<ol_man>
from /proc interrupts i can see that there are IRQs constantly firing. do i need to disable them? how would i do that?
<Turl>
ol_man you have usb unplugged right?
<ol_man>
i compiled the kernel without usb. and yes, i unplugged it too
<Turl>
what uboot are you using?
<ol_man>
being unable to boot a kernel with the stock one i installed a "novo elf 4.0.3" rom which boots an android boot.img. there i put a 3.0.8 kernel (sunxi-lichee) and a debian initrd
<ol_man>
i don't know the exact version of uboot
<Turl>
that uboot probably has usb support
<Turl>
so maybe it leaves the controller in a way that it's waking the device or something
<Turl>
just a theory though
<ol_man>
with usb enabled in kernel it said that usb would go to suspend correctly. didn't crosscheck by /proc/interrupts
<ol_man>
there are interrupts e.g. from AX20 coming through i2c. must it be shut down?
<ol_man>
and there is also a "aw clock event irq"
<Turl>
no idea, axp20 is the power chip
<ol_man>
i thought i would need that during suspend
aholler_ has joined #arm-netbook
aholler has quit [Ping timeout: 248 seconds]
<ol_man>
turl can i mask selected interrupts so they don't wake the cpu?
ol_man_ has left #arm-netbook [#arm-netbook]
hipboi has joined #arm-netbook
<ol_man>
turl thankyou for your input, need to leave now
ol_man has left #arm-netbook [#arm-netbook]
<nemik>
are there any current livesuite images for linaro/ubuntu? i'm trying with scripts/mk_livesuit_img.sh and a linaro alip rootfs, not sure if this can work?
<br->
rm: what's the diff between mk802 and mk802 ii?
<rm>
mostly those things they have outlined in the description
<rm>
a power controller IC, a different antenna, and ESD protection on HDMI
<rm>
also it doesn't seem to have a DC adapter or socket
<rm>
needs to be powered via one of the MiniUSBs
<rm>
or are those micro
<rm>
then it's one more difference, the original one use mini
<rm>
used*
<br->
aha
Alex1269_ has joined #arm-netbook
rellla has joined #arm-netbook
<hansg>
Note I'
<hansg>
I've both and I'm not convinced the wifi antenna on the mk802ii is better, on the contrary. Same for the dc-plug which has been replaced with a micro-usb whose only purpose is powering the device
<hansg>
OTOH the mk802ii does seem to run cooler, likely because of it having the pmu
<hansg>
Both work fine for me.
<br->
so for under 400 euro, can buy enough car batteries + sd card + cheap tablet to fit entire text contents of english wikipedia and read it 6 hours a day every weekday for a full year
<hansg>
I'm personally a big fan of the mini-x, that tends to be somewhat more expensive, but does give you composite video + normal (non hdmi) audio out as well as an ir receiver + remote
<hansg>
And it is prettier :)
pcat has joined #arm-netbook
merbanan has quit [Read error: Operation timed out]
<jelly-home>
it would be ideal for an STB if it also had wired ethernet
<jelly-home>
er, they have images from the same PoV, it's just white
<hansg>
jelly-home, take a look at the first picture on which it is twice, once standing on its side, you can see the back-side there (although not very clear) and you can see the rj-45 there, also it having an rj-45 is listed in both item descriptions
<hansg>
jelly-home, if you want to be sure you can wait till I get mine :)
<jelly-home>
ooh
* hansg
just had his birthday and asked money from everyone, and now has a lot of a10 devices coming in the mail
<rm>
...and apparently Google needs to release a "Google PC" that would be like any of those boxes, except targeted for desktop keyboard+mouse usage
<rm>
before the chinese stop marketing them as only "TV boxes"
popolon has joined #arm-netbook
<hansg>
rm, yeah in me experience with cheap hardware (I've also written a lot of webcam drivers), having the some box / outside does not necessarily mean a device also having the same inside.
<hansg>
They will happily put whatever is cheapest that month inside a box they already have, since designing a box + the tools to create the housing is more expensive then a pcb re-design only
<hansg>
Luckily with these "tv" boxes they tend to advertise the soc used, with webcams no such luck, there I would just order a few and it was like the lottery. Meaning I've a ton of pac207 based cams, and my webcam collection does not cover all the sn9x102 variants I would like it to cover ...
gsilvis has quit [Read error: Connection reset by peer]
<hansg>
mnemoc, I assume you've seen my patches from the weekend? Note that the sunxi-hdmi patch fixes a somewhat serious regression which is present since the dynamic mode changing patches landed (basically running anything other then 1920x1080 breaks X)
<rm>
although the same tablet w/o phone/3g is $53, not $125
<hansg>
But only yesterday, so it will be a while before I'll receive it
<hansg>
I also have 2 different A10s based "tv-sticks" coming. Which interestingly enough, assuming the build-in wifi is usb connected, have more USB ports then the wiki says the A10s has ...
<hansg>
And these will be my first sun5i devices
<mnemoc>
hansg: I'll start merging things now
<hansg>
mnemoc, great, note no hurry on the fex / u-boot stuff. But it would be good to get the hdmi fix out there. Is the dynamic mode stuff still only in staging ? Otherwise you should probably also directly push the hdmi fix to stable
<andoma>
hansg: no the android install on it does not seem to be willing to talk adb at all over USB. Can't see any USB target when connecting it to my linux box
<andoma>
so i don't have a working fex file
<andoma>
nor can i find any UART pads on the PCB
<andoma>
i guess next step is to root exploit the android install
<andoma>
but as usual there's lack of time
<hansg>
andoma, you can get the fex file without adb, that is what I've done for all my devices:
<hansg>
1) install terminal app from marketplace
<hansg>
2) start terminal
<hansg>
su -
<hansg>
insert usb-stick into otg port
<mnemoc>
hansg: his mk802+ comes su-less
<hansg>
(you will need the real usb port for hub + keyboard + mouse, those tend to not work in otg because to many interrupt endpoints
<hansg>
su less, oh that is a bummer ...
<andoma>
yeah i didn't manage to su on it
<andoma>
i'm a total android-noob so i can't really tell how this "usually" are supposed to work on android
<rm>
might need to do "busybox su blah"
<hansg>
Anyways to finish my story, what I do is terminal then su, then use "cat /.../script.bin > /mnt/usbhost0/script.bin"
<rm>
or sudo
<rm>
on mine even cp did not exist/work
<hansg>
You need to use cat because there is no cp in my experience
<rm>
only "busybox cp"
<hansg>
rm, ah busybox cp is a good trick, I just used cat :)
<mnemoc>
don't tell it loud or allwinner will remove bb from their android sdk too
<andoma>
i started on writing some kind of NAND reader in FEL mode but realized after a while what a mess the AW NAND driver is
<hansg>
andoma, the pictures of your pcb are interesting the wireless module has much more pins then an usb wireless module, so it is likely spi, and indeed googling for rtl8189es shows that is an spi chip...
<hansg>
Getting that working with a custom kernel might be a challenge...
<andoma>
i kinda gave op on these mk802+ devices for now and ordered some mk802ii instead
<hansg>
andoma, try rm's busybox trick to su from a terminal, or try installing a terminal-app which allows you to start a terminal directly as root, I think some allow that (assuming your android is rooted and just missing su)
<hansg>
andoma, then hopefully you can get the script.bin ...
<mnemoc>
and a10meminfo!
<hansg>
Right
<andoma>
yeah, that's a start for sure and then hook up wired ethernet and i might be able to boot an ssh server at least
<mnemoc>
[dram_para] in recent script.bin is full of useless null entries
<hansg>
mnemoc, did you see my mail about dram para ? It seems most of them are very constant and we can predict emr1
<mnemoc>
have it stared, but haven't read yet
<hansg>
mnemoc, ofcourse a10meminfo is better, but I wonder how much thinking has gone into the settings we read with a10meminfo, judging from the script.bin quality not a whole lot.
<mnemoc>
adding some probing and thinking in u-boot spl's dram_init might not be that bad indeed
lkcl_ has joined #arm-netbook
<hansg>
(clearing my stack) "<mnemoc> hansg: I was thinking in merging the stage, then your fix, and tag" tag as in add to stable and tag there ? Not sure about that otoh I feel the dynamic mode stuff needs more testing otoh getting it into stable may be the best way to get it more testing
<hansg>
"<mnemoc> adding some probing and thinking in u-boot spl's dram_init might not be that bad indeed"
<hansg>
I'm not really talking about probing, I'm not sure that is possible, but opening up devices, looking at the pcb + dram chip numbers and deriving stuff from there is possible, not ideal, but possible
<mnemoc>
ah. ok
<hansg>
specifically it seems that the number of dram chips (2 or 4) determines the emr1 value
lkcl has quit [Ping timeout: 272 seconds]
jelly has quit [Remote host closed the connection]
<mnemoc>
dynamic mode switching is still stage-only
<e-ndy>
hansg, are you able to run selinux on fedora + a10?
<e-ndy>
hansg, with 3.4 kernel
<mnemoc>
-r1 pushed, sunxi stage braches rebased
sspiff has quit [Read error: Operation timed out]
marcan has quit [Read error: Operation timed out]
focus_well has quit [Read error: Operation timed out]
drachensun has quit [Read error: Operation timed out]
focus_well has joined #arm-netbook
drachensun has joined #arm-netbook
marcan has joined #arm-netbook
sspiff has joined #arm-netbook
sspiff has quit [Changing host]
sspiff has joined #arm-netbook
<hansg>
e-ndy, yes, although currently I'm using enforcing=0 on the cmdline to make sure selinux is not getting in the way, but I'm not seeing much avc-s. I hope to remove the enforcing=0 soon
<hansg>
e-ndy, actually I hope to release F-18 beta images ready to dd to sdcard for A10 devices before the end of the day, which should be using selinux
pcat has joined #arm-netbook
<e-ndy>
hansg, and what about tar.xz archive? i use ssd for rootfs and sd card only for uboot
<hansg>
Or you could loopback mount and then cp -a or rsync
<oliv3r>
mnemoc: you mean like an A31s or sumat
<hansg>
One problem with using anything but images is that selinux attributes will get lost.
<mnemoc>
oliv3r: the A31 is a tablet SoC
<mnemoc>
just like the A13
<oliv3r>
mnemoc: that's why it's sun6i; makes sense, but is it not even done yet?
<mnemoc>
hansg: fair point
<oliv3r>
they do frame it as something more powerfull; 'the quad'
<mnemoc>
oliv3r: they are selling A31 devices already. but not A20
<oliv3r>
I find a very appealing feature of the A20, that it is pin-compatiible
<mnemoc>
and mali400-mp2
<oliv3r>
oh, well I don't want A31, but are there links on the wiki
<mnemoc>
A20 is far more likely to be compatible with all our sources
<oliv3r>
well mali400 '-mp1' would be enough for me, but yes, lima > powervr :)
<andoma>
i want a dongle with a20. Then i'm happy
<oliv3r>
with it being pin-compatible, chances are huge yes
<oliv3r>
dongle a20, settop box with A20
<oliv3r>
swap my A10 from my tablet with an A20 >: ) :p
<ssvb>
mnemoc: I would not be so sure about A7 vs. A8
<ssvb>
mnemoc: A7 is going to be definitely a lot faster for floating point calculations and maybe for memory accesses (improved TLB, automatic data prefetcher, etc.)
<mnemoc>
ssvb: 1 vs 1 or 2 vs 1?
<ssvb>
mnemoc: 1 vs. 1
<oliv3r>
well the A20 does have VFPv3 AND VFPv4 (or v4 that is compatibel with v3) :p
<oliv3r>
so that should speed up things aswell
<mnemoc>
hard to tell until we get devices to benchmark :p
<ssvb>
mnemoc: based on the slides and the documentation I have seen, A7 is more like a reworked A8 with obvious bottlenecks corrected (FPU, memory subsystem) and also removed less useful functionality (somewhat worse dual issue capabilities)
<mnemoc>
ssvb: I was only considering the DMIPS/MHz thing. but yet, it could be faster
<ssvb>
mnemoc: btw, they also downgraded NEON in A9, and nobody complained
<oliv3r>
to bad the A20 isn't an A7 + A15 instead of 2x A7 :D
<mnemoc>
uh
<oliv3r>
anyway, we want devices :(
<oliv3r>
for manufacturers it should be so easy to make them, just swap SoC's
hipboi has left #arm-netbook ["Leaving"]
<oliv3r>
oh hipboi was here! damn :S well atleast he's back :)
<mnemoc>
allwinner doesn't seem to be interested in higher end. they seem quite confortable in fast design-to-market cheapest-manufacturing-possible segment
<jinzo>
oliv3r, I think that if A20 would be that easy to swap with A10, we would'we seen devices allready
pcat has quit [Read error: Connection reset by peer]
<oliv3r>
jinzo: it's pin compatible!
<jinzo>
So they say!
<jinzo>
doesen't mean it could couse problems
<mnemoc>
pin compatibility doesn't tell much about the sw side
<oliv3r>
well mabe some capacitors coul dneed change, but I doubt they changed anything/much on that end. THey really probably just swapped the arm IP, added some dualcore glue, same for the GPU, and that's probably all their work done. I wouldn't be supprised if existing bugs remain
<mnemoc>
i simply hope they don't decide to modify the irq and clock stuff
<mnemoc>
oliv3r: yes, that is my hope. A10 with newer ARM and mali
<mnemoc>
A20 been 2xA10
<mnemoc>
which opens the door for an A40, as A20 with quad-a7
<mnemoc>
but dreaming is free
<oliv3r>
lol that it is
<oliv3r>
would be nice to have an idea when an A20 device would become available
<mnemoc>
if we are lucky tom will be able to get us some sample cubieboard with replaced soc once they are available
<oliv3r>
i'll wait for tom then :)
<oliv3r>
and find someone who knows a little about templates to 'fix' the broken current one :(
<oliv3r>
so I can continue, not that I hav emuch free time
<mnemoc>
write your own
<mnemoc>
kidnapping mediawiki templates is a bad idea
pcat has joined #arm-netbook
<oliv3r>
I wouldn't even know where to start!
<mnemoc>
{|
<mnemoc>
:)
<mnemoc>
see the templates I wrote for PIO
<oliv3r>
those are templates? I saw you did table in table
<oliv3r>
what I tried to accomplish, was have one 'table' which you can pull data from in a few forms
Ershov has joined #arm-netbook
<oliv3r>
using that template you could more easily define registers but also have idfferent usages for them, and drawing graphs would be easy
<oliv3r>
oh the PIO table has changed a lot
leowt has joined #arm-netbook
<leowt>
hi m8s
<oliv3r>
mnemoc: mux I its most appearant what you did, right?
<oliv3r>
Port Bank I*
<mnemoc>
Port Bank I = PI
<oliv3r>
mmm pi
<mnemoc>
PA, PB, PC, .... PI
<oliv3r>
hmm, PA, PB didn't seem to use any assignments though? I may just finish the lists using only tables and leave the templating up to someone else :)
<oliv3r>
if only there was debug output, it would be a little easier to see what's going wrong
<mnemoc>
the only way I know for mediawiki templates is to write them iteratively
<mnemoc>
once grown, they are imposible to understand
ganbold_ has joined #arm-netbook
Ershov has left #arm-netbook [#arm-netbook]
Ershov has joined #arm-netbook
<oliv3r>
how fast ist he mmc controller in tthe A10? If i'd get 'the fastest card' would that be a waste?
rellla2 has joined #arm-netbook
rellla has quit [Ping timeout: 244 seconds]
rellla2 is now known as rellla
<oliv3r>
e.g. if I'd get an UHS-I card, will the a10 actually be able to use it 'fast'?
datagutt has joined #arm-netbook
<buZz>
no
<buZz>
a10 is slow
<mnemoc>
f* DX refunded me the uhost :(
<jinzo>
likewise
<mnemoc>
and they had the face to reply my ticket (where I asked them not to do so) telling me they had refunded it and a "anything else I can help you?"
<hramrach>
they said the will be rufunding them since they are 'defective'
<hramrach>
maybe they could not get them at that price in fact
<hramrach>
but whatever the case they don't have the sticks
<hramrach>
oliv3r: I don't get much apparent IO with a class 6 card
<hramrach>
but since the leds have to be lit in userspace until a block device trigger is available they may just note get lit when the CPU is busy
mentis has joined #arm-netbook
Jef91 has joined #arm-netbook
SouL_ has joined #arm-netbook
SouL_ has left #arm-netbook [#arm-netbook]
sgtpepper has quit [Quit: leaving]
leowt has quit [Quit: leowt]
opm_70 has joined #arm-netbook
hg_5 has joined #arm-netbook
ganbold_ has quit [Remote host closed the connection]
sudani has joined #arm-netbook
leowt has joined #arm-netbook
drachensun_ has joined #arm-netbook
drachensun_ has quit [Client Quit]
<drachensun>
I just found out something I think might be of interest
<drachensun>
that service.i-onik.de link
<drachensun>
actually contained some nand init code I couldn't find anywhere else
<drachensun>
that was vital for getting nand working on a newer A10 board design
<drachensun>
so it might have newer code than we have elsewhere, did anyone else take a look at that?
Jef91 has joined #arm-netbook
<mnemoc>
there are several source dumps with interesting stuff. unfortunatelly no one with time to properly submit the real changes/fixes into our tree
<mnemoc>
the most annoying part is that changes in the tree for one mach-sun?i tend to be destructive for the other
<hansg>
Maybe we should archive those dumps somewhere, and make a wiki page with links to said archive? That way if / when we may need info from there we have the trees available. Otherwise the trees may be gone before we get around to looking at them ...
<mnemoc>
I have them stored
<hansg>
Good!
<mnemoc>
wanted to make a unified lichee3-sunxi branch, stuck at 3.0.8, integrating only allwinner source dumps
<mnemoc>
but $work$ hasn't allowed
<hansg>
possible dump question, but what is this "lichee" thing I keep seeing every now and then ?
<hansg>
s/dump/dumb/
<ibot>
hansg meant: possible dumb question, but what is this "lichee" thing I keep seeing every now and then ?
<mnemoc>
lichee is the name of allwinner's SDK
<hansg>
Ah, ok
hg_5 has quit [Read error: Connection reset by peer]
<mnemoc>
the gpl-violating thing they give to manufacturers
<mnemoc>
fortunatelly they have such a mess with repos that they unintentionally leak parts of the code they intend to distribute bin-only
<drachensun>
lol
<hansg>
How is it gpl-violating, because they only give it to manufacturers, or .... ? All the allwinner code I've seen in the sunxi-3.4 tree seems to have a proper GPL header
<drachensun>
so is that source leak the latest?
<mnemoc>
critical parts are .o/.ko only
<mnemoc>
also, availability is subject to a gpl-violating NDA
<mnemoc>
drachensun: o-ink's 1.5 whatever tarballs seem to be pretty recent, yes
<hansg>
Hmm, but the accidentally leaked sources then do have GPL copyright headers on top ?
<mnemoc>
no. they GPLed the code we have in march (released by a dutch company)
<mnemoc>
after that they faked a company and changed all headers to be owned by that fake company instead of allwinner
<hansg>
mnemoc, hmm, so to be clear what we'v in the sunxi-3.0 and sunxi-3.4 branches is based on their GPL release from march + own work ?
<mnemoc>
hansg: yes
<hansg>
ok, thats good :)
<drachensun>
I had downloaded it before but hadn't gone through it
<drachensun>
I was searching for the nand init parameters and google actually found the xDE, x97, etc string in the file
<drachensun>
lucky for me
<drachensun>
I'll submit that nand table has a patch today
<drachensun>
once its down unzipping and I can scan through my local copy
<drachensun>
s/down/done/
<ibot>
drachensun meant: once its done unzipping and I can scan through my local copy
<specing>
Has anyone ever tried contacting the chinese authorities about the violations?
pcat has quit [Read error: Connection reset by peer]
<buZz>
lol
<buZz>
'hello this is FSF.cn yes?' 'yes i would like to report a GPL violation'
<buZz>
'hahahahahahahahaha' *click*
<jinzo>
buZz, they would probably hold the line - for the lulz ofcourse
<buZz>
:P
<mnemoc>
this can only end if ARM decides to get involved and talk their licensees into gpl compliance
<mnemoc>
neither inviduals, distrobutors, manufacturers or even copy right owners can in practice force them to do anything
<hramrach>
arm does not really need to bother
<mnemoc>
arm profits from it
<mnemoc>
like companies making stuff in asian factories without bothering to care if children work there or not
<hramrach>
they provide blobs and whatnot but they do not violate GPL
<hramrach>
unless they provide kernel code with a NDA
<hramrach>
but I suspect they provide only blobs and manuals
<mnemoc>
manuals? hahahaha
<hramrach>
what do they provide to the Chinese then?
<mnemoc>
./build.sh
<hramrach>
if they provided code under NDA they would be in GPL violation themselves
<hramrach>
which they likely want to avoid
<hramrach>
because then you could sue them
<specing>
Has anyone ever tried contacting ARM about this, then?
<mnemoc>
afaik the answer is "not our problem"
<hramrach>
if the Chinese receive code and only pass it on as blob and violate GPL
<hramrach>
you can sue them
<mnemoc>
good luck trying
<hramrach>
not ARMs problem
<mnemoc>
it's not ARMs problem. but only ARM can solve it
<hramrach>
of course, suing Chinese is pointless
<hramrach>
specing: suggests that to FSF
<hramrach>
I am sure they would if there was a case to make
<mnemoc>
the legal path has zero future unless there is intention to comply
<hramrach>
but after the Linksys case I would expect the western based companies would be more careful about licensing
<hramrach>
well, you can have them pay some random money posibly
<mnemoc>
there are so many hops involved it's imposible to do in a soc-lifetime
<hramrach>
which might make them slightly uncomfortable
<mnemoc>
device manufacturers don't get the gpl-violating SDK directly from allwinner, but from wits. factories only receive images
<hramrach>
given that you get userspace blob and kernel code I suspect that ARM made the SDK so that they have their ass covered
<hramrach>
so factories or users could ask for source
<hramrach>
but factories don't care, especially in China
<hramrach>
and users don't get anything from Chinese factory
<slapin>
evil people everywhere :( there is no place lazy developer can go :(
<mnemoc>
the legal path is useless
mentis has quit [Quit: Konversation terminated!]
<mnemoc>
slapin: get a gov. job :p
<slapin>
mnemoc: I'm not that lazy, I passed this stage already - low salary and nothing to do, boring... you get to find yourself work or go wash tables... ewwww...
<mnemoc>
:)
ccssnet has joined #arm-netbook
gimli has joined #arm-netbook
<slapin>
I just don't like when people find lots of excuses to do nothing, but call me lazy, when I don't do something for them.
<drachensun>
looks like the source included for the A10 has the community patches through 8/17/2012
<jinzo>
the ionik one?
<drachensun>
yup
hg_5 has joined #arm-netbook
<slapin>
drachensun: please don't use this date format, it makes me dizzy :(
<jinzo>
community patches = the linux-sunxi effort? or the mainline kernel?
<drachensun>
that standard where I live, how do you do it?
<jinzo>
day/month/year probably
<jinzo>
couse y'know, it kinda makes sense once you think about it :P
opm_70 has quit [Remote host closed the connection]
<mnemoc>
month/day/year is the only way to have 8/17/2012 as a valid date
<drachensun>
Henrik Nordstrom <henrik@henriknordstrom.net>Date: Fri Aug 17 22:47:25 2012 +0200
<mnemoc>
doh
<jinzo>
but I can't believe it, that the official allwinner SDK would sync up with the linux-sunxi effort
<mnemoc>
not sync, but maybe kidnap a fix
hansg has quit [Remote host closed the connection]
<jinzo>
still interesting
<mnemoc>
indeed
<jinzo>
I really hope someday someone from Allwinner would show up here
<specing>
hipboi did
<jinzo>
yeah, but he was not employed for that long
<mnemoc>
he came because *he* believed
<slapin>
mnemoc: this is why it makes me dizzy - my validator is failing me, and 17-th month made buffer overflow :(
<mnemoc>
y-m-d is best
<drachensun>
mnemoc: thanks
<slapin>
mnemoc: yeah, even without dashes.
<mnemoc>
drachensun: eh?
<mnemoc>
slapin: yes
<drachensun>
mnemoc: I had asked the preferred format and got no answer :-)
<slapin>
datagutt: anything which can be read straight without conversion should be fine, like Jan 20, 1980, or 2010-14-15 or 20101475 is good to parse, later is best for sorting, too.
<slapin>
x/x/x or c.c.c are often require additional knowledge to parse
<slapin>
damn patch is still in moderation queue, and I think it will be ill-reviewed again, as it is too big.
<slapin>
datagutt: sorry not for you, that was intended for drachensun, but I missed again
<mnemoc>
drachensun: :)
<specing>
slapin: I kinda doubt the validity of '2010-14-15'
<mnemoc>
*g*
<slapin>
specing: yeah, sure, you doubt it
<specing>
and '20101475'
<slapin>
specing: how many days in month you say?
<mnemoc>
not earth date ;-)
<specing>
in the silly system we use today, [28, 31]
<drachensun>
is that how you get so much done? you have more months than the rest of us?
<slapin>
sorry guys, I must shut up now to not say anything silly
<specing>
time representation is even worse than the whole american imperial measurement system thingie
<drachensun>
sorry, I couldn't resist
* slapin
hopes u-boot submission will not turn into troll-show again
<specing>
We don't even have new year on a SOLSTICE!!
<drachensun>
specing: yeah, some super villian should make a plot to slightly change the earths rotation
<specing>
but on some rediculous date that has no scientific meaning whatsoever
<drachensun>
so some multiple of it will exactly match our orbit period around the sun
<specing>
also we add one day every four years, WTF?
<slapin>
I know what facepalm is, just your meaning escapes me
* slapin
go 20 of wr703n's and needs to get an idea on how to use them
<specing>
"...disappointment..."
<slapin>
s/go/got/
<jinzo>
slapin, IoT
<slapin>
specing: disappointment in what?
<specing>
the current time-keeping methods in use by homo dumbians
* slapin
uses NTP and TZ=... date for all his timing needs
* slapin
got a contract last year - Linux device failed to display proper time on web page.
<slapin>
Lots of people fought it before me and failed
<slapin>
Actually, I tried to dismiss it to after what I seen
<slapin>
thay made their own custom timezones library, and made separate javascript application to calculate dates, and had crap level in both
<slapin>
this stuff was "fixed" by returning to normal tools. I think it is normal in embedded - produce monsters and reinvent basic things again and again 'for good'.
<nemik>
are there any current livesuite images for linaro/ubuntu? i'm trying with scripts/mk_livesuit_img.sh and a linaro alip rootfs, not sure if this can work? it's booting but just has a blank screen, not loading the OS
<nemik>
would any changes to the linaro rootfs need to be made for this to work?
KoH__ has joined #arm-netbook
rz2k has quit []
<hramrach>
you can run Linux in Javascript
<leowt>
rm, ARM_PATCH_PHYS_VIRT, why did you enabled it?
eebrah|away is now known as eebrah
<rm>
don't remember; checking now, and I certainly don't remember this being on the top level and labeled experimental in menuconfig...
nomadic has quit [Ping timeout: 245 seconds]
nomadic has joined #arm-netbook
<drachensun>
nemik: I attempted this, any image with X is too large it seems, livesuit throws errors
<drachensun>
nemik: Well the ones I tried were too large, it might be possible to squeeze one in but I did not pursue it because I don't know the goal size to make it work
* mnemoc
wonders if ithamar's flashing tool is operational
ibrah has joined #arm-netbook
eebrah has quit [Ping timeout: 252 seconds]
ibrah is now known as eebrah
<nemik>
drachensun: it works for me fine, no errors from livesuit. i tried the 'alip' image
<nemik>
so it's only 250MB or so. i think that includes X too, just not full desktop
<drachensun>
interesting
<drachensun>
I thought alip was what I had too
Skoti has joined #arm-netbook
<rm>
f***ck..........
<rm>
cubie shop tainted my paypal :D
<rm>
now it comes up chinese even from links on non-chinese sites/stores
<rm>
with no way to switch on screen
<mnemoc>
o.o
voronaam has joined #arm-netbook
<drachensun>
maybe paypal's friendly, responsive, easy to get a hold of customer support can help
<rm>
ok, sorted
<rm>
logged in, then logged out, seems to have worked
<mnemoc>
are you still located in .tw?
<leowt>
CONFIG_CMDLINE=”mem=512M@0×40000000″ is enough to get full memory ?
<mnemoc>
from time to time paypal insist in showing stuff in spanish in mine.... only because i live in .es
<leowt>
mnemoc: spanish?
<nemik>
drachensun: did you just download and use the latest alip image from the linaro site?
<mnemoc>
leowt: don't mess with mem=. disable the feature you don't want in your kernel
<drachensun>
nemik: pretty much, its not the latest though, I got it a few months ago
<drachensun>
nemik: Since you said that I started looking though it. A lot of apt cache junk can go
<leowt>
mnemoc: disabling features liberates dedicated memory?
<drachensun>
nemik: but I think it would still be around 400M did you do anything else?
<mnemoc>
leowt: mali, ve, g2d and fb_reserve
<leowt>
mnemoc: gracias hermano ;)
<mnemoc>
o.o
<mnemoc>
de nada
<leowt>
mnemoc: are you spanish native?
<nemik>
drachensun: not really, just downloaded it. i'm using sunxi-bsp to make a hw-pack and kernel. then running the mk-image script.
<mnemoc>
leowt: chilean
<nemik>
doing it with a custom mk802ii_defconfig based on Toby's for his SD card stuff. so that could be wrong too. i'm targeting this all for an MK802ii, how about you?
<leowt>
i very often go to gallicia, they treat us like no other spanish ppl do. We can talk portuguese to them and they understand. The others dont. And we understand all of the spanish form anywere
<leowt>
mnemoc: Porto
<Skoti>
rm: how about linux-sunxi kernel support for a10s ? is working ?
<mnemoc>
it's so annoying there is not direct train to porto... I really want to visit that city
hg_5 has quit [Ping timeout: 256 seconds]
<mnemoc>
Skoti: olinuxino people just booted their A10s board with it today
KoH__ has quit [Ping timeout: 272 seconds]
<mnemoc>
Skoti: so, yes
<Skoti>
mnemoc: ok. i only need wifi and debian userland working
<mnemoc>
the first an most important step is done from the preinstalled android. get script.bin and a10meminfo data
<drachensun>
nemik: i didn't need the sunxi flash tool, I just boot up on a working sd image
<leowt>
mnemoc: there was something like a TGV Porto<-->Vigo. but its stuck
<drachensun>
nemik: overwrite the u-boot env partition, the kernel partition and the rootfs partition
<leowt>
i mean, they were thinking about constructing
<drachensun>
nemik: There is probably a better way to do it that recoverers more of the nand for use but I haven't messed with that yet
eebrah has quit [Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121213185036]]
<drachensun>
s/recoverers/recovers/
<ibot>
drachensun meant: nemik: There is probably a better way to do it that recovers more of the nand for use but I haven't messed with that yet
<mnemoc>
leowt: 2027? :(
<leowt>
mnemoc: i think there will never be xD no money
<leowt>
mnemoc: where do you live?
focus_it has joined #arm-netbook
<mnemoc>
coruña... on the north
<mnemoc>
there was a guy from vigo here... but he left some months ago
<focus_it>
If they pull of the drivers for graphics, then Freescale IMXQ6 would be a serious competitor to all other SoCs
<focus_it>
I did plan an A10 SoM, but I'm thinking of giving Freescale the higher priority. But obviously wait until the dev board arrives and it it proves a lot more than the PRO A10 supplied by Wits Tech and their limited support, then I think I might have to go with Freescale
<drachensun>
I think that sabre lite board is IMXQ6 right? its already got support with Linaro
<mnemoc>
yes
<focus_it>
dranchensun: It is sabre board lite quad core. Thanks for the head up on Linaro. My boss tentatively agrees to change of plan.
freakazoid0223 has quit [Quit: Leaving]
<focus_it>
I spend 6 months chasing after A10. I don't know what to say.
<mnemoc>
it is doable. but you have to ignore any documentation coming from allwinner :p
<focus_it>
All I want is to build an LCD + touch with Linux running on it so I can get rid of buttons and just use Gambas and speech to do the communications
<focus_it>
memoc: funny!
<mnemoc>
olimex got their A10S olinuxino prototype working today
<focus_it>
I plan to release it all in KiCAD so anyone can build it for their Internet of Things projects.
datagutt has quit [Quit: kthxbai]
<mnemoc>
eagle files on their github
<mnemoc>
hopefully an A10 based will finally follow
<focus_it>
what I want is a tablet out of its case, with many ports including sata and good speakers, no buttons and just touch interface, and a good Linux with all sources such as Lubuntu at lowest cost - then its easy to customize it to any application of the future
<focus_it>
Tomorrow would be fine if some one has plans :)
<mnemoc>
get a table an remove the back part of the case?
<mnemoc>
tablet*
Jef91 has quit [Remote host closed the connection]
<focus_it>
forget to mention - need the thing in KiCAD to allow it to be modified, and Linux must be very solid with all the sources. A10 nearly there - just the problems with Allwinner and Wits tech holding it back
Jef91 has joined #arm-netbook
<focus_it>
If everybody start to use it, they need all that 100% working to get their projects going instead of being stuck doing debugging. Luke is stuck now doing hardware debugging.
<focus_it>
And software debugging.
<leowt>
my old 512 mmc dont work with uboot :/
<focus_it>
Spend $100, buy 20 pieces of 8GB class 10's :)
<focus_it>
Never have to look back then.
<leowt>
focus_it: since im only using sd to load u-boot and kernel
<leowt>
no need for a gb sd
dyoung-away is now known as dyoung
<mnemoc>
leowt: u-boot spl is board specific
<mnemoc>
due to dram initialization
<leowt>
mnemoc: whats ur point?
<mnemoc>
leowt: misread mmc for ram
<mnemoc>
thought your old 512MB *device* wasn't working with current u-boot
<mnemoc>
ignore me
<drachensun>
leowt: I found an updated nand_id list today, that might help you
<leowt>
im getting confused xD
<leowt>
spl: mmc init failed: err - -1
<leowt>
whith an 512MB MMC
<drachensun>
yeah, oops
<drachensun>
nm
hg_5 has joined #arm-netbook
<leowt>
well
<leowt>
got to go
<leowt>
brb
leowt has left #arm-netbook [#arm-netbook]
aexl has joined #arm-netbook
aexl has quit [Quit: Page closed]
rellla has joined #arm-netbook
rellla has quit [Quit: rellla]
<mnemoc>
mripard: nice to see the right uart already implemented :)
<mripard>
mnemoc: yeah
<mripard>
actually, you called a spell on me, I only see that
<mripard>
because after talking to you a bit about it on saturday
<mripard>
I had the precise oops on sunday...
<mripard>
and never had it before.
<mnemoc>
:D
Skoti has quit [Quit: Page closed]
tinti_ has joined #arm-netbook
vinifm has joined #arm-netbook
gimli has quit [Ping timeout: 272 seconds]
<KoH_>
i'm trying to get otg device mode working on sunxi-3.4; i now know that usb_add_gadget_udc is never being called from anywhere within sun5i_usb and so the controller will never be known to the udc-core. however i'm a bit at a loss making sense of the usb_hardware_scan_thread while trying to figure out where an appropriate place for that call might be.
<KoH_>
why is it there and needed? does anyone know why proper initialization isn't performed during sw_udc_probe_otg?
<mnemoc>
KoH_: poke hansg, he is the usb expert
<KoH_>
thanks
<KoH_>
will do
<KoH_>
am i right that otg hasn't been working, yet? i've read some threads on google-groups but i'm not sure that those were about sunxi-3.0
<techn_>
:D
<techn_>
newbie Co.Ltd :D
<techn_>
xxx@newbietech.com :)
<rm>
does anyone boot a Mele A1000/A2000 from SATA?
<traeak>
i use an ancient 16MB card -p
* WarheadsSE
doesn't have one with sata
<rm>
traeak, does the kernel+initrd still have to be on the card?
<traeak>
rm: you'll have to ask mnemoc but i suspect the uboot may require that to override the internal nand
<traeak>
rm: not sure how to make the internal nand redirect to the sata
<traeak>
rm: and yes that'sall i have on the card is the uboot, kernel and initird
<rm>
I was wondering can only uboot be on the SD card
<rm>
could it read kernel from sata
<rm>
probably not
<rm>
but well, it's not a huge problem
<traeak>
that i can't tell you :(
<traeak>
probably depends on what the uboot is able to do
<traeak>
might require too many drivers getting bootstrapped (sata driver)
<rm>
how do you point the kernel to boot from SATA?
<hramrach>
you can set the kernel location in boot.scr
<rm>
root=/dev/sdaX?
<hramrach>
but what locations are supported
<traeak>
yes
<hramrach>
is unclear
<rm>
traeak, upload your SD card image somewhere :>
<hramrach>
yes, to boot from sata you set the root device
<traeak>
uEnv.txt i have root=/dev/sda1 on the last line
<hramrach>
did not try but should work
<hramrach>
boot.scr has kernel commandline too
<traeak>
from what i recall some months ago (yuk early october), uEnv.txt was the simplest way to do the root on /dev/sda1
<traeak>
extraargs=rootwait perhaps also required for uEnv.txt
dfletcher has joined #arm-netbook
dfletcher has quit [Changing host]
dfletcher has joined #arm-netbook
<hramrach>
if it's text fil then it's simplest
dfletcher is now known as drgreenthumb
<hramrach>
boot scr has to be compiled wiht u-boot tools