Re: [PATCH] platform/x86: Fix NULL deref in amd_isp_probe

Mario Limonciello <[email protected]>
Newsgroups org.kernel.vger.platform-driver-x86,org.kernel.vger.linux-kernel
Message-ID <[email protected]>

On 8/10/26 16:52, Nirujogi, Pratap wrote:
> 
> 
> On 8/9/2026 11:24 PM, Xueqin Luo wrote:
>> [You don't often get email from [email protected]. Learn why this 
>> is important at https://aka.ms/LearnAboutSenderIdentification ]
>>
>> Caution: This message originated from an External Source. Use proper 
>> caution when opening attachments, clicking links, or responding.
>>
>>
>> amd_isp_probe() uses ACPI_COMPANION() without NULL check. Add it to
>> prevent crash when driver_override matches a device without ACPI
>> companion.
>>
>> Fixes: e1af6f0146d6 ("platform/x86: Update swnode graph for amd isp4")
>> Signed-off-by: Xueqin Luo <[email protected]>
>> ---
>>   drivers/platform/x86/amd/amd_isp4.c | 7 +++++++
>>   1 file changed, 7 insertions(+)
>>
>> diff --git a/drivers/platform/x86/amd/amd_isp4.c b/drivers/platform/ 
>> x86/amd/amd_isp4.c
>> index 0d494899502c..0026d6663c5c 100644
>> --- a/drivers/platform/x86/amd/amd_isp4.c
>> +++ b/drivers/platform/x86/amd/amd_isp4.c
>> @@ -379,6 +379,11 @@ static int amd_isp_probe(struct platform_device 
>> *pdev)
>>                  goto error_unregister_sw_node;
>>
>>          adev = ACPI_COMPANION(&pdev->dev);
>> +       if (!adev) {
>> +               ret = -ENODEV;
>> +               goto error_unregister_notifier;
>> +       }
>> +
>>          /* initialize root amd_camera_node */
>>          adev->driver_data = (void *)pinfo->swnodes[0];
>>
>> @@ -388,6 +393,8 @@ static int amd_isp_probe(struct platform_device 
>> *pdev)
>>          platform_set_drvdata(pdev, isp4_platform);
>>          return 0;
>>
>> +error_unregister_notifier:
>> +       bus_unregister_notifier(&i2c_bus_type, &isp4_platform->i2c_nb);
>>   error_unregister_sw_node:
>>          software_node_unregister_node_group(isp4_platform->pinfo- 
>> >swnodes);
>>          return ret;
> LGTM. I agree that adding the NULL check is good practice. That said, I 
> believe it's not strictly needed as the device_get_match_data() in 
> line#366 is expected to fail if there is no ACPI companion.

It sounds like it's essentially dead code then.

> 
> https://github.com/torvalds/linux/blob/master/drivers/platform/x86/amd/ 
> amd_isp4.c#L366
> 
> Reviewed-by: Pratap Nirujogi <[email protected]>
> 
>> -- 
>> 2.43.0
>>
>
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.