linux-kernel - Re: [PATCH v3 1/2] i2c: davinci: Add PM Runtime Support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <f185afb9-e877-1916-fcbc-c2a28f23b0cc@ti.com>
Date:   Mon, 11 Sep 2017 15:07:15 -0500
From:   Franklin S Cooper Jr <fcooper@...com>
To:     Sekhar Nori <nsekhar@...com>, <wsa@...-dreams.de>,
        <robh+dt@...nel.org>, <linux@...linux.org.uk>,
        <linux-i2c@...r.kernel.org>, <devicetree@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>,
        <linux-arm-kernel@...ts.infradead.org>, <grygorii.strashko@...com>,
        <vigneshr@...com>
Subject: Re: [PATCH v3 1/2] i2c: davinci: Add PM Runtime Support



On 09/11/2017 06:29 AM, Sekhar Nori wrote:
> On Friday 08 September 2017 11:24 PM, Franklin S Cooper Jr wrote:
>> 66AK2G has I2C instances that are not apart of the ALWAYS_ON power domain
>> like other Keystone 2 SoCs and OMAPL138. Therefore, pm_runtime
> 
> unlike ?
> 
>> is required to insure the power domain used by the specific I2C instance is
>> properly turned on along with its functional clock.
>>
>> Signed-off-by: Franklin S Cooper Jr <fcooper@...com>
>> ---
>> Version 3 changes:
>> Remove several statements that set clk to NULL
>> Fix error path
>>
>>  drivers/i2c/busses/i2c-davinci.c | 64 +++++++++++++++++++++++++++++++++-------
>>  1 file changed, 53 insertions(+), 11 deletions(-)
> 
>> @@ -802,12 +821,22 @@ static int davinci_i2c_probe(struct platform_device *pdev)
>>  	dev->clk = devm_clk_get(&pdev->dev, NULL);
>>  	if (IS_ERR(dev->clk))
>>  		return PTR_ERR(dev->clk);
>> -	clk_prepare_enable(dev->clk);
>>  
>>  	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>>  	dev->base = devm_ioremap_resource(&pdev->dev, mem);
>>  	if (IS_ERR(dev->base)) {
>> -		r = PTR_ERR(dev->base);
>> +		return PTR_ERR(dev->base);
>> +	}
>> +
>> +	pm_runtime_set_autosuspend_delay(dev->dev,
>> +					 DAVINCI_I2C_PM_TIMEOUT);
>> +	pm_runtime_use_autosuspend(dev->dev);
>> +
>> +	pm_runtime_enable(dev->dev);
>> +
>> +	r = pm_runtime_get_sync(dev->dev);
>> +	if (r < 0) {
>> +		dev_err(dev->dev, "failed to runtime_get device: %d\n", r);
>>  		goto err_unuse_clocks;
> 
> You end up doing a pm_runtime_put_sync() on failure here, instead of
> pm_runtime_put_noidle() like rest of the patch.
> 
> May be handle this failure here instead of relying on the goto path.

Ok
> 
>>  	}
>>  
>> @@ -849,27 +878,40 @@ static int davinci_i2c_probe(struct platform_device *pdev)
>>  	if (r)
>>  		goto err_unuse_clocks;
>>  
>> +	pm_runtime_mark_last_busy(dev->dev);
>> +	pm_runtime_put_autosuspend(dev->dev);
>> +
>>  	return 0;
>>  
>>  err_unuse_clocks:
>> -	clk_disable_unprepare(dev->clk);
>> -	dev->clk = NULL;
>> +	pm_runtime_dont_use_autosuspend(dev->dev);
>> +	pm_runtime_put_sync(dev->dev);
>> +	pm_runtime_disable(dev->dev);
>> +
>>  	return r;
>>  }
> 
> Thanks,
> Sekhar
>