boards/chromebooks: Update board status for coral, grunt, rammus

Signed-off-by: Ricardo Cañuelo <ricardo.canuelo@collabora.com>

content/boards/chromebooks/boards/common_issues.md

+---
+title: Common issues
+---
+
+This is a list of the issues we found that were common for all the
+Chromebooks tested in the lab.
+
+
+## Issues related to Servo v4
+
+The Servo v4 debugging interface causes different issues depending on
+the Chromebook board and the lab setup.
+
+### Network interface issues
+
+#### R8152 driver not reliable at Gigabit speeds
+
+The Realtek R8152 USB-Eth adapter can be found in the Servo v4 and in
+other standalone dongles.
+
+The R8152 Ethernet driver in Depthcharge is not reliable when working at
+Gigabit speeds. It's recommeded to configure the link as Fast Ethernet
+(100 Mbps) when booting over TFTP.
+
+
+#### Ethernet frame reception issues
+
+Ethernet frame reception may fail and hang Depthcharge when using the
+Servo v4 as an USB-Eth adapter (error: `R8152: Bulk read error`). Bigger
+file transfers make this more likely to happen, although it's not
+completely deterministic.
+
+
+### Stability issues
+
+#### Debugging interface unstable when using Servo v4 connected to power supply
+
+Observed in:
+
+- `coral`: R64-10068.111.0.serial
+- `rammus`: R72-11275.110.0.serial
+- `octopus`: R72-1197.144.0.serial
+
+with Servo v4 FW 2.3.22 and 2.4.35.
+
+When the Servo v4 is connected to a power supply the debugging
+interface becomes unstable. Some of the effects seen:
+
+  - Charging is unreliable.
+  - Serial consoles stop working. The following message appears in the
+    kernel log:
+
+```
+google ttyUSB4: usb_serial_generic_write_bulk_callback - urb stopped: -32
+```
+  - Ethernet data transfer through the Servo fails.
+  - Cr50 commands lost sometimes.
+
+Basically, all communication going through the Servo starts failing
+intermittently.

content/boards/chromebooks/boards/coral.md

@@ -18,13 +18,17 @@ and devices, but they are based on Intel Apollo Lake 64 bit CPUs
 The Collabora LAVA lab contains the following `coral` devices:
 
   - ASUS Chromebook C523NA
-    - `asus-C523NA-A20057-coral`
+    - [`asus-C523NA-A20057-coral-cbg-0`](https://lava.collabora.co.uk/scheduler/device/asus-C523NA-A20057-coral-cbg-0)
+    - [`asus-C523NA-A20057-coral-cbg-1`](https://lava.collabora.co.uk/scheduler/device/asus-C523NA-A20057-coral-cbg-1)
+    - [`asus-C523NA-A20057-coral-cbg-2`](https://lava.collabora.co.uk/scheduler/device/asus-C523NA-A20057-coral-cbg-2)
+    - [`asus-C523NA-A20057-coral-cbg-3`](https://lava.collabora.co.uk/scheduler/device/asus-C523NA-A20057-coral-cbg-3)
+
+
 
 ### Debugging interfaces
 
 `coral` boards have been flashed and tested with both [SuzyQ and Servo
-v4](../../01-debugging_interfaces) interfaces. No issues reported so
-far.
+v4](../../01-debugging_interfaces) interfaces.
 
 In an Asus C523N, the debug port is the USB-C port in the left side
 (also used for the power supply).
@@ -36,9 +40,12 @@ supported by Depthcharge (R8152).
 
 #### Known issues
 
-The R8152 Ethernet driver in Depthcharge doesn't seem reliable when
-working at Gigabit speeds. It's recommeded to configure the link as Fast
-Ethernet when booting over TFTP.
+  - [`coral-1`](https://lava.collabora.co.uk/scheduler/device/asus-C523NA-A20057-coral-cbg-1)
+    fails to boot in around 5% of the attempts. One of the Coreboot
+    stages resets the board and then it never comes out of reset (last
+    console message: `CSE timed out. Resetting`).
+
+See also [Common issues](../common_issues.md).
 
 ### Example kernel command line arguments
 

content/boards/chromebooks/boards/grunt.md

@@ -41,14 +41,26 @@ supported by Depthcharge (R8152).
 
 #### Known issues
 
-The R8152 Ethernet driver in Depthcharge doesn't seem reliable when
-working at Gigabit speeds. It's recommeded to configure the link as Fast
-Ethernet when booting over TFTP.
-
-Depending on the firmware version of the Servo v4 interface used, some
-issues may be present, such as the loss of the cr50 serial console when
-issuing a cold reset or Depthcharge hanging when receiving network
-packets.
+  - When using a Servo v4 as a debugging interface, the Cr50 interface
+    shuts down after issuing a `cold_reset:on` command. However, this
+    doesn't seem to happen if the Servo v4 is connected to a power
+    supply and charging the Chromebook.
+
+  - In some setups the serial console gets stuck and freezes during one
+    of the Coreboot stages. Closing and reopening the serial terminal
+    recovers the problem, since the Chromebook keeps running. Most of
+    the times this happens is at this point during the boot process:
+
+   ```
+   POST: 0x40
+   agesawrapper_amdinitpost() entry
+   DRAM clear on reset: Keep
+   variant_mainboard_read_spd SPD index 9
+   CBFS: 'Master Header Locator' located CBFS at [df0000:ffffc0)
+   CBFS: Locating 'spd.bin'
+   CBFS: Found @ offset 79bc0 size 2000
+   ```
+See also [Common issues](../common_issues.md).
 
 ### Example kernel command line arguments
 

content/boards/chromebooks/boards/rammus.md

@@ -27,9 +27,7 @@ supported by Depthcharge (R8152).
 
 #### Known issues
 
-The R8152 Ethernet driver in Depthcharge doesn't seem reliable when
-working at Gigabit speeds. It's recommeded to configure the link as Fast
-Ethernet when booting over TFTP.
+See also [Common issues](../common_issues.md).
 
 ### Example kernel command line arguments