**** BEGIN LOGGING AT Mon Jan 25 02:59:57 2010 Jan 25 16:22:15 pytey_ ?? Jan 25 19:11:31 Hi, does anyone have any experience running ocd on an ARM host? Jan 25 19:12:37 I'm trying to bring up OCD on the beagle board, using the ft2232 based tincan tools flyswatter. Jan 25 19:13:44 I get a "usb bulk write failed" from the ft2232_write call in flyswatter_init(). Jan 25 19:15:19 The same config works fine on a PC host, and the flyswatter's serial port seems to be fine run from the beagle (vB4) over the OTG USB port through my USB hub. Jan 25 19:19:55 AlTheKiller: uh why would you want to run openocd on the beagle ? Jan 25 19:20:45 Prototyping a portable debugger. Jan 25 19:21:18 which ftdi lib are you trying to use? Jan 25 19:21:32 The open one, v0.16 Jan 25 19:22:00 AlTheKiller: and you have cross compiled the lib for arm? Jan 25 19:22:44 Yes, I've updated .bb files for libftdi and OCD under openembedded. Jan 25 19:22:55 The linker doesn't complain or anything Jan 25 19:23:11 AlTheKiller: you have procfs and usbfs mounted? Jan 25 19:23:23 Yes Jan 25 19:24:16 are you testing it running as root? Jan 25 19:24:31 Sure am. Jan 25 19:25:09 how are you powering your beagle? via a wall-wart? Jan 25 19:25:48 Yeah, peripherals through a powered usb hub. Jan 25 19:25:59 *and peripherals ... Jan 25 19:27:43 you've covered all the basic items Jan 25 19:27:50 :) Jan 25 19:27:56 That's why I'm bothering you. Jan 25 19:28:38 i assume you have the correct libusb installed as well Jan 25 19:28:47 I'm not sure where to dig, I don't think this is an OCD problem per-sey. Jan 25 19:28:59 More likely a lftdi lusb one. Jan 25 19:29:36 Yeah I've got lusb1 1.04 and lusb-compat 0.1.12 Jan 25 19:30:42 I was hoping it was the lusb > lusb1 +lusb-compat jump. But I dragged the sources out of my OE work directory and configured them on a PC host and all was well. Jan 25 19:32:20 i suspect it might be an endian issue with the lftdi since it hasn't been tested to any extent on arm systems Jan 25 19:35:11 Seems plausible. Jan 25 19:35:37 I'll have a dig, thanks. **** ENDING LOGGING AT Tue Jan 26 02:59:56 2010