Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <bbc432e7-c52f-7e35-03d1-401cb68501cb@linux.intel.com>
Date:   Mon, 14 Nov 2022 10:14:19 -0600
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Firefox/102.0 Thunderbird/102.4.2
From:   Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Subject: Re: [PATCH 3/4] soundwire: debugfs: Switch to sdw_read_no_pm
To:     Charles Keepax <ckeepax@opensource.cirrus.com>, vkoul@kernel.org
Cc:     yung-chuan.liao@linux.intel.com, sanyog.r.kale@intel.com,
        alsa-devel@alsa-project.org, linux-kernel@vger.kernel.org,
        patches@opensource.cirrus.com
References: <20221114102956.914468-1-ckeepax@opensource.cirrus.com>
 <20221114102956.914468-4-ckeepax@opensource.cirrus.com>
Content-Language: en-US
In-Reply-To: <20221114102956.914468-4-ckeepax@opensource.cirrus.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 11/14/22 04:29, Charles Keepax wrote:
> It is rather inefficient to be constantly enabling/disabling the PM
> runtime as we print out each individual register, switch to holding a PM
> runtime reference across the whole register output.

the change is good, but technically the pm_runtime resume happens for
the first read and suspend with a delay if use_autosuspend() is enabled,
so presumably we'll see the same number of resume/suspend with the
existing code and the suggested change.

Maybe update the commit message to mention that we constantly change
reference counts, as you did in the next patch?

> 
> Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
> ---
>  drivers/soundwire/debugfs.c | 11 ++++++++++-
>  1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/soundwire/debugfs.c b/drivers/soundwire/debugfs.c
> index 49900cd207bc7..0718e9cda138a 100644
> --- a/drivers/soundwire/debugfs.c
> +++ b/drivers/soundwire/debugfs.c
> @@ -4,6 +4,7 @@
>  #include <linux/device.h>
>  #include <linux/debugfs.h>
>  #include <linux/mod_devicetable.h>
> +#include <linux/pm_runtime.h>
>  #include <linux/slab.h>
>  #include <linux/soundwire/sdw.h>
>  #include <linux/soundwire/sdw_registers.h>
> @@ -35,7 +36,7 @@ static ssize_t sdw_sprintf(struct sdw_slave *slave,
>  {
>  	int value;
>  
> -	value = sdw_read(slave, reg);
> +	value = sdw_read_no_pm(slave, reg);
>  
>  	if (value < 0)
>  		return scnprintf(buf + pos, RD_BUF - pos, "%3x\tXX\n", reg);
> @@ -55,6 +56,10 @@ static int sdw_slave_reg_show(struct seq_file *s_file, void *data)
>  	if (!buf)
>  		return -ENOMEM;
>  
> +	ret = pm_runtime_resume_and_get(&slave->dev);
> +	if (ret < 0 && ret != -EACCES)
> +		return ret;
> +
>  	ret = scnprintf(buf, RD_BUF, "Register  Value\n");
>  
>  	/* DP0 non-banked registers */
> @@ -112,6 +117,10 @@ static int sdw_slave_reg_show(struct seq_file *s_file, void *data)
>  	}
>  
>  	seq_printf(s_file, "%s", buf);
> +
> +	pm_runtime_mark_last_busy(&slave->dev);
> +	pm_runtime_put(&slave->dev);
> +
>  	kfree(buf);
>  
>  	return 0;