#qi-hardware IRC log for Friday, 2012-02-24

uncloudedxiangfu: is there any source material for the Ben character printed on to outside of the NanoNote case?03:44
xiangfuunclouded, yes. I also try to find out that one. then will let you know. 03:52
uncloudedxiangfu: thanks.  So is the character I used slightly different then?03:53
xiangfuunclouded, it's different. 03:53
jyfl987xiangfu: ftp://ftp.ingenic.cn/3sw/01linux/01loader/u-boot/u-boot-1.1.6-jz-20110719-r1728-add-jz4770.patch.bz207:33
xiangfujyfl987, wow. that u-boot-1.1.6 patch07:34
jyfl987xiangfu: so it could be used for xburst-boot?07:34
xiangfuwhat you mean?07:35
xiangfuxburst-tools07:35
jyfl987yep its xburst-tools07:35
jyfl987xiangfu: i see someone are trying to port debian on ainol novo7(mips edition)07:35
xiangfujyfl987, we have the jz4770 usbboot source code. sent by someone work for Ingenic.07:36
pabs3jyfl987: do you have a link about that?07:36
jyfl987pabs3: do you mean what i said about debian port?07:37
xiangfujyfl987, http://downloads.openmobilefree.net/Ingneic/07:37
xiangfupabs3, it's under 'debian-mips@lists.debian.org'07:38
jyfl987xiangfu: got it07:38
xiangfupabs3, this one: http://lists.debian.org/debian-mips/2012/02/msg00014.html07:38
pabs3jyfl987: yeah07:38
jyfl987lists.debian.org/debian-mips/2012/02/msg00014.html07:38
jyfl987xiangfu: this guys's destination is closed to mine :]07:39
xiangfujyfl987, I have replied that email thread :)07:39
jyfl987xiangfu: just saw that.07:39
jyfl987xiangfu: and i hope those ac100 hacker can give help, i have an ac100 which have ubuntu installed,07:40
xiangfujyfl987, what is 'ac100'? I think all you need is JZ4770 program manual. :D07:41
whitequarkxiangfu: toshiba ac10007:41
whitequarka very crappy ARM netbook, which is also very cheap07:41
jyfl987whitequark: you got it07:41
whitequarkbut the GPU is unsupported by Linux07:41
whitequarkwhich is why it's crappy07:41
whitequarkit is also proprietary and undoc'd07:41
xiangfu:)07:41
jyfl987whitequark: well it has tegra for linux07:41
jyfl987whitequark: so its now have supported GPU :]07:42
whitequarkjyfl987: maybe you have other ac100? because a lot of people in my country bought it07:42
whitequarkand then found out that it won't actually work good07:42
whitequarkmaybe there was a somewhat working blob07:42
whitequarkbut no proper driver07:42
whitequarkI don't really remember, but there was a lot of disappointment over it and when I checked, it was awfully closed07:43
jyfl987whitequark: you can go to #ac100 for help, i just use their ppa source at launchpad to get support07:43
whitequarknah, I don't have one07:43
xiangfujyfl987, you needs jz4770 program manual. and maybe you can re-use the code from 'http://projects.qi-hardware.com/index.php/p/qi-kernel/'.  there is v3.2 support for jz4740. 07:43
jyfl987xiangfu: how about SIMD? since ingenic use mips isa, its not a problem for me to learn that but SIMD is just another thing07:44
xiangfujyfl987, well. step by step. 1. maybe your jz4770 boot linux 2. then make it works better. :)07:45
xiangfus/maybe/make07:45
jyfl987xiangfu: ok07:46
jyfl987xiangfu: just wonder if your kernel could use the kernel module provide by android kernel?07:46
whitequarkjyfl987: simd is not really documented but somewhat reverse engineered07:46
jyfl987via ABI07:46
jyfl987whitequark: can you show me a link?07:46
whitequarkjyfl987: sure, wait a few minutes07:47
whitequarkjust fyi: in the mplayer port, they first assemble an accelerated .c codec to the asm source (.s)07:47
whitequarkthen insert SIMD instructions with `.byte' and a weird AWK script07:47
jyfl987whitequark: i have download that, if i got an exists document , i wont want to rework :]07:48
xiangfu( if your kernel could use the kernel module provide by android kernel) jyfl987 you mean the ben nanonote kernel?07:50
xiangfujyfl987, ben nanonote use upstream kernel. and some patches. you can find those patches at openwrt-xburst.git.07:51
xiangfujyfl987, sorry. I never look into android kernel. 07:51
jyfl987xiangfu: yes, i mean your qi-kernel 07:51
jyfl987xiangfu: if it could, then it wont need the source code for touch screen mcu driver07:52
jyfl987xiangfu: just for my target, :] , i know its unacceptable for you 07:52
whitequarkjyfl987: http://www.gp32x.com/board/index.php?/topic/50216-jzsimd-mxu-instructions-list/07:57
whitequarkthis07:57
whitequarkand also this would help you: http://mosquito.dyndns.tv/freesvn/trunk/nanonote/forth/xburst-media.fs07:57
jyfl987whitequark: aha, its a forth?07:57
jyfl987i love forth system07:57
whitequarkjyfl987: er, I think that's some unrelated example code07:58
jyfl987whitequark: thanks very much07:58
whitequarkthe CPU is very certainly not a forth machine ;)07:58
jyfl987whitequark: yes, maybe we could got a try of chunk moore's green array :]07:59
Action: jyfl987 afk07:59
qi-bot[commit] Xiangfu: m1 patches: add fix-vga-vsync-pulses.patch (master) http://qi-hw.com/p/wernermisc/91da74709:34
xiangfuwpwrak, I git push my commit on video parameters. :)09:34
xiangfuhave any idea on test plan of Milkymist One: http://en.qi-hardware.com/wiki/Milkymist_One_Firmware#Test_Plan ?10:40
wolfspraulxiangfu: nice, test plan! :-)11:50
wolfspraulthat's my only idea so far11:50
wolfspraul:-)11:50
wolfspraulTEST PLAN - YEAH!11:50
wpwrak(usb issues) lemme write a post ...12:13
T44hi13:09
T44xiangfu: zear just told me that you might be able to help me, i am looking for someone that can read chinese13:10
T44just a few words13:10
T44from a schematic design13:10
xiangfuyes13:10
T44http://uphotos.eepw.com.cn/jiajunpeng/pics/1249971381_0.jpg13:11
T44not sure if there is anything helpful13:11
T44google translate helps me through various forum threads, but it cant help with pictures :)13:12
xiangfustill loading the picture. 13:13
zearT44, it's relatively easy to transcribe the characters13:14
zearthe key is to count the number of strokes in each character and then look in lists of characters with that number of strokes13:15
xiangfuleft top (,þ¸ÅqžE5ïïýe) This drawing is only a reference, may be slightly different with the product.13:15
zearalthough it's probably easier for japanese where they use only ~10 000 kanji, so the lists are relatively small13:16
T44xiangfu: heh, actually already figured as much, the schematic doesnt exactly match the device :)13:16
xiangfuright(Ý13:16
xiangfuright(#h:þ) the buzzer13:17
T44aaah, thats what the mystery device is :)13:17
T44that part didnt make much sense13:17
viricxiangfu: what terminal emulator do you use in your workstation?13:17
viricI wonder with what font13:17
DocScrutinizer51mhm, panasonic.eu adverises mobile phone with 5730 modem chip13:18
DocScrutinizer51might actually be semi-interewsting at least for me13:18
xiangfubottom(Ñ^ÊæG:þ)  Metal ring and the contacts film13:19
xiangfuviric, gnome-terminal. but I am use the xchat.13:19
DocScrutinizer51*allegedly* I couldn't flash any sw I build here locally to any MP device13:19
T44right... thanks13:19
xiangfuviric, font. I am use a Chinese font. :)13:19
viricxiangfu: ah, do you use antialiased fonts in the terminal?13:20
T44that wasnt very useful unfortunately :)13:20
DocScrutinizer51still I could know what the modem is *actually* doing13:20
T44xiangfu: thanks for your help, guess i'll just have to try to understand the rest of the schematic :)13:21
xiangfuviric, I am use a free Chinese font: (http://wenq.org/) seems there is no english page. included by most linux system.13:21
T44its a very confusing design13:21
xiangfuviric, ( antialiased fonts) what this mean? 13:22
viricxiangfu: not only with foreground and background color, but also with pixels within the range13:22
xiangfuT44 understand the rest of the schematic. yes. :)13:22
viricxiangfu: http://www.graphicsacademy.com/what_antialiastext.php13:22
DocScrutinizer51anybody interested in the precise URL?13:22
whitequarkDocScrutinizer51: what could I (or any other hw hacker) do with that phone?13:23
whitequarkit's tivoized anyway isn't it?13:23
DocScrutinizer51NFC13:23
T44bye13:24
DocScrutinizer51afk for 60s13:24
whitequarkNFC as in "near field communication"?13:24
DocScrutinizer51back13:24
wpwrakdoes NFC have any use outside marketing ? ;-)13:25
wpwrakwell, marketing and horror13:26
xiangfuviric, I think my is "Anti-Aliased" font13:26
xiangfus/my/mine13:26
viric:) ok13:26
DocScrutinizer51NoF*Clue13:27
viric(I dislike antialiased glyphs for terminals)13:27
wpwrakas in companies proposing some very scary lose-your-money-fast (codename "payment") systems13:27
wpwrakDocScrutinizer: heh ;-)13:27
DocScrutinizer51~dict tivoize13:27
whitequarkDocScrutinizer51: bootloader which checks rsa signature13:28
whitequarkor: the great strawman of gpl313:28
DocScrutinizer51http://smartphones.panasonic.eu13:28
xiangfuviric, I guess wpwrak is using "Non Anti-Aliased" font :-)13:29
DocScrutinizer51actually has NFC ;-P13:29
viricxiangfu: why would he?13:29
wpwrakxiangfu: yeah. mainly X11 fonts 7x14 and 6x10 ;-)13:31
xiangfuI saw his screenshot only. 13:31
xiangfus/only/once13:31
DocScrutinizer51modem: http://www.stericsson.com/products/m5730-thor.jsp13:33
whitequarkDocScrutinizer51: as you're now into that modem stuff, can you enlighten me...13:37
whitequarkI ran `strings' (and IDA, but that was less interesting) on modem firmware, and it looks like it's some 20 year old codebase from ten different companies including stericsson, siemens, someone bought by intel and more13:38
whitequarkall loosely screwed together and patched three times in a row13:38
whitequark(on Samsung Galaxy SII modem fw)13:39
DocScrutinizer51ooops 5780 inside, not 573013:41
viricxiangfu: ah ok :)13:41
viricDocScrutinizer51: what's good in a 5730?13:41
viricor 578013:42
DocScrutinizer51the fact that I know it 'from inside' :)13:42
viricahh13:43
DocScrutinizer51actually I'm the one working on the tickets against this modem here 13:43
jivsHi15:05
rz2kwhitequark: you and grindars published any info about that coding horror on #xda-devs or #cyanogenmod channels/forums? wonder if anyone did same to sgs2 modem.15:05
jivstrying to get dirtpan working with the latest qi-image15:05
jivsthe izcoordinator seems to run fine, but when i run iz assoc with another ben, there is no resonse15:06
jivscan anyone  here plz help to debug the problem15:06
wpwrakdoes dmesg reveal anything ?15:10
jivsit doesn;t load spi_atben.ko, so i am manually doing insmod 15:11
jivsafter that it says spi_atben is ready for mischief15:11
jivsis there anything else i could check15:11
wpwrakhmm, with spi_atben not loaded, you could check whether communication works at all - by using the tools from ben-wpan15:15
jivshttp://pastebin.com/xTsLagaf  :-result of lsmod after doing insmod15:15
wpwrakthat looks reasonable15:16
wpwrak(tools) e.g., atrf-txrx15:16
jivsok15:17
jivshow imp. are kmod-fakehard and kmod-fakelb for atben to work?15:18
jivsiz listphy doesn;t show any interface w/o these packages15:19
wpwrakjivs: i think they don't matter at all15:19
wpwrakoh, are these devices that worked before ? or are they new ones ? (the atbens and/or the bens)15:20
jivsand shows wpan-phy0 and wpan-phy1 with those packages15:20
jivsthe atbens and ben work well when we use the dirtpan compiled separately in a kernel15:21
jivsnow we are trying to get it working with new image 15:21
jivssilly question, is iz required at all for dirtpan to work now?15:21
wpwrakah, then there's no point in testing with atrf-txrx. sorry, didn't realize what you were doing15:21
jivsokay15:22
wpwrakyes, you still need the IEEE 802.15.4 link layer15:22
jivsokay15:22
jivscould it be any thing missing on kernel or any package missing?15:25
wpwrakdon't know. nothing obvious comes to mind15:27
wpwrakyou seem to have everything i can think of15:27
jivshas anyone ever got two atben working with the new qi image?15:27
jivsthe izcoordinator seems to run fine and listen for clients but clients never recieved any short address15:28
wpwrakdoes the system have an MMC driver running ? if yes, you'd have to disable it before doing anything with WPAN15:29
jivsi tried to run izcoordinator on ben with new image and iz assoc on ben with old kernel, doesn;t work15:29
jivshmm15:29
jivshow can i check that?15:29
jivsi have kmod-mmc_3.0-1_xburst.ipk kmod-mmc-spi_3.0-1_xburst.ipk installed on ben now15:30
wpwrakecho jz4740-mmc.0 >/sys/bus/platform/drivers/jz4740-mmc/unbind15:30
wpwrakshould disable it15:30
wpwrakalso ks8995 looks suspicious15:31
jivsdoes it give any message?15:31
wpwrakand the at86rf230 driver is compiled as a module ? if yes, it seems that you didn't load it. but then i don't understand where your wpan-phy0 comes from15:32
jivswpan-phy0 might be coming from fakehard as well15:32
wpwrak(mmc driver) hmm yes, it should spit out some errors because it can't figure out what to do with atben15:32
jivsdidn;t give any error for me15:33
wpwraki'd get rid of fakehard. it can only make things worse :)15:33
jivsat86rf230 driver will be displayed on lsmod?15:33
wpwrakassuming it's a module, yes. does  dmesg | grep -i rf23  show anything ?15:34
jivsi doubt i have at86rf23015:34
jivsnothing15:34
wpwrakthen there's very little reason for atben to do more than sit there, looking pretty ;-)15:35
wpwrakbut you also have to get rid of the conflicting drivers. anything else that uses MMC will clash.15:36
jivslsmod has mmc_spi15:36
wpwrakperhaps just unloading the critters will make them clean up properly15:36
jivsdo you thing that might be an issue?15:36
antoniodariushwpwrak, do you have a list of the minimum required packages that we need to select from the toolchain to get it working15:36
wpwrakmmc_spi is a little strange. not sure what it tries to do. i'd get rid of it.15:37
jivsokay15:37
wpwrakantoniodariush: i don't how things map into packages. i always build things from scratch, without going though openwrt15:38
antoniodariushwe used to do that as well but we'd like to avoid that and get it working out of the box15:40
wpwrak(mmc_spi) that's an MMC driver that works on top of SPI. what's weird about it is that we have a perfectly good MMC controller. do if it uses the same pins, it would be a grossly inefficient alternative to the proper controller (and driver). if it uses other pins, i'd kinda wonder which.15:40
wpwrak(mmc_spi) so that's another item where nothing good can come from having it15:40
jivsi am recompiling w/o these packages now15:41
jivsthe at86rf230 driver is selected on the toolchain15:41
wpwraktoo bad xiangfu is probably asleep now. he must have tested the packaging and have a working sequence.15:41
jivsbut i am not sure if it is being build/loaded properly15:41
wpwrakks8995 is that sdio-wlan device, right ? do you actually have that ? if not, one more item to get rid of15:42
jivsno we dont use that 15:43
wpwrakokay. so trading all the devices that don't make sense for the one that does ought to improve the situation :)15:43
jivs:-)15:44
wpwraki don't know if there are internal dependencies in the openwrt build process that would knock out at86rf230. chances are that it's just not loaded15:44
jivsis spi_gpio is needed?15:45
wpwraksholdn't be. atben talks to the ports directly15:55
wpwrakwell, spi_atben15:56
wpwrakah yes, that's another driver that's missing15:56
jivsok15:57
jivs[  561.980000] at86rf230: probe of spi32766.0 failed with error -1615:59
jivsactually when we are doing insmod on at86rf230.ko, it is giving that error16:00
jivshints anything?16:00
wpwrakthat's EBUSY. perhaps something else is in the way ?16:18
jivsgot rid of that error with echo jz470-mmc...16:25
jivstesting further now16:25
jivsyeh it works now16:33
jivsgot short address from coordinator16:33
jivsthanks wpwrak 16:33
jivswill test dirtpan now16:34
wpwrakwhee ! :)16:35
jivsthanks.. :-)16:39
wpwrakso it all works now ?16:50
antoniodariushwpwrak, so far so good we only have a problem on the dirtpan now17:00
antoniodariush"/dev/net/tun: No such file or directory"17:00
Action: zenlunatic just got his BNN... about to unbox19:09
zenlunaticis the screen supposed to be a little flickery?19:17
viriczenlunatic: not that I remember. but maybe my retina has a longer integration time19:22
virictalking about flickering...19:23
viricThese christmas, the governmet put the street decoration lights a lot using LEDs19:23
viricwhich had such a power supply, that were lighting only half of the 50Hz AC cycle19:23
viricand a long street blinking at 50Hz flashes instead of 100Hz was quite nocieable19:24
viricnoticeable19:24
viric(I exposed the opposite way I was experiencing and deducing the situation)19:24
rz2klol19:26
zenlunaticseems to be a driver problem19:34
zenlunaticcool... wheezy sees it19:35
DocScrutinizer51whitequark: I have no idea what Samsung flashes to the modem, but I dwefinitely not seen such crap you mentionee, anywhere in gthe image I build every day20:26
DocScrutinizer51whitequark: maybe OSE has some special curls for Siemens etc, and this might show up in a binary20:28
DocScrutinizer51whitequark: also are you sure you've not scanned the APE stuff Samsung added? you know, I'm just talking about modem20:30
--- Sat Feb 25 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!