This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM2434: AM2434 EtherNET/IP Test Question.

Part Number: AM2434
Other Parts Discussed in Thread: LP-AM243

Hi,

     Customer is setting EtherNet/IP Conformance Test tool and using AM243-LP as DUT(Device Under Test) to run test.

In order to reproduce the test environment, They have some questions about setting.

Do they need to check the below checkbox?

  1. Type 1 or 2 Reset
  2. Profile Verification
  3. Encapsulation

  • Hi ,

    Let me come back by Wednesday next week.

    BR

    Nilabh A.

  • Hi  ,

    Do they need to check the below checkbox?

    1. Type 1 or 2 Reset
    2. Profile Verification
    3. Encapsulation

    Yes, These are compulsory test as per ODVA and we test our DUT for above three.

    BR

    Nilabh A.
  •      Customer report they have problem after they turn on Profile Verification and Encapsulation. They are using SDK 8.4.

    Can you please provide some suggestions to pass tests?

    Regards

    Andre

    CT19_am243x-lp_tiEtherNetIP_freeRTOS_Identity_ProfileVerificationTest_Encapsolation_Selection.log  

  • Hi Andre,

     Customer report they have problem after they turn on Profile Verification and Encapsulation. They are using SDK 8.4.

    TI highly recommends using latest SDK version, as this comes with bug fixes which are reported on previous SDK.

    I will check regarding the conformance test results on SDK 8.04. Let me get info from internal team on this. I will try to get back by early next week. Thanks

  • Nilabh

    Customer report they have the same issue when they use SDK 8.5. Can you please check or  provide our own test log?

    Regards

    Andre

  • Nilabh,

       Customer has used the version 8.5 to re-test all cases below with the .soc file setting from the SDK, and update the test issues below:

    They got 6 errors and 3 warning from the conformance test.

    The error below happens when running the “Profile Verification” test.
    1> *** EtherNet/IP Devices undergoing certification for the first time and multi-port devices receiving an update that adds CIP functionality are required to implement LLDP protocol and LLDP Management Object starting from May 2022.

     

    Another 2 errors happen when running the “Time Sync” test in line 1217 and line 1234.

    1217      1> *** This attribute is required if 1. A device supports more than one CIP port and it's impossible for the PTP port and CIP port numbers to be the same; or 2. More than one PTP port is associated with a single CIP port (e.g. PRP). Otherwise the error can be waived.

     

    When they did the “DLR” and “QoS” tests, they also receive another error at line 1781.

    1781      1> *** Didn't receive valid Learning_Update Frame from DUT within 100 milliseconds.

     

    The “TCP/IP” and the “Ethernet Link” tests shows another error as below.

    3357      1> *** Attribute data value incorrect attr 3, Srv Code x0E

     

    There is 1 warning they received from the “Identity” test.

    434       1> ** Warning: It is highly recommended to support Identity Object Class Revision 2.

     

    Another 2 warning they received from the “Assembly” test.

    2041      1> ** Warning: No data in the Data attribute. Please verify if it's acceptable.2572      1> ** Warning: No data in the Data attribute. Please verify if it's acceptable.

     

    The log files are as attachments.

     

    They would like to know if these issues are caused by known firmware issues or by test configurations?

    How to fix these problems?

    Regards

    AndreAssembly+ConnectionMessage.logDLR+QoS.logIdentity.logMessageRouter.logTCPIP+EtherNetLink.logTimeSync.log

  • Hi Andres,

    are we sure that they have configured their host PC (on which the test runs) and the IP addresses in the test tool itself correctly?
    On the host PC one is required to configure 2 consecutive IP addresses on the network port that is connected to the target, for example 192.168.1.1/24 and 192.168.1.2/24. Also on the Run Test tab page, there are the Select DUP IP Addresses and Select Host IP Addresses pull-down menus that are frequently wrong when one starts CT19. They need to make also sure that MAC addresses are correct. 
    The LLDP error message is known, but addressed with later revisions of the stack (the did not made it into the SDK yet due to problems with the SDK RC for 8.4 and 8.5)

    Regards

    Nilabh

  • Nilabh,

         After customer corrects the MAC setting, the error in DLR test is fixed, but other 3 test issues are remained. They also confirm all NIC configuration follow the descriptions in CT19 help manual.

        Can you please help for the following problems in “Time Sync” and “Ethernet Link” tests?

    Time Sync test: 

        2 errors happen when running the “Time Sync” test in line 1217 and line 1234.

    1217      1> *** This attribute is required if 1. A device supports more than one CIP port and it's impossible for the PTP port and CIP port numbers to be the same; or 2. More than one PTP port is associated with a single CIP port (e.g. PRP). Otherwise the error can be waived.  

    The “Ethernet Link” tests shows another error as below.

    3357      1> *** Attribute data value incorrect attr 3, Srv Code x0E

     

    Additionally, to those warning messages they need TI’s answers if it will be fixed or not.

    There is 1 warning I received from the “Identity” test.

    434       1> ** Warning: It is highly recommended to support Identity Object Class Revision 2.

     

    Another 2 warning I received from the “Assembly” test.

    2041      1> ** Warning: No data in the Data attribute. Please verify if it's acceptable.2572      1> ** Warning: No data in the Data attribute. Please verify if it's acceptable.

     

     

    They also suddenly get a “License Expired” message from the UART port. It seems causing lots of errors when doing the encapsulation test and never happened before.

    Could you help customer  to figure out what’s going on and how to fix it?

     

    Thank you for your help.

    Regards

    Andre

  • Hi Andrea,

    Can you please help for the following problems in “Time Sync” and “Ethernet Link” tests?

    Let me check on the error with the team internally.

    hey also suddenly get a “License Expired” message from the UART port. It seems causing lots of errors when doing the encapsulation test and never happened before.

    Could you help customer  to figure out what’s going on and how to fix it?

    For this could you please tell me if you are using F variant of AM243 or not?

  • Nilabh,

        They are using   XAM2434BSFGGIALX(HS-FS) device. Any issue with this sample?

    Regards

    Andre

  • Hi Andreas,

      They are using   XAM2434BSFGGIALX(HS-FS) device. Any issue with this sample?

    F samples should not have given this issue. Let me check this internally.

    Can you please help for the following problems in “Time Sync” and “Ethernet Link” tests?

    Concerning the Time Sync test, this error message can be ignore, i.e. 'waived' as ODVA calls it. This was confirmed to us by ODVA. as we have only one PTP port that is used by both Ethernet Link instances

    Concerning the Ethernet Link test error, we strongly assume that the physical MAC address is not configured correctly. It has to be set both in  246 Ethernet Link - Instance Group 1 and under Physical Data - TCP Interface 1.

    Concerning the Warning on the Identity Object, this is work in progress, as we need this to support Vol 7C IO-Link gateways.

    Finally, the Assembly test warning, this can be ignored.

  • Nilabh,

          More questions from customer, they still have problems when they test QoS by ODVA conformance tool to CT19.1. 

    1. There are 5 test fails in 26 tests when they did the ACD test.. Please referring to the full test log in attachment.

    ACD full test report.xmlACD full test report.html 

    All failing tests are relevant to two issues of “the response behavior when a conflict is detected” and “ACD timing verifying”. The screen clips are as attachments of the two .PNG files.

    Do you know what may cause these behaviors fail?

    Regards

    Andre

         

  • Hi Andrea, 

    we are looking into the issues, let me get back on it once I have an update

  • Hi Andrea,

    the issues in the ACT test reported above we cannot reproduce here. We see one ACD issue here reported by the CT19.1 scripts, but this is not one of those above. This is currently investigated.

    Concerning the messages above, do you happen to have a Wireshark trace, as from the HTML file we don't get enough detail.

    The ACD test seems to be tricky once in a while (from the user perspective) thus a Wireshark trace may help to identify the real issue for the messages above. As mentioned, with our current software status we don't see those messages and nothing had been changed related to ACD since the 8.4 release.

    Can you please help with the wireshark trace here.

  • I asked customer to provide it, will update later.

    Regards

    Andre

  • Customer  re-tested the ACD fail test cases 3.7, 3.8, and collected the Wireshark data.
    Please see the attachments. The Wireshark data is between the PC and the Hub A, according the ACD hardware setup from ODVA.

    ACD 3.7_3.8 .pcapng

    ACD 3.7_3.8.html
    Fullscreen
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    <html xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xs="http://www.w3.org/2001/XMLSchema-datatypes">
    <head>
    <META http-equiv="Content-Type" content="text/html; charset=utf-8">
    <style>
    table {
    border-collapse: collapse;
    width: 100%;
    }
    table, th, td {
    border: 1px solid black;
    }
    th, td {
    text-align: left;
    padding: 8px;
    }
    th {
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

    Regards

    Andre

  • Thanks Andreas for the info, Our team is analyzing the info. Let me get back by Thursday on this. 

  • One more issue:

    If they test “Ethernet Link” case more than one time, this test case probably has error at switching between 10M and 100M.

    When error occur, they can’t PING AM243-LP by computer anymore.

    The test log and AM243 log are attached, and test result as below:

    Regarding  license issue, do you  have feedback?

    customer sometimes reproduce this issue in “Ethern

    et Link” case:

    Regards

    Andre

  • log as attachmentsCT19_am243x-lp_tiEtherNetIP_freeRTOS.log

    am243_comport_log.txt
    Fullscreen
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    DMSC Firmware Version 8.5.3--v08.05.03 (Chill Capybar
    DMSC Firmware revision 0x8
    DMSC ABI revision 3.1
    [BOOTLOADER_PROFILE] Boot Media : NOR SPI FLASH
    [BOOTLOADER_PROFILE] Boot Media Clock : 100.000 MHz
    [BOOTLOADER_PROFILE] Boot Image Size : 875 KB
    [BOOTLOADER_PROFILE] Cores present :
    r5f0-0
    [BOOTLOADER PROFILE] SYSFW init : 10850us
    [BOOTLOADER PROFILE] System_init : 5367701us
    [BOOTLOADER PROFILE] Drivers_open : 268us
    [BOOTLOADER PROFILE] Board_driversOpen : 153626us
    [BOOTLOADER PROFILE] Sciclient Get Version : 10022us
    [BOOTLOADER PROFILE] CPU load : 10945us
    [BOOTLOADER_PROFILE] SBL Total Time Taken : 184707us
    Image loading done, switching to application ...
    Pruicss max =3 selected PRU:3
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
    .

  • hi Andrea,

    thanks for the trace. We can now confirm that there is an issue in the ACD implementation. The problem with the failing conflict detection showed up when we migrated from the older 0.7 lwIP stack to the newer lwIP socket based interface between our version 1.04.10 (which had been ODVA certified with CT18.1) and version 2.01.00 which used the new interface between the EtherNet/IP stack and the lwIP stack. 

    We also looked though the SDK examples (08.04.00.17) . We don't see the reported error. The Wireshark trace you sent however clearly shows an issue the the conflict pretended by the conformance test tool is not detected.
    In the trace that we produced here (with the TMDS64GPEVM) you see our ARP probe in Line 21 and the response from the tool in line 22 (23 is just a duplicate made up by Wireshark), and the next test case starts in line 27. So we are a bit puzzled concerning the results in your trace.


    .ACDtest_SDK_08.04.00.17.pcapng.zip

  • Hi Nilabh

    Thanks for your support.

    Our SDK version is 08.05.00.24. Does this version have the LwIP problem?

    Is it a stack problem or an ODVA's problem?

     

    By the way, Andre ever mentioned a version 8.6 will release to solve the “Profile verification” test fail issue. May I know when the version will be released?

    What are the test issues that TI already knows on the CT19.1 of ODVA conformance test tool?

    Best regards

    Bo-Cheng

     

  • Hi Andrea/Bo-Cheng

    As  I mentioned before with our setup here we did not see this error messages. Therefore it would be important to exactly reproduce your setup.

    We are trying to reproduce your setup on our side for this we need:

    1. If could we get their appimage?
    2. Is there any way to figure out the network interface settings of the PC you are using for CT19.
  • Nilabh,

       Here is reply from customer:

     Ethernet and app setting are as attachments. All are according to the instruction of the ODVA CT19.1 user manual. All firewalls are disabled in the test PC.

    If other setting need to be confirmed?

    Regards

    Andre

  • Hello Nilabh,

    Do you have any update?

    By the way, do you have the schedule of when the stack V8.6 will release?

    Thank you 

    Bo-Cheng

  • As I had requested I need the .out and .appimage file to reproduce the setup.

    Could you please share that.

  • Thanks Andrea for sharing this.

    I would also like to know is the usage of RGMII mode intentional, if yes may I know the reason?

    We have earlier seen some instability with RGMII.

    For the time being please use MII mode binaries to test/ Attaching the binaries.

    /cfs-file/__key/communityserver-discussions-components-files/908/ti_2D00_arm_2D00_clang.7z

  • Nilabh,

       Customer have test the MII mode with v8.5 stack on EVM: LP-AM243 / PROC109E2 (GP device).

    1. The firmware is built by ourselves with the SDK version: 08_05_00_24 and Project:{MCU_PLUS_SDK}/examples/industrial_comms/ethernetip_adapter_demo/mii/am243-lp, because the firmware that you offered on e2e is not functional on our EVM when we ping the IP address.
    2. All the ACD  issues are the same as before we used RGMII. Please see the attachments.
    3. They also found when they run many times of the conformance test, the encapsulation test will occasionally have the following errors.
      These error can be easily reproduced by continuously running 5 times of the Ethernet-Link test. See the attachment.

    The used .soc file and the UART output are also in attached.

     

    Customer  want to know,

    1. What the EVM and the stack version and the project that you are used for testing these issues?
    2. Is there any idea to figuring out what may cause the ACD and the encapsulation issues?

     

    Since the testing issues are happened on the evaluation board, such unstable and failure testing result is already affect the product development.

    We hope the root cause can be figuring out as soon as possible.

    Regards

    Andre'

    ACD test.html

    am243x-lp_tiEtherNetIP_freeRTOS.socCT19_am243x-lp_tiEtherNetIP_freeRTOS-5timesEthernetLinkTest.log

  • Hi Andrea,

    I just got feedback from the development team. They use the appimage that you sent us and flashed it into a non HS_FS Launchpad. We had seen no error in the ACD test with this image. Attached you find the Log and the Wireshark trace.

    Regards,
    Martin

    ACD full test report.zip

  • Martin,

        Anything wrong with their setup? Can you please help to check anything different than our test setup so their result is bad ?

    Regards

    Andre

  • Hi Martin, Nilabh

    Thank you for the ACD full test report. I have compared the report with ours, and I found our ARP Probe behaviors are very different then yours.

    May I know what's the brand and model of hubs/switches did you use for the ACD tests?

    Thank you

    Bo-Cheng

  • Hi Bo-Cheng

    Since our team is OOO next week, please expect a response by 19th June.

  • Hi Bo-Cheng

    we use the following test setup as described in the conformance test example report (Excel)

    The switch used by the tester is a tp-link TL-SG108PE.

    Attached you find the test results obtained with an E2 revision launchpad.

    ACD full test report_boardE2.zip

    Regards,

    Nilabh A.

  • Could you also please share your setup info, like the block diagram and the equipment's used.

  • Hi Nilabh,

    Our ACD test setup is like the figure below.
    I also tested with the I219-V newwork card, and the result was the same.


    Regards
    Bo-Cheng

  • Hi Bo-Cheng

    We don't see anything that would explain they behavior they seem to see. There's one thing we haven't asked yet. The network traffic, are you recording it via a mirror port on the switch or directly on the network port on the test PC. The second thing that comes to our mind would be if you have checked the firewall logs. Possibly something got blocked there. Firewalls get picky when they see duplicate use of IP addresses.

    Could you please check the above points?

  • Hi Andrea,

    Could you please arrange a debug call with customer to look into this issue, We will have our team joint yeh call. Wednesday next week 21/06/23 seems to be good for us

  • Hi Andrea, Ant update here?

  • I sent you web conference already, and I saw you accepted  invitation last week. 

    Any issue???

    Regards

    Andre

  • Customer wants to discuss anther e2e issue which did not get any feedback in the past two weeks. https://e2e.ti.com/support/microcontrollers/arm-based-microcontrollers-group/arm-based-microcontrollers/f/arm-based-microcontrollers-forum/1232056/am2434-ethernet-ip-cip-motion-feature-support

    Please also prepare the answer to address these issues in the meeting.

    Regards

    Andre 

  • Customer wants to discuss anther e2e issue which did not get any feedback in the past two weeks. https://e2e.ti.com/support/microcontrollers/arm-based-microcontrollers-group/arm-based-microcontrollers/f/arm-based-microcontrollers-forum/1232056/am2434-ethernet-ip-cip-motion-feature-support

    Please also prepare the answer to address these issues in the meeting.

    Regards

    Andre 

  • Hi Nilabh,

    Let me update a latest test result.

     

    1. After we replace the D-Link DGS-1210-10MP switches by a PHOENIX CONTACT 2891001 FL switch, only one test fail of the case 4.3.4 and a difference of response behavior are remained. Since we have no idea what may cause the test fail of the 4.3.4, we ask for your help. Please refer to the attach files for the full and single test record data.
    2. The difference of response behavior is happened on the case 3.2.1. I noticed that our full test behavior is different than yours, like the figure below. It only happens on doing a full test, but performs pass when doing a single test independently. I also attach the single test result of ACD_3.1_3.2_Pass. Do you know what may cause shch phenomenal?

    TestResult.zip

    Best Regard

    Bo-Cheng

  • Hi Bo

    We will try to debug this over next debug session.

  • Hello Nilabh,

    May I ask for a WireShark data between the DUT and the Switch of the ACD behavior test 4.3? 
    I want to compare our behavior with yours. 

    Thank you

    Bo-Cheng

  • I've made the request since Nilabh is out of office.

    Regards,
    Mike

  • Hi Bo-Cheng

    We rerun now ACD tests on both LP-AM243 PROC109E2 (with our release 2.1 for SDK 08.05) and with the PROC109E3 HS_FS with the official MCU+ SDK 08.06.00.43. Attached you find the test results. We have not observed any issues. The error messages we had seen this morning on the E2 revision board was a typical 'Change User' scenario (i.e. change me), We just forgot to modify the MAC address in the SOC editor of the CT19.1. Once the correct MAC address is set in SOC editor (in the tool it's still call STC Editor) under 246Ethernet Link - Instance Group 1 and under Physical Data - TCP Interface 1, tests are running with result PASS.

    There are however the know issues in the 08.06.00.43 release (calloc, time sync) that need an update.

    best regards,

    Martin

    ACD Behavior Test V2.4.13 TC 4.3 Generic Device Port 1 LP-AM243 2023-07-05T15_MCU_PLUS_SDK_08.06.00.43.zipACD Behavior Test V2.4.13 TC 4.3 Generic Device Port 1 LP-AM243 2023-07-05T15_MCU_PLUS_SDK_08.05.00.01.zipACD Behavior Test V2.4.13 Run All Generic Device Port 1 LP-AM243 2023-07-05T15_MCU_PLUS_SDK_08.05.00.01.zip

  • May I ask for a WireShark data between the DUT and the Switch of the ACD behavior test 4.3? 
    I want to compare our behavior with yours. 

    LP-AM243_Interactive_ACD_Behavior_Test_V2.4.13_Test_section_4.3.pcapng.zip

  • Hello Nilabh,

    Our Wireshark logs between the DUT and the swirch as attachment. One is tested with a PHOENIX CONTACT switch, and another is with a TP Link switch.

    In the end of last week I got a TP Link switch TL-SG108E. I used it on the test 4.3, and the test is passed. However, as I know, ODVA use the D-Link DGS-1210-10P switch, not the TP-Link one. I ever used different switches for the test, but only the TP-Link can pass this test.

    Could you ask KUNBUS to use other brand's switch for the ACD behavior test? In current situation, it's hard to have confidence that the product can pass the test at ODVA labs.

    The list of switches we ever used:
    1. D-Link DGS-12210-10MP (Fail)
    2. D-Link DES1005A (Fail)
    3. PHOENIX CONTACK FL SWITCH SFNB 5TX (Fail)
    4. SIEMENS SCALANCE 204IRT (Fail)
    5. HP (Fail)
    6. TP-Link TL-SG108E (Pass)

    ACD Behavior test 4.3 with diff brand switches.zip

  • Sure Bo-Cheng,

    Let me get back with a feedback by Thursday latest

1 2