linux-kernel - Re: [PATCH v2 3/5] bus: mhi: host: pci_generic: Remove MHI driver and ensure graceful device recovery

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-prev] [day] [month] [year] [list]

Message-ID: <ba37a9c9-3fd3-46fc-8d72-e57be45fc562@oss.qualcomm.com>
Date: Wed, 23 Jul 2025 14:51:06 +0200
From: Konrad Dybcio <konrad.dybcio@....qualcomm.com>
To: Vivek Pernamitta <quic_vpernami@...cinc.com>, Vivek.Pernamitta@...cinc.com,
        Manivannan Sadhasivam <mani@...nel.org>
Cc: mhi@...ts.linux.dev, linux-arm-msm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/5] bus: mhi: host: pci_generic: Remove MHI driver and
 ensure graceful device recovery

On 7/23/25 2:11 PM, Vivek Pernamitta wrote:
> 
> 
> On 7/10/2025 6:12 PM, Konrad Dybcio wrote:
>> On 7/10/25 10:58 AM, Vivek.Pernamitta@...cinc.com wrote:
>>> From: Vivek Pernamitta <quic_vpernami@...cinc.com>
>>>
>>> When the MHI driver is removed from the host side, it is crucial to ensure
>>> graceful recovery of the device. To achieve this, the host driver will
>>> perform the following steps:
>>>
>>> 1. Disable SRIOV for any SRIOV-enabled devices on the Physical Function.
>>> 2. Perform a SOC_RESET on Physical Function (PF).
>>>
>>> Disabling SRIOV ensures that all virtual functions are properly shut down,
>>> preventing any potential issues during the reset process. Performing
>>> SOC_RESET on each physical function guarantees that the device is fully
>>> reset and ready for subsequent operations.
>>>
>>> Signed-off-by: Vivek Pernamitta <quic_vpernami@...cinc.com>
>>> ---

[...]

>> It seems rather unlikely that out off all MHI devices, only QDU100
>> needs this quirk when working under SR-IOV
> The reset_on_driver_unbind flag has been explicitly added for the QDU100
> device. While other devices might enter RAMDUMP mode and wait when a SOC
> reset is issued, the QDU100 device will pass through without entering
> RAMDUMP mode

Rather inconveniently, this is not something that you mentioned in
the commit message.. Especially in the middle of a series that focuses
on enabling SR-IOV which suggests it's strictly related to virtualization,
making it difficult or impossible to understand the problem that this
patch is actually solving.

Please rewrite the commit message.

Konrad