<drachensun>
leowt: I run it on most of my devices
<leowt>
drachensun: all at 1ghz?
<leowt>
all the time?
<drachensun>
generally yeah
Alex1269 has joined #arm-netbook
<leowt>
drachensun: no issues?
<drachensun>
I've noticed it will cause long g++ compiles to have errors
<drachensun>
like when building xbmc
<drachensun>
so I popped the case off and point a fan at the chip while I did that
<drachensun>
I think even ondemand would have had the same effect though as it was just an hour long high intensity operation
<drachensun>
for normal use though, no problems
<leowt>
drachensun: did it work normally with the fan?
<drachensun>
yup
<leowt>
drachensun: so, maybie a simple alu cooler on top of a10 is enough
<drachensun>
I would think so, an attached heat sink seems like it should work as well as a fan blowing from 5 inches away
<Turl>
drachensun: what's mean ambient temp over there?
<Turl>
just for reference
<drachensun>
Florida, USA so today, in January our high was around 80F
<drachensun>
I usually keep the AC in the house set at 78F
<Turl>
~26.5C, ~25.5C
<Turl>
pretty cool temps
<Turl>
I've ran on performance gov but I didn't do any long term stress testing
<drachensun>
yeah, I haven't tried that same test outside in summer when its 37C
<Turl>
:)
<leowt>
how do you set the cpufreq tweaks cript?
<leowt>
on boot?
<leowt>
or by hand?
<leowt>
i mean, manually
<Turl>
I run android usually so I just set it on the settings
<Turl>
when I'm using debian I add the echo commands to rc.local
<leowt>
dirty old rc.local
<leowt>
xD
<Turl>
it works
<Turl>
what else is there to say? :)
<leowt>
true =P
<leowt>
compiling k modules and installing packages stays fast
<leowt>
cool performance
tinti has quit [Quit: Leaving]
stefanro has joined #arm-netbook
stefanro1 has quit [Ping timeout: 252 seconds]
voronaam has quit [Quit: Leaving.]
leowt has quit [Quit: leowt]
L84Supper has quit [Ping timeout: 252 seconds]
L84Supper has joined #arm-netbook
TheLarch has joined #arm-netbook
TheLarch has quit [Changing host]
TheLarch has joined #arm-netbook
TheLarch has quit [Client Quit]
lerc has joined #arm-netbook
freakazoid0223 has joined #arm-netbook
wingrime_ has joined #arm-netbook
<drachensun>
techn: you mentioned cleaning up the ft5x_ts driver. I'm having to do that to, I needed to add some ioctls for screen rotation since it uses absolute position
<drachensun>
techn: Please let me know what all you had in mind, I started moving things to the header file that belong there but I didn't want to run into your changes if you are going to do the same thing
aholler_ has joined #arm-netbook
aholler has quit [Ping timeout: 248 seconds]
freakazoid0223 has quit [Ping timeout: 245 seconds]
Jef91 has quit [Remote host closed the connection]
Jef91 has joined #arm-netbook
Jef91 has quit [Remote host closed the connection]
Jef91 has joined #arm-netbook
Jef91 has quit [Remote host closed the connection]
Jef91 has joined #arm-netbook
Quarx has joined #arm-netbook
gimli has quit [Ping timeout: 245 seconds]
rz2k has joined #arm-netbook
pcat has quit [Ping timeout: 260 seconds]
ZaEarl has quit [Ping timeout: 248 seconds]
sv has quit [Changing host]
sv has joined #arm-netbook
RITRedbeard has quit [Quit: Leaving]
RITRedbeard has joined #arm-netbook
wingrime has quit [Ping timeout: 248 seconds]
sv is now known as discopig
rellla has joined #arm-netbook
discopig has quit [Quit: discopig]
Sv has joined #arm-netbook
Sv is now known as discopig
cheng has joined #arm-netbook
TestModule has joined #arm-netbook
Jef91 is now known as Jef91|AFK
pcat has joined #arm-netbook
merbzt has joined #arm-netbook
hansg has joined #arm-netbook
aholler_ is now known as aholler
Quarx has quit [Read error: No route to host]
Quarx has joined #arm-netbook
pcat has quit [Ping timeout: 256 seconds]
<buZz>
what 'video decoder core' is on those Rockchip devices like RK3066?
xman has joined #arm-netbook
<buZz>
sad that dx.com doesnt carry any RK3066 devices with 2GB ram :P
<br->
do those things work automatically? i almost bought one
<mnemoc>
with the preinstalled gpl-violating android, sure, they work out of the box
<mnemoc>
not sure how mature the open/build-your-own mk3066 support is already. but some sources were released recently
<br->
something i've wondered about regularly.. it should be generally possible to run a regular, say debian, linux userspace on an android kernel
<br->
and additionally, even without binary-only patches made by vendor, it should be possible to build a large subset of standard kernel modules against that kernel by ensuring various ABI parameters match.. right?
<br->
as in, just copy the android bzImage over to a vanilla linux system, not running it in a chroot on android or anything like that
<buZz>
br-: yeah, chrooted debian works on any android with root
<br->
yep.. but will android kernel happily start regular glibc+/sbin/init from say, debian. or is there some weird android shit going on during boot that would prevent that scenario
kaspter has quit [Ping timeout: 245 seconds]
<br->
ah, blergh. all the kernel<->user apis are different on android kernels. platform drivers aren't using ALSA etc
<mnemoc>
the only real weirdness in the "android" kernel (which is just a messed up linux) is paranoid network
popolon has joined #arm-netbook
<mnemoc>
you have to put every user which will need network access in a magic group id
<buZz>
there are some android specific things you should disable in the kernel
<buZz>
like android's logger
<br->
buZz: this is assuming the source is unavailable
<mnemoc>
i mean when running a bin-only gpl-violating stock android kernel. if you can disable stuff, just disable CONFIG_ANDROID
<buZz>
yeah
cheng has quit [Quit: Leaving]
merbzt has quit [Ping timeout: 252 seconds]
<slapin>
how can sound drivers not use alsa?
<slapin>
hi, all!
<buZz>
by using OSS?
<br->
slapin: ask the korean devs porting android to new devices
<br->
on android you just need audioflinger to understand how to talk to your device, no mention of alsa or oss
<br->
so they probably patch the shit out of that too, or something. never looked into it
tinti has joined #arm-netbook
<slapin>
br-: isn't it harder than just use alsa?
<br->
slapin: at least nexus one does not use alsa. perhaps alsa interface requires your driver provide some minimal functional guarantees that requires effort to implement
<mac->
hey
<mac->
it is possible to boot tablet from a SD Card with i.e. Linux ?
<rm>
yes
<rm>
if you mean an Allwinner CPU-based tablet
pcat has joined #arm-netbook
wingrime has joined #arm-netbook
<slapin>
br-: where can I see samples of such driver? I wonder, because otherwise there would no alsa drivers for these devices, and my older s3c6410 android-1.5 tablet uses alsa, at least it is there.
<slapin>
I can use aplay and arecord from chroot
<slapin>
and it has normal Linux + android on the same flash out of the box (/usr/bin is Linux, /system/bin is Android)
<slapin>
I wonder if it is possible to do the same with 4.x and Allwinner...
<slapin>
and most cool thing is getty on serial port out of the box, and accessible u-boot, too bad it has only 256MB of RAM, If install some older embedded Linux distro, like Angstrom with GPE, that would be plenty...
<slapin>
hno: when you will be available for some hacking?
<br->
slapin: on my nexus one, there are no ALSA char devices registered in /proc/devices
<br->
ergo, the phone cannot possibly be using ALSA
<br->
does your alsa-supporting device have a char device registered in /proc/devices with major 116?
ganbold_ has joined #arm-netbook
<slapin>
br-: I have /dev/snd* and aplay -l shows device, I don't have the device at the moment, so can't see /proc/devices. It is not phone, it is tablet.
<slapin>
br-: and why phone can't use alsa? my n900 thinks otherwise...
SouL_ has joined #arm-netbook
SouL_ has left #arm-netbook [#arm-netbook]
hipboi has quit [Quit: Leaving]
<br->
slapin: it's not about capability, its about design choices
<br->
if developer can avoid writing 10,000 lines of code just to satisfy some interface he doesn't care about, he might not use it. at least looks like that way on nexus one
dyoung-away is now known as dyoung
<br->
the android audio api is audioflinger, it's not alsa. maybe some implementations may use alsa underneath, but it's not required
<slapin>
but if your soc is supported in alsa (as with most SoCs except most modern ones from China) it is easier to just use alsa as it just works
<slapin>
I've tested my own android build now on Pandaboard - sound works with alsa.
merbanan has joined #arm-netbook
<slapin>
so only available APIs are ALSA and OSS, as i see.
<oliv3r>
hmm, i thought audioflinger was some device atop of alsa, just as pulseaudio is ontop of alsa aswell (alsa or oss-old/oss-new)
<oliv3r>
also, it would be quite some duplication effort, to have all drivers (oss/alsa) be ported to audioflinger, when it really is not needed. Alsa is a hardware audio drive framework; audioflinger is (i think) a sound mixer
<slapin>
audioflinger is like pulse audio - userspace libraries and daemons, see above URL.
<oliv3r>
yeah
<oliv3r>
the main reason for audioflinger to exist, is pulse was no wear near as complete as it is now
rellla2 has joined #arm-netbook
rellla has quit [Ping timeout: 248 seconds]
datagutt has joined #arm-netbook
rellla2 is now known as rellla
<br->
slapin: as mentioned on the url you pasted, the only requirement for the platform vendor is they implement AudioHardwareInterface. in fact it even mentions if you want to use ALSA/OSS, you stil have to implement that interface
<br->
considering some of these vendors are exposing /dev/ram chmod 777'd just to avoid writing code, its not hard to imagine some don't bother with alsa either
hansg has quit [Ping timeout: 260 seconds]
pcat has quit [Read error: Connection reset by peer]
rasusto has joined #arm-netbook
xman has quit [Remote host closed the connection]
pcat has joined #arm-netbook
hansg has joined #arm-netbook
leowt has joined #arm-netbook
<leowt>
helloz
dyoung is now known as dyoung-away
<slapin>
br-: it is extremely hard to program audio device via /dev/ram :) so writing kernel driver is unavoidable; so why not copy-paste some alsa driver instead of implementing something new? have you seen ASoC drivers? minimal driver is just a few lines of code, if your codec is supported.
<slapin>
AudioHardwareInterface is not kernel stuff, and you will have to write code to implement it. If you use alsa, you don't have tow rite any more code as it is there out of the box
<mnemoc>
slapin: mmap + .so saves them from violating the GPL
<slapin>
br-: for very new devices I've seen lame character device audio solutions, but these all were replaced with alsa drivers then, crappy though.
<slapin>
mnemoc: that is not worth writing, as code over /dev/ram should be very specific, not worth hiding either (hello, d-link).
Jef91|AFK is now known as Jef91
<slapin>
mnemoc: you can't DMA over /dev/ram, and no sound, sorry
<mnemoc>
good point
<hramrach>
why can't you DMA over /dev/ram
<hramrach>
when backed by disk you can DMA over any fd
<hramrach>
oliv3r: PA is nowhere near complete
<hramrach>
see how awesome it works when they tried to make a PA driver in Wine
<hramrach>
maybe just the docs are lame because the PA plugin for alsa works better than direct PA Winde driver
<slapin>
hramrach: works for me though
<hramrach>
PA works for you?
<hramrach>
it works for many people
<hramrach>
so does alsa
<hramrach>
it does not change that the interface is horrible, hard to use from both sides, and probably underdocumented
<hramrach>
for some simple cases it works
<hramrach>
for some cases when you want to control what it does - it just ignores what you tell it
<hramrach>
and stuff breaks
<slapin>
hramrach: that's with everything more complex than helloworld, shit happens, so someone needs to fix it.
<leowt>
can anyone give me a help putting uboot on nand?
<slapin>
btw, is there some central place for Atheros u-boot, or something? I've got lots of their funny hardware with funny u-boot versions (1.1.3 or something) of various malfunction state - on one device I can report device hang 50% times on u-boot countdown if console pins are connected (works if interrupted then "boot"'ed).
<leowt>
where do i start?
pcat has quit [Read error: Connection reset by peer]
jelly has joined #arm-netbook
<hramrach>
slapin: sure, stuff happens. you can at least make the interface sane, though.
<hramrach>
neither ALSA nor PA developers bothered with that, apparently
<hramrach>
leowt: does your nand not come with u-boot?
<hramrach>
and on what device?
<leowt>
yes it does, but not ready to boot from sda1
<leowt>
mele a1000
<hramrach>
is ther *any* u-boot ready to boot from sda1?
<hramrach>
first thing is to make full backup of the nand
<hramrach>
because there are some proprietary bootloader bits which may not be easy to replicate
<mnemoc>
boot0/boot1 is outside the logical nand space
<leowt>
hramrach: tnks 4 the hirony
<leowt>
1st i want to know if nand u-boot and the rest of files like boot.cmd and stuff, works like from SD
<leowt>
if not
<leowt>
where do i start
<leowt>
learning that
<slapin>
leowt: what do you need this u-boot for?
<slapin>
leowt: read Buildroot instructions on wiki
<leowt>
slapin: get rid of SD
<leowt>
and boot rootfs from sda
<slapin>
leowt: sunxi-current u-boot is not ready for that (yet) you need original one.
<slapin>
leowt: which sda? nanda maybe? not possible.
<hramrach>
not so sure
<leowt>
slapin: sata disk
ZaEarl has joined #arm-netbook
<hramrach>
maybe you could use the current u-boot but store the files on SATA - if it can read that
<slapin>
leowt: well, SPL is not nand-bootable yet.
<slapin>
hramrach: a problem is SPL
<hramrach>
no
<hramrach>
or depends on the way the stuff fits togethre
<hramrach>
what reads u-boot?
<mnemoc>
u-boot-sunxi's lichee-dev for uImage + initramfs from nand, and then root=/dev/sda1
<slapin>
leowt: if you can make original boot0 boot current u-boot, then it might work (if sata is supported in there, as I don't have any board to test).
<slapin>
leowt: USB works for me
<hramrach>
or you can make a stub rootfs on nand which just loads kernel from sata and kexecs
ganbold_ has quit [Remote host closed the connection]
<leowt>
i must not use usb or sd to boot
<leowt>
thats my goal
<leowt>
just nand + hdd
<leowt>
sata
<hramrach>
kexec is probably easiest way atm
<hramrach>
you can try to load current u-boot from the nand and have it read the rest from sata
<hramrach>
but I have no idea if any u-boot supports sata at all
* slapin
can't agree to hramrach view on ALSA interface quality, though
<hramrach>
slapin: I have talked to a few people who tried to use that
<slapin>
hramrach: I've used interface myself lots of times for various embedded hardware
lkcl_ has quit [Ping timeout: 276 seconds]
Jef91 has quit [Remote host closed the connection]
Jef91 has joined #arm-netbook
wingrime has quit [Ping timeout: 246 seconds]
rellla has quit [Ping timeout: 276 seconds]
<hno>
slapin, don't know. Don't expect much this or next week. Barely have working Internet at the moment (surf works, but not much else, damn corporate firewalls and broken hotel hotspots)
<slapin>
hno: can you read mail?
dyoung-away is now known as dyoung
<slapin>
hno: I've sent you email
dyoung is now known as dyoung-away
<hno>
slapin, when I manage to bypass the firewalls yes..
<hno>
slapin, I have two branches. sunxi-current is only forwarded. sunxi-patchqueue is the one that is rebased.
SouL_ has joined #arm-netbook
SouL_ has left #arm-netbook [#arm-netbook]
<slapin>
hno: do you keep these up to date?
Jef91 is now known as Jef91|Out
<hno>
slapin, pushing sunxi-current now.
<nemik>
so i booted into a linaro SD image after using the livesuite generated by ./scripts/mk_livesuit_image.sh with a linaro rootfs.
<slapin>
hno: and about patchqueue? because I want to base my test branch on it (autocreated).
<nemik>
it still has the same nanda/b/c/d/e/f partition structure as Android, not like linux. it also does not boot
pcat has joined #arm-netbook
<nemik>
the other very weird thing is that when i have "mem=448M.." or whatever in the CMDLINE part of the kernel config, the windows livesuit software rejects the image as invalid. when that param isn't present, it flashes the image fine(but of course does not boot).
Yaku has joined #arm-netbook
<hno>
slapin, sunxi-patchqueue pushed.
<hno>
hopefully it still builds.
<hno>
slapin, be warned that the NAND clock setting code do not exists in sunxi-patchqueue as no nand patch have been queued yet.
freakazoid0223 has joined #arm-netbook
<slapin>
hno: I will make NAND patch, thanks (but IIRC clock setting code belongs somewhere in CPU, no?)
<hno>
there is some NAND clock setting code in sunxi-current armv7/sunxi/clock.c
<hno>
but not in sunxi-patchqueue.
tinti has quit [Ping timeout: 248 seconds]
<slapin>
hno: ok, thanks a lot, that should save a lot of time.
TestModule has quit [Ping timeout: 245 seconds]
Quarx has quit [Read error: Connection reset by peer]
<techn_>
drachensun: Hi, I finished that patch redo yesterday.. but started "emulate cursor" patch
<techn_>
I was thinking to send them at same time
<techn_>
but sure.. I can send them separetly
tinti has joined #arm-netbook
<drachensun>
ok, I was just curious what all you were moving
<drachensun>
I was moving a bunch of defines over to the header as my latest project needs to call new ioctls on the driver
<drachensun>
and I started thinking that will probably make it a pain to merge depending on how big your overhaul is
<techn_>
not big.. since there is no code reordering.. just removed flags + some code.. + that earlier patch
<techn_>
but I'm not yet sure how much I have to do with pointer event emulation
<techn_>
Thinkgs will be clear during this evening
<drachensun>
I think I might have been wrong about the point thing
<drachensun>
I'm setting everything up now to have much larger icons
<drachensun>
and that seems to work pretty well
<drachensun>
and then the pointer doesn't hang out in your view
<drachensun>
for precise task it is still nice though
popolon has quit [Quit: Quitte]
servili007 has joined #arm-netbook
<nemik>
drachensun: do you have that script you wrote for installing linux to the internal NAND flash?
<nemik>
i don't see a pengpod image for internal, just ones for SD cards
lkcl_ has joined #arm-netbook
<drachensun>
its not online yet
<drachensun>
it isn't done yet
Gujs has quit [Ping timeout: 260 seconds]
vinifm has joined #arm-netbook
<nemik>
ok, thanks. if you have any partial work, i could try to help on it too.
gimli has joined #arm-netbook
toxicpsion has joined #arm-netbook
<servili007>
what's the proper way to add UVC camera support to android aside from adding support to the kernel?
Gujs has joined #arm-netbook
freakazoid0223 has quit [Read error: Connection reset by peer]
Yaku has quit [Ping timeout: 260 seconds]
Kerbonaut_Jarrod has joined #arm-netbook
<Kerbonaut_Jarrod>
hi all
<Kerbonaut_Jarrod>
I would want to discuss/know about: exynos 4 quad (4412 prime) & exynos 5 (5250)
Kerbonaut_Jarrod is now known as sky770
<sky770>
actually not able to decide between ODROID-U2 & ARMBRIX :|
* sky770
coughs up *ahem*ahem*
<ZaEarl>
you won't find much love for exynos in this channel.
<sky770>
^ aye :D
<sky770>
just curious :)
<specing>
You want whatever has SATA and Ethernet
<sky770>
a15 is fkin $$ ^ :(
<sky770>
a15 == exynos 5
<ZaEarl>
chromebook seems like a pretty good deal
aexl has joined #arm-netbook
<toxicpsion>
to boot fom nand i gotta use lichee uboot? right?
<mnemoc>
lichee-dev branch of u-boot-sunxi, yes
<mnemoc>
dropping it over nanda's linux/u-boot.bin
<nemik>
servili007: the only way i could think of is to use the USB host API in androi and re-writing all the linux kernel drivers to userland stuff with the NDK
voronaam has joined #arm-netbook
<toxicpsion>
mnemoc: thx. just wanted to make sure.
XenGi is now known as XenGi_
sky770 has quit [Ping timeout: 248 seconds]
sky770 has joined #arm-netbook
leowt has quit [Quit: leowt]
freakazoid0223 has joined #arm-netbook
<vinifm>
hi, A13 supports more than 512MB?
<mnemoc>
no
<mnemoc>
A13 is craplet-optimized
<voronaam>
is that 512MB of RAM?
<mnemoc>
yes...
aexl has quit [Quit: Page closed]
aexl has joined #arm-netbook
hansg has quit [Quit: Leaving]
Yaku has joined #arm-netbook
TestModule has joined #arm-netbook
TestModule has quit [Ping timeout: 245 seconds]
rz2k has quit []
sky770 has quit []
<servili007>
nemik: Well, when I say proper, I mean to the level of support found in stock roms and some mods
TestModule has joined #arm-netbook
<servili007>
so the work has mostly been done
<TestModule>
Is 3.4 supporting x11, hdmi, and usb-host at this moment?
<mnemoc>
yes
TestModule_ has joined #arm-netbook
TestModule has quit [Ping timeout: 245 seconds]
aexl has quit [Ping timeout: 245 seconds]
Alex1269 has quit [Ping timeout: 245 seconds]
rellla has joined #arm-netbook
dyoung-away is now known as dyoung
Jef91|Out is now known as Jef91
datagutt has quit [Quit: kthxbai]
dyoung is now known as dyoung-away
dyoung-away is now known as dyoung
<nemik>
servili007: ahh ok, sorry. i thought you meant something you could distribute to any phone without modifications to it
<drachensun>
mnemoc: Are you sure they A13 wont take 1G ram? I've got one sitting on a shelf here which is supposed to be A13 with 1G
<rm>
so what is the problem to turn it on and verify?
<mnemoc>
drachensun: I wouldn't put my hands on fire for it, but the A13's datasheet says "Memory Capacity up to 512MB
<mnemoc>
otoh the A10S (also sun5i) claims 2GB
<mnemoc>
there is an A12 too, but i've never seen any spec for it
hramrach has quit [Remote host closed the connection]
freakazoid0223 has quit [Quit: Leaving]
hramrach has joined #arm-netbook
Yaku has quit []
dyoung is now known as dyoung-away
rasusto has quit [Read error: Connection reset by peer]