HPI Conf Test: ControlStateSet failure

[openhpi2]

When running the HPI conformance test under the Blade center, we get an error from the saHpiControlStateGet_Set_T1.test

"Does not conform to expected behavior cannot set the specified control state! Return Value: SA_ERR_HPI_INVALID_CMD"

Reported by: carlmc1

[openhpi2]

• assigned_to: nobody --> renierm

Original comment by: carlmc1

[openhpi2]

• assigned_to: renierm --> carlmc1

Original comment by: renierm

[openhpi2]

Logged In: YES user_id=660960

Work on this with Kevin Gao, because this test must be corrected. The test fails if it gets anything but OK. It tries to set a value in this case that is not supported by the hardware, so the snmp_bc plugin returns SA_ERR_HPI_INVALID_CMD. This should be valid and fine. The test needs to change by setting more reasonable/probable values based on the state read at the beginning. For example, if it was a digital control and it was off, then try to set it to ON, not pulse on. <snip> case SAHPI_CTRL_TYPE_DIGITAL: state->StateUnion.Digital = SAHPI_CTRL_STATE_PULSE_ON; break; </snip>

Original comment by: renierm

[openhpi2]

• assigned_to: carlmc1 --> renierm

Original comment by: renierm

[openhpi2]

• milestone: --> 359430

Original comment by: sdague

[openhpi2]

• status: open --> closed-fixed

Original comment by: pdphan

[openhpi2]

Logged In: YES user_id=901950

In addition to changes made by Steve and Renier last week,

1. Change initial snmp timeout value to 2 seconds, snmp_bc_session.c, to communicate with OIDs that are I2C devices (BLADE-MIB::ledBladeIdentity.x and BLADE-MIB::ledBladeInformation.x) Because I2C devices are slower than others, the default timeout value is not sufficient.

2. Correct OID for BLADE-MIB::bootCountPowerOnTimeBoots, bc_resources.c and sim_resources.c

Original comment by: pdphan

[openhpi2]

• milestone: 359430 --> Future
• status: closed-fixed --> pending-fixed

Original comment by: pdphan

[openhpi2]

Logged In: YES user_id=901950

Change status to pending and group to future because I am working with Raleigh BC-I mm firmware to resolve the timing problem seen by Renier with BC-I mm firmware BRET59E.

$ snmpget -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 SNMPv2-SMI::enterprises.2.3.51.2.2.8.2.1.1.11.1 = INTEGER: 0 $ snmpset -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 i 2 Timeout: No Response from bcImm. $ snmpget -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 Timeout: No Response from bcImm. $ snmpget -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 Timeout: No Response from bcImm. $ snmpget -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 Timeout: No Response from bcImm. $ snmpget -c bcpublic3 -v 1 bcImm .1.3.6.1.4.1.2.3.51.2.2.8.2.1.1.11.1 SNMPv2-SMI::enterprises.2.3.51.2.2.8.2.1.1.11.1 = INTEGER: 2

Original comment by: pdphan

[openhpi2]

• status: pending-fixed --> closed-fixed

Original comment by: carlmc1