Opened 9 years ago
Closed 4 years ago
#3237 closed defect (wontfix)
Returning the cache in Mod.elabModValue breaks some models
Reported by: | perost | Owned by: | somebody |
---|---|---|---|
Priority: | high | Milestone: | never |
Component: | Backend | Version: | trunk |
Keywords: | Cc: | lochel, wbraun, vwaurich |
Description
Mod.elabModValue previously didn't return the cache even though it used Ceval, which might be causing some issues with structural parameters (since information about them is stored in the cache). Returning the cache causes issues with some models though, so these issues needs to be fixed before we can return the cache. The only thing that needs to be done to return the cache from Mod.elabModValue is to replace the _ with outCache in the call to Ceval.ceval. The models that fail are:
IndexReduction.setVarKind failt because of wrong input: simulation/libraries/3rdParty/Exercises/ServoSystem1.Aufgabe3_2.mos simulation/libraries/3rdParty/Exercises/ServoSystem1.Aufgabe3_3.mos simulation/libraries/3rdParty/Exercises/ServoSystem3.Aufgabe3_2.mos simulation/libraries/3rdParty/Exercises/ServoSystem3.Aufgabe3_3.mos Different result: simulation/libraries/msl31/Modelica.Electrical.Machines.Examples.DCSE_Start.mos simulation/libraries/msl32/Modelica.Electrical.Machines.Examples.Transformers.Rectifier12pulse.mos Error: When solving linear system: simulation/libraries/msl32/Modelica.Mechanics.MultiBody.Examples.Constraints.PrismaticConstraint.mos Works fine, but shouldn't?: simulation/modelica/others/ParameterCycle.mos
A lot of other test cases fail because of slightly different output, but none of them seem to be wrong from what I can see.
Change History (1)
comment:1 Changed 4 years ago by perost
- Milestone changed from Future to never
- Resolution set to wontfix
- Status changed from new to closed
Might still be an issue, but we no longer care about fixing issues in the old frontend.