Two drives appear to be represented by the same FileX media pointer, and access fails after an overcurrent event.
Hi,
split from this thread USBX Debug Logging | Community
I am running:
- CubeIDE 2.2.0
- CubeMX 6.18.1
- H5 MCU Package 1.7.0
- FileX 6.4.0
- USBX 6.4.0
- ThreadX 6.4.0
I guess I do not understand what the UX_ENABLE_DEBUG_LOG macro does, I assumed it would cause
the USBX stack to generate a slew of debug messages, which I could post-process, and maybe shine some
light on a problem I as trying to debug. Now that I know it does nothing, I can understand why the
code doesn’t compile, as there is no need to enable UX_ENABLE_DEBUG_LOG.
The change I made was just to get it to compile, in usbx/common/core/src/ux_utility_debug_log.c I disabled
the following statements, as the variable ‘source’ does not exist.
#if 0
/* Calculate the string length. */
debug_location_string_length = UX_DEBUG_LOG_SIZE;
_ux_utility_string_length_check(source, &debug_location_string_length, UX_DEBUG_LOG_SIZE);
debug_message_string_length = UX_DEBUG_LOG_SIZE;
_ux_utility_string_length_check(source, &debug_message_string_length, UX_DEBUG_LOG_SIZE);
#endif
Anyways, my problem is as follows:
I have to USB sticks plugged into ports on the USB hub chip. The USB hub and USB sticks (flash drives)
get enumerated by USBX (correctly or not?).
I cause an Over Current event on one of the USB sticks, causing the Hub to disable VBUS 5 volt power to the overcurrent port. Then I try and create,write,read a file on the other USB stick, and it fails? That is my issue, I did verify the VBUS 5V is active for the second USB stick. I was expecting the second USB stick to be able to create,read,write a FileX file.
When I process the Event callbacks from the USBX stack, I get a DEVICE_INSERTION event for both flash drives,
and I save the FX_MEDIA pointer so that I can later use FileX API’s.
Befor the Over Current event, I printed out both of the FX_MEDIA pointers (one for each USB port/stick), and noticed that they were the exact same pointers? I would assume they would be unique pointers. Could there be a bug in the HUB Host class software which causes this?
Thanks
Ken
