I occasionally get my libftdi application locking up with an error on ftdi_write_data(). I would like to recover from this automatically, but I don't know what caused the error. The code for ftdi_write_data() calls libusb_bulk_transfer().
If it gets an error it returns a generic (-1).
Would it be reasonable to pass through the error from libusb_bulk_transfer() directly so I can diagnose the error?
Here’s a snippet of the existing code. I’ve highlighted the problem area in yellow.
int ftdi_write_data(struct ftdi_context *ftdi, const unsigned char *buf, int size)
{
int offset = 0;
int actual_length;
if (ftdi == NULL || ftdi->usb_dev == NULL)
ftdi_error_return(-666, "USB device unavailable");
while (offset < size)
{
int write_size = ftdi->writebuffer_chunksize;
if (offset+write_size > size)
write_size = size-offset;
if (libusb_bulk_transfer(ftdi->usb_dev, ftdi->in_ep, (unsigned char *)buf+offset, write_size, &actual_length,
ftdi->usb_write_timeout) < 0)
ftdi_error_return(-1, "usb bulk write failed");
Could we change to something like this where we pass through the specific negative error code from libusb_bulk_transfer()? It seems that it might make it easier to figure out how to recover from
an ftdi_write_data error.
int err;
if ((err=libusb_bulk_transfer(ftdi->usb_dev, ftdi->in_ep, (unsigned char *)buf+offset, write_size, &actual_length, ftdi->usb_write_timeout)
< 0)) {
Ftdi_error_return(err, “usb bulk write failed”);
}
Any suggestions for how to recover from ftdi_write_data() errors would be appreciated. It seems like knowing the lower level error code might be helpful..
Currently, I just flush the buffers but it rarely fixes the problem.
Kind regards,
--
Rick Walker
