mnemoc changed the topic of #arm-netbook to: EOMA: Embedded Open Modular Architecture - Don't ask to ask. Just ask! - http://elinux.org/Embedded_Open_Modular_Architecture/EOMA-68 - ML arm-netbook@lists.phcomp.co.uk - Logs http://ibot.rikers.org/%23arm-netbook or http://irclog.whitequark.org/arm-netbook/ - http://rhombus-tech.net/
hg_5 has quit [Ping timeout: 252 seconds]
gimli has quit [Remote host closed the connection]
HACKhalo2 has quit [Ping timeout: 260 seconds]
tuliom has quit [Ping timeout: 256 seconds]
tuliom has joined #arm-netbook
tinti has joined #arm-netbook
merbzt has quit [Ping timeout: 255 seconds]
<drachensun> focus_it thanks I'll see what we can do ;)
stefanro1 has joined #arm-netbook
stefanro has quit [Ping timeout: 260 seconds]
tuliom has quit [Quit: Konversation terminated!]
Turl has quit [Excess Flood]
Turl has joined #arm-netbook
popolon has quit [Quit: Quitte]
Jef91 has joined #arm-netbook
<Jef91> Anyone know what board the MK808 devices use?
<WarheadsSE> it's a rockchip i think
<Jef91> Yea, looks like the RK3066 WarheadsSE - thanks
mSquare has joined #arm-netbook
itamar_ has joined #arm-netbook
Jef91 has quit [Quit: Instantbird 1.3 -- http://www.instantbird.com]
Jef91 has joined #arm-netbook
freakazoid0223 has joined #arm-netbook
tinti has quit [Quit: Leaving]
itamar_ has quit [Remote host closed the connection]
Quarx has joined #arm-netbook
Quarx has quit [Read error: No route to host]
<niftylettuce> If anyone has anything to contribute to MK802 II, I found this thread here: http://slickdeals.net/f/5612748-Rikomagic-MK802-II-New-Generation-Mini-Android-4-0-PC-Google-TV-Box-HD-IPTV-Player-44-90
freakazoid0223 has quit [Quit: Leaving]
<xxiao> niftylettuce: nice, from the review it's not very reliable
<niftylettuce> xxiao: yeah but are any of them? lol
<xxiao> mele2000 so far works fine
<niftylettuce> xxiao: do you have resources/links for that?
<niftylettuce> xxiao: isn't that same as mk803 iii ?
<niftylettuce> err mk802*
<xxiao> google mele 2000
<xxiao> i bought it at aliexpress i think
Jef91 has quit [Remote host closed the connection]
Ershov has joined #arm-netbook
Turl has quit [Excess Flood]
Turl has joined #arm-netbook
Turl has quit [Changing host]
Turl has joined #arm-netbook
<Turl> lundman: ping
<Turl> lundman: my computer crashed and now I get "SPL panic" on boot and /dev/zfs doesn't appear
<Turl> lundman: any hints? :)
<lundman> what?
<lundman> ok, so you boot, and it loads spl, then panics?
<lundman> what is the panic?
<lundman> also, remind me, you running offical zol, or uhm the special branch I worked on?
tzafrir_laptop has quit [Read error: Connection reset by peer]
<Turl> lundman: official
<Turl> VERIFY(0 == dmu_bonus_hold(spa->spa_meta_bjset, obj, FTAG, &db)) failed
<Turl> SPLError: 735:0:(spa.c:1353:load_nvlist()) SPL PANIC
<lundman> ok I wasnt sure. not many care about crypto :)
<Turl> can't open /tmp/spl-log.... for dump: -30
<lundman> seems to be in hold l2cache
<lundman> fails on import? you could import with -N
<Turl> it fails on boot, idk what is it doing
<Turl> what does N do?
<lundman> import pool without mounting anything
<Turl> "sudo zpool --help" makes it crash
<lundman> but that is after the panic yes?
<lundman> rc11? rc12?
<Turl> booting with init=/bin/bash
<Turl> and "zpool -N"
<Turl> makes it crash
Sv2 has joined #arm-netbook
<lundman> yeah so it is the pool import
<lundman> amusing, Sun has a patch for it
<Turl> I see zvol_init.. _create_minors, _replay_write, _create_minor, spa_open_common, _load_best, spa_activate, spa_load, zap_lookup, load_nvlist.isra, spl_debug_bug, spl_debug_dumpstack
<lundman> so it dies importing the pool due to zvol? you have zvols?
<Turl> yeah my swap is a zvol
<Turl> can I disable it somehow?
<lundman> I feel a little vindicated, but it doesnt help you
sv has quit [Ping timeout: 276 seconds]
<lundman> I spent last week arguing with ZOL that -N is ignored for zvols
<lundman> I can tell you what line to comment out though?
<Turl> anything that'll get my data back I guess
<lundman> data isnt lost in anyway, just zvol load code. there are 2 tickets on to fix it as we speak
<Turl> I took a pic of the full crash, once I get my machine operational I can attach it to whatever ticket you need
<lundman> well, do you want to wait for them to fix it, in next couple of days, or try commenting out a line of code and see if you can import pool without swap
<Turl> well, I need to finish a final project for university
<Turl> and I did a lot of progress today
<Turl> and of course I didn't back it up
<Turl> :)
tzafrir_laptop has joined #arm-netbook
<Turl> I can edit whatever code and rebuild with dkms
<lundman> wicked. we arent doing anything scary, just telling import pool to NOT create the minor dev nodes for zvol
<lundman> zfs/module/zfs/zfs_ioctl.c
<Turl> sounds good
<Turl> got nano running on dual monitor 1080p, what do I edit? :)
<lundman> zfs_ioc_pool_import: 1314
<lundman> if (error == 0)
<lundman> zvol_create_minors(zc->zc_name);
<lundman> comment that call out (2 lines)
<Turl> anything else?
<lundman> nope
<lundman> compile, rmmod zfs, and insmod
<lundman> or, reboot
<lundman> actually you paniced, so reboot required
<Turl> how do you invoke a build with dkms?
<lundman> whats a dkms?
<lundman> I'm kidding, its a package or something
<lundman> actually
<lundman> wait
<lundman> spl.ko has an argument
<lundman> MODULE_PARM_DESC(zvol_inhibit_dev, "Do not create zvol device nodes");
rellla has joined #arm-netbook
<lundman> yes, just rmmod spl.ko ; insmod spl.ko zvol_inhibit_dev=1
svp1 has joined #arm-netbook
<lundman> oh actually, you cant just rmmod spl, you have to "rmmod zpios zfs zcommon zunicode zavl znvpair spl"
<lundman> then "insmod spl.ko zvol_inhibit_dev=1" insmod znvpair && insmod zavl ..
<Turl> I need to reboot because it panic'ed
<lundman> yes
<Turl> I echo options spl zvol...=1 > /etc/modprobe.d/spl.conf
<Turl> seems to boot fine now
<lundman> cool, is that how it works in linux
<lundman> just delete the zvol, and recreate it
<lundman> but it wont mount it until next reboot
<lundman> when you have time, please create the ticket
<Turl> erm, I lied, it booted because I had run dkms remove >:<
<lundman> doh
<Turl> and of course kernel source is on zfs...
<lundman> ouch
* Turl tries old kernel
<lundman> oh heh its not spl
<lundman> its for zfs
<lundman> insmod ./module/zfs/zfs.ko zvol_inhibit_dev=1
<Turl> luckily I keep a 3.2 from debian installed
<lundman> ok I can confirm that it did not create the zvols here
<lundman> but I cant confirm it will stop panic
<Turl> it still seems to panic :(
<lundman> you changed the line from spl.ko to zfs.ko ?
<Turl> yeah, the trace is different
<lundman> what is it ?
<lundman> was it with -N ?
<lundman> wait what, you arent even issuing the command to import?
<Turl> "zpool" loads the modules or whatever, maybe it tries to autoimport
<lundman> ah i guess that is possible
<lundman> why dont you manually insmoding the modules?
<lundman> "zpool import -N" should list the pool, then "zpool import -N name" to import
<Turl> "zpool import -N" crashes
<Turl> I modprobed zfs manually with that parameter
<lundman> can you "zdb -l $device" where $device is one of your raw hdds?
<lundman> you can use specific txg numbers when importing pools too
<lundman> for example: zpool import -T-510254
<lundman> if you have txg=510256 in zdb output
<lundman> if you have txg=510254 in zdb output
<Turl> zdb -l /dev/sda causes SPL panic lundman
<lundman> ffs
<lundman> what did you do to the pool :)
<Turl> lundman: there was a huge electric storm outside
<Turl> my pc is shiny
<Turl> so I was all like fuck I'll shut it down
<lundman> you might have to join us in #zfsonlinux and create a ticket
<Turl> and I was working so I hit hibernate
<Turl> but then I remembered it has 16G ram
<Turl> and after it was like 10m hibernating and didn't respond I halted it :|
<Turl> lundman: doesn't zfs have a fsck tool or something of sorts? :/
<lundman> no, it self heals, in that, once it is imported, it will run through your data, and chsum and repair problems
<hno> hi lundman, there is quite a bit of gcc warnings in u-boot zfs driver still.
<lundman> but irony is, it dies in SPL (solaris porting layer)
<lundman> hno: really? I made it warning free when I submitted it, but you know how much gcc changed between versions
<Turl> "self healing" is not much issue if you can't even mount it :\
<Turl> much help*
<hno> lundman, using 4.7.
<lundman> turl: I mean, you could probably just boot the Sol11 ISO, go to root shell, import it and it will be fixed
<lundman> 4.5 I believe i was on
<Turl> where do I get such an iso? :/
<lundman> free download from sun
<Turl> can I liveusb it?
<lundman> yes, it has usb
<lundman> but do *not* upgrade the pool :)
<Turl> FUU asks for login
<lundman> luckily that is a separate command "zpool upgrade"
<lundman> really?
<Turl> what do I download?
<lundman> I use the text CD ISO usually
<lundman> but mount from IPMI and boot.
<lundman> dont know your hardware though
<Turl> the first x86 link?
<lundman> yep
<lundman> usb is further down on the page
<hno> odd... building fs/zfs/libzfs.o is clean. But full build shows the error. It's about strict aliasing and that have indeed changed in reporting in later gcc versions.
<lundman> lets not forget uboot would not accept it with warnings normally. but if there is something I can help with
<Turl> lundman: can I chainload it from grub?
<lundman> yes
<Turl> so I got 9min before the iso is downloaded
<Turl> what do you suggest once I dd it to my pendrive and chainload from grub?
<lundman> create the ticket so it can be fixed :)
<lundman> hno: you have to low black on black text
<Turl> lundman: just booting it will fix the pool?
<lundman> trul: boot, pick option 6) Root shell, then it will ask if you want to import existing pool on /a
<lundman> say yes, then check it with zpool status
<lundman> hno: that is quite a lot of warnings
<hno> lundman, it's a lot of warnings but all the same. A typecasted pointer in a way not compatible with pointer aliasing rules potentially breaking with optimization enabled.
<hno> lundman, http://fpaste.org/3v6y/raw/ should work fine
<lundman> all macros with DN_BONUS
rellla2 has joined #arm-netbook
rellla has quit [Ping timeout: 264 seconds]
Guest82090 has quit [Ping timeout: 260 seconds]
fredy has joined #arm-netbook
Turl has quit [Ping timeout: 260 seconds]
fredy is now known as Guest50081
Marex has quit [Ping timeout: 260 seconds]
Turl has joined #arm-netbook
arokux has quit [Ping timeout: 260 seconds]
<Turl> I seem to have timeouted
<Turl> lundman: how can I boot it from grub? :P
arokux has joined #arm-netbook
Marex has joined #arm-netbook
<lundman> usually I do "root (cd0); chainloader +1 ; boot
<Turl> I did set root=(hd2)
<Turl> chainloader +1
<Turl> and that gets me invalid signature
<lundman> your ISO image is on hard disk2 somehow?
<Turl> on usb drive
<Turl> hd0 & 1 are internal
<lundman> not sure you can dd an iso image to usb and boot it as hdd? can you?
<Turl> usually you can, when they're hybrid images
<Turl> I guess solaris one isn't
rz2k has joined #arm-netbook
mSquare has quit [Read error: Connection reset by peer]
<rz2k> the guy who decided to use waf python2 based build system as a default one for glmark2 must be really high. it fails at every step, starting from unacceptable configuration speed on embedded systems, ending with missing option for defining own EGL configuration (what if I dont have pkg-config around?).
mSquare has joined #arm-netbook
<Turl> lundman: ok so I burned a dvd and booted into it
<Turl> I chose 3) shell
<Turl> but it didn't ask to import
<Turl> zpool status shows no pools, zpool import shows my pool, zpool status again shows no pools
<Turl> lundman: "zpool import home" says pool may be in use from another system , -f to import anyway
<Turl> lundman: is that because it crashed?
<Turl> lundman: zpool import -f home crashes solaris >.<
<lundman> back
<lundman> hah, no way?
<lundman> guess you could try and see what the fbsd RC does, since it slightly different paths
<lundman> but that needs some expert insight
<Turl> so much for a "stable" filesystem >.<
<lundman> dont suppose "zpool import -o readonly=on" makes any difference
<Turl> I tried "zpool import -f -N home"
<lundman> apaprently readonly stops it from trying to reply the transaction logs
<lundman> *replay
<Turl> should I use -N still lundman?
<lundman> sure why not
<lundman> is it a mirror? raid? single disk pool?
<Turl> still exploded
<Turl> HDD, with ZIL and L2ARC on two partitions on a SSD
<lundman> yeah, create the ticket and I'll chat to brian and see if we can come up with fixes
<Turl> what should I write on the ticket?
<Turl> "zfs/spl fails and doesn't wanna import my pool?"
<lundman> Panic during pool import in load_nvlist
<lundman> maybe?
<Turl> against spl or zfs?
<lundman> zfs
<lundman> zfs does bad code, caught by the ASSERT, defined in SPL
<Turl> should I link to all the pictures?
<lundman> you can, if you dont want to type it in
<lundman> but its appreciated if you at least give shorthand call stack
cat_x301 has joined #arm-netbook
<Turl> on which one?
ibrah has joined #arm-netbook
popolon has joined #arm-netbook
brooc has joined #arm-netbook
<brooc> Hello
<brooc> hno, are you here?
<rz2k> ssvb: I've pasted hardfloat results for OpenSSL, you requested them some days ago. http://linux-sunxi.org/Benchmarks#Results_2
<lundman> great, I will encourage the guys from here
cat_x301 has quit [Read error: Connection reset by peer]
<rz2k> ssvb: and yeah, ghash 14806.49k vs. ghash 52904.92k
cat_x301 has joined #arm-netbook
<ssvb> rz2k: -DAES_ASM is the most interesting, because it is used for ssh/scp, but yeah, it's good to see that linaro/ubuntu are the only slowpokes and the other distros are fine :)
<Turl> lundman: maybe I'll try with zdb http://forums.freenas.org/archive/index.php/t-7095.html
<lundman> interesting, there is a recover switch
<lundman> :MODULE_PARM_DESC(zfs_recover, "Set to attempt to recov
<lundman> there sure is
<lundman> try that
<lundman> the fbsd kernel has a reputation of being very robust too
pwhalen has quit [Ping timeout: 276 seconds]
<brooc> Hi
<brooc> I am trying to build the kernel using the make commads in the WiKi
<brooc> For some reason the makefile is using the arm-linux-gnueabi-gcc and not arm-linux-gnueabihf-gcc
<brooc> I went over to the Linaro repository and they only ofer this for download in previous releases
<brooc> Is there a reason why this is?
<brooc> And why the makefile assumes arm-linux-gnueabi-gcc is installed?
<brooc> Should I just make softlinks to all the binaries with the name without hf?
<brooc> anyone?
<RaYmAn> use CROSS_COMPILE=your_arm_gcc_prefix
<RaYmAn> env. variable
<RaYmAn> e.g. if you wanted to set arm-linux-gnueabi-gcc you'd do CROSS_COMPILE=arm-linux-gnueabi-
<brooc> got it
<brooc> I now see that was a stupid question
<brooc> "mkimage" command not found - U-Boot images will not be built
<brooc> Why am getting this?
<RaYmAn> you're missing u-boot tools package
<RaYmAn> it varies what it's called depending on your distribution
<RaYmAn> uboot-mkimage package on debian
<brooc> The distribtution of the host or the destination?
<brooc> I am compiling on Ubuntu
<brooc> Do I need to install something?
<RaYmAn> host
<brooc> Thanks installing
<brooc> why is this not in the WiKi?
<RaYmAn> I'm pretty sure it is.
<RaYmAn> If not, do feel free to add it :P
<RaYmAn> "why" something is not in the wiki is usually because no one has added it. It's rarely more complicated than that ;)
<brooc> Where do you think it is? Maybe I am not looking in the right place
<RaYmAn> I dunno, maybe it's not
<RaYmAn> Do feel free to add it to linux kernel page
<brooc> Just did
madmalkav has joined #arm-netbook
tinti has joined #arm-netbook
<Gumboot> Hang about... what's this AES talk, here?
<Gumboot> ssvb: What has Ubuntu done wrong?
<rz2k> Gumboot: they build openssl softfloat on linaro armhf rootfs
<rz2k> key word is armhf
<Gumboot> How'd they manage that?
<rz2k> no idea
<Gumboot> What kind of softfloat? Just the ABI, or complete removal of floating-point instructions?
eFfeM has joined #arm-netbook
<Gumboot> For just the ABI I've never seen such huge performance differences. Is that actually NEON versus scalar, then?
<Gumboot> There's a #linaro channel on this network if you want to ping them.
<brooc> For development purposes would anyone use MK802?
<rm> development of what
<rm> and "for", or just "on" the mk802
<brooc> just on
<brooc> Development of kernel and uboot
<mnemoc> RaYmAn: my chromebook was totally ignored by the customs :p took a bit longer but ended up getting it for 220E :)
<rm> brooc, no
<rm> first, do you expect to compile stuff right on the SD card?
<rm> Mele A1000/A2000 at least have a SATA slot
<rm> second, a slow CPU
<rm> you would really fare better by getting a cross-compile environment going on a N-core x86 box
<rm> where's the fun in development when each compile takes you 5 hours
<rm> (vs 15 minutes)
<rz2k> <Gumboot> What kind of softfloat? Just the ABI, or complete removal of floating-point instructions? - see compiling options yourself http://linux-sunxi.org/Benchmarks#Results_2 first is linaro-alip, second is archlinux-arm
madmalkav has quit [Quit: Leaving]
jquip has joined #arm-netbook
<brooc> rm, I intend to compile on my linux PC
<brooc> not on the dev board itself
<brooc> I want it just for testing
<jquip> how does one enable the goodix touchscreen? its a gt811 i think... are all variants merged into the goodix_touch module? I can't seem to find gt811_ts in my compiled modules.. only goodix_touch..
<brooc> BTW, I want to setup my toolchain for building a simple hello world program for the ARM cortex-A8/9
<brooc> Want to try and run this on my Android phone
<brooc> Where can I find a tutorial that walks through the environment setup?
<rz2k> android has own libc
<rz2k> you cant do that, afaik.
<brooc> Will opening a terminal on the device wont work?
<brooc> Well then I'll try it on qemu
<brooc> rz2k, what do you think of using the MK802 as dev board?
ibrah has quit [Quit: ChatZilla 0.9.89 [Firefox 17.0/20121119183901]]
<rz2k> brooc: that differs on definition of devboard for you
<brooc> What do you mean?
<brooc> I am looking to develope kernel and uboot for the sunxi
<rz2k> I suggest you a cubieboard because all pins are exposed there
<rz2k> on mk802 you will suffer with soldering UART to tiny pins and etc.
<RaYmAn> mnemoc: nice! :)
<brooc> I realize that, but the problem is that the Cubieboard is not being sold
itamar_ has joined #arm-netbook
<rz2k> you can go for mele a2000
<RaYmAn> if you really don't care about hardware, stability, pins etc, going for whatever is the cheapest a10 device is probably a good idea
<rz2k> it has UART exposed on pinheader and JTAG can be accessed thru SD card slot hack (ask hno)
<RaYmAn> =P
RITRedbeard_ has quit [Ping timeout: 246 seconds]
RITRedbeard_ has joined #arm-netbook
<brooc> the mele a2000 is too expensive
<brooc> What is the cheapest a10 device?
<RaYmAn> I think you can get some for like 50$
SouL_ has joined #arm-netbook
SouL_ has left #arm-netbook [#arm-netbook]
<brooc> My price range is around 35
<brooc> like the raspberry pi
CaCtus491 has quit [Ping timeout: 264 seconds]
<orly_owl> will you buy any accessories for the rpi?
CaCtus491 has joined #arm-netbook
tzafrir_laptop has quit [Ping timeout: 252 seconds]
<brooc> What do you mean?
tinti has quit [Quit: Leaving]
<orly_owl> a case, memory card
<orly_owl> etc
<brooc> nope
<brooc> why do you ask?
jelly has joined #arm-netbook
ssspiff has joined #arm-netbook
sspiff has quit [Ping timeout: 252 seconds]
Sv2 is now known as sv
ssspiff has quit [Ping timeout: 252 seconds]
sspiff has joined #arm-netbook
alcides has joined #arm-netbook
<sv> if you want something like melee but cheaper
<sv> check out hackberry on www.minand.com
ssspiff has joined #arm-netbook
<sv> $55 if i remember right
<sv> same specs as melee a1000
<sv> A10, 1gb ram, 4gb nand, wifi, etc
<sv> should be $65 with shipping
<sv> not that much pricier than raspberrypi
sspiff has quit [Ping timeout: 246 seconds]
<rm> bleh
<rm> how about $39 shipped
<jinzo> there's a lot of boards outta there :D
<jinzo> rm, shipped next year doesen't count :P
<RaYmAn> mnemoc: which n7 was it you had?
<jinzo> ught that reminds me, I'm due on some comparison tables.
<rm> jinzo, what do you mean by "next year"
<rm> Ships out within 2 days
<mnemoc> RaYmAn: 16GB
<jinzo> rm, oh I tought you're aiming at cubieboard
<rm> the question was about the chapest A10 device
<sv> cubieboard is nice if you dont want wifi and want more expansion ports
<sv> but i thought cubieboard was $49+$10 shipping?
<jinzo> rm, yes I know I know - I blurped.
<RaYmAn> mnemoc: rooted?
<mnemoc> RaYmAn: yes
vinifm has joined #arm-netbook
<mnemoc> RaYmAn: and got 4.2.1/JOP40D bugfix yesterday
<RaYmAn> okay
<vinifm> hno: are you doing NAND enabled u-boot version?
<mnemoc> the lichee-dev branch is a drop-in replacement for stock's nanda's u-boot.bin
<brooc> That is why I started with the MK802 to begin with
<mnemoc> vinifm: nand work for mainline is in the process of understanding the NFC well enough to write an mtd driver
pwhalen has joined #arm-netbook
svp1 has left #arm-netbook [#arm-netbook]
vinifm has quit [Quit: Ex-Chat]
jquip has quit [Quit: jquip]
madmalkav has joined #arm-netbook
Jef91 has joined #arm-netbook
<brooc> So
<brooc> hno, are you here?
gimli has joined #arm-netbook
<ssvb> Gumboot: that was just the case of not enabling armv4 hand optimized assembly in openssl, nothing to do with VFP/NEON or floating point ABI
Quarx has joined #arm-netbook
<WarheadsSE> ssvb: openssl 1.0.1c+ has an annoying bug in it for < v7 neon
<WarheadsSE> has a bad habit of trying to test neon on non-neon systems and taking a crap due to a missed sigfault handler.
mSquare has left #arm-netbook [#arm-netbook]
madmalkav has quit [Read error: Connection reset by peer]
alcides has quit [Quit: > SELECT * FROM users WHERE clue > 0]
Jef91 has quit [Quit: Instantbird 1.3 -- http://www.instantbird.com]
cat_x301 has quit [Ping timeout: 248 seconds]
<techn> w00t.. samsungs 1.7Ghz quad core with 2gigs ram with 90$
<techn> x86 is outdated in year or two.. :p
<mnemoc> +30 for shipping, +?? for storage ;-)
* xenoxaos loves his odroid
<techn> yeah.. after chinese pushes their variant in half price.. there is samekind devices for consumers with 100$
<techn> and performance is enough for everyday surfing
<xenoxaos> pretty decent build quality as well
<techn> and power consumption is 5W instead of 50W :p
<techn> or how much x86 computers draw average?
<techn> 100W?
<xenoxaos> my desktop at home sits 250-300
<xenoxaos> well..that's my entire office though
<xenoxaos> router, modem, desktop, monitors, raid enclosures, goflexes,
* xenoxaos has way too muhch arm shit
<L84Supper> Hardkernel boards sell out in days, they only make a few hundred. Is there anything new there?
<L84Supper> was there a "Worst Form Factor" contest that I missed out on? http://www.hardkernel.com/renewal_2011/products/prdt_info.php?g_code=G135341370451 :)
<xenoxaos> i'm glad it's not a bare board
jquip has joined #arm-netbook
<hno> hi brooc?
ibrah has joined #arm-netbook
<focus_well> Anyone got latest A10 pinout details? The pinouts I am looking at on cubieboard http://dl.cubieboard.org/hardware/cubieboard_schematic_2012-08-08.pdf doesn't tally with all the ones on the A10 data sheets I have to hand. Sata in particular not listed on datasheet dated 2012-04-06
<specing> techn: that ... is... so .... ugly
<L84Supper> focus_well: the A10 EVB schematics have the same pinout for SATA as the cubieboard
<L84Supper> probably cut-n-paste
<L84Supper> is SATA left out of the data sheets due to incompetence or was it intentionally left out along with many other sections?
<focus_well> L84Supper: I search the evaluation board cd - A10 user manual has exactly 1 mention of the word "sata"!!! WTF???
<focus_well> 495 pages!!!
<L84Supper> focus_well: yes, welcome to dealing with allwinner
<L84Supper> their motto should be " Hey, but it's only $7, what do you expect?"
avernos has quit [Ping timeout: 246 seconds]
vinifm has joined #arm-netbook
<focus_well> L84Supper: I'm looking at the evaluation board schematics - its similar to cubie's mentions the sata pins, and it could be reverse engineered to get a pinout of the A10
<focus_well> what a load of cobbers!!
<focus_well> cobblers even!
<L84Supper> focus_well: the design philosophy might be "if it works, it works"
<L84Supper> who knows what their goal actually was
<focus_well> These SoC vendors cut their own throats by not publishing datasheets . No datasheet, no product, no sale of chips = die quickly!!
<mnemoc> they are also "pcb design house" and "software integration house"
<L84Supper> cram the ip blocks into the design, do half assed job of clock and power distribution, tape out, hope for the best
<mnemoc> by limiting software and documentation they promote their other business within china borders
<L84Supper> focus_well: allwinner already made their $$ on the a10, they are off to the next gen device already
<L84Supper> they just sell a BOM + schematics, a reference layout, they don't even have customers that would know what to do with a complete data sheet
<L84Supper> I expect no more from them on the next gen part
* mnemoc has a very bad feeling about sun6i's GPL sources availability
<L84Supper> they sold millions of a10's and nobody needed complete specs to make crappy tablets
<focus_well> I'm looking at the evaluation board - its full of crap design and excess components - the excess components taking care of bad design - doh!!
<mnemoc> also consider they aim at crappy factories
<L84Supper> fine for monkey see, monkey do manufacturing, if it works, don't fix it
<mnemoc> so maybe a better design would lead to more bad boards
<L84Supper> they are already on the next gen device, they sell more based on their low price, not on their docs
<mnemoc> i've been asking them to buy a sun6i devkit for months... they don't even reply now :|
<L84Supper> it was an SOC designed for products with a 1 year life
<L84Supper> and if your a10 tablet doesn't die soon, you probably won't want it next year anyway or might not be able to upgrade to a later version of Android
<focus_well> If any of us make money we should make our own SoC and open source it - put an end to this miserable perverts who value NDAs above chip sales
avernos has joined #arm-netbook
<focus_well> got to go - talk later
merbzt has joined #arm-netbook
gimli has quit [Ping timeout: 246 seconds]
gimli has joined #arm-netbook
Quarx|2 has joined #arm-netbook
arokux1 has joined #arm-netbook
cat_x301 has joined #arm-netbook
Workboot_ has joined #arm-netbook
Quarx|2 has quit [Ping timeout: 250 seconds]
arokux has quit [Ping timeout: 246 seconds]
Quarx has quit [Ping timeout: 246 seconds]
Workboot has quit [Ping timeout: 246 seconds]
Quarx|2 has joined #arm-netbook
<rz2k> I finally got myself to test xserver 1.13+ xf86-video-mali and fixed probably everything, linux-sunxi repo for xf86-video-mali is updated.
<rz2k> I've checked on Linaro-alip and ArchLinux-ARM 1.11.x and 1.13+ xservers accordingly
avernos has quit [Ping timeout: 246 seconds]
avernos has joined #arm-netbook
avernos has quit [Changing host]
avernos has joined #arm-netbook
mikey_w has quit [Ping timeout: 246 seconds]
mikey_w has joined #arm-netbook
itamar_ has quit [Remote host closed the connection]
popolon has quit [Quit: Quitte]
RITRedbeard_ has quit [Quit: Leaving]
Quarx|2 has quit []
RITRedbeard_ has joined #arm-netbook
madmalkav has joined #arm-netbook
vinifm has quit [Ping timeout: 246 seconds]
vinifm has joined #arm-netbook
rz2k has quit []
sv has quit [Quit: sv]
sv has joined #arm-netbook
drachensun has quit [Remote host closed the connection]
j1nx_ has joined #arm-netbook
RITRedbeard_ has quit [Ping timeout: 256 seconds]
rellla2 has quit [Remote host closed the connection]
eFfeM has quit [Quit: Leaving.]
hg_5 has joined #arm-netbook
<WarheadsSE> woo, decent drivers soon :)
drachensun has joined #arm-netbook
<Turl> lundman: ping
vinifm has quit [*.net *.split]
mikey_w has quit [*.net *.split]
specing has quit [*.net *.split]
The-Compiler has quit [*.net *.split]
rmull has quit [*.net *.split]
destinal has quit [*.net *.split]
exuvo has joined #arm-netbook
focus_it has joined #arm-netbook
mikey_w has joined #arm-netbook
specing has joined #arm-netbook
The-Compiler has joined #arm-netbook
destinal has joined #arm-netbook
vinifm has joined #arm-netbook
rmull has joined #arm-netbook
<WarheadsSE> I look forward to xorg-server 1.13 compatible A10 mali drivers
pwhalen has quit [Excess Flood]
pwhalen has joined #arm-netbook
<focus_it> L84Supper: I love Allwinner and their A10. I just want them to keep on succeeding. Their chips are the cheapest
<focus_it> and their demand is endless - mostly for Android which is why they can't service other customers.
<focus_it> If I try to buy a chip at same price in EU or US, I get low power good for nothing chips at that price.
<focus_it> The EU and US chip makers live in the past.
Workboot_ is now known as Workboot
<focus_it> The only thing they have going is the reliability of the supply chain.
<focus_it> In EU and US, but not the performance for the same price.
exuvo has quit [Quit: The end draws near]
<focus_it> mnemoc: to get datasheets, and more interest out of Allwinner, I suggest you go meet them at their offices if you can. The CN way of doing business is face to face. They are incredibly polite people.
<WarheadsSE> Except he's going to have a helluva trip to get there.
<WarheadsSE> I agree that overall I like the chip, ost of that because of the work by peopl in this channel
<focus_it> Its not difficult to get to China - either go to Hong Kong and cross into Shenzhen and take a train to your destination, or take plane directly if you have a budget.
ibrah_ has joined #arm-netbook
ibrah has quit [Ping timeout: 265 seconds]
<mnemoc> focus_it: why directing that to me?
ibrah_ is now known as ibrah
<focus_it> Someone wrote "mnemoc has a very bad feeling about sun6i's GPL sources availability "
<mnemoc> I did
madmalkav has quit [Quit: Leaving]
<focus_it> The best way to get datasheets is go meet them if you can.
<mnemoc> but I'm cheap offshore labour, not someone who can go to negotiate GPL violations with a company in china
<focus_it> Sorry, and I mean also the gpl'd code if you can talk with them face to face
<mnemoc> focus_it: olimex, a company and paying customer of allwinner gets GPL violating kernel and u-boot trees
<focus_it> mnemoc: not to worry - this is advise for future if you make it big
<mnemoc> i?
<mnemoc> I'm not a business person, neither interested in becoming one
<L84Supper> focus_it: about a year ago somebody started spouting off about the A10 and the great deal they made with allwinner about exchanging community support for open docs and availability of a10's to the community
<focus_it> age heals all such desires :-)
<L84Supper> so far only some of the docs are open and only some of the source
<L84Supper> the a10 went from $6 to ~$25 and now back down to <$10
<mnemoc> docs/sources availability for sun5i is far worse than sun4i. and I'm sure sun6i will be even worse
<L84Supper> so just make the best of the docs and the reference schematics if you want to use the a10
<mnemoc> focus_it: luke was supposed to have talked and agreed wonderful things with allwinner
<brooc> hno, is the remote connection to a dev board s posibillity?
<brooc> s=still
focus_it has quit [Ping timeout: 246 seconds]
focus_it has joined #arm-netbook
<focus_it> L84Supper: its possible some companies went to Allwinner and offered fabulous deals with real money that left luke cut out. It won't be the fault of Allwinner that this has happened.
<focus_it> You go to these sweaty back room deal making meetings, you get to know what really happens.
<mnemoc> mass murdering list for a better world by prio: 1) lawyers, 2) bankers, 3) MBAs
<L84Supper> focus_it: real money appears to be the motivation behind allwinner, not open source or community support, but I'm sure they like the free advertisements and any community contributions for the very little they released
<Turl> mnemoc: you seem to have fogotten about politicians
<mnemoc> Turl: they are mostly lawyers
<mnemoc> or MBAs
<mnemoc> either way, they are covered
<L84Supper> sociopaths in general, just start acknowledging them and try to rehabilitate them vs awarding them with power
<focus_it> L84Supper: its a lot less to do with money and more to do with deal making - the average China manager is honour bound to accept best deals for his workers and if better deals come along it all gets massaged in. There is no black and white as we have here.
MMlosh has joined #arm-netbook
<L84Supper> here = China for me
<focus_it> The way to get in more rapidly is to go in with money and offer to set up additional companies and hire workers as a subsidiary.
<focus_it> L84Supper: where? I didn't know. I was in Guanzhou City for 2 years.
<L84Supper> not far from where you were
<focus_it> Shenzhen? My favourite haunt?
<L84Supper> but usually farther north near Shanghai
<xxiao> focus_it: where are you now...SZ/GZ?
<focus_it> xxiao: in UK, missing Shenzhen and Guanzhou dearly!
* xxiao plans to live there in 2013 for a few months
<xxiao> AW is in Zhuhai, which is a few hours away by car or by boat
<focus_it> xxiao: in London? - give me a ring I buy you a meal!
<mnemoc> sounds more like .hj
<L84Supper> when we asked around 6-8 months ago there was little interest for the a10 except for tablets
<mnemoc> .hk
<xxiao> thanks. i'm in us now
<focus_it> xxiao: if you plan coming to London, give me a ring.
<xxiao> focus_it: sure :)
<L84Supper> I've been waiting to see what Samsung and Freescale would end up offering
<xxiao> mnemoc: what's missing for A10 as far as doc/datasheet are concerned
<xxiao> i thought AW did some nice jobs on releasing them
<xxiao> other than codec portioni
<xxiao> portion
<L84Supper> i.mx6 is still hard to come by and was very late, and Samsung doesn't want to sell to China except for large phone or tablet OEM's
<mnemoc> xxiao: the datasheet is merely ~90 pages, barely listing features and pins... which are incorrect compared to the reference designs
<L84Supper> wasn't somebody AW related up in Beijing?
<mnemoc> xxiao: the leaked "user manual" is ~400 pages and only covers an small set of the IPs
<xxiao> for ARM, the top is Apple-Samsung-Qualcomm, they do not exist for small players. middle is TI/FSL, they're exiting; low-end are AW, Rockchip,etc, they don't speak English
<focus_it> xxiao: I been trying to make pinouts of A10. I have datasheet with evaluation kit, but the pinouts for chip are inaccurate on the witstech datasheet dated 2012-04-06 - do you a better one with sata and other stuff 100% accurate?
<xxiao> L84Supper: AW is a southern china company, it may have applied start-up funding from Beijing though
<mnemoc> focus_it: 100% accurate is only the reference design, and the derivatives... like eoma68-a10, olimex's and cubieboard.
<xxiao> i think hipboi(?) should have all those documents
<mnemoc> xxiao: I was told the founders of allwinner left actions-semi to "do things better"
<mnemoc> xxiao: he doesn't
<xxiao> otherwise how could he make cubieboard/mele2000 working
<L84Supper> focus_it: I wouldn't put too much faith in the a10 data sheets
<focus_it> I can reverse engineer the cubieboard as they release release their board schematic - http://dl.cubieboard.org/hardware/cubieboard_schematic_2012-08-08.pdf
<xxiao> the pinout should be easy to reverse-engineer, when you have multiple schematic available now
<L84Supper> the a10 data sheet is more of a suggestion than a set of facts
<xxiao> it's those registers....
<focus_it> L84Supper: I'm coming from the spoon fed world of PICs and other CPUs and this situation with NDAs and the like is absolutely new and disgusting
<xxiao> focus_it: for 8/16 bit there is no IP, for 32bit/64bit esp with codec they do exist, thus the closed-info with NDA
<L84Supper> focus_it: my background is x86 motherboards since the 8086, so it's nothing new
<L84Supper> it just evolved differently in the east vs the west
<xxiao> sometimes, it's not the SOC holding the secrets, it's the IP owner where they licensed from
<focus_it> I'm doing a lot of stuff in NXP arms and the datasheets are complete. So its still a surprise to find the situation so different for these bigger ARMs and the SoCs.
<L84Supper> lots of IP is borrowed :)
<L84Supper> focus_it: AW just whipped the a10 together to be sold as a kit to OEMs with virtually no engineering talent, they just copy the design for 6-12 months and then move on to the next hot ARM SOC
<xxiao> right, it's after all a small company, even TI/FSL lags in documentation
<xxiao> the consumer market just evolves too fast
<focus_it> I realized that instead of making NXP ARMs and LCD product, I should make A10 ARM + LCD + Linux + Gambas and then I don't have to make the window managers and the like - its already done in Linux!
<xxiao> before it's fully documented the next new one comes out
<xxiao> focus_it: NXP is a loser, other than its NFC, sorry
<focus_it> certainly I agree! :-)
<focus_it> That why I want A10
<focus_it> The hardware companies would make more money if they document and sold the chip with documentation me thinks
<L84Supper> not much more and possible less
<L84Supper> they would get more low volume customers that will suck up design resources
<xxiao> the way in south china, where volume rules, is either the SOC vendor, or some IDH, get the gerber and images ready for others to manufature and ship
<L84Supper> they get lots of "oh yeah we can build that" when they hardly know which end of a soldering iron to hold
<L84Supper> one of the ways x86 chipset vendors could easily qualify new buyers was to ask them what layout tools they used
<L84Supper> if it wasn't allegro they would end the conversation
<focus_it> wow thats crude!
<xxiao> no wonder intel is sinking
<focus_it> how much is allegro?
<L84Supper> they can't hold customers hands during the design and layout
<L84Supper> not sure what the current price is, it used to be >$20K
<focus_it> I assume a KiCAD drawing is a conversation killer then?
rellla has joined #arm-netbook
<L84Supper> they supply Orcad Capture schematics and Allegro board files, also some gerber format since Allegro can import them back into layout
<xxiao> was the 4-core A7 from AW attractive ?
<xxiao> normally new designs will be on sale at about the same time when the chip is announced...
<L84Supper> as HW sure, but only if they share the source
<L84Supper> otherwise it's like another rockchip or amlogic soc
<xxiao> in the west, SOC vendors like FSL announce their new chip at some quarter earning meetings years before new stuff are order-able
<focus_it> Well there is opencores.org and their open source core running Linux. They asked for kickstarter funding but didn't get the full amount.
<focus_it> Someone could take that and start making SoCs. There is plenty of open source IP at opencores.org
<slapin> hno: ping
<focus_it> 16800 member
<slapin> hno: any particular register settings?
<focus_it> That is a lot of members
<L84Supper> focus_it: yeah, we've had this discussion before, there's just not that big a market for open hardware
<L84Supper> even open source advocates tend to choose low cost over open
<focus_it> L84Supper: you be right. Money first and then do instead of pontificating. I just want to get things done and I hate being powerless on the techie front.
<L84Supper> it's frustrating I know
<xxiao> i though Europe favors ATMEL? pretty open it seems
<L84Supper> IMHO it's back to the sociopath being put in charge too often, too much short term thinking
<xxiao> microchip should acquire mips for a good future, now it's too late
<focus_it> Atmel is not cheap compared to the big SoCs like A10, plus slow I believe last time I checked.
<L84Supper> they all use the same fabs
<L84Supper> the silicon costs the same
<L84Supper> it's matter of overhead and how they are financed
<focus_it> Microchip sole 1 billion PICs to Samsung (just one customer) in 9 months. ARM sells 6 billion processors in whole year.
<L84Supper> the 5 people at AW don't cost much :)
<xxiao> talking about china, the real chip vendor is hisilicon
<focus_it> Microchip "sold"
<xxiao> Huawei owns Hisilicon, it may compete with TI/Qualcomm one day
<L84Supper> look at the mess at Freescale, will they even be here in a year?
<xxiao> again it does not play with small shops
<focus_it> Microchip a big force in small embedded - they try to become big by making 32 bit PICs using MIPS cores.
<L84Supper> TI got out of the handset business
<focus_it> But 32 bit PICs are waaay too slow compared to ARMs out there
<xxiao> i noticed a sad truth, probably mentioned here before
<xxiao> that companies are nicer to FOSS, including TI/FSL, are losing the battle
<xxiao> those that do not, such as Qualcomm/Broadcom/Marvel/Nvidia, are doing better
<focus_it> TI make good arm - but not at good price and too slow their chips
<L84Supper> look at nvidia
rellla has quit [Quit: rellla]
revident has joined #arm-netbook
<L84Supper> a few years ago AMD offered to buy nvidia, but the CEO of nvidia only wanted to if he became the CEO of AMD/nvidia, now nvidia could buy AMD
<xxiao> nvidai can buy N*AMD
* xxiao knows too many people in AMD
<L84Supper> even VIA missed the boat every time
<L84Supper> the could have paired ARM with unichrome years ago
<xxiao> x86 is so yesterday anyway
<L84Supper> same with freescale, always a year too late
<xxiao> FSL will be a car-IC company only soon
<xxiao> hmm maybe they should buy OBD-II IC companies
<xxiao> TI will be a new born National
<L84Supper> Renesas is another
<focus_it> mnemoc: thanks for the olimex link - those schematics are a treat!
newbie has joined #arm-netbook
<L84Supper> ST also has some multi arm core soc's
<xxiao> got another email, why schools always push kids to buy more coke/junkie to earn points for their own funding?
<xxiao> sick of it
<L84Supper> whatever happened to the snowball dev boards?
newbie is now known as Guest3493
<xxiao> ST has the best line of M3/M4
<techn> now back to sata problem
<L84Supper> pin numbers?
<L84Supper> aren't all the schematics the same for sata
hp__ has quit [Ping timeout: 276 seconds]
<techn> Turl: drachensun: could you test ft5x_ts patches with android?
<Turl> I don't have any ft5x_ts device
<techn> hmm.. who had ft5x_ts android :/
<mnemoc> d
<mnemoc> err
<mnemoc> the pengpod guy iirc
<drachensun> it will take me a little bit, I'll have to fix some other things with my Android build
<drachensun> but yeah, I can
<techn> thanks
ajmitch has quit [Ping timeout: 264 seconds]
ajmitch has joined #arm-netbook
ajmitch has joined #arm-netbook
ajmitch has quit [Changing host]
jquip has quit [Quit: jquip]
<zoobab> there is now an openwrt port for A10
<zoobab> at least for cubieboard
<techn> huaah.. baseline implementation of achi_platform init seems to be buggy .. atleast it leaves memleak if init call fails
jquip has joined #arm-netbook
brooc has quit [Ping timeout: 245 seconds]
bsdfox_ has quit [Read error: Connection reset by peer]
tuliom has joined #arm-netbook
popolon has joined #arm-netbook
DEAT_ has quit [Ping timeout: 250 seconds]
jquip has quit [Quit: jquip]
eFfeM has joined #arm-netbook
alcides has joined #arm-netbook
alcides has joined #arm-netbook
alcides has quit [Changing host]
revident has quit [Quit: Combustible lemons? Bah, I bring you weaponized asparagus!]
vinifm has quit [Remote host closed the connection]
freakazoid0223 has joined #arm-netbook
j1nx_ has quit [Quit: Beware of programmers who carry screwdrivers.]
Gumboot has quit [Ping timeout: 248 seconds]
Workboot has quit [Ping timeout: 250 seconds]
Gumboot has joined #arm-netbook
hg_5 has quit [Ping timeout: 252 seconds]
eFfeM has quit [Quit: Leaving.]
tuliom has quit [Remote host closed the connection]
<focus_it> "is there an asterisk port for A10" sorry I don't know how best to ask that question! :-)
focus_it has quit [Quit: Leaving]
Guest3493 is now known as hp__
merbzt has quit [Remote host closed the connection]