Message-ID: <2680bb04-0a58-1914-44a4-a51d5aebb8f1@linux.intel.com>
Date:   Thu, 9 Feb 2023 09:33:08 -0600
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Firefox/102.0 Thunderbird/102.7.1
Subject: Re: [PATCH 5/5] soundwire: qcom: set clk stop need reset flag at
 runtime
Content-Language: en-US
To:     Srinivas Kandagatla <srinivas.kandagatla@linaro.org>,
        vkoul@kernel.org
Cc:     yung-chuan.liao@linux.intel.com, alsa-devel@alsa-project.org,
        linux-kernel@vger.kernel.org, steev@kali.org,
        johan+linaro@kernel.org, quic_bjorande@quicinc.com
References: <20230209131336.18252-1-srinivas.kandagatla@linaro.org>
 <20230209131336.18252-6-srinivas.kandagatla@linaro.org>
From:   Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
In-Reply-To: <20230209131336.18252-6-srinivas.kandagatla@linaro.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 2/9/23 07:13, Srinivas Kandagatla wrote:
> WSA Soundwire controller needs an full reset if clock stop support
> is not available in slave devices. WSA881x does not support clock stop
> however WSA883x supports clock stop.
> 
> Make setting this flag at runtime to address above issue.
> 
> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
> ---
>  drivers/soundwire/qcom.c | 13 ++++---------
>  1 file changed, 4 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/soundwire/qcom.c b/drivers/soundwire/qcom.c
> index 74e38c0d651b..0224a5a866de 100644
> --- a/drivers/soundwire/qcom.c
> +++ b/drivers/soundwire/qcom.c
> @@ -536,10 +536,14 @@ static int qcom_swrm_enumerate(struct sdw_bus *bus)
>  
>  		sdw_extract_slave_id(bus, addr, &id);
>  		found = false;
> +		ctrl->clock_stop_not_supported = false;
>  		/* Now compare with entries */
>  		list_for_each_entry_safe(slave, _s, &bus->slaves, node) {
>  			if (sdw_compare_devid(slave, id) == 0) {
>  				qcom_swrm_set_slave_dev_num(bus, slave, i);
> +				if (!slave->prop.simple_clk_stop_capable)
> +					ctrl->clock_stop_not_supported = true;

IIRC the 'simple_clk_stop_capable' for a peripheral refers to the
Simplified_SCSP_SM

see Figure 35 "Slave Clock Stop Prepare State Machine (SCSP_SM)"

In addition, there's a requirement that all peripherals shall support
ClockStopMode0. Only ClockStopMode1 is optional, and that case is
handled with a different property:

 * @clk_stop_mode1: Clock-Stop Mode 1 is supported

I think you overloaded one concept with another, or used the wrong property?

> +
>  				found = true;
>  				break;
>  			}
> @@ -1500,15 +1504,6 @@ static int qcom_swrm_probe(struct platform_device *pdev)
>  		 (ctrl->version >> 24) & 0xff, (ctrl->version >> 16) & 0xff,
>  		 ctrl->version & 0xffff);
>  
> -	/* Clk stop is not supported on WSA Soundwire masters */
> -	if (ctrl->version <= 0x01030000) {
> -		ctrl->clock_stop_not_supported = true;
> -	} else {
> -		ctrl->reg_read(ctrl, SWRM_COMP_MASTER_ID, &val);
> -		if (val == MASTER_ID_WSA)
> -			ctrl->clock_stop_not_supported = true;
> -	}
> -
>  #ifdef CONFIG_DEBUG_FS
>  	ctrl->debugfs = debugfs_create_dir("qualcomm-sdw", ctrl->bus.debugfs);
>  	debugfs_create_file("qualcomm-registers", 0400, ctrl->debugfs, ctrl,