Circuit_Get_Losses segfaults

[keegit]

I have the following Rust binding to the Circuit_Get_Losses function. Calling it produces a segmentation fault and core dump.

dss_c::Circuit_Get_Losses(result_ptr, result_cnt);

output:

Segmentation fault (core dumped)

I'm assuming there's some state initialization that I haven't done, but I'm unsure what needs to be initialized.

[PMeira]

Which branch/commit are you using?

[PMeira]

I'm assuming there's some state initialization that I haven't done, but I'm unsure what needs to be initialized.

Yep. See also #56

[PMeira]

If you're using the master branch and 0.12 beta releases, please provide the sequence of calls so I can try to reproduce.

If it's 0.10.x, avoid opening new issues for that. Although there are a few more places that need state checks (per #56), half of the code has been rewritten since 0.10.x.

Worst case scenario, link to the debug version of DSS C-API and you should get detailed info -- either started through a debugger like GDB or load the core dump afterwards. Some general info at https://github.com/dss-extensions/dss_capi/blob/master/docs/debug.md

[keegit]

I'm using the master branch, here is the sequence of calls I make:

DSS_Start(0);
Text_Set_Command("Redirect <path to dss file>");
Solution_Set_Mode(0); // 0 for snapshot mode
Circuit_Set_ActiveElement();
Loads_Set_kW();
Loads_Set_kvar();
// conditionals
if device == reg {
  RegControls_Set_Name();
  RegControls_Set_TapNumber();
  RegControls_Set_Max_TapChange();
}
if device == cap {
  Capacitors_Set_Name();
  Capacitors_Set_States([0]);
}
Solution_Solve();
Circuit_Get_Losses(); <-- Segfault

[PMeira]

hm... seems fine. Are you initializing the two pointers for that function?

Solution_Set_Mode(0); // 0 for snapshot mode
There is an enum for that:

dss_capi/include/dss_capi.h

Lines 47 to 64 in cd3cf22

	enum SolveModes {
	SolveModes_SnapShot = 0, // Solve a single snapshot power flow
	SolveModes_Daily = 1, // Solve following Daily load shapes
	SolveModes_Yearly = 2, // Solve following Yearly load shapes
	SolveModes_Monte1 = 3, // Monte Carlo Mode 1
	SolveModes_LD1 = 4, // Load-duration Mode 1
	SolveModes_PeakDay = 5, // Solves for Peak Day using Daily load curve
	SolveModes_DutyCycle = 6, // Solve following Duty Cycle load shapes
	SolveModes_Direct = 7, // Solve direct (forced admittance model)
	SolveModes_MonteFault = 8, // Monte carlo Fault Study
	SolveModes_FaultStudy = 9, // Fault study at all buses
	SolveModes_Monte2 = 10, // Monte Carlo Mode 2
	SolveModes_Monte3 = 11, // Monte Carlo Mode 3
	SolveModes_LD2 = 12, // Load-Duration Mode 2
	SolveModes_AutoAdd = 13, // Auto add generators or capacitors
	SolveModes_Dynamic = 14, // Solve for dynamics
	SolveModes_Harmonic = 15 // Harmonic solution mode
	};

I noticed you're not checking for DSS errors. Check Error_Get_Number() -- if that is non-zero, Error_Get_Description() should list the error message.

[keegit]

Figured it out - the issue was that I was not properly initializing the pointers passed to the function.