USB superspeed peripherals Forum Discussions

The GPIF2 user guide, has the following description for the COMMIT action:

 

The COMMIT action commits the data packet / buffer on the selected Ingress DMA channel. This action is typically used to force “buffer / packet end” using state machine. For committing short buffers, this action should be used along with the IN_DATA action. If the buffer is partially filled and the COMMIT action is performed without IN_DATA action, a zero length buffer will be committed in addition to the partially filled buffer.

 

 

It mentions what happens when the COMMIT action is performed when the current thread's buffer is only partially filled. However, it is not immediately clear what happens when the COMMIT action is performed WITH an IN_DATA action where the IN_DATA action will fill the buffer. Can I get a clarification on what happens in this scenario?

In this scenario, does the COMMIT action do anything since the IN_DATA action will completely fill a buffer and that should automatically commit the filled buffer to the consumer socket?

In this scenario, does the COMMIT action additionally commit a ZLP after the filled buffer is committed to the consumer socket?

 

Furthermore, what does COMMIT action do if the current buffer is empty (i.e. no IN_DATA action was done beforehand)? Does it commit a ZLP to the consumer socket?

 

Thanks

Show Less

What is the behavior of the ARM926EJ-S core when a division by zero or an underflow/overflow occurs during arithmetic operations. I know some architectures have defined behavior or can be configured to trigger an interrupt in response to one of these arithmetic exceptions. Does this processor have any such capabilities to detect one of these conditions at runtime?

Show Less

I am trying to use FX3 with Sigrok PulseView. I came across KBA233652 but was not able to find a simple way to compile the binaries for Windows. Has someone tried this or is there a compiled installer for PulseView available somewhere? Show Less

I'm struggling with a CX3 on a custom design that comes up properly on the USB bus (2.0) when powered on with PMODE pins properly set but appears to be having its internal pull down resistors enabled by default (unexpected or at least I have not seen any hints in any documentation to that effect). The (preliminary) reason for this conclusion is based on the example behavior of the I2C I/Os - While they initially track VIO (1.8V) through their (external) 4.75 kOhm pull-up resistors, they later on drop down to somewhere around 0.9 V and 1.3 V (data and clock respectively) and stay there (see attached scope shot).

I2C SDAI2C SCL

I have already checked whether this might be an issue with the chip reset but couldn't find any evidence (manually issued a hard reset while device was already up-and-running with no change to the I/O behavior).
In addition, I'm also able to successfully download custom firmware into RAM whereupon the device correctly re-enumerates and re-attaches to the USB bus as two devices (one USB 2.0 StreamerExample and one USB 3.0 UVC endpoint). I can also issue I2C (master) transaction but due to the mentioned issue they abort as per the I2C protocol (master monitoring its own data; see scope screenshot).

I2C Bus Master Access

Any other idea(s) on where to look for possible reasons of this behavior?

Show Less

I'm struggling with a CX3 on a custom design that comes up properly on the USB bus (2.0) when powered on with PMODE pins properly set but appears to be having its internal pull down resistors enabled by default (unexpected or at least I have not seen any hints in any documentation to that effect). The (preliminary) reason for this conclusion is based on the example behavior of the I2C I/Os - While they initially track VIO (1.8V) through their (external) 4.75 kOhm pull-up resistors, they later on drop down to somewhere around 0.9 V and 1.3 V (data and clock respectively) and stay there (see attached scope shot). I have already checked whether this might be an issue with the reset but couldn't find any evidence (see attached scope shot).
In addition, I'm also able to successfully download custom firmware into RAM wherupon the device correctly renumerates and re-attaches to the USB bus as two devices (one USB 2.0 StreamerExample and one USB 3.0 UVC endpoint). I can also issue I2C (master) transaction but due to the mentioned issue they abort as per the I2C protocol (master monitoring its own data; see scope screenshot).

Show Less

Hi, I implement the "Spi-Firmware update" feature in my project or u can say enable the Write/Read on Spi Flash. On application side I am using the BeginDataXfer() api for data-transfer on control endpoint. like this :

 

And corresponding Firmware implementation is done by referring example comes with sdk.

under the vendor request

 Why this BeginDataXfer() struck,  never return  the error or success status ??
 what changes should I do on application or Firmware side is needed ?

Show Less

I am trying to open cx3config.cycx in KBA236855.

If open it before importing the project into EZ USB suite, it shows error as : 

 

If open it after importing the project into EZ USB suite, it shows error as : 

Then, I created a new CX3 configuration project, and set parameters as shown in thread: https://community.infineon.com/t5/USB-superspeed-peripherals/The-problem-with-AR0234CS-sensor-KBA236855/td-p/651267

But it reports error as below,

My questions,

1. what is wrong with my configuration for AR0234 to stream  video 1920x1080 @100fps as in KBA236855?

2. Can you send me a copy of cx3config.cycx file that I can open in EZ USB suite?

3. Can you help send me a .cycx file or screen shot CX3 MIPI configuration for work with AR0234 to steam  video 1920x1200@120fps? I tried several configurations , it keeps reporting error. 

Thank you very much. 

Kelly

Show Less