RTC time conversion test - reduce test execution time.

[mprse]

Description

Reduce number of tested cases as follows:

For each of the following years test example time of the first and last day of each month:

• first - 1970
• example not leap year (not divisible by 4)
• example leap year (divisible by 4 and by 100 and by 400)
• example leap year (divisible by 4 and not by 100)
• example not leap year (divisible by 4 and by 100)
• last fully supported - 2105

Test execution time on K64F is now ~39 sec.

Status

READY

Migrations

NO

[0xc0170]

/morph build

[mbed-ci]

Build : SUCCESS

Build number : 812
Build artifacts/logs : http://mbed-os.s3-website-eu-west-1.amazonaws.com/?prefix=builds/5802/

Triggering tests

/morph test
/morph uvisor-test
/morph export-build

[mbed-ci]

Exporter Build : SUCCESS

Build number : 462
Build artifacts/logs : http://mbed-os.s3-website-eu-west-1.amazonaws.com/?prefix=builds/exporter/5802/

[mbed-ci]

Test : SUCCESS

Build number : 652
Test logs :http://mbed-os-logs.s3-website-us-west-1.amazonaws.com/?prefix=logs/5802/652

[cmonr]

Nice job on lowering the execution time.

Is there a reason that self.notify_complete(False) isn't present after e725b4c#diff-f51e22d57406e551f8f327cbf2119b77R120 ?

Also, is there a reason that a single timestamp is compared at once, instead doing all comparisons with a single kv transaction?

[mprse]

@cmonr

Is there a reason that self.notify_complete(False) isn't present after e725b4c#diff-f51e22d57406e551f8f327cbf2119b77R120 ?

There is no reason. Do we need this?

Also, is there a reason that a single timestamp is compared at once, instead doing all comparisons with a single kv transaction?

Yes, there is a reason. This test verifies two functions _rtc_maketime and _rtc_localtime. First one converts calendar time to timestamp. The result is then send to the host and verified against timestamp created using python library for the same calendar date/time. If the result is correct, then "passed" key is send to the device with encoded value which holds expected week day and year day for the tested calendar date/time (this is needed to test _rtc_localtime). Second function converts timestamp back to calendar date/time, returned structure except date and time provides also week day and year day. So now the verified timestamp value is converted back to calendar date/time and check is made if result structure has date and time same which have been given to _rtc_maketime and valid week day and year day (provided from host).

[cmonr]

Is there a reason that self.notify_complete(False) isn't present after e725b4c#diff-f51e22d57406e551f8f327cbf2119b77R120 ?

There is no reason. Do we need this?

I was thinking that we did, but it looks I was mistaken.

The reasoning for the test messaging makes sense. I had originally wanted to simply the back and forth traffic to make logging a bit cleaner, but it's not a high priority.

[adbridge]

This is dependent on #5363 scheduled for 5.8. Thus this also needs to go to 5.8.