[PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[Vishnu Sankar]

On systems where the USCS ACPI method is absent, acpi_evalf() returns
AE_NOT_FOUND which maps to -EIO. This caused tpacpi_usbc_security_init()
to propagate the error and thinkpad_acpi failed to probe entirely on
unsupported platforms.

Fix this by checking for USCS method presence with acpi_has_method() before
attempting to call it, returning -ENODEV immediately if absent.
This follows the same pattern used by other subdrivers in thinkpad_acpi.c.

Fixes: 67e8d1e9cacd ("platform/x86: thinkpad_acpi: Add USB-C Security (USCS) support")
Reported-by: Oliver Lin <oliver@email.com>
Closes: https://lore.kernel.org/platform-driver-x86/239c8162-e1e6-4b49-8292-35547c5a525c@liuxiaozhen.dev/
Signed-off-by: Vishnu Sankar <vishnuocv@gmail.com>
---
 drivers/platform/x86/lenovo/thinkpad_acpi.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/platform/x86/lenovo/thinkpad_acpi.c b/drivers/platform/x86/lenovo/thinkpad_acpi.c
index 91db4c8eeac2..85ed506516f0 100644
--- a/drivers/platform/x86/lenovo/thinkpad_acpi.c
+++ b/drivers/platform/x86/lenovo/thinkpad_acpi.c
@@ -11368,6 +11368,9 @@ static int tpacpi_usbc_security_init(struct ibm_init_struct *iibm)
 {
 	int err;
 
+	if (!acpi_has_method(hkey_handle, "USCS"))
+		return -ENODEV;
+
 	err = usbc_security_query(&tp_features.usbc_security_enabled);
 	if (err == -ENODEV)
 		return 0;
-- 
2.51.0

Re: [PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[Oliver Lin]

Hello Vishnu,

It looks like my email address was recorded incorrectly. My address is:

     oliver@liuxiaozhen.dev

I'm currently testing this patch and will report back once the testing 
is complete.

Thanks,
Oliver Lin

Re: [PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[Oliver Lin]

Hi Vishnu,

I tested the following two commits:

  platform/x86: thinkpad_acpi: Add USB-C Security (USCS) support
  platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on 
unsupported platforms

The tests were performed with Linux 7.2.0-rc2 on:

Lenovo ThinkPad P15 Gen 1
BIOS N30ET47W 1.30, EC N30HT33W
Lenovo ThinkPad P50s
Machine type 20FLCTO1WW

On the ThinkPad P15 Gen 1, the firmware WMI query still produces an
AE_AML_PACKAGE_LIMIT error while evaluating _SB.WMI1.WQA0:

ACPI BIOS Error (bug): AE_AML_PACKAGE_LIMIT, Index (0x2) is beyond
end of object (length 0x2)
ACPI Error: Aborting method _SB.WMI1.WQA0 due to previous error

However, with the probe-failure fix applied, thinkpad_acpi continues
initializing successfully. The battery extension, secondary fan control,
rfkill switch and ThinkPad extra buttons were all registered normally.

I also tested the USB-C data path with a Xiaomi Redmi Note 13 5G:

Initial device detection: PASS
Repeated unplug/replug: PASS
S3 deep suspend/resume: PASS
USB-C device detection after resume: PASS

The device was successfully enumerated again after resume. The
thinkpad_acpi module remained loaded and the kernel taint value was 0.

On the ThinkPad P50s, which does not expose the new USB-C Security
interface, thinkpad_acpi also initialized successfully. Existing
thinkpad_acpi functionality, including battery registration, rfkill,
hotkeys and suspend/resume, continued to work normally. This confirms
that an unsupported platform no longer causes the complete
thinkpad_acpi probe to fail.

I did observe existing errors on the P15 Gen 1 from the NVIDIA UCSI and
xHCI paths:

ucsi_ccg 2-0008: probe with driver ucsi_ccg failed with error -110
xhci_hcd 0000:01:00.2: xHC error in resume, USBSTS 0x401, Reinit

The test kernel was built without the NVIDIA display driver. These
messages did not prevent the ACPI UCSI Type-C ports from being created,
and did not prevent the tested USB-C device from working before or
after suspend.

Therefore, for this series:

Tested-by: Oliver Lin oliver@liuxiaozhen.dev

Thanks,
Oliver

Re: [PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[Ilpo Järvinen]

[-- Attachment #1: Type: text/plain, Size: 459 bytes --]

On Sun, 12 Jul 2026, Oliver Lin wrote:

> Hello Vishnu,
> 
> It looks like my email address was recorded incorrectly. My address is:
> 
>      oliver@liuxiaozhen.dev
> 
> I'm currently testing this patch and will report back once the testing 
> is complete.

Hi Vishnu,

To make things easier for me (and me less likely to forget while 
applying), could you please send v2 with corrected email + the tested by 
from the other email.

-- 
 i.

Re: [PATCH] platform/x86: thinkpad_acpi: Fix USB-C Security probe failure on unsupported platforms

[Vishnu Sankar]

Hi Oliver,

Thanks for the quick follow‑up.
Sorry for the wrong email address.

Thanks a lot for the testing.
The other errors you encounter seem to be coming from the think-lmi driver.
I will note this down and discuss it with Mark internally.

Ilpo,
Thanks for the quick follow up.
Sure, I’ll send a v2 of the patch with Oliver’s corrected email
address and include his Tested-by tag from the follow-up message. I’ll
get that out shortly.

Thank you.

Vishnu


On Mon, Jul 13, 2026 at 8:17 PM Ilpo Järvinen
<ilpo.jarvinen@linux.intel.com> wrote:
>
> On Sun, 12 Jul 2026, Oliver Lin wrote:
>
> > Hello Vishnu,
> >
> > It looks like my email address was recorded incorrectly. My address is:
> >
> >      oliver@liuxiaozhen.dev
> >
> > I'm currently testing this patch and will report back once the testing
> > is complete.
>
> Hi Vishnu,
>
> To make things easier for me (and me less likely to forget while
> applying), could you please send v2 with corrected email + the tested by
> from the other email.
>
> --
>  i.



-- 

Regards,

      Vishnu Sankar