As mentioned in my new year's plan a motivation for my current project is having an End-of-life (EOL) Chromebook and wanting to bring it up to date.
Acer C720, as pulled from the helpful Developer Information for Chrome OS Devices
|Release||September 12, 2013|
|Model||Acer C720/C70P/C740 Chromebook|
|Code name||Peppy & Pepto|
|Closed Case Debugging||No|
Prior to power washing the device it showed:
76.0.3809.136 (Official Build) (64-bit)
12239.92.0 (Official Build)
- Build Date:
Tuesday, August 27, 2019
Linux localhost 3.8.11 #1 SMP Tue Aug 27 16:53:30 PDT 2019 x86_64 Intel(R) Celeron(R) 2955U @ 1.40GHz GenuineIntel GNU/Linux
Interestingly while power washing I was provided an option to “Update firmware for added security”:
Even more interesting was this did not result in a higher firmware version.
After the power wash and logging in as a guest it still reports
Google_Peppy.4389.117.0. To cap it all off a second power wash did not prompt
for a firmware update???
While looking for more information on this I found the Trusted Platform Module
firmware vulnerability: technical documentation (fortunately my machine was
properly patched. Helpfully though this referenced
chrome://system as a place
to get detailed firmware information. It has a ton of great information.
For example the
vendor | ti name | Unknown-10de fw_version | peppy_v1.5.133-9078375
This refers to the embedded controller, which how cool is it that all this information is publicly available.
tpm_family 312e3200 spec_level 0000000200000003 vendor 49465800 tpm_model ffffffff firmware_version 0000000000000422 vendor_specific 042203f20274706d3338ffffff version_fingerprint 487001745 1d070e91
0000000000000422 - 4.34 is fixed. Yay.
As a final note the
dmesg output had a few interesting lines. The first of
which is perhaps the aforementioned firmware update (though oddly it seem to get
applied every boot)?
[ 12.150597] hda-intel 0000:00:1b.0: Applying patch firmware 'peppy_alsa.fw'
The other of which, is probably an interesting rabbit hole:
[ 0.097482] chromeos: tpm attack callbacks: -19
Some quick references are probably Issue 473051: Unexpected error is seen while recovering on peppy and falco and the linked commit. As well as the remark
CHROMIUM: ignore errors when registering tpm attack callbacks.
This feature is only used on some platforms and only for logging purposes, so it is safe to let it not work.
Enough on the firmware rabbit hole, but one final tidbit I found interesting is that even to get to guest mode seemed to required network connectivity and accepting some Terms of Service. A quick search did not immediately reveal a way around this. If would be interesting to log this connection and see what, if any, traffic is actually sent.
On to Developer Mode
Consulting the Debug Button Shortcuts has all kinds of interesting key
combinations. The specific one we want is Recovery Mode (
Esc + Refresh and press Power) then Developer Mode (
Ctrl + D).
It really is that easy.
I have power washed chromebooks and worked in developer mode a number of times previously, but this all goes to show that you can always dig deeper and learn more, even doing tasks that seem straight forward.
I've actually gone ahead and build a chromiumos image, but I'm going to leave describing that (and testing it to another session).