<lundman>
So if one was to add ISO support (say, dvd/bluray) to android, I would guess one should use ContentProvider() API
<Marex>
lundman: is android really suited to be used as a DVD player ?
<lundman>
ISO file support, but sure, it could have an attached DVD player too, why not
<WarheadsSE>
Why not, its only MPEG2 for DVDs
<WarheadsSE>
decss isn't that heavy Marex
<lundman>
and trivial, already re-written the libdvdcss/libdvdnav libraries a few times, mostly to add bluray support (udf2.60)
<lundman>
so, I just need to android specific magic that adds another mediatype
<lundman>
pretty sure i saw mele do just that for iso files
<lundman>
wonder if that source is available
<WarheadsSE>
maybe, somewhere
* WarheadsSE
hasn't booted his Mele A100 in too long
<lundman>
i booted mine, but didnt attach console :)
<Marex>
lundman: that's the shitty part about BSD/MIT licensing ... GPL is much better at that
<lundman>
alas, i dont give a rats arse about licenses
<lundman>
:)
<lundman>
source to start from speed up work true, but doing it yourself increases understanding
<Marex>
WarheadsSE: it's not about the media decoding ... I think the part that handles the DVD drive might be missing ... and it's still a touch interface, it's annoying to control with <mouse/rc>
<lundman>
the dvd drive is standard usb storage, should be fine
<lundman>
as for the dvd menu triggers, should work with touch screen, if someone writes it
<Marex>
lundman: I actually do care about these ... basically because I did hack on both OpenBSD and Linux kernels and figured the difference the hard way
<lundman>
sure, but if I can get it, I'll boot IDA. it'd be nice if I didnt have to, this is true
<Marex>
lundman: dvd drive isn't a standard usb storage, at least not in the way of "sd" driver
<Marex>
lundman: it's I think sg or something ... something a bit different
<Marex>
you need to push the atapi command through, no ?
<Marex>
"sr" even
<lundman>
sure, if you mean CSS, it has something like 3 extra atapi commands, after that, it is regular work
* Marex
is fine with having almost every FSL arm chip in possession :S
<hno_>
cat1, no. had hoped to but not sure where mr cubieboard is..
<cat1>
looks qute chip and looks to be better equipped that paspberry pi.
<cat1>
s/qute/quite/
<ibot>
cat1 meant: looks quite chip and looks to be better equipped that paspberry pi.
Quarx has quit []
kaspter has quit [Ping timeout: 264 seconds]
<mnemoc>
talking about mr cubie... he seems alive
<calris>
Why would you think he wasn't?
hp_ has quit [Ping timeout: 246 seconds]
jlj has joined #arm-netbook
<mnemoc>
there is no life outside irc
<Marex>
mnemoc: life is overrated ;-)
<mnemoc>
:)
gsilvis has quit [Read error: Connection reset by peer]
gsilvis has joined #arm-netbook
gsilvis has quit [Ping timeout: 246 seconds]
XenGi has quit [Read error: Connection reset by peer]
gsilvis has joined #arm-netbook
hp_ has joined #arm-netbook
<calris>
boing boing
<Marex>
calris: you're too alive ;-)
<calris>
:)
<Marex>
calris: how's the mele ?
<Marex>
oh ... but wait
<Marex>
calris: does that mean you're starting on uboot/arm hacking ?!
<calris>
Still sitting on the bench
<calris>
Yep
<calris>
Housework taking priority :(
<Marex>
calris: I'm sure you're a good housewife :-)
<Marex>
calris: who's gonna maintain x86 then? You ain't gonna give up on it, are you ?
<calris>
It's running on idle atm
<calris>
If someone needs some x86 development done
<calris>
I can restart
jas-hacks has quit [Remote host closed the connection]
<Marex>
calris: we still need x86-bios machine, you know ...
<Marex>
calris: I have some sample code for that, but it needs cleanup
kocka has joined #arm-netbook
<kocka>
hoi
<Marex>
actually ... why is x86 init done in 3 stages (f f_r r) ?!
<calris>
To get cache enabled super early
<calris>
Faster boot
<calris>
Mares: BIOS....Blech !!!
<calris>
s/Mares/Marex
<hno_>
The A10 have ~20KB "BIOS" ROM.
walrus_ has joined #arm-netbook
<mnemoc>
"BOOT" ROM sounds better :p
<hno_>
That's 9 KB.
<mnemoc>
uh.. I thought they were the same
<hno_>
FEL and BROM is two completely separate programs.
<mnemoc>
so FEL is stored in a place called "BIOS" or "BOOT" ROM?
<mnemoc>
I'm clear FEL is a separated app... only trying to get the terms right
<hno_>
There is no right, only left.
<mnemoc>
:)
<hno_>
neither is a BIOS in the PC sense, but the term BIOS is used outside PC as well.
<hno_>
however, would not say the A10 ROM qualifies as a BIOS.
<calris>
Bios is a set of real-mode interrupt handlers to perform low leave I/o like disk and video access
MISHU has joined #arm-netbook
<MISHU>
hi there!
<hno_>
hi MISHU
<MISHU>
any updates on the project?
Nu7s has joined #arm-netbook
<MISHU>
ahow much is shipping to Spain is gonna cost?
<mnemoc>
MISHU: what project in particular?
<hno_>
nothing at that level yet.
<MISHU>
the cubieboard
<MISHU>
ლ(ಠ_ಠლ)
<mnemoc>
afaik the first batch went on production last friday, but that's presold to an investor
Quarx has joined #arm-netbook
<Nu7s>
mnemoc, are you involved in the cubieboard?
<specing>
0.o We have reached 100 users!
<MISHU>
and to the general public?
<mnemoc>
nope. that's what Tom told me
<mnemoc>
MISHU: end of the month probably
<Nu7s>
oh ok
<specing>
Tom (aka hipboi, aka Mr. Cubie, aka Tom Cubie) is the one in charge
<MISHU>
is it going to be a discount for buying more than one ? 10 - 20 for example?
<specing>
and he is not here at the moment
<MISHU>
how many are you going to make?
<Nu7s>
I bought a raspberry PI, but it lacks in certain areas I think the cubieboard will not lack.
<mnemoc>
again. it's a one man project, and that man is not here
<specing>
MISHU: ask tom when he comes by
<MISHU>
sorry for he question rampage. I'm excited
<MISHU>
one man? really?
<mnemoc>
yes. one man
<mnemoc>
tom
<mnemoc>
aka hipboi, aka Mr. Cubie, aka Tom Cubie
<mnemoc>
that also means he is overwhelmed. beside been very busy at work getting sun6i out and preparing his life for his offspring who is supposed to arrive in 2 months
<Nu7s>
and don't forget his Aliexpress store ;-)
<mnemoc>
true :)
* calris
has a penny drop moment
* Nu7s
picks up the penny
* calris
didn't realise Tom == Cubieboard
* calris
feels stupid
<RaYmAn>
mnemoc: the site talks about a team :> lol
<calris>
Cubieboard just got 1000% more awesome :)
<mtd>
cubieboard looks cool!
<mnemoc>
RaYmAn: well... he has a friend doing the EE design and i'm sure he isn't doing the factory work :p
<calris>
A real community member, a SoC manufacturer employee, an online store owner
<mtd>
...especially in the "shipping" part.
<Nu7s>
Now if only XBMC had Allwinner support...
<Marex>
calris: that x86 crap only complicates the DM effort though ;-)
<calris>
Nu7s: I think it will get there
<Nu7s>
I hope so, it would be awesome
<mnemoc>
the a10-olinuxino looks awesome too. /me wants both :p
<calris>
Mares: do not be afraid to break x86... between you, me, Wolfgang and everyone here, x86 has officially entered bit-rot mode
<calris>
s/Mares/Marex
<calris>
:(
<mnemoc>
calris: tab is your friend
<calris>
tab?
<mnemoc>
a key, next to the Q in keyboards, used to auto-complete nicknames
<mnemoc>
do you think people actually types RaYmAn ?
<calris>
I'm on a HP TouchPad
<calris>
No auto-complete :(
<calris>
Using AndChat
<RaYmAn>
just need a suitable softkeyboard that supports it :P
<RaYmAn>
mnemoc: surprisingly, a lot do :P
<mnemoc>
RaYmAn: :D
<mnemoc>
calris: there are virtual keyboards for android with tab key too
<calris>
Owwwwww
<calris>
For CM10?
<mnemoc>
yes
<calris>
Tell me more....
<Marex>
calris: tab key ... two keys below the VIM key (also often called escape key) ;-)
<calris>
:P
<RaYmAn>
even swype has tab ;P (long press space)
* hno_
misses his N900...
<mnemoc>
hno_: not happy with your new lumia? *cough*
<calris>
Any good keyboard suggestions?
<hno_>
haven't bought a new phone yet. Borrowing an android crap thing while tryig to figure out what I want..
<calris>
Get an iPowne
<orly_owl>
wait for the iphone 25
popolon has quit [Ping timeout: 246 seconds]
<mnemoc>
it's sad there are so few phones with sliding keyboard
<calris>
It's sad that some brain dead jury thinks Samsung owe crapple $1B
popolon has joined #arm-netbook
<calris>
popolon: nice IP :)
<popolon>
calris, sorry :D
<popolon>
ipv6
calris1 has joined #arm-netbook
<calris>
Hmm
calris is now known as calris_touchpad
<cat1>
hno: what did you do to your n900? i was working in integration team at the time it was about to release.. good old days..
calris1 is now known as calris
<calris>
Ahhhhh
<calris>
that feels better :)
<mnemoc>
cat1: it seems it wasn't waterproof enough :p
<cat1>
ah, i see :D
<mnemoc>
i know people who lost their n900 on water and people who wasn't able to charge it anymore
<mnemoc>
not sure what's hno's case
* cat1
confused n900 with n770, in n900 times he joined harmattan team..
hp_ has quit [Ping timeout: 246 seconds]
<cat1>
cat1: you need to take your omega3 stuff.
<mnemoc>
o.o
<cat1>
anyway both times were good /me thinks..
* mnemoc
wishes a painful death to Elop and those in the board who let him destroy nokia
<popolon>
calris, about complete keyboard on android, you can try 'hacker's keyboard', a free (as GPL) applcation, an be downloaded on GNU f-droid.org market and perhaps on google market (don't know, don't use it)
<The-Compiler>
Marex: otherwise they probably would've used something more easy to reach thane sc
<The-Compiler>
than esc*
<The-Compiler>
popolon: well, it all depends on if the application actually takes Tab as an input for "complete nickname". If it doesn't, having a Tab key doesn't help much
<cat1>
Just found out that compilation of sun3i on 3.6 kernel is broken because of missing sun3i_pm.h. Cannot find it in other branches as well. Anybody, any hint where to find this file?
* cat1
actually does not have any sun3i board so it is not critical for him, but he just wants to keep things tidy
cat_n9 has joined #arm-netbook
hno_ has quit [Ping timeout: 276 seconds]
jlj has joined #arm-netbook
revident has joined #arm-netbook
hp_ has joined #arm-netbook
RITRedbeard has joined #arm-netbook
ssalevan has joined #arm-netbook
<mnemoc>
hno: right. the board has 2012-08-08 written on it. the previous picture had 2012-07-01. so it's just a newer prototype
<mnemoc>
specing: btw, my problem yesterday with atxmega128a3 was solved by upgrading the firmware of the programmer :|
<specing>
interesting
hno_ has joined #arm-netbook
<hno_>
cat1, are you sure it's sun3i_pm.h?
<cat1>
hno_: this is what i see
<cat1>
arch/arm/mach-sun3i/pm/sun3i-pm.c:36:34: fatal error: linux/power/sun3i_pm.h: No such file or directory
<mnemoc>
cat1: we have a tested r3p0 (from Turl) to pull in
<hno_>
hipboi, can you answer specific questions regarding A13 registers or is that secret?
<cat1>
great, what about 3p1? :)
<mnemoc>
cat1: no libs
<hipboi>
hno_: you can ask
<mnemoc>
cat1: but also, we need to integrate mali/sun5i before that
<hipboi>
let's see if it's secret
<mnemoc>
cat1: it doesn't help to let sun4i escape farther
<mnemoc>
cat1: or it becomes unmaintainable
<cat1>
mnemoc: agreed, though i started fork with 3.6 + linux-sunxi-3.4 merge
<cat1>
mnemoc: sun4-5 are compilable, sun4i is bootable, tested it with mk802
<cat1>
mnemoc: some hacks had to be done though..
<mnemoc>
cat1: we can't move forward without fixing the broken windows first. 3.4 has broken usb gadget and incomplete serial driver
<cat1>
mnemoc: i understand, i will keep merging as long as fixes are coming into 3.4
<hno_>
hipboi, ok, here is two questions for you. What is the MBUS clock? and what is DRAM controller register 0x23c?
<mnemoc>
ok
<cat1>
mnemoc: besides, i just wanted to play around with 3.6
<mnemoc>
cat1: fair enough
<mnemoc>
cat1: but help fixing 3.4 troubles is more welcomed ;-)
cavera has quit [Ping timeout: 244 seconds]
<RaYmAn>
mnemoc: slave driver ;)
<hipboi>
hno_: this is very specfic, i need to look at the data sheet
<cat1>
mnemoc: sure, as soon as i start using this heavily :) btw, any plans to get it allwinner upstream? the code demands a lot cosmetics!
<mnemoc>
RaYmAn: :)
pawel5870 has quit [Ping timeout: 245 seconds]
<hipboi>
hno_: i am not at zhuhai these days
<hno_>
hipboi, changed job=
<hno_>
?
<mnemoc>
cat1: it's not merely about cosmetics. we need to have devicetree and mtd support before even dreaming about mainlining
<cat1>
mnemoc: right.. but we'd rather keep code in shape anyway
<hno_>
We don't need mtd support to mainline. Obviously we can't mainline the existing NAND driver, but..
<mnemoc>
ok, devicetree and gpiolib to mainline the core
<hno_>
yes
hipboi_ has joined #arm-netbook
<cat1>
mnemoc: do you have any mailing list dedicated for patch review? it would be easier to track issues.
<RaYmAn>
axp209 also needs some love esp. for a13
<RaYmAn>
though, should be relatively simple
<mnemoc>
cat1: we currently use github's crap tracker
<hipboi_>
hno_: can you email the question so i can reply you when i go back to the office
<hno_>
Sure. @allwinner or @gmail?
hipboi has quit [Ping timeout: 244 seconds]
<hno_>
hipboi_?
<hipboi_>
hno_: yes
<mnemoc>
hipboi_: when do you return from wuhan?
<hno_>
hipboi_, which address you prefer, or does not matter?
<hipboi_>
mnemoc: i think tomorrow, or the day after tomorrow
<hipboi_>
hno_:gmail, please
tombola has joined #arm-netbook
kaspter has quit [Ping timeout: 240 seconds]
<hno_>
hipboi_, ok, sent.
<hipboi_>
hno_: ok. thanks
<cat1>
mnemoc: /me ponders that it could be possible to think of manilining mach-sun?i already; of course some polishing must be done before, still might be good starting point.. well, never mind :)
<RaYmAn>
mnemoc: does it make sense without a usable gpio implementation?
<RaYmAn>
err
<RaYmAn>
cat1:
<cat1>
RaYmAn: well, it depends on how much one wants to do with device.
<cat1>
s/how much/how many things/
<ibot>
cat1 meant: RaYmAn: well, it depends on how many things one wants to do with device.
<RaYmAn>
yeah, but it wouldn't even partially boot then...
<cat1>
RaYmAn: hmm, how come it boots now?
<mnemoc>
cat1: using a gpio implementation which upstream won't ever accept
<RaYmAn>
^^
<cat1>
mnemoc: oh-oh..
<cat1>
mnemoc: well, then minimalistic goal is set: we need to improve just this.
* RaYmAn
nominates cat1 for rewriting gpio implementation
<cat1>
hooray!!
<RaYmAn>
pinmux will most likely also need some serious changes to be somewhat in line with everything else
* cat1
is not a master though
<hno_>
rz2k, no news yet. But have a feeling it's very close.
<rz2k>
hno_: thanks, it just seems like I will need to etch out a board for my JTAG connection, this will take time and I can do this only on weekend.. so I cant even check whats going on.
<mnemoc>
cat1: the most early bit of allwinner hackery we need to get rid of is in arch/arm/kernel/head.S
<hno_>
rz2k, u-boot starts just fine. But Linux crashes.
<mnemoc>
cat1: the map for script.bin and another 3MB map for io space
<cat1>
mnemoc: btw, i removed this part from head.S in my tree -- nothing is broken afterwards
<hno_>
mnemoc, the head.S crap can be killed without any pain.
<mnemoc>
hno_: the r0/r1 sure... but the memory maps?
<hno_>
the script.bin map is for script.bin. Will die with device tree.
<hno_>
the i/o map is only for probing the dram controller, which also should not be done.
traeak has quit [Remote host closed the connection]
<cat1>
hno_: btw, anyone who is working on device tree for allwinner?
<hno_>
not yet
<hno_>
but there is nothing strange about it.
<cat1>
hno_: yeah, but one must be really in details about hardware
<cat1>
hno_: or you mean that there is nothing strange that none is working on it?
<hno_>
the hardware is not complex, and the data we need in devicetree is almost all in script.bin + platform.h
<RaYmAn>
cat1: script.bin is essentially a custom device tree... So all the details are pretty much there - just needs to be done right
_hipboi_ has joined #arm-netbook
<cat1>
RaYmAn: good
traeak has joined #arm-netbook
hno_ has quit [Quit: leaving for home]
<mnemoc>
cat1: code to play with the pins can be found in sunxi-tools' pio.c
* cat1
thinks of learning how to create dt
hipboi_ has quit [Ping timeout: 246 seconds]
<cat1>
mnemoc: is this part of kernel source tree? can't recall seeing such file.
<mnemoc>
sunxi-tools has it's own repo, also in github
<cat1>
mnemoc: ah, riiight, thanks!
<cat1>
mnemoc: btw, i can't wait your mali merge. any estimate?
* cat1
tries to run mer+nemo on mk802, gpu support is badly needed.
eFfeM_work has quit [Ping timeout: 245 seconds]
<mnemoc>
r3p0? I want to merge the nand and video unification into 3.4 first. then update to the latest 3.0.x, and r3p0 comes next
walrus_ has quit [Quit: Page closed]
<cat1>
mnemoc: can i see this unification changes already somewhere, just to get the idea or it is in your local repo?
hp_ has quit [Ping timeout: 246 seconds]
<mnemoc>
cat1: they are in the default branch of linux-allwinner....
<mnemoc>
cat1: skip the commits that say "restore .config compatiblity"
<mnemoc>
those are only relevant to our "stable" branch, 3.0
hipboi_ has joined #arm-netbook
<cat1>
mnemoc: ok, thanks, will take a look.
hno_ has joined #arm-netbook
<mnemoc>
you may also want to look to the plat and plat-gpio wip branches of 3.4. the first explores the merge of mach-sun?i and the second explores gpiolib upon the other
_hipboi_ has quit [Ping timeout: 276 seconds]
Nu7s has quit [Ping timeout: 260 seconds]
<hno_>
hipboi_, you tried to ping me some time ago (23 aug). Do you remember what it was about?
hipboi_ has quit [Ping timeout: 240 seconds]
cat_n9 has quit [Ping timeout: 244 seconds]
MISHU has quit [Quit: Page closed]
j1nx_ has joined #arm-netbook
<mnemoc>
wb j1nx_
<j1nx_>
thx
<j1nx_>
Do I quickly saw correctly the u-boot sun5i branche of hno has proper memory detection support?
<hno_>
not that I know.
<j1nx_>
then I guess not :D
<hno_>
it has parameterized dram setup so you can feed it what dram you have. But boot info is still hardcoded.
<j1nx_>
before I ran home, I quickly saw some stuff changed in dram.c
<j1nx_>
but did not read it
<hno_>
all of dram.c changed.
<j1nx_>
Will look into it later then, plans to push it to sun4i as well?
<hno_>
this dram.c is based on pm/standby/ and observation of what boot0 does.
<hno_>
first need to get it working proper.
rellla has quit [Quit: rellla]
<j1nx_>
you are deving on olixino A13 at the moment?
<hno_>
yes
jlj has quit [Ping timeout: 244 seconds]
cat_n9 has joined #arm-netbook
<j1nx_>
Am looking at the new pictures of the Cubieboard. K1ck A$$
<j1nx_>
Any dates from Tom?
hno_ has quit [Ping timeout: 246 seconds]
hno_ has joined #arm-netbook
<hno_>
j1nx_, none yet, but he was here some minutes ago.
<j1nx_>
propbably to bed already
<mnemoc>
he is in wuhan because of family matters, but he said he will be returning to work (in zhuhai) in 2/3 days.
pawel5870 has joined #arm-netbook
cat_n9 has quit [Ping timeout: 264 seconds]
Nu7s has joined #arm-netbook
<cat1>
hno_: tried to use linux/power/aw_pm.h as you suggested, now compilation fails like this:
<cat1>
arch/arm/mach-sun3i/pm/sun3i-pm.c:37:69: fatal error: ../../../../drivers/power/sun3i_power/regulator/machine.h: No such file or directory
<mnemoc>
techn: unfortunatelly the "full thing" gives Quarx very ugly colors. but we probably can integrate the lcd_spi part first, and get the rest of the patch in a branch for further analysis
<mnemoc>
cat1: :)
<mnemoc>
RaYmAn: techn is who fixed fbcon and unified the video drivers
<RaYmAn>
mnemoc: I know
<mnemoc>
:)
furan- has quit [Quit: Bye]
<techn>
4, modify hdmi avi packet for hd;
<techn>
5, support deinterlace;
<techn>
those I can do.. is there separate patches for those?
<techn>
or just that blob?
<mnemoc>
techn: just that blob
<mnemoc>
techn: the pull request take the spi part out
<mnemoc>
takes*
<techn>
yep.. I don't have device to test those 3 first :(
<techn>
.. but I'll hope that I'll have in couple days.. since I bid for a10 and a13 tablets (partly broken) :)
<mnemoc>
the color problem comes out of hdmi, not lcd
<mnemoc>
lcd seems to work fine
<techn>
on that full or single patch?
<hno_>
no, the A13 is still irritated over something.
<techn>
ah.. just re-read
<mnemoc>
techn: both. the patch from the pull request is a unification of the lcd_spi support in that large patch
<mnemoc>
techn: it seems that in some part of the large patch they changed the default color system to something weird
<mnemoc>
techn: but I'm sure some safe improvements can be taken out of it
* cat1
wonders if that would be possible to run checkpatch.pl before applying any patch. Lindent would please eyes too :)
<mnemoc>
cat1: the dirty commits come from the internal gpl-violating git repos of allwinner
<cat1>
mnemoc: what can i say.. nice! :D
eFfeM has joined #arm-netbook
<mnemoc>
at this point commits fixing whitespace and "style" only harm history
<mnemoc>
so we fix it slowly on the parts where we fix/improve actual code
<The-Compiler>
anyone in here interested about a broken pandaboard? Connected it to 12V, RS232 voltage IC started to smoke, but I guess something else (maybe even the CPU) is also damaged because it doesn't boot anymore
<cat1>
mnemoc: right.. sorry for whining..
<mnemoc>
cat1: I share your feeling
<The-Compiler>
just pay me the shipping (or if it's not that much, nevermind) and it's yours.
<mnemoc>
cat1: and my vim actually sets the background color of bad whitespacing to red. hurts even mroe
<cat1>
mnemoc: same here
<mnemoc>
:)
<The-Compiler>
mnemoc: that sounds interesting, how do you do that?
<mnemoc>
and some patterns to define WrongWhitespace
cat_n9 has joined #arm-netbook
<cat1>
mnemoc: however cannot stop thinking of somewhat more collaborative way with allwinner guys: e.g. having one common malining list for patch review/exchange..
<RaYmAn>
I don't think allwinner is in any way comitted to being part of anything community. afaik tom does it out of interest and stuff, with just barely permission for it?
<mnemoc>
cat1: i wish... but their management seems to prefer the oposite direction. afaik they changed the copyright notes in their tree to some random third company (probably to reduce liability), and newer SDKs get .git dirs removed and key parts (like mach-* and cedarx kernel drivers) shared as bin only
<mnemoc>
cat1: that commit above was gracefully shared by Tom to fix ainol aurora support
<mnemoc>
on the bright side our code was all GPLed and it actually works :)
<cat1>
match RedundantWhitespace /\s\+$\| \+\ze\t/
<traeak>
hmmm
<The-Compiler>
actually.. not sure if I should four spaces or a tab displayed as for spaces in languages other than python...
<The-Compiler>
s/should/should use/
<ibot>
The-Compiler meant: actually.. not sure if I should use four spaces or a tab displayed as for spaces in languages other than python...
<The-Compiler>
cat1: what does \+\ze\t match?
<traeak>
cat1: how do you trigger this condition?
<cat1>
just put it into .vimrc. I also made shortcut that forces it.
<traeak>
i think my way of using vim may not typically force this (I'm trying to see what it does)
<mnemoc>
traeak: turns the bg color of evil spaces to red
<traeak>
yeah, i guess one way: space followed by tab then code
<traeak>
not lighting up for me, using standard xterms
ssvb has quit [Quit: Leaving]
<cat1>
The-Compiler: do not remember anymore, besides, it is not my-hand-made :)
<hno_>
Ok, some progress. It's something boot0 does after DRAM setup that makes it stable.
<mnemoc>
\o/
<mnemoc>
how many bytes to decode about DRAM setup? :p
<mnemoc>
s/about/after/
<The-Compiler>
traeak: hmm, same here with urxvt
<ibot>
mnemoc meant: how many bytes to decode after DRAM setup? :p
<hno_>
The while DRAM setup in boot0 is decoded.
<hno_>
s/while/whole/
<ibot>
hno_ meant: The whole DRAM setup in boot0 is decoded.
<hno_>
now this is confusing.. still unstable after boot0 attempted loading boot1 from nand.
popolon has quit [Ping timeout: 246 seconds]
lkcl has quit [Ping timeout: 264 seconds]
<hno_>
Hmm... disable instruction cache, that should not make this difference.
popolon has joined #arm-netbook
TomNL has quit [Ping timeout: 252 seconds]
<hno_>
or maybe it's just openocd that dislikes instruction cache.
pawel58701 has joined #arm-netbook
pawel5870 has quit [Read error: Connection reset by peer]
eFfeM has quit [Quit: Leaving.]
jlj has joined #arm-netbook
ssvb has joined #arm-netbook
<hno_>
Hmm... maybe it actually works and only console settings are wrong?
<hno_>
cpu still running now.
rellla has joined #arm-netbook
Quarx has quit []
hammer has joined #arm-netbook
hammer_ has quit [Ping timeout: 244 seconds]
<mnemoc>
call something...
<hno_>
?
<The-Compiler>
hmm.. is it possible to use android to modify the NAND uboot so it boots from SD?
<The-Compiler>
(or maybe even boot the kernel from NAND and have the rootfs on the SD?)
<hno_>
The-Compiler, sure, just replace it with an MMC capable u-boot.
<The-Compiler>
I want to try the stock uboot because of the weird issues I'm having
<hno_>
does it have the mmc command?
<mnemoc>
hno_: if only console is not printing you should be able to jump to a function that does something you can then verify reading memory... or not?
<hno_>
easier to just fix the console. A bit of confusion between uart0/uart1 in my setup.
<The-Compiler>
hno_: nope. Can I put the kernel on the NAND and the rootfs somewhere else then?
<hno_>
Yes
<The-Compiler>
any guide on what I should do? I really don't want to render it unbootable by some stupid mistake so I'd have to fire up a VM to get livesuit
<hno_>
The-Compiler, isn't there several FAT partitions in Android? If there is then just dump the kernel in one of them and use fatload to load it from u-boot.
<The-Compiler>
hmm, can't see any fat partition I could use
<Turl>
the last partition is vfat
<Turl>
the one labeled "UDISK"
<The-Compiler>
Turl: last one where? I only see /dev/block/nandi on /mnt/private but I can't access that it seems
L84Supper has quit [Quit: puff of smoke]
lkcl has joined #arm-netbook
<Turl>
The-Compiler: on android it's mounted on /mnt/sdcard
<Turl>
usually it's nandi1 I think
<The-Compiler>
oh right, I just saw sdcard and assumed it was the external sd
<hno_>
The-Compiler, that's android idea of consistent naming.
<Turl>
hno_: it's more like legacy baggage I'd say
<Turl>
many stuff has /sdcard hardcoded
<The-Compiler>
so where on earth is the device for the external sd
<The-Compiler>
ah, /dev/block
<hno_>
The-Compiler, that's a mmc block device somewhere..
<Turl>
not sure where a10 devices have it, but it should be somewhere in /mnt/ too
<Turl>
the device itself is /dev/block/mmcblk0pX
<RaYmAn>
at least they didn't make their own custom mmc stuff
<Turl>
RaYmAn: they did do custom nand stuff => no yaffs :(
<RaYmAn>
I know
<RaYmAn>
it annoys me :P
j1nx_ has quit [Quit: A fine is a tax for doing wrong. A tax is a fine for doing well.]
<Turl>
I was like hmm I'm gonna try yaffs
<The-Compiler>
sh: cp: not found
<traeak>
heh
<The-Compiler>
pfff, okay, cat then.
<RaYmAn>
It's one of the things I'll certainly put effort towards when/if I have time and no one else do it before (going nand -> mtd)
<Turl>
then noticed it depended on mtd
<Turl>
yeah
<hno_>
you may have busybox even if cp is not found
<Turl>
The-Compiler: yeah android has no cp, use busybox cp if you need it
<Turl>
or just cat/dd
L84Supper has joined #arm-netbook
<The-Compiler>
so I have /mnt/sdcard/uImage now... let's see if I manage to get it to boot that
<pawel58701>
The-Compiler: I'm using kernel from nightly build
<pawel58701>
but booting doesn't work from SD u-boot
<The-Compiler>
I have random crashes and funny paging exceptions no matter what I do.
tombola has left #arm-netbook [#arm-netbook]
Almamuetya has quit [Ping timeout: 246 seconds]
<Turl>
lerc: yeah, just saying I don't see the cubieboard on slashdot
<pawel58701>
It's possible to change linaro resolution at runtime ? I changed output resolution to 1600x1024 but linaro resolution is still 1200x...
<mnemoc>
pawel58701: script.bin
<mnemoc>
or ioctl
<The-Compiler>
hno_: booting via stock uboot doesn't fix my issues so it's not uboot-mmc's fault
<mnemoc>
Turl: how do you check for io wait stuff?
<Turl>
top shows a % of io wait mnemoc
<The-Compiler>
running mtest now in uboot which also looks good... what the heck could be wrong with my board then
<pawel58701>
mnemoc: I used ioctl (vga_on.c) but I have black border. xrandr shows 1280x720 and ioctl was set to 1680x1050
j1nx_ has joined #arm-netbook
<j1nx_>
does anybody already have pictures of the inside of the Mele A3600 ?
<j1nx_>
Propbably the same, but just to be sure.
* mnemoc
wants to see it's guts too
<j1nx_>
All outputs are positioned at the exact same location, so I guess no real suprises
<lerc>
Turl: Slashdot will eventually link to the engadget article that links to the MAKE article that links to the Liliputing article which links to the cnxSoft article which links to the linux-sunxi page.
<mnemoc>
:p
<j1nx_>
lol
<Turl>
mnemoc: hm doesn't look like io wait
<Turl>
but php is using quite a bit of cpu, wonder what is it doing :P
<mnemoc>
Turl: I just stopped php-fpm
<mnemoc>
and letting it rest a bit
<mnemoc>
load down to 3 now
<Turl>
yeah
<mnemoc>
php started
<hno_>
Yes! it boots!
<mnemoc>
Turl: suggestions to improve the php-fpm pools config is highly welcomed
<Turl>
mnemoc: cubieboard is a drupal right?
<mnemoc>
Turl: yes
<Turl>
mnemoc: did you enable the caches?
rellla has quit [Quit: rellla]
<mnemoc>
Turl: I didn't set it up...
<The-Compiler>
I guess I'll just post on the mailinglist to see if anyone can help me.. and if not I spent $100 on hardware I can't use :(
<j1nx_>
hno_ : Congratz
<Turl>
well getting slashdotted *and* not caching is a combination set up for disaster
<mnemoc>
hno_: kudos!
<Turl>
hno_: great!
<j1nx_>
The-Compiler : wassup?
<mnemoc>
Turl: congratulations, you have become sudoer :p
<Turl>
mnemoc: :)
<lerc>
Throw open the switches on the sonic oscillator... and step up the reactor power three more points!
<mnemoc>
Turl: but please keep me posted of changes :p
<The-Compiler>
j1nx_: spend like three evenings getting my Mele A2000 to boot anything apart of the stock android, no luck. I get random lockups and memory paging kernel exceptions, but memtest in uboot and the stock android run fine
<mnemoc>
Turl: if you know how to do that without hipboi's drupal credentials...
<Turl>
mnemoc: it works over drupal, it's like a transparent proxy
<Turl>
over anything really
<mnemoc>
Turl: so it's nginx setup?
<Turl>
it's server independent
<mnemoc>
.oO
<Turl>
you move the NS records to point to them
<The-Compiler>
mnemoc: yes, yes
<mnemoc>
The-Compiler: :<
<The-Compiler>
mnemoc: even tried booting the kernel with the stock uboot
<mnemoc>
Turl: in that case we need hipboi
<mnemoc>
Turl: do you use them somewhere?
<Turl>
mnemoc: I own cubieboard.org, so I can change the NS at will
<mnemoc>
Turl: setting up nginx's reverse proxy's caching sounds like a good idea
<mnemoc>
Turl: oh, didn't know that
<mnemoc>
messy world
<Turl>
mnemoc: CM uses it on some pages to offload quite a bit of stuff
<hno_>
u-boot changes pushed to sun5i branch if someone wants to play with them.
<steev>
boo i missed the announce for the cubieboard :(
<mnemoc>
steev: it hasn't been announced...
<steev>
cubieboard.org has it
<Turl>
mnemoc: if you agree with it I'll enable it for cubieboard.org
<mnemoc>
Turl: +1
<hno_>
will clean things up a bit in the next days.
<mnemoc>
steev: it got known from the irc logs when tom mentioned his project here some months ago
<mnemoc>
steev: then he added some specs to the wiki
<steev>
mnemoc: ah, so it's technically not available yet
<steev>
not even preorders
* steev
really wants one
<mnemoc>
steev: and cnx-soft found it there some days ago....
revident has quit [Quit: Combustible lemons? Bah, I bring you weaponized asparagus!]
<steev>
there was a comment on the HN site with the link to the board, and the guy is basically saying it's a scam because Tom's nickname is hipboi
<RaYmAn>
lol
<steev>
i mean seriously, that's what you're gonna complain about? someone's nickname?
<hno_>
What wonderful background research on who is who.
<ccssnet>
ya seems cubieboard.org is getting hammered. takes a while for pages to load for me
<mnemoc>
ccssnet: load of the server is over 18 again :<
<mnemoc>
damn php
<ccssnet>
lol. ya php is not super fast, but i do use it myself here: http://atccss.net
<traeak>
hehe
<traeak>
cubieboard is toast
<traeak>
i probably just helped DOS him
<ccssnet>
well i closed my browser. no longer taking part myself
<Turl>
mnemoc: DNS moved, hopefully it'll propagate soonish :)
<steev>
it's on the front page of news.ycombinator.org
<hno_>
It still answers fine, only a bit slow.
<ccssnet>
thats cause for a hammering also
Nu7s has joined #arm-netbook
<hno_>
And mentioning the R-Pi. ofcourse it gets hammered.
<hno_>
11.
<hno_>
Watch out raspberry pi - 1ghz 1gb ram 4gb flash 96 gpio - $49 (cubieboard.org)
<steev>
yep
<traeak>
hmm....so they plan on mass manufacturing these after all?
* mnemoc
wonders why no one compares it with the beaglebone
<traeak>
if cubieboard would have come out in april/may ....
<hno_>
sweclockers also have it..
<traeak>
the beagle* stuff never had as much marketing as the rpi did
<hno_>
and wikipedia. and slashgear. and techspot, and makezine, ang geeky-gadgets and .....
<RaYmAn>
I doubt rpi would have gotten anywhere near the attention it did, if it wasn't because RPI is a foundation with a fairly noble public goal, heh
<RaYmAn>
It's kind of a pity it has mostly turned into an xmbc board.
Nu7s has quit [Read error: Connection reset by peer]
<RaYmAn>
</minor rant>
<RaYmAn>
;)
<Turl>
mnemoc: can you reset my pw on the server? :< I forgot it
<RaYmAn>
Turl: public key auth! :P
<steev>
RaYmAn: oh i agree
<Turl>
RaYmAn: yeah that's why I forgot it :P
<steev>
it was supposed to get into the hands of school kids... i've not heard of any kids using them
<Turl>
RaYmAn: sudo doesn't use my key though
<mnemoc>
Turl: same as login
<RaYmAn>
Turl: good point, lol
<RaYmAn>
steev: rpi foundation actually posts almost daily about kids doing cool stuff with them :)
<RaYmAn>
steev: on twitter
<Turl>
thx mnemoc
<mnemoc>
yw
<mnemoc>
Turl: I'm setting up nginx's cache too
<steev>
RaYmAn: i'm far to busy to watch people twatting all day
tzafrir has quit [Ping timeout: 244 seconds]
tuliom has joined #arm-netbook
hno_ has quit [Quit: Lämnar]
j1nx_ has quit [Quit: Friends help you move. Real friends help you move bodies.]
<mnemoc>
Turl: not sure if it's the CDN or nginx's cache but load is falling now
<mnemoc>
spoke too soon... up at 11 again
<mnemoc>
12
<mnemoc>
13
<mnemoc>
narf
<mnemoc>
damn drupal sending everything with no-cache
pawel58701 has quit [Quit: Leaving.]
<mnemoc>
most of the IO comes from pg
<mnemoc>
which is kind of expectable... if nothing is been cached
<mnemoc>
narf
<mnemoc>
Turl: it seems you killed the drupal :p
<mnemoc>
Turl: nevermind, 500 is gone
<RaYmAn>
wow, scammers/spammers are getting really lazy
<RaYmAn>
"I will want to discuss something with you about your family and it has some
<RaYmAn>
form of relationship with you going by the similarity in the name."
<mnemoc>
o.o
<RaYmAn>
It's like they aren't even trying anymore :(
<Turl>
it's a wordpress mnemoc :P
<mnemoc>
RaYmAn: not nice that even them lost the hope...
<mnemoc>
Turl: uhm
<Turl>
mnemoc: in any case, all of the images and css and stuff is now being served by cloudflare
<mnemoc>
Turl: that's the easy part
<mnemoc>
Turl: even when the load was over 18 I got static content in less than a second
<Turl>
yeah, but as long as the site keeps sending a 1980 exp date it won't be cached :P
<Turl>
I would set up wp super cache but I need the wp credentials
<Turl>
load is ~3 now and going down
<mnemoc>
yes, looks pretty decent now
<mnemoc>
or maybe the "attack" finished
<Turl>
the wiki works normally
<Turl>
it's still on ycombinator home mnemoc
* mnemoc
didn't even know that site
<mnemoc>
I guess it's popular
<Turl>
same
<Turl>
I thought ycombinator was one of those 'angel investors'
tuliom has quit [Remote host closed the connection]