“firmware_UserRequestRecovery.ec_wp” failure.

[Martin Yan]

As tester runs suite:faft_ec, it fails at firmware_UserRequestRecovery.ec_wp, and test separate “firmware_UserRequestRecovery.ec_wp” job, the failure information as, and DUT is stuck and displays “Chrome OS is missing or damaged”,

 

-------------------------------------------------------------------------------------------------------------------------

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp                                    [  FAILED  ]

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp                                      FAIL: Not succeed: calling crossystem_checker with args ({'recovery_reason': '193', 'mainfw_type': 'recovery'},) returning False

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp [  FAILED  ]

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp   FAIL: Not succeed: calling crossystem_checker with args ({'recovery_reason': '193', 'mainfw_type': 'recovery'},) returning False

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp   12/12 10:01:04.908 ERROR|             utils:0280| [stderr] mux_client_request_session: read from master failed: Broken pipe

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp   12/12 10:01:04.963 ERROR|             utils:0280| [stderr] Warning: Permanently added '192.168.1.4' (ED25519) to the list of known hosts.

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp   12/12 10:02:05.476 ERROR|             utils:0280| [stderr] mux_client_request_session: read from master failed: Broken pipe

/tmp/test_that_results_YJy2Rk/results-1-firmware_UserRequestRecovery.ec_wp/firmware_UserRequestRecovery.ec_wp   12/12 10:02:06.564 ERROR|             utils:0280| [stderr] Warning: Permanently added '192.168.1.4' (ED25519) to the list of known hosts.

 

Also attached DEBUG files, any suggestions?

 

The USB device attached on servo board has image extracted from “chromiumos_test_image.bin” and it passes firmware_FAFTSetup.

 

[Shawn N]

12/12 10:01:10.004 INFO | test_runner_utils:0199| autoserv| calling

crossystem_checker with args ({'recovery_reason': '193',
'mainfw_type': 'recovery'},)

12/12 10:01:10.063 INFO | test_runner_utils:0199| autoserv| Expected
'recovery_reason' value '193' but got '2'

/* User manually requested recovery via recovery button */
VB2_RECOVERY_RO_MANUAL = 0x02,
...
/* Recovery mode test from user-mode */
VB2_RECOVERY_US_TEST = 0xc1,

Maybe EC_HOST_EVENT_KEYBOARD_RECOVERY was set in a previous boot and
never cleared?

> --
> You received this message because you are subscribed to the Google Groups
> "faft users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to faft-users+...@chromium.org.

[Tom Wai-Hong Tam (談偉航)]

If your system boot time takes longer than the normal, the recovery-related test

logic will time out and try 3-finger recovery boot (0x02), which will overwrite the

expected recovery reason (0xc1 or 193).

The default timeout value is delay_reboot_to_ping = 30. You can check this FAFT

config to a larger value, like 60.

> email to faft-users+unsubscribe@chromium.org.

--
You received this message because you are subscribed to the Google Groups "faft users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+unsubscribe@chromium.org.

[Martin Yan]

I reset tester, servo board and DUT, the result is still same. thanks.

[Martin Yan]

I modified

chroot/build/reef/usr/local/build/autotest/server/cros/faft/config/DEFAULTS.py

delay_reboot_to_ping = 60 /* original value is 30 */

but as I ran “firmware_UserRequestRecovery.ec_wp” again, the value was set back to 30 in this DEFAULTS.py file.

where can I modify this default value to take effect? somewhere mentions "refer to config object implementation", but couldn't find it.

[Tom Wai-Hong Tam (談偉航)]

Under chroot, you can change the ~/trunk/src/third_party/autotest/files/server/cros/faft/config

When running test_that, pass the argument: --autotest_dir ~/trunk/src/third_party/autotest/files

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+unsubscribe@chromium.org.

[Martin Yan]

Modify delay_reboot_to_ping = 90, and rerun “firmware_UserRequestRecovery.ec_wp” job, test still fails. Per following log information,

 

11:00:44 INFO | autoserv| Set/get all: ['warm_reset:on', 'sleep:0.5000', 'warm_reset:off']

11:00:45 INFO | autoserv| -[ModeSwitcher]-[ end simple_reboot('warm') ]-

11:00:45 INFO | autoserv| -[bypass_rec_mode]-

11:01:38 INFO | autoserv| Enabled dut_sees_usb

 

// Note: delay 90s, but still ping time out, because during this 90s delay, DUT is stuck and displays “Chrome OS is missing or damaged, Please remove all connected devices and start recovery”. DUT doesn’t boot up to cause “ping time out” next;

 

11:03:10 INFO | autoserv| ping timed out, try REC_ON

11:03:10 INFO | autoserv| Setting power_state to rec

11:03:26 INFO | autoserv| -[FAFT]-[ start wait_for_client ]---

 

 

Any reason to cause “[ end simple_reboot('warm') ]-” not to bring up, and DUT displays “Chrome OS is missing or damaged, Please remove all connected devices and start recovery”?

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+...@chromium.org.

[Tom Wai-Hong Tam (談偉航)]

There are some recent changes on servod, related to the power_state to rec.

Just want to confirm if you get it updated.

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+unsubscribe@chromium.org.

[Martin Yan]

Under chroot to update servod,

~/trunk/src/scripts $ sudo emerge hdctools

~/trunk/src/scripts $ sudo emerge openocd

 

Test is still failure.

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+...@chromium.org.

[Tom Wai-Hong Tam (談偉航)]

Looks like it is a bug at the recent servod changes.

You may cherry-pick the following patch to fix it (or simply

wait the patch merged and repo sync again).

  https://chromium-review.googlesource.com/830253/

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+unsubscribe@chromium.org.

[Martin Yan]

Has the patch been merged into googlesource? such that I can re-sync.

To unsubscribe from this group and stop receiving emails from it, send an email to faft-users+...@chromium.org.

[SEAN_FS...@wistron.com]

It was merged, I can see it in ToT.

--------------------------------------------------------------------------------------------------------------------------------------------------------------- This email contains confidential or legally privileged information and is for the sole use of its intended recipient. Any unauthorized review, use, copying or distribution of this email or the content of this email is strictly prohibited. If you are not the intended recipient, you may reply to the sender and should delete this e-mail immediately. ---------------------------------------------------------------------------------------------------------------------------------------------------------------