Refactor IXXAT PR

[zariiii9003]

I thought it might be easier to just create a PR against your PR instead of writing a long review. Could you take a look and test on your hardware?

[zariiii9003]

@MattWoodhead Regarding the extended-Parameter, are there really any non-ancient interfaces, that do not support both standard and extended IDs at the same time?

[MattWoodhead]

Sorry @zariiii9003, thanks for pinging me. I got the email notifications, but the links kept sending me to the PR in the main repo so I wasn't sure where your comments were! I will take a look at this after work today.

[MattWoodhead]

Apologies @zariiii9003, there were some careless copy paste errors in this.

Just to keep you updated, I am getting some test failures:

FAILED test/test_interface_ixxat.py::TestHardwareCaseStd::test_send_periodic - can.exceptions.CanOperationError: No messages have been received
FAILED test/test_interface_ixxat.py::TestHardwareCaseStd::test_send_single - ValueError: Response does not match the sent message
FAILED test/test_interface_ixxat.py::HardwareTestCaseFd::test_send_after_shutdown - AssertionError: CanOperationError not raised

I have added the following line to both my branch and your PR (on my machine locally), as part of investigating the failures, and have found that something is causing the VCI canControlInitialize function call to fail in this branch.

INFO     can.ixxat:canlib.py:473 CAN Filters: None
INFO     can.ixxat:canlib.py:715 Searching for first available device
INFO     can.ixxat:canlib.py:534 Using unique HW ID b'HW509182'
INFO     can.ixxat:canlib.py:536 Initializing channel 0 in shared mode, 16 rx buffers, 16 tx buffers
INFO     can.ixxat:canlib.py:638 Initialising Channel 0 with the following parameters:
dwMode=0, dwBPS=250000, wTS1=6400, wTS2=1600, wSJW=1600, wTDO=0
dwMode=0, dwBPS=2000000, wTS1=1600, wTS2=400, wSJW=400, wTDO=1600
WARNING  can.ixxat:canlib.py:131 VCI_E_ACCESSDENIED error raised when calling VCI Function canControlInitialize
WARNING  can.ixxat:canlib.py:131 VCI_E_ACCESSDENIED error raised when calling VCI Function canControlStart
WARNING  can.ixxat:canlib.py:131 VCI_E_ACCESSDENIED error raised when calling VCI Function canControlStart
WARNING  can.ixxat:canlib.py:131 VCI_E_ACCESSDENIED error raised when calling VCI Function canControlReset

I will investigate more tomorrow.

[zariiii9003]

Why are the TS1, TS2 and SJW values so big in your log file? That's weird. Which test is that?

[zariiii9003]

@MattWoodhead i rebased onto your branch.

[zariiii9003]

Could you try again?

[MattWoodhead]

Why are the TS1, TS2 and SJW values so big in your log file? That's weird. Which test is that?

It is the same or similar values (depending upon chosen bitrate) in any of the tests that use one of the predefined bit timing settings from the constants file - these were added in PR hardbyte#1119 (although I did add in some more predefined bitrates). Based on dwMode being 0, I think the timing get set by the ratio of the quanta value to the sum of the two time segmant quanta - see page 136 (138 in a pdf reader) of the API docs

[MattWoodhead]

Hi @zariiii9003.
I have tested the latest version from this branch but unfortunately still getting the same errors. I have attached the pytest results json from my branch and this branch below (as txt files unfortunately as it wouldnt allow me to attach a JSON file).
I have also noticed that my branch introduced a regression compared to the main repo - it does not allow multiple connections to the same channel (raising the same VCI_E_ACCESSDENIED error in the logs) - I am just going to push a test case for this to my branch shortly.

ixxat_unified_class2_pytest_result.txt

ixxat_unified_class2_refactor_pytest_result.txt

Do you think that the change in the VCI function call order from
_canlib.canControlOpen -> _canlib.canControlInitialize -> _canlib.canControlGetCaps
to
_canlib.canControlOpen -> _canlib.canControlGetCaps -> _canlib.canControlInitialize
may have caused any of these issues?

[zariiii9003]

I think VCI_E_ACCESSDENIED was just silenced. See here: hardbyte#488

Can you show bOpMode and bExMode on your branch? I really don't understand, why my branch does not work 😩

[MattWoodhead]

I think VCI_E_ACCESSDENIED was just silenced. See here: hardbyte#488

Can you show bOpMode and bExMode on your branch? I really don't understand, why my branch does not work 😩

INFO     can.ixxat:canlib.py:707 Initialising Channel 0 with the following parameters:
bOpMode=7, bExMode=0
pBtpSDR=dwMode=0, dwBPS=500000, wTS1=6400, wTS2=1600, wSJW=1600, wTDO=0
pBtpFDR=dwMode=0, dwBPS=2000000, wTS1=1600, wTS2=400, wSJW=400, wTDO=1600

I don't really understand why either :(
I shall continue to try and investigate.

[zariiii9003]

@MattWoodhead did you find anything?