All of lore.kernel.org
 help / color / mirror / Atom feed
From: Benjamin Tissoires <benjamin.tissoires@redhat.com>
To: Jason Gerecke <killertofu@gmail.com>
Cc: linux-input@vger.kernel.org, Jiri Kosina <jkosina@suse.cz>,
	Ping Cheng <pinglinux@gmail.com>, Aaron Skomra <skomra@gmail.com>,
	Jason Gerecke <jason.gerecke@wacom.com>
Subject: Re: [PATCH] HID: wacom: Fix sibling detection broken by 345857b
Date: Wed, 18 Jan 2017 09:00:27 +0100	[thread overview]
Message-ID: <20170118080027.GI8644@mail.corp.redhat.com> (raw)
In-Reply-To: <20170117233858.21546-1-killertofu@gmail.com>

On Jan 17 2017 or thereabouts, Jason Gerecke wrote:
> Commit 345857b included a change to the operation and location of the call
> to 'wacom_add_shared_data' in 'wacom_parse_and_register'. The modifications
> included moving it higher up so that it would occur before the call to
> 'wacom_retrieve_hid_descriptor'. This was done to prevent a crash that would
> have occured when the report containing tablet offsets was fed into the
> driver with 'wacom_hid_report_raw_event' (specifically: the various
> 'wacom_wac_*_report' functions were written with the assumption that they
> would only be called once tablet setup had completed; 'wacom_wac_pen_report'
> in particular dereferences 'shared' which wasn't yet allocated).
> 
> Moving the call to 'wacom_add_shared_data' effectively prevented the crash
> but also broke the sibiling detection code which assumes that the HID
> descriptor has been read and the various device_type flags set.
> 
> To fix this situation, we restore the original 'wacom_add_shared_data'
> operation and location and instead implement an alternative change that
> can also prevent the crash. Specifically, we notice that the report
> functions mentioned above expect to be called only for input reports.
> By adding a check, we can prevent feature reports (such as the offset
> report) from causing trouble.
> 
> Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com>
> Tested-by: Ping Cheng <pingc@wacom.com>
> ---

Reviewed-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>

I agree this should go in v4.10 given that it introduces a regression
for Wacom generic devices.

Cheers,
Benjamin

>  drivers/hid/wacom_sys.c | 16 ++++++++--------
>  drivers/hid/wacom_wac.c | 10 ++++++++++
>  2 files changed, 18 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/hid/wacom_sys.c b/drivers/hid/wacom_sys.c
> index b9779bcbd140..8aeca038cc73 100644
> --- a/drivers/hid/wacom_sys.c
> +++ b/drivers/hid/wacom_sys.c
> @@ -740,6 +740,11 @@ static int wacom_add_shared_data(struct hid_device *hdev)
>  		return retval;
>  	}
>  
> +	if (wacom_wac->features.device_type & WACOM_DEVICETYPE_TOUCH)
> +		wacom_wac->shared->touch = hdev;
> +	else if (wacom_wac->features.device_type & WACOM_DEVICETYPE_PEN)
> +		wacom_wac->shared->pen = hdev;
> +
>  out:
>  	mutex_unlock(&wacom_udev_list_lock);
>  	return retval;
> @@ -2036,10 +2041,6 @@ static int wacom_parse_and_register(struct wacom *wacom, bool wireless)
>  	if (error)
>  		goto fail;
>  
> -	error = wacom_add_shared_data(hdev);
> -	if (error)
> -		goto fail;
> -
>  	/*
>  	 * Bamboo Pad has a generic hid handling for the Pen, and we switch it
>  	 * into debug mode for the touch part.
> @@ -2080,10 +2081,9 @@ static int wacom_parse_and_register(struct wacom *wacom, bool wireless)
>  
>  	wacom_update_name(wacom, wireless ? " (WL)" : "");
>  
> -	if (wacom_wac->features.device_type & WACOM_DEVICETYPE_TOUCH)
> -		wacom_wac->shared->touch = hdev;
> -	else if (wacom_wac->features.device_type & WACOM_DEVICETYPE_PEN)
> -		wacom_wac->shared->pen = hdev;
> +	error = wacom_add_shared_data(hdev);
> +	if (error)
> +		goto fail;
>  
>  	if (!(features->device_type & WACOM_DEVICETYPE_WL_MONITOR) &&
>  	     (features->quirks & WACOM_QUIRK_BATTERY)) {
> diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c
> index b1a9a3ca6d56..0884dc9554fd 100644
> --- a/drivers/hid/wacom_wac.c
> +++ b/drivers/hid/wacom_wac.c
> @@ -2187,6 +2187,16 @@ void wacom_wac_report(struct hid_device *hdev, struct hid_report *report)
>  
>  	wacom_report_events(hdev, report);
>  
> +	/*
> +	 * Non-input reports may be sent prior to the device being
> +	 * completely initialized. Since only their events need
> +	 * to be processed, exit after 'wacom_report_events' has
> +	 * been called to prevent potential crashes in the report-
> +	 * processing functions.
> +	 */
> +	if (report->type != HID_INPUT_REPORT)
> +		return;
> +
>  	if (WACOM_PAD_FIELD(field)) {
>  		wacom_wac_pad_battery_report(hdev, report);
>  		if (wacom->wacom_wac.pad_input)
> -- 
> 2.11.0
> 

  reply	other threads:[~2017-01-18  8:01 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-17 23:38 [PATCH] HID: wacom: Fix sibling detection broken by 345857b Jason Gerecke
2017-01-18  8:00 ` Benjamin Tissoires [this message]
2017-01-19 13:24 ` Jiri Kosina

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=20170118080027.GI8644@mail.corp.redhat.com \
    --to=benjamin.tissoires@redhat.com \
    --cc=jason.gerecke@wacom.com \
    --cc=jkosina@suse.cz \
    --cc=killertofu@gmail.com \
    --cc=linux-input@vger.kernel.org \
    --cc=pinglinux@gmail.com \
    --cc=skomra@gmail.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.