iamfrankenstein has quit [Ping timeout: 276 seconds]
iamfrankenstein has joined #linux-sunxi
<lennyraposo>
ha ha
<lennyraposo>
I just saw tkaisers super cooling method for the pine
avph has quit [Ping timeout: 246 seconds]
avph has joined #linux-sunxi
<lvrp16>
if h3 is on 40nm, why is the a31s gigantic compared to it?
<lvrp16>
oh it's a respin of a33
yann|work has quit [Ping timeout: 240 seconds]
vagrantc has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
avph has quit [Ping timeout: 246 seconds]
avph has joined #linux-sunxi
IgorPec has joined #linux-sunxi
adj_ has joined #linux-sunxi
IgorPec has quit [Ping timeout: 252 seconds]
ganbold has quit [Ping timeout: 244 seconds]
ganbold has joined #linux-sunxi
reinforce has joined #linux-sunxi
zuikis has joined #linux-sunxi
petr_ has quit [Remote host closed the connection]
avph has quit [Ping timeout: 246 seconds]
petr has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
ganbold has quit [Ping timeout: 244 seconds]
avph has joined #linux-sunxi
hulu1522 has joined #linux-sunxi
robogoat has quit [Ping timeout: 276 seconds]
<wens>
A64 patches for u-boot!
avph has quit [Ping timeout: 246 seconds]
JohnDoe_71Rus has quit [Ping timeout: 276 seconds]
jernej has joined #linux-sunxi
IgorPec has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
robogoat has joined #linux-sunxi
avph has joined #linux-sunxi
popolon has quit [Quit: leaving]
avph has quit [Ping timeout: 246 seconds]
avph has joined #linux-sunxi
vishnup has quit [Quit: Connection closed for inactivity]
nove has joined #linux-sunxi
zuikis has left #linux-sunxi [#linux-sunxi]
zuikis has joined #linux-sunxi
avph has quit [Ping timeout: 246 seconds]
bpi-user has joined #linux-sunxi
mane has joined #linux-sunxi
<bpi-user>
Hi! I have a question about U-Boot (mainline). Maybe someone can clarify that for me. The Debian wiki says:
<bpi-user>
"AHCI support allows u-boot to boot the kernel, initrd and dtb from a SATA harddisk. U-boot itself has still to be installed on an SD card, but the rest of the system can be put onto a (much faster) harddisk."
<bpi-user>
What exactly does that mean? Can the boot.scr file be placed on the SATA disk as well, or does it have to be on the SD card? This would basically mean, I could leave the SD card unpartitioned and just flash the Uboot binary to it, right?
avph has joined #linux-sunxi
<plaes>
wens: \o/
mane has quit [Client Quit]
<NiteHawk>
bpi-user: right. you just need the sd-card to get the bootloader up and running (SPL, U-Boot and possibly U-Boot environment). once that u-boot has initialized your hw, including sata, it shouldn't be a problem to load boot.scr from there
<zuikis>
bpi-user: i think mainline u-boot does not have boot.src file anymore
<bpi-user>
NiteHawk: Thanks. And how does that work if you have connected a SATA drive and a USB drive to the device at the same time? Does U-Boot have an internal order in which it probes storage devices for boot files (boot.scr, vmlinuz/zImage, dtb and possibly an initrd) or will it just pick "randomly"?
avph has quit [Ping timeout: 246 seconds]
<bpi-user>
zuikis: I think you mean the script.bin files (generated from fex). The boot.scr (generated from boot.cmd) is still required. At least all the Debian based systems, I've seen still use it and the sunxi wiki still mentiones them, too.
juri_ has joined #linux-sunxi
<juri_>
iscsitarget-dkms? ;)
avph has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
<NiteHawk>
bpi-user: iirc then u-boot has a predefined search order in its default environment - check http://git.denx.de/?p=u-boot.git;a=blob;f=include/configs/sunxi-common.h;h=a3994e1a6ad479292fe23cdf4c6d0dbcfabd3ef8;hb=HEAD#l416 onwards. but that's easily customized e.g. by reordering or removing boot devices from the env settings
avph has quit [Ping timeout: 246 seconds]
<zuikis>
bpi-user: indeed, wiki mentions it, however i'm not sure what it is for, since i've built my image without boot.scr or if i did it was built automagically
<zuikis>
nor i used any other file besides .bin to make it boot, i just added extlinux/extlinux.conf with info about zImage
<bpi-user>
NiteHawk: Thanks. I'll have a look at the code. But I guess that's already all I need to know. Just for the record, while I do have a SATA drive and a USB drive connected to my device, only the SATA drive has boot files on it. I was just asking because of the device naming issues one may encounter when you have SATA and USB drives connected at the same time (i.e. one time /dev/sda is the SATA drive, the other time, it's the US
tkoskine_ has joined #linux-sunxi
uwe__ has joined #linux-sunxi
<lennyraposo>
wens
mossroy has joined #linux-sunxi
robogoat_ has joined #linux-sunxi
<lennyraposo>
what about the u-boot patches?
steev has quit [Ping timeout: 268 seconds]
uwe_ has quit [Ping timeout: 268 seconds]
MoeIcenowy has quit [Ping timeout: 268 seconds]
robogoat has quit [Ping timeout: 268 seconds]
cosm has quit [Ping timeout: 268 seconds]
edolnx has quit [Ping timeout: 268 seconds]
akaWolf has quit [Ping timeout: 268 seconds]
phylophyl has quit [Ping timeout: 268 seconds]
tkoskine has quit [Ping timeout: 268 seconds]
andoma has quit [Ping timeout: 268 seconds]
lemonzest has quit [Ping timeout: 268 seconds]
tgaz has quit [Ping timeout: 268 seconds]
azend|vps_ has quit [Ping timeout: 268 seconds]
lerc has quit [Ping timeout: 268 seconds]
akaWolf has joined #linux-sunxi
MoeIcenowy_ has joined #linux-sunxi
edolnx has joined #linux-sunxi
akaWolf has joined #linux-sunxi
akaWolf has quit [Changing host]
lerc_ has joined #linux-sunxi
andoma_ has joined #linux-sunxi
avph has joined #linux-sunxi
phylophyl has joined #linux-sunxi
cosm has joined #linux-sunxi
tgaz has joined #linux-sunxi
lemonzest has joined #linux-sunxi
enrico_ has quit [Quit: Bye]
azend|vps has joined #linux-sunxi
<bpi-user>
zuikis: Ah, interesting. I haven't seen any mention of extlinux.conf files on sunxi devices. So, do you basicall place U-boot variables in there to let U-Boot know where to boot from, kernel parameters, etc.?
<bpi-user>
zuikis: Is there a good ressource explaining such a setup?
<lvrp16>
tkaiser: phoronix will have new number for orange pi and banana pi in two weeks
<zuikis>
yes, it reminds syslinux/grub config a lot
<edolnx>
Note that u-boot does not implement syslinux directly, but it is designed to parse the same configuration files
tkaiser has joined #linux-sunxi
<tkaiser>
bpi-user: I would do a web search for 'root partuuid gdisk'
<tkaiser>
lvrp16: Great! With or without heatsink? With or without meaning? ;)
<lvrp16>
tkaiser: with heatsink and active cooling
<lvrp16>
tkaiser: i originally sent him boards but he used orange pi images
<lvrp16>
tkaiser: and no heatsinks, so it was pretty slow
<lvrp16>
tkaiser: i am rerunning the numbers using armbian for him and it will be published probably along with the c2
<tkaiser>
lvrp16: More importantly the Xunlong throttling settings killed CPU cores that's the main reason some results are so low
<lvrp16>
i would imagine so, even armbian 5.05 will kill processors
avph has joined #linux-sunxi
steev has joined #linux-sunxi
<lvrp16>
i didn't follow the opione development but i keep getting cpu freq set errors on 5.05, should i try 5.06?
<tkaiser>
lvrp16: Relinking script.bin should be enough. The whole board auto detection is broken and we must change this since when a user connected an USB hub at first boot then the board will be wrongly identified
<lennyraposo>
tkaiser you should patent your cooling kit for the pine ;)
yann|work has joined #linux-sunxi
<lvrp16>
ahh, thats what happened to me.
<tkaiser>
lvrp16: And then thermal problems are also quite common
<lvrp16>
i mean the usb hub, mine came up as a 2 mini
<lvrp16>
i have an otg hub attached
<tkaiser>
But we have to rework that anyway since the new NanoPi M1+ is 100% identical to OPi PC (less RAM though)
<lvrp16>
kind of useless when orange pi one is already $9
<tkaiser>
lennyraposo: Yes, pretty ugly and loud solution. But with a large and slow fan that is mounted vibration free and controlled airflow this should work pretty well
<tkaiser>
lvrp16: What's useless?
<lvrp16>
nanopi m1+
<lvrp16>
m1*
<tkaiser>
The volume discounts are interesting... starting at +50 pieces ;)
<lvrp16>
how much?
<tkaiser>
I think price per piece incl. shipping is then $12.40 or something like that
massi has quit [Quit: Leaving]
<tkaiser>
But still OPi PC is the better choice due to the voltage regulator. I don't like SoCs being driven with a voltage too high
<lvrp16>
yeah steven is coming out with better versions of the pc and plus
<lvrp16>
exciting :D
<lvrp16>
but $12.4 is still higher than opione
<lvrp16>
although ir is cool
<tkaiser>
lvrp16: Not for me since with shipping included it's $13.5
<lvrp16>
where are you based?
<tkaiser>
lvrp16: Does Steven also tries to sell a board with GbE, 2 GB RAM but without bullshit (WiFi and the horrible GL830)?
<tkaiser>
lvrp16: Germany
<lvrp16>
should be better than that even
avph has quit [Ping timeout: 246 seconds]
<lennyraposo>
just got my pine board
<bpi-user>
zuikis nad edolnx: Thanks for the links! I'll have a look around.
<tkaiser>
bpi-user: Are you talking about 'root=...' in kernel cmdline?
avph has joined #linux-sunxi
<bpi-user>
tkaiser: Thanks. I know that already (and I boot my device with root=PARTUUID=... already to bypass this issue). I just mentioned it because I wasn't sure if tehre are similar issues if you booted directly off a SATA disk without having the boot.scr on the SD card.
<lennyraposo>
nothing a little tension cannot fix when afixed to the case I am designing for a cluster of these things
avph has quit [Ping timeout: 246 seconds]
avph has joined #linux-sunxi
zkirill has joined #linux-sunxi
avph has quit [Ping timeout: 246 seconds]
<lennyraposo>
time to test images
mosterta|2 has joined #linux-sunxi
hulu1522 has quit [Remote host closed the connection]
hulu1522 has joined #linux-sunxi
avph has joined #linux-sunxi
tomboy65 has quit [Ping timeout: 260 seconds]
tkaiser has quit [Read error: Connection reset by peer]
tomboy65 has joined #linux-sunxi
tkaiser has joined #linux-sunxi
avph has quit [Ping timeout: 246 seconds]
matthias_bgg has left #linux-sunxi ["Leaving"]
tomboy65 has quit [Ping timeout: 260 seconds]
avph has joined #linux-sunxi
tomboy65 has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 260 seconds]
avph has quit [Ping timeout: 246 seconds]
avph has joined #linux-sunxi
<lennyraposo>
android image seems ok
<lennyraposo>
needs a little more work but they should have it by end of april
<lennyraposo>
gonna try the arch & ubuntu images next
<montjoie>
yes I am near to solve the last TX ring bug in sun8i-emac
<montjoie>
due to some undocumented feature:)
<lvrp16>
woot
<lennyraposo>
:)
adj_ has quit [Quit: Leaving]
adj_ has joined #linux-sunxi
nove has quit [Quit: nove]
<montjoie>
now I successfully compile something over NFS
bpi-user has quit [Quit: Page closed]
<lennyraposo>
what was the hurdle in transmit?
adj__ has joined #linux-sunxi
Netlynx has quit [Quit: Leaving]
<montjoie>
in short write order of the ring descriptor
<montjoie>
emerge -e world in progress for full proof of work
adj_ has quit [Quit: Leaving]
adj__ has joined #linux-sunxi
avph has quit [Read error: Connection reset by peer]
avph has joined #linux-sunxi
<KotCzarny>
ssvb, remember the last time i recompiled fbturbo with libump and wasnt at the place to check the results? missing redraws are the result (again)
zkirill has quit [Quit: zkirill]
avph has quit [Ping timeout: 246 seconds]
<lennyraposo>
longsleep you about?
Gerwin_J has quit [Quit: Gerwin_J]
tipo has joined #linux-sunxi
<tipo>
anyone got a problem with the banana pro? need something with true sata for a torrent box
zkirill has joined #linux-sunxi
<juri_>
the pcduino3 nano lite has sata.
<KotCzarny>
tipo: what problem?
<tipo>
KotCzarny, just don't want to have a repeat of my experience with the banana pi m1+ (DOA)
<KotCzarny>
tipo, totally dead? o.O
<tipo>
no just the 5v was dead. the board booted fine. just no 5v for sata/fan/etc
zuikis has left #linux-sunxi [#linux-sunxi]
<KotCzarny>
did you try changing the diode?
<tipo>
don't have diodes / small soldering irons laying around
ricardocrudo has joined #linux-sunxi
<tipo>
plus the seller will pay for return shipping
<KotCzarny>
i guess you can steal 5v from the charger anyway
<KotCzarny>
and DOA can happen to anyone
<KotCzarny>
also, olimex is european company if you happen to live in .eu
<KotCzarny>
and they design/make their own boards locally, not via chinese manufacturing
<tipo>
not in the eu
<tipo>
umm.... what is the better option, pcduino3 nano or banana pro? in terms of.... general kernel compatibility. if either can boot openbsd, it's a winner
avph has joined #linux-sunxi
<tkaiser>
tipo: There are no differences between any A20 boards in this regard.
<tipo>
tkaiser, i was under the impression that some manufacturers don't release the uboot source code
<tkaiser>
A20 is A20, there's no difference (just slightly, for example Olimex Lime2 had problems with GbE since they used another GbE PHY than most/all other boards)
<tkaiser>
If that's the case a device with slower I/O but more CPU horsepower will perform better (eg. OPi PC)
<tipo>
mmmm... interesting that you say that, because I have an orange pi pc arriving in the mail fairly promptly
<tkaiser>
tipo: But OPi PC is only Fast Ethernet so in case you expect more than 9MB/s you're lost
tipo has quit [Read error: Connection reset by peer]
tipo has joined #linux-sunxi
<tkaiser>
(but cheap Ethernet dongles with USB 2.0 exist ;)
<zkirill>
I can’t get USB to work (possibly power related) on A20 with USB OTG running FreeBSD and came across the docs for linux-sunxi and Inventra known issue. I was wondering if anyone thinks that these are related and how the Inventra issue was originally diagnosed?
<tipo>
tkaiser, what is this "Fast Ethernet" you refer to? some sort of false gigabit?
<tkaiser>
tipo: Nope, that's just an older standard maxing out at 100 Mbits/sec. 25 years ago that was bleeding edge
<tipo>
mmm... didn't notice it wasn't gigabit, or I forgot since I ordered it
<KotCzarny>
well, for home torrenting 100mbit is good enough
<tipo>
KotCzarny, says you
* tipo
lols
* tipo
wants to have a killer seed ratio to make up for years of leeching due to bandwidth caps
<tipo>
i owe the network like 1TB
<KotCzarny>
pah, data junkie :P
<tipo>
nah; it's years of bandwidth caps, and about 400gb of wikileaks junk data that they never gave me the key for lol
* tipo
stores le precious insurance files
<tipo>
i love the first comment you get, tkaiser. "I tried the test, 7 hours later it was still writing files. I closed the test and turned transmission back on."
<lennyraposo>
ubuntu image is working quite well longsleep
<lennyraposo>
performance is pretty good
<lennyraposo>
installing mate atm
<tkaiser>
tipo: Gets even better when I calculate it will last 15 hours when he's doing wrong and he still thinks I'm asking him for doing a 15h test
<tipo>
tkaiser, i know lol
<tkaiser>
tipo: In case Ethernet bandwidth becomes a problem USB2.0-GbE dongles exist
iamfrankenstein has quit [Ping timeout: 260 seconds]
<lennyraposo>
is the sata controller sata to usb bridge
<lennyraposo>
or ?
<lennyraposo>
?
<tkaiser>
lennyraposo: The SATA controller of Pine64? ;)
<lennyraposo>
nope the device you are referring to
<lennyraposo>
not the pine
<lennyraposo>
now sata there ;)
<lennyraposo>
no*
<tkaiser>
lennyraposo: Nope, Cubietruck IIRC, A20 therfore native SATA but with a write speed limitation
<tkaiser>
40-45 MB/s sequential speed max. But random I/O is pretty fast when used together with a fast SSD
<lennyraposo>
that's decent enough
<lennyraposo>
for a seedbox
reinforce has quit [Quit: Leaving.]
<tkaiser>
lennyraposo: But _if_ the bottleneck isn't storage but CPU instead then A20 is not the device of choice.
<tipo>
anyone tried aria2 as a torrent client on an A20?
<tkaiser>
And interestingly H3 with USB and mainline kernel (and then being able to use UASP -- USB Attached SCSI protocol) is also pretty fast when it's about random I/O
<tkaiser>
tipo: When your OPi PC arrives simply try it out, limit clockspeed to 1008 MHz, kill 2 CPU cores and give it a try
<tipo>
k
iamfrankenstein has joined #linux-sunxi
<lennyraposo>
interesting
IgorPec has quit [Ping timeout: 248 seconds]
<lvrp16>
i just tested the pi one and pc, pi one is slightly slower than the pi 3
<lvrp16>
pi pc is faster than pi 3
<KotCzarny>
:)
<lvrp16>
it does have a clock speed advantage :)
<KotCzarny>
did you test yours with limatester?
<lennyraposo>
anyoen play with the s905s?
<lennyraposo>
the c2s
<lvrp16>
v7 code against v6 raspbian though, so technically pi 3 is faster
<lvrp16>
yes, they are fast
<lvrp16>
especially actively cooled
<lennyraposo>
they aren't leveraging the cpu's capabilities
<lennyraposo>
the pi3
<lennyraposo>
won't see anything probably until the end of year from Pi in terms of a new raspbian for the pi3
<lennyraposo>
sticking to 32 bit for the time being
<lennyraposo>
;)
<tkaiser>
lvrp16: You're testing with PTS?
<lvrp16>
tkaiser: yes
<lennyraposo>
gotta test arch images next
<tkaiser>
lvrp16: Quite uninteresting then since not optimised and partially only single threaded
afaerber has joined #linux-sunxi
<lvrp16>
takiser: Pi One with Heatsink and Active cooling and Armbian is about 5% slower than Raspbian Pi 3 stock with Raspbian, I know it's not quite fair. Benchmarks that need 512MB RAM were horrible though, to be expected.
<tkaiser>
If I want to do number crunching I have to use optimised settings. And if I don't want to to this then it depends on the use cases. I/O or network: Forget about any RPi. Clueless user searching for great community... different story ;)
<lennyraposo>
actually
<lennyraposo>
fairs better
<tkaiser>
lvrp16: Did you have an eye on throttling on all 3 boards?
alain_ has joined #linux-sunxi
<lvrp16>
pi 3 is not actively cooled ;) not fair like i said
<tkaiser>
PTS monitors only temperatures IIRC and does not record cpufreq.
<lvrp16>
pi 3 throttles
<lennyraposo>
sorry was thinking mips
<tkaiser>
LOL, then it's obvious why OPi One gets close to RPi 3
<lvrp16>
with PTS C2 is like 1.5x faster with active cooling
<lennyraposo>
get a giggle from looking back at the pentium - pentium 4 era
<tkaiser>
lvrp16: Banana Pi M3 is twice as fast with heatsink+fan compared to just a heatsink when running the usual PTS stuff.
<lvrp16>
yes, M3 is insanely fast when I benched it
<lvrp16>
i have some cello boards booked, i want to see how fast these fat cores will go
<tkaiser>
lvrp16: At least you won't need a heatsink with Cello
<lvrp16>
tkaiser: am i missing sarcasm? can't tell
<tipo>
what is this crap about "at least you won't need a heatsink"? i have spare heatsinks for every board I own. they're so cheap.
<lvrp16>
irc needs emoticons
<lvrp16>
or accents
<lennyraposo>
wow
<tipo>
lvrp16, no.
<lennyraposo>
just realized the card I used for the android image was a class 4
<lennyraposo>
and it still booted in under a minute
<tkaiser>
lvrpi16: Compare the specs for the 'little' 4 core AMD used on the Cello with the bigger one. Just a few W difference. Then think about the engines for 14 x SATA and 2 x 10GbE (that can't be used on this moronic board). Then you know what to expect
<tkaiser>
2 x SATA and 1 x GbE and 4 slow A53 cores. Not that challenging ;)
<lennyraposo>
and no arm-ored transport ;)
<lvrp16>
tkaiser: they're 4xA57
<lennyraposo>
it is an a57 for the amd
<lennyraposo>
at 1.7
<lennyraposo>
1.9 and 2.0 ghz (depending on model)
<lennyraposo>
the thing will make for a decent storage node
<lennyraposo>
if only dedicated for it
<lennyraposo>
what is the mak ram that can be installed
<lennyraposo>
ddr4 8gb?
alain_ has quit [Quit: Page closed]
<lennyraposo>
ddr3
lemonzest has quit [Quit: Leaving]
<lennyraposo>
128gb ram
<lennyraposo>
ddr3 - 4
<lvrp16>
1 A57 @ 2GHz core is 2W. 4 is 8W. 8 is 16W. Calculate in interconnect power consumption.
<willmore>
lennyraposo two C2's here.
<tkaiser>
lvrp16: I'm curious regarding consumption if all the interesting stuff can't be used
<KotCzarny>
and banana pi is 1.2W total, including micro audio amp
<lvrp16>
tkaiser: the 10GBe and PCIE is the biggest power consumption, i'd imagine ~5-10W idle though
mossroy has quit [Quit: Quitte]
<NiteHawk>
bpi with 240mA @ 5V? That's mostly idle, or?
<KotCzarny>
nitehawk: playing audio from sshfs
<NiteHawk>
ah, okay
<willmore>
I can't get a c2 to throttle running any of the PTS code. I can get to throttle with cpuburn only. I'm not using any active cooling.
<tkaiser>
lvrp16: According to AMD A1120 (4 x A57 @ 1.7GHz) vs. A1170 (8 x A57 @ 2.0Ghz) is 25W vs. 32W TDP.
cptG_ has joined #linux-sunxi
tomboy65 has quit [Ping timeout: 276 seconds]
<tkaiser>
lvrp16: Please report back but this Cello thingie (and all the other LeMaker instruments) isn't interesting anyway...
<lvrp16>
tkaiser: TDP is funny :) there's whole threads on TDP
<lvrp16>
gotta meeting have a good night if i don't catch you guys later
<lvrp16>
tkaiser: the interesting part of the AMD ARM chip is not the features, it's the power management and ACPI. they're basically porting stuff from the desktop side into ARM space.
<tkaiser>
lvrp16: I want 10 GbE! :P
<lvrp16>
tkaiser: formalizing the various bus specifications under ARM so that you can have common code initializing the hardware
tomboy65 has joined #linux-sunxi
<lvrp16>
tkaiser: large board manufacturers wouldn't spend a couple of hundred thousand unless it's a sure thing
cptG has quit [Ping timeout: 260 seconds]
<lvrp16>
tkaiser: and once it becomes a thing, all the other manufacturers jump in and take all the profits out of it
<lvrp16>
tkaiser: so it's really tricky finding a balance, if you do it too early, you're burning cash like calxeda, if you do it too late, you miss the profits.
tomboy65 has quit [Ping timeout: 252 seconds]
<lennyraposo>
bummer to hear that willmore
ricardocrudo has quit [Ping timeout: 246 seconds]
<KotCzarny>
-ECHAN
<willmore>
lennyraposo, what's a bummer?
tomboy65 has joined #linux-sunxi
<lennyraposo>
getting the c2 to throttle
<willmore>
I remounted the HS on one of them and not it never throttles no matter what I do.
<willmore>
lennyraposo, you have to try really hard.
<willmore>
Not like an Opi1 where you just need to look at it sideways and it'll throttle.
<lennyraposo>
lmao
<lennyraposo>
that's what I am hearing
<willmore>
I need to try an Opi1 with some agressive cooling. :)
paulk-collins has quit [Quit: Leaving]
<willmore>
they just didn't layout that board to make large heatsinks practical. I'd need to mill some reliefs out of the bottom of the HS to fit over that silly torroid right next to the H3.
<willmore>
This is why I would advise anyone using a SoC like these to put the SoC on the *bottom* of the board. Then you can easily use a metal baseplate/case as a heatsink.
<KotCzarny>
i run mine in a metal box without vent holes
<KotCzarny>
at 1.2W it's not heating much
<tkaiser>
willmore: Banana Pi/Pro are designed like that. And guess what: People put these in the official acrylic cases with 2 mm airflow around and then started to drill a large hole into the acrylic plate to mount an annoying fan ;)
<willmore>
tkaiser, *facepalm*
<KotCzarny>
tkaiser: my bpim1 churns happily without heatsink/fan
<willmore>
You can lead people to knowledge, but you can't make'em think.
<willmore>
Also acrylic for a case? That's a bad idea. That stuff build up static like nobody's business.
jernej has quit [Ping timeout: 250 seconds]
<willmore>
polycarbonate is way better.
jernej has joined #linux-sunxi
<willmore>
Carbon black colored ABS is probably the best.
<KotCzarny>
full metal boxinist
<willmore>
ODROID U2 did it right. Make the heatsink the case. ;)
<willmore>
tkaiser, that's sort of pretty. They went to a lot of effort to make it pretty.
<tkaiser>
willmore: And functional (great heat dissipation too). Good example for great design
<willmore>
Wow, I added some air moving over my Opi1 and it made a massive difference.
<willmore>
Running cpuminer on four cores used to make it bang around 648, 800-something and 1000-something as it overheated and throttled back down. It was getting 1.4KH/s.
<willmore>
Now, with a little airflow, it's getting 2KH/s and sitting around 52C at 1.2GHz.
<willmore>
Wow.
<willmore>
I wonder if it would survive cpuburn...
<KotCzarny>
nothing survives cpuburn
<willmore>
Don't be so sure. I"m running ssvb's newer A7 version and it's looking good so far...
<KotCzarny>
ahm, thought you have something h3/a64 based
<willmore>
Nope, just an H3 based board.
<willmore>
Looks like it took it to 70C running four cores.
<willmore>
KotCzarny, the C2 can handle cpuburn all day long with no airflow other than natural convection is you fiddle with the HS a tiny bit. If you don't a tiny bit of airflow will probably be enough. I'll go test that shortly.
<KotCzarny>
not bad then
tomboy65 has quit [Ping timeout: 244 seconds]
tomboy65 has joined #linux-sunxi
<Keziolio>
does the C2 has mainline kernel support?
<Keziolio>
oh wait is an amlogic soc
avph has quit [Ping timeout: 246 seconds]
jernej has quit [Ping timeout: 250 seconds]
avph has joined #linux-sunxi
<lennyraposo>
hats off to longlseep on his images
<lennyraposo>
they work
<lennyraposo>
rather nicely
<lennyraposo>
now time for me to build something from scratch
<lennyraposo>
cause at this moment he needs a longsleep ;)
<lennyraposo>
I must say I am impressed with the pine thus far
<agraf>
i need a working kernel :(
<lennyraposo>
for what device?
<agraf>
the pine of course :)
<lennyraposo>
ther eis the 3.10.65-bsp
<lennyraposo>
not mainline but it works
<lennyraposo>
;)
<agraf>
i doubt tumbleweed would be incredibly happy about a 3.10 kernel ...
<agraf>
i can try though
<willmore>
Keziolio, yes, it does. It just doesn't support all that much hardware. :(
<agraf>
also 3.10 doesn't have efistub yet
<agraf>
bleks
<lennyraposo>
since 3.3
<lennyraposo>
;)
<Keziolio>
willmore: well the only thing i care about is the ethernet port and usb, i asked in some channel some time ago and someone said "maybe in the future"
<willmore>
Keziolio, yeah, many ARM SoCs are in the "basic is in mainline, but you'll be headless and maybe not have much I/O". But, it's better than it used to be. I appreciate the work of all the people who have gotten us this far.
<Keziolio>
the raspberry pi has an excellent support, i don't want to be tied to a 2012 kernel until the end of times
<willmore>
KotCzarny, still stable at 70C long term. yay!
<willmore>
I need to look into a tiny fan for lvrp16's Opi1 case. I feel bad cutting it up to put in a fan, though. It's just so cute!
<lennyraposo>
cpuburn willmore?
<willmore>
lennyraposo, yes, sir.
<lennyraposo>
all cores?
<lennyraposo>
hwo long has it been runing for?
<willmore>
With the tiny HS that lvrp16 included in some of their kits. Yes, all cores.
<lennyraposo>
nice
<willmore>
Uhh, since I last mentioned it...... 1:15 or so?
<willmore>
With just a little air moving over it. I have a huge 120mm low RPM fan blowing over the general area.
<lennyraposo>
your current time
<willmore>
No, that's elapsed time.
<lennyraposo>
nm
<lennyraposo>
lol
<lennyraposo>
feel ike an idiot right now
<lvrp16>
willmore: the great thing is that it's secured by screws, so you can remod the entire front and back without worrying about it holding to gether
<lennyraposo>
haha
<willmore>
EDT, FWIW.
<willmore>
lvrp16, it's a very nice little case!
<lennyraposo>
1:15 minutes
<willmore>
1h15m
<lennyraposo>
very nice
<lvrp16>
:D i thought so too, orange pi case is seriously the best case aesthetically
<willmore>
Stable at full 1.2GHz clock the whole time running four cores of ssvb's new A7 cpuburn.
<willmore>
Gotta got for dinner. BBL.
<Keziolio>
so the H3 should be mainlined in 4.5? If i run the command to convert from the zImage that I got from the package manager to uImage i should be able to reboot to the new working kernel?
tipo has quit [Remote host closed the connection]
<Keziolio>
i'm hosting something on the board i'm not very keen on experimenting
tipo has joined #linux-sunxi
<lennyraposo>
longlseep
<lennyraposo>
I am gonan be putting together a few Debian Images