Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Content-Type: text/plain;
        charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 13.0 \(3608.60.0.2.5\))
Subject: Re: [PATCH] i2c: nvidia-gpu: Handle timeout correctly in
 gpu_i2c_check_status()
From:   Kai-Heng Feng <kai.heng.feng@canonical.com>
In-Reply-To: <BYAPR12MB2727B1862E56594E4B6A3E0ADCF00@BYAPR12MB2727.namprd12.prod.outlook.com>
Date:   Tue, 24 Mar 2020 01:20:31 +0800
Cc:     Wolfram Sang <wsa@the-dreams.de>,
        Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
        "open list:I2C CONTROLLER DRIVER FOR NVIDIA GPU" 
        <linux-i2c@vger.kernel.org>,
        open list <linux-kernel@vger.kernel.org>
Content-Transfer-Encoding: 8BIT
Message-Id: <E6C41EAE-C56D-4FAB-969F-0842F4FF5C18@canonical.com>
References: <20200311165806.12365-1-kai.heng.feng@canonical.com>
 <DA56E103-CAA7-4C28-9528-9EAFF2C717CD@canonical.com>
 <BYAPR12MB2727B1862E56594E4B6A3E0ADCF00@BYAPR12MB2727.namprd12.prod.outlook.com>
To:     Ajay Gupta <ajayg@nvidia.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Ajay,

> On Mar 24, 2020, at 00:47, Ajay Gupta <ajayg@nvidia.com> wrote:
> 
> Kai-Heng
> 
>> -----Original Message-----
>> From: Kai-Heng Feng <kai.heng.feng@canonical.com>
>> Sent: Sunday, March 22, 2020 10:38 PM
>> To: Ajay Gupta <ajayg@nvidia.com>
>> Cc: Wolfram Sang <wsa@the-dreams.de>; Andy Shevchenko
>> <andriy.shevchenko@linux.intel.com>; open list:I2C CONTROLLER DRIVER FOR
>> NVIDIA GPU <linux-i2c@vger.kernel.org>; open list <linux-
>> kernel@vger.kernel.org>
>> Subject: Re: [PATCH] i2c: nvidia-gpu: Handle timeout correctly in
>> gpu_i2c_check_status()
>> 
>> External email: Use caution opening links or attachments
>> 
>> 
>>> On Mar 12, 2020, at 00:58, Kai-Heng Feng <kai.heng.feng@canonical.com>
>> wrote:
>>> 
>>> Nvidia card may come with a "phantom" UCSI device, and its driver gets
>>> stuck in probe routine, prevents any system PM operations like suspend.
>>> 
>>> Let's handle the unaccounted case that the target time equals to
>>> jiffies in gpu_i2c_check_status(), so the UCSI driver can let the
>>> probe fail as it should.
> If status is not seen in 999.5 ms then I don't see any reason why it will come
> exactly at 1000ms. In fact,  we expect status to be seen within 160ms as per
> I2C_MST_I2C0_TIMING_TIMEOUT_CLK_CNT (16 cycle) and 
> I2C_MST_I2C0_TIMING_SCL_PERIOD_100KHZ (10ms/cycle) programmed in 
> I2C_MST_I2C0_TIMING Register. We already have enough extra time to look
> For response.

This is to handle when there's no response.

When the while loop terminates because of "time_is_after_jiffies(target)" (i.e. target <= jiffies), we also need to to handle "target == jiffies" case in the following if clause to properly timeout.

I don't think I2C timings here can affect jiffies.

Kai-Heng

> 
> Thanks
>> nvpublic
>>> 
>>> Fixes: c71bcdcb42a7 ("i2c: add i2c bus driver for NVIDIA GPU")
>>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
>> 
>> A gentle ping...
>> 
>>> ---
>>> drivers/i2c/busses/i2c-nvidia-gpu.c | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>> 
>>> diff --git a/drivers/i2c/busses/i2c-nvidia-gpu.c
>>> b/drivers/i2c/busses/i2c-nvidia-gpu.c
>>> index 62e18b4db0ed..1988e93c7925 100644
>>> --- a/drivers/i2c/busses/i2c-nvidia-gpu.c
>>> +++ b/drivers/i2c/busses/i2c-nvidia-gpu.c
>>> @@ -88,7 +88,7 @@ static int gpu_i2c_check_status(struct gpu_i2c_dev
>> *i2cd)
>>>              usleep_range(500, 600);
>>>      } while (time_is_after_jiffies(target));
>>> 
>>> -     if (time_is_before_jiffies(target)) {
>>> +     if (time_is_before_eq_jiffies(target)) {
>>>              dev_err(i2cd->dev, "i2c timeout error %x\n", val);
>>>              return -ETIMEDOUT;
>>>      }
>>> --
>>> 2.17.1
>>> 
>