IRC log for #arm-netbook on 20120330

01:58.44*** join/#arm-netbook QingPei (~qingpei@114.241.176.48)
03:53.29*** join/#arm-netbook hp__ (~hp_@p5B38A957.dip.t-dialin.net)
04:51.53*** join/#arm-netbook dukie (~dukie@djkie.static.corbina.ru)
04:52.40*** join/#arm-netbook dungeonduke (~dukie@djkie.static.corbina.ru)
04:53.44*** join/#arm-netbook dungeonduke (~dukie@djkie.static.corbina.ru)
06:36.27*** join/#arm-netbook acidfu (~nib@modemcable094.94-70-69.static.videotron.ca)
07:02.04hp__anyone flashed melee with the latest Linaro?
07:57.34*** join/#arm-netbook lkcl (~lkcl@host217-43-148-3.range217-43.btcentralplus.com)
07:57.34*** join/#arm-netbook poptire (~quassel@64.31.59.70)
08:10.50*** join/#arm-netbook acidfoo (~nib@modemcable094.94-70-69.static.videotron.ca)
08:54.21mnemochp__: a good thing about the a10-based devices is that you can boot and keep the root in the SD card, without risking the preinstalled OS at all
08:54.48mnemochp__: so you don't need to flash the device to try linaro's rootfs
09:00.11*** join/#arm-netbook arokux2 (~arokux@merkur178.inf.uni-konstanz.de)
09:01.14*** join/#arm-netbook arokux1 (~arokux@merkur178.inf.uni-konstanz.de)
10:46.28hp__mnemoc cool
11:17.48*** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk)
11:36.17CIA-91rhombus-tech: Omar master * r9c1afa901627 /allwinner_a10/orders/omar.mdwn:
12:01.23hp__CIA-91 ?
12:04.45mnemochp__: notifications of changes on the wiki
12:21.20hp__mnemoc just wondering.. do you think windows8 will ever run on the A10? :)
12:29.17mnemocnot interested at all on anything made by microsoft or apple
12:30.49hp__:)
12:59.40*** join/#arm-netbook Gumboot (~gumboot@rev.bovine.muck.net.nz)
13:22.00*** join/#arm-netbook eflatun (~eflatun@46.1.24.219)
14:04.11*** join/#arm-netbook QingPei (~qingpei@114.241.189.145)
15:00.35traeakcool thx, decided to grab uart board as well, got that all taken care of thorugh tom
15:01.51mnemoc:)
15:06.04traeaknow finding time....ugh
15:57.48traeakthis system is absolutely much cooler and far more useful to me than a raspberry pi
15:58.05*** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk)
16:11.01hp__traeak why do you need a uart board?
16:15.37traeakhp__: serial console ?
16:16.05traeakhp__: misname...wait one TTL to usb debug board is name of it :-p
16:18.24*** join/#arm-netbook destinal (~chatzilla@ip68-108-26-102.lv.lv.cox.net)
16:34.37CIA-91rhombus-tech: lkcl master * r3ebc07f44adf /allwinner_a10/orders/ (Slipcougar.mdwn mmcv2008.mdwn): order cleanup
16:35.59CIA-91rhombus-tech: lkcl master * rb1a2fb0a6837 /allwinner_a10/orders/ (oldwanderer.html oldwanderer.mdwn): order processing cleanup
16:51.07*** part/#arm-netbook QingPei (~qingpei@114.241.189.145)
17:21.45*** join/#arm-netbook destinal (~chatzilla@ip68-108-26-102.lv.lv.cox.net)
17:35.43steevhp__: i thought windows 8 would require an A15
17:36.15steev(A15 as in Cortex A15, with A10 being a Cortex A8)
17:39.26traeaki think windows8 just needs armv7
17:39.32traeakforget how much ram that is
17:40.26steev512
17:41.11traeakso min 512MB for windows8
17:41.27traeaknot that i care honestly
17:41.40traeakbut i dont' wantto see MS destroy another market like they did with netbooks
17:41.54traeakalthough linux wasn't really quite ready for netbooks when they were released
17:43.49steevnot really ready currently
17:45.12steevdevelopers of the majority of desktops use 23" monitors, and don't design for anything but a larger resolution.  using many programs is really bad at present
17:47.10mnemocthe netbooks market died because the hardware was cheap junk
17:47.58hnonetbooks marked died because hardware worth the name costs more than people are willing to pay when there is similar cheap junk.
17:48.32mnemoc:)
17:48.58traeakheh
17:49.00hnoplus vendors played a Windows like attitude "we know what you want to have, don't think anything else"
17:49.05traeakat home i pretty much just use netbooks
17:49.15traeakarch linux on them
17:50.06steevhno: hey! i work for a netbook vendor, we do not!
17:50.17mnemocthe netbooks I tried were hardly capable of opening any modern web app
17:50.26steevactually our netbooks are our best sellers
17:50.36mnemocand those bending keyboards were unusable
17:50.55traeaki really don't do work on netbooks
17:51.15traeakbut they work just fine for web browsing, email, and yeah i do some side projects on them for coding
17:51.17traeakveyr small
17:51.22steevi program on them, and occasionally do native compiles, but i don't use them while i compile on them
17:51.29hnosteev, that's cool, you surived the death of the netbook.
17:51.36mnemoc:D
17:51.45traeaki've got one of the original acer aspire ones with the crappy ssd (kitchen computer)
17:51.46steevhno: only cuz ours doesn't suck :P
17:51.50penguin42I've done builds on a netbook; it's not great - but with 1GB of RAM and a hard drive it's usable
17:51.50traeakand then one with a hard drive
17:51.57traeakbatteries are finally toasting on them
17:52.08mnemocsteev: url to yours?
17:52.10traeakyeah, i upgraded them both
17:52.20steevmnemoc: i work for Genesi
17:52.29mnemocsteev: :D
17:52.46mnemocsteev: those are *smart*books not netbooks
17:52.49mnemoc:p
17:53.01traeakhehe, sry, i just ordered one of those dirt cheapy nokia 3g netbooks with the 9 cell
17:53.02steev:P
17:53.12hp__:)
17:53.12steevtraeak: url? i've not heard
17:53.17steevcollects hardware
17:53.24hp__nokia has netbooks?
17:53.28hnofor what it's worth tablets is heading the same way unfortunately. Next 2 years in the tablet market will be messy.
17:53.46traeakhttp://lastcall.dailysteals.com/
17:53.51traeaki got that one for like $180 or something
17:54.21steevtraeak: oh, it's an atom, bleh, nothx, i have plenty of x86 boxen
17:54.45traeakyeah, i know...just saying i don't have any *real* computers at my house anymore
17:55.08traeak2 x86 netbooks with arch and 2 tablets with android
17:55.25traeaksadly i haven't seen a compelling non x86 netbook yet
17:55.31steevspiffy, they have the 1280x720 screen in em
17:55.40steevi have that in my personal smartbook
17:56.59hnohave 4 working x86 boxes, 5 ARM "things" (two boxes, 2 phones, 1 headless), 2 PPC, 1 SPARC32, plus a lot of other junk not really used or with closed firmware which can not be touched (usually ARM)
17:57.58hnoprobably some MIPS stuff in the junk pile as well.
17:58.01steevi only have 1 PPC :(
17:58.17hno1 MAC something, 1 PS3.
17:58.27specinghno: routers are MIPS usually
17:59.04hnospecing, the ones I have is mostly ARM, but I think one may be MIPS.
17:59.56penguin42has quite a lot of old junk, but I have a few SPARCs, ancient ARM, ancient PPC, ancieint HP-PA, an Alpha, a couple of 68020 (Sun3), a MIPS or two and a few x86
18:00.09penguin42all have run Linux at one time or another
18:00.45hnoonly counted stuff which runs today.
18:00.53specinghttp://dmitry.co/index.php?p=./04.Thoughts/07.%20Linux%20on%208bit :D
18:01.07penguin42hno: I've not had a lot of my stuff on for years
18:01.50penguin42hno: chatting to some guys at work I think we came to the conclusion that most geeks had an SGI Indy that they'd got years ago and booted a few times and hadn't used for ages :_0
18:02.17steevi used to have an Indy!
18:02.40traeaksad mips is so far behind
18:03.18specingI had hope in chinese Loongsons untill they started emulating x86
18:04.05hnoNever even used an SGI actually. But most other stuff..
18:04.21hnoAlpha was nice.
18:04.45hnoAnd grew up with M68K.
18:05.11hnowell, some 6502 before that.
18:06.32traeakintel sucked up alpha
18:07.01traeakbut yeah they were cool (well not temperature wise)
18:07.40hnoNot sure I would blame Intel for that.. Digital managed to deprecate the Alpha quite well on their own and quite likely even it Intel had not existed..
18:08.41traeakwell yeah they fell into the same stupidity that killed off the mainstream unix vendors
18:27.52hp__time for a reboot
18:29.48CIA-91rhombus-tech: Matthijs master * r74044c371635 /allwinner_a10/orders/snita.mdwn:
18:42.56*** join/#arm-netbook hp_ (~hp_@p5B38A957.dip.t-dialin.net)
18:51.45*** join/#arm-netbook MaDMaLKaV (~IceChat9@188.86.253.69)
19:02.44*** join/#arm-netbook destinal__ (~chatzilla@ip68-108-26-102.lv.lv.cox.net)
19:10.37hnomnemoc, some progress. early print functions do work.
19:10.54hno## Transferring control to Linux (at address 40008000) ...
19:10.54hnoStarting kernel ...
19:10.54hnoHello world
19:11.59mnemoc\o/
19:13.24mnemocdid you fix the intereaction problem between uboot and the kernel?
19:13.42hnoHaven't had time to do anything until now.
19:14.52mnemocoh... I thought earlyprintk run after those lookups.
19:15.24hnoI just added a printascii immediately on entry to verify the early print functionality.
19:16.05mnemoc:o
19:16.45mnemocvery smart move :)
19:16.56mnemocit's great to know it actually does work
19:19.59hnoactually it didn't. The register address is wrong hitting UART1 and not 0.
19:20.07hnobut does not help...
19:21.48hnohmm... expected two "Hello world" from this kernel, only got one.
19:27.57hnoWoot! earlyprintk even works! It talks to me! but then dies...
19:28.38hnomnemoc, http://fpaste.org/MfU6/ !
19:29.17mnemoc\o/
19:29.36mnemocdiff! diff! diff!
19:29.42hnoIt's coming.
19:31.11hnoThe most important fix is
19:31.12hno-#define PA_UARTS_BASE          0x01c28400
19:31.12hno+#define PA_UARTS_BASE          0x01c28000
19:31.47traeakwhat device is this on?>
19:31.54hnomele
19:32.01traeakcoolio
19:32.23mnemocbut the same kernel binary covers any a10-based device
19:32.39traeakexcept for any peripherals
19:32.56mnemocfor low level details there is script.bin
19:33.08mnemocfor higher level, linux itself does the magic
19:33.34traeakand i do see flatmushis pretty active in hacking on the lima driver
19:33.50traeakso there may be hope for a linux image when i get my mele
19:34.00mnemocthere is a 2d driver so for basic ui we don't need mali
19:34.34traeaknot immediately
19:35.54mnemocunfortunatelly for reverse engineering the cedar driver we are alone
19:36.04traeakhmm...wonder what the chance is now for the a10 platform to take a chunk away from the rpi
19:37.19mnemocI see no point these days in using anything below armv7
19:37.59hnoargh! write protect tab on my SD-card just died. Cant write to it any more.. ;(
19:38.57traeakmnemoc: exactly
19:39.52mnemocand with the prices of armv7-m chips arduino and all those 8/16bit fpgas are senseless too
19:40.26hnoarent't the arduino ones M profile? (only THUMB)
19:40.45hnoah, right, you said so.
19:42.04traeakhow long will the allwinner a10 stay in production?
19:42.25mnemocthey haven't yet announced the sucessor
19:42.39mnemocso probably a10 will be in production for a while
19:43.29traeaknow with the kde plasma tablet...does that mean that the ali a9 soc has also been opened up?  although it doesn't have sata
19:43.57mnemocalso a soc with the specs of the allwinner a10 has lot of market even after it's cortex-a9 sibling hits the market
19:44.32traeaki guess the safe part is both platforms are armv7
20:06.53specingI just got some time to plug my tablet into adb to see why it doesent boot
20:07.21specinglogcat: dalvik keeps complaining about no space being left on the device
20:08.22hnomnemoc, change pushed to github
20:12.15hnomnemoc, do your 2.6.36-sun4i tree work?
20:12.31specingThere is no space because /data is not mounted
20:13.27mnemochno: yes, that's the one that worked after adding the-evil-commit
20:13.38hnook
20:27.52mnemochno: btw, how did you notice it was going to uart1?
20:29.31traeakhttp://latest.mistertao.com/taobao-products/taobao-item-14603567785.html
20:29.46hnomnemoc, by reading the code.
20:30.23mnemochno: :)
20:32.02traeakinteresting at the bottom are all the mele devices
20:32.09mnemocdidn't know taobao had an international shop
20:32.12traeakmore than a few have usb3
20:32.22traeakmakes me wonder what those other devices are based on
20:33.11traeakhave to make my wife translate, although she doesn't know what to look for
20:33.53*** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk)
20:34.57hnotraeak, mele lists all their products except the A1000 in their english site.
20:35.16traeakahh okay
20:35.31hnohttp://www.mele.cn/en/web/pro_list.aspx
20:35.56hnothey have quite many.
20:36.07hnosee left menu.
20:37.33traeakunless i'm wrong teh stuff on tabao doesn't show up on their website
20:40.29traeakthe mele a1000 shows up as am image on their front page flash but no way to get to them
20:47.35*** join/#arm-netbook destinal__ (~chatzilla@ip68-108-26-102.lv.lv.cox.net)
20:52.01hnotraeak, there is plenty of A1000 info in mele chineese pages.
20:53.33traeakhno i was lokgin at thte a2000 and was curious what else mele is making
20:59.12specingHmmm, how would one find if a program is hardfloat?
20:59.37mnemoctry using `file`
20:59.44hnogood question. It's nore mentioned in the file header.
21:00.51traeakokay the "other" mele stuff is based on the realtek 1185 500mhz mips processor with 256MB ram
21:02.53hnoHmmm.. 2.6.36 did not boot at all for me.
21:03.07hnonot even early_printk stuff.
21:03.09specingmnemoc: this example (hardfloat gentoo) doesen't tell me anything: /bin/bash: ELF 32-bit LSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.16, stripped
21:04.36hnospecing, correct.There is nothing in the file header telling which ARM ABI the binary is using, which i a major headache for anyone needing to support both.
21:05.45hnoit's missing entirely from the ARM ELF specification.
21:07.07specingtraeak: Your wife is/knows chinese?
21:07.22traeakspecing: yes
21:12.03specingLooks like there is quite an international community here
21:12.39steevis the a10 boot from mmc still correct in the wiki? i can't seem to get my ly-f1 to boot using mmc
21:16.12specingI haven't even tried yet
21:16.18hnomnemoc, 2.6.36 plain refuses to cooperate with me. printascii works, but no sign of the kernel console output.
21:16.26specingDoes mmc override internal nand?
21:16.39hnospecing, yes.
21:16.40steevspecing: yes
21:16.45specingnice
21:16.57steevi did something wrong, just don't know what
21:17.05specingHow easy is it to relocate ext4 partitions, then?
21:17.14specings/ext4/ext2/
21:17.23hnospecing, what do you mean?
21:17.45specingI have one massive ext2 partition on my SD card with Gentoo on it
21:17.51hnook
21:18.06hnoand you want to make more space infront of it?
21:18.17specingyes
21:18.33specingactually, I need to shorten the partition
21:19.01hnoyou need to rewrite the whole partition, and resize2fs the filesystem to desired size.
21:19.30hnosafest to backup the partition, resize2fs the backup, repartition, and write back.
21:19.39specingD:
21:20.54destinal_https://1saleaday.com/  $80 for a droid lapdock which I think is basically keyboard, trackpad, hdmi monitor, battery, speakers, USB hub, and probably USB sound card.   could be handy for modifying to work with EOMA68 cards or something...
21:21.20hnothe alternative is to resize2fs the filesystem on th SD card, and then use some tool that moves it in-place, but if that crashes then all gone.
21:23.44hnodestinal_, EOMA68 do not have HDMI.
21:23.53specingdestinal_: they could have put the phone where the touchpad is located, now it just looks retarded
21:24.17hnospecing, you don't see the phone when in use.
21:24.18destinal_hno: true, you'd need to put an HDMI encoder chip on an IO board or something
21:24.50destinal_but EOMA-CF has HDMI in the spec, right?
21:24.59hnodon't remember
21:25.03specinghno: It could still act as a touchpad...
21:26.11hnospecing, yes, but not a nive one, and getting all dirty.
21:26.20hnos/nive/nice/
21:26.54specingdamn sed bots
21:26.54destinal_interesting thing the lapdock connector on bionic seems to not be one solid proprietary connector but two.  perhaps male mini USB  and micro HDMI or something
21:29.52destinal_yup  "At first, we’re a bit hesitant to attach the adapter to the handset – mainly due to the fact that it’s only held in position by the handset’s microUSB and microHDMI ports."
21:30.05destinal_nice.  glad not too proprietary
21:31.46hnodestinal_, Micro USB and Micro HDMI says the manual
21:33.28destinal_yup micro not mini
21:34.10hnoas it should. They agreed on the micro-usb as universal charge port.
21:35.27hnoSo the EOMA64 card should dock quite nicely using the external connectors. Not how it's meant to dock but still...
21:37.18hnoEOMA64 A10 card.
21:37.34hno68.
21:58.35mnemochno: https://github.com/amery/linux-allwinner/commits/amery/linux-2.6.36-sun4i <--- this one? note the amery/ prefix
22:17.58mnemochno: just tried a fresh clone and works :-/
22:18.50CIA-91rhombus-tech: rasiemse master * r12c3c2a8b2df /allwinner_a10/orders/tronetix.mdwn:
22:20.21traeaksun4i
22:20.22traeakstrange
22:20.51mnemoctraeak: that's the machine type of the allwinner a10
22:22.08mnemocthere was sun3i before and the next will be sun5i (cortex a9)
22:22.50traeakk
22:23.03traeakjust reminds me of the old days of sunos and solaris
22:23.04traeakheh
22:23.10mnemoc:)
22:23.51traeakbut that woudl be ":sparc" architecture
22:24.00penguin42has a couple of Sun3's here if mnemoc wants them
22:24.16mnemocpenguin42: -v
22:24.19penguin42you can just about lift them by yourself
22:25.05traeakwell the sp
22:25.14traeakwell the sparc 5/10/20 pizza boxen were cool
22:25.31mnemocpenguin42: you mean sun machines or allwinner' sun3i-based?
22:25.50penguin42mnemoc: A Sun 68020 machine
22:26.10hnomnemoc, yes that one.
22:26.26mnemochno: did you load script.bin?
22:26.32hnoyes
22:26.35mnemocweird
22:26.39hnoyes
22:27.22hnoI only fixed the UART address macro.
22:27.40mnemocmy test was after including your commit too
22:28.08hnoanything particular in .config?
22:28.21mnemocsun4i_defconfig
22:28.55hnosun4i_defconfig + CONFIG_EARLY_PRINTK here..
22:35.45mnemocit doesn't help, but this is the output I get http://dpaste.com/724473/
22:35.52hnomnemoc, can you upload your working kernel somewhere?
22:36.45mnemoc1m
22:37.01hnojust to check if it's something with my u-boot setup or changes.
22:39.34mnemochttp://dl.dropbox.com/u/24940371/A10/uImage http://dl.dropbox.com/u/24940371/A10/script.bin
22:41.15mnemochno: it might be that your uboot gives the assigned machine number while mine gives the one allwinner stole
22:41.43hnothe hack overrides machine arch number.
22:41.53mnemoctrue
22:41.56hnoand original uboot sends a completely different id.
22:42.45hnooriginal kernel still boots fine..
22:45.12mnemocand the uImage from the link?
22:45.34hnoworks.
22:49.18mnemocand you are super sure you built amery/linux-2.6.36-sun4i from git@github.com:amery/linux-allwinner.git ?
22:49.26mnemocpuzzled
22:50.06hnotrying again with just sun4i_defconfig
22:50.57hnogit diff to your tree gives nothing.
22:51.35hnogit diff remotes/amery-allwinner/amery/linux-2.6.36-sun4i
22:52.39hno* remote amery-allwinner
22:52.39hno<PROTECTED>
22:53.27mnemocI didn't know that remotes/foo syntax
22:53.42mnemocalways used: git diff amery-allwinner/amery/linux-2.6.36-sun4i..
22:54.05hnogit searches at the end if no exact match is found
22:54.58hnoit's possible to create a local branch that is named  amery-allwinner/amery/linux-2.6.36-sun4i
22:55.28mnemocand "remotes" is reserved?
22:55.43hnoNot really, but special.
22:56.08hnoas fetch updates any refs there.
22:59.10hnocompiling...
23:00.03hnopoor old computer.. anno 2005 I think.
23:00.54mnemocI rent a server at hetzner.de mostly only for building :p
23:04.00mnemocI can make it burn without affecting my local the responsiveness of my laptop or it's battery :)
23:05.18hnonow it works. Odd.
23:16.28hnoand it still boots after killing the r0-r2 hack.
23:17.04mnemocuhm... so it's on the table allocation part?
23:17.31mnemocI'm sure I tried that... :-/
23:18.13hnowith the right arch id set by uboot?
23:18.37mnemocmaybe one or the other but not both
23:20.29mnemoccould you port the table allocation evilness to the android-3.0 tree?
23:20.46hnoI have that already, still does not work.
23:21.14hnobut not tried the android tree. only the -sun4i tree.
23:23.49hno2.6.36 with early_printk enabled hangs hard.
23:25.36hnoif earlyprintk boot option is given. Otherwise it works.
23:26.04hnoexplains the hang seen earlier.
23:26.28mnemocweird^3
23:27.08hnovery
23:27.57hnoesp considering that printascii do work fine.
23:28.17hnoand early_printk uses printascii
23:33.05traeakhttp://liliputing.com/2012/03/this-is-what-a-50-android-tablet-looks-like.html
23:35.36hnotraeak, yes, the A10 devices are pulling the plug on tablets market.
23:54.33mnemocgood night

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.