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/
<hno> mnemoc, please use the "stable" branch (sunxi) for any released binaries. -current is really -current and can have any amount of upstream + our breakage in it.
<hno> It's fine for testing, but it's not something I would like random users to try installing.
<mnemoc> hno: those are intended for testing only...
<mnemoc> kernel is also build from the test branch
<mnemoc> built*
Sternennebel1 has quit [Quit: Leaving.]
<mnemoc> I'll rename the dir from stage-3.0 to testing-3.0
<mnemoc> hno: the purpose is to make easier for people to test changes before they go to the "stable" branches
avernos has quit [Quit: Leaving]
<hp__> darn should i or shouldn't i pre order an ouya.. tegra 3 - quad core.. hmmm
<Turl> tegra 3 yuck
<hp__> :)
<hp__> maybe you are right :)
<Turl> s/maybe //
<lundman> I'll get one i think, can't be worse than broadcom :)
<Turl> lundman: broadcom has open source userspace components :)
<lundman> yes, bare minimum
<lundman> slightly less than allwinner
<lundman> so, everyone is about the same, woo
<Turl> lundman: well they released their gpu stuff the other day right? :)
<Turl> lundman: tegra has 0 userspace source - literally
<lundman> just the binary blobs yep
* Turl reinstalls his MIPS vm after breaking it
benjamin__ has quit [Ping timeout: 240 seconds]
former has joined #arm-netbook
<former> /me
<former> 会让她
former has left #arm-netbook ["(><((">BokaIRC<"))><)"]
<lundman> ちょっと
<WarheadsSE> Turl: saw my message
<Turl> WarheadsSE: about xbmc? yep
<Turl> WarheadsSE: how fresh are the rootfs's these days?
<lundman> 2, maybe a 3
<WarheadsSE> Turl: I need to build a new kernel, otherwise I can get you a systemd converted rootfs tomorrow.
<WarheadsSE> Any tidbits wanted on the kernel never hurts to do a pull request :p
<Turl> :)
mysteryname has joined #arm-netbook
popolon has quit [Quit: Quitte]
<Gumboot> I guess that on an SoC one could just as easily veil the drivers by calling into a hypervisor.
<lundman> does ARM have the MMU calls to allow a hypervisor
<Gumboot> Umm... they're instructions I never had to bother with so I don't really remember them. There's an alternate SWI (which has changed its name to something I can't remember) to get into trustzone-managed space.
<lundman> its possible.. wasnt that long ago ARM didnt have an MMU at all :)
<lundman> RE was real easy back then
<Gumboot> SVC is standard, SMC is the security thing.
stefanro1 has joined #arm-netbook
<Gumboot> What have allwinner opened up, by the way? I'm a bit confused on that.
<Gumboot> I downloaded a "datasheet" from their site a day or two ago, but that was no use.
RITRedbeard_ has joined #arm-netbook
stefanro has quit [Ping timeout: 268 seconds]
<ZaEarl> Gumboot, lots of the drivers for the components on the A10.
<Marex> hp__: tegra 3 is icky and crap
<Marex> and it often misbehaves
Thomas42 has quit [Ping timeout: 252 seconds]
NAiL has quit [Ping timeout: 252 seconds]
NAiL__ has joined #arm-netbook
Thomas42 has joined #arm-netbook
drachensun has quit [Quit: Leaving]
Quarx has joined #arm-netbook
Guest65804 has quit [Ping timeout: 245 seconds]
ZaEarl has quit [Ping timeout: 260 seconds]
cheng has joined #arm-netbook
tzafrir_laptop has quit [Ping timeout: 260 seconds]
ZaEarl has joined #arm-netbook
tzafrir_laptop has joined #arm-netbook
tekzilla has quit [Ping timeout: 255 seconds]
slash_random has quit [Ping timeout: 256 seconds]
tekzilla has joined #arm-netbook
furan has joined #arm-netbook
furan is now known as Guest83996
t0dbld has quit [Ping timeout: 260 seconds]
<rm> > could make the board a pretty good native ARM build machine under Ubuntu
<rm> what's with everyone's obsession with Ubuntu
<rm> especially as he later says "e.g. for Debian packages"
mysteryname has quit [Remote host closed the connection]
mSquare has joined #arm-netbook
eebrah has quit [Remote host closed the connection]
oliv3r has quit [Ping timeout: 248 seconds]
ol1ver has quit [Ping timeout: 240 seconds]
ZaEarl_ has joined #arm-netbook
ZaEarl has quit [Read error: Connection reset by peer]
oliv3r has joined #arm-netbook
ol1ver has joined #arm-netbook
Sv is now known as discopig
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 260 seconds]
mSquare has joined #arm-netbook
L84Supper has quit [Ping timeout: 256 seconds]
mSquare1 has quit [Ping timeout: 265 seconds]
L84Supper has joined #arm-netbook
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 260 seconds]
mSquare1 has quit [Ping timeout: 246 seconds]
mSquare has joined #arm-netbook
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 240 seconds]
mSquare1 has quit [Ping timeout: 264 seconds]
mSquare has joined #arm-netbook
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 260 seconds]
mSquare has joined #arm-netbook
mSquare1 has quit [Ping timeout: 260 seconds]
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 260 seconds]
von_fritz has joined #arm-netbook
<oliv3r> Gumboot: look at the a10 manual, its on the wiki and on the interwebs
mSquare1 has quit [Ping timeout: 255 seconds]
mSquare has joined #arm-netbook
pawel5870 has joined #arm-netbook
toxicpsion has quit [Ping timeout: 260 seconds]
mSquare1 has joined #arm-netbook
mSquare has quit [Ping timeout: 240 seconds]
AndChat138129 has joined #arm-netbook
mSquare has joined #arm-netbook
mSquare has left #arm-netbook [#arm-netbook]
mSquare1 has quit [Ping timeout: 260 seconds]
AndChat138129 has quit [Ping timeout: 256 seconds]
AndChat|138129 has joined #arm-netbook
AndChat|138129 has quit [Ping timeout: 256 seconds]
pawel5870 has quit [Ping timeout: 255 seconds]
AndChat138129 has joined #arm-netbook
AndChat|138129 has joined #arm-netbook
AndChat138129 has quit [Ping timeout: 256 seconds]
AndChat|138129 has quit [Ping timeout: 256 seconds]
AndChat138129 has joined #arm-netbook
RITRedbeard_ has quit [Ping timeout: 244 seconds]
AndChat138129 has quit [Ping timeout: 256 seconds]
rellla has joined #arm-netbook
AndChat138129 has joined #arm-netbook
AndChat|138129 has joined #arm-netbook
AndChat138129 has quit [Read error: Connection reset by peer]
AndChat|138129 has quit [Ping timeout: 265 seconds]
slapin_nb has quit [Ping timeout: 256 seconds]
AndChat138129 has joined #arm-netbook
The-Compiler has quit [Quit: WeeChat 0.4.0-dev]
The-Compiler has joined #arm-netbook
AndChat138129 has quit [Ping timeout: 265 seconds]
pawel5870 has joined #arm-netbook
slapin_nb has joined #arm-netbook
Sternennebel has joined #arm-netbook
AndChat138129 has joined #arm-netbook
cheng has quit [Quit: Leaving]
AndChat138129 has quit [Ping timeout: 265 seconds]
AndChat138129 has joined #arm-netbook
AndChat|138129 has joined #arm-netbook
AndChat138129 has quit [Read error: Connection reset by peer]
AndChat|138129 has quit [Ping timeout: 244 seconds]
AndChat138129 has joined #arm-netbook
AndChat138129 has quit [Ping timeout: 244 seconds]
<oliv3r> hipboi is alive! he just responded to an e-mail on the ML :)
<Gumboot> oliv3r: The datasheets are strictly electrical data. Is there more?
<mnemoc> \o/
<oliv3r> Gumboot: the manual has register information
lundman has quit [Ping timeout: 255 seconds]
<Gumboot> Link?
AndChat138129 has joined #arm-netbook
AndChat|138129 has joined #arm-netbook
AndChat138129 has quit [Read error: Connection reset by peer]
AndChat|138129 has quit [Client Quit]
<Gumboot> What's this "DECRYPTED" business? Should I even be looking at that?
<jelly-home> Gumboot: well it's not tegra!
<oliv3r> Gumboot: the original leaked pdf was 'encrypted' as in, you couldn't print it nor copy paste it
<oliv3r> it was simply annoing to work with or to select text from it
<oliv3r> but it is pure A10 stuff, nothing related to tegra :p
<Gumboot> I think it's best if I restrict myself to looking at public documentation. Not leaks.
jlj has quit [Ping timeout: 256 seconds]
<Gumboot> (and, obviously, private documentation which I'm fully entitled to view)
jlj has joined #arm-netbook
popolon has joined #arm-netbook
pawel5870 has quit [Ping timeout: 252 seconds]
<ZaEarl_> Gumboot, the pdf includes a "For WITS only" watermark, so if you have any concerns you best not download it.
<Gumboot> Noted. Thank you.
specing has joined #arm-netbook
specing has quit [Changing host]
specing has joined #arm-netbook
lundman has joined #arm-netbook
tzafrir_laptop has quit [Ping timeout: 260 seconds]
lundman_ has joined #arm-netbook
lundman has quit [Ping timeout: 255 seconds]
lundman_ is now known as lundman
<mnemoc> someone leaked it to the cubieboard ML a while ago
specing has quit [Read error: Connection reset by peer]
<stefanro1> mnemoc: is there a reason why we don't squash/combine the two u-boot images while compiling/generating into one single image?
<stefanro1> mnemoc: thats how its usually done for platforms needing those two images - less error prone etc
<mnemoc> i guess it's because we have spl-less boards too
<mnemoc> but don't now
<stefanro1> perhaps hno knows for sure
specing has joined #arm-netbook
<stefanro1> mnemoc: we don't need to combine those images for all sunxi boards - it can be configurable on a board-per-board basis
<mnemoc> it might just be legacy from the initial port hipboi did
<stefanro1> perhaps i'll just give it a try on the cubieboard and send an initial patch...
slapin_nb has quit [Ping timeout: 276 seconds]
slash_random has joined #arm-netbook
<slapin> rehi, all
<mnemoc> wb slapin
<slapin> that pdf contains only some information. No NAND nor MMC controller information there
ZaEarl_ has quit [Ping timeout: 260 seconds]
<mnemoc> you might find something more in the datasheet of it's predecessor (the F20) but basically the only documentation we have is the code
specing has quit [Ping timeout: 255 seconds]
merbanan has joined #arm-netbook
<slapin> ah, yeah, I remember, I just can't get rid of frustration
<slapin> by the way, any AW A10 device with CAN? Is there any industrial version?
<merbanan> yey, working console on the mini-x
<andoma> merbanan: !
tzafrir_laptop has joined #arm-netbook
<mnemoc> slapin: we don't have any knowledge about the can controller. they might use it on their RTOS, but for linux we have nothing
<mnemoc> slapin: if you have a $project$ allwinner might be willing to give you registers info to write one
mSquare1 has joined #arm-netbook
mSquare1 has left #arm-netbook [#arm-netbook]
Sternennebel has quit [Quit: Leaving.]
<slapin> mnemoc: too bad for them, since I better use expernal CAN in such case. Since I prefer doing planning stage before all other stages. I have lots and lots of $projects$ and only 1-2% of them end up being real production devices. But to gain money as developer I need to make platform choice to start develop on, I like AW SoCs and I'd like to be able to convince people to use them. But to make such choice I need to see datasheet.
<slapin> but this non-disclosure stuff prevents me from doing research on market, which frustrates me
<slapin> then I'll go to fsc or ti or atmel or whatever and will get information I need without such weird stuff
<slapin> and without promises about buying 1000s of devices - my generic volume is 100 devices per project. I have really low chance to interest AllWinner or any other high-volume supplier.
<slapin> blah-blah-blah, etc.
specing has joined #arm-netbook
QingPei has joined #arm-netbook
atef has joined #arm-netbook
rellla has quit [Quit: rellla]
<libv> 533Mhz mali-400?
<mnemoc> the odroid2 thingy?
<libv> ~60Mtriangles, 2.132Gpixels
<libv> pretty respectable for mobile hw
<libv> and yes :)
<atef> pandora PI
atef has quit [Ping timeout: 245 seconds]
<oliv3r> Gumboot: you should however be able to look at http://linux-sunxi.org/A10/registers
specing has quit [Read error: Connection reset by peer]
Almamuetya10 has joined #arm-netbook
orly_owl has quit [Quit: leaving]
<libv> does sunxi-tools/bootinfo provide anything needed for uboot?
<libv> because i appear not to be allknowing enough to find out how to use it :)
<mnemoc> boot from nand, get into u-boot, and dump boot1
<libv> ah, it uses the first 2058 sectors
<libv> 2048 even
<libv> right
<mnemoc> it will be more useful after slapin gets us a fel-able raw nand extractor :)
<libv> slapin: what's keeping you :p
<mnemoc> because we currently can only get the rests of boot1 left on memory
<RaYmAn> you can dump boot0 (and 1) with jtag quite easily, but yeah - doesn't really help users much :P
<mnemoc> needing people to get a serial console or jtag on each device we want to liberate doesn't scale well
<libv> mnemoc: does sunxi-boards just need the raw dumped .fex, or is more work needed before that is accepted in?
<mnemoc> we currently only have use for .fex
<mnemoc> not sure what else can be useful
<mnemoc> boot1 lets us "fix" the dram_para sections in the .fex, then used to generate u-boot's dram initialization data
<libv> i get some errors with dram generation
<mnemoc> paste?
<mnemoc> libv: can you show me the [dram_para] section of .fex?
<mnemoc> those 5 fields giving you errors are the fields probed by livesuit when generating boot1
<mnemoc> they are usually NULL.... but that shouldn't give you that error
<libv> oh, yes, nulled indeed
<mnemoc> but with strings.... yuck
<libv> well, empty-stringed
<libv> email the .fex to the list?
<mnemoc> not yet
<libv> ok
<mnemoc> first we need to get those fixed
<mnemoc> option 1) steal those values from another 1GB board. cubie/hackberry ?
<mnemoc> option 2) try to get the boot1 dump from nand's u-boot
<libv> what needs to be fixed? these dram params?
<mnemoc> those ""
<mnemoc> we need real numbers
<mnemoc> `md.b 0x42400000 0x82d0` on nand's u-boot should do the trick
<libv> ok, then i first have to find out how to make it into u-boot :)
<mnemoc> after fixing the .fex, and testing the dram.c actually boots, the .fex is welcomed :)
<mnemoc> option 3) is guessing the value from the pictures you took
QingPei has quit [Quit: Leaving.]
QingPei has joined #arm-netbook
QingPei has quit [Ping timeout: 244 seconds]
specing has joined #arm-netbook
<oliv3r> PLL6 up on the wiki
<mnemoc> \o/
popolon has quit [Quit: Quitte]
QingPei has joined #arm-netbook
QingPei has quit [Client Quit]
<libv> i have too many toys
<slapin> libv: same here, and work
<libv> the hyundai has been on my desk for quite a while now, and my gf still hasn't noticed that this is a new device
<libv> i still had no fingerwagging yet
<mnemoc> :)
<slapin> libv: I have stack of ~20 tablets on my table, my wife hates them, my kid played with them a bit and have not managed to wreck any. I still fali to beleive what happens these days. $50 tablets with $100 screens.
<slapin> s/fali/fail/
<ibot> slapin meant: libv: I have stack of ~20 tablets on my table, my wife hates them, my kid played with them a bit and have not managed to wreck any. I still fail to beleive what happens these days. $50 tablets with $100 screens.
<L84Supper> libv: have a favorite hackable phone yet?
<slapin> ah damn
* slapin still fails to figure out NAND init sequence
<slapin> L84Supper: is there any except n900?
<slapin> L84Supper: phones from China are too closed, without kernel source anywhere :(
<L84Supper> slapin, I have a collection of arm11 and older tablets. I only bought two a8 tablets on closeout that are post close/lock the bootloader
<slapin> L84Supper: tablets are different thing, I've got only one with closed u-boot (protected by password)... and some with somewhat b0rked (never stops booting)
<L84Supper> slapin, I have associates there with the source since they work for the handset makers. They have done some pretty fancy reprogramming of their phone features
<L84Supper> still have to see what can make it out into the wild
<slapin> L84Supper: I've got 4 different 2/4 SIM card Android-based phones, all on the same processor, without any documentation found anywhere
<slapin> L84Supper: I'd make these into personal trackers, or special vdr, or as motion capture devices (one have 3 cameras), if I had access to kernel/boot loader source code
<slapin> as android phones they are very poor, due to lack of RAM, but with normal Linux I could make them into specialized but very powerful beasts
<L84Supper> I haven't been following the secure bootloader and unlock vectors in tablets and phones too closely the past year
<L84Supper> somebody had reversed the omap's in the Milestones
<slapin> these doesn't contain any seen secure booting, at least I see u-boot and Linux kernel booting on soldered serial port
<L84Supper> oh nice, ca you hare the make and models, I'd like to look at them
<L84Supper> can you share
<slapin> L84Supper: these are not here, probably can dig them out from trash pile on weekend
<L84Supper> slapin: heh, what cpu/soc?
<libv> my feeling is that nanda/boot.ini is the key to entering uboot, right?
<slapin> L84Supper: MTKsomething, don't remember...
<L84Supper> mediatek
<oliv3r> Did I say pll7 was up? PLL7 is up, time for home now :)
<slapin> L84Supper: MT6573
<libv> aha... hold '1' while pressing board reset
* mnemoc still waiting for his JiaYu G3 (mtk6577, 720p, 4.3")
<L84Supper> slapin, mediatek is going after the $100 and lower smartphone market in China, not many of these are sold outside of China
<L84Supper> slapin: there are very few devs there that could unlock them or dare to share how, unless it becomes common to make $
<jelly-home> mnemoc: sounds vaporwarish :-|
<arete74> mnemoc:who you buy JiaYu G3?
<mnemoc> arete74: http://lightake.com/
<mnemoc> but i think i'll cancel it and look for other mtk6577 720p device
<libv> mnemoc: actually, are there no full images available for these devices, full images that can be pried apart with some tools?
<libv> mnemoc: could the necessary information be found from those?
<libv> another route would be to get this information out of registerspace after boot
<mnemoc> don't know.... i only want a cheap but good phone for my daugher
<hno> stefanro1, you mean combinind SPL and the main u-boot image? Sure we can combine them, but there is people who use SPL with other versions of u-boot.
RITRedbeard has quit [Ping timeout: 264 seconds]
<libv> hno: as our main u-booter, would the register route be feasible?
<libv> finding out ram setup from registers?
<hno> steev, and when moving to nand they are most likely going to be flashed separately.
<hno> libv, what?
<steev> wat
<steev> oh
<libv> crap, it lost it already
<jelly-home> haste indeed
<mnemoc> :)
<hno> libv, are we talking about A10 DRAM conifiguraiton, or something else?
<libv> yes
<mnemoc> libv: these guys seem to have images and tools, http://mtksj.com/
<libv> mtk?
<L84Supper> mtk = mediatek
t0dbld1 has joined #arm-netbook
<libv> mnemoc: i am talking about the a7hd here, not some mediatek thing
<hno> Don't really trust script.bin any longer. But the missing fields can easily be determined by looking at the board.
<libv> hno: ok, just the make and model of the ramchips?
<mnemoc> libv: :)
<hno> libv, the missing parameters are basically: How much DRAM, and in how many chips?
<libv> 1GB, looking up hynix partnrs right now
<hno> io_width is the I/O data bus width of each chip.
<lkcl> hynix have a good web site which allows you to decode the info off their ICs
<hno> bus_width is the total data bus width. Generally 32 for A10 devices. It also supports running with crippled 16-bit memory bus but haven't seen any of those.
<libv> 4 chips, as 2Gb each
<libv> ddr3
<hno> 8 or 16 data bits each?
<lkcl> libv: is it the H5TQ2G863BFR
<lkcl> yep then it's the x8
t0dbld1 is now known as t0dbld|work
<t0dbld|work> identify t0dbld toddpatrick
t0dbld|work has quit [Changing host]
t0dbld|work has joined #arm-netbook
* libv sticks his head in the mele and curses
<lkcl> that's the same as in the A10 CPU Card, and also wits-tech's EVB.
<lkcl> :)
<libv> same as in the mele :p
<L84Supper> mnemoc, http://mtksj.com/ yeah, looks like somebody has access to docs there
<hno> libv, then rank=1, io_width=8, bus_width=32, density=2048, size=1024
<libv> hno: smashing, thanks
<hno> It's silly that DRAM have no auto-identification.
<libv> on standard pc simms, there is a special chip for that, go figure :)
<libv> it's i2c to a separate chip
<libv> cheaper than doing i2c to each individual chip
<mnemoc> but somehow livesuit knows to probe them
<libv> because even that will not tell you the complete pictuer of the simm
<L84Supper> quad core MT6589 ships in Q1, mt6573 is EOLed Q1
<libv> in any case, cool, i have a .fex and i have a dram.c
<libv> seems like i should now be able to get uboot going on this device
RITRedbeard has joined #arm-netbook
<mnemoc> libv: \o/
<hno> libv, maybe. But yours have a battery... have ignored anything related to batteries so far.
<libv> slapin: how are you running your kernel on a7hd anyway?
<libv> hno: would that change things for uboot then?
<L84Supper> on DRAM ID would never work since the vendors would never cooperate, dimm/simm ID's aren't even trustworthy and they actually have to be programmed by the vendor that makes the module
<hno> I hope not, but the Allwinner bootloader messes a bit with battery settings.
<L84Supper> RAM init in x86 always requires probing and verifying what RAM is actually there even if you read the modules ID and info
<hno> But I think the kernel also resets all of the battery settings.
<L84Supper> it will have to be the same with ARM mainboards that use dimm/simm, if they ever appear
<hno> L84Supper, ARM mainboards that use dimm modules is already here.
<hno> obviously not in tablets, but on the server side.
<L84Supper> yeah, nothing mainstream yet, it will be interesting to see what the major players decide to do with ARM 64b
Sternennebel has joined #arm-netbook
<libv> now let's see whether it wants to boot this sd card
<specing> L84Supper: just forget it, AMD will fail to deliver again
<libv> nope, choses android :(
<libv> ah, but i now have found how to get into uboot
<L84Supper> specing: unfortunately that has been the pattern, I'm wondering what we are actually going to see for ARM server boards. All the nice parts are currently kept out of the hands of mere mortals
<L84Supper> is any vendor shipping ARM servers systems besides Dell and Calxeda?
<libv> hah, the kernel boot halts exactly where my mele halts
<libv> watchdog keeps hitting the same bit
<Turl> mnemoc: get a nexus 4 for yourself and hand her over your gnex :P
<mnemoc> nah
<L84Supper> libv: is the watchdog not getting reset? or?
wingrime has joined #arm-netbook
<libv> wait a second, could it be that just the serial vanishes?
<libv> last line: <6>serial8250.0: ttyS0 at MMIO 0x1c28000 (irq = 1) is a U6_16550A
<Turl> yeah I've seen it doing that
<Turl> usually on reboot though, not on first boot :<
<Turl> libv: 3.0 or .4?
<libv> first time round, the watchdog kicks in, and you get the same "lock" next round, and then nothing anymore until the powerbutton is held for a while
<libv> 3.0
<Turl> unplug serial from PC, unplug power, plug power, plug serial
<libv> ok, a bit tough on a tablet
<Turl> s/unplug power/halt using ~10s power button press/
<ibot> Turl meant: unplug serial from PC, halt using ~10s power button press, plug power, plug serial
<Turl> s/plug power/boot tablet/
<ibot> Turl meant: unplug serial from PC, unboot tablet, plug power, plug serial
<Turl> ha, unboot tablet :P
<libv> heh, got alip in front of me now, seems to have worked
<libv> and it seems that now the sd card is being booted from
<libv> nice.
<libv> slapin: i have panel.
* libv spends a minute doing happy dance
<mnemoc> libv: now your fixed .fex and u-boot board patches are welcomed :p
<libv> mnemoc: gee :p
<libv> Turl: that kernel halting issue is quite irritating though
<Turl> yeah it is :<
<libv> but it happens on the mele as well
<Turl> there's a worse one though
<mnemoc> mine halts much earlier
<Turl> if you leave serial connected while you powercycle uSD slot doesn't get power
<Turl> or something
<Turl> and it boots from nand :<
<mnemoc> haven't seen that one in a while
<Turl> I get it every single time the device hard locks and needs to be powercycled
<Turl> and if you don't unplug serial before, it'll boot from nand
<Turl> normal reboots are unaffected
* libv attaches a working a7hd picture to the wiki
specing_ has joined #arm-netbook
<xxiao> .
slapin has quit [Remote host closed the connection]
specing has quit [Ping timeout: 260 seconds]
slapin has joined #arm-netbook
<slapin> rehi, all!
<Gumboot> oliv3r: Yeah, that's cool. I was just trying to figure out what level of support was coming out of them, right now.
<slapin> how am I supposed to use bootinfo from sunxi-tools?
popolon has joined #arm-netbook
<L84Supper> "if you don't unplug serial before, it'll boot from nand" whats the point of hardware with this design? is this some intended boot order checker?
<Turl> L84Supper: sounds like a hardware bug to me
<L84Supper> weird stuff
<mnemoc> they didn't expect users to connect to the uart
<L84Supper> the load on the serial port is next to nothing
<L84Supper> I see so much bad design these days, I really don't know what they teach EE's anymore or where they get some ideas
<mnemoc> to deliver "something" fast and cheap
<Turl> boots? SHIPIT
<mnemoc> :)
<libv> wifi works :)
<Turl> nice pic libv :)
<Turl> libv: what chip does that one have, realtek?
<libv> rtl
<libv> yeal
<Turl> what driver did you use?
<Turl> rtlwifi or the softwinnerized one?
<libv> the one built as standard sun4i defconfig
<Turl> the softwinnerized one then
<libv> same one in the mele i guess
<libv> so this really is a bit of a mele with a panel
<Turl> hm new driver release from realtek, gotta update it :)
<libv> driver ain't half noisy though
<hno> slapin, bootinfo /path/to/binary/with/bootheader
<hno> BROM/boot0/boot1
<L84Supper> almost works™
<RaYmAn> woo, my chromebook just shipped :D 10 days early :D
* RaYmAn <3 amazon
<t0dbld|work> nice RaYmAn
<t0dbld|work> haha me too
<t0dbld|work> my laptop no ship though
<t0dbld|work> :-(
<t0dbld|work> ill have to go on my vacation iwth out a new laptop its looking like
<t0dbld|work> hmm well i can get a 8gb stick for like $30 so if thats true not bad and get 10gb mem in it lol
<t0dbld|work> upps worng window
<t0dbld|work> :-)
* slapin managed to execute fel on stick
<slapin> a problem is there's nothing to talk with it
<hno> slapin, there is our little test1.
specing_ has quit [Ping timeout: 246 seconds]
* slapin damns people who designed cheap cp2102 cables and managed to ignore common ground requrement
<hno> slapin, you can also upload u-boot-spl I think.
* slapin had to extract board from compound "enclosure" with solder fan and fix it
Quarx has quit []
<slapin> hno: I need some small app which I could quickly build and use it to read/write nand regs
<slapin> hno: is there something to dump memory locations?
<slapin> anything with text output?
<mnemoc> look at fel-pio.c and fel-gpio
specing has joined #arm-netbook
<slapin> hno: is NAND on dedicated pins or these have to be configured for function?
<hno> fel can read/write anything accepting byte accesses. Most I/O only accepts word aligned I/O.
<hno> slapin, pins ned to be configured, clocks enabled.
<techn> libv: nice.. you got trace of my problem with tablets! :)
merbanan has quit [Ping timeout: 246 seconds]
<hno> and nand controller configured
<libv> techn: not sure that your sentence parses correctly?
<libv> techn: are you talking about the reboot issue with mmc with serial attached?
<techn> libv: to get it boot you need to press 7 seconds power and then press power again twice :p
<hno> slapin, for an easier ride I would recommend using u-boot mr/mw commands. Boot nand u-boot and you'll get clocks and pins configured already.
<slapin> hno: so to configure nand from fel helper program is needed?
lkcl has quit [Read error: No route to host]
<techn> libv: I'm talking about boot problem with tablets with hno's u-boot
<slapin> hno: my nand u-boot disallows entering command prompt
<techn> .. or let me read to end.. did you get it solved
<libv> techn: it booted
<hno> slapin, yes.
<slapin> hno: so I can't do it
<hno> slapin, build your own and drop in nanda
<libv> techn: so what boot problems are you having then?
<slapin> hno: do not want
<hno> why?
<slapin> hno: too complex task for my current mood, I'd better have fun with sources and numbers
<techn> libv: I have two tablets(a10/a13) and both of them has problem that they wont start on first power cycle.. I need to do that 7second reset
<techn> libv: or did you solve that problem already?
<slapin> hno: awwww!!! I remember, I have almost configured u-boot, showing me nice NAND device id, might be ok with values.
<libv> it refuses to boot on simply pressing power after a full powerdown, yes
<slapin> hmmm, it'd be interesting to use NAND using u-boot scripting...
<techn> libv: and that is same problem as with mele?
<libv> not sure
<libv> i guess it might be due to how the power management chip is wired up
<libv> simply pressing power on android also did not immediately give me a booting android
<libv> techn: do you have data points for the opposite?
Turl has quit [Changing host]
Turl has joined #arm-netbook
<rz2k> techn: you have #error reset the board# or just panic at random point?
<rz2k> on A10 tablet
<rz2k> because random panic was fixed with I2C initialization of AXP209
<rz2k> atleast on mele/cubieboard
<Maqs> does the sunxi-uboot stuff have menu support?
<mnemoc> o.o
<Turl> libv: how do you kill the ultimate "<3>[ 126.470000] disp: process 321 (mScreenBrightne) has skipped the version handshake." spam? :)
<libv> Turl: ?
<libv> where does that thing live?
<Turl> # dmesg|grep handshake -c
<Turl> 196
<libv> again, where does that thing live?
<Turl> it's userspace, probably liblights
<libv> what userspace?
<libv> android?
<Turl> yep
<Turl> more /dev/disp usage for your collection
<Turl> :P
<hno> slapin, u-boot scripting is quite nice for banging I/O registers. But is missing a read into variable.
<libv> ok
<libv> Turl: does liblights open/close /dev/disp constantly?
<libv> because this message comes only once normally.
<slapin> hno: there were some mem to variable thing IIRC...
<slapin> or was it was
<libv> anyway, currently setting up git send-email for the hyundai stuff
<libv> then i am splitting out the disp reload fixes, and pushing that out
<libv> then i am fixing the version.c
<Turl> libv: possibly, it opens one per lights hal instance
<libv> seems like we need a sysfs entry for this btw
<libv> but debugfs, sysfs and all other sane kernely things were a bit alien to vito and whonot
<hno> slapin, I don't remember any.
<Turl> yeah, usually lcd brightness is exposed on the leds thingy on sysfs
<Turl> libv: can you test the top 3 commits from https://github.com/allwinner-dev-team/linux-allwinner/commits/allwinner-3.0 and see if wifi still works?
ZaEarl has joined #arm-netbook
<techn> rz2k: libv: I dont have serial cable
<libv> Turl: unless it is burning hard, please give me a few h still, will do so before i update version.c :)
<Turl> libv: sure, no hurry :)
<libv> techn: serial is pretty crucial imho :)
<libv> techn: well worth the destruction of at least the looks of a device :)
<techn> libv: true.. but I have mini-x which works.. so no need to break things :p
<techn> I can wait that someone fixes that problem.. During that time I can whine ;)
freakazoid0223 has quit [Remote host closed the connection]
specing has quit [Ping timeout: 260 seconds]
specing has joined #arm-netbook
<libv> someone please foster the allwinner usb stack...
* libv just got a nice kernel panic with otg and a mouse
<libv> hrm, my laptops system time is off :(
<Turl> libv: a bug report on the issue tracker with the panic output would be good I guess
* Turl runs ntp on all the things
<libv> yeah, fixed now
<libv> why desktop distros do not come with ntp pre-installed and enabled, i do not know
<jelly-home> perhaps they prefer chrony
<jelly-home> perhaps ntpd still fails to work well when a machine isn't up 24/7
slash_random has quit [Ping timeout: 256 seconds]
Turl has quit [Quit: >.<]
Turl has joined #arm-netbook
Turl has quit [Excess Flood]
Turl has joined #arm-netbook
Turl has quit [Client Quit]
Guest95881 has joined #arm-netbook
Guest95881 is now known as Turl
slash_random has joined #arm-netbook
<slapin> do anybody remembers latex markup? I try to make datasheet-style nand registers description
<Turl> \o/
<Turl> \LaTeX FTW :)
<RaYmAn> latex ftw indeed
<mnemoc> slapin: oliv3r uses latex in our wiki pages about the PLLs :)
<mnemoc> slapin: problem with that is we are still learning about the A10 registers, the data we have is not set in stop
<mnemoc> stone
arete has joined #arm-netbook
<mnemoc> even info in the official datasheets and user manual contradicts the reality we see on the code
<RaYmAn> mnemoc: chromebook arrives tomorrow! :D
<mnemoc> RaYmAn: what? *arrives*?
<mnemoc> to .dk?
<RaYmAn> according to UPS
<RaYmAn> They say delivery is tomorrow
<mnemoc> NICE
<RaYmAn> it's awesome :D
<mnemoc> when I tried amazon.co.uk didn't want to ship it to .es
<RaYmAn> first amazon ships it ~9 days early, then UPS expedites the shipment so it's 1 day delivery :D
<mnemoc> lucky bast*
<RaYmAn> (I only paid for the next step up from free shipping - so like 7GBP)
<mnemoc> :)
<RaYmAn> too bad they didn't ship there
<RaYmAn> heh
<RaYmAn> but you can get n4, n7 and n10 directly from play store :(
<RaYmAn> I have to order by proxy
<hno> mnemoc, it is not OK to copy material from the leaked user manual to linux-sunxi wiki.
<mnemoc> hno: afaik he is not copying
<mnemoc> but using it together with the sources to get better understanding and then document
<hno> http://linux-sunxi.org/A10/TZPC can not have any other source.
<mnemoc> :(
<hno> It is not OK to use the user manual as base for what is written in the wiki. Using it to clarify one or two values is ok, but not copying noticeable amounts.
<mnemoc> oliv3r: ---^
<hno> If there is no other good source, then document the differences found relateive to the user manual.
<mnemoc> techn: does that solve the purple screen problem?
<techn> mnemoc: That's still solution proposal.
<techn> I dont have hdmi-dvi dongle
slash_random has quit [Ping timeout: 256 seconds]
<techn> ..so cant test. But hdmi-hdmi seems to work.. but blacklevel's seems to be bit wrong :/
<techn> .. atleast YUV had better black levels imho :/
<mnemoc> techn: tell the guy of the ticket to try it :)
<techn> mnemoc: Will NB activate automaticly for new boards?
<WarheadsSE> So, anyone got a particuluar commit that would be good for the tag on the kernel build?
<mnemoc> techn: could make it so.... but we aren't ready to deprecte cnxsoft's nightly builds yet
<mnemoc> need to add support for variants first
wingrime has quit [Ping timeout: 240 seconds]
* techn starts to study livesuit
<mnemoc> been able to make livesuit image from any rootfs will be nice
<WarheadsSE> mnemoc, Turl any point of suggestion, or just pull current master?
<mnemoc> WarheadsSE: i planned to merge the stage branches tomorrow. you could test that (stage/sunxi-3.0)
<techn> hopefully that merge can be tagged :p
<WarheadsSE> Mkay.. but I was looking for best *stable*
<WarheadsSE> so i was about to try the last patch to v3.0 android v2
arete has quit [Quit: Leaving]
<mnemoc> i personally consider stage/sunxi-3.0 more stableish than sunxi-3.0 at the moment
<WarheadsSE> and we're looking at amery/linux-allwinner/?
<mnemoc> techn: depends on the fix for your regression :)
<mnemoc> WarheadsSE: no, linux-sunxi/linux-sunxi
<WarheadsSE> ah
<WarheadsSE> moved.
<mnemoc> same for u-boot
<WarheadsSE> yay wiki out of date.
<WarheadsSE> yeah I have word I should be using a newer uboot :p
<WarheadsSE> sata clock and other jazz
<mnemoc> if you go to amery/linux-allwinner/ you'll see some notes about the move
<WarheadsSE> I didnt read down far enouch :p
<mnemoc> https://github.com/amery/linux-allwinner .... not down at all :)
<mnemoc> the default branch is empty :)
<WarheadsSE> my browser has settings svaed to not-default :p
<WarheadsSE> stage3.0 tou say
<mnemoc> yes
<mnemoc> after a pending fix from techn that will be tagged the last 3.0.42-rN and then jump forward to 3.0.50
<WarheadsSE> going to have to redo my config :po
<mnemoc> is it me or it looks awfully thick?
<libv> vivante :(((
<techn> which .fex file is this?
<mnemoc> cubieboard's I would guess
<mnemoc> techn: ah, no. that's the partitioning info
<mnemoc> techn: it's pretty much generic
<hno> libv, the article writer pretty much have the same opinion there.
<libv> hno: well, if anyone comes around to me again in future and wants to RE the pvr again, i will point him to vivante
<libv> but there is currently noone working on vivante
<techn> mnemoc: verified :)
jgay has joined #arm-netbook
<libv> as for performance, the odroid-x2 with its 533 mali-400 is pretty top range
<hno> libv, is there anyone noticeable using vivante?
<mnemoc> techn: have a url for me?
<libv> hno: some previous gen ainols also used vivante
<libv> they are reasonably common
<libv> but not as common as pvr or mali
<hno> slapin, any progress?
<mnemoc> techn: it would be nice to have so text to add to the body of that commit
<jgay> I have a vivante in my ainol novo 7 (paladin)
<jgay> GC860
<techn> mnemoc: Ok. Next time then :)
<mnemoc> techn: give me one here
<mnemoc> techn: text
<hno> jgay, thats one of the MIPS based ones, right?
<jgay> hno, yeah
<jgay> the processor is made by ingenic
<libv> bomb :)
Turl has quit [Changing host]
Turl has joined #arm-netbook
<jgay> same processor that is used in the Ben NanoNote
<mnemoc> libv: ow
<jgay> o wait, i lied
<jgay> just checking now ... i thought it was but it is a differnt one
<mnemoc> libv: i wanted to merge the stage tomorrow :)
<libv> now you can merge with unloadable disp modules
<jgay> the ben nanonote uses the JZ4720 the novo7 paladin uses JZ4770
<libv> all these codepaths only get used during unloading
<hno> jgay, same processor series, but the ben nanonote one is much older and do not have GPU.
Scepterr has quit [Changing host]
Scepterr has joined #arm-netbook
<WarheadsSE> mnemoc: off goes that
<libv> there's going to be quite a few more patches tonight
<libv> now that i am not attaching turbo switches to tablets, i have to do something with my evening
<mnemoc> ok, i'll forget about 3.0.50
<WarheadsSE> hrm
<mnemoc> :)
<mnemoc> libv: keep them coming :)
<WarheadsSE> heh, steev freescales response to me looking for X11/video stuff on the iMX6 .. links to 1 day trainings! hah!
<mnemoc> free? :)
<WarheadsSE> sadaly, $99 for the day, but comes with lunch and a coupon for a sabrelight.
<mnemoc> :)
* WarheadsSE glares at Nitrogen6x on desk
<Turl> how much worth of a coupon? a 99$ lunch is pretty expensive ;)
<WarheadsSE> 50% off coupon
<mnemoc> doesn't look bad
<mnemoc> where is it?
<Turl> how much does the board cost?
<WarheadsSE> this one? Rockville MD
<WarheadsSE> $199
<WarheadsSE> so
<WarheadsSE> in the end, your paying for "training", lunch, and a half a sabre
<Turl> so basically
<Turl> if you were to buy a board
<WarheadsSE> that would all come free
<Turl> you can either buy the board, or buy the board and get a free lunch and training
<Turl> yeah :P
Almamuetya10 has quit [Ping timeout: 260 seconds]
<hno> A bit far to travel for a free lunch.
ddanila has joined #arm-netbook
<WarheadsSE> hrmm
<WarheadsSE> /linux/drivers/gpu/mali/mali/Kbuild:59: "CONFIG_TRACEPOINTS required for USING_PROFILING"
<WarheadsSE> and a whole bunch of gcc: error: directory": No such file or directory
<Turl> did you download using git or a zip?
* Turl recalls an issue with people downloading zips and missing symlinks
<mnemoc> yes, mali has a symlink inside, and zip doesn't support that
gsilvis has quit [Read error: Connection reset by peer]
gsilvis has joined #arm-netbook
* mnemoc wonders what happened to cat_x301
<jelly-home> maybe the packaging tool didn't use InfoZIP's -y / --symlinks For UNIX and VMS (V8.3 and later), store symbolic links as such in the zip archive, instead of compressing and storing the file referred to
<WarheadsSE> Turl: the tarball from github
<WarheadsSE> its a tar.gz
<Turl> WarheadsSE: no idea then
<libv> WarheadsSE: git clone then
<WarheadsSE> yeah
<WarheadsSE> thinking I may have to
<steev> wha
<mnemoc> invite him to join us :)
<techn> libv: I have that tablet :)
<techn> a13_mid :)
<libv> techn: ah, but you did not bother to expose serial yet i guess
<techn> libv: no :p
pawel5870 has joined #arm-netbook
<hno> mnemoc, is that sufficient? (lichee-dev)
<mnemoc> hno: machine type registers and proper atags mem info? yes, should be. I can test
<mnemoc> oh, ticket closed :)
<mnemoc> hno: will test and let you know
<hno> Should probably merge sun5i support as well. Not much.
<hno> but not tonight.
ddanila has quit [Ping timeout: 245 seconds]
<hno> techn, just do it.
<hno> techn, and please also post hires pictores of the board on wiki.
gimli has joined #arm-netbook
<mnemoc> Turl: have you noticed mali rebuilds itself every single time you make uImage or modules? :(
<mnemoc> Turl: any chance you can take a look on that?
<andoma> mnemoc: yeahs, that's anoying :/
<mnemoc> andoma: even more when you do it for every board we have a .fex, every time you add a change to the tree :<
<libv> rz2k: yikes, that seems like some painful soldering on the q88
<libv> techn: same
<libv> device as rz2k has? http://linux-sunxi.org/Q88 ?
<techn> mnemoc: that rebuilding problem is propably becouse ump and mali components uses same binaries.. and builds them twice
<techn> *binaries => objects
<libv> ooh, disp_version.c is really quite a broken pos.
<libv> fixes will come soon.
<libv> what idiot wrote this crap.
<mnemoc> techn: but it's not just twice, it's always
jgay has quit [Quit: Ex-Chat]
<mnemoc> libv: you? :)
<libv> mnemoc: never!
<libv> must've been my crap code writing evil twin
<mnemoc> :)
<mnemoc> fresh new set of hwpacks out of the oven. stage/sunxi-3.0 + latest u-boot-sunxi http://dl.linux-sunxi.org/amery/sunxi-3.0/latest/
gsilvis has quit [Read error: Connection reset by peer]
gsilvis has joined #arm-netbook
<hno> mnemoc, I can reproduce the reboot problem on my cubieboard.
<mnemoc> hno: but without jtag it sounds kind of impossible to fix :<
<hno> well, console works.
<hno> and there really isn't much besides the PMU that isn't reset on a reboot.
<mnemoc> i see
<libv> Turl: after this latest patchset disp_version.c is actually a useful guide on what to do
<libv> Turl: for the brightness thing, where /dev/disp seems to be opened/closed all the time, you should add the versioning handshake and ignore the return value if it is -1
<libv> this way, the brightness code will not spam the logs when an older kernel is used
<libv> hrm... htitp://
<libv> new protocol for ... special websites?
<libv> or maybe i didn't press ESC in time in vi
alcides has joined #arm-netbook
alcides has joined #arm-netbook
pawel5870 has quit [Ping timeout: 268 seconds]
von_fritz has quit [Quit: vonfritz leaves, don't panic]
<libv> there seems to be no place in the .fex where the available connectors is listed...
benjamin__ has joined #arm-netbook
<libv> namely: mele: lcd, hdmi, vga, cvbs. hyundai a7hd: lcd, hdmi, olinuxino: vga, lcd
<jelly-home> so mele wins if you want to make a STB with SCART for poor europeans
<libv> most of the other entries are quite useless though, except for the lcd panel timing
<libv> jelly-home: anything with a vga connector really, you just need to have the syncs embedded in R/G/B instead of pushed out over the separate lines
<jelly-home> oh, didn't know that (and, I guess, not all cards can push out pixels as slowly as PAL wants them)
<libv> usually they can, but not all cards support interlacing these days
<mnemoc> libv: btw, anything to say about adrien tetar's last patch to fix copyright notes with capital C and killing poor danling?
<libv> just typing up the connector email
<mnemoc> :)
slash_random has joined #arm-netbook
drachensun has joined #arm-netbook
<Turl> mnemoc: how big is cubieboard? /me wonders if it would fit inside wrt54g
<mnemoc> 10x6
<libv> mnemoc: when did he send it?
<mnemoc> today at 12:14
<libv> i have a pair of empty emails from him
<libv> ah, before that
<libv> big C is wrong anyway for (c)
<libv> as it is supposed to mimic ©
<mnemoc> if the rest is fine I can fix that detail with sed
<libv> let him do that
<libv> i think he likes being kept busy :p
<mnemoc> :)
<libv> it's amazing that he did not spot the copy paste <tab>->spaces issue in disp_version.c
<mnemoc> and `git am` doesn't trigger pre-commit so my checkpatch hook was ignored :|
<libv> well, seems that this pedanticness is quite selective
<libv> in actual good news, the lcd module will vanish soon.
<WarheadsSE> mnemoc: staging you said, yes?
<mnemoc> WarheadsSE: stage/sunxi-3.0 includes only fixes and cleanup, and unless someone finds something weird beofre tomorrow night, it will get merged and taged
<WarheadsSE> K, so I will just go ahead and use sunxi-3.0
<WarheadsSE> then rebuild tomorrow
<mnemoc> :)
<libv> i will not be around tomorrow evening in case something turns up, and over the weekend i will only have access to the a7hd
<mnemoc> i'm pretty confident nothing wrong will be found and i'll be able to merge, tag and upgrade to 3.0.50 during this weekend
<libv> ok :)
<mnemoc> i wish i had more time... i've not been able to touch the pinctrl drive in like a month :<
<mnemoc> meh
gimli has quit [Quit: Verlassend]
<mnemoc> but my brain is on strike anyway... so I better go to sleep
<libv> ok, goodnight :)
<mnemoc> s/drive/driver/
<mnemoc> good night