Re: [PATCH] bus: mhi: core: Move to polling method to wait for MHI ready

[Jeffrey Hugo <jhugo@codeaurora.org>]

On 2/18/2021 7:18 PM, Bhaumik Bhatt wrote:
> In certain devices, it is likely that there is no incoming MHI
> interrupt for a transition to MHI READY state. One such example
> is the move from Pass Through to an SBL or AMSS execution
> environment. In order to facilitate faster bootup times as there
> is no need to wait until timeout_ms completes, MHI host can poll
> every 25 milliseconds to check if device has entered MHI READY
> until a maximum timeout of twice the timeout_ms is reached.
> 
> Signed-off-by: Bhaumik Bhatt <bbhatt@codeaurora.org>
> ---
>   drivers/bus/mhi/core/pm.c | 32 ++++++++++++++++----------------
>   1 file changed, 16 insertions(+), 16 deletions(-)
> 
> diff --git a/drivers/bus/mhi/core/pm.c b/drivers/bus/mhi/core/pm.c
> index 65ebca8..ec0060c 100644
> --- a/drivers/bus/mhi/core/pm.c
> +++ b/drivers/bus/mhi/core/pm.c
> @@ -9,6 +9,7 @@
>   #include <linux/dma-direction.h>
>   #include <linux/dma-mapping.h>
>   #include <linux/interrupt.h>
> +#include <linux/iopoll.h>
>   #include <linux/list.h>
>   #include <linux/mhi.h>
>   #include <linux/module.h>
> @@ -157,30 +158,29 @@ int mhi_ready_state_transition(struct mhi_controller *mhi_cntrl)
>   	struct mhi_event *mhi_event;
>   	enum mhi_pm_state cur_state;
>   	struct device *dev = &mhi_cntrl->mhi_dev->dev;
> -	u32 reset = 1, ready = 0;
> +	u32 reset, ready;
>   	int ret, i;
>   
> -	/* Wait for RESET to be cleared and READY bit to be set by the device */
> -	wait_event_timeout(mhi_cntrl->state_event,
> -			   MHI_PM_IN_FATAL_STATE(mhi_cntrl->pm_state) ||
> -			   mhi_read_reg_field(mhi_cntrl, base, MHICTRL,
> -					      MHICTRL_RESET_MASK,
> -					      MHICTRL_RESET_SHIFT, &reset) ||
> -			   mhi_read_reg_field(mhi_cntrl, base, MHISTATUS,
> -					      MHISTATUS_READY_MASK,
> -					      MHISTATUS_READY_SHIFT, &ready) ||
> -			   (!reset && ready),
> -			   msecs_to_jiffies(mhi_cntrl->timeout_ms));
> -
>   	/* Check if device entered error state */
>   	if (MHI_PM_IN_FATAL_STATE(mhi_cntrl->pm_state)) {
>   		dev_err(dev, "Device link is not accessible\n");
>   		return -EIO;
>   	}
>   
> -	/* Timeout if device did not transition to ready state */
> -	if (reset || !ready) {
> -		dev_err(dev, "Device Ready timeout\n");
> +	/* Wait for RESET to be cleared and READY bit to be set by the device */
> +	ret = readl_relaxed_poll_timeout(base + MHICTRL, reset,
> +					 !(reset & MHICTRL_RESET_MASK), 25000,
> +					 mhi_cntrl->timeout_ms * 1000);
> +	if (ret) {
> +		dev_err(dev, "Device failed to clear MHI Reset\n");
> +		return -ETIMEDOUT;
> +	}
> +
> +	ret = readl_relaxed_poll_timeout(base + MHISTATUS, ready,
> +					 (ready & MHISTATUS_READY_MASK), 25000,
> +					 mhi_cntrl->timeout_ms * 1000);
> +	if (ret) {
> +		dev_err(dev, "Device failed to enter MHI Ready\n");
>   		return -ETIMEDOUT;
>   	}
>   
> 

I think this gets a NACK from me.

mhi_read_reg_field() uses the controller defined accessor method 
read_reg().  readl_relaxed_poll_timeout() does not.  If, say, MHI was 
implemented over I2C, direct readl() accesses wouldn't be valid, and 
this would break.

Also, if the link ever goes down in the middle if the timeout, the 
controller read_reg() should detect that, but 
readl_relaxed_poll_timeout() would not.

I'm fine with this concept, but I think the implementation needs to be 
adjusted.

-- 
Jeffrey Hugo
Qualcomm Technologies, Inc. is a member of the
Code Aurora Forum, a Linux Foundation Collaborative Project.