lucascastro has quit [Remote host closed the connection]
Nyuutwo has joined #linux-sunxi
lucascastro has joined #linux-sunxi
victhor has quit [Remote host closed the connection]
imcsk8 has joined #linux-sunxi
anarsoul|2 has quit [Ping timeout: 240 seconds]
lurchi__ is now known as lurchi_
detectiveaoi has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
vagrantc has quit [Quit: leaving]
cnxsoft has quit [Remote host closed the connection]
cnxsoft has joined #linux-sunxi
ep__ has joined #linux-sunxi
ep_ has quit [Ping timeout: 248 seconds]
lkcl has joined #linux-sunxi
lurchi_ is now known as lurchi__
detectiveaoi has quit [Remote host closed the connection]
detectiveaoi has joined #linux-sunxi
lkcl has quit [Ping timeout: 260 seconds]
lkcl has joined #linux-sunxi
aleksandr has joined #linux-sunxi
aleksandr has quit [Ping timeout: 260 seconds]
detectiveaoi has quit [Remote host closed the connection]
detectiveaoi has joined #linux-sunxi
lurchi__ is now known as lurchi_
nuuuciano_ has quit [Ping timeout: 248 seconds]
lkcl has quit [Ping timeout: 240 seconds]
dddddd has quit [Remote host closed the connection]
Nyuutwo has quit [Ping timeout: 240 seconds]
nots has quit [Quit: Page closed]
wens has joined #linux-sunxi
ganbold has joined #linux-sunxi
lurchi_ is now known as lurchi__
[7] has quit [Ping timeout: 260 seconds]
mavkhimenia has joined #linux-sunxi
mavkhimenia has quit [Ping timeout: 240 seconds]
imcsk8 has quit [Read error: Connection reset by peer]
imcsk8 has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
TheSeven has joined #linux-sunxi
IgorPec has joined #linux-sunxi
lucascastro has quit [Remote host closed the connection]
lurchi_ has joined #linux-sunxi
lucascastro has joined #linux-sunxi
lurchi__ has quit [Ping timeout: 268 seconds]
reinforce has joined #linux-sunxi
foxx_ has joined #linux-sunxi
lkcl has joined #linux-sunxi
detectiveaoi has quit [Read error: Connection reset by peer]
lkcl has quit [Ping timeout: 260 seconds]
florianH has joined #linux-sunxi
Mr__Anderson has joined #linux-sunxi
imcsk8 has quit [Ping timeout: 248 seconds]
Mr__Anderson has quit [Remote host closed the connection]
aleksandr has joined #linux-sunxi
imcsk8 has joined #linux-sunxi
<lwh`>
I checked fex files of other sunxi boards and gemei_g9 looks almost like that of mine (inet-3f-rev06). What bothers me is dram_clock, which is 408 and on gemei, 432. Is it possible that devs underlocked inet-3f and it really can handle 432?
<KotCzarny>
doesnt mattery much
<KotCzarny>
408 is probably because it's more stable safe value
<lwh`>
So I should keep it that way?
<KotCzarny>
sure, first make something that boots, then you can tweak
<lwh`>
It got me thinking, I just have to recompile dts after modificiation, right? I don't need to rebuild entire uboot?
<KotCzarny>
are you working on uboot or kernel?
<KotCzarny>
both have their own dts
<KotCzarny>
in case of uboot it's compiled in
<lwh`>
Really?!
<KotCzarny>
in case of kernel it's loaded during uboot boot
sr-digitronic has joined #linux-sunxi
<lwh`>
oh, yes, you're right. Forgot about that.
sr-digitronic has quit [Client Quit]
djakov has quit [Ping timeout: 256 seconds]
djakov has joined #linux-sunxi
clemens3 has joined #linux-sunxi
matthias_bgg has joined #linux-sunxi
BenG83_ has quit [Ping timeout: 264 seconds]
BenG83_ has joined #linux-sunxi
djakov has quit [Ping timeout: 256 seconds]
djakov has joined #linux-sunxi
djakov has joined #linux-sunxi
djakov has quit [Changing host]
msimpson has joined #linux-sunxi
victhor has joined #linux-sunxi
fkluknav has joined #linux-sunxi
mavkhimenia has joined #linux-sunxi
mavkhimenia has quit [Ping timeout: 260 seconds]
lkcl has joined #linux-sunxi
lwh` has quit [Remote host closed the connection]
lwh` has joined #linux-sunxi
tuxillo has quit [Ping timeout: 240 seconds]
dddddd has joined #linux-sunxi
tuxillo has joined #linux-sunxi
leviathan has joined #linux-sunxi
ganbold has quit [Ping timeout: 240 seconds]
fkluknav has quit [Ping timeout: 248 seconds]
victhor has quit [Remote host closed the connection]
<libv>
KotCzarny: the idea behind being device centric is that you google for linux + some device info, and you hit the page meant for your device
<KotCzarny>
yes, but from linux/mainline point of view devices are just lego blocks (soc drivers)
<libv>
pointing you to uboot config, dtb name (or fex or whatever), and then to how to build uboot and kernel for them
<montjoie>
libv: I still use wiki for uboot and some board info (where are uart?)
<libv>
but it seems that we are doing development boards only these days
<libv>
which then also means that people have some clue as to which SoC is being used
<KotCzarny>
to add devices you would need users that have them
<libv>
linux-sunxi was also meant to protect against short-lived $shitpi sellers with their desire to create their own "communities" each and every time
<KotCzarny>
and it's that most devs contribute, users not that much
<pmpp>
can't even find a "CHIP" page on elinux site
<KotCzarny>
libv, $hit pi effect is specific because they made sure they have monopoly on those particular socs
nuuuciano_ has joined #linux-sunxi
<libv>
anyway... back to banging my head against the cargoculted wiki that is confluence.
fkluknav has joined #linux-sunxi
mavkhimenia has joined #linux-sunxi
mavkhime_ has joined #linux-sunxi
mavkhimenia has quit [Ping timeout: 246 seconds]
mavkhime_ has quit [Ping timeout: 260 seconds]
aleksandr has joined #linux-sunxi
mavkhimenia has joined #linux-sunxi
mavkhime_ has joined #linux-sunxi
mavkhimenia has quit [Read error: Connection reset by peer]
gnarface has joined #linux-sunxi
wasutton3 has joined #linux-sunxi
lwh` has quit [Quit: ERC (IRC client for Emacs 25.3.1)]
dddddd has quit [Remote host closed the connection]
jbrown has joined #linux-sunxi
afaerber has quit [Quit: Leaving]
lkcl has quit [Ping timeout: 264 seconds]
lkcl has joined #linux-sunxi
afaerber has joined #linux-sunxi
IgorPec has quit [Ping timeout: 256 seconds]
lkcl has quit [Ping timeout: 260 seconds]
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
foxx_ has joined #linux-sunxi
foxx_ has quit [Remote host closed the connection]
merbanan has quit [Ping timeout: 240 seconds]
demfloroub has joined #linux-sunxi
demfloroub has quit [Remote host closed the connection]
f0xx has joined #linux-sunxi
LargePrime has joined #linux-sunxi
mavkhime_ has quit [Read error: Connection reset by peer]
mavkhimenia has joined #linux-sunxi
merbanan has joined #linux-sunxi
BenG83_ has quit [Ping timeout: 246 seconds]
<elros_>
I built image using a33 lichee sdk. Is there any way to boot it from sdcard? I would like to not break existing android for now
IgorPec has joined #linux-sunxi
BenG83_ has joined #linux-sunxi
megi has quit [Ping timeout: 245 seconds]
elros_ has quit [Remote host closed the connection]
f11f12 has quit [Quit: Leaving]
aalm has quit [Ping timeout: 265 seconds]
marble_visions has quit [Quit: bye]
gnarface has quit [Quit: Leaving]
marble_visions has joined #linux-sunxi
cnxsoft has quit [Quit: cnxsoft]
<adj_>
libv, even now that rockchip have, in my opinion, better SOC options for mainline, linux-sunxi wiki and community are much better than anything else in the ARM world
<adj_>
I suppose that the better supported SOC'S by rockchip is because the chromebook effect
ep_ has quit [Quit: Leaving]
hanni76 has joined #linux-sunxi
mavkhime_ has joined #linux-sunxi
mavkhime_ has quit [Read error: Connection reset by peer]
mavkhime_ has joined #linux-sunxi
mavkhimenia has quit [Ping timeout: 252 seconds]
victhor has joined #linux-sunxi
gnarface has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
BenG83_ has quit [Remote host closed the connection]
dddddd has joined #linux-sunxi
lkcl has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
argulp has joined #linux-sunxi
kaspter has quit [Ping timeout: 246 seconds]
kaspter1 has joined #linux-sunxi
kaspter1 is now known as kaspter
fkluknav has quit [Ping timeout: 264 seconds]
aleksandr has quit [Remote host closed the connection]
aleksandr has joined #linux-sunxi
Putti has quit [Remote host closed the connection]
<mirko>
is it normal that - when using `stress -m X` - the device doesn't react anymore at all? Not even via serial. When trying cancel stres svia CTRL+c it takes almost a minute until stress got killed and everythign is back to normal
<mirko>
i'm not stressing the cpu, just memory - stressing all 4 cores is fine, it's just mem
<mirko>
taling about H3 here
<KotCzarny>
running off the sdcard ?
<KotCzarny>
then yes
<KotCzarny>
check swap usage
msimpson has quit [Quit: Leaving]
<mirko>
eeks, forgot about Debian created a swap partition..
<mirko>
thanks, makes sense
<KotCzarny>
disable swap and try again
<DonkeyHotei>
enable zram swap
<KotCzarny>
having mmc trigger bound to led helps
<mirko>
yes, without swap everything seems fine
<KotCzarny>
erm, zram would bring it further down to its knees in low mem usage
<mirko>
the ram chips stay surprisingly cold
<KotCzarny>
clock them higher?
argulp has joined #linux-sunxi
<mirko>
stress always gets killed by OOM - i figured that's because malloc() (almost) never fails and turned off overcommitment. Then stress doesn#t even start -.-
<mirko>
KotCzarny: will try
<KotCzarny>
if you are using armbian, they are using safe values
<KotCzarny>
some boards are better than others tho
<mirko>
using vanilla debian testing
<mirko>
i'm aware of the board issues (having an opi zero and an opi lite)
<mirko>
still, what's the best way to stress memory?
<KotCzarny>
limamemtester
<DonkeyHotei>
recompile the linux kernel in a loop
<KotCzarny>
no, seriously, limamemtester
<KotCzarny>
it uses gpu to drive it hard
JohnDoe_71Rus has joined #linux-sunxi
<KotCzarny>
requires monitor to visually assert results though
sehraf has joined #linux-sunxi
<mirko>
KotCzarny: do i need the proprietary blobs?
<mirko>
which one am i supposed to use and - if "one"-ones - where can i find them?
fkluknav has quit [Ping timeout: 240 seconds]
<KotCzarny>
use the one that works?
<mirko>
i just wanted to make sure to create comparable results
<mirko>
for the table on the wiki page
<KotCzarny>
i thought you wanted to stress test memory
lucascastro has quit [Remote host closed the connection]
<mirko>
yes, and i figured it might make sense to add my obersavtions/results to the wiki table - which asks for more results
<mirko>
which i think only makes sense when using the same images - and as there seem to be "one/lite"- and "pc"-specific ones, i thought i should use the "right" one
fl__0 has joined #linux-sunxi
[7] has quit [Ping timeout: 245 seconds]
fl_0 has quit [Ping timeout: 265 seconds]
TheSeven has joined #linux-sunxi
<ssvb>
mirko: this is just convenience package for Orange Pi PC, somebody may rebuild it for Orange Pi One too
<ssvb>
it bundles the lima-memtester tool with a suitable U-Boot & kernel
fl_0 has joined #linux-sunxi
fkluknav has joined #linux-sunxi
<ssvb>
if you have any bootable image with a sunxi-3.4 legacy kernel for your board, then you can try to run lima-memtester on it directly
<mirko>
ack, thanks
TheSeven has quit [Client Quit]
<mirko>
i was just confused by the wiki table mentioning images they used having "one" instead of "pc" in their filenames
fl__0 has quit [Ping timeout: 256 seconds]
<KotCzarny>
ssvb, maybe limamemtester should have own page?
<ssvb>
KotCzarny: what kind of page? do you mean a linux-sunxi wiki page?
<KotCzarny>
yes
tom_nov has quit [Quit: Leaving]
<KotCzarny>
there is a lot of h3 devices, and it doesnt make much sense to duplicate same paragraphs
TheSeven has joined #linux-sunxi
<ssvb>
why not?
<KotCzarny>
because only results tables would differ
<ssvb>
only Orange Pi PC has been ever properly tested, the reliability of other boards was very much hit and miss
<KotCzarny>
i think people done it on other oranges too
lkcl has quit [Ping timeout: 264 seconds]
clemens3 has joined #linux-sunxi
<ssvb>
do we have a place to host a large number of lima-mentester fel-bootable binaries for various boards?
<KotCzarny>
what about dl.linux-sunxi.org ?
<KotCzarny>
or github?
<ssvb>
isn't it so that the linux-sunxi server runs out of free space periodically?
<KotCzarny>
since it's already on gh, just add binaries subdir?
<KotCzarny>
don't mind me, i have one installer for all h3 devices, saves time, space and bandwidth
<ssvb>
so if we build a bunch of binaries, upload them somewhere and then have a link to the right binary from the board page at the linux-sunxi wiki, then it will work
TheSeven has quit [Ping timeout: 260 seconds]
<ssvb>
this is more reliable compared to having a link to the same binary from every board page PLUS a requirement for the user to select the right board type when running it
nuuuciano_ has joined #linux-sunxi
TheSeven has joined #linux-sunxi
aalm has joined #linux-sunxi
<ssvb>
but the biggest problem is finding volunteers to run these tests, because nobody gives rat about reliability unless their board fails in a very obvious way
<mirko>
ssvb: can't we just buy those boards and do them ourselves?
<mirko>
and yes, by "us", i include myself and wouldn't mind spending 100$ on a AliExpress order
<KotCzarny>
mirko: it's for statistics
<KotCzarny>
you would need lots of them
<KotCzarny>
best would be coercing vendors to do that
<mirko>
my stats and lots you mean: lot's of the very same kind of board?
<pmpp>
i'm willing to integrate memtest to fel-installer if anyone need ( not a GUI but TUI vt100 and indeed is python (with msys2 on windows )
<ssvb>
mirko: it would be beast if board vendors actually did all these reliability tests themselves and contributed the right DRAM config to U-Boot for their board
<pmpp>
btw fel installer help people to launch zadig on windows
<KotCzarny>
pmpp, if you add limamemtester loading it would allow lots of windoze people to contribute results too
<pmpp>
KotCzarny: it could for now nobody failed to setup zadig so it is promising
fl__0 is now known as fl_0
<ssvb>
mirko: the biggest problem with board vendors is that they tend to overclock their hardware
fkluknav has quit [Ping timeout: 260 seconds]
adj_ has joined #linux-sunxi
<ssvb>
because if your board clocks DRAM at 432MHz while your competitor's board clocks DRAM at 408MHz, then you kinda have more attractive specs on paper
<pmpp>
haha that reminds me off people giving extra money to overclock pentium 166 200 mmx
<tkaiser>
ssvb: The problem is those board vendors who do these tests use a patched u-boot
<mirko>
i see
<tkaiser>
FriendlyELEC defines 432 MHz being a good value for some of their boards, uses this value, write 672 somewhere else and this gets picked up by mainline u-boot later
<tkaiser>
So FriendlyELEC's own OS images run stable but boards relying on upstream projects not
<mirko>
so 128M instead of 64M let's H3 boot with a 4k screen attached via HDMI
adj_ has quit [Remote host closed the connection]
adj_ has joined #linux-sunxi
<ssvb>
tkaiser: yes, board maintainers in the mainline U-Boot are generally doing a very bad job
nuuuciano_ has quit [Ping timeout: 246 seconds]
<ssvb>
normally it would be best if board vendors maintained configs for their hardware in the mainline U-Boot
<tkaiser>
But I really don't understand why we constantly end up with those stupid 672 MHz DRAM clockspeeds since the tests showed that we can't rely on such numbers
<ssvb>
what we have now is that some random dude just submits a poorly tested config to the mainline U-Boot, and sometimes even mentions that he did not even test it on real hardware
<ssvb>
nevertheless such patches are accepted because "it is always possible to fix it later" or something like this
<ssvb>
but of course nobody was interested in testing anything or improving the DRAM config on it
lucascastro has joined #linux-sunxi
<ssvb>
borderline reliable 432MHz setup was perfectly fine for the U-Boot Cubietruck maintainer(s)
adj_ has quit [Remote host closed the connection]
<ssvb>
KotCzarny: autotuning?
<KotCzarny>
ssvb,yeah, but it might be distro specific
<KotCzarny>
s/distro/os/
<ssvb>
it's not OS specific, the hardware itself just needs to be configured properly
<KotCzarny>
i meant the integrating testing app and storing result somewhere, maybe sid?
<tkaiser>
KotCzarny: Nobody will test anyway...
dlan has quit [Ping timeout: 246 seconds]
<ssvb>
tkaiser: I see only one possible solution: assume that every board is poorly configured (unless proven otherwise) and add a warning banner at its linux-sunxi wiki page
<ssvb>
then remove these banners one at a time if we make some progress
leviathan has quit [Ping timeout: 268 seconds]
<ssvb>
then the users will maybe start asking questions and contributing something
leviathan has joined #linux-sunxi
adj_ has joined #linux-sunxi
nuuuciano_ has quit [Ping timeout: 246 seconds]
dlan has joined #linux-sunxi
dlan has joined #linux-sunxi
dlan has quit [Changing host]
<tkaiser>
ssvb: Other alternative: give up on Allwinner boards since too much hassles (my approach soon)
<tkaiser>
KotCzarny: Tired of dealing with the same stability issues again and again... and since ssvb's approach to get better DRAM initialization was fruitless simply trash performance a little bit further
dlan has joined #linux-sunxi
dlan has joined #linux-sunxi
dlan has quit [Changing host]
<tkaiser>
On the other hand with some boards a nice DRAM speed boost is there too, eg. with NanoPi M1.
<mirko>
why the neo plus 2 goes up?
<KotCzarny>
if it means stable board.. sure
<tkaiser>
mirko: NanoPi NEO DRAM clock speeds in upstream u-boot are all just fantasy values. No idea why they are as they are
<tkaiser>
In Armbian we use what the device vendor (FriendlyELEC) used after testing
<ssvb>
tkaiser: yes, I don't care about allwinner hardware anymore exactly because it is not a productive use of my time
<mirko>
i was surprised of the success of the bootlin kickstarter campaign - wouldn't have thought it would be backed
<KotCzarny>
mirko: in a big part it's thanks to board vendors
<KotCzarny>
few notable ones contributed
<mirko>
don't get me wrong - highly appreciated. but i didn't expect end users to pledge, and the vendors I'd have assumed, would think similar about allwinner as you do
<ssvb>
tkaiser: and now 4 years later the sunxi board maintainers are still doing a sloppy job
<mirko>
resellers would need to pressure the vendors
fl__0 has joined #linux-sunxi
scream has joined #linux-sunxi
<ssvb>
mirko: the kickstarter campaign will be a success if we get some usable results from it
<ssvb>
it's actually a very interesting project because now it involves money
fl_0 has quit [Ping timeout: 240 seconds]
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
netlynx has joined #linux-sunxi
<ssvb>
and money implies at least some accountability
<mirko>
ssvb: does money change much here? it might hurt the image of bootlin if they don't deliver, that's quite some motivation i guess.. on the other hand: who in 2018 really expects a crowdfunding campaign being successful?
<mirko>
i don't (and i pledged)
<mirko>
i think the only accountability comes with the association to bootlin
fl_0 has joined #linux-sunxi
fl__0 has quit [Ping timeout: 246 seconds]
<adj_>
mirko, all crowdfunding campaigns I funded success, most of them hardware campaigns, and the wort with one or two months delay
<adj_>
best advice is to fund only people you trust
<adj_>
if you do that, few will fail
fkluknav has joined #linux-sunxi
jbrown has joined #linux-sunxi
hardfalcon has joined #linux-sunxi
fl_0 has quit [Quit: STRG + Q]
Mr__Anderson has joined #linux-sunxi
aalm has quit [Ping timeout: 268 seconds]
aalm has joined #linux-sunxi
fl_0 has joined #linux-sunxi
tlwoerner has quit [Ping timeout: 256 seconds]
nuuuciano_ has joined #linux-sunxi
tlwoerner has joined #linux-sunxi
tlwoerner has quit [Excess Flood]
tlwoerner has joined #linux-sunxi
tlwoerner has joined #linux-sunxi
phipli has joined #linux-sunxi
nots has joined #linux-sunxi
aleksandr has joined #linux-sunxi
fkluknav has quit [Ping timeout: 245 seconds]
<mirko>
adj_: hm, maybe i should switch to trusting the right people ;)
aleksandr has quit [Ping timeout: 240 seconds]
<adj_>
:)
leviathan has quit [Read error: Connection reset by peer]
<adj_>
maybe you are too nice
leviathan has joined #linux-sunxi
<adj_>
many times the projects are over ambitious for the people and resources they have
leviathan has quit [Remote host closed the connection]
netlynx has quit [Quit: Ex-Chat]
leviathan has joined #linux-sunxi
<mirko>
how come btw. the H3 on the nano-pi can be clocked so much higher than the H3 on the OPi Zero? PCB surface (as in: mm²) seems to be about the same(?)
<KotCzarny>
voltage of soc
<KotCzarny>
voltage of cpu i mean
<mirko>
but doesn't the zero also have a regulator?
<KotCzarny>
as for dram, its about different type of ram
<mirko>
afaik it is only safe to operate at around 6xx MHz
hardfalcon has quit [Ping timeout: 246 seconds]
<tkaiser>
mirko: Are you confusing DRAM clockspeeds with CPU?
<tkaiser>
Wrt DRAM clockspeeds: same board, different SoC. E.g. Orange Pi Zero Plus 2. Exists with H3 and with H5 SoC.
<KotCzarny>
ah, right, that too
<tkaiser>
If a H3 is on the PCB the DRAM can be clocked at high clockspeeds but consumption and temperatures rise a lot (that's why Armbian is using 408 MHz)
<tkaiser>
But if there's a H5 then the same DRAM module can be clocked with 624 MHz. I have not the slightest idea why
scream has quit [Remote host closed the connection]
<tkaiser>
But reduced DRAM clockspeeds on single channel DRAM H3 boards is mostly due to consumption savings and lower temperatures
scream has joined #linux-sunxi
<tkaiser>
For whatever reasons situation with H5 is different even on the same PCB with same DRAM module
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
<mirko>
KotCzarny: i was talking about H2+ (OPi zero) which is almost a H3 (as far as i understood)
libv_ has joined #linux-sunxi
libv has quit [Ping timeout: 260 seconds]
<tkaiser>
OPi Zero can clock up to 1.2GHz since DVFS implemented 1.1V and 1.3V
vagrantc has joined #linux-sunxi
wasutton3 has quit [Ping timeout: 255 seconds]
jbrown has quit [Ping timeout: 264 seconds]
nuuuciano_ has quit [Ping timeout: 246 seconds]
lucascastro has quit [Remote host closed the connection]