<avafinger>
ouch, cannot believed, removed the USB TTL debug and rebooted, voila! works
<avafinger>
must limit the clk
<avafinger>
Android partition are there but /dev/mmcblk0p16 and /dev/mmcblk0p11 cannot be mounted, /system and /volumn are there
<avafinger>
i get "lost page write due to I/O error on /dev/mmcblk0p16" error
<avafinger>
just pushed a log from the new situation, looks like there is something needed on DTS to force DDR mode
<avafinger>
is there a way to extract the DTS from the eMMC (i can read it)?
<avafinger>
mean DTB
deskwizard is now known as AlphaSierraSierr
AlphaSierraSierr is now known as deskwizard
interrobangd has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
popolon has quit [Quit: WeeChat 1.4]
jbrown has quit [Ping timeout: 240 seconds]
BenG83 has quit [Ping timeout: 242 seconds]
jbrown has joined #linux-sunxi
avafinger has quit [Quit: Page closed]
apritzel has quit [Ping timeout: 258 seconds]
cnxsoft has joined #linux-sunxi
ErwinH has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
ErwinH has quit [Ping timeout: 255 seconds]
cnxsoft has quit [Ping timeout: 245 seconds]
cnxsoft has joined #linux-sunxi
NiteHawk has quit [Quit: Bye.]
NiteHawk has joined #linux-sunxi
ninolein has quit [Ping timeout: 245 seconds]
ninolein has joined #linux-sunxi
jbrown has quit [Ping timeout: 240 seconds]
jbrown has joined #linux-sunxi
pg12 has quit [Ping timeout: 258 seconds]
pg12 has joined #linux-sunxi
terra854 has joined #linux-sunxi
yann-kaelig has quit [Quit: Leaving]
perr has joined #linux-sunxi
perr has joined #linux-sunxi
perr has quit [Changing host]
victhor has quit [Ping timeout: 272 seconds]
jbrown has quit [Ping timeout: 240 seconds]
<MoeIcenowy>
plaes: of course I got sdcard working...
<MoeIcenowy>
is your master 4.9 or 4.10?
jbrown has joined #linux-sunxi
interrobangd has quit [Ping timeout: 240 seconds]
crzyp3ck has joined #linux-sunxi
TheSeven has quit [Ping timeout: 245 seconds]
TheSeven has joined #linux-sunxi
perr has quit [Ping timeout: 246 seconds]
premoboss has joined #linux-sunxi
<plaes>
MoeIcenowy: 4.10
fkluknav has joined #linux-sunxi
jbrown has quit [Ping timeout: 240 seconds]
leviathanch has joined #linux-sunxi
<montjoie>
plaes: happy to see i am not the only one to have this error:)
jbrown has joined #linux-sunxi
<montjoie>
all my bisect go to a mripard patch in 4.9 rc, but 4.9 works. It is really strange
premoboss has quit [Ping timeout: 245 seconds]
IgorPec has joined #linux-sunxi
<plaes>
which patch?
<crzyp3ck>
which is better? for usb <->uart connection. cp2102 adapter or pl2303hx?
<plaes>
no difference for regular uart
<crzyp3ck>
plaes: considering that my pc is linux system and for the future use (who knows what I want to do with it in the future) which one would be the better choice?
<crzyp3ck>
plaes: ????
<plaes>
both are fine
<plaes>
for 99% of sunxi you only need GND, RX, TX
<crzyp3ck>
plaes: no difference?
<plaes>
nope
JohnDoe_71Rus has joined #linux-sunxi
<crzyp3ck>
plaes: do any of them need driver to work? software driver?
|Jeroen| has joined #linux-sunxi
<plaes>
what distro are you using?
<plaes>
anyway, looks like I use them both because the modules are loaded on my machine
<plaes>
lsmod |egrep -e '(cp21|pl23)'
<plaes>
pl2303 20480 0
<plaes>
cp210x 28672 0
<plaes>
on at least Ubuntu/Fedora they get loaded automatically when inserting hw
<plaes>
so don't worry
<KotCzarny>
crzyp3ck: they are cheap, get both
<KotCzarny>
;)
<KotCzarny>
but as plaes have said, for regular work it doesnt matter
<KotCzarny>
as long they are 3.3V versions
<crzyp3ck>
it is slackware 14.2 and openbsd -current
<crzyp3ck>
KotCzarny: np
<crzyp3ck>
:)
<KotCzarny>
and they should be detected automagically, unless you've compiled your own kernel and forgot to include serial dongles drivers
leviathanch has quit [Remote host closed the connection]
<KotCzarny>
do lsusb or dmesg first to confirm insertion was detected
leviathanch has joined #linux-sunxi
<crzyp3ck>
ok
<crzyp3ck>
dmesg kernel modules ; will do
<crzyp3ck>
tnx
cnxsoft1 has joined #linux-sunxi
|Jeroen| has quit [Quit: dada]
cnxsoft has quit [Ping timeout: 248 seconds]
cnxsoft1 is now known as cnxsoft
mzki_ has quit [Ping timeout: 248 seconds]
lerc has quit [Ping timeout: 258 seconds]
<montjoie>
plaes: the cmmit is f233dbca6227703eaae2f67d6d9c79819773f16b pinctrl: sunxi: Rework the pin config building code
<montjoie>
but clearly git bisect bug
<montjoie>
bisecting between 4.9 and 4.10rc1 CANNOT go outside this interval
<montjoie>
"must not" is the better word
lerc has joined #linux-sunxi
<montjoie>
the problem come with "Merge tag 'pinctrl-v4.10-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl"
<plaes>
but it might actually configure pins wrong
<KotCzarny>
still, dishonor for real hackers/crackers
<willmore>
Agreed
the_reverend has joined #linux-sunxi
<the_reverend>
Good evening linux-sunxi people
<the_reverend>
I have a quick question about mac address stability on a machine like the OPi0
<KotCzarny>
the_reverend: no max in hardware, uboot sets it from machine id
<KotCzarny>
s/max/mac/
<KotCzarny>
and linux picks it up from uboot (or generates own)
<the_reverend>
So, if I duplicated a mainline Armbian installation and placed the sd card in a new Orange Pi Zero, the new machine would obtain a new mac to the old machine
<the_reverend>
?
<KotCzarny>
should.
<the_reverend>
Ok sweet, I'll try it. There's a project I'm working on that might need tens/hundreds of the little buggers (all with same IP config, etc). I'm hoping that I can get an image working perfectly, mount it on a Linux VM and just script the small differences that each image needs (hostname)
<the_reverend>
I always thought MAC addresses were written in stone before encountering these little board. I've stuff to learn here.
<KotCzarny>
cheap hardware is cheap
<the_reverend>
Cheap hardware is great. I think I must enjoy the heartache. You guys certainly must ;)
<KotCzarny>
we will yet have to see about longevity
<KotCzarny>
if you want powerful, look for socs with a15/17 cores
<KotCzarny>
might be old, but much faster
<the_reverend>
I'm running one as an ad-blocking DNS server for 100+ simultaneous users. Doesn't even break sweat
<the_reverend>
Opi0
<KotCzarny>
yeah, for that its more than enough
<KotCzarny>
even old rpi would suffice i guess
<the_reverend>
Next project involves flying a Orange Pi, an LTE modem and a Ubiquiti 5GHz radio 60 metres up in the air under a helium balloon at sea :)
<the_reverend>
Streaming HD video back from a rib boat to a main ship :D
<KotCzarny>
o.O
iamfrankenstein has quit [Quit: iamfrankenstein]
<KotCzarny>
if you do that, dont forget to drop the link :)
<the_reverend>
drop a link?
<KotCzarny>
of the video
<the_reverend>
For sure, I certainly will
<the_reverend>
It's for volunteer lifeboats
<apritzel>
the_reverend: MACs aren't even set in stone on PCs
<KotCzarny>
apritzel: they are set in firmware though
<the_reverend>
We're going to have an omnidirectional antenna on the balloon and a directional antenna on the speedboat on a gimbal,
<KotCzarny>
which is almost the same
<apritzel>
you can set them in Linux, if you like
<the_reverend>
there will be an orange pi on that to make sure the directional antenna always points at the balloon!
<KotCzarny>
apritzel: but they will change back on reboot
<apritzel>
and the driver usually programs them from some non-volatile storage
<apritzel>
and this is the part that cheap vendors save
<the_reverend>
yeah, i learned that only recently. its why using mac addresses to control access to networks is a bad idea ;)
<apritzel>
it's good to get longer access to airport WiFis, for instance ;-)
<the_reverend>
yeah, a good idea for users :D
<the_reverend>
so does u-boot use some kind of hash to generate the macs from a unique id, like a cpu serial number or something?
<apritzel>
the_reverend: exactly
<KotCzarny>
which recently changed, because that unique id was read bugged
<KotCzarny>
;)
<KotCzarny>
for h3
<KotCzarny>
but all is well now
<the_reverend>
great! thanks for the info people, i asked on the Armbian forums but I think the knowledgeable folks there are constantly overwhelmed by slightly more basic questions
iamfrankenstein has joined #linux-sunxi
ErwinH has joined #linux-sunxi
<the_reverend>
have a great weekend, adios!
the_reverend has left #linux-sunxi [#linux-sunxi]
the_reverend has joined #linux-sunxi
the_reverend has quit [Client Quit]
<apritzel>
was that comment about knowledgeable people ironic?
<KotCzarny>
maybe he met tkaiser
<KotCzarny>
he does that to people
ErwinH has quit [Ping timeout: 240 seconds]
the_reverend has joined #linux-sunxi
<the_reverend>
the mods there know what they're doing
<the_reverend>
many of the users, really, really don't though
<KotCzarny>
that's true, and that's what forums and search function is for
<KotCzarny>
ie. shared knowledge gets shared
<the_reverend>
low signal to noise on occasion, low modulation, low throughput :D
<the_reverend>
I try to help along, like suggesting that people ask new questions in new threads with clear titles
<the_reverend>
so that *others* can find the answers
<the_reverend>
the entirety of my knowledge is based on googleable information :"D
<the_reverend>
But I honestly think that people playing with these cheap boards are going to do incredibly cool stuff... if they don't get on the wrong side of Herr T Kaiser :p
<the_reverend>
tkaiser: if I'm reading that code right it seems that the OPi0 has a mac address set randomly at first boot, so is an exception to what KotCzarny said about uboot :(
<the_reverend>
why is that?
apritzel has quit [Ping timeout: 258 seconds]
<tkaiser>
the_reverend: that's how the MAC address is set in Armbian now on OPi Zero. No idea why anyone tells you different things.
yann-kaelig has quit [Quit: Leaving]
<the_reverend>
cheers for the headsup. the code seems to target only for OPi0, are the addresses for the others set by uboot?
<KotCzarny>
funny, why you do exception for opi0?
<tkaiser>
KotCzarny: Since the driver sets a random address on each boot. At least the legacy driver. Same with Pine64 BTW (legacy / Ethernet).
<KotCzarny>
isnt uboot setting same address on h3/h2+ based on unchanging SID ?
<tkaiser>
the_reverend: It depends always on the driver what has to be done. Armbian images with mainline kernel are not meant for end users ATM.
apritzel has joined #linux-sunxi
f0xx has quit [Ping timeout: 240 seconds]
<tkaiser>
KotCzarny: Who cares? If the driver doesn't pick up the address, the address has to be set somewhere. And no one is touching the smelly XR819 legacy driver.
<tkaiser>
So you described the future but now users have to take care when they want to copy over OS images from one device to another (same with Pine64!)
<KotCzarny>
we are talking about ethernet or wifi?
<KotCzarny>
my understanding with reverend's question was about ethernet
<KotCzarny>
if he was asking about wifi, then my bad
<the_reverend>
Yup, only using etherenet. Not touching OPi0 Wi-Fi for a long time
<the_reverend>
If ever :D
<tkaiser>
Ok, then I was wrong. With Ethernet MAC address depends on SID (and algorithm changes)
<KotCzarny>
yeah. but using same image on different boards should be safe
apritzel has quit [Ping timeout: 260 seconds]
<the_reverend>
No worries, but useful info indeed :tkaiser we may at some point make use of the Wi-Fi on those boards so it's very useful to know that currently I'd end up with cloned Wi-Fi macs
<the_reverend>
hahaha I spent 3 days over Christmas nearly throwing a router out of the window
<the_reverend>
I ran out of swear words
<the_reverend>
I hardly slept
<the_reverend>
The answer, a limitation in Ubiquiti/VyOS load-balancing set-up and a Huawei USB modem that when without SIM pretends to be every site on the internet
ErwinH has quit [Ping timeout: 258 seconds]
<the_reverend>
Responds to pings for everything
<the_reverend>
o_O
<KotCzarny>
kind of windowsish thing, hide the problem, here see this very colorful, yet zero information page
ErwinH has joined #linux-sunxi
terra854 has quit [Quit: Connection closed for inactivity]
ErwinH has quit [Ping timeout: 240 seconds]
the_reverend has left #linux-sunxi [#linux-sunxi]
popolon has joined #linux-sunxi
ErwinH has joined #linux-sunxi
crzyp3ck has quit [Quit: Leaving.]
ErwinH has quit [Ping timeout: 240 seconds]
avph has quit [Ping timeout: 240 seconds]
avph has joined #linux-sunxi
jbrown has quit [Ping timeout: 240 seconds]
iamfrankenstein has quit [Read error: Connection reset by peer]
Ntemis has quit [Remote host closed the connection]
ErwinH has quit [Ping timeout: 248 seconds]
|Jeroen| has joined #linux-sunxi
victhor has joined #linux-sunxi
my123 has quit [Read error: Connection reset by peer]
iamfrankenstein has joined #linux-sunxi
AneoX has joined #linux-sunxi
my123 has joined #linux-sunxi
my123 has joined #linux-sunxi
my123 has quit [Changing host]
ErwinH has joined #linux-sunxi
ErwinH has quit [Ping timeout: 240 seconds]
ErwinH has joined #linux-sunxi
ErwinH has quit [Ping timeout: 240 seconds]
<AneoX>
Hi! Please, maybe somebody could help me, with A20 uart high baud? I have changed apb1 clock, and got strange behavior. All baud rates calculation become twice bigger. Apb1 was 24mhz, i have set to 96(also 144 tried). Now if i set 250000 rate in my app, actually got ~510638, instead of 1000000, got 2181818, etc
<AneoX>
f1c28c00 register always is 0
<AneoX>
f1c28c04 always 0x5
<AneoX>
its UART_DLL and UART_DLH addresses of uart 3
<AneoX>
i have tried mannually change UART_DLH, but when app starts to communicate, UART_DLH become 0x5 back
jbrown has quit [Ping timeout: 240 seconds]
IgorPec2 has quit [Ping timeout: 272 seconds]
<apritzel>
AneoX: which SoC was that again?
<apritzel>
Are you sure that APB1 is actually at 96 MHz?
<AneoX>
apritzel : now [ccu-err] clk_set_parent: set apb1 parent to sata_pll failed
<AneoX>
sata_pll its AW_SYS_CLK_PLL6
ErwinH has quit [Ping timeout: 255 seconds]
<apritzel>
ElBarto: can you try: tools/mkimage -E -f sun50i_atf.its sun50i-a64.itb
<apritzel>
ElBarto: for creating the FIT image
<apritzel>
ElBarto: so adding -E, to append the binaries to the DT and not embed them?
paulk-collins has quit [Quit: Leaving]
<apritzel>
ElBarto: I think that's missing in the README
<AneoX>
apritzel: if take a look at ccm sources in legacy kernel, looks like it want to use AW_SYS_CLK_PLL62 with apb1, AW_SYS_CLK_PLL62 it is a half clock of AW_SYS_CLK_PLL6, example from sources http://pastebin.com/hwXFwNwN
<ElBarto>
apritzel: that worked thanks !
<ElBarto>
apritzel: also path to the its is wrong in your readme
<ElBarto>
(but this I got before :P)
<apritzel>
ElBarto: I know, but then again the branch is called -wip ;-)
<ElBarto>
apritzel: thanks a lot, having ethernet would be easier for me for FreeBSD dev
<apritzel>
eventually the FIT source should end up in that directory
<ElBarto>
ok, good to know
<ElBarto>
apritzel: is this branch is what you've sent upstream recently ? (I haven't really reading the mails)
<apritzel>
ElBarto: no, actually it's quite outdated, but the only branch I published with FIT support
<apritzel>
the manual says the frequency of PLL6 is (24*N*K)/2 and PLL6*2 is 24*N*K
<apritzel>
AneoX: I don't know if those values are real, as they reflect the modelled frequency, not the actual one
<apritzel>
AneoX: and the A64 for instance uses the doubled frequency as the clock source
<AneoX>
288 is a half of 576, so kernel is divide pll6 to 2, not multiply
<AneoX>
comment in ccm_sys_clk.c line 994: /* rate of AW_SYS_CLK_PLL62 is always the half of the rate of AW_SYS_CLK_PLL6 */
<AneoX>
case AW_SYS_CLK_PLL62: return (__s64)24000000*aw_ccu_reg->Pll6Ctl.FactorN*(aw_ccu_reg->Pll6Ctl.FactorK + 1)>>1;
<AneoX>
shift right way, its divide as i remember
utente_ has joined #linux-sunxi
<AneoX>
of sorry, it was sun5i . for sun7i i found case AW_SYS_CLK_PLL62: return (__s64)24000000 * aw_ccu_reg->Pll6Ctl.FactorN * (aw_ccu_reg->Pll6Ctl.FactorK + 1) >> 2;
<AneoX>
also case AW_SYS_CLK_PLL6X2: return (__s64)24000000 * aw_ccu_reg->Pll6Ctl.FactorN * (aw_ccu_reg->Pll6Ctl.FactorK + 1);
jernej has quit [Ping timeout: 240 seconds]
<AneoX>
so as you said, PLL6 is (24*N*K)/2, but PLL62 PLL6 is (24*N*K)/4
premoboss has quit [Ping timeout: 240 seconds]
<apritzel>
AneoX: mainline seems to use 24*N*K/2 for the APB1 source
<apritzel>
and there is a /4 clock as well, as a possible input for AHB
<apritzel>
AneoX: so PLL62 seems to mean "divided by 2" in the BSP
<apritzel>
AneoX: is there a normal AW_SYS_CLK_PLL6 clock in that code?
<apritzel>
so my hunch is that 24*N*K/2 is the actual APB1 input
<apritzel>
although you
<AneoX>
so i need to change some kernel code
<apritzel>
...r clk_set_parent(clk, tmpClk) seems to set PLL62 as the source
<apritzel>
that may be modelled wrong in the kernel
<AneoX>
yes, but if i tryed pll6, its fails. need some correction
<apritzel>
so that seems to be the bug:
<AneoX>
will try now
<apritzel>
the kernel thinks that PLL62 is the clock source and computes the frequencies accordingly
<apritzel>
but actually it's PLL6, so it's off by factor 2
<apritzel>
AneoX: I assume it only accepts certain clocks as the parent of APB1, so setting PLL6 will be denied
<apritzel>
AneoX: as the kernel believes that is not possible to program
<AneoX>
i understood, need correct possibles parents
<apritzel>
exactly
cptG has joined #linux-sunxi
<AneoX>
looks it works, pll6 now is a parent clock 576 of apb1. uart0 works at 115200 as expected. BUT, half of kernel modules are failed. usb wifi not works, mali, ump, lcd and others )))
<AneoX>
apritzel: any ideas?)))
<apritzel>
because you shouldn't change PLL6, really
<apritzel>
I was wondering about that already
<apritzel>
half of the system depends on PLL6
<apritzel>
which in turn means you can't have every possible frequency for the UART, unfortunately
cptG_ has quit [Ping timeout: 248 seconds]
<apritzel>
but 100 MHz should work, for instance
<AneoX>
it was 600, i have changed to 576 and all systems works, until i change apb1 clock
<apritzel>
so does it actually program PLL6 to 96 MHz instead of 576 and setting the APB1 divider to 6?
<AneoX>
no )))
<AneoX>
PLL6 is 576, but APB1 is 96
<apritzel>
can you dump the actual APB1 clock divide ratio register and check that?
<AneoX>
cant modprobe sunxi_dbgreg in this state, as another modules. libkmod: ERROR ../libkmod/libkmod.c:554 kmod_search_moddep: could not open moddep file '/lib/modules/3.4.103-gb56f56d-dirty/modules.dep.bin'
<apritzel>
that seems to be correct as well: / 3 / 2 = /6