#656 saHpiResourceTagSet can't work for some sensors

[wang, jing]

When using saHpiResourceTagSet to set tag for
resource, it always return
SA_ERR_HPI_INVALID_CMD

Following are test case piece:

memset(&tag, 0, sizeof(tag));
tag.DataType = SAHPI_TL_TYPE_BINARY;
tag.Language = SAHPI_LANG_ENGLISH;
tag.DataLength = STR_LEN;
memcpy(tag.Data, TEST_STR, sizeof
(tag.Data));
val = saHpiResourceTagSet(session_id,
resource_id, &tag);
if (val != SA_OK) {
printf(" Does not conform the expected
behaviors!\n");
printf(" Set new tag failed!\n");
printf(" Return value: %s\n",
get_error_string(val));
val = HPI_TEST_FAIL;
}
else
{
val = saHpiRptEntryGetByResourceId
(session_id,

rpt_entry.ResourceId,
&rpt_entry);
if (val != SA_OK) {
ret = HPI_TEST_NA;
}
}

output:
src/rpt_entry_process/saHpiResourceTagSet_T1:
execution: FAILED: Output: 255
Does not conform the expected behaviors!
Set new tag failed!
Return value: SA_ERR_HPI_INVALID_CMD

[Sean Dague]

Logged In: YES
user_id=21949

this is a plugin bug, guessing you are testing with OpenIPMI
for now, but if not please report which you are.

[wang, jing]

Logged In: YES
user_id=1152287

This bug related to critical bug#1080960. I use hpi_shell to
set tag to non-present resource, it fail. As tariq talked, only
in ipmi level, user can see the resource is present or not. In
OpenHPI level, we can't tell this, that is really a big issue.

[wang, jing]

Logged In: YES
user_id=1152287

If resource is MC, the settag will fail. So it should be test
case bug

[Sean Dague]

Logged In: YES
user_id=21949

This is not a test case bug.

This function is required to be supported for every resource
in HPI. If a plugin isn't supporting that than that plugin
is non conformant.

Note, the resource tag doesn't need to be persistant (i.e.
stored to nvram or anything), however an end user is
required to be allowed to set the resource tags for all
their resources.

[Tariq Shureih]

Logged In: YES
user_id=118691

Duplicate of: 1082666