linux-kernel - Re: [PATCH v1] Input: elants_i2c - fix division by zero if firmware reports zero phys size

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:   Sun, 28 Mar 2021 20:21:01 +0300
From:   Dmitry Osipenko <digetx@...il.com>
To:     Dmitry Torokhov <dmitry.torokhov@...il.com>
Cc:     Michał Mirosław <mirq-linux@...e.qmqm.pl>,
        Johnny Chuang <johnny.chuang.emc@...il.com>,
        Jasper Korten <jja2000@...il.com>,
        Svyatoslav Ryhel <clamor95@...il.com>,
        linux-input@...r.kernel.org, linux-tegra@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1] Input: elants_i2c - fix division by zero if firmware
 reports zero phys size

28.03.2021 07:44, Dmitry Torokhov пишет:
> Hi Dmitry,
> 
> On Tue, Mar 02, 2021 at 01:08:24PM +0300, Dmitry Osipenko wrote:
>> Touchscreen firmware of ASUS Transformer TF700T reports zeros for the phys
>> size. Hence check whether the size is zero and don't set the resolution in
>> this case.
>>
>> Reported-by: Jasper Korten <jja2000@...il.com>
>> Signed-off-by: Dmitry Osipenko <digetx@...il.com>
>> ---
>>
>> Please note that ASUS TF700T isn't yet supported by upstream kernel,
>> hence this is not a critical fix.
>>
>>  drivers/input/touchscreen/elants_i2c.c | 8 +++++---
>>  1 file changed, 5 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/input/touchscreen/elants_i2c.c b/drivers/input/touchscreen/elants_i2c.c
>> index 4c2b579f6c8b..a2e1cc4192b0 100644
>> --- a/drivers/input/touchscreen/elants_i2c.c
>> +++ b/drivers/input/touchscreen/elants_i2c.c
>> @@ -1441,14 +1441,16 @@ static int elants_i2c_probe(struct i2c_client *client,
>>  
>>  	touchscreen_parse_properties(ts->input, true, &ts->prop);
>>  
>> -	if (ts->chip_id == EKTF3624) {
>> +	if (ts->chip_id == EKTF3624 && ts->phy_x && ts->phy_y) {
>>  		/* calculate resolution from size */
>>  		ts->x_res = DIV_ROUND_CLOSEST(ts->prop.max_x, ts->phy_x);
>>  		ts->y_res = DIV_ROUND_CLOSEST(ts->prop.max_y, ts->phy_y);
>>  	}
>>  
>> -	input_abs_set_res(ts->input, ABS_MT_POSITION_X, ts->x_res);
>> -	input_abs_set_res(ts->input, ABS_MT_POSITION_Y, ts->y_res);
>> +	if (ts->x_res > 0)
>> +		input_abs_set_res(ts->input, ABS_MT_POSITION_X, ts->x_res);
> 
> There is absolutely no difference between setting respluton to 0 vs not
> setting it at all, so I dropped the conditionals and applied.
> 
>> +	if (ts->y_res > 0)
>> +		input_abs_set_res(ts->input, ABS_MT_POSITION_Y, ts->y_res);
>>  	if (ts->major_res > 0)
> 
> We could drop this condition as well.
> 
>>  		input_abs_set_res(ts->input, ABS_MT_TOUCH_MAJOR, ts->major_res);

I'll make a follow up patch to clean up this condition, if you haven't
done it yet. Thanks!