linux-kernel - Re: [PATCH v2] bus: mhi: host: don't free bhie tables during suspend/hibernation

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <5dcf3ddf-6b0a-4b4c-b4a3-6e015a7e2f53@collabora.com>
Date: Fri, 18 Apr 2025 13:10:50 +0500
From: Muhammad Usama Anjum <usama.anjum@...labora.com>
To: Jeff Hugo <jeff.hugo@....qualcomm.com>,
 Krishna Chaitanya Chundru <quic_krichai@...cinc.com>,
 Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>,
 Johannes Berg <johannes@...solutions.net>, Jeff Johnson
 <jjohnson@...nel.org>, Jeffrey Hugo <quic_jhugo@...cinc.com>,
 Yan Zhen <yanzhen@...o.com>, Youssef Samir <quic_yabdulra@...cinc.com>,
 Qiang Yu <quic_qianyu@...cinc.com>, Alex Elder <elder@...nel.org>,
 Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
 Kunwu Chan <chentao@...inos.cn>
Cc: kernel@...labora.com, mhi@...ts.linux.dev, linux-arm-msm@...r.kernel.org,
 linux-kernel@...r.kernel.org, linux-wireless@...r.kernel.org,
 ath11k@...ts.infradead.org
Subject: Re: [PATCH v2] bus: mhi: host: don't free bhie tables during
 suspend/hibernation

On 4/14/25 7:14 PM, Jeff Hugo wrote:
> On 4/14/2025 1:32 AM, Muhammad Usama Anjum wrote:
>> On 4/12/25 6:22 AM, Krishna Chaitanya Chundru wrote:
>>>
>>> On 4/12/2025 12:02 AM, Muhammad Usama Anjum wrote:
>>>> On 4/11/25 1:39 PM, Krishna Chaitanya Chundru wrote:
>>>>>
>>>>>
>>>>> On 4/11/2025 12:32 PM, Muhammad Usama Anjum wrote:
>>>>>> On 4/11/25 8:37 AM, Krishna Chaitanya Chundru wrote:
>>>>>>>
>>>>>>>
>>>>>>> On 4/10/2025 8:26 PM, Muhammad Usama Anjum wrote:
>>>>>>>> Fix dma_direct_alloc() failure at resume time during bhie_table
>>>>>>>> allocation. There is a crash report where at resume time, the
>>>>>>>> memory
>>>>>>>> from the dma doesn't get allocated and MHI fails to re-initialize.
>>>>>>>> There may be fragmentation of some kind which fails the allocation
>>>>>>>> call.
>>>>>>>>
>>>>>>>> To fix it, don't free the memory at power down during suspend /
>>>>>>>> hibernation. Instead, use the same allocated memory again after
>>>>>>>> every
>>>>>>>> resume / hibernation. This patch has been tested with resume and
>>>>>>>> hibernation both.
>>>>>>>>
>>>>>>>> The rddm is of constant size for a given hardware. While the
>>>>>>>> fbc_image
>>>>>>>> size depends on the firmware. If the firmware changes, we'll
>>>>>>>> free and
>>>>>>> If firmware image will change between suspend and resume ?
>>>>>> Yes, correct.
>>>>>>
>>>>> why the firmware image size will change between suspend & resume?
>>>>> who will update the firmware image after bootup?
>>>>> It is not expected behaviour.
>>>> I was trying to research if the firmware can change or not. I've not
>>>> found any documentation on it.
>>>>
>>>> If the firmare is updated in filesystem before suspend/hibernate, would
>>>> the new firwmare be loaded the next time kernel resumes as the older
>>>> firmware is no where to be found?
>>>>
>>>> What do you think about this?
>>>>
>>> I don't think firmware can be updated before suspend/hibernate. I don't
>>> see any reason why it can be updated. If you think it can be updated
>>> please quote relevant doc.
>> I've not found any documentation on it. Let's wait for others to review
>> and it it cannot be updated, I'll remove this part.
>>
> 
> Wouldn't this be trivial to test?  Boot the device, go modify the
> firmware on the filesystem, then go through a suspend cycle.
I just tested this. I've used an old firmware from last year vs the
latest one.

Firmware A: old firmware size: 5349376
Firmware B: new firmware size: 5165056

A here has bigger size.

1. I loaded A at boot and then replaced the firmwares in filesystem with
B before syspend. At resume time, B was loaded fine by freeing the
bigger memory area and allocating the smaller one.

2. I loaded B and then replaced A in its place before suspend. At resume
time, memory was freed and larger memory was allocated. But driver
wasn't able to initialize correctly:

[  184.051902] ath11k_pci 0000:03:00.0: timeout while waiting for
restart complete
[  184.051916] ath11k_pci 0000:03:00.0: failed to resume core: -110
[  184.051923] ath11k_pci 0000:03:00.0: PM: dpm_run_callback():
pci_pm_resume returns -110
[  184.051945] ath11k_pci 0000:03:00.0: PM: failed to resume async:
error -110
[  187.251911] ath11k_pci 0000:03:00.0: wmi command 16387 timeout
[  187.251924] ath11k_pci 0000:03:00.0: failed to send
WMI_PDEV_SET_PARAM cmd
[  187.251933] ath11k_pci 0000:03:00.0: failed to enable dynamic bw: -11

So should we generalize above that changing firmware at
suspend/hibernation time isn't supported. If firmware package is
updated, does user restarts every time?

Attaching the dirty logs for reference. There are a lot of debug logs in it.

-- 
Regards,
Usama
View attachment "log" of type "text/plain" (138525 bytes)