**** BEGIN LOGGING AT Fri Apr 29 02:59:58 2016 Apr 29 17:24:02 OpenMoko wiki says that the usb serial connection should show up as ttyUSB1 (given that no other usb serial converters are present), but I only get ttyUSB0. I've had no luck getting the serial console to work on either ttyACM0 (NOR u-boot) nor ttyUSB0 (Neo booted to Debian) Apr 29 17:36:55 lide: probably USB0 is expected since the first interface might have been added to exception because it's usually used for JTAG on this board. Check your dmesg. Apr 29 17:37:01 lide: what speed are you using on USB0? Apr 29 17:37:06 lide: and what terminal emulator? Apr 29 17:37:15 lide: do you disable all flow control? Apr 29 18:00:41 PaulFertser: in Minicom hardware flow control was on by default, software was off. I tried turning hardware flow control off to no avail. Apr 29 18:01:20 I am using usb 3.0 ports, 2.0 was behaving erratically Apr 29 18:02:43 I am booted to Debian right now on Neo with the debug board connected. Apr 29 18:04:49 lide: of course both hw and sw flow control need to be off for usb uart when you use it that way. Apr 29 18:04:53 another thing... I tried flashing u-boot and rootfs but dfu-util gave "no such alternate method" (or some such) error. I don't remember seeing that one before. it tried to use device 0x0000:0x0000 by default so I gave the correct id as a parameter, but got this error. Apr 29 18:05:03 lide: also, take close attention at baud rate and other related settigs. Apr 29 18:05:21 I used baud rate 115200,8N1 as given in the wiki Apr 29 18:05:24 lide: usb 2.0 controllers are usually way more stable, as xhci driver is still buggy. Apr 29 18:05:27 Right Apr 29 18:05:55 I think when it says about "alternate device" you need to ask dfu-util to "reset" it or somesuch. Apr 29 18:06:11 You can use "dfu-util -l" to see if it gets correct list Apr 29 18:06:38 that has been my experience also (both on Windows and Linux), but in this instance usb 2.0 was behaving crazy for some reason, spamming something in my dmesg that I didn't understand (something about debounce, as if it was firing constantly or turning itself on and off rapidly?) Apr 29 18:07:00 I'll check that out Apr 29 18:07:04 Oh, debounce. That might suggest HW issues with your usb cable or debug board. Apr 29 18:07:45 the thing is, the jtag connection is working and I can halt the CPU using the telnet interface (on top of openocd) Apr 29 18:07:52 (on 3.0) Apr 29 18:08:21 3.0 is so ancient Apr 29 18:08:37 But if it works then yes, interface B for UART should work as well. Apr 29 18:08:45 In fact, they both can work at the same time. Apr 29 18:08:46 usb 3.0 I mean :) Apr 29 18:08:49 Ah Apr 29 18:08:52 I thought openocd 3.0 Apr 29 18:40:31 ok now I get: ./dfu-util -a u-boot -d 0x1d50:0x5119 -R -D u-boot-gta02v5-1.3.1+gitr650149a53dbdd48bf6dfef90930c8ab182adb512-r1.bin: "error get_status: error sending control message: Protocol error" Apr 29 18:40:41 lsusb: "Bus 003 Device 013: ID 1d50:5119 OpenMoko, Inc. GTA01/GTA02 U-Boot Bootloader" Apr 29 18:42:02 ./dfu-util -l: Found Runtime: [0x1d50:0x5119] devnum=14, cfg=0, intf=2, alt=0, name="UNDEFINED" Apr 29 18:42:50 I think I saw many more devices before when I ran dfu-util -l... like 7 or so. Apr 29 18:42:59 with different alt=n Apr 29 18:44:26 derp, I didn't notice before that you could give a number to the -a parameter. where might it be documented how the numbers are mapped? Apr 29 18:45:27 and what might be the issue if all the names are undefined Apr 29 18:46:12 lide: are you in NOR or NAND u-boot? Apr 29 18:46:27 lide: i think all alt numbers should be visible in -l Apr 29 18:46:27 I used a precompiled stand-alone dfu-util binary from https://people.debian.org/~timo/NeoFreeRunner/ Apr 29 18:46:30 NOR Apr 29 18:46:53 I haven't even gotten into NAND u-boot afaik... Apr 29 18:47:06 If you see "NOR" there on the LCD, then yes. Apr 29 18:55:53 missed this: " lide: and what terminal emulator?" -- screen, minicom, cu (launched from xterm) Apr 29 18:56:42 lide: I usually use minicom if it's installed and so far it worked nicely both for ACM and for USB serial devices. Apr 29 18:59:36 I actually see NOR in the status bar but the terminal is unresponsive to input Apr 29 19:00:15 (not sure if that is even related to the host being NOR u-boot) Apr 29 19:13:30 when I plug Neo into the usb 2.0 port and go into NOR u-boot menu, after a few seconds the screen turns white and it boots... Apr 29 19:14:01 lide: how do you actually "go into NOR u-boot menu"? Apr 29 19:14:25 I hold AUX while powering the device on Apr 29 19:15:35 Right Apr 29 19:19:25 it looks like I only get 1 device on the usb 3.0 port and multiple on 2.0 Apr 29 19:21:39 new full-speed USB device number 1 ... all the way up to 127 Apr 29 19:21:45 in my dmesg. wtf. Apr 29 19:22:42 also, it rebooted itself again. I was holding AUX because I was expecting it, and now it's in the NOR u-boot menu, dmesg shows 80 new devices having registered, but lsusb shows no Neo at all, nor does dfu-util -l list anything. Apr 29 19:39:56 lide: if you get "disconnected" then it means something is wrong usb-wise. Apr 29 19:46:03 PaulFertser: it does. at times. here's (a glimpse of) my dmesg: http://pastebin.com/raw/W24e3nH6 Apr 29 19:47:08 this only happens in the usb 2.0 port. if I plug it into a usb 3.0 port it "works" but only gives me one device. Apr 29 19:49:41 here is the dmesg after plugging into usb 3.0: http://pastebin.com/raw/QFFRPfm6 Apr 29 19:53:12 I don't get the disconnects & reboots on usb 3.0, usbnet and jtag work, but that's that. I'm guessing something else is wrong then because I can't flash the device. Apr 29 20:23:09 "[17397.935318] cdc_ether 1-1.2:1.0 enp0s26u1u2: renamed from usb0" -- why yes, that is a much nicer name, thank you very much. Apr 29 20:29:03 lide: dmesg looks sane Apr 29 20:29:30 lide: the enp name is supposed to be better because it's persistent and depends on exact bus topology. Apr 29 21:28:53 well... I reconnected the debug board, got into the NOR u-boot menu, and seems like I can flash the device now. Apr 29 21:35:04 here's hoping the rootfs that's taking a million years to download actually works Apr 29 21:36:23 having fun despite the fact that due to a number of obstacles I haven't gotten any real work done yet. god bless unpolished tech! :) **** ENDING LOGGING AT Sat Apr 30 02:59:58 2016