linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Maximilian Luz <luzmaximilian@gmail.com>
To: Hans de Goede <hdegoede@redhat.com>, linux-kernel@vger.kernel.org
Cc: "Mark Gross" <mgross@linux.intel.com>,
	"Andy Shevchenko" <andy.shevchenko@gmail.com>,
	"Barnabás Pőcze" <pobrn@protonmail.com>,
	"Arnd Bergmann" <arnd@arndb.de>,
	"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
	"Rob Herring" <robh@kernel.org>,
	"Jiri Slaby" <jirislaby@kernel.org>,
	"Rafael J . Wysocki" <rjw@rjwysocki.net>,
	"Len Brown" <lenb@kernel.org>, "Blaž Hrastnik" <blaz@mxxn.io>,
	"Dorian Stoll" <dorian.stoll@tmsp.io>,
	platform-driver-x86@vger.kernel.org,
	linux-serial@vger.kernel.org, linux-acpi@vger.kernel.org
Subject: Re: [PATCH v2 1/9] platform/surface: Add Surface Aggregator subsystem
Date: Tue, 8 Dec 2020 15:54:18 +0100	[thread overview]
Message-ID: <b77b74d2-b854-7e8f-fd8e-abee6cdd22ff@gmail.com> (raw)
In-Reply-To: <9748d778-b5e9-c80c-5968-a77b3203d769@redhat.com>



On 12/8/20 3:43 PM, Hans de Goede wrote:
> Hi,
> 
> On 12/8/20 3:37 PM, Maximilian Luz wrote:
> 
> <snip>
> 
>>>> +
>>>> +    obj = acpi_evaluate_dsm_typed(handle, &SSAM_SSH_DSM_GUID,
>>>> +                      SSAM_SSH_DSM_REVISION, func, NULL,
>>>> +                      ACPI_TYPE_INTEGER);
>>>> +    if (!obj)
>>>> +        return -EIO;
>>>> +
>>>> +    val = obj->integer.value;
>>>> +    ACPI_FREE(obj);
>>>> +
>>>> +    if (val > U32_MAX)
>>>> +        return -ERANGE;
>>>> +
>>>> +    *ret = val;
>>>> +    return 0;
>>>> +}
>>
>> [...]
>>
>>>> +/**
>>>> + * ssam_controller_start() - Start the receiver and transmitter threads of the
>>>> + * controller.
>>>> + * @ctrl: The controller.
>>>> + *
>>>> + * Note: When this function is called, the controller should be properly
>>>> + * hooked up to the serdev core via &struct serdev_device_ops. Please refer
>>>> + * to ssam_controller_init() for more details on controller initialization.
>>>> + *
>>>> + * This function must be called from an exclusive context with regards to the
>>>> + * state, if necessary, by locking the controller via ssam_controller_lock().
>>>
>>> Again you are being a bit hand-wavy (I assume you know what I mean by that)
>>> wrt the locking requirements. If possible I would prefer clearly spelled out
>>> locking requirements in the form of "this and that lock must be held when
>>> calling this function". Preferably backed-up by lockdep_assert-s asserting
>>> these conditions.
>>
>> The reason for this is that this function specifically is currently only
>> called during initialization, when the controller has not been published
>> yet, i.e. when we have an exclusive reference to the controller.
>>
>> I'll change this to fully enforce locking (with lockdep_assert).
>>
>>> And maybe if you are a bit stricter with always holding the lock when
>>> calling this, you can also drop the WRITE_ONCE and the comment about it
>>> (in all places where you do this).
>>
>> The WRITE_ONCE is only there to ensure that the basic test in
>> ssam_request_sync_submit() can be done. I always try to be explicit
>> about access that can happen without the respective locks being held.
> 
> Yes I saw the matching READ_ONCE later on (as the comment indicated
> I would), which made it more obvious to me why the WRITE_ONCE is here,'
> so maybe I should have gone back and updated this comment.

No worries, always good to have another look at these kinds of things.

> Anyways, keeping the WRITE_ONCE + READ_ONCE for this is fine.
> 
>> Unfortunately it's not feasible to hold the reader lock in
>> ssam_request_sync_submit() due to reentrancy. Specifically, as the lock,
>> if at all (i.e. if this is not a client driver bound to the controller),
>> must be held not only during submission but until the request has been
>> completed. Note that if we would hold the lock during submission, this
>> is just a smoke-test.
> 
> Ack.
> 
> <more snip>
> 
> Regards,
> 
> Hans
> 

  reply	other threads:[~2020-12-08 14:55 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-03 21:26 [PATCH v2 0/9] Add support for Microsoft Surface System Aggregator Module Maximilian Luz
2020-12-03 21:26 ` [PATCH v2 1/9] platform/surface: Add Surface Aggregator subsystem Maximilian Luz
2020-12-08 13:01   ` Hans de Goede
2020-12-08 14:37     ` Maximilian Luz
2020-12-08 14:43       ` Hans de Goede
2020-12-08 14:54         ` Maximilian Luz [this message]
2020-12-03 21:26 ` [PATCH v2 2/9] platform/surface: aggregator: Add control packet allocation caching Maximilian Luz
2020-12-15 13:42   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 3/9] platform/surface: aggregator: Add event item " Maximilian Luz
2020-12-15 13:44   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 4/9] platform/surface: aggregator: Add trace points Maximilian Luz
2020-12-15 14:20   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 5/9] platform/surface: aggregator: Add error injection capabilities Maximilian Luz
2020-12-15 14:43   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 6/9] platform/surface: aggregator: Add dedicated bus and device type Maximilian Luz
2020-12-15 14:49   ` Hans de Goede
2020-12-15 14:50   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 7/9] docs: driver-api: Add Surface Aggregator subsystem documentation Maximilian Luz
2020-12-15 16:25   ` Hans de Goede
2020-12-03 21:26 ` [PATCH v2 8/9] platform/surface: Add Surface Aggregator user-space interface Maximilian Luz
2020-12-15 16:35   ` Hans de Goede
2020-12-15 20:00     ` Maximilian Luz
2020-12-03 21:26 ` [PATCH v2 9/9] platform/surface: Add Surface ACPI Notify driver Maximilian Luz
2020-12-15 17:18   ` Hans de Goede
2020-12-06  7:07 ` [PATCH v2 0/9] Add support for Microsoft Surface System Aggregator Module Leon Romanovsky
2020-12-06  8:32   ` Greg Kroah-Hartman
2020-12-06  8:35     ` Leon Romanovsky
2020-12-06 11:13     ` Maximilian Luz
2020-12-06  8:41   ` Hans de Goede
2020-12-06  8:56     ` Leon Romanovsky
2020-12-06 10:04       ` Hans de Goede
2020-12-06 10:33         ` Leon Romanovsky
2020-12-06 10:41           ` Hans de Goede
2020-12-06 11:41             ` Leon Romanovsky
2020-12-06 13:43               ` Maximilian Luz
2020-12-06 10:51         ` Maximilian Luz
2020-12-06  8:58     ` Blaž Hrastnik
2020-12-06  9:06       ` Leon Romanovsky
2020-12-06 10:33         ` Maximilian Luz
2020-12-06 10:43           ` Hans de Goede
2020-12-06 10:56             ` Maximilian Luz
2020-12-06 11:30           ` Leon Romanovsky
2020-12-06 13:27             ` Maximilian Luz
2020-12-06 15:58   ` Maximilian Luz
2020-12-07  6:15     ` Leon Romanovsky
2020-12-07  8:49     ` Hans de Goede
2020-12-07  9:12       ` Greg Kroah-Hartman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=b77b74d2-b854-7e8f-fd8e-abee6cdd22ff@gmail.com \
    --to=luzmaximilian@gmail.com \
    --cc=andy.shevchenko@gmail.com \
    --cc=arnd@arndb.de \
    --cc=blaz@mxxn.io \
    --cc=dorian.stoll@tmsp.io \
    --cc=gregkh@linuxfoundation.org \
    --cc=hdegoede@redhat.com \
    --cc=jirislaby@kernel.org \
    --cc=lenb@kernel.org \
    --cc=linux-acpi@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-serial@vger.kernel.org \
    --cc=mgross@linux.intel.com \
    --cc=platform-driver-x86@vger.kernel.org \
    --cc=pobrn@protonmail.com \
    --cc=rjw@rjwysocki.net \
    --cc=robh@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).