Messages in this thread | | | Date | Thu, 22 Sep 2022 10:29:06 +0800 | Subject | Re: [PATCH 1/2] ACPI: PCC: replace wait_for_completion() | From | "lihuisong (C)" <> |
| |
在 2022/9/21 23:47, Sudeep Holla 写道: > On Tue, Sep 20, 2022 at 05:44:59PM +0800, Huisong Li wrote: >> Currently, the function waiting for completion of mailbox operation is >> 'wait_for_completion()'. The PCC method will be permanently blocked if >> this mailbox message fails to execute. So this patch replaces it with >> 'wait_for_completion_timeout()'. And set the timeout interval to an >> arbitrary retries on top of nominal to prevent the remote processor is >> slow to respond to PCC commands. >> > Sounds good to me. The only concern(may be not serious) is what happens > if we receive response from the platform after the timeout. I have tested If OS still cann't receive response in noramal latency(must be filled accurately as protocol said) + retries, there is a high probability that an exception occurs. Even if we receive response after the timeout, I think there may be no impact, but the response data in PCC share memory is ignored. > for that in non ACPI non PCC context. I don't have a setup to trigger that > with ACPI PCC + this patch to test. Other than that, I am fine with this: > > Reviewed-by: Sudeep Holla <sudeep.holla@arm.com> >
| |