Custom Query (74 matches)

The check for over specific runs on clauses is erroneously triggered in certain cases.

Considering the example below, if the last called function is over-specific on its own then the calling is flagged as over-specific as well (incorrectly). Additionally, this applies only if it is the last otherwise the calling one is not flagged. Finally, there has to be at least one level in between.

module Issue1m {

	type port P1 message {
		inout integer;
	}
	
	type component C1 {
		var integer vc_III;
	}

	type component C2 {
		var integer vc_I;
		port P1 PP;
	}

	function f_F1 () runs on C2 {
		f_F1_1();
		f_F2();
		f_F2_1(); 
		//if the last one is over specific 
		//then the calling is flagged as over-specific as well (incorrectly)
		// - if it is not the last then the calling one is not flagged
		// - there has to be at least one level in between
	}

	function f_F2 () runs on C2 {
		//second level reference
                f_F3();
	}
	function f_F3 () runs on C2 {
		PP.send(1);
	}

	function f_F1_1 () runs on C1 {
		//reference to an element of another component
                vc_III := 1;
	}
	function f_F2_1 () runs on C1 {
		//no reference
	}
}

The check for over specific runs on clauses is not triggered in obvious cases

Considering the example below, calling a function that runs on a different component and references its elements will cause the check for over specific runs on to fail on the calling function.

module Issue1m {

	type component C1 {
	}

	type component C2 {
		var integer vc_I;
	}

	function f_F1 () runs on C1 {
		f_F2(); //This fixes the over-specific runs on even though it has nothing to do with it
	}

	function f_F2 () runs on C2 {
		vc_I := 1;
	}
}

The issue is related to incorrect termination criteria when searching for references to component definition recursively. It can terminate prematurely if the last called function does not have a runs on clause and is the same on several levels. In the example below, if the order of f_FX and f_F0 is changed it will be fine, but not as it is.

module Issue1 {
	type component C1 {
		var integer vc_V1;
	}

	function f_F0() runs on C1 {
		vc_V1 := 1;
	}

	function f_F1() runs on C1 {
		f_F0();
		f_FX();
	}

	function f_F2() runs on C1 {
		f_F1();
		f_FX();
	}

	function f_F3() runs on C1 {
		f_F2();
	}
	
	function f_FX() {
	}
}

Original report:

We have a bug report on a possible issue regarding T3Qv2 (latest working version, T3Q v2.0.0b18) on rule "checkNoOverSpecificRunsOnClauses"

Comparing the execution on module EUTRA_CommonFunctions_MBMS run with both T3Q and T3Qv2, there is no error message on T3Q but on T3Qv2 there is an error message that we think that should not be happening:

Validation output of C:\Users\arroyona\Documents\STF160\Trunk\ttcn\Common\EUTRA\EUTRA_CommonFunctions_MBMS.ttcn

C:\Users\arroyona\Documents\STF160\Trunk\ttcn\Common\EUTRA\EUTRA_CommonFunctions_MBMS.ttcn: 218: WARNING: Code Style: Definition for "f_EUTRA_CellInfo_GetMBMS_NotRepCoeff" contains an over specific runs on clause! (6.13, checkNoOverSpecificRunsOnClauses)

Where the function raising the error message is as follows:

/*

• @desc Get the notification Repetition Coeff of the cell

• @param p_CellId

• @return integer

• @status APPROVED (LTE_AR10R11)

*/

function f_EUTRA_CellInfo_GetMBMS_NotRepCoeff(EUTRA_CellId_Type p_CellId) runs on EUTRA_PTC return integer

{

var SystemInformationBlockType13_r9 v_Sib13 := f_EUTRA_CellInfo_GetSib13(p_CellId);

var MBMS_NotificationRepCoeff_Type v_MCCH_NotRepCoeff := v_Sib13.notificationConfig_r9.notificationRepetitionCoeff_r9;

var integer v_RepCoeff;

select (v_MCCH_NotRepCoeff) {

case (n2) {

v_RepCoeff:= 2;

}

case (n4) {

v_RepCoeff := 4;

}

case else {

FatalError? (FILE, LINE, "invalid value for Notification Co-eff");

}

}

return v_RepCoeff;

}

The runs on message is necessary here as the call to function @EUTRA_CommonFunctions_MBMS.f_EUTRA_CellInfo_GetSib13' runs over component type@…_PTC'

This can be checked using the source code of our last delivery of 17wk10, on ATS LTE_AR10R11IWD17wk10:

​http://www.3gpp.org/ftp/tsg_ran/WG5Test_ex-T1/TTCN/Deliveries/TTCN3/iwd-TTCN3-B2016-06D17wk10.zip

or directly testing following file of the ATS with T3qv2:

.\iwd-TTCN3-B2016-06D17wk10\LTE_AR10R11IWD17wk10\LTE_AR10R11IWD_17wk10\Common\EUTRA\EUTRA_CommonFunctions_MBMS.ttcn

Please let us know your thoughts and in case you agree, we will wait for a fixed version of the tool.