EtherCAT Problem on AM3357 (Patch on ETG Stack V5.10 fails)

Torsten Koester

Hello,

i´ve got a problem applying the TI provided Patch from SDK 1.1.0.4 to the source of the ETG Stack.

The offsets in the Patch doesn't match with the current SSC V5.10 sources.

I firstly imported the TI_AM335x_1i1i0i4.xml to the SSC Tool which works fine.
The activated some additional features like Diag History, Emergency and FoE. Then i created the source file and tried to apply the TI_ECAT.Patch with mostly failed hunks. I could adapt the most changes from the patch by hand. But on the file ecatappl.c I could not find the corret lines. I now have a running Version of the stack but i always get a warning (Diagcode 0x000F: Calculate bus cycle time failed (Local timer too slow)) in the diag history when running in Freerun and SM Synchron mode. DC not tested yet. This message is generated from pdi_isr() function. But i don't now why. So i asume this might be a problem from the patch. It even doenst matter if i use the pdi and sync task or the interrupt variant. Is it possible to get a correct version of the ecatappl.c file (We are ETG member).

Regrads

Torsten

over 11 years ago

0 Biser Gatchev-XID over 11 years ago

TI__Guru**** 393215 points

Moving this to the SYS/BIOS forum.

0 Vinesh Balan over 11 years ago

TI__Expert 7360 points

Torsten,

What issues are you getting while trying to apply the patch?

Also, which version of SSC Tool are you using?

Regards,
Vinesh

0 Torsten Koester over 11 years ago in reply to Vinesh Balan

Prodigy 80 points

Hi Vinesh,

thae patch can't be applied because the linenumbers given in the patch does not match the real file ones. So patch.exe says hunk failed.

I'm using Slave Stack Tool V1.3.1.0

Do you know what the message "Calculate bus cycle time failed (Local timer too slow)" should be for? I doesn't understand the meaning of this warning...

Regards,

Torsten

0 Vinesh Balan over 11 years ago in reply to Torsten Koester

TI__Expert 7360 points

Torsten,

The SSC tool automatically applies the patch to the source(if you have configured the path to the patch in TI_AM335x_1i1i0i4.xml or as mentioned in this). So you don't have to apply the patch separately.

I would also suggest to re-download the SSC tool(if you did not download it recently)

Regards,
Vinesh

0 Torsten Koester over 11 years ago in reply to Vinesh Balan

Prodigy 80 points

Hi Vinesh,

thanks that helped. I now have a version of the ecatappl.c file which is slightly different from the one i got before.

But i still get the Warning in the diag history. Can you explain me whats the meaning of it?

Regards

Torsten

0 PratheeshGangadhar over 11 years ago in reply to Torsten Koester

TI__Mastermind 47851 points

Hi,

Are you running master in RUN mode or DC mode?

As I can see from SSC code, this warning is issued in FreeRun mode i.e when sSyncManOutPar.u16GetCycleTime = 0

#if DIAGNOSIS_SUPPORTED
Diag_CreateNewMessage(DIAG_CODE_APPL_SYNC,DIAG_MSG_TYPE_WARNING,DIAG_TEXT_ID_CALC_BUS_CYCLE_FAILED,0,NULL);
#endif

I am not a stack expert, it will be good to ask in EtherCAT forums if you want to know more

0 Torsten Koester over 11 years ago in reply to PratheeshGangadhar

Prodigy 80 points

Hi Pratheesh,

it doesn´t matter if the master runs in RUN or in FREERUN mode.

The Slave runs in SM Synchron Mode and as i understand the SSC Code the Diagnosis is created in SM Synchronmode when bEscIntEnabled = true and !DcSyncActive. That seems to be a correct behaviour but is not realy helpfull for a user. Maybe the master must start a cycle time calculation before. I already asked in the ETG Forum but yet no response.

Thank you.

Regards

Torsten

Processors

Processors forum

EtherCAT Problem on AM3357 (Patch on ETG Stack V5.10 fails)