librambutan/core, branch newprint [no description] https://git.bnewbold.net/librambutan/atom?h=newprint 2010-05-21T02:47:59+00:00 moved the usb directory to its proper home in ./libmaple 2010-05-21T02:47:59+00:00 AJM poslathian@poslathian.(none) 2010-05-21T02:47:59+00:00 urn:sha1:dcc4c457f0cad15b9d60e8fc3afa913bacacdbc6 also, removed some old usb file, bootVect.h, which setup the static table for the runtime usb lib that no longer exists and was provided by the bootloader rev 1 added a delay to the resrt slide to allow enough time for graceul closures of the serial port 2010-05-21T02:09:16+00:00 AJM poslathian@poslathian.(none) 2010-04-25T21:42:34+00:00 urn:sha1:ba9ff3861bbcc17c78086b1887691c324f13ba0e fixed the receive 0 bytes bug, where full buffers block the usb port and never get re-enabled. 2010-05-21T02:09:16+00:00 AJM poslathian@poslathian.(none) 2010-04-25T20:13:07+00:00 urn:sha1:fc4b9b386a80d06c73e03d70767643aae3961e2a added carraige feed to newline send on the test function usbSendHello 2010-05-21T02:09:16+00:00 AJM poslathian@poslathian.(none) 2010-04-25T19:55:04+00:00 urn:sha1:7b1e00bca888247efddddd28912d51e9ba7112ba fixed the FIFO bug, tx/rx now seems stable and ready for integration. 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-25T19:52:04+00:00 urn:sha1:cd172b2d5ed1ff486eab399e4da5af8315cddc8b theres still a remaining bug whereby fast consecutive read/writes to serial might crash the fifo, but other than that it seems to work. see usbSendHello for an example. include usb.h in your application, and call setupUSB() to turn everything on. 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-25T19:19:41+00:00 urn:sha1:8d6602547bc7beb39652a35a1cfdddc250f683ff FINALLY figured out the tx/rx bugs, so the whole stack is officially working. for future reference its ESSENTIAL that you set the EPNUM flag of the global device struct appropriately regardless of whether or not you enable or configure more endpoints. 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-25T04:51:43+00:00 urn:sha1:4dcbd9612275d7957484f4a87e16f0bb09484ee6 finally got TX working without any delay, there is some weird timing involved in what happens when the line state is toggled (host port disconnected) and theres still data waiting to go out, this will hang any while(txCount != 0) calls. similarly, sending data without checking txCount != 0 will hard fault the chip. This is all handled now by usbSendBytes, which returns -1 for unconnected, and 0 for simply 'waiting to send' 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-25T02:23:50+00:00 urn:sha1:6ad3acd15e9859ac9bc77c1eb04d67108dcc5a03 c++ ified all of the usb_core files, added the auto-reset feature to the virtual com port, got the descriptors in functional although not pretty order that can be fixed using the attribute packing 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-23T03:27:09+00:00 urn:sha1:a86ec7c81d7ad2037e900899a0b32c5592cae7c0 added virtual com port echoing as the test configuration for the usb serial code 2010-05-21T02:09:15+00:00 AJM poslathian@poslathian.(none) 2010-04-20T20:48:29+00:00 urn:sha1:a082defb795cc4495139c8b05de4fd48bd66dacd