updata CH347 #424
updata CH347 #424
Conversation
Added CH347F, which is a new package version of CH347T. BUG fix: Sometimes writing bit files may fail. Added functions: Added VID, PID and other settings, the default selected model is CH347T.
src/ch347jtag.cpp
Outdated
| if (bus_addr != 0 && dev_addr != 0) { | ||
| cnt = libusb_get_device_list(NULL, &devs); | ||
| if (cnt < 0) goto err_exit; | ||
| while ((dev = devs[i++]) != NULL) { | ||
| if (libusb_get_device_descriptor(dev, &desc) < 0){ | ||
| continue; | ||
| } | ||
| if (desc.idVendor == CH347_VID && desc.idProduct == CH347_PID && libusb_get_bus_number(dev) == bus_addr && \ | ||
| libusb_get_device_address(dev) == dev_addr){ | ||
| if (libusb_open(dev, &dev_handle) < 0){ | ||
| printError("libusb init failed"); | ||
| goto err_exit; | ||
| } | ||
| } | ||
| } | ||
| }else { | ||
| dev_handle = libusb_open_device_with_vid_pid(usb_ctx, CH347_VID, | ||
| CH347_PID); | ||
| } |
There was a problem hiding this comment.
The code used here to search for a device using bus/dev addr is quite similar to libusb_open_device_with_vid_pid implementation. It's maybe possible to merge both to:
- for each device compare vid/pid
- if
bus_addranddev_addrare provided do comparison too
Something similar to:
[...]
if (desc.idVendor != CH347_VID || desc.idProduct != CH347_PID)
continue;
if (bus_addr != 0 && dev_addr != 0 && (libusb_get_bus_number(dev) != bus_addr || libusb_get_device_address(dev) != dev_addr))
continue;
src/ch347jtag.cpp
Outdated
| uint32_t speed_clock[8] = { | ||
| KHZ(468.75), KHZ(937.5), MHZ(1.875), MHZ(3.75), | ||
| MHZ(7.5), MHZ(15), MHZ(30), MHZ(60) | ||
| }; |
There was a problem hiding this comment.
Clock speed/index depends on device (STANDARD_PACK and LARGER_PACK). This must be keep in mind/used here
There was a problem hiding this comment.
Sorry, I don't understand what you mean STANDARD_PACK.
There was a problem hiding this comment.
seems my previous comment was lost.
https://github.com/YTEC-info/CH347-Softwares/blob/main/Datasheet%20%26%20Manual/WCH-CH347-JTAG-Interface(1.3).pdf
explains according to model and firmware the is two mode.
Each of them as specific frequencies.
| for (unsigned i = 0; i < size; ++i) { | ||
| if (ibuf[3 + i] == 0x01) { | ||
| *rptr |= (0x01 << i); | ||
| }else{ | ||
| *rptr &= ~(0x01 << i); | ||
| } |
There was a problem hiding this comment.
Here buffer received is in bit mode: one byte received contains the value for on state of TDO.
So each byte must be reconstructed.
Also rptr is never updated so you always write at the same position.
Why replacing original code?
There was a problem hiding this comment.
This represents the value of each bit on the last byte of TDO, and the value of the rptr binary bit has been changed.
There was a problem hiding this comment.
*rptralways point to the same adress, you have to increment memory position each 8bits.
Also you shift using i, if size == 256the bit will be lost because*rptr` is a 8bit.
Again I don't see (but maybe I'm wrong) any issue with the original code.
There was a problem hiding this comment.
The size here will not be greater than 8. If chip returns 10bit data 0b1100111101, then 347 will be packaged as D4 00 01 CF D2 02 00 00 01, so according the original code size(here is 2) / 8, the last 2 bits will be missed.
src/ch347jtag.cpp
Outdated
| if (bus_addr != 0 && dev_addr != 0) { | ||
| cnt = libusb_get_device_list(NULL, &devs); | ||
| if (cnt < 0) goto err_exit; | ||
| while ((dev = devs[i++]) != NULL) { | ||
| if (desc.idVendor != vid || desc.idProduct != pid) | ||
| continue; | ||
| if (bus_addr != 0 && dev_addr != 0 && (libusb_get_bus_number(dev) != bus_addr || libusb_get_device_address(dev) != dev_addr)) | ||
| continue; | ||
| libusb_open(dev, &dev_handle); | ||
| break; | ||
| } | ||
| }else { | ||
| dev_handle = libusb_open_device_with_vid_pid(usb_ctx, vid, pid); | ||
| } |
There was a problem hiding this comment.
In fact here my idea is to remove libusb_open_device_with_vid_pidand the test in l.159.
There was a problem hiding this comment.
You mean like this?
There was a problem hiding this comment.
Yes!
maybe a break after libusb_open to only select the first when bus_addr and dev_addr are not provided?
Also libusb_free_device_list(devs, 1) must be used just after the loop and before if (!dev_handle) {
There was a problem hiding this comment.
You are right, thanks for pointing out the error.
|
Applied. Thanks @ZhiyuanYuanNJ ! |
Added CH347F, which is a new package version of CH347T.
BUG fix: Sometimes writing bit files may fail.
Added functions: Added VID, PID and other settings, the default selected model is CH347T.