Akaizen has quit [Remote host closed the connection]
Akaizen has joined #linux-sunxi
specing has quit [Ping timeout: 276 seconds]
specing has joined #linux-sunxi
popolon has quit [Quit: Quitte]
newleaves_jason has joined #linux-sunxi
<hno>
oliv3r, yes NAND boot uses the exact same boot header. u-boot SPL loads fine from NAND once you manage to get it there.
<hno>
but not much point without NAND driver in u-boot SPL.
<hno>
Regarding livesuit compatibility. Each of the may different NAND drivers we have are mutually exclusive and will wipe out the other. MTD, old Allwinner, new Allwinner. Simply beause they do not recognise the blocks as valid.
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
FreezingCold has quit [Ping timeout: 258 seconds]
FDCX_ has joined #linux-sunxi
newleaves_jason has quit [Quit: Leaving]
newleaves has joined #linux-sunxi
tomboy65 has joined #linux-sunxi
<Gerwin_J>
So strange that i can nothing found about allwinner on CES..
<Gerwin_J>
i want have more information about a80
tomboy64 has quit [Ping timeout: 252 seconds]
FreezingCold has joined #linux-sunxi
pirea has quit [Quit: pirea]
pwhalen has quit [Quit: Leaving]
pwhalen has joined #linux-sunxi
FDCX_ has quit [Ping timeout: 252 seconds]
andhe_ has joined #linux-sunxi
moofree_ has joined #linux-sunxi
hramrach has quit [Ping timeout: 240 seconds]
y0g1 has quit [Ping timeout: 272 seconds]
t3st3r has quit [Ping timeout: 240 seconds]
woprr has quit [Ping timeout: 272 seconds]
andhe has quit [Ping timeout: 272 seconds]
xenoxaos has quit [Ping timeout: 272 seconds]
moofree has quit [Ping timeout: 272 seconds]
bsdfox has quit [Ping timeout: 272 seconds]
maz has quit [Ping timeout: 272 seconds]
motley has quit [Remote host closed the connection]
bsdfox_ has joined #linux-sunxi
bsdfox_ is now known as bsdfox
woprr has joined #linux-sunxi
maz has joined #linux-sunxi
FDCX_ has joined #linux-sunxi
xenoxaos has joined #linux-sunxi
y0g1 has joined #linux-sunxi
xenoxaos has quit [Excess Flood]
<wens>
andhe_: did you see my latest reply on bluetooth in the ML?
xenoxaos has joined #linux-sunxi
motley has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
jinzo has quit [Quit: Leaving]
<wens>
I asked a co-worker to go by allwinner's booth at CES
<wens>
it was closed yesterday
<libv>
heh
<libv>
rumour has it that it is pvr, some magazines videoblog talked about that
<mrnuke>
either theyre busy designing the next awesome SoC, or they're busy fixing their damn bugs
<libv>
yay, u-boot from nand on uart
<libv>
mrnuke: they are preparing for yet another disastrous high end release because they used pvr again
<mrnuke>
WTF is pvr?
<libv>
powervr
<mrnuke>
is that also full of blobs to get any sort of performance out of it?
<libv>
heh, how on earth did hramrach get anything to boot with that howto he wrote
<libv>
at no point does it try to read a kernel from anywhere
<libv>
it's just not set up to do that
<libv>
it tries to read a whole partition to try to do that
<libv>
allwinner is on its own with that one, as far as it concerns me
newleaves has quit [Ping timeout: 252 seconds]
pwhalen has quit [Read error: Operation timed out]
newleaves has joined #linux-sunxi
<libv>
wow, segfault writing and then unmounting
pwhalen has joined #linux-sunxi
pwhalen has joined #linux-sunxi
pwhalen has quit [Changing host]
<libv>
this is really not too promising for it being intended to be a development board
<libv>
yay, read script.bin, but apparently it lost uImage along the way
<libv>
yay. kernel booted til rootfs
[7] has quit [Disconnected by services]
TheSeven has joined #linux-sunxi
HdkR_ has joined #linux-sunxi
HdkR has quit [Ping timeout: 252 seconds]
JohnDoe_71Rus has joined #linux-sunxi
pacopad_ has joined #linux-sunxi
<libv>
hrm, usb-otg also seems broken on a13, or there is another wiring problem with another inet device
<libv>
ah no, i was just using that.
<libv>
ah.
* libv
plugs in the cable
bamvor_ has joined #linux-sunxi
gzamboni_ has joined #linux-sunxi
xnxs has joined #linux-sunxi
specing_ has joined #linux-sunxi
JohnDoe_71Rus has quit [*.net *.split]
xenoxaos has quit [*.net *.split]
specing has quit [*.net *.split]
pacopad has quit [*.net *.split]
bamvor has quit [*.net *.split]
gzamboni has quit [*.net *.split]
specing_ is now known as specing
pacopad_ is now known as pacopad
mturquette has quit [Ping timeout: 246 seconds]
newleaves has quit [Quit: Leaving]
newleaves has joined #linux-sunxi
rz2k has quit []
pwhalen has quit [Ping timeout: 252 seconds]
pwhalen has joined #linux-sunxi
pwhalen has joined #linux-sunxi
pwhalen has quit [Changing host]
Black_Horseman has quit [Ping timeout: 252 seconds]
Gerwin_J has quit [Quit: Gerwin_J]
spv__ has quit [Quit: Ex-Chat]
Quarx has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
HeHoPMaJIeH has quit [Changing host]
HeHoPMaJIeH has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 260 seconds]
Md has joined #linux-sunxi
rellla has joined #linux-sunxi
<oliv3r>
nice, cubietruck uses 0.005 Amps when its not charging, but 'off'
<oliv3r>
libv: remember, abdroid kernels are in a single fs-less partition
panda84kde has joined #linux-sunxi
<oliv3r>
so hamrach's howto probably assumes you use an android kernel + initramfs 'image'
newleaves has quit [Quit: Leaving]
newleaves has joined #linux-sunxi
bbrezillon has joined #linux-sunxi
<ssvb>
oliv3r: have you tried to do some measurements for sunxi-3.4?
<ssvb>
oliv3r: I'm just curious if you can check/confirm the power consumption difference between idling on "ondemand" and "performance" governors
ZetaNeta has quit [Ping timeout: 252 seconds]
<ssvb>
oliv3r: please reply to the mailing list if you get some data
eyal has quit [Ping timeout: 260 seconds]
wolfy has joined #linux-sunxi
wolfy has left #linux-sunxi [#linux-sunxi]
mturquette has joined #linux-sunxi
<oliv3r>
AC100 probably is the 'versatile audio codec processor' AC, audio codec, or allwinner codec
<oliv3r>
and the audio codec processor may be so that the extra core inside the A80 can be in standby, yet still decode and play audio!
jemk has joined #linux-sunxi
<oliv3r>
ssvb: i'm writing hansg's fedora image right now to an usb stick
<oliv3r>
ssvb: i can tell you though that cubietruck uses < 0.001 amps when it's off and not charging though ;)
<ssvb>
oliv3r: sorry for nagging you, I just want to know if my current draw results are bullshit or not :)
<oliv3r>
ssvb: no nagging; only arrived at work an hour ago :p
<oliv3r>
and backlog reading
<oliv3r>
finding allwinner CES info
<oliv3r>
all very important stuffs
<oliv3r>
i'll start with hansg's image and his kernel; as that's probably a safe baseline
<ssvb>
trying to do the power consumption measurements with cpuburn-a7 would be also very interesting
<oliv3r>
yeah i'll try to download that
<oliv3r>
i doubt though that with his 3.4 kernel i'll get working wifi :(
<oliv3r>
so i'll see if i can get it transferd via lan
<oliv3r>
ssvb: cpuburn-a7 is precompiled in your repo?
<ssvb>
no, but if you have gcc, then it's very easy to compile
<oliv3r>
i don't know if there's a compiler in his sd image
<oliv3r>
and i'm @work
<oliv3r>
but i'll see if i can crosscompile it on my home pc
<ssvb>
it's something like "arm-linux-gnueabihf-gcc -o cpuburn-a7 cpuburn-a7.S" with a crosscompiler
<oliv3r>
aww cute kitty!
ssvb has quit [Remote host closed the connection]
<oliv3r>
the kitty scared him!
<ccaione>
oliv3r: they said powervr in a80
<oliv3r>
ccaione: yeah :( I cried a little
_massi_ has joined #linux-sunxi
<andhe_>
wens: you mention you fixed a pin muxing error in bt.... could you point out what you fixed so I can learn? I've stared at the docs and the dt alot lately but didn't spot any error, but I'm just trying to learn dt.
<wens>
well, the driver core will use the "default" state pin mux for devices
<wens>
you name it "default" with pinctrl-names in the DT
bbrezillon has quit [Read error: Operation timed out]
<wens>
if you have multiple pinctrl-N entries, you need multiple name entries to match
<wens>
I didn't know this, so the external clock output pin was never muxed
<andhe_>
ok, so if I understand you correctly, the fix was on the line:
<andhe_>
pinctrl-names = "default", "default";
<wens>
correct
ykchavan has joined #linux-sunxi
<andhe_>
(don't have my/your old tree available at the moment)
<andhe_>
thanks
<wens>
I don't either :p
<andhe_>
I still think dt contains a bit too much magic for me to really understand it.
<wens>
you need the manual to match the pins
<wens>
and interrupts
<andhe_>
I've looked alot at the pins and how they are described in dt... I think I've atleast understood that part...
<andhe_>
(compared your dt changes to the a20 user manual)
tomboy65 is now known as tomboy64
<andhe_>
took me a bit to understand "7 18" means PH18 ..... (I think your last patches didn't have the comments for that)
<andhe_>
anyway, nice to hear you've found a way forward .... I guess I'll continue digging into how to get it working on 3.4, which might benefit someone....
<wens>
bluetooth?
<wens>
or wifi?
<andhe_>
I guess bt
<wens>
bt should be quite straight forward
<andhe_>
I looked at a patch in the cubieboard/linux-sunxi tree that I think is needed for the rfkill bits in cubietruck for bt
<wens>
i thought the bits were already in there?
<andhe_>
(yesterday) then I fell asleep....
<wens>
anyway i have to get my stmmac patches out first
<andhe_>
wens: I don't know, don't trust anything I say... but sunxi-rfkill only seems to handle the case of (external?) sdio devices....
<andhe_>
I _guess_ that part is still needed in linux-sunxi 3.4 tree.... will try tonight.
notmart has joined #linux-sunxi
notmart has joined #linux-sunxi
notmart has quit [Changing host]
<andhe_>
anyway, don't let me distract you too much from your mainlining effort. :)
shineworld has joined #linux-sunxi
ykchavan has quit [Ping timeout: 264 seconds]
popolon has joined #linux-sunxi
popolon has joined #linux-sunxi
<oliv3r>
oh wow, writing an SD card at 237kb/s is extremly slow :S
shineworld has quit [Quit: Leaving]
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
ykchavan has joined #linux-sunxi
arokux has joined #linux-sunxi
ZetaNeta has joined #linux-sunxi
ZetaNeta has quit [Changing host]
ZetaNeta has joined #linux-sunxi
Night-Shade has quit [Ping timeout: 252 seconds]
LoCoZeNoz_ZUE has joined #linux-sunxi
ZetaNeta has quit [Read error: Connection reset by peer]
JohnDoe71rus has joined #linux-sunxi
FR^2 has joined #linux-sunxi
bfree has joined #linux-sunxi
Night-Shade has joined #linux-sunxi
<oliv3r>
what's the difference between sunxi-next and sunxi-devel?
<oliv3r>
next being what has been merged; devel what could be merged but it WiP?
<ccaione>
mripard: given the latest email I think I'll refactor the driver to make it compatible with A31 also
HdkR_ has quit [Read error: Connection timed out]
<andhe_>
oliv3r: AIUI next is what's sitting in different maintainers repos and will go to torvalds when he opens the next merge window.... devel is wip/rebased branch with useful but not yet ready/clean stuff.
HdkR has joined #linux-sunxi
<oliv3r>
thought so ;) just wanted to know which branch had my latest patches ;)
<andhe_>
not everyone seems to be so strict about their naming though, so I guess it's different from person to person....
Seppoz_nowork is now known as Seppoz
Seppoz has quit [Changing host]
Seppoz has joined #linux-sunxi
<oliv3r>
it makees sense, naming wise, but i thought we'd only have 1 tree, maybe with 2 branches
<oliv3r>
but it's fine this way i suppose
<oliv3r>
we also have experimenta/sunxi-3.10 to spice it up
hero100 has quit [Remote host closed the connection]
<mnemoc>
oliv3r: hi, when you have a minute, can you take a look into the soc-detect and script.bin stuff I made for 3.10? in wip/sunxi-3.10/* of my github? at least to decide the location of the files :\
<oliv3r>
i will do it now
Night-Shade has quit [Read error: Operation timed out]
<oliv3r>
i'm waiting for the FEodra 19 image to get written to sd card so have a few
<mnemoc>
/script parses script.bin and it's going to spawn the missing devices or inject them into the device-tree
HdkR has quit [Remote host closed the connection]
<oliv3r>
sweet
HdkR has joined #linux-sunxi
<oliv3r>
you took your old soc_detect code and then patched it later on right?
<oliv3r>
debating if i should comment on the first patch; and see if it is fixed later or not
<mnemoc>
final state please
<oliv3r>
pff, then i need a squashed patch to review on! :p
<oliv3r>
lets see if i can do that
lynxis has joined #linux-sunxi
<mnemoc>
diff against experimental-3.10 for now
ccaione has quit [Ping timeout: 245 seconds]
ccaione has joined #linux-sunxi
<oliv3r>
you've updated experimental right? so the base is the same
<mnemoc>
yes
<oliv3r>
we have two experimental branches? wip/exp and exp?
<mnemoc>
exp in the official repo
<mnemoc>
and reference-3.10 there too
<oliv3r>
yeah i noticed we have 2 3.10 experimental branches; i'll go with the newest one
hero100 has joined #linux-sunxi
<mnemoc>
oliv3r: it would be nice of we could give a "generic" 3.10 and "generic" dts to people with new devices and be able to do research with it
<mnemoc>
oliv3r: not assuming that much
<oliv3r>
yeah safe bet
<mnemoc>
oliv3r: also comparing the detected soc with what's provided by DT and warning if they defer
<oliv3r>
i would love to see this soc_detect work show up in /proc/cpuinfo, it doesn't do that right now does it
<mnemoc>
differ*
<mnemoc>
u32 system_rev allows it
<oliv3r>
1632 etc
<mnemoc>
sadly that only allows the sunNi family
<mnemoc>
but we can encode it decently
<mnemoc>
CCXY
<mnemoc>
CC=32
<mnemoc>
X for the SoC
<mnemoc>
Y for the rev
rz2k has joined #linux-sunxi
<oliv3r>
mnemoc: probably; but yeah that's where I imagine it would go
<mnemoc>
and "agreeing" in CCX mapping to A??
<mnemoc>
"A??"
<mnemoc>
babelfish should probably also use soc-detect instead of trusting the mach-id from the bootloader, which is usually BS
Night-Shade has joined #linux-sunxi
hipboi has quit [Remote host closed the connection]
<mnemoc>
as allwinner's kernel ignores the machine id provided and sets a CONFIG_CHIP_ID there
agath_at_work_pd is now known as agath_at_work
<mnemoc>
allwinner's bootloader doesn't pass the meminfo either, so babelfish should query the dramc
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
Black_Horseman has joined #linux-sunxi
deasy has joined #linux-sunxi
FunkyPenguin has quit [Ping timeout: 246 seconds]
FunkyPenguin has joined #linux-sunxi
FunkyPenguin has quit [Excess Flood]
FunkyPenguin has joined #linux-sunxi
<oliv3r>
i think mainline kernel currently completly relies on the DT doesn' tit
FunkyPenguin has quit [Excess Flood]
<oliv3r>
the question then arises, is it the kernels responsibility to find mistakes int he dt
<oliv3r>
for our 3.10 absolutly, as script.bin's tend to be bs often too
FunkyPenguin has joined #linux-sunxi
<oliv3r>
hmm, 3.3 kernel gives me all sorts of ata errors; i hope those are all fixed with later versions, otherwise either my disk is broken (quite likly) or the PHY is not right
<oliv3r>
it usually points strongly toward a hardware problem (often a bad SATA cable or a bad or inadequate power supply).
<oliv3r>
heh
<oliv3r>
voltage is dropping significantly
<oliv3r>
so maybe that's why
<oliv3r>
4.8V may be not enough; i assume the sata power mapped from USB port
<oliv3r>
0.4 amps and crash!
Xaros has joined #linux-sunxi
pwhalen has quit [Ping timeout: 260 seconds]
<oliv3r>
writing to the ssd, 0.62 amps @ 5.1V
Xaros has quit [Client Quit]
Xaros has joined #linux-sunxi
Xaros has quit [Remote host closed the connection]
<mnemoc>
oliv3r: I believe 3.10 should be smarter and rely the lesser possible on the DT and script.bin
<mnemoc>
oliv3r: been eventually able to have a multi-sunxi kernel
<mnemoc>
and use that as research tool to add proper mainline support for unknown boards/chips
<oliv3r>
we also need some more unification i noticed, now we have compile time dependancies; single kernel is os nice
<oliv3r>
but i agree
Xaros has joined #linux-sunxi
rz2k has quit []
<mnemoc>
oliv3r: on 3.10 we also need to be able to just import new crap drivers from new SDKs for new SoCs, so and that probably means a driver providing glue symbols to be able to get started quickly
FunkyPenguin has quit [Ping timeout: 260 seconds]
<mnemoc>
once there and running, refactoring, cleanup and mainlinization beging
<mnemoc>
begins*
FunkyPenguin has joined #linux-sunxi
<oliv3r>
it's probably more parallel then that :p
<oliv3r>
half the drivers are mainlined allready :p
<oliv3r>
i expect that disp, audio will need the refactoring etc
ykchavan has quit [Ping timeout: 264 seconds]
Xaros has quit [Quit: Zwi se logou mou!!!]
<rellla>
how does fb0_framebuffer_num correlate with libvdpau_sunxi? is "2" sufficient for displaying video + osd?
<oliv3r>
jemk: ^
<jemk>
rellla: it should be irrelevant, at least I thought so. turbofb has some needs there, but vdpau allocates all memory from ve reserved area.
<rellla>
video is ok. using osd with output_surface isn't failing, but not displaying osd, using bitmap_surface for osd stops VDR itself. maybe some implementations are missing though.
<mnemoc>
oliv3r: I was thinking in A23, A43, A70, A80, A90, ...
<mnemoc>
oliv3r: but yes, backport them!
<oliv3r>
mnemoc: hopefully, those will mostly use the same IP
<oliv3r>
a80 will be spat uppon
<oliv3r>
powervr, pfew!
Xaros has joined #linux-sunxi
Xaros has quit [Client Quit]
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
Black_Horseman has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
LoCoZeNoz_ZUE has quit [Ping timeout: 260 seconds]
<jemk>
rellla: i didn't check the bitmap surface things, i even stopped all osd related work at the moment, too little time
<rellla>
jemk: i know. therfore pkendall's tree ;)
sgo11 has joined #linux-sunxi
<jemk>
but all this osd stuff is very hacky, at least my part makes a lot of assumptions about things it shouldn't do
<sgo11>
hi, I am new to cubieboard. I am trying to install ubuntu to my new cubietruck. by reading this link: http://docs.cubieboard.org/tutorials/a20-cubietruck_lubuntu_desktop_releases, I am very confused. it says "ct_lubuntu_desktop_12.10_20130909.img.gz and ct_lubuntu_desktop_12.10_card0_20130909.img.gz are built without sound modules. but ct_lubuntu_desktop_12.10_20130910.img.gz is ok". but where is ct_lubuntu_desktop_12.10_20130910.img.gz?
JohnDoe71rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<oliv3r>
libv: just found that patch it said 'sunxi doesn't support mode = 4
<libv>
i was using vga on disp on A10 just a few weeks
bbrezillon has joined #linux-sunxi
<oliv3r>
cool
<juanfont>
hi. i'm using a wit m2 dev kit (a20 powered). after 10 minutes or so the display goes black. i've already added consoleblank=0 to uEnv.txt. any other idea?
<oliv3r>
er, probably some standard term evn var to prevent that?
<libv>
juanfont: smells like console blanking
<sgo11>
hi, i am reading http://goo.gl/iWPFLK . I am not sure if this installation method will provide sound or not. in the end, it says "If we want to change to VGA output default, we need to modify the script.bin at the 1st partition". I never use arm board before. this sentence confuses me. do I have to choose VGA or HDMI? can't I use VGA and HDMI? I think if I connect vga, then vga should work. if I connect hdmi, hdmi will work. why should I choose one as de
<sgo11>
fault? thanks.
jemk has quit [Ping timeout: 272 seconds]
<sgo11>
why should I choose one as default. thanks.
<oliv3r>
sgo11: i don't think our disp code does auto detection atm
<libv>
Turl: but it still needs a little nudge, which i will go into next week if it hasn't improved fully
<libv>
oliv3r: and even then...
<libv>
oliv3r: no ddc on vga on existing hw
<oliv3r>
yep; ontop fo that
<oliv3r>
so sgo11 no, auto detect does not work/exist on sunxi
<juanfont>
oliv3r, libv, the thing is that when it goes blank (both in X or in terminal) i have to reboot it to get it back (ssh still works)
<libv>
kms driver does do load detection, but that might be unreliable if the hw doesn't have vga
juanfont has left #linux-sunxi ["Saliendo"]
juanfont has joined #linux-sunxi
<libv>
juanfont: disp isn't the nicest bit of code
<sgo11>
oliv3r, really? ... I was so naive. ARM world is very different. I just can't believe it.
<oliv3r>
sgo11: not so much arm world, more cheap chinese SoC world
<oliv3r>
sgo11: and horrible chinese drivers
<libv>
sgo11: PC world <-> embedded
<oliv3r>
libv: well with the KMS, you could check for ddc, and revert to manual options if missing though?
<libv>
oliv3r: i am wrong though, cubieboard does ddc
Gerwin_J has joined #linux-sunxi
<oliv3r>
libv: cubietruck?
<libv>
oliv3r: i am going to require proper layout info for the vga connector
<libv>
ah, yes, i do need caffeine
<libv>
for device tree
<libv>
(the layout info)
<oliv3r>
libv: well for cubie* we have schematics at the least
<oliv3r>
benn might be willing to give you the VGA section of the board layout if that's needed
<libv>
script.bin does not describe which ddc lines are wired to the connector, nor does it describe which h/vsync lines
<sgo11>
oliv3r, thanks. this is just unbelievable! I cannot believe I have to compile the code if I change my monitor from VGA to HDMI or vice versa.
<libv>
oliv3r: cubie is documented, and tom cubie helped me with finding out that the mele's blue connector and extra wires simply go dead on the board (after the on board connector)
<libv>
sgo11: welcome to the embedded world
Black_Horseman has quit [Remote host closed the connection]
<libv>
underdeveloped, backwards and secretive
<libv>
sgo11: be glad that we are here, and that you happened to have chosen sunxi hw
<oliv3r>
well booting fedora 19 on CT and vga worked for the console
<oliv3r>
lets see if it comes up with X
<sgo11>
libv, oliv3r thanks for the help. I am sure I will have many questions.
<oliv3r>
libv: should be tom radxa now :p
<libv>
sgo11: our wiki answer a great many of them
<libv>
oliv3r: in case you didn't read the backlog... i am running off of nand fully now
<libv>
no normal human being will manage that, even if it is properly documented
<oliv3r>
libv: oh really; i must have missed that; i did see you got u-boot onto the nand
<oliv3r>
libv: yeah it's not inteded for normal humans :p
<oliv3r>
libv: btw, the kernel booting image, that's the android kernel; 1 partition for kernel + initramfs without fs
<libv>
i will split out the changes needed to u-boot and send them in
<oliv3r>
libv: and hno said, aw 3.11, 4.11 and mtd are mutually exclusive
<libv>
yeah
<oliv3r>
yay, i see a mouse cursor!
<oliv3r>
great, now i have a platform to work on :D
<sgo11>
oliv3r, wait a second. I tried the build-in android. it can auto-detect vga or hdmi. why not linux? it should be the same.
<libv>
i hope i spurred enough people on to actually fix the mtd stuff
<oliv3r>
on my tablets, brining up fedora is easy; it 'just works'
<libv>
sgo11: can it autodetect it?
<libv>
sgo11: which hw is this again?
<oliv3r>
libv: not just you ;)
<oliv3r>
libv: someone even offered sponsorship for mtd developer
<sgo11>
libv, yeah. cubietruck. I used both vga and hdmi yesterday.
<sgo11>
vga to my monitor and hdmi to my tv. both works. why not linux?
<oliv3r>
sgo11: simultaniously?
<libv>
sgo11: because we have code we have been working on for years, which supports multiple SoCs
<oliv3r>
i expect that the default config sets up dualhaed
<libv>
sgo11: not something some vendor quickly threw together from bits and bobs provided by allwinner
<sgo11>
oliv3r, not at the same time.
<libv>
cubietruck has ddc
<oliv3r>
disp driver should be the same though
<oliv3r>
sgo11: try it at the same time, i'm curious if it is setup as dualhead (mirror) by default
<libv>
all of that should get fixed after fosdem
<oliv3r>
yeah we have a big 'todo' list
<sgo11>
oliv3r, back to my original question, http://goo.gl/iWPFLK says I have to compile the code to set VGA as default. does it mean linux can not auto-detect vga or hdmi for cubietruck?
Seppoz has quit [Remote host closed the connection]
Seppoz has joined #linux-sunxi
<oliv3r>
sgo11: you have to edit a 'config' file, called script.bin
<oliv3r>
sgo11: our wiki should explain everything :p
<oliv3r>
sgo11: script.bin gets decompiled with 'bin2fex' and you can recompile it with fex2bin
<sgo11>
oliv3r, sorry, that is not my question. I can follow that guide. but android can autodetect vga or hdmi. can linux do that for cubietruck? thanks.
<oliv3r>
libv: sunxi room takes 800 seats; what will we do with the 795 seats that are left?
<oliv3r>
sgo11: no.
<libv>
sgo11: either it worked out of the box, or it didn't
<libv>
sgo11: that's what autodetection is about.
<sgo11>
why not? if android can.
<libv>
sgo11: the source code is available, good luck with fixing it
cajg has quit [Ping timeout: 264 seconds]
<libv>
sgo11: because it hates you?
<sgo11>
i don't have any experiences on kernel and driver development.
<oliv3r>
sgo11: our current linux disp driver can't do it;
<oliv3r>
sgo11: patches welcome
<libv>
more likely, our current linux disp does not have any cubietruck specific changes
<oliv3r>
i think default ct runs 3.3 kernel or something ancient
<oliv3r>
i still think they do some trickery; like vga enabled by default, and if hdmi is plugged in, disable vga
<oliv3r>
but no clue what they changed etc
<libv>
sgo11: in a very nearby future, things will be better.
<libv>
as i already have vga and hdmi wired up in my kms driver
<sgo11>
thanks.
<libv>
both lack ddc still, hdmi is up next, once i have sent out my nand patches and my nand capable u-boot patches, and documented the process, i hopefully should be able to finally work on hdmi i2c again
<sgo11>
is rootfs-xxx.tar.gz the actual ubuntu release? thanks.
<oliv3r>
sgo11: no, it's crafted by smart hackers to be as good as possilbe
<oliv3r>
sgo11: but since the guys behind cubieboard are behind it, all these questions are better handled there. we worry mostly about u-boot + kernel
<arete74>
hi, in mainline the option for enable sunxi processor is "Allwinner A1X SOCs" , is more correct "Allwinner SUNXI SOCs"
<sgo11>
oliv3r, ok. thanks a lot for the explanation even if I am still confused. :)
<oliv3r>
arete74: yeah we should really rename that in one of the next patches, AXX sounds stupid too
<arete74>
i can send patch?
<oliv3r>
arete74: for mainline? of course, dunno if SUNXI is the best term, sunxi might be better
<oliv3r>
mripard: ^
nove has joined #linux-sunxi
<arete74>
oliv3r, send to linux-sunxi and lkaml?
<oliv3r>
and cc mripard etc
<arete74>
oliv3r, etc is for ?
<oliv3r>
et cetera?
cajg has joined #linux-sunxi
<sgo11>
another question. before buying cubietruck, I asked in #ubuntu-arm. people there told me if I have an old ubuntu version in arm, I can simply do "do-release-upgrade" to update to the latest build. is this true in cubietruck? I can follow the tutorial and install ubuntu 12.10 and then do-release-upgrade to 13.10? thanks.
<oliv3r>
sgo11: again #cubieboard; but from what I expect, they used a fairly recent ubuntu, added packges for our kernels and our modules; and left the rest 'pristine'
<oliv3r>
so i assume apt-get update/upgrade; do-releaease-update all should work well
<oliv3r>
i assume you are running from an SD card; so "whats the worst that can happen" :)
<oliv3r>
flash a new sd card :)
<sgo11>
oliv3r, yeah, I will run in microsd. thanks. hopefully, I can install lubuntu 12.10 successfully first. thanks.
jemk has joined #linux-sunxi
<oliv3r>
ohh the irony
<oliv3r>
i have told people a few times what could potentially fix the load 1.0+ thing; and now i run into it and don't remember
<libv>
"load 1.0+" ?
<libv>
oliv3r: if only you had wikied it :p
<oliv3r>
load is always 1.00 when idle
<oliv3r>
i do recall it being a script.bin issue; and i think it's the og
<libv>
btw, sun4/5i_defconfig doesn't have cpufreq
<libv>
which is why i ran into that yesterday
<oliv3r>
ah!
<oliv3r>
why not?! strange
<libv>
same for g_ether
<libv>
which imho is a pretty important one if you are working with tablets
<libv>
it isn't in u-boot, but it is for the proper os
<libv>
it gives me networking on the go, as wifi point-to-point is flaky on the ICE
<libv>
(german high speed rail)
<libv>
and once you have set it up, you just use it all the time
<libv>
just stick in the cable, done.
<oliv3r>
oh nice
<oliv3r>
ICE has ethernet?
<libv>
g_ether is ethernet over usb
<libv>
gives you like 1-2MBps on NFS
<oliv3r>
usb otg btw causes the high cpu load; it probbably does an IDLE poll consuming 100% cpu
<oliv3r>
but you don't see it in top as a process, just the load avg 1.00
<libv>
who has been working upstream usb support?
<libv>
and did he at all care about OTG?
<oliv3r>
usb != otg; we have seperate controlle
<oliv3r>
otg is mentor graphics Inventra IP
<libv>
great. more upstream and development board shortsightedness hitting us
<oliv3r>
well the loadavg is an old bug; from the AW kernels
hani has joined #linux-sunxi
<oliv3r>
but luckly, the mentor graphics IP is mainlined and wens will work on adapting it for sunxi
<oliv3r>
then backporting should be easy
<libv>
ok, so wens is working on it
pwhalen has quit [Ping timeout: 252 seconds]
<libv>
wens: OTG is important if we do not want to fully limit ourselves to development boards
<oliv3r>
aw took the inventra driver, or some reference driver, and mutalated it
<oliv3r>
(added gpio support for the detect pin)
<wens>
just started looking at it
<wens>
have to post my other patches first
<libv>
wens: good, thanks :)
<wens>
ccaione: nice drawing
hani has quit [Ping timeout: 252 seconds]
<ccaione>
wens: mripard's :)
pwhalen has joined #linux-sunxi
hansg has joined #linux-sunxi
<ccaione>
arete74: your email is empty
hani has joined #linux-sunxi
<nove>
oliv3r: isn't olimex in fosdem will have stand and sell boards, ask them to put banner to advertise the sunxi talk
<libv>
ah
<libv>
oliv3r: don't worry
<libv>
oliv3r: it will be ok, as arm is hot and this is the only interesting ARM talk
hani has quit [Remote host closed the connection]
<oliv3r>
crap i put my cpufreq on 4 GHz and i dunno what it did, but it crashed :)
<oliv3r>
it's still drawing power, but not responding to input anymore
<oliv3r>
nove: i don't think he intends to bring lots of hardware
* ccaione
wants tshirts
Gerwin_J has quit [Quit: Gerwin_J]
hero100 has quit [Quit: Leaving]
<nove>
oliv3r: when someone ask olimex what is the level of support for the boards with sunxi, they could refer the existence of the talk, this sort of things (just saying, in the case this little things doesn't come to mind)
<oliv3r>
hold the press, change all engines; default governor for sunxi should be performance.
<oliv3r>
nove: i'm pretty sure they will tbh
<oliv3r>
don't think the governor will influence mali much
<oliv3r>
anybody familiar with yum, fedora and ideally with hansg's image?/
<oliv3r>
i'm trying to install something via yum through a proxy, but i'm not sure if yum likes my proxy settings
<hansg>
oliv3r, is your date correct ?
<hansg>
If your date is of the https cert validation will fail
<oliv3r>
hansg: there's no proxy setting in the GUI installer is there? i looked but couldn't find it
<ccaione>
hansg: the patch for changing the trigger type has been submitted to mainline?
<hansg>
oliv3r, no not that I know of
sgo11 has quit [Quit: Leaving]
<hansg>
ccaione, yes
<hansg>
I think it has already gone into some stable releases too
<oliv3r>
hansg: i guess i can't use ntpdate via a proxy either :(
<ccaione>
hum, I based the code on mripard's sunxi-next
<mripard>
ccaione: hmm, I might not have updated sunxi-next lately then :(
<ccaione>
:) ok, time to switch to mainline then
<oliv3r>
ccaione: your rtc driver will let me save the clock from the system right? hwclock --systohc right?
<ccaione>
oliv3r: don't remember the syntax but yes
<oliv3r>
ccaione: just something that came up in my mind
<oliv3r>
hansg: date fixes it :)
<oliv3r>
well i got a basic sd card setup; now time to setup all this via an ssd :D
<Turl>
morning
<ccaione>
moin Turl
<oliv3r>
hansg: i assume the image will work just as well on an sdd as it did on the mmc?
<oliv3r>
Turl: mornin'
<WarheadsSE>
since I have many of you active .. A20, best branch for up'ing the package in Arch?
<hansg>
oliv3r, it does for me, after just cp -pr from mmc to ssd, and update /etc/fstab a the / fs uuid will change
<hansg>
oliv3r, and you need to relabel selinux attributes on reboot, do touch .autorelabel in the new root before rebooting
<hansg>
(note the relabel is going to take a while)
<oliv3r>
hansg: oh i simply xzcat the image to the sdd
<oliv3r>
wont' that work? the auto resize etc?
nikrou has joined #linux-sunxi
<nikrou>
Hi all
<oliv3r>
hansg: obviously i'll boot mmc do a root=/ change
<nikrou>
I've got an a13 devices semi-bricked after installing CWM
<oliv3r>
nikrou: a13 is unbrickable
<nikrou>
I tried to update it through fel mode
<hansg>
oliv3r, you will still need an mmc to boot, with a root=/dev/sda3 argument, then what you suggest should work I guess. I've never tried. I move rootfs from one disk to another semi regularly so I just did what I always do
<nikrou>
i'm on debian
<oliv3r>
nikrou: livesuit only works from windows
<nikrou>
sure ?
<nikrou>
nobody succeeded on linux ?
<Turl>
oliv3r: there's also a linux version, but it's a bit of a pain I hear
<arokux>
ccaione: hi, you were n01 earlier, weren't you? :)
<ccaione>
arokux: yep
<arokux>
ccaione: are you still working on axp?
<ccaione>
yep
<arokux>
cool
<ccaione>
while(1) printf("yep\n");
<wens>
arokux: you're back
<arokux>
yep :)
<oliv3r>
yay
* arokux
hugs oliv3r
<WarheadsSE>
arokux: hey, there he be
<arokux>
yep, I was on vacation and although I had two sunxi boards with me it was difficult to force myself to work on something...
<arokux>
WarheadsSE: you want to update the kernel for A20?
<WarheadsSE>
That'd be nice.. yeah
<WarheadsSE>
Looking at which tree/branch is best option for all peripheral support at this time
<WarheadsSE>
the other sunxi kernels could use a bump too
<arokux>
sunxi-3.4 was always *the* branch, but I wasn't around for some time.
<nikrou>
I cannot compile awutils
<nikrou>
many errors
<WarheadsSE>
arokux: there we some issues with the A20 in the past, IIRC.
<WarheadsSE>
I know that there has also been some decent progress towards mainline
<nikrou>
Read to fast
<arokux>
yes, thanks to hansg who has gathered all our hacks into one branch
<nikrou>
I already have boot.img, recovery.img, system.img and bootloader.img
hansg has quit [Quit: Leaving]
* nikrou
is sorry for noise
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 272 seconds]
pacopad has joined #linux-sunxi
<arokux>
wens: is everything sorted with gmac?
jemk has quit [Ping timeout: 272 seconds]
pacopad_ has quit [Ping timeout: 272 seconds]
pacopad_ has joined #linux-sunxi
<wens>
arokux: cleaning and rebasing to do a final test
pacopad has quit [Ping timeout: 264 seconds]
<arokux>
wens: I see
<arokux>
hm.. there are some patches for NAND against mainline. what are they? somebody is working towards adding full support of the NAND controller to mainline? o_O
pacopad_ has quit [Ping timeout: 264 seconds]
pacopad has joined #linux-sunxi
<wens>
only basic support atm
<arokux>
good to know
jemk has joined #linux-sunxi
pacopad has quit [Ping timeout: 260 seconds]
pacopad has joined #linux-sunxi
<libv>
i got annoyed by the android figurine that is being shown when booting from nand
<cajg>
Well, sort of. But I would be very grateful if anyone can explain what this means for me as a user of A10 devices
<cajg>
and as a patcher and compiler of kernels for them
pacopad_ has quit [Ping timeout: 272 seconds]
<cajg>
in my neophyte imagination, this is for recent kernels and I need not woryy for a while, if ever
pacopad__ has joined #linux-sunxi
<buZz>
i bet coreboot will just be a replacement for uboot
<cajg>
hmm, but uboot is already open, I thought, so I can't see great advantage
<buZz>
i think coreboot's only goal is booting the kernel faster after powerup
pacopad_ has joined #linux-sunxi
<cajg>
oic, I only reboot my MK802 about twice a year
pacopad has quit [Ping timeout: 272 seconds]
pacopad_ is now known as pacopad
mkutsevol has joined #linux-sunxi
pacopad__ has quit [Ping timeout: 272 seconds]
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 272 seconds]
bbrezillon has joined #linux-sunxi
bbrezillon has quit [Client Quit]
Quarx has quit []
pacopad has joined #linux-sunxi
pacopad_ has quit [Ping timeout: 272 seconds]
pacopad_ has joined #linux-sunxi
pacopad__ has joined #linux-sunxi
pacopad has quit [Ping timeout: 264 seconds]
pacopad__ is now known as pacopad
notmart has quit [Quit: notmart terminated!]
pacopad_ has quit [Ping timeout: 252 seconds]
popolon has quit [Quit: Quitte]
pacopad has quit [Ping timeout: 252 seconds]
pacopad has joined #linux-sunxi
FR^2 has quit [Quit: Connection reset by peer]
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 272 seconds]
pacopad_ is now known as pacopad
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
pacopad_ has joined #linux-sunxi
pacopad has quit [Read error: Operation timed out]
pacopad_ is now known as pacopad
panda84kde has quit [Quit: Konversation terminated!]
<oliv3r>
ccaione: nope, nothing yet
Gerwin_J has joined #linux-sunxi
<oliv3r>
libv: that logo is epic
_massi_ has quit [Remote host closed the connection]
<libv>
oliv3r: it looks fantastic on a crappy tablet screen
<libv>
black/black is pretty daft
<libv>
but the blue hue makes it really nice on a tablet
<oliv3r>
hehe yeah
<oliv3r>
i'm gonna put it on my tablet
<libv>
if your tablet works properly, you only see it for a second :p
<oliv3r>
i have a 3 second u-boot delay on mine :p
cajg has quit [Ping timeout: 272 seconds]
<oliv3r>
sunxi 3.4 doesn't need uart?
arokux2 has joined #linux-sunxi
FR^2 has joined #linux-sunxi
cajg has joined #linux-sunxi
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
jinzo has joined #linux-sunxi
ykchavan has quit [Ping timeout: 260 seconds]
jemk has quit [Remote host closed the connection]
ssvb has joined #linux-sunxi
<oliv3r>
ssvb: i did do all those power tests with the battery installed; but it was fully charged, so shouldn't influence it all
<ssvb>
oliv3r: thanks a lot
<oliv3r>
ssvb: i do see I should have used slightly different frequencies for testing
<oliv3r>
ssvb: next time, shall we use the following table for frequency choice:
<oliv3r>
<6>[cpu_freq] INF: voltage = 1450mv frequency = 1008MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1400mv frequency = 912MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1200mv frequency = 720MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1100mv frequency = 528MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1300mv frequency = 864MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1050mv frequency = 312MHz
<oliv3r>
<6>[cpu_freq] INF: voltage = 1050mv frequency = 144MHz
<oliv3r>
with optionally 60 MHz I suppose as rock bottom :)
<oliv3r>
though the cpu_freq driver goes as low as 0 MHz @ 1V
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
jemk has joined #linux-sunxi
FreezingCold has joined #linux-sunxi
pacopad_ has joined #linux-sunxi
<oliv3r>
if i'd want to boot from sda, but it's not connected or powerd up; i'd get a kernel panic right? unable to mount vfs bal bla? not a plain old 'hang'
pacopad has quit [Ping timeout: 264 seconds]
pacopad_ is now known as pacopad
<oliv3r>
hmm, waiting for root device ..
<WarheadsSE>
yeah, if you have rootwait ..
<WarheadsSE>
it'lls just sit there
rellla has joined #linux-sunxi
<mnemoc>
re-moin
<oliv3r>
maybe i don't have power
<WarheadsSE>
also possible, since it is unaware of the device apparently
<oliv3r>
can't be power issue
<oliv3r>
i have a battery backup + external power
<oliv3r>
but lets try a 1 amp charger
<oliv3r>
same
<oliv3r>
strange, very strange
<oliv3r>
actually; when booting the exact same kernel; but with rootfs on mmc; i can mount sda just fine and browse it
rose28357 has quit [Remote host closed the connection]
ZetaNeta has quit [Read error: Connection reset by peer]
ZetaNeta has joined #linux-sunxi
rose28357 has joined #linux-sunxi
nove has quit [Quit: nove]
penghb has quit [Ping timeout: 265 seconds]
jemk has quit [Remote host closed the connection]
netlynx has quit [Quit: Leaving]
penghb has joined #linux-sunxi
woprr has quit [Remote host closed the connection]
<plaes>
soo.. which board with SATA and LAN would you recommend?
<oliv3r>
olimexino micro-a20
<oliv3r>
or olimex-lime
<plaes>
cool, that's what I was aiming to ask next :)
Black_Horseman has joined #linux-sunxi
<oliv3r>
lime is hard to get now; and might only be a10
<oliv3r>
olimexino a20 is ... wel a20
<plaes>
has there been any signs that allwinner starts integrating recent submissions with their code?
<oliv3r>
or stole something because it was better then their own crap :)
<plaes>
well, this would be an improvement :)
pwhalen_ has quit [Quit: Leaving]
pwhalen has joined #linux-sunxi
<jinzo>
afaik - but I'm not that sharp today - there were some concrete proofs/signs that they're including patches from linux-sunxi even in te A10 days?