HeHoPMaJIeH has quit [Remote host closed the connection]
avsm has joined #linux-sunxi
<oliv3r>
ssvb: we painted the board black, now cpuburn alone generates 106C
<ssvb>
oliv3r: revA or revC?
<oliv3r>
revC
<oliv3r>
with lima-texture added, we go down 1 or so degree
<oliv3r>
ambient temp, 28.1C
<oliv3r>
40% rH
<oliv3r>
but this is painted black
<oliv3r>
drawing .9 amps
<oliv3r>
my concern is short circuit with the black paint somewhere
<ssvb>
why painting it, btw?
<oliv3r>
we painted all hardware matte black for the infrared camera
<oliv3r>
most reliable and accuratetemperature measurement
<ssvb>
so are you saying that running the textured-cube demo together with the cpuburn is actually *reducing* the heat generation compared to the cpuburn alone?
<oliv3r>
temperate is very slowly going up to 107 though
<oliv3r>
initially it did
<oliv3r>
because cpuburn was told to 'slow down'
<oliv3r>
but it's slowly going up higher
<oliv3r>
108 C now with both
<oliv3r>
no crash yet though
<oliv3r>
the strange thing is, sunxi_tp says 106.8 C
<oliv3r>
109C now for the SoC
<oliv3r>
how likly is paint causing a short
<oliv3r>
almost 110C :p
<oliv3r>
lunch time now, i'll let it run a little longer
<ssvb>
are you not afraid that it may melt or catch fire? ;)
<oliv3r>
bah
<oliv3r>
:p
<oliv3r>
but the paint cause 50% increased temp
<oliv3r>
if it breaks, we got 7 more :p
dlan has quit [Ping timeout: 255 seconds]
dlan has joined #linux-sunxi
wickwire has joined #linux-sunxi
<oliv3r>
ssvb: aww cpuburn-a8 died :(
<oliv3r>
i think it overheated :)
<oliv3r>
so i don't know the final temp
<JohnDoe_71Rus>
maniacs
kivutar has quit [Ping timeout: 252 seconds]
<ssvb>
oliv3r: so, basically A10-Lime (or any other A10 based device) in the default 1008MHz CPU clock frequency configuration may fail if stressed by some really demanding workload
<ssvb>
the x86 systems employ tricks like turbo boost for a reason :)
<ssvb>
we also need thermal throttling support to dynamically adjust the clock frequency and prevent overheating
afaerber_ is now known as afaerber
<ssvb>
if this is not done, then everything (CPU clock speed, Mali clock speed, ...) has to be configured very conservatively
<ssvb>
oliv3r: have you considered using heatsinks?
xinj has joined #linux-sunxi
<oliv3r>
ssvb: we are, we are concidering active cooling too
<oliv3r>
but this may all be due to paint
<oliv3r>
it idles at 50C now
<oliv3r>
the entire board is painted though (got pic)
<oliv3r>
i have a board now with only the SoC painted
<oliv3r>
so will test that now
<oliv3r>
and i'll test a board within the olimex cover
<ssvb>
libv: that's why we want clearly redistributable mali blobs
<ssvb>
libv: until this happens, we can only safely refer to the cubietech website for downloads and should not attempts to mirror them
<RaYmAn>
You have to wonder whether they even live up to this " I certify, under penalty of perjury, that the information in this notice is accurate."
<libv>
ssvb: jem will never deliver that
<libv>
he's too powerhungry and afraid of losing control
<libv>
then there is the way in which arm distributes its drivers, through SoC vendors only
<libv>
the only reason we are seeing arm giving out drivers semi-directly is because they need to establish the mali t series as gpgpu capable
kuldeepdhaka_ has quit [Ping timeout: 244 seconds]
<libv>
i think Cyveillance just committed suicide
<ssvb>
libv: I'm having difficulties trying to understand if Jem Davies is genuinely evil or just clueless about how the software licenses work
<arokux>
libv: [a question not connected to the discussion] do mobile gpu support some openCL or something similar?
<ssvb>
libv: if the former is true, then he can cause a lot of DMCA inconveniences for us, rockchip, exynos and related communities
<libv>
ssvb: ARM should be smarter than this
<libv>
ssvb: jem does not run the whole of ARM, and there are a lot of cool and very intelligent people in his department under him, but those guys sadly get blocked a lot
<ssvb>
libv: if the latter is true, then everyone is mostly safe and the ARM lawyers just did a sloppy job either intentionally to keep everyone on the hook, or because they are just lazy
<libv>
ssvb: there is a lot of that going on yes
<libv>
ssvb: this passing on of distribution responsibility is a laywers wet dream and a nightmare for everyone else
<libv>
also, we did not breach anything with accepting the cubie binaries, we were not the leak
<libv>
the split repos are just a safe-guard really
<libv>
if arm really wants to shoot itself in the foot, they can take the proprietary repo
<libv>
(shoot itself in the foot, like qualcomm just did, i just gave phoronix a heads up btw)
<libv>
qualcomm will be _very_ unhappy about this horrible marketing they just got
afaerber has joined #linux-sunxi
<ssvb>
how does the github dmca takedown work? are they blocking individual repositories or the whole accounts?
<libv>
just repos
<ssvb>
ok
<libv>
ssvb: why, are you scared?
<libv>
this qualcomm thing is going to be a huge shitstorm
<ssvb>
I just don't want to get in any trouble
<libv>
and it all will be undone in a week or so, and there will be a huge "i'm sorry" from qualcomm
<libv>
and cyveillance will have to go in receivership as it will never have any customers again
<libv>
plus, it's github
<libv>
meaning that we all use git and have the repos sitting safely on hundreds of computers across the world
<libv>
this is also bad marketing for github
wickwire has joined #linux-sunxi
<libv>
if needs be, we have a mirror project set up on gitorious in a matter of hours
<libv>
it's really not a time to be scared
<libv>
just get out the popcorn and watch half qualcomm screaming internally and loudly apologising publically
<arokux>
"It also records details about visits to a popular internet journal for Linux operating system users called "the Linux Journal - the Original Magazine of the Linux Community", and calls it an "extremist forum"."
dack has joined #linux-sunxi
<libv>
arokux: we have to be less scared of that today than we should've been a year or so ago
<ssvb>
wow
<oliv3r>
ssvb: with heatsink it looks stable :)
<wens>
wonder what they call the FSF and other organizations
<libv>
arokux: .us businesses are losing a lot of revenue over the nsa thing
<libv>
and they are beginning to act up
<wens>
oliv3r: you should consider adding heatsinks to the ram chips as well
<ssvb>
are the ram chips really hot?
<libv>
since .us businesses singularily drive .us politics, this will get resolved in time
<libv>
but the damage done to the ciscos, junipers, ibms, etc, that's long lasting
<ssvb>
oliv3r: if you want to also stress the ram chips and measure their temperature, then the lima-memtester program is the right choice :)
<wens>
ssvb: my cb2 kind of crashes without them
<wens>
they aren't as hot as the soc, but still hot
<arokux>
libv: what about that other question, can embedded GPUs be used to offload computations?
<libv>
arokux: sure
<libv>
that's what androids renderscript is for
<arokux>
libv: so there is something like OpenCL supported?
<libv>
on the t-series, and other recent embedded gpus, of course
<libv>
on the mali-400, no
<libv>
but you can offload some computation to it
<arokux>
libv: Ok, thanks
<libv>
but the question really is: is your usecase really worth the extra power, management and code overhead?
<arokux>
libv: for now I just need to know if there is only CPU or something else, that should be tried also.
<dack>
I have a weird situation that hopefully someone on here can help with...
alexvf has joined #linux-sunxi
<dack>
I have a cubieboard that works fine when running linux off of a sdcard, but when copying the exact same linux over to the NAND then the wifi stops working.
<dack>
The only difference between the two cases is uboot. Is it even possible for a difference in uboot to be causing something like this?
<dack>
The logs show the proper kernel modules being loaded in both cases, but only the sdcard allows for the wifi to actually work.
<arokux>
dack: what about kernel log. have you compared working and non-working ones?
paulk-collins has quit [Ping timeout: 260 seconds]
<dack>
arokux: yes. the kernel logs are identical exept the one has information about connecting to the AP and doing a dhcp negotiation
<dack>
arokux: the uboot logs are different, though... Is it possible to use u-boot-sunxi to create a new u-boot image for the nand?
wickwire_ has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
<arokux>
dack: no idea. in the non-working case, have you tried to setup connection manually? maybe it will give you some additional hints
<dack>
I've tried to manually get the wifi dongle working after booting, but nothing seems to work. I "ifconfig wlan0 up" and it shows, but no LED action on the dongle. I try to do a scan and it says "not available"
<dack>
arokux: :)
<dack>
arokux: I've tried unplugging and replugging in with no joy
<arokux>
dack: I had a problem where networkmanager was somehow blocking the wlan adapter. i needed to disable networkmanager, then manual config worked.
<dack>
the weird thing is, I did have it working before, but I recently updated the filesystem
<dack>
arokux: I'm not using network manager, I'm using connman. I tried disabling and re-enabling wifi in connman, but no change
<arokux>
dack: I see. I do not have any specific knowledge so I can only give some advice on getting more log output.
<dack>
arokux: k.. I guess I could turn on the debugging for the wifi kernel module...
<dack>
arokux: so, it's probably not the difference in uboot, then?
<arokux>
dack: who knows... it could be.
<arokux>
dack: u-boot isn't directly talking to wlan adapter but it does enable/disable different clocks and does pin configuration.....
<dack>
arokux: ^_^ okay, can I just copy the u-boot.img from u-boot-sunxi on to the nand? I haven't been able to find any docs about updating the uboot on the nand.
<arokux>
I do not know, sorry.
<dack>
alright, I'll see what happens.. worse case, I have to reflash
<arokux>
dack: wait
<dack>
arokux: I'm still here.. :)
<arokux>
dack: u-boot-sunxi cannot talk to NAND, so....
<dack>
arokux: I thought I read something about it only being for sdcards... but I couldn't find that info again yesterday
<arokux>
dack: so you cannot boot anything from NAND using u-boot-sunxi.
<arokux>
dack: well, not only. USB sticks work for example ;)
<arokux>
dack: network boot too
<arokux>
dack: maybe even SATA, I'm not sure.
<dack>
arokux: :) nand seems pretty important for these kinds of devices...
<oliv3r>
arokux: speaking off, i may have to forward port the nand patches to sunxi u-boot
<arokux>
oliv3r: I thought libnand was so ugly it had no chance to be mainlined in u-boot?
<oliv3r>
it won't
<oliv3r>
but for us, and sunxi possibly, there is reason to keep it alive
<arokux>
oliv3r: what about SATA, will u-boot-sunxi boot from SATA ?
<arokux>
dack: well yes. but there were no volunteer yet to add NAND support... maybe you could add it? :p
<dack>
arokux: I know nothing about uboot... just enough to compile it and follow instructions to install it
<arokux>
dack: my knowledge was only enough to teach u-boot to boot from a USB stick...
<dack>
how are all these 3rd party manufactures using uboot on nand in their devices?
<arokux>
dack: they take that old u-boot which is supplied by Allwinner, it supports NAND. u-boot-sunxi was forked from it, but support for NAND was removed.
<dack>
arokux: removed because it was crap?
<arokux>
dack: yes. because u-boot-sunxi was going to be mainlined. and having crap there wouldn't have allowed this.
<dack>
arokux: makes sense... but still annoying for my purposes. ;)
<oliv3r>
either u-boot sunxi, or boot0 for our project :)
<oliv3r>
arokux: if you add SPI, you can make u-boot boot the kernel from sata
<arokux>
dack: well, you are not sure this is u-boot what is causing you problems with wifi ;)
<oliv3r>
arokux: i don't think we removed it
<oliv3r>
arokux: i think we took stock u-boot and added specific stuff, and nand was never added
<oliv3r>
but i could be wrong :)
<oliv3r>
anyway, i very well may have to port nand support
<arokux>
oliv3r: ah, ok. :)
<oliv3r>
mainline kernel may be to risky
<arokux>
oliv3r: yep, could be this way.
<oliv3r>
and too little supported devices
<dack>
arokux: very true..
<arokux>
oliv3r: why you think you need to port? why now?
<oliv3r>
we need to boot from nand
<oliv3r>
i don't wanna use boot0+boot1 etc
<oliv3r>
compiling it is a huge pain
<arokux>
oliv3r: and what do you mean by "our project"?
<oliv3r>
ssvb: with 3.4, lowering the freq. doesn't lower the voltage, does it
<oliv3r>
arokux: secret!
<arokux>
oliv3r: oh...! :)
<arokux>
oliv3r: who is "us" then? :)
<oliv3r>
i work for ultimaker :)
<arokux>
oliv3r: cool! so you aren't on that hotline thing anymore? :)
<dack>
arokux: hmm.. I connman doing all of that for me. It seems like I'd end up with a load of stuff to go through.
<arokux>
dack: you need to disable connman and everything else that configures network automatically. then you do everything manually putting strace every time in front of each command
<dack>
arokux: k.. I'm going to try a few other things first
<arokux>
dack: well, this is just an idea. you can try to find something else.
<arokux>
dack: which ones? just interesting...
<arokux>
dack: btw, all other usb-devices are working fine?
Gerwin_J has quit [Quit: Gerwin_J]
<dack>
arokux: well, before I had some other kernel modules installed.. I was going to try putting those back in, though I didn't think they were doing anything.
<dack>
arokux: hmm.. good point, I should try that first.
<arokux>
dack: I see
<arokux>
oliv3r: now with mainline stuff NAND is in accessible, right?
<arokux>
not accessible*
_massi has quit [Read error: Connection reset by peer]
Gerwin_J has joined #linux-sunxi
ssvb has quit [Ping timeout: 240 seconds]
_massi has joined #linux-sunxi
<dack>
arokux: I have a bluetooth dongle that appears to be working... but none of the wifi dongles are working..
<dack>
arokux: before I had set up all the crda stuff, but since removed it. I'm not sure why that would only affect it on the NAND, though
nabblet has joined #linux-sunxi
FR^2 has quit [Quit: Connection reset by peer]
ssvb has joined #linux-sunxi
<oliv3r>
arokux: not, bbrezillon made some patches, but those are of course not compatible with libnand (different formatting)
<oliv3r>
and the mtd-nand stuff is to experimental i'm affraid
avsm has quit [Quit: Leaving.]
nabblet has quit [Ping timeout: 255 seconds]
xinj has quit [Ping timeout: 240 seconds]
libcg has quit [Quit: rtfm]
physis_ has joined #linux-sunxi
physis has quit [Ping timeout: 264 seconds]
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
petr has quit [Ping timeout: 240 seconds]
petr has joined #linux-sunxi
avsm has joined #linux-sunxi
bertrik has joined #linux-sunxi
bbrezillon has quit [Ping timeout: 248 seconds]
avsm has quit [Quit: Leaving.]
avsm has joined #linux-sunxi
w00tc0d3 has joined #linux-sunxi
w00tc0d3 has quit [Changing host]
w00tc0d3 is now known as netchip
_massi has quit [Quit: Leaving]
rgenoud has quit [Ping timeout: 240 seconds]
bluse has quit [Remote host closed the connection]
bluse has joined #linux-sunxi
lynxis has joined #linux-sunxi
yann_s has quit [Ping timeout: 255 seconds]
yann_s|AFK has joined #linux-sunxi
yann_s|AFK is now known as yann_s
prahal has joined #linux-sunxi
diego_r has quit [Quit: Konversation terminated!]
avsm has quit [Quit: Leaving.]
kuldeepdhaka has joined #linux-sunxi
paulk-collins has quit [Ping timeout: 260 seconds]
dlan has quit [Ping timeout: 264 seconds]
paulk-collins has joined #linux-sunxi
dlan has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 240 seconds]
afaerber has quit [Quit: Verlassend]
amitk has quit [Quit: leaving]
wickwire has quit [Remote host closed the connection]
kivutar has joined #linux-sunxi
wickwire_ has quit [Ping timeout: 260 seconds]
tomboy64 has quit [Ping timeout: 264 seconds]
kivutar has quit [Client Quit]
tomboy64 has joined #linux-sunxi
kuldeepdhaka has joined #linux-sunxi
paulk-collins has quit [Remote host closed the connection]
paulk-collins has joined #linux-sunxi
ninolein has quit [Ping timeout: 272 seconds]
ninolein_ has joined #linux-sunxi
ecelis has quit [Quit: leaving]
kuldeepdhaka has quit [Ping timeout: 255 seconds]
zoobab_ has joined #linux-sunxi
xavia has joined #linux-sunxi
prahal has quit [Quit: prahal]
jukivil1 has quit [Quit: Lost terminal]
netlynx has quit [Quit: Leaving]
ecelis has joined #linux-sunxi
robb83 has quit [Ping timeout: 248 seconds]
robb83 has joined #linux-sunxi
<zombu2>
anyone here that can help e out with an error i m getting on boot ?
<zombu2>
start_linux=====268499
<zombu2>
E:Invaid tag(0x74656874)!
<zombu2>
radxa rock trying to boot aruntu from sd
<ssvb>
zombu2: that's a wrong channel, try to ask for help in #linux-rockchip
<zombu2>
k thnx figured i ask maybe someone had a radxa
bonbons has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
afaerber has joined #linux-sunxi
prahal has joined #linux-sunxi
akaizen has quit [Remote host closed the connection]