hno changed the topic of #linux-sunxi to: /Allwinner/sunxi development discussion - Don't ask to ask. Just ask and wait! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi
arokux2 has quit [Remote host closed the connection]
BJfreeman has quit [Quit: had a good time]
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
robb83 has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
robb83 has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
robb83 has quit [Client Quit]
<Legitsu> arokux no
<Legitsu> usb hubs are 5v
<Legitsu> unlessy ou are talking about a differnt kind of hub
<Legitsu> ifcase then ignore me
<Legitsu> acceptable range for usb devices is 5.0v to 5.2v
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
iamfrankenstein has joined #linux-sunxi
iamfrankenstein has quit [Excess Flood]
<robertjw> Is the MK802 IV quad-core with Rockchip RK3188 supported?
<Legitsu> I know the 802 III works
<Legitsu> I see no reasion that the IV should't
<Turl> supported by what?
<robertjw> by fedora19/sunxi
<Turl> you just said it has a rockchip SoC, so no
<Turl> it only supports sunxi (Allwinner)
<robertjw> I just didn't know if there were any issues with the RK3188 processor
<robertjw> ok, then I won't order one
<Legitsu> avoid rikomagic and tronsmart
<robertjw> I have a Mele a2000 (512MB) and sunxi/f19 loaded up on it fine except I've got a problem with no audio feed to the hdmi
<Legitsu> if you want a fuctional stick that runs android get the CX919
<robertjw> I'll check that out
<robertjw> I'm not really interested in android except as a "boot loader" to bring up a full linux.
<Legitsu> never gonna happen
<Legitsu> not with rockchip
<Legitsu> theres hybrid kernels that will run insert flavor of disto here
<robertjw> The Mele A2000G looks interesting
<Legitsu> but on the pure non-android side of things good luck
<Legitsu> no hardware accleration for the gpu makes it pretty well gimpd
<robertjw> are you talking about the mele?
<Legitsu> rockchip in general
<robertjw> then I'll x rockchip off my search list
<Legitsu> video decoding is slightly better on rockchip
<Legitsu> but thats under android
<robertjw> like I mentioned, I have a Mele A2000 w/ 512MB memory. I'm happy with it except no audio.
<Legitsu> I have a MK908 that hasa Rk3188 theres a ton of kernel bugs to be worked out
<Legitsu> but its fairly quick for that it is
<robertjw> running android on your 908?
<Legitsu> tbh freescale cpus are where you wanna look
<Legitsu> yea
<Legitsu> got Cyanogen 10.1 running
<Legitsu> its livable
<Legitsu> some slight issues playing 1080p MKV
<Legitsu> but its "acceptable"
<Legitsu> V3 sticks are garbage tho the V1.0s are better
\\Mr_C\\ has quit []
<Legitsu> they aren't as quick as the rockchip hardware in terms of RAW power
<Legitsu> but they get the job done
<robertjw> that looks interesting
<robertjw> "supports running Linux" not too many details in that statement
<robertjw> I see that fedora-arm is commiting to support rockchip hardware directly in F20.
\\Mr_C\\ has joined #linux-sunxi
<Legitsu> huh
<Legitsu> how they gonna manage that with no pure linux kernel sources
<robertjw> well they have until November
* Legitsu giggles
<Legitsu> unless they dont support hardware accerlation
<Legitsu> either that or they will use a hybrid-droid kernel
<robertjw> lack of hardware accelleration would make it pretty uninteresting
<Legitsu> as far as I know currently there is no way outside of android
<Legitsu> to get the mali400 gpu up and running
<Legitsu> which is a shame
<Legitsu> the mali400mp is a nice little graphics engine
<Legitsu> granted its not a PowerVR533
<Legitsu> but it gets along
<robertjw> no way because of restricted specs?
<Legitsu> dun know
<Legitsu> libv was making some progress
<Legitsu> I dont know exactly what he has working or how fast it is
* Legitsu can barely add if defines to existing code you should take anything he says with a grain of salt
<Legitsu> if Libv keeps after it freedora might have something by nov
<Legitsu> tho I seriously doubt any changes will make the release at this late stage
<robertjw> you're probably right, they already released F20-Alpha-TC4 yesterday
hipboi has joined #linux-sunxi
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
rz2k has joined #linux-sunxi
* Legitsu doesn;t much care as he uses windows as his primary os anyway
atiti has joined #linux-sunxi
wingrime has joined #linux-sunxi
Legitsu has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
stulluk has joined #linux-sunxi
hipboi has quit [Ping timeout: 260 seconds]
hipboi has joined #linux-sunxi
stulluk has quit [Ping timeout: 264 seconds]
hipboi has quit [Quit: Leaving]
robb83 has joined #linux-sunxi
geecko has joined #linux-sunxi
robb83 has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
pitillo has quit [Changing host]
pitillo has joined #linux-sunxi
<hno> arokux, 3.3V is enough for powering a hub chip, but not for powering the hub ports vusb pin. But it is ńot the hub chip that powers vusb (it only controls it).
arokux2 has joined #linux-sunxi
<arokux2> hi hno, thanks
<hno> Fedora do not support or ship proprietary blobs on any hardware, with the exception for certain firmware.
<arokux2> I tried a different approach to port usb host, but it doesn't work again
<arokux2> I've isolated the code in 3.4, so that it does not use any of the AW infrastructure, then just copied it to mainline
<arokux2> but it works only partially
<hno> mnemoc, sun7i defconfig is a bit of a mess..
<hno> i2c_sunxi built as module by default which means no AXP control unless i2c_sunxi.ko is loaded.
<hno> the defconfigs need a cleanup in general, too much differences. But the sun7i is just too different from the others.
stulluk has joined #linux-sunxi
<wingrime> hno: you can try take sun4i's
<wingrime> hno: and change soc to sun7i
<wingrime> hno: and try build
<arokux2> cat /proc/ccmu in case you haven't seen it
<wingrime> arokux2: there is phy part
<wingrime> arokux2: in usv
<wingrime> *usb
<wingrime> arokux2: have you touched it?
<arokux2> as I said. I've isolated the code in 3.4
<arokux2> it won't use any infrastructure, but mess up with bits
<arokux2> it works in sunxi-3.4
<arokux2> but does not in mainline. something is done externally
<arokux2> maybe PLL6 should be enabled.
<arokux2> but next I need to set up tftp
<arokux2> wingrime, what are you working on?
<arokux2> hno, you should be an expert in this: will tftp work with:
<arokux2> U-Boot 2012.10-04280-g1a8ac55 (Aug 04 2013 - 16:05:02) Allwinner Technology
<arokux2> it seems not to.
<arokux2> also, is the newest u-boot compatible with sunxi-3.4?
<arokux2> I think no, because of script.bin, right?
fredy has quit [Excess Flood]
<wingrime> arokux2: mostly cedarx
stulluk has quit [Ping timeout: 260 seconds]
<wingrime> arokux2: about mainline , I have IR driver that still not sended to mainline
<arokux2> wingrime, why? :p
<wingrime> arokux2: also I will try make sata work
<wingrime> arokux2: clocks not in kernel
<arokux2> everybody is so crazy about GPU, everything else could be done much easier
fredy has joined #linux-sunxi
<arokux2> anybody with working tftp setup? dhcp command hangs for me in u-boot shell, restart of the board needed..
<arokux2> now it does not hang... after 10th retry.
pacopad has joined #linux-sunxi
akaizen has quit [Remote host closed the connection]
<arokux2> now I get "software interrupt" and hangs agin
<leviathanch_> Turl: it's doing something! yeah! :-D
* leviathanch_ has made the mmc controller do something in linux-next :-D
popolon has joined #linux-sunxi
geecko has quit [Quit: Quitte]
<arokux2> leviathanch_, cool!
<leviathanch_> arokux2: not long anymore and we will have mmc on upstream
<leviathanch_> arokux2: as soon as I've cleaned it up and moved it from plsdsif
<leviathanch_> "Please don't shout in Finish"
<leviathanch_> ;-)
<arokux2> leviathanch_, not to discourage you but I thought the same several weeks ago with usb
<arokux2> :(
<arokux2> hno, can I flash our u-boot onto the board? my mmc seems not to be stable anymore :(
<arokux2> leviathanch_, btw, do you use tftp? :)
<leviathanch_> arokux2: I'm using the most recent u-boot-sunxi
<leviathanch_> and I'm loading u-boot from SD card
<hno> arokux2, you can drop our u-boot.bin in nanda, but it can't read from nand.
<leviathanch_> arokux2: I'm using SD card
<leviathanch_> the rootfs is within the initrd
<arokux2> leviathanch_, I recommend you setup tftp... after 100+ plugs my mmc seems to be unstable..
<arokux2> hno, no need to read from nand, would it talk to tftp?
<hno> yes
<hno> if you tell it to. Not automatically.
<hno> you can build one with a default environment that will automatically tftp by modifying include/configs/sunxi-common.h
<arokux2> hno, that is clear. where are the instructions to put u-boot.bin onto nand?
<arokux2> and hno, can you fix this "env set fdt_high ffffffff"? :)
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
<arokux2> hno, also, any idea why u-boot shell freezes after some time? :(((
<hno> no
notmart has joined #linux-sunxi
notmart has quit [Changing host]
notmart has joined #linux-sunxi
akaizen has joined #linux-sunxi
akaizen has quit [Ping timeout: 240 seconds]
notmart has quit [Quit: notmart terminated!]
<jukivili> arokux2: did you have chance to test musb yet?
<arokux2> jukivili, not yet, sorry. yesterday I had another try of porting of the usb-host, but again it didn't worked :(
<arokux2> jukivili, I did some measurements with multimeter and now my board seems not to be stable any more :(
rellla has joined #linux-sunxi
<arokux2> jukivili, what are you doing atm?
<jukivili> reading libgcrypt source
<arokux2> :)
<arokux2> hno, is there one u-boot that handles script.bin and dtb?
mouchon has quit [Ping timeout: 260 seconds]
atiti has quit [Ping timeout: 264 seconds]
massi has joined #linux-sunxi
mouchon has joined #linux-sunxi
<libv> fedora and their claims of software freedom...
<libv> redhat was far too keen, back in the day, to attack radeonhd and to side with ATI to help make sure that fglrx could continue as is
massi has quit [Remote host closed the connection]
<mnemoc> hno: hramrach has tried a bunch of time to make them all consistent.... but never got an ack... now we have yet another mess on the list
<arokux2> hi, mnemoc
arokux2 has quit [Remote host closed the connection]
arokux2 has joined #linux-sunxi
ZetaNeta has joined #linux-sunxi
ZetaNeta has joined #linux-sunxi
<arokux2> jukivili, around? booted kernel with our config
<arokux2> jukivili, usb still wont show up
<arokux2> I've measured the power, only 3.3V it is normal?
<arokux2> jukivili, the power of the usb host ports is 5V
<jelly-home> s/power/voltage/
akaizen has joined #linux-sunxi
<arokux2> jelly-home, thanks
<arokux2> jukivili, still = stick
deasy has joined #linux-sunxi
akaizen has quit [Ping timeout: 240 seconds]
<jukivili> arokux2: did the allwinner otg-host driver work?
<arokux2> jukivili, you'd need to tell me how to check it
<steev> bah, there i was thinking that the cubieboard2 left sweden back on 08-23 since it says Posted through Sweden Post, but apparently it didn't leave til 09-05 :/
<arokux2> steev, still waiting for it? where did you order?
<steev> arokux2: from r0ckstore
<steev> why it took like 2 weeks to leave sweden, i'll never know. such is international shipping
<arokux2> jukivili, could you plz stay around so we can test it quickly? :p
<jukivili> arokux2: test with sun4i_defconfig, it has aw otg (in host-only mode) enabled
<arokux2> jukivili, hm.. need to enable it in fex too.....
<arokux2> jukivili, do you respect it?
<jukivili> yes, sunxi-musb reads the same fex config as aw driver does
<arokux2> ?
<jukivili> yes.. the mele1000 fex does not have usb_id_gpio or usb_det_vbus_gpio, but that should be ok since usb_drv_vbus_gpio should be enough for host-only
<arokux2> jukivili, what about usb_used etc?
<jukivili> try with: usb_used = 1, usb_port_type = 1, usb_detect_type = 0
<arokux2> jukivili, and no pin assignment?
<arokux2> jukivili, (could you plz stay around?)
<arokux2> jukivili, usb_host_init_state ?
<jukivili> usb_id_gpio and usb_det_vbus_gpio empty, and usb_drv_vbus_gpio = port:PB09<1><0><default><0>
<jukivili> usb_host_init_state is not used by sunxi-musb
<jukivili> hm.. but is by aw driver..
<jukivili> set usb_host_init_state=1
<arokux2> ERR: script_parser_fetch usb_controller_type failed
<arokux2> jukivili, ^
hramrach has quit [Remote host closed the connection]
<jukivili> typo in fex?
hramrach has joined #linux-sunxi
<arokux2> jukivili, fex doesn't have it at all
rz2k has quit []
<jukivili> interesting
<jukivili> try just copy pasting [usbc0] from cubie config
atiti has joined #linux-sunxi
<jukivili> arokux2: sorry but I have to go now
<arokux2> jukivili, ok
<arokux2> ping me later
<oliv3r> mripard: http://lxr.free-electrons.com/source/drivers/net/phy/mdio-sun4i.c?a=arm#L162 here you remove the mdio stuff, but you don't disable the regulator. while 3 lines above line 154 you do in case of something breaks. is that intended? Why leave the regulator on when you 'stop' the device?
<oliv3r> mripard: and i just stumbled uppon the fact that the interrupt from the MDIO isn't handled by the emac driver, is this a dedicated connection these two have? I don't see anything in the interrupt table
rellla has quit [Ping timeout: 260 seconds]
<oliv3r> i would assume so oviously
<mripard> oliv3r: hmmm, it's odd, but it works. the regulator is disabled anyway when the driver exits.
<mripard> so, the regulator_disable is here useless.
<oliv3r> mripard: why is the regulator disabled anyway?
<mripard> hmmmm... because?
<oliv3r> mripard: because of devm_regulator_get?
<mripard> it's the point of using devm_regulator_get :)
<oliv3r> since devm things get cleaned up when you exit the driver?
<oliv3r> ok, that makes sense :)
<oliv3r> tns
<oliv3r> tnx
<mripard> and I don't understand your MDIO question. Why would the EMAC driver handle the MDIO irq?
<oliv3r> it says there PHY interrupts not supported :p
<mripard> hmmm, yeah, ok. :)
<oliv3r> what i saw from the schematics, the PHY is connected to portA (emac)
<mripard> that must be a stefan comment :)
<arokux2> oliv3r, is your otg working in host mode?
<oliv3r> so if the PHY generates an interrupt, would that be done via the regular interrupt controller, or does the emac do something internally with that?
<oliv3r> arokux2: no clue, haven't used it in ages
<oliv3r> mripard: if not, then what's the name of the PHY interrupt int he list of interrupts :p
<mripard> maybe it's on a gpio?
<oliv3r> i'll check the schematic and pinout of the phy again
<oliv3r> i'm guessing one of the MDII pins is responsible
<oliv3r> so the emac controller (the IP) gets to do it
<oliv3r> but that's just a strong uninformed guess :)
tinti has quit [Quit: Leaving]
<mripard> honestly, I don't really know the EMAC stuff, I mostly carried stefan's patches
<mripard> and all my attempts to refactor it have been a huge failure :)
<oliv3r> hmm, of course, MDIO is just 1 pin :) so sun4i-mdio just handles that one pin i guess; the rest is emac connection
<oliv3r> yeah it looked a little messy in places
<arokux2> has anybody otg in host mode working?
<oliv3r> mripard: i don't even find the word interrupt or IRQ in the datasheet
<arokux2> jukivili, I haven't got otg in host mode with aw driver too (using cubieboard config)
<mripard> arokux2: btw, you pinged me?
<arokux2> mripard, yes, but I've solved the problem. yesterday I had another attempt to port usb hosts, but no luck, works partially only
<mripard> ok :)
<mripard> so the patch you have on your github don't work?
<arokux2> mripard, they work better :)
<arokux2> mripard, but also do not work.
<arokux2> mripard, if they were, I would shout louder :)
<oliv3r> arokux2: have you tried identifying the used IP to see if there maybe exists a USB driver for it allready?
<oliv3r> arokux2: also, you working on OTG or the regular USB driver?
<oliv3r> arokux2: since i find it strange OTG is proven so troublesome, with the driver allready in mainline
<arokux2> oliv3r, on regular. this is standard ehci/ohci host, so only bus glue needed
<arokux2> but still, haven't managed to get it working
<arokux2> but I'm close..
<rah> mripard: have you any idea what the phoenixcard software does?
<arokux2> oliv3r, maybe you could help me? :p
<rah> mripard: I have an a1000g like Bamvor Zhang, who I understand has been in contact with you
<rah> mripard: it apparently needs to be flashed with an SD card created using the PhoenixCard software
<wingrime> oliv3r: I think USB have PHY part that need configre
<rah> mripard: I've tried creating a card with a buildroot image and an A31 kernel but it doesn't load u-boot
<rah> mripard: have you any idea what the PhoenixCard software does?
<oliv3r> arokux2: heh, sorry, gotta go ina few mintues; but maybe a good approach was tot ry and see if we can find similarities with other USB host controllers
<arokux2> oliv3r, as I said, only bus glue is needed.
<arokux2> oliv3r, there is no driver what so ever.
popolon has quit [Quit: Quitte]
<oliv3r> arokux2: so it uses the default platform usb host driver/framework
<oliv3r> arokux2: in that case, that should be doable :)
<oliv3r> arokux2: do you have anything published on git?
<arokux2> oliv3r, YES!
<arokux2> oliv3r, check: [RFC] USB EHCI support for mainline
<arokux2> on ML
<arokux2> guys, is there u-boot that can manage both, script.bin and dtb?
<arokux2> so that only one SD card can be used for mainline and sunxi-3.4?
jemk has joined #linux-sunxi
<arokux2> hno, ^
<oliv3r> arokux2: of course
<arokux2> oliv3r, which one?
<oliv3r> arokux2: u-boot simply loads data, even the dtb version loads the script.bin by default if it can
<oliv3r> arokux2: any new version
<oliv3r> arokux2: you can even load both files/tables if you wan
<arokux2> oliv3r, hm.. but how do the aw code knows what is where?
<mripard> rah: I built up my own by hands (and I need to document it somewhere), so I never used the phoenixcard stuff
<oliv3r> arokux2: fatload mmc 0 0x43000000 script.bin; fatload kernel; bootm 0x${kernel} - 0x43000000
<oliv3r> arokux2: or fatload mmc 0 0x43000000 cubie.dtb; fatload kernel; bootm 0x${kernel} - 0x43000000
<oliv3r> arokux2: you load either/or to the same address i think
<arokux2> oliv3r, that is awesome. will try it out!!
hramrach has quit [Remote host closed the connection]
vinifr has joined #linux-sunxi
hramrach has joined #linux-sunxi
<hno> arokux2, allwinner code simply looks for scirpt.bin at 0x43000000
<arokux2> thanks hno
<arokux2> hno, are you hacking on smth?
<hno> arokux2, not today. but hacking a lot on another arm based board lately at work, some 5000 of them.. and everything needs to be in place in about 2 weeks..
<arokux2> hno, what SoC?
<leviathanch_> Turl: your clock implementation is lacking the phase shift options
<leviathanch_> Turl: without these the sd card only delivers garbage
<leviathanch_> since it's out of clock
<arokux2> oliv3r, still there? your instruction won't work. http://sprunge.us/RDGJ
<hno> arokux2, Ti Sitara this time. Was hoping to get A10s but was to tight shedule at wrong time to make it.
<arokux2> hno, maybe you can comment on the issue? newest u-boot seems to expect dt. can I feed it with script.bin somehow too?
<hno> I haven't changed u-boot to expect dt.
<hno> why do you think it does?
<hno> that's not correct booting of an allwinner kernel.
<hno> just "bootm 0x46000000"
<arokux2> hno, you are right. checking your wiki now, thanks. oliv3r ^
<arokux2> hno, last question, from the safety point of view, how would you reboot a board if kernel crashed?
<arokux2> I quite ofte get: ### ERROR ### Please RESET the board ###
<hno> arokux2, when do you get that?
<hno> doing what on what board?
<hno> using what u-boot?
<hno> as for rebooting on kernel crash use panic=10
<hno> at kernel kommand line.
<hno> or enable watchdog.
<arokux2> 1. if I press power button and then release it, 2. if i unplug/plug the power. I've noticed that if I wait some time, I won't get this error. I use newest U-Boot
<hno> is the error from SPL or from main u-boot? Anything said before?
<hno> and any difference if you unplug the console RX wire?
<arokux2> hno, http://sprunge.us/GRKW
<arokux2> hno, how could i see the output without RX wire :)
<hno> Ok, DRAM failed to initialize. What board are you using?
<hno> RX is the pin you send data to the board on.. (TX on your UART dongle)
<arokux2> Mele
<hno> A1000?
<arokux2> yep
\\Mr_C\\ has quit []
<hno> arokux2, that is not a new u-boot, that one is some snapshot in june some time before v2013.07-sunxi
<arokux2> hno, should I upgrade?
<hno> arokux2, if you want help debugging some issue then yes.
<arokux2> hno, of course. compiling...
<arokux2> hno, you should update your wiki, make 'boardtype'_config needs CROSS_COMPILE too.
\\Mr_C\\ has joined #linux-sunxi
<arokux2> hno, even clean/distclean targets need it.
<hno> arokux2, feel free to edit. It's open for anyone with a github account.
<arokux2> hno, ok, will do.
<arokux2> hno, but another problem with new u-boot: http://sprunge.us/bSaF
<hno> arokux2, but I never speak about make 'boardtype'_config, only make 'boardtype'
<hno> which does everything in one go.
<arokux2> hno, and btw, same with new U-Boot: http://sprunge.us/Xbah
<arokux2> but maybe it is hardware...
<arokux2> hno, what to do with? http://sprunge.us/UZaV
<vinifr> the address to start the mainline kernel has changed for sunxi?
<arokux2> vinifr, no, why?
<vinifr> the kernel does not boot
<arokux2> vinifr, which branch?
<arokux2> vinifr, works for me. changed anything in uboot? forgot fdt_high?
hipboi has joined #linux-sunxi
<vinifr> so maybe the problem is my driver :)
<arokux2> vinifr, does it boot without it? :)
<arokux2> hno, why would spl need to touch uImage?
<vinifr> strangely also not
<rah> mripard: dude
<rah> mripard: woah!
<rah> mripard: :-)
<rah> mripard: such documentation would be a life saver for me
<rah> mripard: at present, this a1000g is the only A31 device I have
<rah> mripard: with a serial console
<rah> mripard: I'm trying my best to get anything up and running but the only luck I seem to have is bad luck :-)
<rah> mripard: I just need to get a kernel to load on it and I'll be fine
<rah> mripard: getting to that point is turning out to be a real nightmare
<rah> mripard: I beg and emplore you to write down how it can be done :-)
<arokux2> vinifr, want me to test it?
<vinifr> yes, I would appreciate :)
stulluk has joined #linux-sunxi
<hno> arokux2, what do you mean?
<arokux2> hno, http://sprunge.us/IfBB
<hno> arokux2, see install instructions.
<hno> this is NOT the same error, just an incorrect install.
<hno> well, the first one is the same..
<hno> arete74, was that DRAM failure with RX pin disconnected or connected=
_enrico_ has joined #linux-sunxi
<hno> sorry meant arokux2 ^
<arokux2> hno, I followed the instructions
<arokux2> hno, why it says something about uImage? it should start u-boot
<arokux2> hno, that is what I did: http://sprunge.us/YXTA
<arokux2> hno, and yes, "DRAM: ? ? 0 MiB"
<arokux2> happens even if TX is disconnected.
hramrach has quit [Ping timeout: 240 seconds]
hramrach has joined #linux-sunxi
massi has joined #linux-sunxi
wingrime_ has joined #linux-sunxi
<arokux2> hno, fyi: initialization of ram fails with the supplied u-boot on nand too (without TX pin connected).
<arokux2> hno, but what to do with: spl: not an uImage at 1600?
_enrico_ has quit [Read error: Operation timed out]
_enrico_ has joined #linux-sunxi
wingrime_ has quit [Ping timeout: 240 seconds]
<arokux2> vinifr, any luck booting the kernel?
<vinifr> I'll try again in a few minutes
<vinifr> I am studying English now :)
<arokux2> hno, it seems this is the problem: Boards with (SPL) include SPL loader to boot directly from SD-card.
<arokux2> hno, but I do not want to boot directly from SD card I want normal u-boot shell
stulluk has quit [Ping timeout: 264 seconds]
akaizen has joined #linux-sunxi
akaizen has quit [Ping timeout: 260 seconds]
xDR1TeK has joined #linux-sunxi
<xDR1TeK> hi all
<xDR1TeK> quick question
<xDR1TeK> when i unpack a tablet firmware
<xDR1TeK> i get bunch of files
<xDR1TeK> are they supposed to constitute the ramdisk and recovery images?
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
<xDR1TeK> those have names like COMMON _SYS_CONFIG100000
<xDR1TeK> which are long name and small file size
<arokux2> xDR1TeK, _SYS_CONFIG100000 could be script.bin?
<xDR1TeK> i was trying to get the ramdisk and recovery from the firmware
<xDR1TeK> i found a way to dig it from my tablet
<xDR1TeK> but never from an img file
<xDR1TeK> img -> firmware i meant
<xDR1TeK> not the unpacked img files like the ones that have bootloader, boot and recovery
<xDR1TeK> how can i make Q88.img firmware blow up to be "base, commandline, pagesize, ramdisk and zImage"
<arokux2> hno, you have left me without working u-boot :(
<xDR1TeK> well i need to get the images from the firmware without flashing the firmware and using Uberizer to dump rom
<xDR1TeK> anyone plz
<xDR1TeK> help
<xDR1TeK> guys\
<xDR1TeK> some feedback plz
<arokux2> xDR1TeK, i suggest you post e-mail to the ml
<xDR1TeK> ml?
<xDR1TeK> short for?
<arokux2> mailing list
<xDR1TeK> that is far
<xDR1TeK> kk
<vinifr> what do you mean? kk?
<arokux2> vinifr, strange. worked for me.
<xDR1TeK> vinifr that is a double affirmative
<xDR1TeK> ok
<xDR1TeK> kk
<vinifr> yes, strange. My source power is 5V@2A
<xDR1TeK> but now i just can't wait need answers fast
<arokux2> vinifr, if it worked try to figure out what is different now
<arokux2> xDR1TeK, otherwise?
<xDR1TeK> long story
<xDR1TeK> i am tired
<xDR1TeK> and sick of this embedded mess
<xDR1TeK> someone must know how to dump the rom from the firmware without flashing the firmware first
<xDR1TeK> perhaps one could use a VM for that sort then?
<vinifr> arokux, could you to post your boot log?
<xDR1TeK> suppose i dont have the tablet in reach and need to edit a file in boot.img
<xDR1TeK> and then need to build the firmware
hipboi has quit [Quit: Leaving]
<xDR1TeK> or that is not doable
<arokux2> vinifr, no, my u-boot is brocken waiting for hno to respond :(((
<vinifr> :0
<arokux2> vinifr, or maybe you can help.... do you run newest u-boot? it tries to boot uImage directly from SD but I want to have a shell..
<arokux2> vinifr, but you have a shell, so you do not have a problem
<vinifr> let me see my version
<vinifr> VERSION = 2013 PATCHLEVEL = 07 SUBLEVEL =EXTRAVERSION = -rc1
<vinifr> i am be able to boot sunxi uImage from SD
<arokux2> vinifr, where have you got it?
<arokux2> shit
<vinifr> it boot linux-sunxi, but does not boot linux-sunxi-mainline
<vinifr> i mean, for me
<arokux2> there is no linux-sunxi, there is either sunxi-3.4 or mainline
<xDR1TeK> this might be my solution
<xDR1TeK> but i have to leave without finding a way to get around it
<xDR1TeK> later all
xDR1TeK has quit [Quit: SleEeEeEp]
<vinifr> arokux, yes :)
<arokux2> vinifr, will post log soon
<arokux2> vinifr, booted! :)
<vinifr> great
jemk has quit [Read error: Operation timed out]
massi has quit [Quit: Sto andando via]
<vinifr> arokux, where you live?
<arokux2> vinifr, germany
massi has joined #linux-sunxi
<arokux2> vinifr,
<arokux2> I do not understand, however, how could it help you
<arokux2> vinifr, my config is also special
<vinifr> before 'Starting kernel ...'?
<arokux2> vinifr, http://sprunge.us/RHNP
<arokux2> vinifr, if you describe how to test your driver i can test
<vinifr> nice, boot from ethernet
<arokux2> but write a nice doc on ml
<arokux2> vinifr, yes, I will update a wiki soon
<arokux2> the* wiki
<vinifr> I hope at least the kernel starts...
<arokux2> vinifr, but fatload is also correct
<arokux2> vinifr, as said: figure out what you have changed
<arokux2> vinifr, did you pull?
<vinifr> then should appear something like /sys/bus/iio/devices/iio:deviceX
<arokux2> vinifr, write e-mail plz, msgs here will get lost
<arokux2> vinifr, just as reply to your patch
<vinifr> it must be tested,but at least the kernel compile :)
<arokux2> vinifr, why are you keep sending messages here? just send an e-mail to ml
<vinifr> I already sent the patches, I'm waiting for some feedback
<arokux2> vinifr, perfect then. I'll download them from ml.
<vinifr> but i sent only for linux-sunxi, for internal review
_enrico_ has quit [Read error: Connection reset by peer]
_enrico_ has joined #linux-sunxi
popolon has joined #linux-sunxi
dwilkins has quit [Ping timeout: 246 seconds]
dwilkins has joined #linux-sunxi
dwilkins has quit [Excess Flood]
dwilkins has joined #linux-sunxi
<hno> arokux2, right.. there is an old unmaintained wiki copy in my personal repo which someone else have edited. Killed.
<arokux2> hno, :(
<arokux2> hno, kill it so nobody can find it anymore :)
<arokux2> hno, btw, setting console= causes logs not to be output to uart.
<hno> arokux, wiki no longer possible to open in my personal repo.
<hno> onlu linux-sunix/u-boot-sunxi/wiki remains.
<arokux2> hno, I was just kidding :)
<arokux2> maybe it could be useful for you guys: http://linux-sunxi.org/Mainline_Kernel_Howto#Booting_kernels_quickly
<arokux2> vinifr, ^
<vinifr> arokux, ok, i got :)
<arokux2> hno, question. what would be the best way to have network boot but with modules? 1. the could be build into uImage, 2. could be copied to initramfs. first option requires manual editing of the configs, second one is automatic, but takes time. anything better than those two?
<vinifr> there a /sys/bus/iio/devices/iio:device0
<vinifr> there is..
<hno> arokux2, I append them to my initrd.
<hno> or initramfs rather
<arokux2> hno, so option #2
<hno> arokux, yes. Not sure how you would add them to the uImage.. built-in drivers are not modules
<Turl> leviathanch_: what are they? :P
<Turl> arokux2: use nfs rootfs
<arokux2> hno, General Setup / Initramfs source file(s) with a path to your initramfs image
<arokux2> Turl, will it be faster? :)
<Turl> arokux2: well, you'd just need to install them to your would-be root
<hno> hard to use nfs rootfs with sunxi emac as a module.
<Turl> emac can be builtin :)
<hno> but yes, nfs root is a lot more flexible than using initramfs for testing.
<hno> all drivers except mali can be builtin.
<hno> which makes them built in drivers, not modules.
<arokux2> yep, the point is often you need to test configs with modules and you need those modules....
<arokux2> so either you make them built-in or you append them to initramfs or... nfs, thanks Turl
<hno> arokux2, yes. My miniroot setup handles modules nicely.
<hno> by appending them to the initramfs.
<arokux2> hno, miniroot needs to be archieved each time. i'm curios what would be faster archieving or nfs.
<hno> nfs is faster if you have a good network. loading the initramfs over tftp is slow.
<hno> and nfs root is ready as soon as you have run "make modules_install"
<arokux2> hno, well, it is local network through a router, should be fast, I think
<hno> arokux2, correction.. nfs is always faster. If network is not good then TFTP performancs plummets completely. If network is good then tftp is only slow.
<hno> but nfs root setup requires a little more preparation on the server.
dwilkins has quit [Ping timeout: 245 seconds]
<arokux2> Turl, can you post your buildroot config?
<vinifr> i wanna add ssh to .cpio
<arokux2> wow.. who has written such extensive doc for buildroot
<Turl> arokux2: sure, sec
<Turl> arokux2: http://sprunge.us/jVjf
<arokux2> Turl, you did it yourself?
<Turl> arokux2: yeah, it's not that hard
<Turl> arokux2: just choose cortex a8, arm eabi, select a suitable compiler
<Turl> the rest is depending on what you need
<arokux2> Turl, ok, thanks!
<hno> uhm... really should change my gravatar I think. That's not me any longer..
<Turl> hno: old picture?
<hno> Turl, a bit.. 2004 I think.
<rah> I keep getting a segfault when I run bin2fex
<rah> does the program work for anyone here?
<hno> rah, yes.
<rah> :-/
<Turl> rah: run it under valgrind --tool=memcheck
<hno> rah, is your bin2fex up to date?
<rah> hno: is it intellgent to detect when it's not actually looking at a compiled .fex?
<arokux2> hah, there is cubieboard def config in buildroot
<rah> s/intelligent/intelligent enough/
<hno> rah, not really.
<rah> :-/
<Turl> arokux2: yes, it selects lots of sunxi stuff, like sunxi-mali
<rah> I'm trying to demystify the contents of LiveSuit images
<rah> without much luck
<arokux2> Turl, I do not need this junk :)
<hno> rah, if it's crashing on an actual script.bin then first of all make sure you have an up to date copy.
<hno> if still crashing then we need your script.bin..
<hno> or you need to debug why it's crashing.
<rah> hno: it's crashing out files dumped using awimage
<rah> s/out/on/
<rah> jesus I can't type today
<hno> LiveSuit images are not very mysterious, if one ignores the embedded Windows DLLs which Livesuit happily executes with no verification or even telling you..
fredy has quit [Excess Flood]
<rah> hno: what are update_boot0, etc?
<hno> That's exactly what it says. A program that updates boot0.
<rah> they just seem to be binaries in the allwinner-pack-tools
<rah> hno: updates it how?
<hno> writing a new copy to the boot area.
<rah> I don't understand
<hno> which part?
<rah> there's a script called "pack" which compiles images for LiveSuit to flash
fredy has joined #linux-sunxi
<rah> it calls a Linux binary called update_boot0
<hno> ah, that one. I thought you talked about the update_boot0 that is inside the image.
<rah> you're saying that binary updates boot0 by "writing a new copy to the boot area"
<rah> what is it writing a copy of?
<rah> what is "the boot area"?
<rah> oh
<rah> no, the update_boot0 program
<rah> there's update_boot1 as well
<rah> and something called "dragon" which is the final step
<rah> all of which are Linux binaries with no apparent source
<hno> It's all programs, but some are for use by pack when making the image, and some are for the sunxi and embedded in the image.
<rah> indeed
<rah> the mystery is in what is used to make the image, and how, and what gets embedded in the image, and how it's used on the sunxi
<rah> so, basically everything -)
<rah> :-)
<rah> to be honest I don't really understand how you can say that the LiveSuit images aren't very mysterious
<rah> they seem to be completely opaque
<arokux2> Turl, have you used external or internal toolchain?
<arokux2> Turl, trying with external now...
<rah> hno: is the pack script just cargo-culting what's done by build scripts from manufacturers that people happen to have got hold of, or does anybody actually understand this stuff?
<Turl> arokux2: external codesourcery
<rah> s/anybody/someone/
<arokux2> Turl, why cs btw? linaro isn't so good?
<arokux2> Turl, I always wanted to know :)
<Turl> arokux2: buildroot didn't like linaro for some reason
<arokux2> Turl, :( you see, nothing could be easy.... (
<arokux2> Turl, i'm using linaro now, will let you know.
<arokux2> we need sunxi_defconfig for buildroot :)
mnemoc has quit [Quit: bye]
mnemoc has joined #linux-sunxi
<hno> rah, the pack tools is allwinners idea of how one builds a flashable image. Not much rational thinking.
<hno> but works for livesuit..
<hno> sometimes.
<arokux2> Turl, worked for me with: gcc-linaro-arm-linux-gnueabihf-4.7-2013.04-20130415_linux
<hno> rah, update_boot0 and update_boot1 in pack is updating the file header of boot0 and boot1 respectively with info collected from your fex file. These early stages of the boot process do not have access to script.bin.
<arokux2> vinifr, this interface is much nicer: http://news.gmane.org/gmane.comp.hardware.netbook.arm.sunxi
<arokux2> vinifr, but your message isnot there yet.
<rah> hno: how is it that you know this?
<rah> hno: is it documented somewhere?
<arokux2> does someone know how often gman syncs?
* arokux2 loves documentation
<Turl> arokux2: does it boot? :)
<hno> rah, no idea if it's documented outside allwinner. just happens to know a bit too much about what allwinner socs does when they boot.
<arokux2> Turl, not tried yet, so it did not boot??
<Turl> arokux2: I got SIGILLs on init, but ymmv
<rah> hno: "just happen to know"?
<hno> rah, yes.
<rah> hno: *how*? :-)
<rah> hno: did you decompile update_boot0 or something?
<rah> os disassemble it?
<rah> s/os/or/
<rah> hno: or did someone at allwinner tell you?
<rah> hno: or do you have some docs you shouldn't have?
<rah> hno: how is it you know this?
<hno> no I have not decompiled update_boot0. Not needed. It's function is obvious.
<rah> it is definitely not obvious to me
<rah> I don't understand how to get to a viewpoint where it is obvious
<rah> hno: that's why I'm asking you how you got to a point where it's obvious
<hno> rah, see sunxi-tools/bootinfo.c for some details on the file headers that needs to be updated in boot0 & boot1.
<hno> to write u-boot SPL that works I had to understand what boot0 & boot1 does.
<hno> that said I have never actually ran the pack tools or even livesuit. Never seen a need to.
<rah> that's like "How to draw an owl: 1. First draw two circles. 2. Then draw the rest of the fucking owl."
<rah> hno: I need to understand what boot0 & boot1 do
<rah> hno: how did you understand what boot0 & boot1 do?
<hno> boot0 configures SDRAM, boot1 configures the rest and loads u-boot.
<rah> I know roughly what they do
<rah> that doesn't help me debug why my machine isn't loading u-boot
<hno> there is full sources available for the A20 version if you want to study them in detail. Part of the A20 SDK releases.
<rah> where?
<rah> are there sources for the binaries used in the pack script?
<hno> No there is no source for the pack tools.
<hno> and not for the other binaries that get embedded in the image either.
<rah> :-/
<rah> but there's source for boot0 and boot1?
<rah> (for some version of boot0 and boot1?)
<hno> yes, but no one have suceeded in building them from those sources yet. And only for A20.
<rah> are there other sources from which boot0 and boot1 *have* been built?
<hno> I am quite optimistic that the sources that is there actually match the shipped binaries reasonably well, but don't know, or much interest in them. My focus is u-boot and it's SPL where we don't need any of this.
<hno> other than as a reference on what boot0 & boot1 does to initialize the board.
<hno> no interst in pack, dragon, livesuit, phoenix, etc etc. They are all proprietary history.
<rah> I don't understand
<hno> what part?
<rah> well, first, what you mean by "SPL" :-)
<rah> but also, how can you avoid boot0 and boot1 and its friends?
<rah> are you replacing them with u-boot?
<hno> The first state bootloader that is part of u-boot. Does the same as boot0+boot1 more or less, and then loads the full u-boot.
<hno> first stage.
<hno> U-Boot SPL is built from the same sources as u-boot.
<arokux2> rah, :) what is your hw?
<vinifr> arokux, what is your software for tftp?
* hno uses tftp-server http://www.kernel.org/pub/software/network/tftp/, but any tftp server does fine.
<arokux2> vinifr, I'm on Arch Linux, recommenden :)
<hno> even the tftp server of dnsmasq works fine.
* hno have tried
<vinifr> i use Debian7
<soletti> where i can find more info about xbmc development status for a20? or theres none?
<arokux2> vinifr, Debian is very good too, but arch wiki is very extensive
jemk has joined #linux-sunxi
dwilkins has joined #linux-sunxi
<hno> Aha.. it's actually tftp-hpa that I am using.
<rah> hno: that sounds great and when I can get to the point of being able to load code on to my machine, I'll happily but this first stage bootloader on it
<rah> hno: unfortunately, at the moment I *can't* load code on to my machine and I need to figure out why
<hno> rah, what machine do you have?
<rah> arokux2: I have a Mele A1000G with an A31
<arokux2> Turl, you once told me this should go into dts? ---> LOADADDR=0x40008000
<hno> Ok. can't help you out very much tere I am afraid. A31 the only Allwinner CPU I don't have and which is also missing u-boot SPL support.
<Turl> arokux2: eh? dt? no
<Turl> arokux2: that should be on your env if you want make uImage to work
<arokux2> Turl, I thought it could be buried somewhere.... :(
<rah> hno: what I don't understand is: why aren't there notes on all this?
<rah> hno: I mean you could at least add a link to that olimex download to the wiki
<rah> the source code for boot0/boot1 is pretty important
<Turl> arokux2: http://sprunge.us/ZQSI
<hno> rah, those boot0/boot1 sources do not work on A31.
<Turl> I put that on ~/bin/ as "kernel-envsetup", then "source kernel-envsetup" before I start working
akaizen has joined #linux-sunxi
<arokux2> Turl, yep, I have similar, but I use "." instead of "source"
<arokux2> :)
<hno> A31 SDK also includes boot0/boot1 sources, but not aware of any public release of A31 SDK yet.
<hno> but have to admit that I have not been looking for it either.
<Turl> arokux2: yeah, but source looks in $PATH so it's easier to type I guess
vinifr has quit [Quit: Saindo]
<hno> rah, linux-sunxi is about promoting open alternatives. Have no interest in promoting proprietary and GPL violating stuff.
<hno> those that perfer the proprietary Android SDK generally know how to find it.
<hno> and don't care much about open.
akaizen has quit [Ping timeout: 260 seconds]
vinifr has joined #linux-sunxi
* hno apologies to certain Android people around here who do in fact care about open.
<rah> hno: I appreciate that but restricting the availablity of information about the proprietary parts prevents work on free alternatives when that work depends on the information
<rah> I need to understand boot0 and boot1 to get free stuff on my machine
* arokux2 thinks Andoid is very open comparing to iOS, though
<arokux2> rah, do you hack just for fun?
<rah> without links to a proprietary SDK, my machine is stuck with its stock Android image
<rah> arokux2: what do you mean?
<arokux2> rah, do you work on sunxi for commercial interest (just curious)
<rah> arokux2: I do, yes
<arokux2> Turl, buildroot booted fine (built with linaro)
<arokux2> rah, how could one monetize such activity, if it is not a secret, except of building hw
<rah> arokux2: but I'm not a suit
<rah> arokux2: I also run Manchester Free Software
<rah> arokux2: there are many business models based on software and hardware whose users are free
<hno> rah, if you have a commercial relation with Allwinner or someone close to them then just ask them to send you the SDK. Else look around a little an I am pretty sure you'll find the A31 SDK somewhere. In there is boot0 & boot1 sources.
<rah> arokux2: as far as I'm concerned the issue is completely orthogonal to trying to get u-boot working on my A31
<arokux2> rah, ah, so "trying to get u-boot running on A31" is just for fun? that was my original question
<hno> rah, if your board do not succeed in loading u-boot in the stock android image then HW is likely broken.
<hno> especially if you have already tried to reflash it with livesuit/phoenixsuit.
stulluk has joined #linux-sunxi
<rah> hno: it loads it in the stock android
<hno> rah, if your problem is only how to get to the u-boot prompt then there is many ways doing that. Usually sufficient to just trash the env pertition.
<rah> hno: it doesn't load it in the buildroot image that I created and flashed via phoenixcard
<rah> it doesn't do anything
<rah> it just sits there
<hno> was it packed with A31 version of pack tools?
<rah> yes
<hno> and a31 version of kernel etc?
<rah> yes
<hno> where did you find these?
<rah> from a vendor who made them available
<rah> (but not a full Allwinner SDK)
<hno> Do you get any output at all?
<hno> if completely silent then I only see two possible causes
<rah> I do get output
<rah> I'll put it up, just a sec
<hno> That boot entered USB device mode.
<hno> I think.. A31 boot is a little different.
<rah> I don't know what that is
<rah> what is "USB device mode"?
<hno> is your boot.ini correct?
<rah> I don't know, I don't know what boot.ini is :-)
<rah> I told you, the LiveSuit images are completely opaque to me
<rah> it's pretty much the case that if it's not documented on the wiki, I don't know about it
<Turl> boot.ini is a file on nanda from what I recall
<Turl> boot0/1's configuration
<rah> there is a file in pack/out/bootfs called boot.ini
<rah> I presume that's it
<oliv3r> back!
<rah> hno: is that boot.ini correct?
<hno> looks reasonable.
<hno> and is your script.bin correct? This boot1 output seems to think there is a battery
<ZetaNeta> Yo, guys. Anyone aware of Logicom TAB 750?
<ZetaNeta> I just got 1
<ZetaNeta> And wanted to get linux there
<rah> hno: where is the script.bin within a LiveSuit image?
<Turl> ZetaNeta: looks line an allwinner A13 tab from a little of googling
<hno> rah, is that script.bin what is in the android image, or just a random A31 EVB one?
<arokux2> ZetaNeta, feel free to read our wiki! :)
<rah> hno: I have the extracted parts of the stock android image so I can compare if I can get the script.bin from it
<ZetaNeta> Turl, SOUNDS GREAT! When i got it, i didnt even hope to ever get linux there
<rah> hno: that is from the vendor's A31 pack tools
<ZetaNeta> And just came here to ask, becouse of my lucky experience with a A10 Mpman MID74C
<vinifr> hno, would be difficult to adapt mmc sunxi driver to mainline?
<hno> rah, your boot log seems to indicate it thnks there is a battery and goes into emergency suspend.
<rah> hno: ok, so it sounds like I need a more realistic script.bin
<hno> rah, you really need to use a script.bin that matches your board. You can not use a random one for another board.
<rah> hno: how can I get a script.bin from an image that runs on the board?
<Turl> ZetaNeta: first step would be identifying if the tab is the same as another one already supported on uboot, or getting the data needed to build uboot otherwise
<arokux2> vinifr, leviathanch_ is working on it
<hno> rah, if you have adb then it's easy to extract from nanda partition. Or just power of, hold down 1 on the console and power on.
<vinifr> great
<hno> with USB OTG cable connected.
<rah> hno: I have no OTG port on the machine
<rah> I have a serial port but no USB OTG
<hno> are you sure? The A1000 did at least.. but on a custom connector.
<rah> no I'm not sure
<ZetaNeta> Turl, I been thinking of either opening it up, and seeing if thats the A1* board i already know, or to "find / | grep sunxi"
<rah> hno: trying to locate the OTG is a future possibility
<rah> hno: but right now I have none
<hno> but you don't need it if you have adb or other Android root shell access.
<rah> hno: however, I do have a LiveSuit image
<rah> hno: can I extract it from a LiveSuit image?
<hno> also you should have script.bin if you have unpacked that livesuit image.
<Turl> ZetaNeta: you could also get the timings as instructed on the wiki and compare after the fact
<rah> hno: where?
<hno> rah, in the folder whenre you unpacked the image..
<hno> it's also easy to extract it from a phoenixcard image.
<rah> hno: where in the folder?
<rah> hno: there is no script.bin file
<vinifr> is someone doing sunxi keypad driver?
<rah> $ ls | head -n 3
<rah> 12345678_1234567890BOOT_0
<rah> 12345678_1234567890BOOT_0.hdr
<rah> 12345678_1234567890BOOT_1
<rah> these are the kinds of files I have
<hno> rah, the a1x images I have looked briefly at have had script.bin files. But if there is none then you need to extract the boot fat image.
<Turl> vinifr: for mainline?
stulluk has quit [Read error: Operation timed out]
<hno> rah, a full listing please.
<hno> image file names is... different from normal files.
<rah> indeed
<hno> rah, it's the SYS_CONFIG ones. Both should be identical.
<rah> hno: you remember we originally started this conversation with me asking about a segfault in bin2fex?
<rah> hno: it was the SYS_CONFIG files that I was running it against
<Turl> vinifr: wiki says wigyori is working on IIO LRADC
<hno> rah, so Allwinner is doing something we are not used to in their A31 script.bin.
<wingrime> Turl: his lradc is adc driver class,
<rah> luckily I'm fucking good
<wingrime> Turl: so we need more one driver that convert values to keys
<rah> codepad.org/0HqQc5ij
<rah> that's from a script.bin contained in a FAT filesystem in "RFSFAT16_BOOT_FEX00000000"
<vinifr> Turl, I had spoken to him. he agreed to use the driver that had started
<vinifr> but I have not spoken to him lately
<Turl> wingrime: no generic one?
<hno> rah, then try using that one.
<rah> hno: I will
<rah> thank you
akaizen has joined #linux-sunxi
<rah> [pm_para]
<rah> -standby_mode = 1
<rah> +standby_mode = 0
<rah> I find that most curious
<rah> "standby_mode = 1" is the working script.bin
Black_Horseman has joined #linux-sunxi
<wingrime> Turl: looks yes
<hno> rah, I do not think boot1 uses that. But not sure.
<hno> most script.bin parameters is for the kernel.
<wingrime> hno: aw's melis should use same script values
<wingrime> hno: but script have parameters for drivers that we have no
<rah> hno: ok
<vinifr> Turl, are you working in clock stuff?
<Turl> vinifr: yes
<vinifr> I can not determine if there is source clock to lradc
<vinifr> then remained pending source clock and button support to IIO LRADC
<Turl> vinifr: "keypad clock" maybe?
<vinifr> do you mean 'keypad clock' for LRADC?
<Turl> vinifr: yes
<vinifr> Turl, I suspect that there is no clock to LRADC, because your max freq. is 250Hz
<Turl> vinifr: lradc on linux-sunxi is 'sun4i-keypad.c' right?
massi has quit [Quit: Sto andando via]
<vinifr> I do not think, where is it the source?
<vinifr> LRADC != Keypad Interface
<Turl> vinifr: drivers/input/keyboard/
<Turl> hm ok
<Turl> is there a LRADC driver on linux-sunxi?
<Turl> sun4i-keyboard.c on that same directory?
<vinifr> no, there was not
<Turl> sun4i-keyboard.c mentions lradc
<Turl> and uses lradc registers
<vinifr> ops. you are right :)
<vinifr> Turl, button support can be leveraged
<arokux2> libv, you see, again politics
<Turl> vinifr: does not look like it uses any clock that I can see
fredy has quit [Excess Flood]
<Turl> vinifr: magic hardware :)
<vinifr> Turl, yes, as I thought. because max freq. is 250Hz
fredy has joined #linux-sunxi
<arokux2> Thu Jan 1 01:24:44 UTC 1970
<arokux2> on my Mele box :)
<ZetaNeta> hi
<ZetaNeta> i got following files: boot.axf boot.ini drv_de.drv font24.sft font32.sft linux magic.bin os_show script.bin script0.bin sprite sprite.axf vendor
<ZetaNeta> from a new and rare tablet
<ZetaNeta> And surely unlisted
<ZetaNeta> (Logicom TAB750)
<ZetaNeta> does anyone needs those files from NANDA
<libv> arokux2: just tapping up a rather massive blog entry about it
<libv> arokux2: first paragraph contains a direct quote from my previous post: "Open source software is about power, politics, corporate affiliation, and loads and loads of noise."
<arokux2> libv, what isn't? everything is politics.
<Turl> ZetaNeta: can you run a10-meminfo on it?
<ZetaNeta> And... what is it?
<Turl> ZetaNeta: a program that dumps the memory controller registers and determines the timings, frequency, etc
<ZetaNeta> Run it on tablet? Ok, where to get it?
<Turl> ZetaNeta: see http://linux-sunxi.org/Submitting_Boards, "bootinfo from tool"
<Turl> there's a static binary so you can run it on android just fine
<vinifr> Turl, omg, KEY_BASSADDRESS(0xf1c22800), but datasheet says 0x01c22800
<Turl> vinifr: yes, virtual vs physical
<Turl> vinifr: allwinner has static mapping on their kernel, 0xf* virtual -> 0x0* physical
<ZetaNeta> Turl, Ok, done it
<ZetaNeta> Anything else?
<vinifr> hum, ok
<Turl> ZetaNeta: you can convert script*.bin to text using bin2fex tool from sunxi-tools if you wish, but it's not a deal breaker
<ZetaNeta> Turl, Device got sure traces of either OEM either manual root
<ZetaNeta> But i am sure there is no chance previous owner could find a custom firmware for it
<ZetaNeta> As its a rare device sold in some specific area of france
<Turl> ZetaNeta: AW devices usually come "prerooted"
<ZetaNeta> Oh....
<ZetaNeta> Nice :D
<Turl> and insecure :p
<ZetaNeta> And with Busybox, and all partitions rw?
<arokux2> oliv3r is silent
<Turl> ZetaNeta: probably not rw, but you can remount them with no issues
<ZetaNeta> Turl, Well, when i saw, most fs were RW
<ZetaNeta> Almost all, which usualy doesnt
<ZetaNeta> Even on custom firmwares
<Turl> ZetaNeta: once you have the script.fex and a10-meminfo tool output, email both to the linux-sunxi list
<ZetaNeta> Turl, To who???
<Turl> ZetaNeta: data and cache are rw on all androids, system should be ro, and there's not many more fs
<ZetaNeta> Here is "mount"
<ZetaNeta> Subscription required?
<Turl> from what I recall, no, dunno why it says that
<Turl> I believe it's moderated though
<Turl> mnemoc: ping ^
<ZetaNeta> How do i put the attachment?
<Turl> as an attachment? I don't understand your question :)
<ZetaNeta> script.bin
<ZetaNeta> Also, there is also a script0.bin
<ZetaNeta> Any good?
<Turl> I'd attach both to be on the safe side
<Turl> if you can convert them to script.fex with bin2fex even better
<Turl> you can get bin2fex from sunxi-tools, https://github.com/linux-sunxi/sunxi-tools
<hno> script.bin and script0.bin should be identical.
* ZetaNeta gone downloading those tools
naobsd has quit [Ping timeout: 250 seconds]
<oliv3r> arokux2: back reading still :p
* ZetaNeta still thinks if he should write a "letter" to the guy who sold this tablet without even logging out of his gmail
<ZetaNeta> Maybe i should write it from his own address?
<ZetaNeta> Or change his avatar?
<oliv3r> oh i've back read
<oliv3r> script0.bin should be identical to script.bin
<ZetaNeta> Yo! Oliv3r
<ZetaNeta> remember me?
<oliv3r> and here i've been calling it script.orig; whoever i told that, it's script0.bin :)
<oliv3r> ZetaNeta: gentoo or slackware was it not
<ZetaNeta> oliv3r, ?
jemk has quit [Ping timeout: 268 seconds]
<ZetaNeta> Ohohoho! He didnt log out of his facebook
<Turl> "LIBV Intentionally Breaks Videodrivers"
<oliv3r> oh i gotta read that
<libv> Turl: that's only been there for about 5 years
<oliv3r> ZetaNeta: remind me :)
<Turl> libv: nice nevertheless
<libv> Turl: with my initials, which is my nickname, it was pretty natural to choose something like that :)
<libv> when i started that blog, i was mostly complaining about via
<libv> so it was libv incessantly bashes $vendor
<Turl> :)
<arokux2> libv, send a link with your new blog entry
<Turl> arokux2: it's on his livejournal
<eagles0513875> hey all
<libv> arokux2: when it's done
<ZetaNeta> oliv3r, I am the guy who was asking about Mpman MID74C
<oliv3r> ZetaNeta: oh that I do not remember ;)
<oliv3r> ZetaNeta: so the previous owner had no clue as what he was doing really
<oliv3r> why did he sell it
<ZetaNeta> oliv3r, I think i once shuted you up, becouse of too much "tips"
<ZetaNeta> And you were sad after that
<ZetaNeta> oliv3r, no idea
<ZetaNeta> Turl, I didnt find bin2fex
<Turl> ZetaNeta: did you compile the tools?
* ZetaNeta didnt
<ZetaNeta> :P
<Turl> get make'ing then
<ZetaNeta> Turl, Done
<ZetaNeta> Where to send now?
<Turl> ZetaNeta: to the mailing list I showed you earlier
* ZetaNeta doesnt want a link where its explained. /me wants a email to send it to
* ZetaNeta didnt understand a single thing in the link you gave
<Turl> linux-sunxi at googlegroups.com
<ZetaNeta> okai
<ZetaNeta> Just send it like that?
<Turl> yeah, also include some info about your tablet as reference
<ZetaNeta> ok
<Turl> eg brand/model, ram size, storage size, screen size, the normal stuff people can use to identify it
<oliv3r> foto's
<oliv3r> inside and out
<Turl> that's better suited for the wiki isn't it?
<oliv3r> yeah it is
<oliv3r> but if zetaneta is having trouble with the ml; the wiki might be really too much :)
<ZetaNeta> oliv3r, :D
<ZetaNeta> oliv3r, Yet, i can send them via netcat to you :D
<Turl> it's caturday, so suitable transfer method I'd say
<oliv3r> caturday! Purrr
<oliv3r> its sunday now
wingrime_ has joined #linux-sunxi
<ZetaNeta> so... what do i do now?
<ZetaNeta> I want linux on this tablet
<ZetaNeta> and its not in the list yet
atsampson has quit [Quit: everything upgrade]
<oliv3r> ZetaNeta: well step 1 is to compare it to the existing onse
<ZetaNeta> do i use the script.fex from my device?
<ZetaNeta> oliv3r, Tried
<ZetaNeta> No matches
<oliv3r> memory timings are the most important to find out which u-boot you need
<oliv3r> actually, that's really the only important thing
<oliv3r> a10-meminfo otuput
<ZetaNeta> there are only 3 a13 devs
<ZetaNeta> oliv3r, Got it
<oliv3r> did you paste your a10-meminfo somewhere?
<ZetaNeta> Nope
<ZetaNeta> Want it to pastebin?
<oliv3r> please
<ZetaNeta> i can give the fex
<oliv3r> fex not important :)
* ZetaNeta is totaly shocked!
<oliv3r> well it is, but not for this
<oliv3r> you can use any a13 u-boot
* ZetaNeta is now less shocked
<oliv3r> ok the EASIEST way, is to download hansg's fedora 19 build
<oliv3r> did you download that iamge and put it on a 3+ gigabyte SD card?
<ZetaNeta> meh?
arokux2 has quit [Remote host closed the connection]
<oliv3r> put that iamge on a SD card :)
<oliv3r> that's THE easiest way
<ZetaNeta> I dont want easiest way, if it leads me to "dirty, evil RPM distros"
<ZetaNeta> Either debian, either gentoo
<Turl> oliv3r: timings look similar to the A10S olinuxino micro
* ZetaNeta wants to open this tablet up, but cant see where
<oliv3r> ZetaNeta: don't worry about the inside
<oliv3r> ZetaNeta: and you start with something, then go with anything
<oliv3r> ZetaNeta: the fedora image is just to test and see :)
<Turl> oliv3r: except the zq, it's completely different
<oliv3r> and fedora is still linux
<Turl> ZetaNeta: you can install the fedora image, then replace the rootfs with a debootstrapped debian or whatever you like
<ZetaNeta> Well, can i just start straight from debian?
<oliv3r> Turl: that looks wrong for a zq
<oliv3r> dram_zq = 0x16b9487b
<ZetaNeta> so... how do i go?
<oliv3r> dd if=fedora.img of=/dev/sdcard
<oliv3r> Turl: those b's can't be there :)
<ZetaNeta> And what about "debian.img"?
<ZetaNeta> Or better "debianwheezy.img"
<oliv3r> Turl: erm i mean, nvm it's still to long
<oliv3r> ZetaNeta: the fedora image is known to work; so best to start off with as a first time; you can always change the rootfs later etc
<ZetaNeta> well, ok
<ZetaNeta> gonna try it
<Turl> ZetaNeta: go read the readme there for fedora 19, use * A10s-OLinuXino-MICRO (Olimex) as board
<oliv3r> ZetaNeta: even I used fedora first, and i'm a gentoo nerd
<Turl> onde you get that working, replace the rootfs with whatever you like :)
<Turl> once*
<ZetaNeta> oliv3r, Only RPM distro i tried is OpenSUSE, and i dont wanna put more dirt in the eyes
<oliv3r> hahaha
<oliv3r> ZetaNeta: learning, baby steps etc :)
<ZetaNeta> DEB works, Portage works, Even just compiling does (LFS for ex.). But not RPM
<ZetaNeta> :D
* ZetaNeta gone to write the SD
<oliv3r> what's so wrong with rpm?
<oliv3r> the only thing I find wrong with RPM, that it hasn't merged with deb to create a universal package format :)
<ZetaNeta> "Name".....
<Turl> oliv3r: there was a time where rpm dep management sucked hard
<ZetaNeta> oliv3r, Also true
<Turl> not sure if it's still the case
<ZetaNeta> Turl, Last time i checked, it was
<oliv3r> Turl: well my first linux i ever tried, was redhat 5.4 i think
<oliv3r> i lie, 5.2
<oliv3r> absolutly deffo 5.2
<ZetaNeta> My em...
<ZetaNeta> and what file should i choose?
<oliv3r> the f19 version
<oliv3r> the f18's are old :)
<Turl> f19
<Turl> ZetaNeta: there's a readme on there, go read it ;)
<ZetaNeta> oliv3r, I had a long story of using different distros.
<oliv3r> 5.2 was interesting, booted up on the terminal, I had a dialup that i had no clue how to use, and i was young and had no idea really what this thing was and how to use it
<ZetaNeta> Turl, Readme? Not hardcore!
<Turl> ZetaNeta: you need to run a script after you xzcat it, go read the readme :)
<ZetaNeta> can i unxz it and then dd? or xzcat does black magic?
<Turl> yeah you can too
<Turl> but that's two commands :)
<oliv3r> mind you, I was one of those rare people that stuck with ms-dos for ages, only ran win31.exe when i had to
<ZetaNeta> Do you want me to show you the "gentoo installing command"?
<ZetaNeta> I am the one who uses NetBSD on mac68k
<ZetaNeta> I am the one who is putting linux on a BT Freestyler 750 DECT phone
<oliv3r> even with win95 it booted into dos :p
<ZetaNeta> s/ler/le
<oliv3r> but when i tried linux for real; i had slackware 7.0
<ZetaNeta> PATRICFAG
<ZetaNeta> brb
<ZetaNeta> need to go
pacopad has quit [Quit: pacopad]
akaizen has quit [Remote host closed the connection]
wingrime_ has quit [Ping timeout: 268 seconds]
fredy has quit [Excess Flood]
[7] has quit [Ping timeout: 256 seconds]
akaizen has joined #linux-sunxi
TheSeven has joined #linux-sunxi
fredy has joined #linux-sunxi
wingrime has quit [Ping timeout: 245 seconds]
rz2k has joined #linux-sunxi
vinifr has quit [Remote host closed the connection]
naobsd has joined #linux-sunxi
_enrico has joined #linux-sunxi
_enrico_ has quit [Ping timeout: 264 seconds]