[etherlab-dev] BUG in the mailbox handling

[Jun Yuan]

Scenario 3: During the running
A user application is running, which uses ecrt_sdo_request_read to read
state data from the slave. Meanwhile the command "ethercat download" is
executed in the terminal => the application suddenly receives
EC_INT_REQUEST_FAILURE/EC_REQUEST_ERROR in ecrt_sdo_request_state(), it
failes to check the slave state.

Cause: ecrt_sdo_request_read and ecrt_master_sdo_download racing


On 30 May 2014 23:57, Jun Yuan <j.yuan at rtleaders.com> wrote:

> Hi,
>
> as Gavin Lambert already mentioned one year ago, see
> http://lists.etherlab.org/pipermail/etherlab-dev/2013/000327.html.
>
> I checked the log of an industrial EtherCAT application I made, and I was
> shocked to discover how deeply and frequently my application is affected by
> this issue. It doesn't only affect those who use FoE, SoE, etc, it also
> makes problem while only CoE is used. It doesn't not only affect when the
> user application is started, but also show its teeth without any
> application.
>
> Scenario 1: CoE, no user application
> One slave on the bus with 7 slaves is turned off for temporary
> maintenance. After 5 - 30 seconds, it's turned on again. =>
>
> [ 4629.242295] EtherCAT ERROR 0-0: Received unknown response while
> uploading SDO 0x1C12:00.
> [ 4629.242297] EtherCAT DEBUG: 00 80 07 00 00 00 11 00 09 06
> [ 4629.242305] EtherCAT ERROR 0-0: Failed to read number of assigned PDOs
> for SM2.
>
> Cause: SDO dictionary fetch was disrupted as the slave comes up.
> Consequence: User application cannot be started later, as no PDO was
> assigned to SM2.
>
>
> Scenario 2: CoE, with user SDO operation
> Soon after all the slaves become online, the user application is started,
> which read/write several SDOs from the slaves. And most of the sdo
> operations fail.
>
> [ 1169.599680] EtherCAT DEBUG 0-0: Uploading SDO 0x10F3:02.
> [ 1169.607575] EtherCAT ERROR 0-0: Received unknown response while
> uploading SDO 0x10F3:02.
> [ 1169.607609] EtherCAT ERROR 0-0: Failed to process SDO request.
> [ 1169.615791] EtherCAT DEBUG 0-0: Invalid entry description response
> while fetching SDO entry 0x4810:08!
> [ 1169.615852] EtherCAT DEBUG 0-0: Downloading SDO 0x10F3:03.
> [ 1169.626209] EtherCAT DEBUG 0-0: Invalid SDO download response!
> Retrying...
> [ 1169.628254] EtherCAT DEBUG 0-0: Invalid entry description response
> while fetching SDO entry 0x4810:09!
> [ 1169.628336] EtherCAT DEBUG 0-0: Uploading SDO 0x10F3:27.
> [ 1169.638630] EtherCAT ERROR 0-0: Received unknown response while
> uploading SDO 0x10F3:27.
> [ 1169.665583] EtherCAT DEBUG 0-0: Invalid entry description response
> while fetching SDO entry 0x4810:0A!
> [ 1169.665852] EtherCAT DEBUG 0-0: Uploading SDO 0x608F:01.
> [ 1169.675979] EtherCAT ERROR 0-0: Received unknown response while
> uploading SDO 0x608F:01.
> [ 1169.684240] EtherCAT DEBUG 0-0: Invalid entry description response
> while fetching SDO entry 0x4810:0B!
>
> Cause: User SDO operation happens simultaneously with the SDO dictionary
> fetch.
> Consequence: User application has to be aborted.
>
> Now I would consider it as a serious bug in the IgH EtherCAT Master, which
> prevents the application to be successfully started every time. I'd like to
> ask for the opinions/comments from the developers.
>
> Regards,
> Jun Yuan
>
>
>


-- 
Jun Yuan
[Aussprache: Djün Üän]

Robotics Technology Leaders GmbH
Am Loferfeld 58, D-81249 München
Tel: +49 89 189 0465 24
Fax: +49 89 189 0465 11
mailto: j.yuan at rtleaders.com

Umlautregel in der chinesischen Lautschrift Pinyin: Nach den Anlauten y, j,
q, und x wird u als ü ausgesprochen, z.B. yu => ü,  ju => dschü,  qu =>
tschü,  xu => schü.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-dev/attachments/20140531/7fd97097/attachment.html>