XO 1.5 Tests

  This page is monitored by the OLPC team.

These are functional tests remaining for the XO 1.5 A1 motherboard.

List of Tests

XO-1.5 A1 Prototype Test Status

Subsystem Test	OFW	Linux	Notes
Suspend/Resume 1	Y	Y	Basic manually triggered suspend/resume
Suspend/Resume 2	Y	Y	Many cycle periodically triggered suspend/resume
Storage Suspend/Resume	Y	Y	Testing SD storage integrity
Storage Error Rate	Y	Y	Testing SD storage bit error rate
SD/MMC Port Testing	Y	Y	Testing SD/MMC driver on Via demo board
Video Capture	Y	Y	Trac #9377
Lid Switch	Y	Y
EBook Switch	Y	Y
I2C Communications with Clock Generator	Y	Y	decreases power consumption
EC Control of Charge Current	Y	Y
USB Power	Y	Y	Trac #9366

Suspend/Resume 2 Tests

This is a repeated suspend of the motherboard, followed by a resume triggered by a RTC event. It performs the test in Open Firmware.

Procedure

Obtain an Open Firmware prompt on the laptop. Type:

rtc-wackup

When you have to stop the test, record the cycle number and reboot. Please record the number below, along with any information about a crash of some sort during this test.

Results

XO-1.5 A1 Prototype Test Status

Motherboard	Num. Consecutive Passes	Crash ?	Notes
Isabel	24000
Isabel	19000
Isabel	196000

Storage Suspend Test

This test exercises the reliability of an SD storage device while continually suspending and resuming the laptop.

Storage Error Rate

This test attempts to measure the bit error rate of the SD interface on the XO-1.5. The SD interface, and not the SD cards themselves, have been implicated in a number of errors across the testing of NAND storage devices on XO-1. We need to both verify that this is really the case and also verify that the SD interface on the XO-1.5 is rock-solid. I propose running the same tests on any idle XO-1.5 prototype to accumulate more data. Contact wad for an SD card and instructions if you have an A1 prototype.

EC Control of Charge Current

We need to verify that the EC can indeed modify the charge current of the laptop.

We failed this test. We can influence the charge current, but only in the wrong direction. Additional work is needed.