Open broth-itk opened 3 years ago
Background:
With the original i2c-tiny-usb there was a problem with the maximum I2C transfer size. In the past I had a TPM module connected to the adapter and the linux kernel was unable to talk to it.
Anyhow, I fixed the issue with a patch in the usbtiny code (see changelog in https://dicks.home.xs4all.nl/avr/usbtiny/usbtiny-1.7.tar.gz) for details.
In i2c-star that can be accomplished much easier. My test environment uses 1024 bytes for the control buffer:
uint8_t usbd_control_buffer[1024];
This works fine. Tested with libusb-1.0 and a custom control message.
The static reply_buf will need to get adjusted as well:
static uint8_t reply_buf[1024];
Proposal: Increase control buffer size or use #define USB_CONTROL_BUFSIZE 1024
#define USB_CONTROL_BUFSIZE 1024
If you think it's worth the effort to provide the change via fork + pull request, please let me know.
Thanks for your great project :)
Yes, this sounds useful.
Background:
With the original i2c-tiny-usb there was a problem with the maximum I2C transfer size. In the past I had a TPM module connected to the adapter and the linux kernel was unable to talk to it.
Anyhow, I fixed the issue with a patch in the usbtiny code (see changelog in https://dicks.home.xs4all.nl/avr/usbtiny/usbtiny-1.7.tar.gz) for details.
In i2c-star that can be accomplished much easier. My test environment uses 1024 bytes for the control buffer:
uint8_t usbd_control_buffer[1024];
This works fine. Tested with libusb-1.0 and a custom control message.
The static reply_buf will need to get adjusted as well:
static uint8_t reply_buf[1024];
Proposal: Increase control buffer size or use
#define USB_CONTROL_BUFSIZE 1024
If you think it's worth the effort to provide the change via fork + pull request, please let me know.
Thanks for your great project :)