[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <92e559b1-d392-015e-1aa1-0497cbd4c071@grandegger.com>
Date: Wed, 14 Mar 2018 10:25:04 +0100
From: Wolfgang Grandegger <wg@...ndegger.com>
To: Jakob Unterwurzacher <jakob.unterwurzacher@...obroma-systems.com>
Cc: Martin Elshuber <martin.elshuber@...obroma-systems.com>,
Philipp Tomsich <philipp.tomsich@...obroma-systems.com>,
Marc Kleine-Budde <mkl@...gutronix.de>,
linux-can@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 1/1] can: ucan: add driver for Theobroma Systems UCAN
devices
Am 14.03.2018 um 10:17 schrieb Wolfgang Grandegger:
>
>
> Am 14.03.2018 um 10:14 schrieb Jakob Unterwurzacher:
>> On 14.03.18 10:11, Wolfgang Grandegger wrote:
>>>> + /* handle error frames */
>>>> + canid = le32_to_cpu(m->msg.can_msg.id);
>>>> + if (canid & CAN_ERR_FLAG) {
>>>> + ucan_handle_error_frame(up, m, canid);
>>>> + /* drop frame if berr-reporting is off */
>>>> + if (!(up->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING))
>>>> + return;
>>>> + }
>>>
>>> You still do not generate error messages for state changes, IIUC.
>>
>> The hardware generates error frames on state changes - is that what
>> you mean?
>>
>> In our testing, the state shows up (and updates) correctly in
>>
>> ip -details -statistics link show can0
>
> Counting the state changes is one thing but you should also generate
> error messages for them.
The usual test here is:
$ candump -td -e any,0:0,#FFFFFFFF
should report proper state changes, if you send messages with
1. no cable connected
2. CAN high and low short-circuited
Also downwards if the hardware error is gone and you continue to send
messages.
Wolfgang.
Powered by blists - more mailing lists