Commit Graph

51 Commits

Author SHA1 Message Date
Krzysztof Bieganski 99ea16d7fd
Fix dynamic triggers for named events (#4571) 2023-10-16 11:06:41 -04:00
Aleksander Kiryk ad3bcbb1bb
Support `disable fork` (#4125) (#4569) 2023-10-16 14:02:29 +02:00
Mariusz Glebocki 28bd7e5b19
Rework multithreading handling to separate by code units that use/never use it. (#4228) 2023-09-24 22:12:23 -04:00
Wilson Snyder 05d04a3959 Internals: Fix misnamed member. No functional change. 2023-09-15 23:02:34 -04:00
Wilson Snyder 10c1653e72 Fix ZERODLY to not warn on 'wait(0)'. 2023-09-15 08:53:29 -04:00
Krzysztof Bieganski ffbbd438ae
Internals: Use runtime type info instead of `dynamic_cast` for faster graph type checks (#4397) 2023-08-31 18:00:53 -04:00
Ryszard Rozak e24197fd16
Don't move function calls before the expression (#4413) 2023-08-28 15:44:41 +02:00
Ryszard Rozak 90079c2974
Fix nested assignments on the LHS (#4435) 2023-08-23 12:08:22 +02:00
Krzysztof Boroński b752faa107
Fix jumping over object initialization (#4411) 2023-08-11 18:28:37 +02:00
Krzysztof Bieganski 9caa79a7ea
Internals: Remove the name field from `AstVarRef` (#4395) 2023-08-03 02:52:52 -04:00
Aleksander Kiryk 1549ff9185
Support more types in wait (#4374) 2023-07-24 08:14:01 -04:00
Krzysztof Boroński 4bdda3f240
Improve propagation of process requirement and decouple it from suspendability (#4321) 2023-07-14 17:12:02 +02:00
Wilson Snyder ff4923cf67 Internals: Make V3MemberMap towards removing member cache (#4350) 2023-07-08 12:27:50 -04:00
Krzysztof Bieganski d5d1cc47e0
Fix multiple edge timing controls in class methods (#4318) (#4320) (#4344)
Multiple edge timing controls in class methods would cause compilation errors on
the generated C++ code. This is because the `SenExprBuilder` used for these
would get recreated per timing control, resulting in duplicate variable names.
The fix is to have a single `SenExprBuilder` per scope.
2023-07-07 08:19:49 -04:00
Wilson Snyder 294d78e0e2 Backout unstable dbb35ab8 (#4318) (#4320) 2023-07-05 14:26:28 -04:00
Krzysztof Bieganski 35d6da391b
Fix multiple edge timing controls in class methods (#4318) (#4320)
Multiple edge timing controls in class methods would cause compilation errors on the generated C++ code. This is because the `SenExprBuilder` used for these would get recreated per timing control, resulting in duplicate variable names. The fix is to have a single `SenExprBuilder` per scope.
2023-07-05 10:36:10 +00:00
Aleksander Kiryk 32019d2bc4
Fix state update for always processes (#4311) 2023-06-29 09:21:49 -04:00
Ryszard Rozak 4522834f7a
Fix duplicate fork names (#4295) 2023-06-22 06:51:53 -04:00
Krzysztof Bieganski 058233fc87
Set the global `usesTiming` flag if forks exist (#4274) 2023-06-06 09:35:12 -04:00
Aleksander Kiryk db7935faf3
Add std::process class (#4212) 2023-06-01 10:02:08 -04:00
Krzysztof Bieganski 77502aeb97
Add warning that timing controls in DPI exports are unsupported (#4238)
Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2023-05-30 09:00:10 -04:00
Krzysztof Bieganski 519792d02b
Fix to commit coroutines immediately on `wait` statements (#4229)
Event-triggered coroutines live in two stages: 'uncommitted' and 'ready'. First
they land in 'uncommitted', meaning they can't be resumed yet. Only after
coroutines from the 'ready' queue are resumed, the 'uncommitted' ones are moved
to the 'ready' queue, and can be resumed. This is to avoid self-triggering in
situations like waiting for an event immediately after triggering it.

However, there is an issue with `wait` statements. If you have a `wait(b)`, it's
being translated into a loop that awaits a change in `b` as long as `b` is
false. If `b` is false at first, the coroutine is put into the `uncommitted`
queue. If `b` is set to true before it's committed, the coroutine won't get
resumed.

This patch fixes that by immediately committing event controls created from
`wait` statements. That means the coroutine from the example above will get
resumed from now on.
2023-05-25 20:20:44 -04:00
Krzysztof Boroński 167a30be1c
Fix deep traversal of class inheritance timing (#4216) 2023-05-23 09:01:57 -04:00
Krzysztof Bieganski 729f8b9334
Move suspendable detection to a separate visitor (#4208)
This makes the implementation of the detection and propagation of the
suspendable property simpler and easier to read. More importantly, there are no
more jumps around the AST with the `visit` functions, which in some cases could
result in incorrect visitor context while in the `visit` function. See the added
test, which would cause Verilator to segfault before this patch.

In testing, verilation performance was not shown to be affected by this change.
Though there is a slight performance improvement from this patch, due to adding
one more check before refreshing class member cache.

Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2023-05-17 17:09:33 +00:00
Krzysztof Bieganski 76c5875912
Fix marking overridden methods as coroutines (#4120) (#4169)
This patch fixes two cases where methods in base classes were not being marked
as coroutines, even though they were being overridden by coroutines.
- One case is the class member cache not getting refreshed for searched classes.
- The other is when the overriding methods are not declared as `virtual`. In
  that case, the `isVirtual()` getter on such a method returns false, which led
  to `V3Timing` skipping the step of searching for overridden methods.

Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2023-05-04 15:27:45 +02:00
Wilson Snyder add68130b8 Internals: Rename to dumpLevel(), to avoid confusion with make-a-dump() 2023-05-03 18:04:10 -04:00
Kamil Rakoczy 65a484e00b
Internal: Update clang_check_annotations conditions (#4134) 2023-04-20 07:02:31 -04:00
Kamil Rakoczy 798d7346cf
Internals: Add VL_MT_SAFE attribute to functions that requires locking. (#3805) 2023-03-17 20:24:15 -04:00
Felix Neumärker a3ff375ce7
Fix timing delays to not truncate below 64 bits (#3973) (#3982) 2023-02-27 21:42:22 -05:00
Wilson Snyder 4703fc39be Add error to avoid @* fatal 2023-02-26 13:13:02 -05:00
Wilson Snyder b24d7c83d3 Copyright year update 2023-01-01 10:18:39 -05:00
Krzysztof Bieganski bb44d4e4f2
Support clocking blocks (#3674) 2022-12-23 07:34:49 -05:00
Krzysztof Bieganski 87bb32fde1
Fix forks without any delayed statements (#3792) (#3801) 2022-12-06 07:14:29 -05:00
Wilson Snyder 66d85b3381 Internals: Fix cppcheck warnings. No functional change intended. 2022-11-21 21:40:49 -05:00
Wilson Snyder 25f970eac2 Internals: Fix constructor style. 2022-11-20 15:06:49 -05:00
Geza Lore 3abb65d732 Strengthen AstNode types to AstNodeExpr
Declare every AstNode children and variables as AstNodeExpr where we
statically know this is the appropriate sub-type.
2022-11-20 19:31:28 +00:00
Wilson Snyder 9d7c4d9af3 Fix wait 0. 2022-11-11 17:18:59 -05:00
Geza Lore 65e08f4dbf Make all expressions derive from AstNodeExpr (#3721).
Apart from the representational changes below, this patch renames
AstNodeMath to AstNodeExpr, and AstCMath to AstCExpr.

Now every expression (i.e.: those AstNodes that represent a [possibly
void] value, with value being interpreted in a very general sense) has
AstNodeExpr as a super class. This necessitates the introduction of an
AstStmtExpr, which represents an expression in statement position, e.g :
'foo();' would be represented as AstStmtExpr(AstCCall(foo)). In exchange
we can get rid of isStatement() in AstNodeStmt, which now really always
represent a statement

Peak memory consumption and verilation speed are not measurably changed.

Partial step towards #3420
2022-11-03 16:02:16 +00:00
Krzysztof Bieganski fcf0d03cd4
Dynamic triggers for non-static contexts (#3599)
In non-static contexts like class objects or stack frames, the use of
global trigger evaluation is not feasible. The concept of dynamic
triggers allows for trigger evaluation in such cases. These triggers are
simply local variables, and coroutines are themselves responsible for
evaluating them. They await the global dynamic trigger scheduler object,
which is responsible for resuming them during the trigger evaluation
step in the 'act' eval region. Once the trigger is set, they await the
dynamic trigger scheduler once again, and then get resumed during the
resumption step in the 'act' eval region.

Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2022-10-22 14:05:39 +00:00
Krzysztof Bieganski 785c51dd0b
Fix emitting timing debug info with `--protect-ids` (#3689) (#3701) 2022-10-21 16:56:44 -04:00
HungMingWu 196f3292d5 Improve V3Ast function usage ergonomics (#3650)
Signed-off-by: HungMingWu <u9089000@gmail.com>
2022-10-21 14:12:12 +01:00
Krzysztof Bieganski 5688d1a935
Internals: Add `V3UniqueNames` consistency assertion (#3692) 2022-10-21 07:05:38 -04:00
Krzysztof Bieganski 8a347248f5
Use `AstDelay` nodes for intra-assignment delays (#3672)
Also fix messy implementation of net delays.

Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2022-10-14 09:35:26 +02:00
Wilson Snyder 10fc1f757c Internals: cppcheck cleanups. No functional change intended. 2022-10-02 23:04:55 -04:00
Krzysztof Bieganski 9c2ead90d5
Add custom memory management for verilated classes (#3595)
This change introduces a custom reference-counting pointer class that
allows creating such pointers from 'this'. This lets us keep the
receiver object around even if all references to it outside of a class
method no longer exist. Useful for coroutine methods, which may outlive
all external references to the object.

The deletion of objects is deferred until the next time slot. This is to
make clearing the triggered flag on named events in classes safe
(otherwise freed memory could be accessed).
2022-09-28 18:54:18 -04:00
Geza Lore ddb678cc5b Merge branch 'master' into develop-v5 2022-09-22 17:33:36 +01:00
Geza Lore 95145038b4 Generate AstNode accessors via astgen
Introduce the @astgen directives parsed by astgen, currently used for
the generation child node (operand) accessors. Please see the updated
internal documentation for details.
2022-09-21 14:05:27 +01:00
Geza Lore 7bc7b5372e Merge branch 'master' into develop-v5 2022-09-17 16:12:28 +01:00
Geza Lore af305bf280 Merge branch 'master' into develop-v5 2022-09-16 16:24:36 +01:00
Krzysztof Bieganski a2e1b32a1c
Fix inlining of forks (#3594)
Before this change, some forked processes were being inlined in
`V3Timing` because they contained no `CAwait`s. This only works under
the assumption that no `CAwait`s will be added there later, which is not
true, as a function called by a forked process could be turned into a
coroutine later. The call would be wrapped in a new `CAwait`, but the
process itself would have already been inlined at this point.

This commit moves the inlining to `transformForks` in `V3SchedTiming`,
which is called at a point when all `CAwait`s are already in place.

Signed-off-by: Krzysztof Bieganski <kbieganski@antmicro.com>
2022-09-05 15:19:19 +01:00