<budovi>
Hey, I was here the other day because of perf record not working on Pine 64+ with apritzel's 4.7 kernel,, has anyone figured out something in the meantime? I've tried longsleep's BSP kernel in the meantime and it behaves exactly the same
<budovi>
IRQ numbers are the same, but no interrupts are delivered .__.
<budovi>
and I have no idea where to continue debugging
<budovi>
(unless I could use some sort of A64 emulator to profile, have no idea what options exist)
Putti has joined #linux-sunxi
zzeroo666 has joined #linux-sunxi
zzeroo666 is now known as zzeroo
<libv>
heh, seems like this community is indeed slowly dying then, everyone has been successfully moved to lkml, and the different board forums.
iamfrankenstein1 has joined #linux-sunxi
<zzeroo>
libv: :)
<zzeroo>
This is the passage of time. But still it is not so far.
<libv>
it has dwindled significantly
<zzeroo>
I'm looking for an arm7 board with good community support and(!) an lvds connector/interface like the "banana pro" board from lemaker.
pekka10 has joined #linux-sunxi
iamfrankenstein1 has quit [Quit: iamfrankenstein1]
sarietta has joined #linux-sunxi
<jemk>
i have to agree with libv there, it got really hard to follow sunxi work and problems
<jemk>
i don't want to watch the whole lakml and a ton of forums just to stay informed about what happens in sunxi world
<jemk>
why can't sunxi related patches be send to a sunxi mailinglist?
<jemk>
(i've to admit, my last one missed linux-sunxi too, just because get_maintainers didn't include it)
ricardocrudo has quit [Remote host closed the connection]
sarietta has quit [Ping timeout: 276 seconds]
<montjoie>
jemk: all my patch are sent to linux-sunxi mailing list, like lots of other patch for other
<montjoie>
I tried to propose to add it to get_maintainer but mripard is against
<montjoie>
perhaps moving to a vger ML is possible
fvogt has quit [Quit: Konversation terminated!]
fvogt has joined #linux-sunxi
<jemk>
montjoie: i know, some people do so always, some sometimes and some almost never
<montjoie>
the wiki is not sufficient to see what is done ?
Mr__Anderson has joined #linux-sunxi
<jemk>
the wiki is good, but someone has to maintain it, and one still has to search the relevant patches in archives
<buZz>
add to the wiki when you see something with not enough data
<buZz>
stop being passive and grumpy, be active and contentsupplying ;)
<KotCzarny>
or add the bot that will monitor lkml and put new things in wiki automatically
ricardocrudo has joined #linux-sunxi
<jemk>
ok, patches are linked in the wip section, missed that
ricardocrudo has quit [Remote host closed the connection]
<wens>
most patches can also be found on linux-arm-kernel patchwork
<wens>
and you can use the patchwork client to apply them
JohnDoe_71Rus has quit [Ping timeout: 276 seconds]
FergusL has joined #linux-sunxi
<jemk>
wens: sure, but one still has to filter them to extract sunxi related stuff
reinforce has joined #linux-sunxi
arete74 has quit [Read error: Connection reset by peer]
arete74 has joined #linux-sunxi
reinforce has quit [Client Quit]
premoboss has quit [Ping timeout: 244 seconds]
arete74 has quit [Quit: leaving]
<wens>
i try to add whatever feature patchsets i see to the wiki
sarietta has joined #linux-sunxi
<jemk>
i know, i don't want to whine, i just wonder if it wouldn't be possible to make it easier without much effort
pekka10 has quit [Read error: Connection reset by peer]
sarietta has quit [Ping timeout: 244 seconds]
pekka10 has joined #linux-sunxi
sarietta has joined #linux-sunxi
iamfrankenstein1 has joined #linux-sunxi
iamfrankenstein has quit [Ping timeout: 244 seconds]
iamfrankenstein1 is now known as iamfrankenstein
sarietta has quit [Ping timeout: 272 seconds]
JohnDoe_71Rus has joined #linux-sunxi
JohnDoe_71Rus has quit [Client Quit]
Worf_ has quit [Quit: Konversation terminated!]
iamfrankenstein has quit [Read error: Connection reset by peer]
Net147 has quit [Ping timeout: 276 seconds]
iamfrankenstein has joined #linux-sunxi
Net147 has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
JohnDoe_71Rus has joined #linux-sunxi
dev1990 has joined #linux-sunxi
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
JohnDoe_71Rus has joined #linux-sunxi
<fvogt>
Hi everyone! Are there patches to get USB Gadget/OTG running anywhere?
<Ixnus>
Obviously they are not sold by cubie anymore
<Ixnus>
I'm getting two tablets to help with mainline
<Ixnus>
Also , who has tried CVBS ?
Putti has quit [Ping timeout: 276 seconds]
<Ixnus>
I can't seem to find such a cable around: audio jack to (yellow) RCA
<Ixnus>
Thanks
budovi has quit [Quit: Page closed]
<jemk>
Ixnus: I don't know where to get breakout boards, but i bought a micro sd extension cable instead, removed the other end and added a pinheader instead
premoboss has joined #linux-sunxi
<Ixnus>
jemk: thanks, that's a plan b.
Ixnus has quit [Quit: Page closed]
JohnDoe_71Rus has joined #linux-sunxi
JohnDoe_71Rus has quit [Changing host]
JohnDoe_71Rus has joined #linux-sunxi
premoboss has quit [Ping timeout: 244 seconds]
<maldata>
Hi all. I got mainline u-boot and a mainline kernel running on the Merrii Hummingbird A20. I used a nightly build of debian sid's rootfs. It seems to boot fine, starts service (some fail), but once the system's up I can't do anything. It says "Timed out waiting for device dev-ttyS0.device.
<maldata>
Plenty of hits on google about that, but they all say to enable CONFIG_FHANDLE in the kernel... and it already is.
<maldata>
So without the tty, I can't do anything. I don't know its IP address, and even if I did, sshd might not be running.
<maldata>
So, I can still see serial output and I know the OS is alive, but I have no way to provide input. Thoughts?
as0d70apf has quit [Ping timeout: 276 seconds]
<feherneoh>
maldata: Do you have anything like OTG available?
fvogt has quit [Quit: Konversation terminated!]
andor2007 has quit [Read error: Connection reset by peer]
<feherneoh>
plaes: I have added a wiki, not complete, and no images yet, but it's there
<feherneoh>
How do I add images without linking from external site?
<plaes>
you should make photos yourself
<plaes>
and then you upload them
<plaes>
and then add links
<feherneoh>
upload where? that's my question
<plaes>
left sidebar
<maldata>
feherneoh: Possibly. What can I do with that? I have never used it. Note that I CAN interact with u-boot via serial, but once the kernel boots, I can't supply output anymore.
sarietta has joined #linux-sunxi
<plaes>
feherneoh: "Left sidebar, 'Upload File'"
<feherneoh>
thanks
andor2007 has joined #linux-sunxi
xalaar has joined #linux-sunxi
sarietta has quit [Ping timeout: 264 seconds]
<feherneoh>
maldata: try attaching a keyboard
<feherneoh>
if it's attached when u-boot detects USB devices, port should stay powered after boot too
Mr__Anderson has quit [Ping timeout: 272 seconds]
ganbold has joined #linux-sunxi
Mr__Anderson has joined #linux-sunxi
xalaar has left #linux-sunxi ["WeeChat 1.5"]
_mamalala_ has quit [Quit: Connection reset by beer]
_mamalala has joined #linux-sunxi
pekka20 has joined #linux-sunxi
pekka10 has quit [Read error: Connection reset by peer]
as0d70apf has joined #linux-sunxi
jernej_ has joined #linux-sunxi
<feherneoh>
plaes: updated the wiki, added UART info and pictures
smooker has joined #linux-sunxi
smooker has quit [Client Quit]
<maldata>
Is there an easy way to boot into single user mode in u-boot? Is there a better way to set the root password and add a normal user after installing a rootfs?
dev1990 has quit [Quit: Konversation terminated!]
<feherneoh>
I usually do image/rootfs maintenance on PC
reinforce has joined #linux-sunxi
sarietta has joined #linux-sunxi
<maldata>
I'm not familiar with that procedure. Is there a particular search term that I should look for?
<maldata>
Or do you know of a website that covers that?
<feherneoh>
what exactly do you want to do? just set password, and create user?
<maldata>
yeah. The newly created system has no users, and I can't even log in as root.
<feherneoh>
Are you using SD card?
<maldata>
yes
<feherneoh>
Then connect it to PC, mount the root partition, and additional ones if needed (like boot)
<feherneoh>
Get qemu-arm-static if you don't have that yet
Mr__Anderson has quit [Quit: Leaving.]
<feherneoh>
copy it over to your rootfs
<feherneoh>
do chroot /sdcard/mount/path/ /path/to/qemu-arm-static /bin/bash
sarietta has quit [Ping timeout: 264 seconds]
<feherneoh>
you should get a root shell, then use passwd to change root password
<feherneoh>
and useradd/adduser (latter is simpler) to create a new user
<maldata>
ah... I think qemu is the missing piece of the puzzle
<maldata>
I wasn't using that at all, so I couldn't chroot
feherneoh has quit [Ping timeout: 240 seconds]
Axl__ has joined #linux-sunxi
<montjoie>
I started to work on RSA for Crypto Engine, it will be my biggest challenge
<montjoie>
documentation is s..t
<montjoie>
latest H3 user manual was cleared about RSA, abd A64 have clearly mising important part
Axl__ has quit [Quit: Ex-Chat]
Axl__ has joined #linux-sunxi
Axl__ has quit [Ping timeout: 244 seconds]
yann has quit [Ping timeout: 240 seconds]
matthias_bgg has quit [Quit: Leaving]
AndroUser2 has joined #linux-sunxi
enrico_ has quit [Quit: Bye]
ssvb has joined #linux-sunxi
netlynx has joined #linux-sunxi
<libv>
jemk: it's not just the lkml only emails, there are less and less real users who could become contributors
<libv>
it's dying.
feherneoh has joined #linux-sunxi
<feherneoh>
What is your favourite distro? What GUI would you suggest? (device is a tablet)
<feherneoh>
I did not notice I have been disconnected
pekka20 has quit [Read error: Connection reset by peer]
<ssvb>
what is going on with freenode lately? I dropped out of the channel yesterday and could not connect back
<plaes>
o_O
<KotCzarny>
ssvb: ddos kids having fun
<plaes>
16:45 [freenode] -christel(christel@freenode/staff/exherbo.christel)- [Global Notice] Hi all, it would appear that Christmas has come early and we're once more receiving a great deal of packets. We'll work with our brilliant sponsors to mitigate as best we can and apologise for the noise and disruption.
<plaes>
UTC ^^
<ssvb>
ok, maybe while they were fighting with the ddos assholes, I got into a crossfire and ended up in some sort of a blacklist :)
lemonzest has quit [Quit: Leaving]
IgorPec has quit [Ping timeout: 272 seconds]
<KotCzarny>
ssvb, bad neighbourhood? ;)
avph has quit [Ping timeout: 272 seconds]
avph has joined #linux-sunxi
<feherneoh>
Any idea about SD not being primary boot device?
yann has joined #linux-sunxi
Putti has joined #linux-sunxi
<KotCzarny>
feherneoh: write uboot to nand but configure it to boot to sd?
<feherneoh>
Yeah, that will be probably what I'll do
<feherneoh>
would be much better if NAND wasn't half dead
<feherneoh>
maybe there's hope, I just found some fw images dropped to gdrive
premoboss has joined #linux-sunxi
sarietta has joined #linux-sunxi
Mr__Anderson has joined #linux-sunxi
iamfrankenstein has quit [Read error: Connection reset by peer]
iamfrankenstein has joined #linux-sunxi
jemk_ has joined #linux-sunxi
jemk has quit [Quit: leaving]
jemk_ is now known as jemk
<maldata>
When I run lshw on this A20, it reports 2 CPUs, one of which is disabled. I don't see anything missing from the kernel config, but I may not be looking in the right place. Is this something any of you have come across before?
<feherneoh>
What is the easiest way to access NAND if stock ROM does not boot?
corecode has quit [Ping timeout: 252 seconds]
Turl has quit [Ping timeout: 252 seconds]
gaby__ has joined #linux-sunxi
Turl has joined #linux-sunxi
ninolein_ has quit [Quit: No Ping reply in 180 seconds.]
<tkaiser>
maldata: Did you create already an own .dts for your A20 device or is there one already?
iamfrankenstein1 has joined #linux-sunxi
ccaione has quit [Ping timeout: 252 seconds]
ccaione has joined #linux-sunxi
<tkaiser>
maldata: Anyway, just one CPU core is most probably an u-boot issue and in case you're happy with a working distro where you might just exchange device tree stuff I would choose one of the images here: http://www.armbian.com/banana-pi/
iamfrankenstein has quit [Ping timeout: 272 seconds]
iamfrankenstein1 has quit [Ping timeout: 244 seconds]
<NiteHawk>
maldata: with "mainline" kernel, it may be related to the U-Boot startup mode - check http://linux-sunxi.org/PSCI
iamfrankenstein has joined #linux-sunxi
TheSeven has quit [Remote host closed the connection]
TheSeven has joined #linux-sunxi
netlynx has quit [Quit: Ex-Chat]
TheSeven has quit [Quit: No Ping reply in 180 seconds.]
TheSeven has joined #linux-sunxi
vagrantc has quit [Quit: leaving]
apritzel has joined #linux-sunxi
paulk-collins has quit [Quit: Leaving]
Mr__Anderson has quit [Quit: Leaving.]
<akaizen>
devs: when testing uboot / kernel / os work: do you repeatedly write to sdcard from host machine => insert sdcard into target device => reboot target device
<akaizen>
ssvb: thats awesome! sadly not SBCs have that (those without allwinner chips) or dont have support in uboot yet. So I was wondering if anyone has figured out how to emulate sdcard over USB or somehting
<akaizen>
that seems like it would be generic enough to work on any SBC
Guest2635 has quit [Quit: Leaving]
solarnetone has joined #linux-sunxi
<ssvb>
akaizen: it is still boot ROM specific, the Allwinner chips have an USB driver in the boot ROM and implement something that is called FEL protocol (also Allwinner specific), which can be used to issue commands to read/write memory, run code, etc.
<ssvb>
other SoCs from the other vendors may have something that provides similar functionality, but we can't assume any kind of compatibility or even feature parity
<ssvb>
there is a recommended/standard protocol called DFU, but Allwinner opted not to use it in their chips :)
<akaizen>
ssvb: yup, thats why I figured going after sdcard. If there was a way to emulate sd card (host -> usb cable -> microSD -> target device), that would make development pretty fast and easy
<akaizen>
since you are developing / compiling on the host anyway, having an emulated sdcard device that you could copy the files over to would be nice to automate that loop instead of having to manually swap sdcards or RE FEL or some other protocol
<apritzel>
just one example of an SD card muxer which AFAIK some embedded board test farms use
<apritzel>
budovi: (please ignore the link above)
<akaizen>
apritzel: omg yes, this is exactly what I was looking for! thanks
<akaizen>
apritzel: you have any more info on embedded board test farms?
<apritzel>
budovi: do you have a chance of testing this on other ARM64 hardware? Just to rule out that perf doesn't work properly (at the moment)
<ssvb>
akaizen: you still have supposedly nice command line tools, which can make working with these USB protocols easy (sunxi-fel for FEL and dfu-util for DFU, there is also fastboot which is Android specific)
<apritzel>
akaizen: I only know that the Linaro LAVA environment was using this
<akaizen>
apritzel: this is neat, but its a multiplexer and relies on a real SDcard which wears out. They explorerd an emulator but heres their reasoning for not using it: "Second solution (emulator) would have been very good one if it had existed. At the time we were facing the problem there was no any emulator market and probably still isn't. Killer feature of emulator is that there is no way to wear it out like physical ones. As
<akaizen>
history shows SD cards wear out quite fast. After few hundreds flashes they become worn out. But the problem of SD emulator is that requires a lot of fast memory, usually RAM, and this inconvenience rules out this approach."
<ssvb>
apritzel: I tried to do some quick and dirty tests with eMMC support, but they were not very successful - https://linux-sunxi.org/Bootable_eMMC
<akaizen>
I dont really get or buy the memory issue, an USB -> microSD emulator would be faster than any SDcard, have unlimited read/write and be simpler
<apritzel>
akaizen: yeah, read this, but was wondering if memory is still so much of a concern these days
<apritzel>
ssvb: yeah, read your Wiki page the other day
<ssvb>
apritzel: but if boot partitions can be supported by this hardware in principle, then we will support them eventually after exploring the BROM code a bit more :-)
<apritzel>
ssvb: and started to read some paper about eMMC and boot partitions
<ssvb>
apritzel: we will need to use something like DFU for flashing, because FEL is clearly not fast enough
<apritzel>
ssvb: this boot partition stuff didn't sound so complicated, and apparently Linux supports access to it already?
<ssvb>
yes, I can read/write boot partition from Linux just fine
<apritzel>
ssvb: so do you want to have firmware support for that just for boot-strapping boards?
<miasma>
akaizen: eh, wearing out after few hundred flashes would mean flashing the whole contents or with very bad wear leveling
<ssvb>
apritzel: the problem is that the BROM does not want to boot from whatever I write there and I did not find a way to convince it yet
<apritzel>
ssvb: oh
<ssvb>
maybe some special boot flags are missing or whatever
<apritzel>
ssvb: are we sure that this is actually supported?
<miasma>
btw, what do think about having a list of kernel configs for a board in the wiki? e.g. i can't really figure out which driver adds the temp monitoring support for a20/h3
<apritzel>
ssvb: I might check back with the Remix Mini, but IIRC this booted just from the normal eMMC part
<apritzel>
ssvb: the BPi-M64 as well
<ssvb>
I suspect that boot partitions are supposed to be supported, because there is a code in the BROM which writes something to the top 4 bits of the boot media identifier when the lower 4 bits are 0x2 - https://gist.github.com/ssvb/84bf3b34d0906fbd43b3261837546f24
<ssvb>
and the boot media identifier is 0x2 when we boot from the data area of the eMMC
<ssvb>
so it is reasonable to expect that there might be something like 0x12 and 0x22 when booting from either first or second dedicated eMMC boot partition
<ssvb>
I did not have time to do a detailed inspection of the BROM yet, and we have some other things to finish first
<apritzel>
ssvb: indeed ;-)
<apritzel>
ssvb: and btw: your SPI flash code works with jemk's SPL DRAM init code, so I can boot without an SD card
<apritzel>
akaizen: btw: if you don't care about U-Boot development for a moment, wouldn't TFTP boot make you happy already?
<apritzel>
akaizen: either with rootfs from SD/eMMC or via NFS?
<akaizen>
kernel would have to support that
<apritzel>
akaizen: NFS root, you mean?
<akaizen>
and during development until you get NFS root / TFTP working, you're swapping SD cards
<apritzel>
all of these just depend on Ethernet support, but yes, for initial development you may be stuck with this
<akaizen>
I ported gentoo to a rockchip device and got uboot working and then ubuntu and then was able to bootstrap and flash maniline + gentoo to the eMMC
<akaizen>
yes once initial development is done, TFTP / NFS / netboot makes it easy to test
<akaizen>
unless you need to test those lower level components
<apritzel>
akaizen: btw: I was wondering if SD wear out is really an issue
<apritzel>
I deliberately took a cheap 1GB uSD for this purpose, so once it shows symptoms of failures, I just ditch it and take another one
<akaizen>
it is if you are half way around the world and cant get to the the machine to replace the sd card
popolon has quit [Quit: WeeChat 1.4]
<akaizen>
automation is about removing all those moving parts that can break
<akaizen>
linux-sunxi build farm for new and old devices would be nice , no?
<apritzel>
well, yes, for an automated approach this may be an issue
<apritzel>
but then I was wondering how much automated new SoC support development would really be
<apritzel>
in the end someone has to maintain and setup this anyway, so swapping an SD card after four weeks may be not an issue
<akaizen>
For companies it matters and thats a non-starter
<akaizen>
No one wants to drive to the datacenter to swap sd cards ;p
<apritzel>
akaizen: well, initial development usually happens with the device on your desk
<apritzel>
akaizen: and once you got TFTP and netboot to work, you usually don't need to swap SD cards anymore
<akaizen>
Yes, and theres no reason not to improve it
<akaizen>
unless you are working on uboot and or kernel
<akaizen>
which if you are working on graphics, hw/soc feature implementation, vpu, etc..
<akaizen>
requires you to rebuild kernels all the time and sometimes you will get kernel panics
<akaizen>
so you are back to reflashing your SD card
<apritzel>
why that? just reboot and reload your fixed kernel via TFTP
<apritzel>
I do that all of the time ...
<apritzel>
connected the reset and/or power button to a relay connected to some Raspi
<akaizen>
? how can you reload your old kernel when the device wont boot? you have to take out the sdcard and reflash your working kernel
cptG_ has joined #linux-sunxi
<ssvb>
akaizen: the kernel is getting loaded from the network by U-Boot (via TFTP)
<apritzel>
akaizen: your SD card contains a working U-Boot with TFTP support (which is admittedly what you need to get this started)
<akaizen>
so we need to spend development time re-porting TFTP / netboot to uboot for different boards we test
<apritzel>
akaizen: why that?
<apritzel>
TFTP is a generic feature on top of the U-Boot network device framework
<apritzel>
all you need is a driver for your network