linux-kernel - USB PD broken on Lenovo P15gen2

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 28 Aug 2023 17:44:54 +0300
From:   Nikolay Borisov <nik.borisov@...e.com>
To:     heikki.krogerus@...ux.intel.com,
        Greg KH <gregkh@...uxfoundation.org>,
        linux-usb@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>
Cc:     Anthony Iliopoulos <ailiopoulos@...e.com>
Subject: USB PD broken on Lenovo P15gen2

Hello,

I'm not able to use usb PD on a Lenovo Thinkpad P15gen2 laptop. It's 
equipped with 2 thunderbolt ports and a usb 3.2 gen2 usb port, all of 
which are supposed to support PD 2.0:

cat /sys/class/typec/port*/usb_power_delivery_revision
2.0
2.0
2.0

Whenever I connect a pd-capable device (i.e a mobile phone or a tablet) 
with a known good cable i.e i'm able to achieve fast charging (~30w) on 
the device when using the same cable but connected to an external 
monitor and not the laptop, I'm not able to get fast charge to work. In 
dmesg I get the following:

[ 1262.933106] usb 3-12: new high-speed USB device number 8 using xhci_hcd
[ 1263.083343] usb 3-12: New USB device found, idVendor=22d9, 
idProduct=2046, bcdDevice= 2.23
[ 1263.083350] usb 3-12: New USB device strings: Mfr=1, Product=2, 
SerialNumber=3
[ 1263.083353] usb 3-12: Product: IV2201
[ 1263.083355] usb 3-12: Manufacturer: OnePlus
[ 1263.083356] usb 3-12: SerialNumber: xxxxxxx
[ 1263.116184] ucsi_acpi USBC000:00: UCSI_GET_PDOS failed (-95)
[ 1268.676590] ucsi_acpi USBC000:00: UCSI_GET_PDOS failed (-95)
[ 1273.816117] ucsi_acpi USBC000:00: UCSI_GET_PDOS failed (-95)
[ 1273.904187] ucsi_acpi USBC000:00: UCSI_GET_PDOS failed (-95)

When looking at the state of the partner device:
$ cat /sys/class/typec/port2-partner/supports_usb_power_delivery
yes

$ cat /sys/class/typec/port2-partner/usb_power_delivery_revision
0.0

After some debugging it seems this is because the lenovo is not 
supporting CCI which seems to be some sort of a communication mechanism. 
This was asserted with the following bpftrace script:

bpftrace -e 'kretprobe:ucsi_acpi_read { printf("ucsi_acpi_read returned: 
%d\n", retval); } kretprobe:ucsi_exec_command { 
printf("ucsi_exec_command returned: %d\n", retval); }'

Attaching 2 probes...
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: 9
ucsi_acpi_read returned: 0
ucsi_exec_command returned: 9
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: 9
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: -95
ucsi_exec_command returned: -110
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: -95
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: -95
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: -95
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0
ucsi_exec_command returned: 0
ucsi_acpi_read returned: 0
ucsi_acpi_read returned: 0

This would correspond to the following snippet in ucsi_exec_command:

if (cci & UCSI_CCI_NOT_SUPPORTED)
             return -EOPNOTSUPP;

I've also attached an acpidump of my tables and the relevant usb one is 
ssdt7. And I'm also using the latest available firmware from Lenovo, 
installed via lvfs. This issue happens on kernel 6.2 based ubuntu kernel 
as well as on upstream 6.4.3 but it seems it's not dependent on the 
kernel version as it's somehow related to lenovo's firmware.

I'm wondering whether a similar quirk like the one for the zenbook is 
also required for lenovo ?
View attachment "acpi.dump" of type "text/plain" (1762077 bytes)