2023-02-21 10:16:27

by Vivek Pernamitta

[permalink] [raw]
Subject: [PATCH V1] bus: mhi: host: Avoid ringing EV DB if there is no elements to process

Avoid ringing Event DB if there is no elements to process.
As mhi_poll function can be called by mhi client drivers
which will call process_event, which will ring DB even if
there no ring elements to process.

Signed-off-by: Vivek Pernamitta <[email protected]>
---
drivers/bus/mhi/host/main.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/mhi/host/main.c b/drivers/bus/mhi/host/main.c
index df0fbfe..8bf8394 100644
--- a/drivers/bus/mhi/host/main.c
+++ b/drivers/bus/mhi/host/main.c
@@ -1031,7 +1031,9 @@ int mhi_process_data_event_ring(struct mhi_controller *mhi_cntrl,
count++;
}
read_lock_bh(&mhi_cntrl->pm_lock);
- if (likely(MHI_DB_ACCESS_VALID(mhi_cntrl)))
+
+ /* Ring EV DB only if there is any pending element to process */
+ if (likely(MHI_DB_ACCESS_VALID(mhi_cntrl)) && count)
mhi_ring_er_db(mhi_event);
read_unlock_bh(&mhi_cntrl->pm_lock);

--
2.7.4



2023-02-21 15:53:49

by Jeffrey Hugo

[permalink] [raw]
Subject: Re: [PATCH V1] bus: mhi: host: Avoid ringing EV DB if there is no elements to process

On 2/21/2023 3:15 AM, Vivek Pernamitta wrote:
> Avoid ringing Event DB if there is no elements to process.
> As mhi_poll function can be called by mhi client drivers
> which will call process_event, which will ring DB even if
> there no ring elements to process.
>
> Signed-off-by: Vivek Pernamitta <[email protected]>
> ---
> drivers/bus/mhi/host/main.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/bus/mhi/host/main.c b/drivers/bus/mhi/host/main.c
> index df0fbfe..8bf8394 100644
> --- a/drivers/bus/mhi/host/main.c
> +++ b/drivers/bus/mhi/host/main.c
> @@ -1031,7 +1031,9 @@ int mhi_process_data_event_ring(struct mhi_controller *mhi_cntrl,
> count++;
> }
> read_lock_bh(&mhi_cntrl->pm_lock);
> - if (likely(MHI_DB_ACCESS_VALID(mhi_cntrl)))
> +
> + /* Ring EV DB only if there is any pending element to process */
> + if (likely(MHI_DB_ACCESS_VALID(mhi_cntrl)) && count)
> mhi_ring_er_db(mhi_event);
> read_unlock_bh(&mhi_cntrl->pm_lock);
>

Seems like this optimization also applies to mhi_process_ctrl_ev_ring().
Why not update that as well?