| 00:59.29 | CIA-91 | rhombus-tech: master * r02b8a378d89e /allwinner_a10/orders/aleks.mdwn: |
| 01:07.06 | *** join/#arm-netbook hipboi (~hipboi@123.89.251.50) |
| 01:26.02 | CIA-91 | rhombus-tech: lkcl master * rab9cea7680c3 /community_ideas/educational_and_games_mc.mdwn: |
| 01:27.36 | CIA-91 | rhombus-tech: lkcl master * r7e01a51d5dfe /community_ideas/educational_and_games_mc.mdwn: |
| 01:27.49 | CIA-91 | rhombus-tech: lkcl master * re7d462a95206 /community_ideas/educational_and_games_mc.mdwn: |
| 03:00.35 | *** join/#arm-netbook QingPei (~qingpei@61.149.86.35) |
| 06:30.54 | *** join/#arm-netbook hipboi (~hipboi@123.89.251.50) |
| 08:09.20 | MaDMaLKaV | morning all |
| 08:46.11 | hno | found a 2mm x 4 connector populated with 3 wires on an old cdrom audio cable. Perfect for UART connection. |
| 08:47.46 | hno | might work for usb-device connection as well actually. |
| 08:48.49 | MaDMaLKaV | cool, I got a Mele unit |
| 08:49.05 | MaDMaLKaV | I have to mess some stuff with a tegra tablet and then I will be able to start working on the mele |
| 08:49.32 | MaDMaLKaV | why everybody trying to get an uart port for mele, kernel debugging? |
| 08:53.52 | hno | UART helps a lot if you want to do anything else than Android. |
| 08:54.30 | hno | and my intention with the mele is kernel hacking. |
| 08:54.37 | hno | plus bootloader hacking. |
| 08:57.22 | MaDMaLKaV | ok. I hoped to avoid soldering but if it must be... |
| 08:58.45 | hno | soldering? |
| 08:59.10 | hno | you only need soldering to add JTAG. |
| 08:59.19 | hno | UART is on a connector. |
| 09:00.21 | hno | only problem is that it's a small connector with 2mm spacing between pins so most logic level serial cables can't fit without adjustment. (2.54mm spacing) |
| 09:00.49 | hno | MaDMaLKaV, http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/A10_device_a1000_inside.jpg̈́ |
| 09:03.11 | MaDMaLKaV | oh, cool, it is one of those old cdrom cables connector as you said, must have some at home |
| 09:03.36 | MaDMaLKaV | anyway, I will have to solder the other side to my ftdi breakout board |
| 09:03.46 | MaDMaLKaV | -if it works, I messes it badly last time I used it- |
| 09:05.44 | hno | MaDMaLKaV, easy to test it, just jump the TX & TX wires together. Measure that you have 3.3V between them (while connected together) and GND, and that it echoes back anything you sen in the terminal. |
| 09:07.02 | MaDMaLKaV | ok, will try, thanks for the info |
| 09:30.54 | hno | wonder if I have a 5-pin 2mm connector somewhere... |
| 10:22.31 | menomc | hno: is it possible that the uart connector of my mele doesn't go to ttyS0 as yours? |
| 10:34.23 | mnemoc | hipboi: is it possible that by software my mele has serial output disabled so not even your uboot-mmc greet? |
| 10:42.07 | A2Sheds | whats with the guy from Olimex? The voice from the backwoods strikes again! |
| 10:43.45 | *** join/#arm-netbook gordan (~gordan@78-86-195-86.zone2.bethere.co.uk) |
| 11:09.51 | *** join/#arm-netbook QingPei (~qingpei@61.149.86.35) |
| 11:29.14 | zenitraM | mnemoc: even the built-in boot1 should greet |
| 11:29.42 | zenitraM | s/boot1/boot0/ |
| 11:29.48 | zenitraM | http://rhombus-tech.net/allwinner_a10/hacking_the_mele_a1000/bootlog_sata_ubuntu_mini_log.txt |
| 11:30.06 | MaDMaLKaV | heh, just saw on the mailing list that there seems to be a guy from my city |
| 11:30.21 | mnemoc | zenitraM: then it has to be greeting a different wire |
| 11:30.49 | mnemoc | unless my boot0 is not the same as yours... |
| 11:31.50 | mnemoc | what version/date does your board say? mine says 2011-11-20 ab10-g32-v1.20-2 |
| 11:32.04 | zenitraM | i don't have it here but it was the same as in the pic on there |
| 11:32.31 | mnemoc | same :-/ |
| 11:33.13 | zenitraM | and there is only one firmware version (the one on the mele site) |
| 11:33.37 | zenitraM | MaDMaLKaV: the number of spaniards around this project is growing indeed.... |
| 11:33.40 | mnemoc | i ran out of ideas |
| 11:34.21 | mnemoc | also without a fancy cable I can't get adb either... |
| 11:34.29 | mnemoc | to check from the other end |
| 11:35.19 | zenitraM | mmm |
| 11:35.23 | zenitraM | install |
| 11:35.24 | zenitraM | adbWireless |
| 11:35.26 | zenitraM | from the market |
| 11:35.36 | zenitraM | it will enable tcp on adb so you can connect to adb over wifi |
| 11:36.11 | mnemoc | .oO didn't see the market within all that chinese text ... I'll power ir again |
| 11:36.54 | zenitraM | it is there |
| 11:37.03 | zenitraM | you can set it to english and even spanish btw |
| 11:37.24 | mnemoc | ok, I'll explore it deeper |
| 11:38.22 | mnemoc | I never really intended to even have to look at the pre-installed system :-/ |
| 11:38.43 | hipboi | mnemoc: what's the problem you meet? |
| 11:38.51 | zenitraM | given that you can install a terminal app... it is somehow useful |
| 11:39.36 | mnemoc | hipboi: I don't get anything on the serial port, neither on normal boot or using your uboot-mmc |
| 11:40.00 | mnemoc | I didn't know I can install a terminal app, I'll do that and test the ports from android |
| 11:40.13 | hipboi | i think the internal uart connector match the cable i sent to you |
| 11:40.43 | mnemoc | the one for the breakout board, 1m |
| 11:40.52 | hipboi | yes |
| 11:41.03 | hipboi | do they match? |
| 11:42.17 | hipboi | can you show a picture how you connect them? |
| 11:42.32 | mnemoc | I don't have it here, I will go to search for it now |
| 11:43.34 | hipboi | if i am not wrong, the usb cable should fit the uart connector on the pcb perfectly |
| 11:44.20 | hipboi | if you connect them, no print, it must be soldered wrong direction |
| 11:44.26 | mnemoc | problem is one can't buy those cables over here :) |
| 11:45.15 | hipboi | you have another option, use the breakout card |
| 11:45.26 | hipboi | use a uSD to SD converter |
| 11:45.49 | *** join/#arm-netbook lkcl (~lkcl@host81-132-45-41.range81-132.btcentralplus.com) |
| 11:46.12 | hipboi | compile a u-boot from lichee-dev, it default prints to sd card pin |
| 11:47.35 | mnemoc | found the dongle you sent me, it's connector is 2.54mm while the mele needs 2mm |
| 11:48.34 | hipboi | i see |
| 11:48.43 | hipboi | damn |
| 11:50.56 | A2Sheds | lkcl : sometimes I wonder what the purpose of the wiki, ML and IRC is if people don't read them and just spout off with confused bits of info |
| 11:51.11 | zenitraM | hipboi: shouldn't the default boot0/boot1 output to the uart by default? |
| 11:51.19 | hipboi | yes |
| 11:51.38 | hipboi | boot0/boot1 output to uart on product |
| 11:52.24 | mnemoc | A2Sheds: the tl;dr syndrom |
| 11:54.27 | zenitraM | he's getting nothing from boot0 even |
| 11:54.58 | zenitraM | so i don't think its a sw problem on the mele |
| 11:59.49 | hipboi | uart pins directly connect to the cpu |
| 12:00.07 | hipboi | very less chance to broken |
| 12:00.55 | hipboi | mnemoc: do you have a multimeter? |
| 12:01.11 | MaDMaLKaV | sorry for the offtopic, but on my cellular, using the windows live mail app for android, some messages on the mailing list doesn't display, they appear as a blank mail with an attatched file, always 218 bytes in size, if anyone knows how to fix it, please tell me |
| 12:01.58 | mnemoc | hipboi: yes, tested everything I could with zenitraM and hno last night |
| 12:02.40 | mnemoc | hipboi: but I can test more stuff if you want |
| 12:03.01 | hipboi | so, is there voltage on tx or rx? |
| 12:03.45 | hipboi | first make sure there is voltage on 3.3V |
| 12:04.11 | mnemoc | on the mele Tx,Rx and Vcc to Gnd are 3.3V |
| 12:05.42 | hipboi | oh... |
| 12:05.50 | mnemoc | the dongle is a bit funkier, but when testing Rx+Tx to G also gives 3.3V |
| 12:05.56 | mnemoc | uh? is that bad? |
| 12:08.43 | hipboi | i don't know the voltage of rx tx when working |
| 12:09.07 | hipboi | i will test it Monday |
| 12:11.02 | zenitraM | if the dongle gives 3.3V... i think the mele should also give 3.3V |
| 12:11.17 | hipboi | i think it should be about 3.3v |
| 12:12.31 | mnemoc | zenitraM: does this adbwireless work on normal ethernet too? |
| 12:13.05 | zenitraM | mnemoc: yes, it should |
| 12:14.41 | zenitraM | it's probably called wireless because... well, cell phones don't have ethernet :P |
| 12:14.47 | zenitraM | but i think it just enables adb over tcp |
| 12:15.33 | zenitraM | also there are probably several ssh servers on the market |
| 12:16.31 | mnemoc | first, the language :) |
| 12:19.52 | zenitraM | go to settings... |
| 12:20.01 | *** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk) |
| 12:20.14 | zenitraM | i think there's a shortcut for it down on the home screen |
| 12:20.23 | mnemoc | yes, changed |
| 12:20.42 | zenitraM | you'll have to add your google account to access the market |
| 12:20.58 | mnemoc | there is something that looks like a market but it's not *the* market |
| 12:21.09 | zenitraM | the market is also there, at least it was on mine |
| 12:21.28 | mnemoc | here it is |
| 12:22.01 | mnemoc | using android with a keyboard is odd :p |
| 12:23.29 | zenitraM | it is even more when you plug a mouse :P |
| 12:27.16 | mnemoc | having the market is a great thing, I can install root explorer and all those nice toys too |
| 12:27.44 | mnemoc | I think I have a usb mouse somewhere, /me searching |
| 12:45.18 | MaDMaLKaV | I thought the mele came with some kind of box for the 2.5 sata drive... |
| 12:51.02 | mnemoc | it does |
| 12:51.37 | mnemoc | remote, case for the 2.5 drive and rca cable |
| 12:58.02 | zenitraM | a usb-to-sata for the case would be nice. but that's an add-on :( |
| 12:59.47 | mnemoc | http://dpaste.com/720833/ <--- me trying a echo a "hello" over a serial loopback on the mele.... |
| 13:00.09 | mnemoc | something doesn't seem right |
| 13:00.45 | hipboi | [ 2206.370000] ===> rtw_ips_pwr_up.............. |
| 13:00.56 | hipboi | this is the right output |
| 13:01.11 | hipboi | looks like something goes to input |
| 13:03.56 | zenitraM | seems like output is going to input as is should.. if i'm not wrong |
| 13:04.08 | zenitraM | and you're just getting the kernel messages on there |
| 13:04.16 | zenitraM | yes |
| 13:04.17 | zenitraM | in fact |
| 13:04.24 | zenitraM | the "not found" errors |
| 13:04.33 | zenitraM | are the kernel console errors going to the shell running on the serial |
| 13:05.00 | mnemoc | yes, trying to find the process listening ttyS0 to kill it :p |
| 13:05.17 | zenitraM | try busybox lsof.. if it has it |
| 13:05.41 | mnemoc | no lsof on my busybox :( |
| 13:05.41 | zenitraM | but anyway this means the serial on the mele is working |
| 13:06.21 | zenitraM | so the problem is... the only thing i can think of is the cables not making good contact or something |
| 13:07.04 | mnemoc | very possible |
| 13:07.49 | zenitraM | mmm.. you could try measuring with the multimeter on one side of the cable, with the other plugged to the mele |
| 13:08.58 | CIA-91 | rhombus-tech: lkcl master * r3edfbcf27333 /community_ideas/educational_and_games_mc.mdwn: |
| 13:09.09 | mnemoc | I did, it worked fine :-/ |
| 13:10.12 | mnemoc | any idea who is /dev/ttyS2 ? |
| 13:10.38 | mnemoc | the SD? |
| 13:10.50 | zenitraM | there are a lot of serial devices on script.bin, software-settable but by default only the 0 one is enabled |
| 13:11.00 | zenitraM | probably the rest are "virtual" serials not hooked anywhere |
| 13:11.43 | zenitraM | but yes, you could probably plug a separate ttySX to the SD |
| 13:12.44 | hno | mnemoc, the A10 have up to 8 UARTs depending on config. |
| 13:14.24 | hno | mnemoc, so now we know your cable works, and the mele works. Only need to get them properly connected. |
| 13:16.11 | mnemoc | I will receive "proper" 2.00mm to 2.54mm cables in 2 weeks, I wish I can get working on this sooner than that :-/ |
| 13:17.23 | mnemoc | _the_ electronic parts shop on my city closed up 2 months ago :| |
| 13:17.36 | mnemoc | anyhow, lunch, back in 30m |
| 13:18.37 | CIA-91 | rhombus-tech: lkcl master * r954be0730898 /community_ideas/educational_and_games_mc.mdwn: |
| 13:25.21 | *** join/#arm-netbook gordan (~gordan@78-86-195-86.zone2.bethere.co.uk) |
| 13:37.03 | *** part/#arm-netbook hipboi (~hipboi@123.89.251.50) |
| 13:41.11 | *** join/#arm-netbook QingPei (~qingpei@114.241.180.108) |
| 13:52.35 | CIA-91 | rhombus-tech: Legendinis master * r1daeef076826 /allwinner_a10/orders/legendinis.mdwn: |
| 14:02.52 | A2Sheds | hmm, if we just used uarts for everything it would really simplify the design |
| 14:23.54 | CIA-91 | rhombus-tech: lkcl master * rfb668b5c1b8c /allwinner_a10/orders/powerage.mdwn: order processing cleanup |
| 14:25.46 | hno | A2Sheds, hard to fit everything in the UART bandwidth. |
| 14:26.37 | hno | even if the A10 have quite good high rate UARTs with DMA and everything. |
| 14:27.48 | mnemoc | hits his head on the wall, again and again |
| 14:54.48 | mnemoc | i really fail to understand my problem. the mele and the dongle both are working at 3.3V, both have functional echo loopback, both edges are electrically connected and still can't get them talking :-/ |
| 14:59.59 | penguin42 | this serial? |
| 15:01.17 | mnemoc | wonder where "this" points to |
| 15:01.57 | penguin42 | mnemoc: The thing you're trying to debug |
| 15:02.45 | mnemoc | I'm trying to have a serial console on the mele a1000 |
| 15:02.54 | mnemoc | but it doesn't want to speak with me |
| 15:03.39 | penguin42 | ah, can you get a console after it's booted and does serial work from that? |
| 15:03.55 | mnemoc | neither |
| 15:04.18 | penguin42 | ah, that does make life harder |
| 15:04.44 | mnemoc | it's awfully frustrating |
| 15:04.49 | penguin42 | got a scope? |
| 15:05.16 | mnemoc | only a multimeter |
| 15:05.50 | mnemoc | I've been a software-only ape since I'm 8 |
| 15:06.04 | penguin42 | it's ok, serial doesn't sting :-) |
| 15:06.13 | mnemoc | :) |
| 15:06.36 | penguin42 | so what's the melee currently running? |
| 15:09.06 | mnemoc | android 2.3 with busybox, sy and market |
| 15:09.09 | mnemoc | su |
| 15:10.31 | penguin42 | so you have a shell on it? |
| 15:11.43 | mnemoc | installed a terminal from the market and then enabled adb over tcp |
| 15:12.09 | penguin42 | ok, if you've got a shell then try getting serial working from there before trying to debug boot time serial console |
| 15:13.02 | mnemoc | /dev/ttyUSB0 seems to be tied to a shell somewhere |
| 15:13.35 | mnemoc | works when bridging the Tx and Rx |
| 15:13.44 | mnemoc | but not when using the f* dongle |
| 15:13.54 | penguin42 | which dongle is that? |
| 15:14.36 | mnemoc | a pl2303hx based ttl to usb |
| 15:16.14 | mnemoc | http://www.ebay.com/itm/170728559489 |
| 15:19.02 | penguin42 | so that's from the TTL on the mele to usb on your PC? |
| 15:19.20 | mnemoc | yes |
| 15:19.56 | penguin42 | ok, so other than booting what have you tried from the mele side? |
| 15:20.50 | mnemoc | busybox microcom -s 115200 /dev/ttyUSB0 |
| 15:21.30 | penguin42 | mnemoc: OK, so I've debugged plenty of serial, but don't know the melle - I thought you said the USB was on the PC end? |
| 15:21.36 | mnemoc | also tried booting from an sd with uboot-mmc |
| 15:21.58 | mnemoc | err, yes, wrong paste |
| 15:22.05 | mnemoc | busybox microcom -s 115200 /dev/tty0 :) |
| 15:22.08 | mnemoc | err |
| 15:22.11 | mnemoc | S0 |
| 15:22.38 | penguin42 | have you tried something simpler like echo Hello > /dev/ttyS0 |
| 15:23.22 | mnemoc | good idea |
| 15:23.44 | penguin42 | and how many ttyS* are there - are you sure it's 0? |
| 15:24.00 | lkcl | penguin42: ahhh, that brings back memories :) even if you get garbage (wrong data rate) you still get something out, it tells you the pins are the right way round :) |
| 15:24.22 | penguin42 | nod |
| 15:24.37 | penguin42 | also has a breakout box for when things really don't play ball |
| 15:25.53 | penguin42 | All boards should have one of those ftd chips soldered on board and a nice little micro USB to make life easier |
| 15:27.05 | penguin42 | preferably with the jtag wired as well |
| 15:31.35 | hno | penguin42, yes. I am really considering embedding a FT2232H board in there.. |
| 15:31.51 | mnemoc | I've used some nice boards that export the jtag and ttys0 in adition to usb gadjet and storage over the otg port |
| 15:31.59 | hno | one channel for serial, one channel for JTAG. |
| 15:33.44 | hno | mnemoc, not really. Those provide a virtual serial port gadget on USB, and require a running USB stack on the device. |
| 15:34.09 | mnemoc | ic |
| 15:34.55 | penguin42 | hno: Yeh I've used the sheevaplugs like that - it's nice being able to serial and jtag them remotely; you can do almost all dev entirely remotely |
| 15:37.16 | hno | penguin42, yes. |
| 15:38.24 | hno | mnemoc, for most practical purposes it's good enough however, but you won't be able to use it easily for kernel debugging, several kinds of kernel crashes etc. |
| 15:38.45 | hno | plus that we do not have that on the A10 yet. |
| 15:39.04 | hno | that virtual console over USB is all software. |
| 15:39.38 | mnemoc | I had the hope it was done by a magic IC |
| 15:43.04 | *** join/#arm-netbook tjoy (~dsockwell@184.73.204.96) |
| 15:43.04 | *** join/#arm-netbook Gumboot (~gumboot@rev.bovine.muck.net.nz) |
| 15:45.37 | mnemoc | oh, I can send to the mele ... just fail to receive from it |
| 15:46.13 | mnemoc | sending "stop adbd" did kick me out |
| 15:46.43 | penguin42 | say that again - which way around works? |
| 15:47.02 | penguin42 | you can send from the PC to the mele but not from the mele to the PC? |
| 15:47.24 | mnemoc | right |
| 15:48.16 | penguin42 | and you've got every conceivable bit of flow control switched off? |
| 15:48.41 | mnemoc | gtkterm says flow control: none ... |
| 15:49.18 | mnemoc | not much room for mistakes on that |
| 15:50.16 | mnemoc | may I need to wire the VCC? |
| 15:50.26 | penguin42 | on where? |
| 15:50.35 | mnemoc | mele to dongle |
| 15:50.52 | penguin42 | hmm doubt it |
| 15:51.14 | penguin42 | I'd assumed the dongle is powered by the USB (by I maybe wrong); so you've got gnd and rx/tx wired? |
| 15:51.27 | mnemoc | yes |
| 15:52.04 | mnemoc | and he multitest confirms the mele's Tx is connected correctly |
| 15:52.13 | mnemoc | multitester* |
| 15:52.46 | penguin42 | and what are you using on the mele to get input? |
| 15:53.42 | mnemoc | nothing, there is a root shell I can't see bond to /dev/ttyS0 |
| 15:54.01 | mnemoc | the commands I type on gtkterm get executed |
| 15:54.06 | penguin42 | haha |
| 15:55.43 | mnemoc | but I don't receive anything If I type on the android's terminal, example, echo sucker > /dev/ttyS0 |
| 15:56.11 | mnemoc | or /dev/console |
| 15:56.53 | mnemoc | will try another jumper wire on the mele's Tx |
| 15:57.12 | penguin42 | I thought you said it was working in that direction? |
| 15:57.33 | mnemoc | gtkterm > Mele works |
| 15:57.41 | mnemoc | Mele > gtkterm doesn't |
| 15:57.44 | penguin42 | ah |
| 15:58.14 | penguin42 | parity etc correct on your gtkterm? |
| 15:58.55 | mnemoc | I'm using the old classic 115200 8N1 |
| 15:59.41 | penguin42 | is the mele? |
| 16:01.48 | mnemoc | I'll look at /proc/cmdline after it finishes booting |
| 16:01.55 | mnemoc | replacing the wire didn't help |
| 16:03.24 | mnemoc | console=ttyS0,115200 |
| 16:05.21 | penguin42 | you could try stty -a < /dev/ttyS0 on the mele |
| 16:06.04 | mnemoc | agrees |
| 16:06.22 | mnemoc | http://dpaste.com/720897/ |
| 16:06.24 | penguin42 | can't remember which ones are which |
| 16:06.51 | mnemoc | cs8 <-- 8 |
| 16:07.00 | mnemoc | -parenb -parodd <-- N |
| 16:08.38 | penguin42 | yeh and I think that's also got all the flow control off - looks good |
| 16:15.30 | mnemoc | uhm... I have continuity between G and Tx |
| 16:15.34 | mnemoc | on the mele |
| 16:18.04 | mnemoc | *headdesk* |
| 16:21.55 | penguin42 | try turning the meter around and see if it's still true? |
| 16:24.22 | mnemoc | I'm searching for tape to avoid getting undesired contacts... this are 2.54mm jumper cables without the black head... without any isolation |
| 16:26.20 | mnemoc | on bare eye they don't seem to touch each other.... but I'm not sure anyumore |
| 16:28.19 | mnemoc | hates his clampsy fingers |
| 16:42.49 | mnemoc | penguin42: you were right, connectivity beeping only happens in one direction |
| 16:43.21 | penguin42 | mnemoc: It might not be a real short, it might just be going through the logic in the chip |
| 16:46.19 | mnemoc | but I keep having the same problem :-/ |
| 16:54.47 | mnemoc | one thing I fail to understand is why do I get an echo from the commands I send to the mele... but not their output |
| 16:55.47 | penguin42 | you sure you don't have a local echo somewhere? |
| 16:56.08 | mnemoc | enabling local echo I see characters twice |
| 16:56.23 | mnemoc | so someone else is echoing them |
| 16:56.23 | penguin42 | hmm |
| 16:56.31 | penguin42 | and you've tried doing echo hello > /dev/ttyS0 on the mele? |
| 16:56.52 | mnemoc | yes |
| 16:56.56 | mnemoc | I don't receive anything |
| 16:57.09 | mnemoc | unless it's the f* switch on the dongle |
| 16:57.20 | penguin42 | there's a switch? |
| 16:57.35 | mnemoc | I thought it was for choosing voltages |
| 16:57.52 | mnemoc | it might be loopback |
| 17:03.17 | mnemoc | penguin42: in the pictures on the description of http://www.ebay.com/itm/170728559489 you can see a jumpered switch without any direct hint about it's purpose but in other similar devices it's used to pick 3.3/5.0 |
| 17:05.57 | zenitraM | mmm is the jumper on or did you remove it? |
| 17:06.27 | mnemoc | now I changed it's position |
| 17:10.00 | mnemoc | bridging Rx and Tx doesn't echo in this position :-/ |
| 17:10.04 | mnemoc | now without jumper |
| 17:10.35 | *** part/#arm-netbook QingPei (~qingpei@114.241.180.108) |
| 17:13.12 | mnemoc | \o/ |
| 17:14.50 | mnemoc | f* jumper |
| 17:16.39 | mnemoc | so at the end the Rx/Tx were "for the dongle" not "for the device" .. and the f* jumper needs to be removed |
| 17:16.59 | mnemoc | happy |
| 17:17.13 | zenitraM | great news :D |
| 17:17.16 | penguin42 | good |
| 17:17.45 | zenitraM | now reboot and you'll get u-boot :P |
| 17:19.57 | mnemoc | also noticed ubuntu makes an "mtp-probe" every time I connect it... probably that is what ruins livesuit's capability to detect the tablets |
| 17:20.12 | mnemoc | in FEL mode |
| 17:20.48 | mnemoc | s/noticed/noted/ |
| 17:56.23 | zenitraM | also happens on Mac. probably the flashing code only does USB initialization/handshaking once |
| 18:01.29 | hno | mnemoc, FEL mode uses a different USB vendor,device number and special device type. |
| 18:02.47 | mnemoc | hno: yes, I know |
| 18:04.17 | mnemoc | but for some reason after handing it over to qemu, windows detects the device and install the driver, but livesuit doesn't see the device at all |
| 18:05.32 | mnemoc | but now I don't need to touch the included OS, so it doesn't matter that I can't reflash it :) |
| 18:06.22 | hno | mnemoc, are we talking about mele or some tablet? |
| 18:06.33 | mnemoc | tablet |
| 18:07.22 | mnemoc | afaik the mele is supposed to be flashed from a self-executing SD card |
| 18:07.29 | hno | which version of qemu? |
| 18:08.18 | hno | mnemoc, flashing procedure is the same for all A10. But it's not yet identified how to force it into FEL USB mode. You can flash the tablet using sdcard method as well. |
| 18:09.50 | mnemoc | qemu-kvm-0.14.1 |
| 18:10.15 | mnemoc | good, so I don't need to care about that usb/qemu problem at all :) |
| 18:27.06 | zenitraM | that also happened with parallels at mac at times. seems that the device makes a usb handshake with the host first and then with the guest, which makes livesuit don't work |
| 19:22.08 | hno | mnemoc, qemu-0.14 do not support USB-2.0 which may cause some problems. |
| 19:24.15 | mnemoc | that's indeed a good point |
| 19:32.08 | zenitraM | going to install tinyxp on virtualbox on linux and see if it works with that |
| 19:36.58 | mnemoc | my alergy to Oracle is too strong to try that |
| 19:39.49 | mnemoc | i've been posponing it until I really have the need to reflash one of the devices... but knowing phoenixcard is enough, I assume I won't ever need to :) |
| 20:57.23 | hno | <PROTECTED> |
| 21:02.42 | zenitraM | hno: at least here you can find adapters easily on electronic/DIY supplies stores |
| 21:06.28 | mnemoc | I have over two dozens of http://www.dealextreme.com/p/flat-to-round-power-plug-convertor-3529 around the house :p |
| 21:07.47 | zenitraM | they are at 1 at the chinese store down my street |
| 21:08.31 | zenitraM | also i got one at 0,11 (shipping included) at ebay |
| 21:08.43 | mnemoc | :D |
| 21:12.02 | hno | It's annying, I do know I have two solutions to the power problem in some drawer, just don't find them. |
| 21:13.10 | hno | either the adapter, or an universal power supply with interchangeable connector. Have found the universal power supply, just not the other connectors for it.. |
| 21:15.13 | hno | and probably a travel main power adapter somewhere as well which should include flat to round conversation. |
| 21:15.37 | mnemoc | you can also just cut the power cable and tape it to a cheap 5V power adapter you don't use |
| 21:22.45 | penguin42 | hno: Don;t you just take the wire cutters to the problem? |
| 21:22.48 | hno | I know. Will probably do so unless I find the plugs to the universal adapter tomorrow. I don't like those flat to round adapters, too risky with kids around. |
| 21:23.12 | penguin42 | hno: Just crossbreed the adapter with the correct mains plug with the one with the right DC plug |
| 21:23.33 | hno | exactly. |
| 21:23.57 | penguin42 | hno: Or wire it to a molex connector onto a PC |
| 21:27.10 | mnemoc | does anyone know what arch/arm/mach-sun4i/foo.aa in the source tree is? |
| 21:28.23 | mnemoc | implements this one: |
| 21:28.27 | mnemoc | arch/arm/mach-sun4i/devices.c:125:extern void sw_get_sid(int *a, int *b, int *c, int *d); |
| 21:29.05 | mnemoc | i wonder if something there is blocking the 3.0 port from booting |
| 21:31.09 | hno | Hmm... bad, that's a GPL violation in the kernel tree. |
| 21:31.14 | hno | but it's a small one so can be pruned. |
| 21:35.57 | hno | mnemoc, it's only used for /proc/sid which can be used for reading something from the chip. |
| 21:41.23 | hno | Ha, that's a silly think to keep silent.Just 4 lines. http://fpaste.org/2GYT/ |
| 21:41.48 | hno | silly thing to keep closed. |
| 21:41.54 | mnemoc | hno: btw, when updating your local checkout of the repo please `git pull --rebase` instead of simply `git pull` ... the second produces very annoying merge commits even when there is no interesting change involved |
| 21:43.08 | mnemoc | that's it? doh |
| 21:44.13 | hno | yes that's it. |
| 21:45.43 | hno | maybe it needs some header which is not public, i don't know. |
| 21:46.00 | hno | only decompiled the object file to C. |
| 21:48.13 | hno | mnemoc, for my local trees I prefer to keep merge history. |
| 21:48.44 | mnemoc | ok |
| 21:48.59 | hno | plus that using rebase screws up any branches. |
| 21:49.51 | mnemoc | :) |
| 21:50.18 | hno | public repositories should never be rebased, period. |
| 21:50.50 | mnemoc | git pull --rebase is to rebase the local working tree |
| 21:50.54 | mnemoc | not the remote... |
| 21:51.29 | mnemoc | but saves the remote from getting marks of every time one updates a local working tree |
| 21:51.30 | hno | I know, but I always push my trees public. |
| 21:51.37 | mnemoc | ic |
| 21:52.44 | hno | the remote do not get any marks unless you merge the by pulling/pushing into the remote. |
| 22:03.42 | mnemoc | hno: http://dpaste.com/720976/ too ugly? for replacing that foo.aa |
| 22:04.34 | mnemoc | been precompiled there is already no chance of changing the base of those addresses... |
| 22:05.41 | hno | Looks fine to me. Might want to look into the memory map if there is a hint on what 0xf1c23800 is |
| 22:05.52 | mnemoc | #define IO_ADDRESS(x) ((x) + 0xf0000000) |
| 22:06.41 | hno | Yes, it's somewhere in the I/O area for sure. |
| 22:07.14 | mnemoc | <PROTECTED> |
| 22:07.18 | mnemoc | :) |
| 22:07.23 | mnemoc | no magic anymore |
| 22:07.55 | hno | cool. Except no info on what "sid" is. |
| 22:08.15 | mnemoc | right, one tiny step at the time :) |
| 22:08.38 | hno | ./core.c: { SW_VA_SID_IO_BASE, __phys_to_pfn(SW_PA_SID_IO_BASE), SZ_1K, MT_DEVICE }, |
| 22:08.38 | hno | ./include/mach/platform.h:#define SW_PA_SID_IO_BASE 0x01c23800 |
| 22:08.38 | hno | ./include/mach/platform.h:#define SW_VA_SID_IO_BASE 0xf1c23800 |
| 22:08.49 | mnemoc | doh, ok |
| 22:09.50 | hno | only reference there is to "sid" except for the /proc/sid code in devices.c |
| 22:13.00 | zenitraM | ehr |
| 22:13.06 | zenitraM | i think you can remove it.. |
| 22:13.12 | zenitraM | # cat /proc/sid |
| 22:13.16 | zenitraM | 00000000-00000000-00000000-00000000 |
| 22:13.41 | mnemoc | :D |
| 22:14.31 | zenitraM | might have something to do with the serial number |
| 22:14.45 | zenitraM | (or.. the lack of) |
| 22:16.30 | hno | and hipboi is not here.. |
| 22:29.43 | mnemoc | hno: if you were to reimplement mach-sun4i, what would you use as initial template? |
| 22:30.07 | hno | not sure. |
| 22:46.03 | mnemoc | problem, uImage doesn't boot. option 1) keep blindly trying to find what may not be right or incomplete in the port to 3.0, option 2) wait until allwinner releases their 3.0.8 and compare, option 3) wait until hno gets the mele connected to jtag and debugged, option 4) start mach-sun4i from zero taking the smallest possible chunks of information at the time, option 5) ..? |
| 22:49.34 | hno | Hm.. the mele is now powered but no output on HDMI or VGA. |
| 22:49.43 | mnemoc | s/boot/want to print anything/ |
| 22:49.49 | zenitraM | i have a 3.0.8 uImage hipboi sent to me on the process of trying to repair the mele. but don't think it is useful.. |
| 22:50.36 | zenitraM | hno: reboot it. the boot logo is displayed at least on vga without having to enable it, at least that way you'll see it working |
| 22:51.02 | zenitraM | there's a button on the remote to toggle the video output |
| 22:51.22 | mnemoc | the 1.4.1 firmware image proably has a functional 3.0.8 binary too, but that doesn't help to fix the code tree... |
| 22:52.22 | zenitraM | the button is the second one on the first column. you have to press it for a few secs once booted |
| 22:52.41 | zenitraM | mnemoc: nope, all 2.3 firmwares use the same 2.6.36 afaik |
| 22:53.29 | mnemoc | true, sorry... i confused the mele and my tablet shortly |
| 22:54.37 | hno | no batteries for the remote. |
| 22:54.57 | mnemoc | hdmi comes enabled by default |
| 22:55.22 | zenitraM | and *yikes* composite output |
| 23:16.12 | hno | console works at least |
| 23:38.28 | hno | looks like my mele is slightly broken with no video output. Complains a lot on the console like "[ 263.680000] [DISP WRN] file:drivers/video/sun4i/disp/de_bsp/de/disp_tv.c,line:333: dac 0 short to ground" (and also for dac 1 & 2) |
| 23:39.30 | hno | and this even with nothing connected. |
| 23:44.16 | mnemoc | :( |
| 23:46.02 | hno | stopped complaining after a hard power cycle. |
| 23:46.41 | mnemoc | got display now? |
| 23:46.56 | penguin42 | nice of it to be able to warn you of that |
| 23:47.19 | hno | no display |
| 23:49.09 | hno | but console works, which is the important bit right now. |
| 23:51.06 | mnemoc | i did have an x-files issue with the hdmi output some days ago. the display stopped working entirely whiler messing with the board's uart... the power led didn't want to go on at all, tried with two laptops and two tablets... but after some hours "resting" unplugged it decided to come back to life |
| 23:52.15 | hno | monitor defenitely on. Same I am using for everything. |
| 23:53.46 | hno | it's not an HDCP capable monitor, but would not expect the mele to require one. |
| 23:54.48 | hno | and VGA should surely work. |
| 23:56.03 | hno | odd system this.. it has tcpdump netcat and strace, but not grep. |
| 23:59.07 | penguin42 | does the busybox have a grep builtin? |
| 23:59.33 | hno | thanks, it does |