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.

AM3715: Issue with pusb.exe on Windows 10

Part Number: AM3715
Other Parts Discussed in Thread: FLASHTOOL

I am trying to RAM load my bootloader on to some existing HW using pusb.exe on Windows 10.  I follow the process I have always used, but I see a dialog on the Windows PC stating a USB device has malfunctioned.

When I repeat on a Windows 7 PC, I have no issues. My company has been using this technique to bring up "virgin" boards for many years.

Is there any changes to the pusb.exe tool, or the driver?

I am wondering if MicroSoft has increased security on the USB ports, rendering pusb.exe incapable of RAM loading a Bootloader.

Any insight would be helpful.

  • Hi Raymond,

    I am not familiar with the tool pusb.exe (I barely use Windows as USB host), is it released by TI? any more information about this tool?

    What exact the dialog is on Windows 10 for the malfunctioned USB?

    To understand the changes from using Windows 7 to Windows 10, do you see if the AM3715 USB is enumerated as different device on both Windows versions?

    I assume AM3715 is running in ROM when you use pusb.exe to load the bootloader, is it correct?

  • "pusb" is a tool written by TI. It is used to RAM load an executable into the internal SRAM of the processor. pusb interfaces with the TI ROM code to download the executable.

    My HW is strapped to boot from USB first. To load my executable, I would turn off the HW. I would enter, "pusb.exe -f <filename>", at the dos prompt and hit return. Then I turn on the HW.

    The TI ROM code then downloads the file into the internal SRAM and executes it.

    On Windows 7, I have no issues. However, on Windows 10, I see an error dialog.

    I was wondering if you have seen this, and if you were able to overcome this.

    Thank you

  • This is the dialog...

  • Hi Raymond,

    Did you download pusb tool from ti.com? If so, can you please share the download link?

    Can you please check the Device Manager in Windows to see if AM3715 USB is enumerated differently in Windows 7 and Windows 10? Don't run the pusb tool, just power on your board and check the usb device enumerated in Windows Device Manager.

  • Sorry I havent responded sooner. I downloaded the source code from https://github.com/nmenon/omap-u-boot-utils/.

    Unfortunately that version uses libusb 0.1.  The tool my company uses has been modified to use libusb 1.0. I can give you a version of the tool, if you want.

    In regards to device manager, on windows 10, I see no device appear. I do see the "device manfunstioned" error dialog.

    On windows 7, I see the omap device on device manager for a second.

  • Hi Raymond,

    Raymond Haas53 said:
    I downloaded the source code from https://github.com/nmenon/omap-u-boot-utils/.

    This is on github, not a TI official link. Do you have a download link from ti.com?

  • I think the original tool I used was built for Linux. I then discovered the source code of the tool on github. I don't remember the original link.

    What tool does TI recommand to use to do a usb boot of the processor? If you have such a tool, you should see the behavior I am seeing.

    Thank you

  • Raymond,

    I contacted the author of the tool pusb.exe, he confirmed he wrote this tool long time ago so that he could program Beagle board in Linux. (He was surprised there is still someone using it...) But this tool is not supported by TI.

    Based on some stories I heard, From Build 1803 Windows10 requires all drivers have to be digitally signed, or they will not be loaded. So I guess this causes the USB boot malfunction in your use case.

    TI supports FLASH TOOL for AM3715 (http://www.ti.com/tool/FLASHTOOL), but since the current version v1.06 was released in 2011, I guess the FLASH TOOL would failed for the same reason. TI doesn't have a plan to update the FLASH TOOL for AM3715.

  • I will give the flashtool a try and see if it behaves differently.

    Thanks