Patches 1 to 5, uefirtvariable: Add new test for UEFI runtime SetVariable, subtest 5

[Colin Ian King]

Thanks Ivan for the UEFI runtime SetVariable tests.

I've ACK'd these tests as they exercise the SetVariable runtime well 
enough.  I suspect we may need to re-visit these at a later date and 
perhaps had some more feedback to the user in the event of failures.

For example, setvariable_insertvariable reports back:

"Failed to set variable with UEFI runtime service."

..which is useful to know the that something failed, but I think we 
could expand this a little more. For example, maybe we could add some 
more feedback to the user such as the length of the variable name, size 
of the data, it's attributes, etc.   Imagine if you are running the test 
and you get a failed error - what does that relate to? As a user, it 
could perhaps be useful to know.

Also, let's look at patch 2, uefirtvariable_test1(), you added:


+    for (index = 0; index < (sizeof(attributesarray)/(sizeof 
attributesarray[0])); index++) {
+        if (setvariable_test2(fw, attributesarray[index], 
variablenametest) == FWTS_ERROR)
+            return FWTS_ERROR;
+    }

Note that one can add multiple fwts_passed() and fwts_failed() messages 
in the test, so for each of these minor tests you could feedback a 
fwts_passed() message..

e.g.

	fwts_log_info(fw, "Testing SetVariable on a variable with two different 
sizes.");
         .. do the setvariable_test2() testss..
	for (index = 0; index < ... etc) {
		if (setvariable_test2(fw, ...) == FWTS_ERROR)
			return FWTS_FAILED;

         fwts_passed(fw, "SetVariable on a variable with two different 
sizes passed.");

Anyhow, I'm OK with the tests as they are, but perhaps we should 
re-visit them and add a little more feedback if it doesn't clutter up 
the output too much.

Colin