On 04.10.24 11:12, Benjamin Tissoires wrote:
> On Oct 04 2024, Linux regression tracking (Thorsten Leemhuis) wrote:
>
> so cleanup is doing:
> cleanup:
> amd_sfh_hid_client_deinit(privdata);
> for (i = 0; i < cl_data->num_hid_devices; i++) {
> devm_kfree(dev, cl_data->feature_report[i]);
> devm_kfree(dev, in_data->input_report[i]);
> devm_kfree(dev, cl_data->report_descr[i]);
> }
> return rc;
>
> Would that means that the memory corruption appears during
> amd_sfh_hid_client_deinit(), or...
>
>>> }
>>>
>>> for (i = 0; i < cl_data->num_hid_devices; i++) {
>>> diff --git a/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c b/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c
>>> index 0c28ca349bcd..1300f122b524 100644
>>> --- a/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c
>>> +++ b/drivers/hid/amd-sfh-hid/amd_sfh_pcie.c
>>> @@ -351,7 +351,9 @@ static void sfh_init_work(struct work_struct *work)
>>>
>>> rc = amd_sfh_hid_client_init(mp2);
>>> if (rc) {
>>> - amd_sfh_clear_intr(mp2);
>>> + if (rc != -EOPNOTSUPP)
>>> + amd_sfh_clear_intr(mp2);
>
> ... or during amd_sfh_clear_intr()?
>
> This very much looks like a band-aid (I know it is because you can not
> reproduce, not blaming anyone), but I'd like to know a little bit more
> if the bug is not appearing anywhere else in the normal processing of
> the driver itself.
>
> Also a comment explaining why this is the only case where
> amd_sfh_clear_intr() should not be called would be appreciated.
>
> So all in all, I have a feeling one of these 2 functions is not making a
> proper check and I would rather fix the root cause, not the symptoms.
>
> Cheers,
> Benjamin
>
> PS: sorry, I know this is a long standing issue, but I'd rather not
> paper over a bigger issue :/
Yeah, me to!
There was a new patch in bugzilla and people tested it (see
https://bugzilla.kernel.org/show_bug.cgi?id=219331 for details) , but
nothing has happened now for 48 hours, so please allow me to ask:
Basavaraj Natikar, is there any hope that a proper fix will emerge soon?
Sorry, normally I would wait a lot longer before asking things like
this, but given the severeness of the problem I thought a inquiry for a
quick status update was in order.
Ciao, Thorsten
