00:00
<
willmore >
Several times what megi said.
00:00
gaston1980 has quit [Remote host closed the connection]
00:01
gaston1980 has joined #linux-sunxi
00:01
Mangy_Dog has quit [Ping timeout: 256 seconds]
00:01
<
megi >
diizzy: ^ hopefully not too late
00:05
* willmore
fears diizzy is on fire.
00:06
<
diizzy >
No, it's crosscompiling
00:06
vagrantc has quit [Quit: leaving]
00:09
<
diizzy >
btw, eth seems to be a bit wonky i mainline (5.7)
00:10
<
diizzy >
well, as close to mainline as libreelec (master) is
00:10
<
willmore >
I'm not sure what I run on my OpiOnePlus.
00:10
<
diizzy >
refuses to boot when eth is attached from cold boot, leds are acting very strange
00:10
<
diizzy >
but it doesn't seem to drop link
00:10
<
willmore >
Power supply sequencing issue?
00:11
<
willmore >
Forgetting to shut things off when powering down?
00:11
<
diizzy >
I tried two different ones 2A+ that runs other SBCs just fine
00:11
<
diizzy >
so I'm quite sure its not a psu issue
00:12
* willmore
is hacking on surplus server power supplies (hulking 870W things). Pretty sure I have enough power...
00:12
<
diizzy >
I mean, it works fine if you wait until you see the slash screen
00:13
<
willmore >
Could be something in the init of the tranceiver chip. I forget the name for them.
00:13
<
diizzy >
Yeah, it seems to hang pretty much instantly
00:14
<
diizzy >
I should probably hook up serial at some point
00:14
<
willmore >
Got kernel debugging on?
00:14
<
willmore >
LOL, yeah, that's where I was thinking you should go.
00:15
<
willmore >
earlyprintk and all that joy.
00:16
random_yanek has quit [Quit: random_yanek]
00:16
<
diizzy >
a bootlog is probably fine
00:16
<
megi >
if you'll see phy attaching issues (timeout) it's probably the same issue that Opi 3 has (the circuit looks identical, for sure)
00:17
<
diizzy >
I'm fairly sure it did work earlier like a few months ago but I might be wrong
00:18
<
diizzy >
megi: ...which I guess is non fixable?
00:18
luke-jr has joined #linux-sunxi
00:18
<
megi >
no, it's fixable, I have it fixed on Opi 3
00:18
<
diizzy >
hammer? :P
00:18
<
megi >
it's just that power supplies for the phy have to be enabled at the same time
00:19
<
megi >
and they are not in some cases
00:19
<
megi >
like every boot :)
00:19
<
diizzy >
dts hacking? :)
00:19
<
diizzy >
or possibly uboot?
00:19
<
megi >
you have to fix ATF
00:19
<
megi >
and the driver
00:20
<
diizzy >
so it boils down to dodgy hw
*sigh*
00:20
<
megi >
no, dodgy sw
00:20
<
megi >
ATF is just wrong
00:21
<
megi >
this should help fixing it
00:22
<
diizzy >
cool, is that 5.7 friendly?
00:23
<
diizzy >
awesome, I'll give that a try
00:26
<
diizzy >
I think I kinda understand the issue
00:36
yann has quit [Ping timeout: 258 seconds]
00:49
ChriChri_ has joined #linux-sunxi
00:52
ChriChri has quit [Ping timeout: 256 seconds]
00:52
ChriChri_ is now known as ChriChri
00:59
lurchi_ is now known as lurchi__
01:23
lurchi__ is now known as lurchi_
01:28
gaston1980 has quit [Ping timeout: 264 seconds]
01:30
gaston1980 has joined #linux-sunxi
01:45
gaston1980 has quit [Ping timeout: 246 seconds]
01:47
gaston1980 has joined #linux-sunxi
02:26
lurchi_ is now known as lurchi__
02:34
cnxsoft has joined #linux-sunxi
03:08
luke-jr has joined #linux-sunxi
03:31
TheSeven has quit [Disconnected by services]
03:31
[7] has joined #linux-sunxi
03:54
xes_ has quit [Read error: Connection reset by peer]
03:55
xes has joined #linux-sunxi
04:35
lurchi_ has joined #linux-sunxi
04:39
lurchi__ has quit [Ping timeout: 256 seconds]
04:48
electrostrong has quit [Read error: Connection reset by peer]
04:57
_whitelogger has joined #linux-sunxi
05:18
gaston1980 has quit [Quit: Konversation terminated!]
05:26
JohnDoe_71Rus has joined #linux-sunxi
05:48
_whitelogger has joined #linux-sunxi
05:54
<
jernej >
megi: LE uses ATF v2.2 which doesn't have this issue
06:04
dddddd has quit [Remote host closed the connection]
06:10
AneoX has joined #linux-sunxi
06:45
_whitelogger has joined #linux-sunxi
07:09
cmeerw has joined #linux-sunxi
07:43
sunshavi has quit [Remote host closed the connection]
07:45
sunshavi has joined #linux-sunxi
07:49
quark_ has quit [Read error: No route to host]
07:51
quark_ has joined #linux-sunxi
08:06
_whitelogger has joined #linux-sunxi
08:31
cnxsoft1 has joined #linux-sunxi
08:31
cnxsoft has quit [Read error: Connection reset by peer]
09:07
jstein has joined #linux-sunxi
09:16
djakov has quit [Ping timeout: 272 seconds]
09:17
djakov has joined #linux-sunxi
09:18
lurchi_ has quit [Ping timeout: 244 seconds]
09:26
lurchi_ has joined #linux-sunxi
09:26
lurchi_ is now known as lurchi__
09:28
<
diizzy >
megi: Awesome, all issues are fixed now!
09:29
<
diizzy >
jernej: it does, it least my board (same guy as on the forums)... I bumped atf and applied cpufreq support, works great! :)
09:30
<
jernej >
megi: smauel: what are improvements in atf v2.3 for sunxi socs?
09:32
<
jernej >
diizzy: if you're able to test stability of cpufreq on each point, please sent patch upstream
09:32
<
jernej >
but that's hard on LE, so I suggest some more general purpose distro
09:34
<
diizzy >
I can give it a try, I'm not a linux-guy however
09:35
<
diizzy >
It's also a lot snappier which is to be expected, doesn't seem to perform much better than amlogic s905x @ 1.5Ghz at least not as far as Hi10p is concerned
09:35
<
jernej >
well, 10-bit videos are supported only for HEVC
09:35
<
diizzy >
But it runs a lot better than without cpufreq =)
09:36
<
diizzy >
Yeah, I honestly didn't expect 300Mhz per core to do much
09:36
<
diizzy >
Not sure if they share the same L2 cache size
09:36
<
jernej >
it should be around 800 MHz
09:36
<
diizzy >
I'm comparing S905X to H6
09:37
<
jernej >
I know that rockchip socs support HW decoding of 10-bit H264 videos, but I'm not sure if it is already supported in LE
09:38
<
diizzy >
In come cases indeed, I'm talking about software decoding on both platform
09:38
<
diizzy >
that said, it would be better to use something more scientific than video playback
09:38
<
jernej >
anyway, 10-bit currently doesn't bring much, because at the end, RK, AML and AW SoCs doesn't support 10-bit display output yet (is in the works)
09:39
<
diizzy >
Hmm... I wonder where Kodi tries to grab cpu speed
09:40
<
jernej >
why would you use SW decoding if HW decoding is available?
09:40
<
jernej >
or you're doying some kind of CPU benchmark?
09:40
<
diizzy >
It's not on H264 10-bit?
09:40
<
diizzy >
No, it was mostly an attempt to see if I needed beefier (faster) hardware
09:42
<
diizzy >
hw should ofc be used if available, I dont see why you wouldn't use it
09:50
<
PPA >
On the topic of HW decoding, is there a linux-noob-friendly introduction anywhere on what components have to be put together to get it working?
09:50
<
PPA >
From cursory reading, I gather that there's Cedrus as the thing that works directly with the hardware,
09:50
<
PPA >
there are libraries (this libvdpau/VA-API stuff) which allows user applications to interface with Cedrus,
09:50
<
PPA >
and finally, the programs used (e.g. Kodi, ffmpeg…) need to be able to work with those libraries?
09:50
<
diizzy >
Seems to hover around 63C with a very beefy heatsink
09:52
<
diizzy >
PPA: probably not, looks like you need patches as pretty much everything is bleeding edge
09:53
<
PPA >
Ah, thanks. That may be a bit too much for a beginner.
09:54
<
diizzy >
I guess the quickest way would be to rip out it from LibreELEC or something
09:55
<
diizzy >
but its probably "not worth it" if you're just going to play around
09:59
<
jernej >
yeah, HW decoding is far from beginner friendly
10:00
<
jernej >
but there is a light of hope that things will stabilize in next or two kernel releases
10:00
<
jernej >
by that I mean 5.9 or 5.10
10:00
<
diizzy >
So ~about 1 year or so heh
10:00
<
jernej >
nope, half a year
10:02
<
jernej >
last week iirc was first serious suggestion to make H264 interface public (along with few modifications)
10:03
<
jernej >
so once interface becames public, it will become official UAPI, which means it will stop changing
10:03
<
diizzy >
one thing is for sure, cpufreq on H6 would for sure overheat/thermal throttle a lot on most H6 platforms
10:03
<
PPA >
Thanks for all the hard work on making these things functional, by the way
10:04
* diizzy
looks the dinky H6 "media players" with abysmal heatsinks
10:05
<
jernej >
diizzy: I will probably make H6 configuration separate from other 64-bit AW SoCs
10:05
<
diizzy >
and the orientation of the IR sensor on this board kinda sucks
10:06
<
jernej >
it's different enough to have different optimum settings
10:06
<
diizzy >
what would that be, different govenor?
10:07
<
jernej >
for one thing, then it has iommu which will be supported by 5.8
10:07
<
jernej >
with iommu, CMA is not needed and kernel can use better memory management
10:08
<
diizzy >
I dont see any obvious improvements as far as libreelec goes but then again I'm not familiar with the inner workings
10:08
<
jernej >
well, thermals are important
10:09
<
jernej >
and one user besides you already reported high temperatures
10:10
<
diizzy >
I thought people already knew that H5/H6 could be quite toasty
10:10
<
jernej >
H6 is also first which supports 10-bit 4K@60Hz (4:2:0) output (not implemented yet)
10:11
<
jernej >
with proper devfreq governor, temperatures can go lower, for sure
10:11
<
jernej >
I think that in the past, I saw cores at 432 MHz (something like that) on H3, because HW did everything else
10:11
<
diizzy >
Not to rain on the parade or anything but I get the impression that Allwinner is loosing ground to rockchip and amlogic quickly :/
10:12
<
diizzy >
I think H3 does 600-something ootb
10:12
paulk-leonov has quit [Ping timeout: 240 seconds]
10:13
dddddd has joined #linux-sunxi
10:13
<
jernej >
I'm pretty sure AW is aware of their position and I heard for some new SoCs in the pipeline
10:13
<
jernej >
but we'll see if they'll stick or not
10:13
<
jernej >
anyway, I have to go
10:14
<
diizzy >
I'll try to do some testing and submit upstream
10:14
<
diizzy >
not sure how to handle the atf patch however
10:14
<
diizzy >
as it seems to be a no go upstream
10:25
yann has joined #linux-sunxi
10:32
Mangy_Dog has joined #linux-sunxi
10:52
paulk-leonov has joined #linux-sunxi
11:24
s3b0 has joined #linux-sunxi
11:26
<
s3b0 >
i have 7" lcd for bananapi (Lamobo-R1)
11:26
<
s3b0 >
and i can't make it work in linux
11:26
<
s3b0 >
i was able to enable it in u-boot
11:26
<
s3b0 >
but linux disaplays only garbage
11:33
luke-jr has joined #linux-sunxi
11:34
_whitelogger has joined #linux-sunxi
11:42
JohnDoe_71Rus has joined #linux-sunxi
11:52
<
megi >
diizzy: throttling != overheating
11:54
<
diizzy >
thats true
11:55
AneoX has quit [Ping timeout: 240 seconds]
11:58
AneoX has joined #linux-sunxi
12:07
kaspter has quit [Ping timeout: 260 seconds]
12:10
random_yanek has joined #linux-sunxi
12:38
jbrown has quit [Ping timeout: 272 seconds]
12:50
jbrown has joined #linux-sunxi
13:05
<
diizzy >
megi: network leds are still wonky but at least it boots :)
13:07
<
megi >
not sure what you mean about wonky leds
13:07
<
megi >
for me on opi3 orange one blinks periodically, and green one on traffic
13:07
<
diizzy >
link led goes on and off within a few seconds
13:08
<
diizzy >
Well, I assume its supposed to be a link led
13:08
<
megi >
I wanted to shut dem off, but apparently, it's controlled by phy, and I haven't found any registers to turn them off
13:09
lurchi__ is now known as lurchi_
13:09
<
diizzy >
its a minor issue
13:30
kaspter has joined #linux-sunxi
13:43
random_yanek has quit [Ping timeout: 244 seconds]
13:50
s3b0 has quit [Quit: Konversation terminated!]
14:22
lurchi_ is now known as lurchi__
14:43
lurchi__ is now known as lurchi_
14:47
<
smaeul >
jernej: difference between v2.2 and v2.3 is 1) support for AXP805 and 2) support for SCP firmware
14:51
dddddd has quit [Ping timeout: 264 seconds]
15:01
cnxsoft1 has quit [Quit: cnxsoft1]
15:18
gnarface has quit [Quit: Leaving]
15:18
lurchi_ is now known as lurchi__
15:22
Wizzup has quit [Quit: brb]
15:24
Mangy_Dog has quit [Ping timeout: 244 seconds]
15:34
Mangy_Dog has joined #linux-sunxi
15:39
gaston1980 has joined #linux-sunxi
15:42
andy25225 has quit [Ping timeout: 256 seconds]
15:52
Wizzup has joined #linux-sunxi
15:54
andy25225 has joined #linux-sunxi
15:57
gnarface has joined #linux-sunxi
16:11
Wizzup has quit [Quit: brb]
16:46
lurchi__ is now known as lurchi_
16:53
gnarface has quit [Quit: Leaving]
17:02
lurchi_ is now known as lurchi__
17:02
lurchi__ is now known as lurchi_
17:15
Wizzup has joined #linux-sunxi
17:16
s3b0 has joined #linux-sunxi
17:17
kiqa has joined #linux-sunxi
17:18
lurchi_ is now known as lurchi__
17:21
lurchi__ is now known as lurchi_
17:30
gnarface has joined #linux-sunxi
17:47
tuxillo has quit [Remote host closed the connection]
18:20
Mangy_Dog has quit [Excess Flood]
18:20
Mangy_Dog has joined #linux-sunxi
18:24
lurchi_ is now known as lurchi__
18:28
lurchi__ is now known as lurchi_
18:49
RichardG867_ has joined #linux-sunxi
18:49
RichardG867_ has quit [Changing host]
18:49
RichardG867_ has joined #linux-sunxi
18:52
RichardG867 has quit [Ping timeout: 244 seconds]
18:53
RichardG867_ is now known as RichardG867
18:58
lkcl_ has quit [Ping timeout: 260 seconds]
19:00
lkcl_ has joined #linux-sunxi
19:05
lkcl_ has quit [Ping timeout: 246 seconds]
19:19
dddddd has joined #linux-sunxi
19:24
lkcl has joined #linux-sunxi
20:16
lkcl_ has joined #linux-sunxi
20:18
lurchi_ is now known as lurchi__
20:19
lkcl has quit [Ping timeout: 256 seconds]
20:21
luke-jr has joined #linux-sunxi
20:26
lurchi__ is now known as lurchi_
21:07
lurchi_ is now known as lurchi__
21:10
lurchi__ is now known as lurchi_
21:13
andy25225 has quit [Ping timeout: 246 seconds]
21:23
andy25225 has joined #linux-sunxi
21:46
Ntemis has joined #linux-sunxi
21:56
yann has quit [Ping timeout: 246 seconds]
21:59
vagrantc has joined #linux-sunxi
22:15
lurchi_ is now known as lurchi__
22:22
[DESCENT] has quit [Quit: No comment.]
22:23
[TEMPO] has joined #linux-sunxi
22:30
cmeerw has quit [Ping timeout: 260 seconds]
22:32
qschulz has quit [Remote host closed the connection]
22:35
qschulz has joined #linux-sunxi
22:41
luke-jr has joined #linux-sunxi
23:11
jstein has quit [Quit: quit]
23:15
ldevulder_ has joined #linux-sunxi
23:16
ldevulder has quit [Ping timeout: 246 seconds]