netdev - Re: [PATCH net-next] net: ipa: use dev PM wakeirq handling

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: <42080e79-e7c6-0e2b-5e8c-95aedb5823a0@linaro.org>
Date:   Tue, 31 Jan 2023 07:04:57 -0600
From:   Alex Elder <elder@...aro.org>
To:     Caleb Connolly <caleb.connolly@...aro.org>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>
Cc:     elder@...nel.org, netdev@...r.kernel.org,
        linux-arm-msm@...r.kernel.org, phone-devel@...r.kernel.org
Subject: Re: [PATCH net-next] net: ipa: use dev PM wakeirq handling

On 1/31/23 6:40 AM, Caleb Connolly wrote:
> 
> 
> On 1/28/23 13:47, Alex Elder wrote:
>> On 1/27/23 2:27 PM, Caleb Connolly wrote:
>>> Replace the enable_irq_wake() call with one to dev_pm_set_wake_irq()
>>> instead. This will let the dev PM framework automatically manage the
>>> the wakeup capability of the ipa IRQ and ensure that userspace requests
>>> to enable/disable wakeup for the IPA via sysfs are respected.
>>>
>>> Signed-off-by: Caleb Connolly <caleb.connolly@...aro.org>
>>
>> Looks OK to me.  Can you say something about how you
>> tested this, and what the result was?  Thanks.
> 
> Ah yeah. This was tested on the SDM845 (IPA 3.5.1) based SHIFT6mq in the 
> UK with an EE SIM card.
> 
> All network connections were disabled except for mobile data which was 
> configured using ModemManager. Then I set up a basic TCP server using 
> netcat on a public IP address and connected to it from the device.
> 
> It is then possible to validate that the wakeirq fires and the interrupt 
> is handled correctly by putting the device into s2idle sleep (echo mem > 
> /sys/power/state) and typing some data into the server terminal.
> 
> Then I disabled the wakeup as follows and repeated the test to ensure 
> that the device would no longer wake up on incoming data, and that the 
> data was received when the device resumes.
> 
> echo disabled > /sys/devices/platform/soc\@0/1e40000.ipa/power/wakeup

Great explanation, thank you.  This looks good.

Reviewed-by: Alex Elder <elder@...aro.org>

>>
>>                      -Alex
>>
>>> ---
>>>   drivers/net/ipa/ipa_interrupt.c | 10 ++++------
>>>   1 file changed, 4 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/drivers/net/ipa/ipa_interrupt.c 
>>> b/drivers/net/ipa/ipa_interrupt.c
>>> index c19cd27ac852..9a1153e80a3a 100644
>>> --- a/drivers/net/ipa/ipa_interrupt.c
>>> +++ b/drivers/net/ipa/ipa_interrupt.c
>>> @@ -22,6 +22,7 @@
>>>   #include <linux/types.h>
>>>   #include <linux/interrupt.h>
>>>   #include <linux/pm_runtime.h>
>>> +#include <linux/pm_wakeirq.h>
>>>   #include "ipa.h"
>>>   #include "ipa_reg.h"
>>> @@ -269,9 +270,9 @@ struct ipa_interrupt *ipa_interrupt_config(struct 
>>> ipa *ipa)
>>>           goto err_kfree;
>>>       }
>>> -    ret = enable_irq_wake(irq);
>>> +    ret = dev_pm_set_wake_irq(dev, irq);
>>>       if (ret) {
>>> -        dev_err(dev, "error %d enabling wakeup for \"ipa\" IRQ\n", 
>>> ret);
>>> +        dev_err(dev, "error %d registering \"ipa\" IRQ as 
>>> wakeirq\n", ret);
>>>           goto err_free_irq;
>>>       }
>>> @@ -289,11 +290,8 @@ struct ipa_interrupt 
>>> *ipa_interrupt_config(struct ipa *ipa)
>>>   void ipa_interrupt_deconfig(struct ipa_interrupt *interrupt)
>>>   {
>>>       struct device *dev = &interrupt->ipa->pdev->dev;
>>> -    int ret;
>>> -    ret = disable_irq_wake(interrupt->irq);
>>> -    if (ret)
>>> -        dev_err(dev, "error %d disabling \"ipa\" IRQ wakeup\n", ret);
>>> +    dev_pm_clear_wake_irq(dev);
>>>       free_irq(interrupt->irq, interrupt);
>>>       kfree(interrupt);
>>>   }
>>
>