linux-kernel - Re: [PATCH v1 2/7] platform/x86: serial-multi-instantiate: Improve autodetection

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] [day] [month] [year] [list]

Date:   Sat, 9 Jul 2022 16:46:00 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
Cc:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        "platform-driver-x86@...r.kernel.org" 
        <platform-driver-x86@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Mark Gross <markgross@...nel.org>
Subject: Re: [PATCH v1 2/7] platform/x86: serial-multi-instantiate: Improve
 autodetection

Hi Andy,

On 7/9/22 13:34, Andy Shevchenko wrote:
> On Sat, Jul 9, 2022 at 1:00 PM Hans de Goede <hdegoede@...hat.com> wrote:
>> On 7/9/22 11:52, Andy Shevchenko wrote:
>>> On Saturday, July 9, 2022, Hans de Goede <hdegoede@...hat.com <mailto:hdegoede@...hat.com>> wrote:
>>>     On 7/9/22 02:06, Andy Shevchenko wrote:
> 
> ...
> 
>>>     So nack for this
>>>
>>> This effectively means nack to the series.
> 
>>> But it’s easy to fix. I can add check for ret == 0.
> 
> So, are you okay with fixing it this way? See below how.
> 
>> I don't see how this is a nack for the series, just drop 1/7 + 2/7
>> and rebase the rest. Yes there will be conflicts to resolve in
>> the rebase, but the rest of the cleanups can still go upstream
>> after the rebase.
> 
> Because patch 3 makes a little sense on its own if we drop the patch
> 2. The rest is the simple cleanups which I do not consider as a core
> of this series.

Patch 3 just removes the adev == NULL check and pushes
the ACPI_COMPANION(dev) calls into the function needing
the adev, I don't see how that relies on this patch ?

>>>     >       case SMI_AUTO_DETECT:
>>>     > -             if (i2c_acpi_client_count(adev) > 0)
>>>     > -                     return smi_i2c_probe(pdev, adev, smi, node->instances);
>>>     > -             else
>>>     > -                     return smi_spi_probe(pdev, adev, smi, node->instances);
>>>     > +             ret = smi_i2c_probe(pdev, adev, smi, node->instances);
>>>     > +             if (ret && ret != -ENOENT)
>>>     > +                     return ret;
> 
> /*
>  * ...comment about why we do the following check...
>  */
>  if (ret == 0)
>   return ret;

I'm ok with doing things this way. Note you then end up with:

             if (ret && ret != -ENOENT)
                     return ret;

             if (ret == 0)
                     return ret;

Which can be simplified to just:

             if (ret != -ENOENT)
                     return ret;

> 
>>>     > +             ret = smi_spi_probe(pdev, adev, smi, node->instances);
>>>     > +             if (ret && ret != -ENOENT)
>>>     > +                     return ret;
>>>     > +             if (ret)
>>>     > +                     return dev_err_probe(dev, ret, "Error No resources found\n");
>>>     > +             break;
> 
> if (ret == -ENOENT)
>   return dev_err_probe(...);
> return ret;

Hmm, we don't do this dev_err for the SMI_I2C / SMI_SPI cases,
I see 2 options to solve this:

1) Drop the return calls from switch (node->bus_type) {} instead
always set ret and then break. And do the:

 if (ret == -ENOENT)
   return dev_err_probe(...);

outside the switch-case

2) Drop the dev_err, the driver-core will already log an error for
the -ENOENT anyways.

Regards,

Hans