void rhsp(AstNode* nodep) { return setOp2p(nodep); }
void thsp(AstNode* nodep) { return setOp3p(nodep); }
void attrp(AstAttrOf* nodep) { return setOp4p((AstNode*)nodep); }
void attrp(AstAttrOf* nodep) { return setOp4p(reinterpret_cast<AstNode*>(nodep)); }
// METHODS
virtual bool same(const AstNode*) const override { return true; }
};
@@ -2488,10 +2642,12 @@ public:
class AstNodeAssign VL_NOT_FINAL : public AstNodeStmt {
protected:
AstNodeAssign(VNType t, FileLine* fl, AstNode* lhsp, AstNode* rhsp)
AstNodeAssign(VNType t, FileLine* fl, AstNode* lhsp, AstNode* rhsp,
AstNode* timingControlp = nullptr)
: AstNodeStmt{t, fl} {
setOp1p(rhsp);
setOp2p(lhsp);
addNOp3p(timingControlp);
dtypeFrom(lhsp);
}
@@ -2502,6 +2658,9 @@ public:
// So iteration hits the RHS which is "earlier" in execution order, it's op1, not op2
AstNode* rhsp() const { return op1p(); } // op1 = Assign from
AstNode* lhsp() const { return op2p(); } // op2 = Assign to
// op3 = Timing controls (delays, event controls)
AstNode* timingControlp() const { return op3p(); }
void addTimingControlp(AstNode* const np) { addNOp3p(np); }
void rhsp(AstNode* np) { setOp1p(np); }
void lhsp(AstNode* np) { setOp2p(np); }
virtual bool hasDType() const override { return true; }
@@ -2531,7 +2690,7 @@ public:
AstNode* bodysp() const { return op4p(); } // op4 = body of loop
virtual bool isGateOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_BRANCH; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstNodeIf VL_NOT_FINAL : public AstNodeStmt {
@@ -2558,11 +2717,14 @@ public:
virtual bool isGateOptimizable() const override { return false; }
virtual bool isGateDedupable() const override { return true; }
virtual int instrCount() const override { return INSTR_COUNT_BRANCH; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
void branchPred(VBranchPred flag) { m_branchPred = flag; }
VBranchPred branchPred() const { return m_branchPred; }
void isBoundsCheck(bool flag) { m_isBoundsCheck = flag; }
bool isBoundsCheck() const { return m_isBoundsCheck; }
bool isFirstInMyListOfStatements(AstNode* n) const override {
return n == ifsp() || n == elsesp();
}
};
class AstNodeCase VL_NOT_FINAL : public AstNodeStmt {
@@ -2600,7 +2762,7 @@ protected:
: AstNodeMath{t, fl}
, m_access{access}
, m_name{name} {
this->varp(nullptr);
varp(nullptr);
}
AstNodeVarRef(VNType t, FileLine* fl, const string& name, AstVar* varp, const VAccess& access)
: AstNodeMath{t, fl}
@@ -3089,6 +3251,7 @@ public:
bool isVirtual() const { return m_virtual; }
void lifetime(const VLifetime& flag) { m_lifetime = flag; }
VLifetime lifetime() const { return m_lifetime; }
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == stmtsp(); }
};
class AstNodeFTaskRef VL_NOT_FINAL : public AstNodeStmt {
+30 -14
View File
@@ -481,7 +481,7 @@ string AstVar::cPubArgType(bool named, bool forReturn) const {
}
} else {
// Newer internal-compatible types
arg += dtypep()->cType((named ? name() : string{}), true, isRef);
arg += dtypep()->cType((named ? name() : std::string{}), true, isRef);
}
return arg;
}
@@ -489,7 +489,7 @@ string AstVar::cPubArgType(bool named, bool forReturn) const {
class dpiTypesToStringConverter VL_NOT_FINAL {
public:
virtual string openArray(const AstVar*) const { return "const svOpenArrayHandle"; }
virtual string bitLogicVector(const AstVar* varp, bool isBit) const {
virtual string bitLogicVector(const AstVar* /*varp*/, bool isBit) const {
return isBit ? "svBitVecVal" : "svLogicVecVal";
}
virtual string primitive(const AstVar* varp) const {
@@ -659,7 +659,7 @@ public:
}
};
string AstNodeDType::cType(const string& name, bool forFunc, bool isRef) const {
string AstNodeDType::cType(const string& name, bool /*forFunc*/, bool isRef) const {
const CTypeRecursed info = cTypeRecurse(false);
return info.render(name, isRef);
}
@@ -673,6 +673,9 @@ AstNodeDType::CTypeRecursed AstNodeDType::cTypeRecurse(bool compound) const {
const CTypeRecursed key = adtypep->keyDTypep()->cTypeRecurse(true);
const CTypeRecursed val = adtypep->subDTypep()->cTypeRecurse(true);
info.m_type = "VlAssocArray<" + key.m_type + ", " + val.m_type + ">";
} else if (const auto* const adtypep = VN_CAST(dtypep, WildcardArrayDType)) {
const CTypeRecursed sub = adtypep->subDTypep()->cTypeRecurse(true);
info.m_type = "VlAssocArray<std::string, " + sub.m_type + ">";
} else if (const auto* const adtypep = VN_CAST(dtypep, DynArrayDType)) {
const CTypeRecursed sub = adtypep->subDTypep()->cTypeRecurse(true);
info.m_type = "VlQueue<" + sub.m_type + ">";
@@ -1309,7 +1312,8 @@ void AstClassPackage::cloneRelink() {
}
void AstClass::insertCache(AstNode* nodep) {
const bool doit = (VN_IS(nodep, Var) || VN_IS(nodep, EnumItemRef)
|| (VN_IS(nodep, NodeFTask) && !VN_AS(nodep, NodeFTask)->isExternProto()));
|| (VN_IS(nodep, NodeFTask) && !VN_AS(nodep, NodeFTask)->isExternProto())
|| VN_IS(nodep, CFunc));
if (doit) {
if (m_members.find(nodep->name()) != m_members.end()) {
nodep->v3error("Duplicate declaration of member name: " << nodep->prettyNameQ());
@@ -1320,7 +1324,15 @@ void AstClass::insertCache(AstNode* nodep) {
}
void AstClass::repairCache() {
clearCache();
for (AstNode* itemp = membersp(); itemp; itemp = itemp->nextp()) { insertCache(itemp); }
for (auto* itemp = membersp(); itemp; itemp = itemp->nextp()) {
if (const auto* const scopep = VN_CAST(itemp, Scope)) {
for (auto* blockp = scopep->blocksp(); blockp; blockp = blockp->nextp()) {
insertCache(blockp);
}
} else {
insertCache(itemp);
}
}
}
bool AstClass::isClassExtendedFrom(const AstClass* refClassp, const AstClass* baseClassp) {
// TAIL RECURSIVE
@@ -1343,7 +1355,7 @@ AstClass* AstClassExtends::classp() const {
return refp->classp();
}
void AstClassRefDType::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeDType::dump(str);
if (classOrPackagep()) str << " cpkg=" << nodeAddr(classOrPackagep());
if (classp()) {
str << " -> ";
@@ -1374,7 +1386,7 @@ void AstEnumItemRef::dump(std::ostream& str) const {
}
}
void AstIfaceRefDType::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeDType::dump(str);
if (cellName() != "") str << " cell=" << cellName();
if (ifaceName() != "") str << " if=" << ifaceName();
if (modportName() != "") str << " mp=" << modportName();
@@ -1423,7 +1435,7 @@ void AstJumpLabel::dump(std::ostream& str) const {
}
}
void AstLogOr::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeMath::dump(str);
if (sideEffect()) str << " [SIDE]";
}
void AstMemberSel::dump(std::ostream& str) const {
@@ -1436,7 +1448,7 @@ void AstMemberSel::dump(std::ostream& str) const {
}
}
void AstMethodCall::dump(std::ostream& str) const {
this->AstNodeStmt::dump(str);
this->AstNodeFTaskRef::dump(str);
if (isStatement()) str << " [STMT]";
str << " -> ";
if (taskp()) {
@@ -1522,7 +1534,7 @@ void AstRefDType::dump(std::ostream& str) const {
}
}
void AstNodeUOrStructDType::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeDType::dump(str);
if (packed()) str << " [PACKED]";
if (isFourstate()) str << " [4STATE]";
}
@@ -1613,7 +1625,7 @@ void AstPackageImport::dump(std::ostream& str) const {
str << " -> " << packagep();
}
void AstPatMember::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeMath::dump(str);
if (isDefault()) str << " [DEFAULT]";
}
void AstNodeTriop::dump(std::ostream& str) const { this->AstNodeMath::dump(str); }
@@ -1675,6 +1687,10 @@ string AstQueueDType::prettyDTypeName() const {
if (boundConst()) str += ":" + cvtToStr(boundConst());
return str + "]";
}
void AstWildcardArrayDType::dumpSmall(std::ostream& str) const {
this->AstNodeDType::dumpSmall(str);
str << "[*]";
}
void AstUnsizedArrayDType::dumpSmall(std::ostream& str) const {
this->AstNodeDType::dumpSmall(str);
str << "[]";
@@ -1758,7 +1774,7 @@ void AstScope::dump(std::ostream& str) const {
str << " [modp=" << reinterpret_cast<const void*>(modp()) << "]";
}
void AstScopeName::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeMath::dump(str);
if (dpiExport()) str << " [DPIEX]";
if (forFormat()) str << " [FMT]";
}
@@ -1825,7 +1841,7 @@ void AstNodeBlock::dump(std::ostream& str) const {
if (unnamed()) str << " [UNNAMED]";
}
void AstBegin::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeBlock::dump(str);
if (generate()) str << " [GEN]";
if (genforp()) str << " [GENFOR]";
if (implied()) str << " [IMPLIED]";
@@ -1851,7 +1867,7 @@ void AstCoverInc::dump(std::ostream& str) const {
}
}
void AstFork::dump(std::ostream& str) const {
this->AstNode::dump(str);
this->AstNodeBlock::dump(str);
if (!joinType().join()) str << " [" << joinType() << "]";
}
void AstTraceDecl::dump(std::ostream& str) const {
+252 -96
View File
@@ -77,10 +77,11 @@ public:
, m_num(this, width, value) {
initWithNumber();
}
class DtypedValue {}; // for creator type-overload selection
AstConst(FileLine* fl, DtypedValue, AstNodeDType* nodedtypep, uint32_t value)
class DTyped {}; // for creator type-overload selection
// Zero/empty constant with a type matching nodetypep
AstConst(FileLine* fl, DTyped, const AstNodeDType* nodedtypep)
: ASTGEN_SUPER_Const(fl)
, m_num(this, nodedtypep->width(), value, nodedtypep->widthSized()) {
, m_num(this, nodedtypep) {
initWithNumber();
}
class StringToParse {}; // for creator type-overload selection
@@ -186,20 +187,20 @@ public:
class AstEmpty final : public AstNode {
// Represents something missing, e.g. a missing argument in FOREACH
public:
AstEmpty(FileLine* fl)
explicit AstEmpty(FileLine* fl)
: ASTGEN_SUPER_Empty(fl) {}
ASTNODE_NODE_FUNCS(Empty)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstEmptyQueue final : public AstNodeMath {
public:
AstEmptyQueue(FileLine* fl)
explicit AstEmptyQueue(FileLine* fl)
: ASTGEN_SUPER_EmptyQueue(fl) {}
ASTNODE_NODE_FUNCS(EmptyQueue)
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual string emitVerilog() override { return "{}"; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual bool cleanOut() const override { return true; }
};
@@ -246,7 +247,7 @@ public:
bool littleEndian() const { return leftConst() < rightConst(); }
virtual void dump(std::ostream& str) const override;
virtual string emitC() { V3ERROR_NA_RETURN(""); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstBracketRange final : public AstNodeRange {
@@ -260,7 +261,7 @@ public:
ASTNODE_NODE_FUNCS(BracketRange)
virtual string emitC() { V3ERROR_NA_RETURN(""); }
virtual string emitVerilog() { V3ERROR_NA_RETURN(""); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
// Will be removed in V3Width, which relies on this
// being a child not a dtype pointed node
virtual bool maybePointedTo() const override { return false; }
@@ -275,7 +276,18 @@ public:
ASTNODE_NODE_FUNCS(UnsizedRange)
virtual string emitC() { V3ERROR_NA_RETURN(""); }
virtual string emitVerilog() { return "[]"; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstWildcardRange final : public AstNodeRange {
// Wildcard range specification, for wildcard index type associative arrays
public:
explicit AstWildcardRange(FileLine* fl)
: ASTGEN_SUPER_WildcardRange(fl) {}
ASTNODE_NODE_FUNCS(WildcardRange)
virtual string emitC() { V3ERROR_NA_RETURN(""); }
virtual string emitVerilog() { return "[*]"; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstGatePin final : public AstNodeMath {
@@ -649,7 +661,7 @@ class AstDynArrayDType final : public AstNodeDType {
// Dynamic array data type, ie "[]"
// Children: DTYPE (moved to refDTypep() in V3Width)
private:
AstNodeDType* m_refDTypep; // Elements of this type (after widthing)
AstNodeDType* m_refDTypep = nullptr; // Elements of this type (after widthing)
public:
AstDynArrayDType(FileLine* fl, VFlagChildDType, AstNodeDType* dtp)
: ASTGEN_SUPER_DynArrayDType(fl) {
@@ -819,6 +831,62 @@ public:
virtual bool isCompound() const override { return true; }
};
class AstWildcardArrayDType final : public AstNodeDType {
// Wildcard index type associative array data type, ie "some_dtype var_name [*]"
// Children: DTYPE (moved to refDTypep() in V3Width)
private:
AstNodeDType* m_refDTypep; // Elements of this type (after widthing)
public:
AstWildcardArrayDType(FileLine* fl, VFlagChildDType, AstNodeDType* dtp)
: ASTGEN_SUPER_WildcardArrayDType(fl) {
childDTypep(dtp); // Only for parser
refDTypep(nullptr);
dtypep(nullptr); // V3Width will resolve
}
ASTNODE_NODE_FUNCS(WildcardArrayDType)
virtual const char* broken() const override {
BROKEN_RTN(!((m_refDTypep && !childDTypep() && m_refDTypep->brokeExists())
|| (!m_refDTypep && childDTypep())));
return nullptr;
}
virtual void cloneRelink() override {
if (m_refDTypep && m_refDTypep->clonep()) m_refDTypep = m_refDTypep->clonep();
}
virtual bool same(const AstNode* samep) const override {
const AstNodeArrayDType* const asamep = static_cast<const AstNodeArrayDType*>(samep);
if (!asamep->subDTypep()) return false;
return (subDTypep() == asamep->subDTypep());
}
virtual bool similarDType(AstNodeDType* samep) const override {
const AstNodeArrayDType* const asamep = static_cast<const AstNodeArrayDType*>(samep);
return type() == samep->type() && asamep->subDTypep()
&& subDTypep()->skipRefp()->similarDType(asamep->subDTypep()->skipRefp());
}
virtual void dumpSmall(std::ostream& str) const override;
virtual AstNodeDType* getChildDTypep() const override { return childDTypep(); }
// op1 = Range of variable
AstNodeDType* childDTypep() const { return VN_AS(op1p(), NodeDType); }
void childDTypep(AstNodeDType* nodep) { setOp1p(nodep); }
virtual AstNodeDType* subDTypep() const override {
return m_refDTypep ? m_refDTypep : childDTypep();
}
void refDTypep(AstNodeDType* nodep) { m_refDTypep = nodep; }
virtual AstNodeDType* virtRefDTypep() const override { return m_refDTypep; }
virtual void virtRefDTypep(AstNodeDType* nodep) override { refDTypep(nodep); }
// METHODS
virtual AstBasicDType* basicp() const override { return subDTypep()->basicp(); }
virtual AstNodeDType* skipRefp() const override { return (AstNodeDType*)this; }
virtual AstNodeDType* skipRefToConstp() const override { return (AstNodeDType*)this; }
virtual AstNodeDType* skipRefToEnump() const override { return (AstNodeDType*)this; }
virtual int widthAlignBytes() const override {
return sizeof(std::map<std::string, std::string>);
}
virtual int widthTotalBytes() const override {
return sizeof(std::map<std::string, std::string>);
}
virtual bool isCompound() const override { return true; }
};
class AstBasicDType final : public AstNodeDType {
// Builtin atomic/vectored data type
// Children: RANGE (converted to constant in V3Width)
@@ -975,7 +1043,7 @@ class AstConstDType final : public AstNodeDType {
// ConstDType are removed in V3LinkLValue and become AstVar::isConst.
// When more generic types are supported AstConstDType will be propagated further.
private:
AstNodeDType* m_refDTypep; // Inherit from this base data type
AstNodeDType* m_refDTypep = nullptr; // Inherit from this base data type
public:
AstConstDType(FileLine* fl, VFlagChildDType, AstNodeDType* dtp)
: ASTGEN_SUPER_ConstDType(fl) {
@@ -1134,7 +1202,7 @@ class AstQueueDType final : public AstNodeDType {
// Queue array data type, ie "[ $ ]"
// Children: DTYPE (moved to refDTypep() in V3Width)
private:
AstNodeDType* m_refDTypep; // Elements of this type (after widthing)
AstNodeDType* m_refDTypep = nullptr; // Elements of this type (after widthing)
public:
AstQueueDType(FileLine* fl, VFlagChildDType, AstNodeDType* dtp, AstNode* boundp)
: ASTGEN_SUPER_QueueDType(fl) {
@@ -1328,7 +1396,7 @@ class AstMemberDType final : public AstNodeDType {
// A member of a struct/union
// PARENT: AstNodeUOrStructDType
private:
AstNodeDType* m_refDTypep; // Elements of this type (after widthing)
AstNodeDType* m_refDTypep = nullptr; // Elements of this type (after widthing)
string m_name; // Name of variable
string m_tag; // Holds the string of the verilator tag -- used in XML output.
int m_lsb = -1; // Within this level's packed struct, the LSB of the first bit of the member
@@ -1517,8 +1585,8 @@ class AstEnumDType final : public AstNodeDType {
// Children: ENUMVALUEs
private:
string m_name; // Name from upper typedef, if any
AstNodeDType* m_refDTypep; // Elements are of this type after V3Width
const int m_uniqueNum;
AstNodeDType* m_refDTypep = nullptr; // Elements are of this type after V3Width
const int m_uniqueNum = 0;
public:
AstEnumDType(FileLine* fl, VFlagChildDType, AstNodeDType* dtp, AstNode* itemsp)
@@ -1641,7 +1709,7 @@ public:
return true;
} // esp for V3Const::ifSameAssign
virtual bool isPredictOptimizable() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
// Special operators
// Return base var (or const) nodep dereferences
@@ -1682,7 +1750,45 @@ public:
return true;
} // esp for V3Const::ifSameAssign
virtual bool isPredictOptimizable() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
};
class AstWildcardSel final : public AstNodeSel {
// Parents: math|stmt
// Children: varref|arraysel, math
private:
void init(AstNode* fromp) {
if (fromp && VN_IS(fromp->dtypep()->skipRefp(), WildcardArrayDType)) {
// Strip off array to find what array references
dtypeFrom(VN_AS(fromp->dtypep()->skipRefp(), WildcardArrayDType)->subDTypep());
}
}
public:
AstWildcardSel(FileLine* fl, AstNode* fromp, AstNode* bitp)
: ASTGEN_SUPER_WildcardSel(fl, fromp, bitp) {
init(fromp);
}
ASTNODE_NODE_FUNCS(WildcardSel)
virtual AstNode* cloneType(AstNode* lhsp, AstNode* rhsp) override {
return new AstWildcardSel{this->fileline(), lhsp, rhsp};
}
virtual void numberOperate(V3Number& out, const V3Number& lhs, const V3Number& rhs) override {
V3ERROR_NA;
}
virtual string emitVerilog() override { return "%k(%l%f[%r])"; }
virtual string emitC() override { return "%li%k[%ri]"; }
virtual bool cleanOut() const override { return true; }
virtual bool cleanLhs() const override { return false; }
virtual bool cleanRhs() const override { return true; }
virtual bool sizeMattersLhs() const override { return false; }
virtual bool sizeMattersRhs() const override { return false; }
virtual bool isGateOptimizable() const override {
return true;
} // esp for V3Const::ifSameAssign
virtual bool isPredictOptimizable() const override { return false; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
};
@@ -1709,7 +1815,7 @@ public:
virtual bool cleanRhs() const override { return true; }
virtual bool sizeMattersLhs() const override { return false; }
virtual bool sizeMattersRhs() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstSelLoopVars final : public AstNode {
@@ -1722,7 +1828,7 @@ public:
addNOp2p(elementsp);
}
ASTNODE_NODE_FUNCS(SelLoopVars)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual bool maybePointedTo() const override { return false; }
AstNode* fromp() const { return op1p(); }
void fromp(AstNode* nodep) { setOp1p(nodep); }
@@ -2142,6 +2248,9 @@ public:
virtual AstNodeDType* getChildDTypep() const override { return childDTypep(); }
// op1 = Range of variable
AstNodeDType* childDTypep() const { return VN_AS(op1p(), NodeDType); }
// op2 = Net delay
AstNode* delayp() const { return op2p(); }
void delayp(AstNode* const nodep) { setNOp2p(nodep); }
AstNodeDType* dtypeSkipRefp() const { return subDTypep()->skipRefp(); }
// (Slow) recurse down to find basic data type (Note don't need virtual -
// AstVar isn't a NodeDType)
@@ -3126,8 +3235,7 @@ public:
AstNodeModule* classOrPackagep() const {
AstNode* foundp = m_classOrPackageNodep;
while (auto* const anodep = VN_CAST(foundp, Typedef)) foundp = anodep->subDTypep();
while (auto* const anodep = VN_CAST(foundp, ClassRefDType))
foundp = anodep->classOrPackagep();
if (auto* const anodep = VN_CAST(foundp, ClassRefDType)) foundp = anodep->classp();
return VN_CAST(foundp, NodeModule);
}
AstPackage* packagep() const { return VN_CAST(classOrPackageNodep(), Package); }
@@ -3249,7 +3357,7 @@ public:
addNOp2p(exprp);
}
ASTNODE_NODE_FUNCS(WithParse)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
//
AstNode* funcrefp() const { return op1p(); }
AstNode* exprp() const { return op2p(); }
@@ -3269,7 +3377,7 @@ public:
, m_name{name}
, m_index(index) {}
ASTNODE_NODE_FUNCS(LambdaArgRef)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual string emitVerilog() override { return name(); }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual bool cleanOut() const override { return true; }
@@ -3296,7 +3404,7 @@ public:
addNOp3p(exprp);
}
ASTNODE_NODE_FUNCS(With)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual bool hasDType() const override { return true; }
virtual const char* broken() const override {
BROKEN_RTN(!indexArgRefp()); // varp needed to know lambda's arg dtype
@@ -3470,19 +3578,20 @@ public:
addNOp2p(bodysp);
}
ASTNODE_NODE_FUNCS(AlwaysPublic)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
//
AstSenTree* sensesp() const { return VN_AS(op1p(), SenTree); } // op1 = Sensitivity list
AstNode* bodysp() const { return op2p(); } // op2 = Statements to evaluate
void addStmtp(AstNode* nodep) { addOp2p(nodep); }
// Special accessors
bool isJustOneBodyStmt() const { return bodysp() && !bodysp()->nextp(); }
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstAssign final : public AstNodeAssign {
public:
AstAssign(FileLine* fl, AstNode* lhsp, AstNode* rhsp)
: ASTGEN_SUPER_Assign(fl, lhsp, rhsp) {
AstAssign(FileLine* fl, AstNode* lhsp, AstNode* rhsp, AstNode* timingControlp = nullptr)
: ASTGEN_SUPER_Assign(fl, lhsp, rhsp, timingControlp) {
dtypeFrom(lhsp);
}
ASTNODE_NODE_FUNCS(Assign)
@@ -3507,8 +3616,8 @@ public:
class AstAssignDly final : public AstNodeAssign {
public:
AstAssignDly(FileLine* fl, AstNode* lhsp, AstNode* rhsp)
: ASTGEN_SUPER_AssignDly(fl, lhsp, rhsp) {}
AstAssignDly(FileLine* fl, AstNode* lhsp, AstNode* rhsp, AstNode* timingControlp = nullptr)
: ASTGEN_SUPER_AssignDly(fl, lhsp, rhsp, timingControlp) {}
ASTNODE_NODE_FUNCS(AssignDly)
virtual AstNode* cloneType(AstNode* lhsp, AstNode* rhsp) override {
return new AstAssignDly(this->fileline(), lhsp, rhsp);
@@ -3801,7 +3910,7 @@ public:
}
ASTNODE_NODE_FUNCS(CoverToggle)
virtual int instrCount() const override { return 3 + INSTR_COUNT_BRANCH + INSTR_COUNT_LD; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return true; }
virtual bool isOutputter() const override {
@@ -3817,15 +3926,18 @@ public:
class AstDelay final : public AstNodeStmt {
// Delay statement
public:
AstDelay(FileLine* fl, AstNode* lhsp)
AstDelay(FileLine* fl, AstNode* lhsp, AstNode* stmtsp)
: ASTGEN_SUPER_Delay(fl) {
setOp1p(lhsp);
setNOp2p(stmtsp);
}
ASTNODE_NODE_FUNCS(Delay)
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
//
AstNode* lhsp() const { return op1p(); } // op2 = Statements to evaluate
AstNode* lhsp() const { return op1p(); } // op1 = delay value
void lhsp(AstNode* nodep) { setOp1p(nodep); }
void stmtsp(AstNode* nodep) { setOp2p(nodep); } // op2 = statements under delay
AstNode* stmtsp() const { return op2p(); }
};
class AstGenCase final : public AstNodeCase {
@@ -3897,6 +4009,7 @@ public:
void condsp(AstNode* nodep) { setOp1p(nodep); }
void addBodysp(AstNode* newp) { addOp2p(newp); }
bool isDefault() const { return condsp() == nullptr; }
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstSFormatF final : public AstNode {
@@ -4026,7 +4139,7 @@ public:
virtual bool isPredictOptimizable() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool cleanOut() const { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
VDumpCtlType ctlType() const { return m_ctlType; }
AstNode* exprp() const { return op1p(); } // op2 = Expressions to output
void exprp(AstNode* nodep) { setOp1p(nodep); }
@@ -4097,7 +4210,7 @@ public:
virtual bool isOutputter() const override { return false; }
virtual bool cleanOut() const { return false; }
virtual int instrCount() const override { return INSTR_COUNT_PLI; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
void fmtp(AstSFormatF* nodep) { addOp1p(nodep); } // op1 = To-String formatter
AstSFormatF* fmtp() const { return VN_AS(op1p(), SFormatF); }
AstNode* lhsp() const { return op3p(); }
@@ -4120,7 +4233,7 @@ public:
virtual bool isPure() const override { return true; }
virtual bool isOutputter() const override { return false; }
virtual int instrCount() const override { return 0; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* lhsp() const { return op1p(); } // op1 = Expressions to eval
void lhsp(AstNode* nodep) { addOp1p(nodep); } // op1 = Expressions to eval
};
@@ -4161,7 +4274,7 @@ public:
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op2p(); }
void filep(AstNodeVarRef* nodep) { setNOp2p(nodep); }
};
@@ -4182,7 +4295,7 @@ public:
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op1p(); }
AstNode* filenamep() const { return op2p(); }
AstNode* modep() const { return op3p(); }
@@ -4203,7 +4316,7 @@ public:
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op1p(); }
AstNode* filenamep() const { return op2p(); }
};
@@ -4223,7 +4336,7 @@ public:
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op2p(); }
void filep(AstNodeVarRef* nodep) { setNOp2p(nodep); }
};
@@ -4251,7 +4364,7 @@ public:
virtual bool isPure() const override { return false; } // SPECIAL: has 'visual' ordering
virtual bool isOutputter() const override { return true; } // SPECIAL: makes output
virtual bool cleanOut() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* memp() const { return op1p(); }
void memp(AstNode* nodep) { setOp1p(nodep); }
AstNode* filep() const { return op2p(); }
@@ -4280,7 +4393,7 @@ public:
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool cleanOut() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op2p(); }
void filep(AstNodeVarRef* nodep) { setNOp2p(nodep); }
};
@@ -4303,7 +4416,7 @@ public:
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool cleanOut() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op2p(); }
void filep(AstNodeVarRef* nodep) { setNOp2p(nodep); }
};
@@ -4329,7 +4442,7 @@ public:
virtual bool isPure() const override { return false; } // SPECIAL: has 'visual' ordering
virtual bool isOutputter() const override { return true; } // SPECIAL: makes output
virtual bool cleanOut() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* filep() const { return op2p(); }
void filep(AstNode* nodep) { setOp2p(nodep); }
AstNode* offset() const { return op3p(); }
@@ -4493,7 +4606,7 @@ public:
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* lhsp() const { return op1p(); }
};
@@ -4514,7 +4627,7 @@ public:
virtual bool isOutputter() const override { return true; }
virtual bool isUnlikely() const override { return true; }
virtual bool cleanOut() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* lhsp() const { return op1p(); }
};
@@ -4535,7 +4648,7 @@ public:
virtual bool isPredictOptimizable() const override { return false; }
virtual bool isPure() const override { return !outp(); }
virtual bool cleanOut() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* searchp() const { return op1p(); } // op1 = Search expression
void searchp(AstNode* nodep) { setOp1p(nodep); }
AstNode* outp() const { return op2p(); } // op2 = Expressions to output
@@ -4545,26 +4658,21 @@ public:
class AstTestPlusArgs final : public AstNodeMath {
// Parents: expr
// Child: variable to set. If nullptr then this is a $test$plusargs instead of $value$plusargs
private:
string m_text;
public:
AstTestPlusArgs(FileLine* fl, const string& text)
: ASTGEN_SUPER_TestPlusArgs(fl)
, m_text{text} {}
AstTestPlusArgs(FileLine* fl, AstNode* searchp)
: ASTGEN_SUPER_TestPlusArgs(fl) {
setOp1p(searchp);
}
ASTNODE_NODE_FUNCS(TestPlusArgs)
virtual string name() const override { return m_text; }
virtual string verilogKwd() const override { return "$test$plusargs"; }
virtual string emitVerilog() override { return verilogKwd(); }
virtual string emitC() override { return "VL_VALUEPLUSARGS_%nq(%lw, %P, nullptr)"; }
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual bool cleanOut() const override { return true; }
virtual bool same(const AstNode* samep) const override {
return text() == static_cast<const AstTestPlusArgs*>(samep)->text();
}
string text() const { return m_text; } // * = Text to display
void text(const string& text) { m_text = text; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstNode* searchp() const { return op1p(); } // op1 = Search expression
void searchp(AstNode* nodep) { setOp1p(nodep); }
};
class AstGenFor final : public AstNodeFor {
@@ -4586,7 +4694,8 @@ public:
AstNode* bodysp() const { return op4p(); } // op4 = body of loop
virtual bool isGateOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_BRANCH; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstRepeat final : public AstNodeStmt {
@@ -4603,7 +4712,8 @@ public:
return false;
} // Not relevant - converted to FOR
virtual int instrCount() const override { return INSTR_COUNT_BRANCH; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstWait final : public AstNodeStmt {
@@ -4615,6 +4725,7 @@ public:
}
ASTNODE_NODE_FUNCS(Wait)
AstNode* bodysp() const { return op3p(); } // op3 = body of loop
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstWhile final : public AstNodeStmt {
@@ -4636,11 +4747,12 @@ public:
void addIncsp(AstNode* newp) { addOp4p(newp); }
virtual bool isGateOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_BRANCH; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
// Stop statement searchback here
virtual void addBeforeStmt(AstNode* newp, AstNode* belowp) override;
// Stop statement searchback here
virtual void addNextStmt(AstNode* newp, AstNode* belowp) override;
bool isFirstInMyListOfStatements(AstNode* n) const override { return n == bodysp(); }
};
class AstBreak final : public AstNodeStmt {
@@ -4756,7 +4868,7 @@ public:
ASTNODE_NODE_FUNCS(JumpBlock)
virtual int instrCount() const override { return 0; }
virtual bool maybePointedTo() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
// op1 = Statements
AstNode* stmtsp() const { return op1p(); } // op1 = List of statements
void addStmtsp(AstNode* nodep) { addNOp1p(nodep); }
@@ -4842,7 +4954,7 @@ public:
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual int instrCount() const override { return widthInstrs() * 2; } // xor, or/logor
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstConsAssoc final : public AstNodeMath {
@@ -4861,7 +4973,7 @@ public:
virtual bool cleanOut() const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
AstNode* defaultp() const { return op1p(); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstSetAssoc final : public AstNodeMath {
// Set an assoc array element and return object, '{}
@@ -4883,7 +4995,48 @@ public:
AstNode* lhsp() const { return op1p(); }
AstNode* keyp() const { return op2p(); }
AstNode* valuep() const { return op3p(); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstConsWildcard final : public AstNodeMath {
// Construct a wildcard assoc array and return object, '{}
// Parents: math
// Children: expression (elements or other queues)
public:
AstConsWildcard(FileLine* fl, AstNode* defaultp)
: ASTGEN_SUPER_ConsWildcard(fl) {
setNOp1p(defaultp);
}
ASTNODE_NODE_FUNCS(ConsWildcard)
virtual string emitVerilog() override { return "'{}"; }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual string emitSimpleOperator() override { V3ERROR_NA_RETURN(""); }
virtual bool cleanOut() const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
AstNode* defaultp() const { return op1p(); }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstSetWildcard final : public AstNodeMath {
// Set a wildcard assoc array element and return object, '{}
// Parents: math
// Children: expression (elements or other queues)
public:
AstSetWildcard(FileLine* fl, AstNode* lhsp, AstNode* keyp, AstNode* valuep)
: ASTGEN_SUPER_SetWildcard(fl) {
setOp1p(lhsp);
setNOp2p(keyp);
setOp3p(valuep);
}
ASTNODE_NODE_FUNCS(SetWildcard)
virtual string emitVerilog() override { return "'{}"; }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual string emitSimpleOperator() override { V3ERROR_NA_RETURN(""); }
virtual bool cleanOut() const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
AstNode* lhsp() const { return op1p(); }
AstNode* keyp() const { return op2p(); }
AstNode* valuep() const { return op3p(); }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstConsDynArray final : public AstNodeMath {
@@ -4904,7 +5057,7 @@ public:
virtual int instrCount() const override { return widthInstrs(); }
AstNode* lhsp() const { return op1p(); } // op1 = expression
AstNode* rhsp() const { return op2p(); } // op2 = expression
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstConsQueue final : public AstNodeMath {
@@ -4925,7 +5078,7 @@ public:
virtual int instrCount() const override { return widthInstrs(); }
AstNode* lhsp() const { return op1p(); } // op1 = expression
AstNode* rhsp() const { return op2p(); } // op2 = expression
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstBegin final : public AstNodeBlock {
@@ -5102,7 +5255,7 @@ public:
: ASTGEN_SUPER_New(fl, false, "new", pinsp) {}
ASTNODE_NODE_FUNCS(New)
virtual bool cleanOut() const { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual bool hasDType() const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
};
@@ -5121,7 +5274,7 @@ public:
virtual string emitVerilog() override { return "new"; }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual bool cleanOut() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
AstNode* rhsp() const { return op1p(); }
};
@@ -5141,7 +5294,7 @@ public:
virtual string emitVerilog() override { return "new"; }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual bool cleanOut() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual int instrCount() const override { return widthInstrs(); }
AstNode* sizep() const { return op1p(); }
AstNode* rhsp() const { return op2p(); }
@@ -5243,15 +5396,15 @@ public:
}
};
class AstTimingControl final : public AstNodeStmt {
class AstEventControl final : public AstNodeStmt {
// Parents: stmtlist
public:
AstTimingControl(FileLine* fl, AstSenTree* sensesp, AstNode* stmtsp)
: ASTGEN_SUPER_TimingControl(fl) {
AstEventControl(FileLine* fl, AstSenTree* sensesp, AstNode* stmtsp)
: ASTGEN_SUPER_EventControl(fl) {
setNOp1p(sensesp);
setNOp2p(stmtsp);
}
ASTNODE_NODE_FUNCS(TimingControl)
ASTNODE_NODE_FUNCS(EventControl)
virtual string verilogKwd() const override { return "@(%l) %r"; }
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
@@ -5574,7 +5727,7 @@ public:
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_PLI; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
bool combinable(const AstRand* samep) const {
return !seedp() && !samep->seedp() && reset() == samep->reset()
&& urandom() == samep->urandom();
@@ -5625,7 +5778,7 @@ public:
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_TIME; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual void dump(std::ostream& str = std::cout) const override;
void timeunit(const VTimescale& flag) { m_timeunit = flag; }
VTimescale timeunit() const { return m_timeunit; }
@@ -5646,7 +5799,7 @@ public:
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_TIME; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
virtual void dump(std::ostream& str = std::cout) const override;
void timeunit(const VTimescale& flag) { m_timeunit = flag; }
VTimescale timeunit() const { return m_timeunit; }
@@ -5671,7 +5824,7 @@ public:
virtual bool isSubstOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual int instrCount() const override { return INSTR_COUNT_PLI; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
//======================================================================
@@ -5992,7 +6145,9 @@ public:
class AstCLog2 final : public AstNodeUniop {
public:
AstCLog2(FileLine* fl, AstNode* lhsp)
: ASTGEN_SUPER_CLog2(fl, lhsp) {}
: ASTGEN_SUPER_CLog2(fl, lhsp) {
dtypeSetSigned32();
}
ASTNODE_NODE_FUNCS(CLog2)
virtual void numberOperate(V3Number& out, const V3Number& lhs) override { out.opCLog2(lhs); }
virtual string emitVerilog() override { return "%f$clog2(%l)"; }
@@ -6257,7 +6412,7 @@ public:
virtual bool cleanOut() const override { return true; }
virtual bool cleanLhs() const override { return true; }
virtual bool sizeMattersLhs() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstFEof final : public AstNodeUniop {
@@ -6299,7 +6454,7 @@ public:
AstNode* filep() const { return op1p(); }
void strp(AstNode* nodep) { setOp2p(nodep); }
AstNode* strp() const { return op2p(); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstFGetC final : public AstNodeUniop {
@@ -8424,7 +8579,7 @@ public:
AstNode* exprp() const { return op1p(); } // op1 = expression
AstSenTree* sentreep() const { return VN_AS(op2p(), SenTree); } // op2 = clock domain
void sentreep(AstSenTree* sentreep) { addOp2p(sentreep); } // op2 = clock domain
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstPast final : public AstNodeMath {
@@ -8447,7 +8602,7 @@ public:
AstNode* ticksp() const { return op2p(); } // op2 = ticks or nullptr means 1
AstSenTree* sentreep() const { return VN_AS(op4p(), SenTree); } // op4 = clock domain
void sentreep(AstSenTree* sentreep) { addOp4p(sentreep); } // op4 = clock domain
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstRose final : public AstNodeMath {
@@ -8468,7 +8623,7 @@ public:
AstNode* exprp() const { return op1p(); } // op1 = expression
AstSenTree* sentreep() const { return VN_AS(op2p(), SenTree); } // op2 = clock domain
void sentreep(AstSenTree* sentreep) { addOp2p(sentreep); } // op2 = clock domain
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstSampled final : public AstNodeMath {
@@ -8487,7 +8642,7 @@ public:
virtual bool cleanOut() const override { V3ERROR_NA_RETURN(""); }
virtual int instrCount() const override { return 0; }
AstNode* exprp() const { return op1p(); } // op1 = expression
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstStable final : public AstNodeMath {
@@ -8508,7 +8663,7 @@ public:
AstNode* exprp() const { return op1p(); } // op1 = expression
AstSenTree* sentreep() const { return VN_AS(op2p(), SenTree); } // op2 = clock domain
void sentreep(AstSenTree* sentreep) { addOp2p(sentreep); } // op2 = clock domain
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
class AstPattern final : public AstNodeMath {
@@ -8532,6 +8687,7 @@ public:
AstNodeDType* childDTypep() const { return VN_AS(op1p(), NodeDType); }
void childDTypep(AstNodeDType* nodep) { setOp1p(nodep); }
AstNode* itemsp() const { return op2p(); } // op2 = AstPatReplicate, AstPatMember, etc
void addItemsp(AstNode* nodep) { addOp2p(nodep); }
};
class AstPatMember final : public AstNodeMath {
// Verilog '{a} or '{a{b}}
@@ -8582,7 +8738,7 @@ public:
void rhsp(AstNode* nodep) { return setOp2p(nodep); }
AstSenTree* sentreep() const { return VN_AS(op4p(), SenTree); } // op4 = clock domain
void sentreep(AstSenTree* sentreep) { addOp4p(sentreep); } // op4 = clock domain
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
//======================================================================
@@ -8803,7 +8959,7 @@ public:
virtual bool isPredictOptimizable() const override { return false; }
virtual bool isPure() const override { return false; }
virtual bool isOutputter() const override { return true; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
};
//======================================================================
@@ -8823,7 +8979,7 @@ public:
ASTNODE_BASE_FUNCS(NodeFile)
virtual void dump(std::ostream& str) const override;
virtual string name() const override { return m_name; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
void tblockp(AstTextBlock* tblockp) { setOp1p(tblockp); }
AstTextBlock* tblockp() { return VN_AS(op1p(), TextBlock); }
};
@@ -9091,7 +9247,7 @@ public:
}
ASTNODE_NODE_FUNCS(CReturn)
virtual int instrCount() const override { return widthInstrs(); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
//
AstNode* lhsp() const { return op1p(); }
};
@@ -9122,7 +9278,7 @@ public:
virtual bool cleanOut() const override { return m_cleanOut; }
virtual string emitVerilog() override { V3ERROR_NA_RETURN(""); }
virtual string emitC() override { V3ERROR_NA_RETURN(""); }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
void addBodysp(AstNode* nodep) { addNOp1p(nodep); }
AstNode* bodysp() const { return op1p(); } // op1 = expressions to print
bool pure() const { return m_pure; }
@@ -9139,7 +9295,7 @@ public:
ASTNODE_NODE_FUNCS(CReset)
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
AstVarRef* varrefp() const { return VN_AS(op1p(), VarRef); } // op1 = varref to reset
};
@@ -9157,7 +9313,7 @@ public:
ASTNODE_NODE_FUNCS(CStmt)
virtual bool isGateOptimizable() const override { return false; }
virtual bool isPredictOptimizable() const override { return false; }
virtual bool same(const AstNode* samep) const override { return true; }
virtual bool same(const AstNode* /*samep*/) const override { return true; }
void addBodysp(AstNode* nodep) { addNOp1p(nodep); }
AstNode* bodysp() const { return op1p(); } // op1 = expressions to print
};
@@ -9253,7 +9409,7 @@ class AstTypeTable final : public AstNode {
AstEmptyQueueDType* m_emptyQueuep = nullptr;
AstQueueDType* m_queueIndexp = nullptr;
AstVoidDType* m_voidp = nullptr;
AstBasicDType* m_basicps[VBasicDTypeKwd::_ENUM_MAX];
AstBasicDType* m_basicps[VBasicDTypeKwd::_ENUM_MAX]{};
//
using DetailedMap = std::map<VBasicTypeKey, AstBasicDType*>;
DetailedMap m_detailedMap;
+16 -18
View File
@@ -27,7 +27,8 @@
#include <utility>
#include <vector>
template <class T_Node, class T_Data, int T_UserN> class AstUserAllocatorBase VL_NOT_FINAL {
template <class T_Node, class T_Data, int T_UserN>
class AstUserAllocatorBase VL_NOT_FINAL {
static_assert(1 <= T_UserN && T_UserN <= 5, "Wrong user pointer number");
static_assert(std::is_base_of<AstNode, T_Node>::value, "T_Node must be an AstNode type");
@@ -35,17 +36,16 @@ private:
std::vector<T_Data*> m_allocated;
inline T_Data* getUserp(const T_Node* nodep) const {
// This simplifies statically as T_UserN is constant. In C++17, use 'if constexpr'.
if (T_UserN == 1) {
if VL_CONSTEXPR_CXX17 (T_UserN == 1) {
const VNUser user = nodep->user1u();
return user.to<T_Data*>();
} else if (T_UserN == 2) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 2) {
const VNUser user = nodep->user2u();
return user.to<T_Data*>();
} else if (T_UserN == 3) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 3) {
const VNUser user = nodep->user3u();
return user.to<T_Data*>();
} else if (T_UserN == 4) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 4) {
const VNUser user = nodep->user4u();
return user.to<T_Data*>();
} else {
@@ -55,14 +55,13 @@ private:
}
inline void setUserp(T_Node* nodep, T_Data* userp) const {
// This simplifies statically as T_UserN is constant. In C++17, use 'if constexpr'.
if (T_UserN == 1) {
if VL_CONSTEXPR_CXX17 (T_UserN == 1) {
nodep->user1u(VNUser(userp));
} else if (T_UserN == 2) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 2) {
nodep->user2u(VNUser(userp));
} else if (T_UserN == 3) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 3) {
nodep->user3u(VNUser(userp));
} else if (T_UserN == 4) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 4) {
nodep->user4u(VNUser(userp));
} else {
nodep->user5u(VNUser(userp));
@@ -71,14 +70,13 @@ private:
protected:
AstUserAllocatorBase() {
// This simplifies statically as T_UserN is constant. In C++17, use 'if constexpr'.
if (T_UserN == 1) {
if VL_CONSTEXPR_CXX17 (T_UserN == 1) {
VNUser1InUse::check();
} else if (T_UserN == 2) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 2) {
VNUser2InUse::check();
} else if (T_UserN == 3) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 3) {
VNUser3InUse::check();
} else if (T_UserN == 4) {
} else if VL_CONSTEXPR_CXX17 (T_UserN == 4) {
VNUser4InUse::check();
} else {
VNUser5InUse::check();
@@ -94,7 +92,7 @@ protected:
public:
// Get a reference to the user data. If does not exist, construct it with given arguments.
template <typename... Args> //
template <typename... Args>
T_Data& operator()(T_Node* nodep, Args&&... args) {
T_Data* userp = getUserp(nodep);
if (!userp) {
@@ -106,7 +104,7 @@ public:
}
// Get a reference to the user data
T_Data& operator()(const T_Node* nodep) {
T_Data& operator()(const T_Node* nodep) const {
T_Data* const userp = getUserp(nodep);
UASSERT_OBJ(userp, nodep, "Missing User data on const AstNode");
return *userp;
+30 -24
View File
@@ -28,9 +28,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Begin.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <algorithm>
@@ -76,6 +77,33 @@ private:
return a + "__DOT__" + b;
}
void dotNames(const AstNodeBlock* const nodep, const char* const blockName) {
UINFO(8, "nname " << m_namedScope << endl);
if (nodep->name() != "") { // Else unneeded unnamed block
// Create data for dotted variable resolution
string dottedname = nodep->name() + "__DOT__"; // So always found
string::size_type pos;
while ((pos = dottedname.find("__DOT__")) != string::npos) {
const string ident = dottedname.substr(0, pos);
dottedname = dottedname.substr(pos + strlen("__DOT__"));
if (nodep->name() != "") {
m_displayScope = dot(m_displayScope, ident);
m_namedScope = dot(m_namedScope, ident);
}
m_unnamedScope = dot(m_unnamedScope, ident);
// Create CellInline for dotted var resolution
if (!m_ftaskp) {
AstCellInline* const inlinep = new AstCellInline(
nodep->fileline(), m_unnamedScope, blockName, m_modp->timeunit());
m_modp->addInlinesp(inlinep); // Must be parsed before any AstCells
}
}
}
// Remap var names and replace lower Begins
iterateAndNextNull(nodep->stmtsp());
}
void liftNode(AstNode* nodep) {
nodep->unlinkFrBack();
if (m_ftaskp) {
@@ -141,30 +169,8 @@ private:
VL_RESTORER(m_namedScope);
VL_RESTORER(m_unnamedScope);
{
UINFO(8, "nname " << m_namedScope << endl);
if (nodep->name() != "") { // Else unneeded unnamed block
// Create data for dotted variable resolution
string dottedname = nodep->name() + "__DOT__"; // So always found
string::size_type pos;
while ((pos = dottedname.find("__DOT__")) != string::npos) {
const string ident = dottedname.substr(0, pos);
dottedname = dottedname.substr(pos + strlen("__DOT__"));
if (nodep->name() != "") {
m_displayScope = dot(m_displayScope, ident);
m_namedScope = dot(m_namedScope, ident);
}
m_unnamedScope = dot(m_unnamedScope, ident);
// Create CellInline for dotted var resolution
if (!m_ftaskp) {
AstCellInline* const inlinep = new AstCellInline(
nodep->fileline(), m_unnamedScope, "__BEGIN__", m_modp->timeunit());
m_modp->addInlinesp(inlinep); // Must be parsed before any AstCells
}
}
}
dotNames(nodep, "__BEGIN__");
// Remap var names and replace lower Begins
iterateAndNextNull(nodep->stmtsp());
UASSERT_OBJ(!nodep->genforp(), nodep, "GENFORs should have been expanded earlier");
// Cleanup
+2 -1
View File
@@ -26,9 +26,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Branch.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <map>
+3 -2
View File
@@ -25,9 +25,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Broken.h"
#include "V3Ast.h"
#include "V3Global.h"
// This visitor does not edit nodes, and is called at error-exit, so should use constant iterators
#include "V3AstConstOnly.h"
@@ -48,7 +49,7 @@ static class BrokenCntGlobal {
uint8_t m_count = MIN_VALUE;
public:
uint8_t get() {
uint8_t get() const {
UASSERT(MIN_VALUE <= m_count && m_count <= MAX_VALUE, "Invalid generation number");
return m_count;
}
+3 -2
View File
@@ -27,10 +27,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3EmitCBase.h"
#include "V3CCtors.h"
#include "V3EmitCBase.h"
#include "V3Global.h"
#include <algorithm>
#include <list>
+2 -1
View File
@@ -25,9 +25,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3CUse.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <set>
+3 -2
View File
@@ -37,9 +37,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Case.h"
#include "V3Ast.h"
#include "V3Global.h"
#include "V3Stats.h"
#include <algorithm>
@@ -496,7 +497,7 @@ private:
V3Case::caseLint(nodep);
iterateChildren(nodep);
if (debug() >= 9) nodep->dumpTree(cout, " case_old: ");
if (isCaseTreeFast(nodep) && v3Global.opt.oCase()) {
if (isCaseTreeFast(nodep) && v3Global.opt.fCase()) {
// It's a simple priority encoder or complete statement
// we can make a tree of statements to avoid extra comparisons
++m_statCaseFast;
+2 -1
View File
@@ -40,9 +40,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Cast.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <algorithm>
+3 -2
View File
@@ -24,13 +24,14 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Cdc.h"
#include "V3Ast.h"
#include "V3Graph.h"
#include "V3Const.h"
#include "V3EmitV.h"
#include "V3File.h"
#include "V3Global.h"
#include "V3Graph.h"
#include <algorithm>
#include <deque>
+3 -2
View File
@@ -29,10 +29,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Ast.h"
#include "V3Changed.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <algorithm>
//######################################################################
+13 -7
View File
@@ -23,9 +23,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Class.h"
#include "V3Ast.h"
#include "V3Global.h"
//######################################################################
@@ -97,6 +98,7 @@ private:
m_prefix = nodep->name() + "__02e"; // .
iterateChildren(nodep);
}
nodep->repairCache();
}
virtual void visit(AstNodeModule* nodep) override {
// Visit for NodeModules that are not AstClass (AstClass is-a AstNodeModule)
@@ -114,13 +116,13 @@ private:
// Move later, or we wouldn't keep interating the class
// We're really moving the VarScope but we might not
// have a pointer to it yet
m_toScopeMoves.push_back(std::make_pair(nodep, m_packageScopep));
m_toScopeMoves.emplace_back(std::make_pair(nodep, m_packageScopep));
}
if (!m_ftaskp && nodep->lifetime().isStatic()) {
m_toPackageMoves.push_back(std::make_pair(nodep, m_classPackagep));
m_toPackageMoves.emplace_back(std::make_pair(nodep, m_classPackagep));
// We're really moving the VarScope but we might not
// have a pointer to it yet
m_toScopeMoves.push_back(std::make_pair(nodep, m_packageScopep));
m_toScopeMoves.emplace_back(std::make_pair(nodep, m_packageScopep));
}
}
}
@@ -136,7 +138,7 @@ private:
m_ftaskp = nodep;
iterateChildren(nodep);
if (m_packageScopep && nodep->lifetime().isStatic()) {
m_toScopeMoves.push_back(std::make_pair(nodep, m_packageScopep));
m_toScopeMoves.emplace_back(std::make_pair(nodep, m_packageScopep));
}
}
}
@@ -151,12 +153,16 @@ private:
virtual void visit(AstInitial* nodep) override {
// But not AstInitialAutomatic, which remains under the class
iterateChildren(nodep);
if (m_packageScopep) { m_toScopeMoves.push_back(std::make_pair(nodep, m_packageScopep)); }
if (m_packageScopep) {
m_toScopeMoves.emplace_back(std::make_pair(nodep, m_packageScopep));
}
}
virtual void visit(AstInitialStatic* nodep) override {
// But not AstInitialAutomatic, which remains under the class
iterateChildren(nodep);
if (m_packageScopep) { m_toScopeMoves.push_back(std::make_pair(nodep, m_packageScopep)); }
if (m_packageScopep) {
m_toScopeMoves.emplace_back(std::make_pair(nodep, m_packageScopep));
}
}
virtual void visit(AstNodeMath* nodep) override {} // Short circuit
+3 -1
View File
@@ -26,9 +26,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Clean.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <algorithm>
@@ -88,6 +89,7 @@ private:
if (VN_IS(nodep, Var)
|| VN_IS(nodep, NodeDType) // Don't want to change variable widths!
|| VN_IS(nodep->dtypep()->skipRefp(), AssocArrayDType) // Or arrays
|| VN_IS(nodep->dtypep()->skipRefp(), WildcardArrayDType)
|| VN_IS(nodep->dtypep()->skipRefp(), DynArrayDType)
|| VN_IS(nodep->dtypep()->skipRefp(), ClassRefDType)
|| VN_IS(nodep->dtypep()->skipRefp(), QueueDType)
+2 -1
View File
@@ -30,9 +30,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Clock.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <algorithm>
+7 -4
View File
@@ -22,12 +22,13 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Combine.h"
#include "V3DupFinder.h"
#include "V3Stats.h"
#include "V3Ast.h"
#include "V3AstUserAllocator.h"
#include "V3DupFinder.h"
#include "V3Global.h"
#include "V3Stats.h"
#include <list>
#include <vector>
@@ -220,7 +221,9 @@ class CombineVisitor final : VNVisitor {
// CONSTRUCTORS
explicit CombineVisitor(AstNetlist* nodep) { iterate(nodep); }
~CombineVisitor() { V3Stats::addStat("Optimizations, Combined CFuncs", m_cfuncsCombined); }
~CombineVisitor() override {
V3Stats::addStat("Optimizations, Combined CFuncs", m_cfuncsCombined);
}
public:
static void apply(AstNetlist* netlistp) { CombineVisitor{netlistp}; }
+10 -9
View File
@@ -24,9 +24,10 @@
#include "verilatedos.h"
#include "V3Common.h"
#include "V3Ast.h"
#include "V3Global.h"
#include "V3EmitCBase.h"
#include "V3Global.h"
//######################################################################
// Common component builders
@@ -50,7 +51,7 @@ static void makeToString(AstClass* nodep) {
funcp->isStatic(false);
funcp->protect(false);
AstNode* const exprp
= new AstCMath{nodep->fileline(), R"(std::string("'{") + to_string_middle() + "}")", 0};
= new AstCMath{nodep->fileline(), R"(std::string{"'{"} + to_string_middle() + "}")", 0};
exprp->dtypeSetString();
funcp->addStmtsp(new AstCReturn{nodep->fileline(), exprp});
nodep->addStmtp(funcp);
@@ -85,11 +86,11 @@ static void makeToStringMiddle(AstClass* nodep) {
}
}
}
if (nodep->extendsp() && nodep->extendsp()->classp()->user1()) {
string stmt = "out += \"";
if (nodep->extendsp()) {
string stmt = "out += ";
if (!comma.empty()) stmt += "\", \"+ ";
// comma = ", "; // Nothing further so not needed
stmt += nodep->extendsp()->dtypep()->nameProtect();
stmt += EmitCBaseVisitor::prefixNameProtect(nodep->extendsp()->dtypep());
stmt += "::to_string_middle();\n";
nodep->user1(true); // So what we extend dumps this
funcp->addStmtsp(new AstCStmt{nodep->fileline(), stmt});
@@ -103,13 +104,13 @@ static void makeToStringMiddle(AstClass* nodep) {
void V3Common::commonAll() {
UINFO(2, __FUNCTION__ << ": " << endl);
// NODE STATE
// Entire netlist:
// AstClass::user1() -> bool. True if class needs to_string dumper
const VNUser1InUse m_inuser1;
// Create common contents for each module
for (AstNode* nodep = v3Global.rootp()->modulesp(); nodep; nodep = nodep->nextp()) {
if (AstClass* const classp = VN_CAST(nodep, Class)) {
// NODE STATE
// Entire netlist:
// AstClass::user1() -> bool. True if class needs to_string dumper
const VNUser1InUse m_inuser1;
// Create ToString methods
makeVlToString(classp);
makeToString(classp);
+120 -19
View File
@@ -17,9 +17,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Config.h"
#include "V3Global.h"
#include "V3String.h"
#include "V3Config.h"
#include <map>
#include <set>
@@ -33,7 +34,8 @@
// as wildcards and are accessed by a resolved name. It rebuilds a name lookup
// cache of resolved entities. Entities stored in this container need an update
// function that takes a reference of this type to join multiple entities into one.
template <typename T> class V3ConfigWildcardResolver {
template <typename T>
class V3ConfigWildcardResolver final {
using Map = std::map<const std::string, T>;
Map m_mapWildcard; // Wildcard strings to entities
@@ -44,13 +46,8 @@ public:
/// Update into maps from other
void update(const V3ConfigWildcardResolver& other) {
typename Map::const_iterator it;
for (it = other.m_mapResolved.begin(); it != other.m_mapResolved.end(); ++it) {
m_mapResolved[it->first].update(it->second);
}
for (it = other.m_mapWildcard.begin(); it != other.m_mapWildcard.end(); ++it) {
m_mapWildcard[it->first].update(it->second);
}
for (const auto& itr : other.m_mapResolved) m_mapResolved[itr.first].update(itr.second);
for (const auto& itr : other.m_mapWildcard) m_mapWildcard[itr.first].update(itr.second);
}
// Access and create a (wildcard) entity
@@ -68,12 +65,12 @@ public:
// Cannot be resolved, create if matched
// Update this entity with all matches in the wildcards
for (it = m_mapWildcard.begin(); it != m_mapWildcard.end(); ++it) {
if (VString::wildmatch(name, it->first)) {
for (const auto& wildent : m_mapWildcard) {
if (VString::wildmatch(name, wildent.first)) {
if (!newp) {
newp = &m_mapResolved[name]; // Emplace and get pointer
}
newp->update(it->second);
newp->update(wildent.second);
}
}
return newp;
@@ -198,8 +195,8 @@ public:
AstNode* const nodep = new AstPragma(modp->fileline(), type);
modp->addStmtp(nodep);
}
for (auto it = m_modPragmas.cbegin(); it != m_modPragmas.cend(); ++it) {
AstNode* const nodep = new AstPragma(modp->fileline(), *it);
for (const auto& itr : m_modPragmas) {
AstNode* const nodep = new AstPragma{modp->fileline(), itr};
modp->addStmtp(nodep);
}
}
@@ -344,25 +341,124 @@ public:
using V3ConfigFileResolver = V3ConfigWildcardResolver<V3ConfigFile>;
//######################################################################
// ScopeTrace tracking
class V3ConfigScopeTraceEntry final {
public:
const string m_scope; // Scope or regexp to match
const bool m_on = false; // True to enable message
int m_levels = 0; // # levels, 0 = all, 1 = only this, ...
// CONSTRUCTORS
V3ConfigScopeTraceEntry(const string& scope, bool on, int levels)
: m_scope{scope}
, m_on{on}
, m_levels{levels} {}
~V3ConfigScopeTraceEntry() = default;
bool operator<(const V3ConfigScopeTraceEntry& other) const {
if (m_on < other.m_on) return true;
if (m_on > other.m_on) return false;
if (m_levels < other.m_levels) return true;
if (m_levels > other.m_levels) return false;
return m_scope < other.m_scope;
}
};
// Tracks what matches are known to hit against V3ConfigScopeTraceEntries
class V3ConfigScopeTraceEntryMatch final {
public:
const V3ConfigScopeTraceEntry* m_entryp;
const string m_scopepart;
V3ConfigScopeTraceEntryMatch(const V3ConfigScopeTraceEntry* entryp, const string& scopepart)
: m_entryp{entryp}
, m_scopepart{scopepart} {}
bool operator<(const V3ConfigScopeTraceEntryMatch& other) const {
if (m_entryp < other.m_entryp) return true;
if (m_entryp > other.m_entryp) return false;
return m_scopepart < other.m_scopepart;
}
};
class V3ConfigScopeTraceResolver final {
std::vector<V3ConfigScopeTraceEntry> m_entries; // User specified on/offs and levels
std::map<V3ConfigScopeTraceEntryMatch, bool> m_matchCache; // Matching entries for speed
public:
void addScopeTraceOn(bool on, const string& scope, int levels) {
UINFO(9, "addScopeTraceOn " << on << " '" << scope << "' "
<< " levels=" << levels << endl);
m_entries.emplace_back(V3ConfigScopeTraceEntry{scope, on, levels});
m_matchCache.clear();
}
bool getEntryMatch(const V3ConfigScopeTraceEntry* entp, const string& scopepart) {
// Return if a entry matches the scopepart, with memoization
const auto& key = V3ConfigScopeTraceEntryMatch{entp, scopepart};
const auto& it = m_matchCache.find(key);
if (it != m_matchCache.end()) return it->second; // Cached
const bool matched = VString::wildmatch(scopepart, entp->m_scope);
m_matchCache.emplace(key, matched);
return matched;
}
bool getScopeTraceOn(const string& scope) {
// Apply in the order the user provided them, so they can choose on/off preferencing
int maxLevel = 1;
for (const auto& ch : scope) {
if (ch == '.') ++maxLevel;
}
UINFO(9, "getScopeTraceOn " << scope << " maxLevel=" << maxLevel << endl);
bool enabled = true;
for (const auto& ent : m_entries) {
// We apply shortest match first for each rule component
// (Otherwise the levels would be useless as "--scope top* --levels 1" would
// always match at every scopepart, and we wound't know how to count levels)
int partLevel = 1;
for (string::size_type partEnd = 0; true;) {
partEnd = scope.find('.', partEnd + 1);
if (partEnd == string::npos) partEnd = scope.length();
const std::string scopepart = scope.substr(0, partEnd);
if (getEntryMatch(&ent, scopepart)) {
const bool levelMatch
= !ent.m_levels || (ent.m_levels >= maxLevel - partLevel);
if (levelMatch) enabled = ent.m_on;
UINFO(9, "getScopeTraceOn-part " << scope << " enabled=" << enabled
<< " @ lev=" << partLevel
<< (levelMatch ? "[match]" : "[miss]")
<< " from scopepart=" << scopepart << endl);
break;
}
if (partEnd == scope.length()) break;
++partLevel;
}
}
return enabled;
}
};
//######################################################################
// Resolve modules and files in the design
class V3ConfigResolver final {
V3ConfigModuleResolver m_modules; // Access to module names (with wildcards)
V3ConfigFileResolver m_files; // Access to file names (with wildcards)
V3ConfigScopeTraceResolver m_scopeTraces; // Regexp to trace enables
std::unordered_map<string, std::unordered_map<string, uint64_t>>
m_profileData; // Access to profile_data records
FileLine* m_profileFileLine = nullptr;
static V3ConfigResolver s_singleton; // Singleton (not via local static, as that's slow)
V3ConfigResolver() = default;
~V3ConfigResolver() = default;
public:
static V3ConfigResolver& s() { return s_singleton; }
static V3ConfigResolver& s() {
static V3ConfigResolver s_singleton;
return s_singleton;
}
V3ConfigModuleResolver& modules() { return m_modules; }
V3ConfigFileResolver& files() { return m_files; }
V3ConfigScopeTraceResolver& scopeTraces() { return m_scopeTraces; }
void addProfileData(FileLine* fl, const string& model, const string& key, uint64_t cost) {
if (!m_profileFileLine) m_profileFileLine = fl;
@@ -379,8 +475,6 @@ public:
FileLine* getProfileDataFileLine() const { return m_profileFileLine; } // Maybe null
};
V3ConfigResolver V3ConfigResolver::s_singleton;
//######################################################################
// V3Config
@@ -434,6 +528,10 @@ void V3Config::addProfileData(FileLine* fl, const string& model, const string& k
V3ConfigResolver::s().addProfileData(fl, model, key, cost);
}
void V3Config::addScopeTraceOn(bool on, const string& scope, int levels) {
V3ConfigResolver::s().scopeTraces().addScopeTraceOn(on, scope, levels);
}
void V3Config::addVarAttr(FileLine* fl, const string& module, const string& ftask,
const string& var, VAttrType attr, AstSenTree* sensep) {
// Semantics: sensep only if public_flat_rw
@@ -540,6 +638,9 @@ uint64_t V3Config::getProfileData(const string& model, const string& key) {
FileLine* V3Config::getProfileDataFileLine() {
return V3ConfigResolver::s().getProfileDataFileLine();
}
bool V3Config::getScopeTraceOn(const string& scope) {
return V3ConfigResolver::s().scopeTraces().getScopeTraceOn(scope);
}
bool V3Config::waive(FileLine* filelinep, V3ErrorCode code, const string& message) {
V3ConfigFile* filep = V3ConfigResolver::s().files().resolve(filelinep->filename());
+5 -3
View File
@@ -20,9 +20,9 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Ast.h"
#include "V3Error.h"
#include "V3FileLine.h"
#include "V3Ast.h"
//######################################################################
@@ -37,19 +37,21 @@ public:
static void addModulePragma(const string& module, VPragmaType pragma);
static void addProfileData(FileLine* fl, const string& model, const string& key,
uint64_t cost);
static void addWaiver(V3ErrorCode code, const string& filename, const string& message);
static void addScopeTraceOn(bool on, const string& scope, int levels);
static void addVarAttr(FileLine* fl, const string& module, const string& ftask,
const string& signal, VAttrType type, AstSenTree* nodep);
static void addWaiver(V3ErrorCode code, const string& filename, const string& message);
static void applyCase(AstCase* nodep);
static void applyCoverageBlock(AstNodeModule* modulep, AstBegin* nodep);
static void applyFTask(AstNodeModule* modulep, AstNodeFTask* ftaskp);
static void applyIgnores(FileLine* filelinep);
static void applyModule(AstNodeModule* modulep);
static void applyFTask(AstNodeModule* modulep, AstNodeFTask* ftaskp);
static void applyVarAttr(AstNodeModule* modulep, AstNodeFTask* ftaskp, AstVar* varp);
static uint64_t getProfileData(const string& model, const string& key);
static FileLine* getProfileDataFileLine();
static bool getScopeTraceOn(const string& scope);
static bool waive(FileLine* filelinep, V3ErrorCode code, const string& message);
};
+200 -89
View File
@@ -23,14 +23,15 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3String.h"
#include "V3Const.h"
#include "V3Ast.h"
#include "V3Width.h"
#include "V3Global.h"
#include "V3Simulate.h"
#include "V3Stats.h"
#include "V3String.h"
#include "V3UniqueNames.h"
#include "V3Width.h"
#include <algorithm>
#include <memory>
@@ -79,14 +80,50 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
// bool indicating if the term is clean (0/1 value, or if the top bits might be dirty)
using ResultTerm = std::tuple<AstNode*, unsigned, bool>;
struct LeafInfo final { // Leaf node (either AstConst or AstVarRef)
class LeafInfo final { // Leaf node (either AstConst or AstVarRef)
// MEMBERS
bool m_polarity = true;
int m_lsb = 0;
int m_lsb = 0; // LSB of actually used bit of m_refp->varp()
int m_msb = 0; // MSB of actually used bit of m_refp->varp()
int m_wordIdx = -1; // -1 means AstWordSel is not used.
AstVarRef* m_refp = nullptr;
const AstConst* m_constp = nullptr;
int width() const {
public:
// CONSTRUCTORS
LeafInfo() = default;
LeafInfo(const LeafInfo& other) = default;
explicit LeafInfo(int lsb)
: m_lsb{lsb} {}
// METHODS
void setLeaf(AstVarRef* refp) {
UASSERT(!m_refp && !m_constp, "Must be called just once");
m_refp = refp;
m_msb = refp->varp()->widthMin() - 1;
}
void setLeaf(const AstConst* constp) {
UASSERT(!m_refp && !m_constp, "Must be called just once");
m_constp = constp;
m_msb = constp->widthMin() - 1;
}
void updateBitRange(const AstCCast* castp) {
m_msb = std::min(m_msb, m_lsb + castp->width() - 1);
}
void updateBitRange(const AstShiftR* shiftp) {
m_lsb += VN_AS(shiftp->rhsp(), Const)->toUInt();
}
void wordIdx(int i) { m_wordIdx = i; }
void polarity(bool p) { m_polarity = p; }
AstVarRef* refp() const { return m_refp; }
const AstConst* constp() const { return m_constp; }
int wordIdx() const { return m_wordIdx; }
bool polarity() const { return m_polarity; }
int lsb() const { return m_lsb; }
int msb() const { return std::min(m_msb, varWidth() - 1); }
int varWidth() const {
UASSERT(m_refp, "m_refp should be set");
const int width = m_refp->varp()->widthMin();
if (!m_refp->isWide()) {
@@ -111,6 +148,15 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
BitPolarityEntry() = default;
};
struct FrozenNodeInfo final { // Context when a frozen node is found
bool m_polarity;
int m_lsb;
bool operator<(const FrozenNodeInfo& other) const {
if (m_lsb != other.m_lsb) return m_lsb < other.m_lsb;
return m_polarity < other.m_polarity;
}
};
class Restorer final { // Restore the original state unless disableRestore() is called
ConstBitOpTreeVisitor& m_visitor;
const size_t m_polaritiesSize;
@@ -280,7 +326,6 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
return ResultTerm{resultp, ops, clean};
}
public:
// CONSTRUCTORS
VarInfo(ConstBitOpTreeVisitor* parent, AstVarRef* refp, int width)
: m_parentp{parent}
@@ -299,7 +344,8 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
LeafInfo* m_leafp = nullptr; // AstConst or AstVarRef that currently looking for
const AstNode* const m_rootp; // Root of this AST subtree
std::vector<AstNode*> m_frozenNodes; // Nodes that cannot be optimized
std::vector<std::pair<AstNode*, FrozenNodeInfo>>
m_frozenNodes; // Nodes that cannot be optimized
std::vector<BitPolarityEntry> m_bitPolarities; // Polarity of bits found during iterate()
std::vector<std::unique_ptr<VarInfo>> m_varInfos; // VarInfo for each variable, [0] is nullptr
@@ -329,32 +375,32 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
UINFO(9, "Increment to " << m_ops << " " << nodep << " called from line " << line << "\n");
}
VarInfo& getVarInfo(const LeafInfo& ref) {
UASSERT_OBJ(ref.m_refp, m_rootp, "null varref in And/Or/Xor optimization");
AstNode* nodep = ref.m_refp->varScopep();
if (!nodep) nodep = ref.m_refp->varp(); // Not scoped
UASSERT_OBJ(ref.refp(), m_rootp, "null varref in And/Or/Xor optimization");
AstNode* nodep = ref.refp()->varScopep();
if (!nodep) nodep = ref.refp()->varp(); // Not scoped
int baseIdx = nodep->user4();
if (baseIdx == 0) { // Not set yet
baseIdx = m_varInfos.size();
const int numWords
= ref.m_refp->dtypep()->isWide() ? ref.m_refp->dtypep()->widthWords() : 1;
= ref.refp()->dtypep()->isWide() ? ref.refp()->dtypep()->widthWords() : 1;
m_varInfos.resize(m_varInfos.size() + numWords);
nodep->user4(baseIdx);
}
const size_t idx = baseIdx + std::max(0, ref.m_wordIdx);
const size_t idx = baseIdx + std::max(0, ref.wordIdx());
VarInfo* varInfop = m_varInfos[idx].get();
if (!varInfop) {
varInfop = new VarInfo{this, ref.m_refp, ref.width()};
varInfop = new VarInfo{this, ref.refp(), ref.varWidth()};
m_varInfos[idx].reset(varInfop);
} else {
if (!varInfop->sameVarAs(ref.m_refp))
CONST_BITOP_SET_FAILED("different var (scope?)", ref.m_refp);
if (!varInfop->sameVarAs(ref.refp()))
CONST_BITOP_SET_FAILED("different var (scope?)", ref.refp());
}
return *varInfop;
}
// Traverse down to see AstConst or AstVarRef
LeafInfo findLeaf(AstNode* nodep, bool expectConst) {
LeafInfo info;
LeafInfo info{m_lsb};
{
VL_RESTORER(m_leafp);
m_leafp = &info;
@@ -363,9 +409,9 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
bool ok = !m_failed;
if (expectConst) {
ok &= !info.m_refp && info.m_constp;
ok &= !info.refp() && info.constp();
} else {
ok &= info.m_refp && !info.m_constp;
ok &= info.refp() && !info.constp();
}
return ok ? info : LeafInfo{};
}
@@ -374,7 +420,10 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
virtual void visit(AstNode* nodep) override {
CONST_BITOP_SET_FAILED("Hit unexpected op", nodep);
}
virtual void visit(AstCCast* nodep) override { iterateChildren(nodep); }
virtual void visit(AstCCast* nodep) override {
iterateChildren(nodep);
if (m_leafp) m_leafp->updateBitRange(nodep);
}
virtual void visit(AstShiftR* nodep) override {
CONST_BITOP_RETURN_IF(!m_leafp, nodep);
AstConst* const constp = VN_CAST(nodep->rhsp(), Const);
@@ -382,12 +431,14 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
m_lsb += constp->toUInt();
incrOps(nodep, __LINE__);
iterate(nodep->lhsp());
m_leafp->updateBitRange(nodep);
m_lsb -= constp->toUInt();
}
virtual void visit(AstNot* nodep) override {
CONST_BITOP_RETURN_IF(nodep->widthMin() != 1, nodep);
AstNode* lhsp = nodep->lhsp();
if (AstCCast* const castp = VN_CAST(lhsp, CCast)) lhsp = castp->lhsp();
AstCCast* const castp = VN_CAST(lhsp, CCast);
if (castp) lhsp = castp->lhsp();
CONST_BITOP_RETURN_IF(!VN_IS(lhsp, VarRef) && !VN_IS(lhsp, Xor) && !VN_IS(lhsp, RedXor)
&& !VN_IS(lhsp, ShiftR),
lhsp);
@@ -396,68 +447,68 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
iterateChildren(nodep);
// Don't restore m_polarity for Xor as it counts parity of the entire tree
if (!isXorTree()) m_polarity = !m_polarity;
if (m_leafp && castp) m_leafp->updateBitRange(castp);
}
virtual void visit(AstWordSel* nodep) override {
CONST_BITOP_RETURN_IF(!m_leafp, nodep);
AstConst* const constp = VN_CAST(nodep->bitp(), Const);
CONST_BITOP_RETURN_IF(!constp, nodep->rhsp());
UASSERT_OBJ(m_leafp->m_wordIdx == -1, nodep, "Unexpected nested WordSel");
m_leafp->m_wordIdx = constp->toSInt();
UASSERT_OBJ(m_leafp->wordIdx() == -1, nodep, "Unexpected nested WordSel");
m_leafp->wordIdx(constp->toSInt());
iterate(nodep->fromp());
}
virtual void visit(AstVarRef* nodep) override {
CONST_BITOP_RETURN_IF(!m_leafp, nodep);
UASSERT_OBJ(!m_leafp->m_refp, nodep, m_leafp->m_refp << " is already set");
m_leafp->m_refp = nodep;
m_leafp->m_polarity = m_polarity;
m_leafp->m_lsb = m_lsb;
m_leafp->setLeaf(nodep);
m_leafp->polarity(m_polarity);
}
virtual void visit(AstConst* nodep) override {
CONST_BITOP_RETURN_IF(!m_leafp, nodep);
UASSERT_OBJ(!m_leafp->m_constp, nodep, m_leafp->m_constp << " is already set");
m_leafp->m_constp = nodep;
m_leafp->m_lsb = m_lsb;
m_leafp->setLeaf(nodep);
}
virtual void visit(AstRedXor* nodep) override {
Restorer restorer{*this};
CONST_BITOP_RETURN_IF(!VN_IS(m_rootp, Xor), nodep);
AstNode* lhsp = nodep->lhsp();
if (const AstCCast* const castp = VN_CAST(lhsp, CCast)) lhsp = castp->lhsp();
const AstCCast* const castp = VN_CAST(lhsp, CCast);
if (castp) lhsp = castp->lhsp();
if (const AstAnd* const andp = VN_CAST(lhsp, And)) { // '^(mask & leaf)'
CONST_BITOP_RETURN_IF(!andp, lhsp);
const LeafInfo& mask = findLeaf(andp->lhsp(), true);
CONST_BITOP_RETURN_IF(!mask.m_constp || mask.m_lsb != 0, andp->lhsp());
CONST_BITOP_RETURN_IF(!mask.constp() || mask.lsb() != 0, andp->lhsp());
const LeafInfo& ref = findLeaf(andp->rhsp(), false);
CONST_BITOP_RETURN_IF(!ref.m_refp, andp->rhsp());
LeafInfo ref = findLeaf(andp->rhsp(), false);
CONST_BITOP_RETURN_IF(!ref.refp(), andp->rhsp());
if (castp) ref.updateBitRange(castp);
restorer.disableRestore(); // Now all subtree succeeded
const V3Number& maskNum = mask.m_constp->num();
const V3Number& maskNum = mask.constp()->num();
incrOps(nodep, __LINE__);
incrOps(andp, __LINE__);
// Mark all bits checked in this reduction
const int maxBitIdx = std::min(ref.m_lsb + maskNum.width(), ref.width());
for (int bitIdx = ref.m_lsb; bitIdx < maxBitIdx; ++bitIdx) {
const int maskIdx = bitIdx - ref.m_lsb;
const int maxBitIdx = std::min(ref.lsb() + maskNum.width(), ref.msb() + 1);
for (int bitIdx = ref.lsb(); bitIdx < maxBitIdx; ++bitIdx) {
const int maskIdx = bitIdx - ref.lsb();
if (maskNum.bitIs0(maskIdx)) continue;
// Set true, m_polarity takes care of the entire parity
m_bitPolarities.emplace_back(ref, true, bitIdx);
}
} else { // '^leaf'
const LeafInfo& ref = findLeaf(lhsp, false);
CONST_BITOP_RETURN_IF(!ref.m_refp, lhsp);
LeafInfo ref = findLeaf(lhsp, false);
CONST_BITOP_RETURN_IF(!ref.refp(), lhsp);
if (castp) ref.updateBitRange(castp);
restorer.disableRestore(); // Now all checks passed
incrOps(nodep, __LINE__);
// Mark all bits checked by this comparison
for (int bitIdx = ref.m_lsb; bitIdx < ref.width(); ++bitIdx) {
for (int bitIdx = ref.lsb(); bitIdx <= ref.msb(); ++bitIdx) {
m_bitPolarities.emplace_back(ref, true, bitIdx);
}
}
@@ -477,29 +528,29 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
for (const bool right : {false, true}) {
Restorer restorer{*this};
LeafInfo leafInfo;
LeafInfo leafInfo{m_lsb};
m_leafp = &leafInfo;
AstNode* opp = right ? nodep->rhsp() : nodep->lhsp();
const bool origFailed = m_failed;
iterate(opp);
if (leafInfo.m_constp || m_failed) {
if (leafInfo.constp() || m_failed) {
// Revert changes in leaf
restorer.restoreNow();
// Reach past a cast then add to frozen nodes to be added to final reduction
if (const AstCCast* const castp = VN_CAST(opp, CCast)) opp = castp->lhsp();
m_frozenNodes.push_back(opp);
m_frozenNodes.emplace_back(opp, FrozenNodeInfo{m_polarity, m_lsb});
m_failed = origFailed;
continue;
}
restorer.disableRestore(); // Now all checks passed
if (leafInfo.m_refp) {
if (leafInfo.refp()) {
// The conditional on the lsb being in range is necessary for some degenerate
// case, e.g.: (IData)((QData)wide[0] >> 32), or <1-bit-var> >> 1, which is
// just zero
if (leafInfo.m_lsb < leafInfo.width()) {
m_bitPolarities.emplace_back(leafInfo, isXorTree() || leafInfo.m_polarity,
leafInfo.m_lsb);
} else if (isAndTree() && leafInfo.m_polarity) {
if (leafInfo.lsb() <= leafInfo.msb()) {
m_bitPolarities.emplace_back(leafInfo, isXorTree() || leafInfo.polarity(),
leafInfo.lsb());
} else if (isAndTree() && leafInfo.polarity()) {
// If there is a constant 0 term in an And tree, we must include it. Fudge
// this by adding a bit with both polarities, which will simplify to zero
m_bitPolarities.emplace_back(leafInfo, true, 0);
@@ -515,46 +566,59 @@ class ConstBitOpTreeVisitor final : public VNVisitor {
const AstConst* const constp = VN_CAST(lhsp, Const);
CONST_BITOP_RETURN_IF(!constp, nodep->lhsp());
const bool maskFlip = isOrTree();
const V3Number& compNum = constp->num();
auto setPolarities = [this, &compNum](const LeafInfo& ref, const V3Number* maskp) {
const bool maskFlip = isOrTree();
int constantWidth = compNum.width();
if (maskp) constantWidth = std::max(constantWidth, maskp->width());
const int maxBitIdx = std::max(ref.lsb() + constantWidth, ref.msb() + 1);
// Mark all bits checked by this comparison
for (int bitIdx = ref.lsb(); bitIdx < maxBitIdx; ++bitIdx) {
const int maskIdx = bitIdx - ref.lsb();
const bool mask0 = maskp && maskp->bitIs0(maskIdx);
const bool outOfRange = bitIdx > ref.msb();
if (mask0 || outOfRange) { // RHS is 0
if (compNum.bitIs1(maskIdx)) {
// LHS is 1
// And tree: 1 == 0 => always false, set v && !v
// Or tree : 1 != 0 => always true, set v || !v
m_bitPolarities.emplace_back(ref, true, 0);
m_bitPolarities.emplace_back(ref, false, 0);
break;
} else { // This bitIdx is irrelevant
continue;
}
}
const bool polarity = compNum.bitIs1(maskIdx) != maskFlip;
m_bitPolarities.emplace_back(ref, polarity, bitIdx);
}
};
if (const AstAnd* const andp = VN_CAST(nodep->rhsp(), And)) { // comp == (mask & v)
const LeafInfo& mask = findLeaf(andp->lhsp(), true);
CONST_BITOP_RETURN_IF(!mask.m_constp || mask.m_lsb != 0, andp->lhsp());
CONST_BITOP_RETURN_IF(!mask.constp() || mask.lsb() != 0, andp->lhsp());
const LeafInfo& ref = findLeaf(andp->rhsp(), false);
CONST_BITOP_RETURN_IF(!ref.m_refp, andp->rhsp());
CONST_BITOP_RETURN_IF(!ref.refp(), andp->rhsp());
restorer.disableRestore(); // Now all checks passed
const V3Number& maskNum = mask.m_constp->num();
const V3Number& maskNum = mask.constp()->num();
incrOps(nodep, __LINE__);
incrOps(andp, __LINE__);
// Mark all bits checked by this comparison
const int maxBitIdx = std::min(ref.m_lsb + maskNum.width(), ref.width());
for (int bitIdx = ref.m_lsb; bitIdx < maxBitIdx; ++bitIdx) {
const int maskIdx = bitIdx - ref.m_lsb;
if (maskNum.bitIs0(maskIdx)) continue;
const bool polarity = compNum.bitIs1(maskIdx) != maskFlip;
m_bitPolarities.emplace_back(ref, polarity, bitIdx);
}
setPolarities(ref, &maskNum);
} else { // comp == v
const LeafInfo& ref = findLeaf(nodep->rhsp(), false);
CONST_BITOP_RETURN_IF(!ref.m_refp, nodep->rhsp());
CONST_BITOP_RETURN_IF(!ref.refp(), nodep->rhsp());
restorer.disableRestore(); // Now all checks passed
incrOps(nodep, __LINE__);
// Mark all bits checked by this comparison
const int maxBitIdx = std::min(ref.m_lsb + compNum.width(), ref.width());
for (int bitIdx = ref.m_lsb; bitIdx < maxBitIdx; ++bitIdx) {
const int maskIdx = bitIdx - ref.m_lsb;
const bool polarity = compNum.bitIs1(maskIdx) != maskFlip;
m_bitPolarities.emplace_back(ref, polarity, bitIdx);
}
setPolarities(ref, nullptr);
}
} else {
CONST_BITOP_SET_FAILED("Mixture of different ops cannot be optimized", nodep);
@@ -652,17 +716,21 @@ public:
}
}
std::map<FrozenNodeInfo, std::vector<AstNode*>> frozenNodes; // Group by FrozenNodeInfo
// Check if frozen terms are clean or not
for (AstNode* const termp : visitor.m_frozenNodes) {
for (const auto& frozenInfo : visitor.m_frozenNodes) {
AstNode* const termp = frozenInfo.first;
// Comparison operators are clean
if (VN_IS(termp, Eq) || VN_IS(termp, Neq) || VN_IS(termp, Lt) || VN_IS(termp, Lte)
|| VN_IS(termp, Gt) || VN_IS(termp, Gte)) {
if ((VN_IS(termp, Eq) || VN_IS(termp, Neq) || VN_IS(termp, Lt) || VN_IS(termp, Lte)
|| VN_IS(termp, Gt) || VN_IS(termp, Gte))
&& frozenInfo.second.m_lsb == 0) {
hasCleanTerm = true;
} else {
// Otherwise, conservatively assume the frozen term is dirty
hasDirtyTerm = true;
UINFO(9, "Dirty frozen term: " << termp << endl);
}
frozenNodes[frozenInfo.second].push_back(termp);
}
// Figure out if a final negation is required
@@ -672,7 +740,12 @@ public:
const bool needsCleaning = visitor.isAndTree() ? !hasCleanTerm : hasDirtyTerm;
// Add size of reduction tree to op count
resultOps += termps.size() + visitor.m_frozenNodes.size() - 1;
resultOps += termps.size() - 1;
for (const auto& lsbAndNodes : frozenNodes) {
if (lsbAndNodes.first.m_lsb > 0) ++resultOps; // Needs AstShiftR
if (!lsbAndNodes.first.m_polarity) ++resultOps; // Needs AstNot
resultOps += lsbAndNodes.second.size();
}
// Add final polarity flip in Xor tree
if (needsFlip) ++resultOps;
// Add final cleaning AND
@@ -681,7 +754,11 @@ public:
if (debug() >= 9) { // LCOV_EXCL_START
cout << "Bitop tree considered: " << endl;
for (AstNode* const termp : termps) termp->dumpTree("Reduced term: ");
for (AstNode* const termp : visitor.m_frozenNodes) termp->dumpTree("Frozen term: ");
for (const std::pair<AstNode*, FrozenNodeInfo>& termp : visitor.m_frozenNodes) {
termp.first->dumpTree("Frozen term with lsb " + std::to_string(termp.second.m_lsb)
+ " polarity " + std::to_string(termp.second.m_polarity)
+ ": ");
}
cout << "Needs flipping: " << needsFlip << endl;
cout << "Needs cleaning: " << needsCleaning << endl;
cout << "Size: " << resultOps << " input size: " << visitor.m_ops << endl;
@@ -724,8 +801,25 @@ public:
resultp = reduce(resultp, termp);
}
// Add any frozen terms to the reduction
for (AstNode* const frozenp : visitor.m_frozenNodes) {
resultp = reduce(resultp, frozenp->unlinkFrBack());
for (auto&& nodes : frozenNodes) {
// nodes.second has same lsb and polarity
AstNode* termp = nullptr;
for (AstNode* const itemp : nodes.second) {
termp = reduce(termp, itemp->unlinkFrBack());
}
if (nodes.first.m_lsb > 0) { // LSB is not 0, so shiftR
AstNodeDType* const dtypep = termp->dtypep();
termp = new AstShiftR{termp->fileline(), termp,
new AstConst(termp->fileline(), AstConst::WidthedValue{},
termp->width(), nodes.first.m_lsb)};
termp->dtypep(dtypep);
}
if (!nodes.first.m_polarity) { // Polarity is inverted, so append Not
AstNodeDType* const dtypep = termp->dtypep();
termp = new AstNot{termp->fileline(), termp};
termp->dtypep(dtypep);
}
resultp = reduce(resultp, termp);
}
// Set width of masks to expected result width. This is required to prevent later removal
@@ -769,6 +863,9 @@ public:
class ConstVisitor final : public VNVisitor {
private:
// CONSTANTS
static constexpr unsigned CONCAT_MERGABLE_MAX_DEPTH = 10; // Limit alg recursion
// NODE STATE
// ** only when m_warn/m_doExpensive is set. If state is needed other times,
// ** must track down everywhere V3Const is called and make sure no overlaps.
@@ -992,7 +1089,7 @@ private:
if (orLIsRedundant && orRIsRedundant) {
nodep->replaceWith(
new AstConst(nodep->fileline(), AstConst::DtypedValue(), nodep->dtypep(), 0));
new AstConst(nodep->fileline(), AstConst::DTyped{}, nodep->dtypep()));
VL_DO_DANGLING(nodep->deleteTree(), nodep);
return true;
} else if (orLIsRedundant) {
@@ -1048,7 +1145,7 @@ private:
bool matchBitOpTree(AstNode* nodep) {
if (nodep->widthMin() != 1) return false;
if (!v3Global.opt.oConstBitOpTree()) return false;
if (!v3Global.opt.fConstBitOpTree()) return false;
string debugPrefix;
if (debug() >= 9) { // LCOV_EXCL_START
@@ -1370,7 +1467,7 @@ private:
return (VN_IS(nodep, And) || VN_IS(nodep, Or) || VN_IS(nodep, Xor));
}
bool ifAdjacentSel(const AstSel* lhsp, const AstSel* rhsp) {
if (!v3Global.opt.oAssemble()) return false; // opt disabled
if (!v3Global.opt.fAssemble()) return false; // opt disabled
if (!lhsp || !rhsp) return false;
const AstNode* const lfromp = lhsp->fromp();
const AstNode* const rfromp = rhsp->fromp();
@@ -1385,7 +1482,7 @@ private:
}
bool ifMergeAdjacent(AstNode* lhsp, AstNode* rhsp) {
// called by concatmergeable to determine if {lhsp, rhsp} make sense
if (!v3Global.opt.oAssemble()) return false; // opt disabled
if (!v3Global.opt.fAssemble()) return false; // opt disabled
// two same varref
if (operandsSame(lhsp, rhsp)) return true;
const AstSel* lselp = VN_CAST(lhsp, Sel);
@@ -1420,11 +1517,12 @@ private:
if (rend == rfromp->width() && lstart->toSInt() == 0) return true;
return false;
}
bool concatMergeable(const AstNode* lhsp, const AstNode* rhsp) {
bool concatMergeable(const AstNode* lhsp, const AstNode* rhsp, unsigned depth) {
// determine if {a OP b, c OP d} => {a, c} OP {b, d} is advantageous
if (!v3Global.opt.oAssemble()) return false; // opt disabled
if (!v3Global.opt.fAssemble()) return false; // opt disabled
if (lhsp->type() != rhsp->type()) return false;
if (!ifConcatMergeableBiop(lhsp)) return false;
if (depth > CONCAT_MERGABLE_MAX_DEPTH) return false; // As worse case O(n^2) algorithm
const AstNodeBiop* const lp = VN_CAST(lhsp, NodeBiop);
const AstNodeBiop* const rp = VN_CAST(rhsp, NodeBiop);
@@ -1434,11 +1532,12 @@ private:
const bool rad = ifMergeAdjacent(lp->rhsp(), rp->rhsp());
if (lad && rad) return true;
// {a[] & b[]&c[], a[] & b[]&c[]}
if (lad && concatMergeable(lp->rhsp(), rp->rhsp())) return true;
if (lad && concatMergeable(lp->rhsp(), rp->rhsp(), depth + 1)) return true;
// {a[]&b[] & c[], a[]&b[] & c[]}
if (rad && concatMergeable(lp->lhsp(), rp->lhsp())) return true;
if (rad && concatMergeable(lp->lhsp(), rp->lhsp(), depth + 1)) return true;
// {(a[]&b[])&(c[]&d[]), (a[]&b[])&(c[]&d[])}
if (concatMergeable(lp->lhsp(), rp->lhsp()) && concatMergeable(lp->rhsp(), rp->rhsp())) {
if (concatMergeable(lp->lhsp(), rp->lhsp(), depth + 1)
&& concatMergeable(lp->rhsp(), rp->rhsp(), depth + 1)) {
return true;
}
return false;
@@ -1666,6 +1765,7 @@ private:
lp->rhsp(lrp);
nodep->lhsp(llp);
nodep->rhsp(rlp);
nodep->dtypep(llp->dtypep()); // dtype of Biop is before shift.
VL_DO_DANGLING(rp->deleteTree(), rp);
VL_DO_DANGLING(rrp->deleteTree(), rrp);
// nodep->dumpTree(cout, " repShiftSame_new: ");
@@ -1698,7 +1798,7 @@ private:
AstNode* const lrp = lp->rhsp()->cloneTree(false);
AstNode* const rlp = rp->lhsp()->cloneTree(false);
AstNode* const rrp = rp->rhsp()->cloneTree(false);
if (concatMergeable(lp, rp)) {
if (concatMergeable(lp, rp, 0)) {
AstConcat* const newlp = new AstConcat(rlp->fileline(), llp, rlp);
AstConcat* const newrp = new AstConcat(rrp->fileline(), lrp, rrp);
// use the lhs to replace the parent concat
@@ -2279,6 +2379,7 @@ private:
iterateChildren(nodep);
}
}
virtual void visit(AstClassOrPackageRef* nodep) override { iterateChildren(nodep); }
virtual void visit(AstPin* nodep) override { iterateChildren(nodep); }
void replaceLogEq(AstLogEq* nodep) {
@@ -2506,7 +2607,7 @@ private:
if (nodep->access().isReadOnly()
&& ((!m_params // Can reduce constant wires into equations
&& m_doNConst
&& v3Global.opt.oConst()
&& v3Global.opt.fConst()
// Default value, not a "known" constant for this usage
&& !nodep->varp()->isClassMember()
&& !(nodep->varp()->isFuncLocal() && nodep->varp()->isNonOutput())
@@ -3368,7 +3469,7 @@ private:
TREEOPV("AstConcat{$lhsp.isZero, $rhsp}", "replaceExtend(nodep, nodep->rhsp())");
// CONCAT(a[1],a[0]) -> a[1:0]
TREEOPV("AstConcat{$lhsp.castSel, $rhsp.castSel, ifAdjacentSel(VN_AS($lhsp,,Sel),,VN_AS($rhsp,,Sel))}", "replaceConcatSel(nodep)");
TREEOPV("AstConcat{ifConcatMergeableBiop($lhsp), concatMergeable($lhsp,,$rhsp)}", "replaceConcatMerge(nodep)");
TREEOPV("AstConcat{ifConcatMergeableBiop($lhsp), concatMergeable($lhsp,,$rhsp,,0)}", "replaceConcatMerge(nodep)");
// Common two-level operations that can be simplified
TREEOP ("AstAnd {$lhsp.castConst,matchAndCond(nodep)}", "DONE");
TREEOP ("AstAnd {$lhsp.castConst, $rhsp.castOr, matchMaskedOr(nodep)}", "DONE");
@@ -3484,7 +3585,11 @@ public:
}
virtual ~ConstVisitor() override {
if (m_doCpp) {
V3Stats::addStat("Optimizations, Const bit op reduction", m_statBitOpReduction);
if (m_globalPass) {
V3Stats::addStat("Optimizations, Const bit op reduction", m_statBitOpReduction);
} else {
V3Stats::addStatSum("Optimizations, Const bit op reduction", m_statBitOpReduction);
}
}
}
@@ -3577,6 +3682,12 @@ AstNode* V3Const::constifyEdit(AstNode* nodep) {
return nodep;
}
AstNode* V3Const::constifyEditCpp(AstNode* nodep) {
ConstVisitor visitor{ConstVisitor::PROC_CPP, /* globalPass: */ false};
nodep = visitor.mainAcceptEdit(nodep);
return nodep;
}
void V3Const::constifyAllLive(AstNetlist* nodep) {
// Only call from Verilator.cpp, as it uses user#'s
// This only pushes constants up, doesn't make any other edits
+3
View File
@@ -39,6 +39,9 @@ public:
static void constifyCpp(AstNetlist* nodep);
// Only the current node and lower
// Return new node that may have replaced nodep
static AstNode* constifyEditCpp(AstNode* nodep);
// Only the current node and lower
// Return new node that may have replaced nodep
static AstNode* constifyEdit(AstNode* nodep);
// Only the current node and lower, with special SenTree optimization
// Return new node that may have replaced nodep
+29 -27
View File
@@ -27,9 +27,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Coverage.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <map>
#include <unordered_map>
@@ -278,7 +279,7 @@ private:
// We'll do this, and make the if(...) coverinc later.
// Add signal to hold the old value
const string newvarname = string("__Vtogcov__") + nodep->shortName();
const string newvarname = std::string{"__Vtogcov__"} + nodep->shortName();
AstVar* const chgVarp
= new AstVar(nodep->fileline(), VVarType::MODULETEMP, newvarname, nodep);
chgVarp->fileline()->modifyWarnOff(V3ErrorCode::UNUSED, true);
@@ -288,9 +289,9 @@ private:
// This is necessarily an O(n^2) expansion, which is why
// we limit coverage to signals with < 256 bits.
ToggleEnt newvec(string(""),
new AstVarRef(nodep->fileline(), nodep, VAccess::READ),
new AstVarRef(nodep->fileline(), chgVarp, VAccess::WRITE));
ToggleEnt newvec{std::string{""},
new AstVarRef{nodep->fileline(), nodep, VAccess::READ},
new AstVarRef{nodep->fileline(), chgVarp, VAccess::WRITE}};
toggleVarRecurse(nodep->dtypeSkipRefp(), 0, newvec, nodep, chgVarp);
newvec.cleanup();
}
@@ -313,11 +314,12 @@ private:
for (int index_docs = bdtypep->lo(); index_docs < bdtypep->hi() + 1;
++index_docs) {
const int index_code = index_docs - bdtypep->lo();
ToggleEnt newent(above.m_comment + string("[") + cvtToStr(index_docs) + "]",
new AstSel(varp->fileline(), above.m_varRefp->cloneTree(true),
index_code, 1),
new AstSel(varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code, 1));
ToggleEnt newent{above.m_comment + std::string{"["} + cvtToStr(index_docs)
+ "]",
new AstSel{varp->fileline(), above.m_varRefp->cloneTree(true),
index_code, 1},
new AstSel{varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code, 1}};
toggleVarBottom(newent, varp);
newent.cleanup();
}
@@ -327,11 +329,11 @@ private:
} else if (const AstUnpackArrayDType* const adtypep = VN_CAST(dtypep, UnpackArrayDType)) {
for (int index_docs = adtypep->lo(); index_docs <= adtypep->hi(); ++index_docs) {
const int index_code = index_docs - adtypep->lo();
ToggleEnt newent(above.m_comment + string("[") + cvtToStr(index_docs) + "]",
new AstArraySel(varp->fileline(),
above.m_varRefp->cloneTree(true), index_code),
new AstArraySel(varp->fileline(),
above.m_chgRefp->cloneTree(true), index_code));
ToggleEnt newent{above.m_comment + std::string{"["} + cvtToStr(index_docs) + "]",
new AstArraySel{varp->fileline(),
above.m_varRefp->cloneTree(true), index_code},
new AstArraySel{varp->fileline(),
above.m_chgRefp->cloneTree(true), index_code}};
toggleVarRecurse(adtypep->subDTypep()->skipRefp(), depth + 1, newent, varp,
chgVarp);
newent.cleanup();
@@ -340,11 +342,11 @@ private:
for (int index_docs = adtypep->lo(); index_docs <= adtypep->hi(); ++index_docs) {
const AstNodeDType* const subtypep = adtypep->subDTypep()->skipRefp();
const int index_code = index_docs - adtypep->lo();
ToggleEnt newent(above.m_comment + string("[") + cvtToStr(index_docs) + "]",
new AstSel(varp->fileline(), above.m_varRefp->cloneTree(true),
index_code * subtypep->width(), subtypep->width()),
new AstSel(varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code * subtypep->width(), subtypep->width()));
ToggleEnt newent{above.m_comment + std::string{"["} + cvtToStr(index_docs) + "]",
new AstSel{varp->fileline(), above.m_varRefp->cloneTree(true),
index_code * subtypep->width(), subtypep->width()},
new AstSel{varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code * subtypep->width(), subtypep->width()}};
toggleVarRecurse(adtypep->subDTypep()->skipRefp(), depth + 1, newent, varp,
chgVarp);
newent.cleanup();
@@ -355,11 +357,11 @@ private:
itemp = VN_AS(itemp->nextp(), MemberDType)) {
AstNodeDType* const subtypep = itemp->subDTypep()->skipRefp();
const int index_code = itemp->lsb();
ToggleEnt newent(above.m_comment + string(".") + itemp->name(),
new AstSel(varp->fileline(), above.m_varRefp->cloneTree(true),
index_code, subtypep->width()),
new AstSel(varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code, subtypep->width()));
ToggleEnt newent{above.m_comment + std::string{"."} + itemp->name(),
new AstSel{varp->fileline(), above.m_varRefp->cloneTree(true),
index_code, subtypep->width()},
new AstSel{varp->fileline(), above.m_chgRefp->cloneTree(true),
index_code, subtypep->width()}};
toggleVarRecurse(subtypep, depth + 1, newent, varp, chgVarp);
newent.cleanup();
}
@@ -367,9 +369,9 @@ private:
// Arbitrarily handle only the first member of the union
if (const AstMemberDType* const itemp = adtypep->membersp()) {
AstNodeDType* const subtypep = itemp->subDTypep()->skipRefp();
ToggleEnt newent(above.m_comment + string(".") + itemp->name(),
ToggleEnt newent{above.m_comment + std::string{"."} + itemp->name(),
above.m_varRefp->cloneTree(true),
above.m_chgRefp->cloneTree(true));
above.m_chgRefp->cloneTree(true)};
toggleVarRecurse(subtypep, depth + 1, newent, varp, chgVarp);
newent.cleanup();
}
+2 -1
View File
@@ -20,9 +20,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3CoverageJoin.h"
#include "V3DupFinder.h"
#include "V3Global.h"
#include "V3Stats.h"
#include <vector>
+4 -3
View File
@@ -36,9 +36,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Dead.h"
#include "V3Ast.h"
#include "V3Global.h"
#include <map>
#include <vector>
@@ -85,7 +86,7 @@ private:
}
if (AstNode* const subnodep = nodep->getChildDTypep()) subnodep->user1Inc();
}
void checkVarRef(AstNodeVarRef* nodep) {
void checkVarRef(AstNodeVarRef* nodep) const {
if (nodep->classOrPackagep() && m_elimCells) nodep->classOrPackagep(nullptr);
}
void checkDType(AstNodeDType* nodep) {
@@ -316,7 +317,7 @@ private:
}
}
}
bool mightElimVar(AstVar* nodep) {
bool mightElimVar(AstVar* nodep) const {
if (nodep->isSigPublic()) return false; // Can't elim publics!
if (nodep->isIO() || nodep->isClassMember()) return false;
if (nodep->isTemp() && !nodep->isTrace()) return true;
+11 -1
View File
@@ -51,9 +51,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Delayed.h"
#include "V3Ast.h"
#include "V3Global.h"
#include "V3Stats.h"
#include <algorithm>
@@ -94,6 +95,7 @@ private:
bool m_inDly = false; // True in delayed assignments
bool m_inLoop = false; // True in for loops
bool m_inInitial = false; // True in initial blocks
bool m_ignoreBlkAndNBlk = false; // Suppress delayed assignment BLKANDNBLK
using VarMap = std::map<const std::pair<AstNodeModule*, std::string>, AstVar*>;
VarMap m_modVarMap; // Table of new var names created under module
VDouble0 m_statSharedSet; // Statistic tracking
@@ -105,6 +107,7 @@ private:
void markVarUsage(AstNodeVarRef* nodep, bool blocking) {
// Ignore if warning is disabled on this reference (used by V3Force).
if (nodep->fileline()->warnIsOff(V3ErrorCode::BLKANDNBLK)) return;
if (m_ignoreBlkAndNBlk) return;
if (blocking) nodep->user5(true);
AstVarScope* const vscp = nodep->varScopep();
// UINFO(4, " MVU " << blocking << " " << nodep << endl);
@@ -485,6 +488,13 @@ private:
}
}
virtual void visit(AstNodeReadWriteMem* nodep) override {
VL_RESTORER(m_ignoreBlkAndNBlk);
m_ignoreBlkAndNBlk = true; // $readmem/$writemem often used in mem models
// so we will suppress BLKANDNBLK warnings
iterateChildren(nodep);
}
virtual void visit(AstNodeFor* nodep) override { // LCOV_EXCL_LINE
nodep->v3fatalSrc(
"For statements should have been converted to while statements in V3Begin");
+3 -5
View File
@@ -26,9 +26,10 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Depth.h"
#include "V3Ast.h"
#include "V3Global.h"
#include "V3UniqueNames.h"
#include <algorithm>
@@ -68,10 +69,7 @@ private:
// Put assignment before the referencing statement
AstAssign* const assp = new AstAssign{
nodep->fileline(), new AstVarRef{nodep->fileline(), varp, VAccess::WRITE}, nodep};
VNRelinker linker2;
m_stmtp->unlinkFrBack(&linker2);
assp->addNext(m_stmtp);
linker2.relink(assp);
m_stmtp->addHereThisAsNext(assp);
}
// VISITORS
+2 -1
View File
@@ -23,10 +23,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3DepthBlock.h"
#include "V3Ast.h"
#include "V3EmitCBase.h"
#include "V3Global.h"
#include <algorithm>
+4 -3
View File
@@ -25,10 +25,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3Descope.h"
#include "V3Ast.h"
#include "V3EmitCBase.h"
#include "V3Global.h"
#include <map>
@@ -248,8 +249,8 @@ private:
VL_RESTORER(m_funcp);
if (!nodep->user1()) {
// Static functions should have been moved under the corresponding AstClassPackage
UASSERT(!(nodep->isStatic() && VN_IS(m_modp, Class)),
"Static function under AstClass");
UASSERT_OBJ(!(nodep->isStatic() && VN_IS(m_modp, Class)), nodep,
"Static function under AstClass");
m_funcp = nodep;
iterateChildren(nodep);
nodep->user1(true);
+2 -1
View File
@@ -17,10 +17,11 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Global.h"
#include "V3DupFinder.h"
#include "V3Ast.h"
#include "V3File.h"
#include "V3Global.h"
#include <algorithm>
#include <iomanip>
+4 -5
View File
@@ -23,8 +23,8 @@
#include "config_build.h"
#include "verilatedos.h"
#include "V3Error.h"
#include "V3Ast.h"
#include "V3Error.h"
#include "V3Hasher.h"
#include <map>
@@ -44,7 +44,7 @@ class V3DupFinder final : private std::multimap<V3Hash, AstNode*> {
using Super = std::multimap<V3Hash, AstNode*>;
// MEMBERS
const V3Hasher* const m_hasherp; // Pointer to owned hasher
const V3Hasher* const m_hasherp = nullptr; // Pointer to owned hasher
const V3Hasher& m_hasher; // Reference to hasher
public:
@@ -52,9 +52,8 @@ public:
V3DupFinder()
: m_hasherp{new V3Hasher}
, m_hasher{*m_hasherp} {}
V3DupFinder(const V3Hasher& hasher)
: m_hasherp{nullptr}
, m_hasher{hasher} {}
explicit V3DupFinder(const V3Hasher& hasher)
: m_hasher{hasher} {}
~V3DupFinder() {
if (m_hasherp) delete m_hasherp;
}

Some files were not shown because too many files have changed in this diff Show More