Turl changed the topic of #linux-sunxi to: All systems go! // Allwinner/sunxi /development discussion - did you try looking at our wiki? https://linux-sunxi.org - Don't ask to ask. Just ask and wait! - https://github.com/linux-sunxi/ - Logs at http://irclog.whitequark.org/linux-sunxi
<sdschulze> Works. :)
Steven__ has joined #linux-sunxi
vishnup has quit [Remote host closed the connection]
<sdschulze> Thanks everyone. :)
Steven__ has quit [Client Quit]
Steven__ has joined #linux-sunxi
orly_owl has joined #linux-sunxi
<T0mW> mripard_: got lcd framebuffer on stock 3.19 working. Turns out the stock 3.19 kernel sun7i-a20.dtsi has no definition for the 'de_be0-lcd0' only for 'de_be0-lcd0-hdmi'.
owamba has joined #linux-sunxi
<owamba> hi guys
<owamba> how well is SDL supported by sunxi??
<el3> Hi everyone. I got a pcduino, that now crashes after a ½-1 hour . Yesterday the problem was that I had no swap partition on. After fixing that, it got alot better. syslog is here http://pastebin.com/SRVc5JCq Could someone help debugging?
<el3> the pcduino reboots, and then runs again for ½ hour or so
<sdschulze> owamba: the graphics library?
<owamba> yup
<owamba> I need SDL and few other such as SDL-fonts
akaizen has joined #linux-sunxi
<el3> NiteHawk, Now I do the mkswap /dev/mmcklb0p4; swapon /dev/mmcblk0p4 automativćally on each reboot. The swap partition is 1G
<sdschulze> AFAIK, SDL is pretty hardware-oblivious.
<owamba> true, but since it has deep bonds with X/fbdev I would like to know how well it works. Since it supports different rending engines which are not so oblivious
<owamba> s/rending/rendering/
popolon has quit [Quit: WeeChat 1.1.1]
<el3> uname -a -> Linux user 3.4.103-00033-g9a1cd03 #1 SMP PREEMPT Sun Dec 21 16:40:30 CET 2014 armv7l GNU/Linux... if that helps
<el3> Feb 12 01:22:36 user wpa_supplicant[1608]: wlan0: WPA: Group rekeying completed with 00:02:61:7b:16:60 [GTK=TKIP] <-- this pops up each 10 min in syslog before crashing
mmarker has joined #linux-sunxi
simosx has quit [Quit: Leaving]
forest has quit [Remote host closed the connection]
<T0mW> mripard_: thank you :)
<T0mW> el3: maybe use top to see what your ram is doing? Check top every 10 minutes to see if you are leaking memory (e.g. look at the free vs buffer vs used) or cat /proc/meminfo?
mmarker has quit [Read error: Connection reset by peer]
leviathanch has joined #linux-sunxi
naobsd has joined #linux-sunxi
forest has joined #linux-sunxi
Steven__ has quit [Quit: Leaving]
<el3> T0mW, http://pastebin.com/q3Z4HqCv <-- cat /proc/meminfo
Steven__ has joined #linux-sunxi
<T0mW> el3: google 'Unable to handle kernel paging request at virtual address', that should hit on 'http://www.makelinux.net/ldd3/chp-4-sect-5', search that page for 'Unable to handle kernel paging request at virtual address'
<T0mW> el3: might give you some insight?
mmarker has joined #linux-sunxi
<el3> T0mW, okey thank you
<T0mW> el3: you might have something goofy going on with a misconfiguration / misbehaving app and it is sucking your memory pool dry.
<el3> T0mW, okey a python script that I made runs sometimes
<el3> top VIRT shows 14000 on that
<el3> T0mW, or should I watch %MEM?
TheSeven has quit [Ping timeout: 250 seconds]
alexxy[home] has quit [Quit: No Ping reply in 180 seconds.]
alexxy has joined #linux-sunxi
TheSeven has joined #linux-sunxi
owamba has quit [Ping timeout: 246 seconds]
naobsd has quit [Quit: naobsd]
wenbin has joined #linux-sunxi
nuuhku has quit [Ping timeout: 246 seconds]
Renard has quit [Quit: Renard]
<el3> all ok on memtester
<el3> 2nd time I get failures on bit flipping and Walking ones
<el3> 3rd ok
<Turl> failures are bad :(
<Turl> are you using the lima memtester?
<Turl> or just the normal cpu one?
naobsd has joined #linux-sunxi
naobsd has quit [Client Quit]
naobsd has joined #linux-sunxi
<el3> I just installed apt-get install memtester. And ran memtester 5 1
<el3> Turl,
<Turl> it's a pretty bad case if the normal cpu one can catch it
<Turl> usually the lima one catches them quickly when they exist
<el3> I will try the lima
viccuad has quit [Read error: Connection reset by peer]
<el3> hmm -bash: cmake: command not found
F1skr has quit [Quit: WeeChat 1.1.1]
focus has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
mynick2 has joined #linux-sunxi
el3 is now known as el3-
mynick2 is now known as el3
khuey_ is now known as khuey|away
el3- has quit [Quit: Leaving]
akaizen has quit [Remote host closed the connection]
akaizen has joined #linux-sunxi
montjoie[home] has quit [Ping timeout: 244 seconds]
montjoie[home] has joined #linux-sunxi
akaizen has quit [Ping timeout: 240 seconds]
DagoRed has quit [Ping timeout: 250 seconds]
orly_owl has quit [Quit: leaving]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
akaizen has joined #linux-sunxi
FreezingCold has quit [*.net *.split]
AnticitizenOne has joined #linux-sunxi
FreezingCold has joined #linux-sunxi
sdschulze has quit [Ping timeout: 250 seconds]
AnticitizenOne has quit [Ping timeout: 244 seconds]
zombu2 has joined #linux-sunxi
<ssvb> el3: you can also download a static binary here - https://github.com/ssvb/lima-memtester/releases
diemaco has joined #linux-sunxi
<ssvb> Turl, el3: also updated the wiki page to suggest downloading it as the primary option :)
Andy-D has joined #linux-sunxi
TheSeven has quit [Ping timeout: 265 seconds]
TheSeven has joined #linux-sunxi
<Turl> ssvb: :)
orly_owl has joined #linux-sunxi
<Turl> ssvb: why --no-check-certificate?
<ssvb> Turl: https download will fail if the time is set wrong on the device
<ssvb> Turl: and we don't have an RTC battery on many boards
<Turl> hm ok
<ssvb> yes, this is ugly
<ssvb> but I guess, some users might be puzzled when the download fails because of the certificate check problem due to a wrong date
<Turl> it's kind of defeating the point of using https
<Turl> but well, I hope they don't do stability testing on a "production" system :)
<ssvb> if people actually care about what they are running on their systems, then they should compile from sources instead of using binaries in the first place :)
<Turl> ssvb: fair enough :)
<Turl> ssvb: you should add a chmod +x too
<ssvb> you are right, thanks
<ssvb> hmm, "signal 6 (Aborted)"?
<Turl> ssvb: there's an abort() on the code
<Turl> ssvb: don't worry, that's my x86 box :)
<ssvb> ok :)
<Turl> qemu is magical
<ssvb> also added the explanations about --no-check-certificate to the wiki, now everything should be foolproof and noob compliant :)
Steven__ has quit [Ping timeout: 245 seconds]
<Turl> ssvb: great, thanks :)
<wens> hmmm... Documentation/driver-model/platform.txt mentions putting the probe function and associated stuff in the .init section to reduce footprint
<wens> works for non-hotpluggable hardware
<wens> but not a whole lot of drivers do this
Steven__ has joined #linux-sunxi
<Turl> wens: some of the init clock functions are/were marked init
<wens> yeah, but i meant for all the other drivers
<wens> surely DT-based hardware, especially the SoC, isn't hot pluggable :p
<wens> anyway, just ran across the stuff while cleaning up the rsb driver
<wens> thought i'd ask
<Turl> wens: yeah, I agree with you
<wens> just so happens the probe/setup functions are 200 LoC, almost 50% of the driver
<Turl> :)
reinforce has joined #linux-sunxi
hipboi has quit [Read error: Connection timed out]
hipboi has joined #linux-sunxi
Steven__ has quit [Ping timeout: 265 seconds]
<wens> bummer, compiler won't notice strings only used in probe function and put them in .init.rodata
cubear has joined #linux-sunxi
forest has left #linux-sunxi ["Leaving"]
reinforce has quit [Remote host closed the connection]
reinforce has joined #linux-sunxi
quitte_ has joined #linux-sunxi
quitte has quit [Ping timeout: 240 seconds]
Steven__ has joined #linux-sunxi
bsdfox has quit [Ping timeout: 246 seconds]
bsdfox has joined #linux-sunxi
bsdfox has joined #linux-sunxi
<a1d3s> hm something went wrong on u-boot i do make Bananapro_defconfig , make CROSS_COMPILE=arm-linux-gnueabihf- , dd if=/dev/zero of=/dev/mmcblk0 bs=1k count=1023 seek=1 and finaly dd if=u-boot-sunxi-with-spl.bin of=/dev/sdb bs=1024 seek=8
<a1d3s> it boots but it searchs every time for sun7i-bananapi.dtb not bananapro.dtb
<a1d3s> if i use make menuconfig it shows correctly bananapro.dtb
<wens> check your bootscript?
<a1d3s> version is u-boot-2015.4-rc1 0202 and patched
<wens> i meant your boot.cmd/boot.scr if you have one
<a1d3s> wens yea it was ;)
<a1d3s> need more coffee
<a1d3s> i copy&paste my build plan and there is pi and not pro
<wens> i use the fdt image name variable in my script
<a1d3s> ?
<wens> load ${devtype} ${devnum}:${bootpart} 0x46000000 boot/uImage
<wens> load ${devtype} ${devnum}:${bootpart} 0x49000000 boot/${fdtfile}
ninolein has quit [Ping timeout: 252 seconds]
ninolein has joined #linux-sunxi
<a1d3s> i got an lvds display from lemaker but dont know how to get it work
<a1d3s> and on hdmi i get only a white striped flickr image
<a1d3s> kernel 3.19-next-20150211
<oliv3r> morin'
weissschloss has quit [Remote host closed the connection]
<a1d3s> morning
weissschloss has joined #linux-sunxi
melonipoika has quit [Remote host closed the connection]
paulk-collins has joined #linux-sunxi
p1u3sch1 has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
gianMOD has joined #linux-sunxi
sehraf has joined #linux-sunxi
f15h has joined #linux-sunxi
selfbg has joined #linux-sunxi
_massi has joined #linux-sunxi
<a1d3s> do i have to define somewhere the con2 for the lvds display?
gianMOD has quit [Remote host closed the connection]
frank has joined #linux-sunxi
f15h has quit [Remote host closed the connection]
sehraf has quit [Ping timeout: 246 seconds]
premoboss has joined #linux-sunxi
sehraf has joined #linux-sunxi
wenbin has quit [Quit: Leaving]
Black_Horseman has quit [Remote host closed the connection]
NiteHawk has quit [Ping timeout: 264 seconds]
NiteHawk has joined #linux-sunxi
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 244 seconds]
naobsd has quit [Quit: naobsd]
Steven__ has quit [Quit: Leaving]
wickwire has joined #linux-sunxi
paulk-collins has quit [Remote host closed the connection]
hehopmajieh__ has joined #linux-sunxi
sdschulze has joined #linux-sunxi
HeHoPMaJIeH has quit [Ping timeout: 246 seconds]
viccuad has joined #linux-sunxi
gianMOD has joined #linux-sunxi
naobsd has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 250 seconds]
popolon has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
Akagi201 has quit []
<wens> module_platform_driver_probe() doesn't work with deferred probing :(
<mripard_> yeah, which kind of makes sense
<mripard_> what if your probe depends on a module ? :)
naobsd has joined #linux-sunxi
<wens> so much for saving space
<mripard_> I'm not sure probe was ever supposed to be in __init though
Renard has joined #linux-sunxi
<wens> from base/platform.c: Note that this is incompatible with deferred probing.
<wens> now i see it...
<mripard_> yeah, it's doable, but it doesn't mean it was meant to be :)
<wens> oh well
<wens> cant think of an easy way to fix it
<mripard_> fix what?
<wens> deferred probing + __init stuff
<mripard_> I don't see how you can
<mripard_> it's fundamentally broken
<wens> yeah, it's like the "when is the kernel done booting" problem
<wens> in theory the kernel should be able to know when it has exhausted the list of builtin drivers
<mripard_> yeah
<mripard_> but again
<mripard_> it doesn't mean that your probe is blocked by a module
<mripard_> that might be loaded afterwards
<mripard_> or might not
<mripard_> so unless you rewrite the __init section related code, to create new sections for each drivers, and removing these init sections only if the probe was successful
<mripard_> then you can't fix this
<wens> :(
<wens> mripard_: btw, there's a sunxi Kconfig dependency patch, don't know if you've seen it or not
<mripard_> yeah, I've seen it
pmattern has joined #linux-sunxi
<mripard_> can't we just select the RESET_CONTROLLER option when we enable the clocks ?
<mripard_> that would seem more natural
<wens> you mean adding Kconfig options for the clocks, right? cause we don't have any atm
premoboss has quit [Ping timeout: 250 seconds]
book` has quit [Read error: Connection reset by peer]
<wens> hmm... limit for stable patches is 100 lines, seems doable
naobsd has quit [Quit: naobsd]
book` has joined #linux-sunxi
zombu2 has quit [Read error: Connection reset by peer]
zombu2 has joined #linux-sunxi
zombu2 has quit [Read error: Connection reset by peer]
wickwire has quit [Ping timeout: 246 seconds]
zombu2 has joined #linux-sunxi
aballier has quit [Quit: leaving]
FR^2 has joined #linux-sunxi
Andy-D has quit [Ping timeout: 264 seconds]
wickwire has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
bengal has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
sdschulze has quit [Quit: Verlassend]
naobsd has joined #linux-sunxi
bengal has quit [Ping timeout: 265 seconds]
cazzacar1a has joined #linux-sunxi
cazzacar1a has quit [Client Quit]
gianMOD has joined #linux-sunxi
Andy-D has joined #linux-sunxi
ssvb has quit [Ping timeout: 240 seconds]
AnticitizenOne has joined #linux-sunxi
<mripard_> wens: hmmm, nevermind then
AnticitizenOne has quit [Ping timeout: 255 seconds]
p1u3sch1 has quit [Ping timeout: 252 seconds]
p1u3sch1 has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
zombu2 has quit [Ping timeout: 245 seconds]
zombu2 has joined #linux-sunxi
selfbg has quit [Quit: Leaving]
melonipoika has joined #linux-sunxi
a1d3s is now known as a1d3s|away
a1d3s|away is now known as a1d3s
gianMOD has quit [Remote host closed the connection]
gianMOD has joined #linux-sunxi
<a1d3s> what i have to do , to get an lvds touch working on banana pro?
<a1d3s> and why flickr the screen on hdmi since 3.19 final , in rc5 and 7 all was ok
gianMOD has quit [Ping timeout: 250 seconds]
cubear has quit [Quit: Leaving]
<a1d3s> where must this https://github.com/LeMaker/fex_configuration/blob/master/fex/banana_pro_7lcd.fex be added or is it not needed? for lcd 7"
ssvb has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
<ssvb> a1d3s: fex files need to go to https://github.com/linux-sunxi/sunxi-boards
<ssvb> a1d3s: then you may try to add the LCD support to u-boot - http://linux-sunxi.org/LCD
cnxsoft has quit [Quit: cnxsoft]
<ssvb> a1d3s: if it works in u-boot, then it will also work in the mainline kernel via simplefb
<a1d3s> this fex with 7" enabled is in sunxi-boards
<ssvb> where?
<a1d3s> BananaPro.fex looks like the same as banana_pro_7lcd.fex
<a1d3s> i see all lcd_ para
<ssvb> what did you use to compare fex files?
<a1d3s> ok i see some little diff https://www.diffchecker.com/2lg7odym
<oliv3r> allo all
<a1d3s> hi oliv3r
<oliv3r> mripard_: i directly solderd an sd card to the pins and I get perfect results now
<oliv3r> mripard_: but mrnuke's patch is required to function
<oliv3r> mripard_: are there plans to upstream it?
<a1d3s> ssvb sinovoip gave a new raspbian an told me they added working mali , but i cant find anything about mali in there
<a1d3s> first of all i want to get hdmi back working without flickering white lines
<a1d3s> and then the lvds lcd display :)
orly_owl has quit [Quit: leaving]
<a1d3s> the lcd doesnt work in uboot
<mripard_> oliv3r: mrnuke needs to be a bit more agressive on making it merged, but yeah, eventually, it will be merged
<oliv3r> mrnuke: get to work! :p ^
<oliv3r> mrnuke: i haven't done 'heavy' testing with it, but using mmc-spi, mounting, listing etc works, so enough data should be easily transfered
<oliv3r> mripard *
<oliv3r> so you can have my tested-by
<ssvb> a1d3s: about what to do with the fex files - http://linux-sunxi.org/New_Device_howto#Step_6:_Add_support_to_sunxi-boards
<ssvb> a1d3s: but now I see that the sunxi-boards repository has a number of open pull requests on github...
<mripard_> oliv3r: it's working, he just has to ping the maintainer so the patch is taken.
<ssvb> libv: what would be your opinion about sunxi-boards repository and the pull requests submitted there?
<ssvb> a1d3s: are you trying to get hdmi working with the mainline kernel?
<a1d3s> ssvb yes , u-boot is ok and at loading kernel configuration it begin flickering
gianMOD has joined #linux-sunxi
<a1d3s> in rc5 it get me my logon screen
<ssvb> a1d3s: do you mean that it looks like a regression in the kernel?
<oliv3r> mripard_: i'll try to find it on the ML then and i'll ping it with a tested-by; woudl that work?
<a1d3s> maybe
<a1d3s> i dont know why it begin to flickering , i got some white line horizontal
<ssvb> a1d3s: if it used to work before, then try to bisect it and ping mripard_ if you find the offending patch
vishnup has joined #linux-sunxi
<a1d3s> i will try an old .config for kernel rebuild and test again
<ssvb> if the hdmi display is fine in u-boot, but breaks in the kernel, then the kernel probably messes up some hardware register
gianMOD has quit [Ping timeout: 264 seconds]
<ssvb> a1d3s: do you know what git bisect is?
<a1d3s> no :)
FR^2 has quit [Quit: Connection reset by peer]
<a1d3s> ssvb i do this all just for fun and learn something :) as i say im not a dev
<a1d3s> will be back at home in 30 minutes
a1d3s is now known as a1d3s|away
<ssvb> a1d3s: however I would assume that hdmi would be broken for a lot of people if there is something wrong with it
gianMOD has joined #linux-sunxi
<ssvb> a1d3s: what is the exact kernel revision that you are using?
FlibberTGibbet has joined #linux-sunxi
<ssvb> oliv3r, mnemoc: you have been taking care of the pull requests for the sunxi-boards repository, right?
ricardocrudo has quit [Ping timeout: 244 seconds]
<libv> ssvb: about the meminfo rename: of course
<libv> pull requests?
<libv> are there device pages for those?
<libv> if not, no inclusion
* libv looks at recent changes
a1d3s|away is now known as a1d3s
<a1d3s> back :)
<libv> pfff, much to review :(
viccuad has quit [Quit: WeeChat 1.0.1]
<mrnuke> mripard_, oliv3r: I've re-sent that patch a bunch of times. The next thing to do is walk into Mark's house with a riple and yell at him "Merge that patch!". I'm out of moves
<mripard_> mrnuke: no, the next thing you need to do is send him an email asking whether he will merge that patch or not
<mripard_> I told you that already
F1skr has joined #linux-sunxi
reinforce has joined #linux-sunxi
wickwire has quit [Ping timeout: 250 seconds]
wickwire has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 250 seconds]
<vishnup> mripard: I saw A20 ps2 patches got included in linus' tree
<vishnup> thanks, my first patch is part of linux now :)
<mripard_> vishnup: cool :)
<vishnup> now, I'm trying to bring up my A33 tablet :)
<vishnup> mripard_: I could not attend FOSDEM, did they record videos? till last year they have all videos uploaded to video.fosdem.com
<libv> yes
<libv> videos were recorded
<libv> but they still need to be processed
<vishnup> libv: great!
<libv> and i think that this is going to be more work than it was last year
simosx has joined #linux-sunxi
bsdfox has quit [Ping timeout: 246 seconds]
<vishnup> they might need more time to upload, fosdem guys are really doing great thing
<libv> 23 devrooms
<ssvb> vishnup: maybe you can ping leviathanch, so that he could probably ask allwinner about the the status of libdram for a33
<libv> each generating 16GB of data per day
<libv> 16-20GB
bsdfox has joined #linux-sunxi
bsdfox has joined #linux-sunxi
<ssvb> vishnup: if I understand it correctly, he is working for allwinner - http://irclog.whitequark.org/linux-sunxi/search?q=leviathancn
wickwire has quit [Ping timeout: 252 seconds]
<Turl> oliv3r: as an alternative to it, you can use dma on spi O:)
<vishnup> ssvb: gr8 thing is I got libdrm source on github
<vishnup> and I merged that in mainline u-boot
<ssvb> vishnup: hmm, where from?
<ssvb> vishnup: does it already work?
<vishnup> ssvb: I got the u-boot command prompt
<vishnup> YES!!
<ssvb> very cool!
hehopmajieh__ has quit [Quit: Konversation terminated!]
<ssvb> A33 SDK 4.4?
<ssvb> would here be the right place for it http://dl.linux-sunxi.org/SDK/ ?
<vishnup> I'll send u-boot patch after cleanup and It'll take some time
<ssvb> excellent
<vishnup> ssvb: sure!, however there is no SDK package for A33
<vishnup> it's not provided by allwinner
<ssvb> this https://github.com/realthunder dude seems to have it though
bsdfox has quit [Ping timeout: 246 seconds]
<vishnup> ssvb: do you know who is he/she?
<ssvb> but that's a nice find
<ssvb> I have no idea :) I thought that you might know
<vishnup> our mutual friend google found him :)
<vishnup> and libdram source
<ssvb> anything for a23 and a80?
<mripard_> Turl: yeah, eventually, that's what we will end up doing
<vishnup> ssvb: no idea, as I don't have those devices I did not sarch
<vishnup> s/sarch/search
wickwire has joined #linux-sunxi
f15h has joined #linux-sunxi
<el3> ssvb: thanks. 12 hours later
Netlynx has joined #linux-sunxi
Netlynx has quit [Changing host]
Netlynx has joined #linux-sunxi
domidumont has joined #linux-sunxi
imcsk8 has quit [Quit: Reconnecting]
imcsk8 has joined #linux-sunxi
<ssvb> el3: you are welcome
<ssvb> please keep us informed about your progress with resolving this memory corruption issue
<mrnuke> mripard_: roger that
<ssvb> el3: it might be a defective memory module, or it might be just dram misconfiguration
<ssvb> el3: the latter actually tends to be not too uncommon, and may be resolved by reducing the dram clock speed in u-boot
<mripard_> mrnuke: it's probably a bad time to do that though since we're in the middle of the merge window
<mripard_> but ping him as soon as 3.20-rc1 is out
<el3> ssvb: I built the kernel myself, so misconfiguration is very likely
<ssvb> el3: the dram controller is configured in u-boot
<el3> ssvb: I will try to edit the u-boot. is it dram=something?
<ssvb> el3: depends on the u-boot version
<el3> setenv bootargs console=ttyS0,115200 root=/dev/mmcblk0p2 rootwait panic=10
<el3> fatload mmc 0 0x43000000 script.bin || ext2load mmc 0 0x43000000 script.bin
<el3> fatload mmc 0 0x48000000 uImage || ext2load mmc 0 0x48000000 uImage uImage
<el3> and at last bootm 0x48000000
<el3> ssvb: this is the boot.cmd file
<mrnuke> mripard_: alrighty
<leviathanch> ssvb: hey
<leviathanch> hi
<mrnuke> oliv3r: can I add you on Tested-By: in the c ommit message on that patch?
<leviathanch> ssvb: libdram is partially closed
nuuhku has joined #linux-sunxi
<leviathanch> but we already have made contact between the guy who has built the DRAM controller and the guy doing the upstream code work
<ssvb> el3: looks like I need to add a new section to http://linux-sunxi.org/Hardware_Reliability_Tests about how to resolve problems when they are detected :-)
<leviathanch> ssvb: so the maintainer doing the mainline work from sunxi will write what he can do, and the engineer from 全志科技 will fill in the gaps
<leviathanch> so that at the end we have a working driver
<ssvb> el3: thanks for demonstrating that it is actually needed
FlibberTGibbet has quit [Quit: Leaving]
<ssvb> leviathanch: do I understand it right that somebody from allwinner going to contribute (a80?) dram support to the mainline u-boot?
<el3> ssvb: :) I probably a bit noob. I did not build the mainline kernel, and wanted to run debian. Thanks for patience
<leviathanch> ssvb: yes
<el3> ssvb: Actually I needed to run g_mass_storage, and the images on linksprites site dont support that
<leviathanch> ssvb: the only problem is language, the engineer who built the DRAM controller doesn't speak English
<leviathanch> ssvb: he only understands Chinese
<ssvb> leviathanch: do they have documentation for the dram controller?
<leviathanch> ssvb: frankly, no
<leviathanch> ssvb: and if there would be, there would only be in Chinese since they only know Chinese
<leviathanch> ssvb: but actually they just didn't write any
<leviathanch> it's the guy who has built the controller who has written the libnand
<leviathanch> and this guy doesn't understand your emails well
<ssvb> leviathanch: based on what we know, Allwinner is likely licensing the DRAM controller IP, with the primary suspect being Synopsys DesignWare
<ssvb> leviathanch: if that's the case, then there must be English documentation for it somewhere
steeve has quit [Read error: Connection reset by peer]
<wens> leviathanch: you're up late :p
wens has quit [Quit: Reconnecting]
wens_ has joined #linux-sunxi
<leviathanch> wens: I'm back in Switzerland again ;-)
<leviathanch> wens: I'm back in Switzerland again ;-)
<leviathanch> so no, it's 6 p.m.
<wens_> hmm
wens_ is now known as wens
<ssvb> leviathanch: now what you are saying about this dram guy is a little bit confusing and does not quite add up
<wens> (why do i have a 2 sec. delay on this channel)
<wens> leviathanch: well i'm off to bed, it's 1am
<leviathanch> ssvb: yes, but they can only give us the parts which they can give us legally
<leviathanch> ssvb: the things they can not give us because they have contracts with another company is prohibited to them to hand it over to us
<wens> i don't suppose that includes naming the vendor and part they got?
<leviathanch> wens: nah, that would be ok
<leviathanch> ssvb: you are Vishnu, right?
<ssvb> leviathanch: no, I'm not Vishnu
<wens> considering synopsys stuff is used a lot in TI chips, and docs are freely available
<leviathanch> ah, ok
<wens> confirming it is a synopsys core would help a lot
<leviathanch> Vishnu is working together with 全志科技 right now in order to mainline the DRAM code
<leviathanch> ok
<wens> or hansg would be banging his head at arm asm
<leviathanch> just keep in mind that they usually only understand 1/3 of your emails
<ssvb> leviathanch: right, vishnup is here in this channel and working on a33 bring up
<quitte_> it sounds like allwinner is actually making good on their promise oc community support?
<leviathanch> so just write simple questions, so that they have a chance to get a dictionary
<leviathanch> quitte_: yes
<el3> ssvb: ./lima-memtester 100M --> -bash: ./lima-memtester: Is a directory
<leviathanch> there is even a pic of me on a Chinese wedding btw ^_^
<quitte_> great. unexpected but excellent
<leviathanch> and I already have a desk reserved there, I just don't have a work permit for China yet
<wens> leviathanch: i actually did bilingual in my last email, but no reply yet
<wens> leviathanch: dealing with the gov can be ... tricky ...
<leviathanch> wens: whom did you write?
<ssvb> el3: hmm, try http://linux-sunxi.org/Hardware_Reliability_Tests#DRAM instructions again
<el3> ok
<wens> leviathanch: kevin and sugar, and someone else they cc-ed last time
<ssvb> el3: maybe you have cloned the git repository into this directory instead of downloading the static binary?
<leviathanch> wens: right now they have Chinese new year, many people have hollidays
nove has joined #linux-sunxi
<wens> leviathanch: yeah, as i suspected :(
<wens> the great migration of people, a very scary thing
<leviathanch> wens: if you wrote during the last two weeks it might be very well possible that he is at his family already and doesn't read emails
<leviathanch> wens: you know, I was at their place, they live in a building next to their office building during working season
<leviathanch> and go home to their families during hollidays
<wens> hmm....
<leviathanch> and there are regions with no internet
<WarheadsSE> ^
<WarheadsSE> large regions
<leviathanch> also they work every day for months, so hollidays means no work
<wens> leviathanch: yeah, i know a lot of people come fron inland to the coastal cities for work
<leviathanch> yes, exactly
<wens> work everyday... wow...
<leviathanch> wens: he will answer
<leviathanch> wens: yes, they totally wanna have all the code mainlined, so that they have less work to do at allwinner
<vishnup> leviathanch: I could get basic mainline u-boot on A33, just command prompt.
<wens> anyways, i'm waiting for the merge window to close before sending out patches
<el3> ssvb: got it. This -> chmod +x lima-memtester should be chmod +x ./lima-memtester.2 and then ./lima-memtester.2 100M
<wens> leviathanch: should be plenty of time :)
<vishnup> now mainlining it will be easy
<leviathanch> perfect
<leviathanch> so communication with Allwinner really worked out
<leviathanch> 我爱 ^^
<leviathanch> vishnup: about the Win10
<leviathanch> I was already having dinner with the cubietech people
<leviathanch> uh, also
<vishnup> great
<leviathanch> and I have done some sales for them in Switzerland as well
<leviathanch> vishnup: Digitec wants Win10 on the cubies then they sell it as desktops
<el3> ssvb: Can you look at this http://pastebin.com/0ynSKP2k After that putty crashed
<leviathanch> up to thousand they said
<wens> sounds like allwinner is much better wrt open source than mediatek, though the workforce size is the other way around
<leviathanch> wens: yes
<leviathanch> wens: Allwinner is heavily overworked
<leviathanch> they lack manpower
<leviathanch> I've already have a desk there, just no work permit for China
<leviathanch> :-/
<ssvb> el3: you can try it again, but it looks like lima-memtester is tough enough to knock down your hardware almost instantly
<WarheadsSE> heh
<el3> yea
<wens> mediatek lacks manpower as well, and i think they doubled over the past year or so :/
<vishnup> leviathanch: great, If I get cubie and windows 10 platform builder stuff, I'll be happy to work it on !
gianMOD has quit [Remote host closed the connection]
<el3> ssvb: should I start another session at same time and tail syslog maybe
<wens> my friend there works 3+ projects, alone, no backup
<leviathanch> vishnup: ask Martin, he already has booted win10 on an a80
<vishnup> leviathanch: why don''t they hire sunxi community guys ? ;)
<vishnup> leviathanch: thanks, I'll
<leviathanch> vishnup: they do hite, now it's about the gov who has to give me a work permit for China
<leviathanch> *hire
<ssvb> el3: well, if you confirm that lima-memtester detects problems very fast, then the next step is to tweak dram configuration in u-boot until it works and can pass the tests
<wens> vishnup: language barrier i guess, plus you would have to want to relocate?
<T0mW> el3: I just resolved a memtester error on my A20 system here, basically, the 'sunxi dram odt_en value' had to be set to '1'
<leviathanch> wens: I'm learning Chinese right now (I can already write and read)
<leviathanch> wens: so they offered me a desk and now it's about a permit
<el3> ssvb: okey I will try one more time
<T0mW> odt_en turns on the SDRAM controller resistive termination.
<wens> leviathanch: :)
<leviathanch> vishnup: start learning Chinese, then you get a job there easily :)
<f15h> I think odt_en turns on on-die termination in the dram chips ...
<vishnup> leviathanch: it's hardest language I think! ;)
<leviathanch> vishnup: just learn the basics, the rest you learn while working with them
<leviathanch> vishnup: noo... actually it's more like a programming language
bonbons has joined #linux-sunxi
afaerber__ has joined #linux-sunxi
<leviathanch> vishnup: for example the comparation "operator" 比
<vishnup> leviathanch: I'll see
<leviathanch> <object1>比<object2><attribute>
<leviathanch> <object1> < <object2> <attribute>
nuuhku has quit [Quit: Page closed]
<leviathanch> vishnup: 你 好 吗
<leviathanch> you good <question particle> <--- read operation
<leviathanch> it's like a programming language
<leviathanch> it's easier to learn than any other language I've ever encountered
<leviathanch> that's why they have such trouble learning English
<leviathanch> and give up
<Turl> my try at learning chinese with leviathanch : sunxi比mediatek偉大
<vishnup> leviathanch: I'll look into it.
<el3> leviathanch: Are you chinese?
afaerber_ has quit [Ping timeout: 245 seconds]
<leviathanch> el3: no, I'm European
<T0mW> el3: I had to edit the u-boot-2015.01 file 'board/sunxi/dram_sun5i_auto.c', comment out the dram_para and replace it with this:http://pastebin.com/xzg752P1
<T0mW> el3: my ram errors went away.
<leviathanch> el3: I'm from Zürich, Switzerland
<el3> leviathanch: okey me too. You say chinese is easy to learn?
<ssvb> T0mW: you could have added the changes to the defconfig file for your board
f15h has quit [Remote host closed the connection]
<ssvb> T0mW: what kind of hardware is that?
<el3> T0mW: Okey I might try that after the test. or u-boot file
<T0mW> ssvb: the tpr0..tpr2 needed changes and the config doesn't have entries for them.
<el3> leviathanch: okey greetings from Denmark
<T0mW> ssvb: also cas
<ssvb> T0mW: what kind of changes?
<T0mW> el3, ssvb: the important parameter seems to be CONFIG_DRAM_ODT_EN.
<leviathanch> el3: yes, Chinese is easy to learn
<ssvb> T0mW: yes, but not tpr0..tpr2 and cas?
<leviathanch> el3: I had a six months basic course (every week 4 hours)
<T0mW> ssvb: I'm not on a cubie board
<leviathanch> el3: and already can talk rudimentary
<leviathanch> el3: you should try it :D
<ssvb> T0mW: you can add CONFIG_DRAM_ODT_EN to defconfig
<T0mW> ssvb: I'll try reverting tpr0..tpr2 back to defaults
<leviathanch> el3: of these 8300 signs you actually only need around 150
<T0mW> ssvb: nod, did that.
<leviathanch> el3: then you can read most of the texts
<el3> leviathanch: :D Maybe I should learn thai instead, since my girlfriend is thai
<T0mW> ssvb: the SDRAM on my board is DDR800
<leviathanch> el3: learning Chinese is awesome too
orly_owl has joined #linux-sunxi
<el3> leviathanch: ok? sounds alot easier than it looks
<leviathanch> el3: jup. absolutely
<T0mW> ssvb: doh, no, it is 1600
<ssvb> T0mW: do you have the exact ddr3 chip name and a datasheet for it?
<ssvb> T0mW: oh, 1600 should be perfectly fine
<T0mW> ssvb: wish I could find a data sheet, lots of taobau hits on GT8UB256M16BP-BH
* T0mW going to default u-boot + ODT_EN = 1
<ssvb> T0mW: well, DDR3-800 would be unusual, but 1600 should be perfectly fine
<ssvb> T0mW: You have seen the http://linux-sunxi.org/A10_DRAM_Controller_Calibration page, right?
HeavyMetal has quit [Quit: BNC Services Provided by The ASoTnet IRC Network.]
<T0mW> ssvb: yeah
paulk-collins_ has joined #linux-sunxi
<ssvb> T0mW: do you have plans to contribute aw-som support to the mainline u-boot?
<ssvb> T0mW: what are you current dram settings based on?
<ssvb> *your
<el3> ssvb: How long should I let memtest run? Its loop 3 now
<T0mW> ssvb: this is the run with the default values for tpr0..tpr2 with ODT_EN=1, emr=0, tpr3=0, dqns=0 http://pastebin.com/hUVMfpCS
<ssvb> T0mW: you can run the meminfo tool from sunxi-tools to check whether the odt_en parameter change has taken effect
naobsd has quit [Quit: naobsd]
paulk-collins has quit [Ping timeout: 255 seconds]
<leviathanch> ssvb, el3: I'm having the newest Cubie4 now, with huggs and greatings from the (female, cute and single) project manager from cubietech :)
<ssvb> el3: so it runs long enough without detecting any problems on the second try?
<T0mW> ssvb: I took the settings from: http://linux-sunxi.org/Mainline_U-boot#Failsafe_DRAM_settings.2C_based_on_standard_JEDEC_timings
<el3> ssvb: there were some kernel messages in loop 1. 2,3,4 no messages
<T0mW> ssvb: added the ODT_EN=1 to enable terminations and it works, turn off terminators and it fails.
<ssvb> T0mW: emr1=4 is usually better than emr1=0
<T0mW> ssvb: AH, thanks, will try that.
<ssvb> T0mW: with the latest u-boot from git, you should be able to set appropriate timings for the clock frequency
<el3> leviathanch: Oh nice. I have pcduino3 here. Trying to debug. And learn alot hopefully in the process
<Turl> leviathanch: is benn still in cubietech?
<ssvb> T0mW: be sure to verify the settings with the meminfo tool
<leviathanch> Turl: No, Benn is now at Allwinner. I met him at the cntrance
<leviathanch> *entrance
<T0mW> ssvb: emr1=4 works fine
<Turl> leviathanch: he went back to AW? interesting
wickwire_ has joined #linux-sunxi
<jero> hi, how can I figure out why my sun4i produces the following when using fbturbo: FBTURBO(0): No sunxi-g2d hardware detected (check /dev/disp and /dev/g2d)
<ssvb> el3, T0mW: I have a meeting in a pub today, and I'm already rather late
<T0mW> ssvb: I tried the default jedec settings and nothing worked. BUT, it is a bit confusing just what those 'sunxi dram timings' selections are for. The u-boot menu fields for EMR1 ODT etc. don't change.
<Turl> jero: did you check /dev/disp and /dev/g2d?
<T0mW> ssvb: I'll submit the AWSOM later
wickwire has quit [Ping timeout: 245 seconds]
<jero> Turl: well, they are here!
<Turl> jero: and you can r/w them?
<ssvb> T0mW: be sure to double check everything and also verify with the meminfo tool (just to be sure that the u-boot settings are properly taken into use)
<jero> ssvb: I run mainline uboot on awsom a-10
<ssvb> T0mW: anyway, I really need to go now, see you later :)
<ssvb> jero: that's cool, T0mW is tweaking the dram settings for it at the moment
<ssvb> jero: maybe you two can cooperate
<el3> ssvb: okey. Maybe T0mW can guide me a bit, how to change the u-boot or something else. I dont even know if memory is the problem. Running loop 6 now. all ok
<T0mW> jero: I've the A20 DIMM here
<jero> ssvb: for the A10 I just had to do http://www.fpaste.org/184824/76335814/
<jero> T0mW: ^
<el3> ssvb: But cheers. Have a nice evening :)
<T0mW> jero: did you run memtester or lima-memtester?
<ssvb> el3: thanks
<jero> Turl: yeah, access to both devices is fine. testing as root. can they be tested individually?
<plaes> el3: are you using mainline kernel?
<jero> T0mW: didn't try, I can test that
<el3> plaes: no
<plaes> does memtester work on mainline?
<T0mW> jero: I thought that everything was okay until I started getting some system deadlocks. Then ran the memtester.
<plaes> mainline-kernel
<el3> plaes: I think lima-memtester does http://linux-sunxi.org/Hardware_Reliability_Tests
wickwire_ has quit [Quit: Leaving]
simosx has quit [Quit: Leaving]
<jero> T0mW: testing now :)
khuey|away is now known as khuey_
<Turl> jero: maybe you have selinux or something being annoying
ssvb has quit [Ping timeout: 252 seconds]
<jero> Turl: no, nothing like that
<Turl> jero: paste a full log and ping ssvb then
<montjoie[home]> selinux is never annoying!!!
<Turl> montjoie[home]: if you do setenforce 0 it's not
<Turl> :P
<el3> Okey the lima-memtester runs ok still. Loop 8.. Wonder what to do next
<plaes> hrm.. memtester is sunxi-3.4 :S
<oliv3r> Turl: the other SPI ports don't do DMA?
<mripard_> oliv3r: none of them do
domidumont has quit [Ping timeout: 246 seconds]
<mripard_> it's the same driver for all of them
quitte_ has quit [Ping timeout: 240 seconds]
bengal has joined #linux-sunxi
gianMOD has joined #linux-sunxi
bengal has quit [Remote host closed the connection]
bengal has joined #linux-sunxi
<el3> lima-memtester ran ok. But as usual sytem runs for 10 to 30 misn and these messages pop up in syslog http://pastebin.com/Ujid7daU
gianMOD has quit [Ping timeout: 250 seconds]
<el3> and every 10'th min wpa_supplicant[1651]: wlan0: WPA: Group rekeying completed with 00:02:61:7b:16:60 [GTK=TKIP]
<jero> Turl: you mean the full Xorg.log ?
arossdotme has quit [Ping timeout: 245 seconds]
_massi has quit [Remote host closed the connection]
<jero> does current xf86-video-fbturbo master run with sunxi-3.4 kernel and maliver r3p0 ?
<jero> on allwinner a10 ?
arossdotme has joined #linux-sunxi
<jero> Turl: is cedarx required in order for fbturbo to work ?
<el3> T0mW: Where do I fin the file that I need to set the dram parameter to 1?
<nove> jero: no
<jero> nove: no for cedarx?
<nove> jero: yes
<T0mW> el3: You mean the ODT_EN ?
<el3> T0mW: Yes
<T0mW> el3: that is in the u-boot 'make menuconfig' under 'ARM architecture' and it is 'sunxi dram odt_en value'
<jero> is G2D acceleration the one I want on sun4i ? I can't get it to activate itself.
<el3> T0mW: Okey can I do that on the pcDuino3 while system is running?
forest has joined #linux-sunxi
<el3> T0mW: or do I have to build a new image?
<T0mW> el3: you can scp the new u-boot into the filesystem of the board while it is running, then dd it onto the boot block. For example, I am booting from MMC0, so I first copy the new u-boot into my board with 'scp if=u-boot-sunxi-with-spl.bin root@192.168.20.10', then ssh into the board (ssh root@192.168.20.10) and install the new u-boot (dd if=if=u-boot-sunxi-with-spl.bin of=/dev/mmcblk0 bs=1024 seek=8)
<T0mW> el3: you don't need to build a new image
vishnup has quit [Remote host closed the connection]
<el3> T0mW: Thank you for your patience. I will try if I can do that
<T0mW> el3: np, this stuff takes time to learn
viccuad has joined #linux-sunxi
<el3> T0mW: Yea, hopefully I will get pro at this :)
<nove> jero: this is just a guess, but i remember something about a script.fex parameter (fb0_framebuffer_num) i think
gianMOD has joined #linux-sunxi
Netlynx has quit [Remote host closed the connection]
<jero> nove: currently my problem is sunxi_disp_init() returns NULL, trying to figure out why.
<nove> jero: also, Turl was asking for full xorg.log
<jero> nove: Turl: I put it here: http://www.fpaste.org/184865/76647414/
<jero> nove: the first bad thing to me seems to be "failed to enable the use of sunxi display controller"
<el3> T0mW: http://linux-sunxi.org/U-Boot#Compilation I did runthe 2 make commands. But now when I try make menuconfig it says: make: *** No rule to make target `menuconfig'. Stop.
<nove> jero: maybe this -> hardware: (video memory: 4500kB)
FlibberTGibbet has joined #linux-sunxi
<jero> nove: ah, is that a setting ?
<jero> or bad?
FreezingAlt has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
orly_owl has quit [Quit: leaving]
HeavyMetal has joined #linux-sunxi
HeavyMetal has quit [Excess Flood]
FreezingCold has quit [Ping timeout: 245 seconds]
HeavyMetal has joined #linux-sunxi
<nove> jero: yes, there are some kernel command options to set the amount http://linux-sunxi.org/Kernel_arguments
Andy-D has quit [Ping timeout: 256 seconds]
<T0mW> jero: looks like the A20 DIMM (and probably the A10) will need sunxi dram emr1 = 4, dram odt_en = 1, tpr3 = 0, dqs_gating_delay = 0, mbus = 300MHz. Then, choose either 'sunxi dram timings' that are not the 'Magic'. Set your dram clock to any multiple of 24 (e.g. 408, 432, 456).
<T0mW> jero: I've yet to verify, but it looks like I've got DDR1600 running at 640MHz.
exslestonec| has joined #linux-sunxi
exslestonec has quit [Read error: Connection reset by peer]
<el3> I am probably in wrong directory
<jero> T0mW: normally it's a 408
<T0mW> jero: yeah, but I think that awsom set it at that because it worked for them.
<jero> ah
<jero> T0mW: can I get this from the ram chips?
<T0mW> jero: 'this' ?
<jero> T0mW: sorry; the proper dram specs and timings
<jero> T0mW: it's weird; besides fbturbo, everything works fine.
gianMOD has joined #linux-sunxi
<T0mW> jero: not sure that you can, IIRC, the "standard" DIMM memories have an EEPROM on them with that data.
gianMOD has quit [Remote host closed the connection]
FlibberTGibbet has quit [Quit: Leaving]
gianMOD has joined #linux-sunxi
<T0mW> ah, I see what has been happening! If you start with a stock u-boot config (e.g. make Cubieboard2_defconfig), then the 'sunxi dram timings' defaults to 'Magic vendor timings' using magic numbers (see: board/sunxi/dram_sun5i_auto.c), those values would only work if you set the ODT_EN (termination enable).
khuey_ is now known as khuey|away
<T0mW> without that, select any of the JEDEC under 'sunxi dram timings', then set your DRAM clock (according to list within: board/sunxi/dram_timings_sun4i.h), then, you get proper values and ODT_EN is irrelevant.
VargaD_ has joined #linux-sunxi
<T0mW> a bit confusing.
<jero> T0mW: i'll try tweaking the kernel cmdline first :)
gianMOD has quit [Remote host closed the connection]
VargaD has quit [Ping timeout: 264 seconds]
VargaD_ is now known as VargaD
<el3> I am considering building a whole new kernel. Mainline this time and mainline u-boot. See if that goes better. But last time I tried to compile it, I got the error something like- This compiler is known to compile errors or something
Zboonet has quit [Remote host closed the connection]
<el3> I installed the tool like this apt-get install gcc-arm-linux-gnueabihf
<T0mW> el3: I'm building my stuff using Yocto
<el3> T0mW: okey. Maybe I should try that
<T0mW> el3: Yocto builds out an entire system. Takes time to understand how to use it, but there is a lot of good info on it. However, I've been using Yocoto (OpenEmbedded) for many years now.
<el3> T0mW: LOL just watching the introduction vid on their page. The frankenstein OS is exactly what I got right now. This must be made for me :D
<el3> T0mW: Thanks for advice. I guess I will be moving to yocto, looks nice
<T0mW> el3: there is an irc channel for it, #yocto
<el3> T0mW: Thank you
<T0mW> el3: learning curve is a bit steep, but it beats the heck out of buildroot.
<T0mW> el3: core-image-minimal should get you going, once you setup your MACHINE definition.
<el3> T0mW: okey. I suppose I could follow some tutorials and youtubeat first. And get some guidance on their irc. I just need a headless system where I can load g_mass_storage
<jero> T0mW: this is what meminfo gives me on my A10: http://www.fpaste.org/184905/71688142/
tmp100 has joined #linux-sunxi
<T0mW> jero: thanks
<T0mW> jero: pretty much what I got from an A20 DIMM they sent me with a Ubuntu image on it.
<jero> T0mW: does meminfo get that from the dram or does it print the current hand-made settings ?
<T0mW> jero: how did memtester work out?
<jero> T0mW: pretty good
<jero> i'm looking for a way to bump the (II) FBTURBO(0): hardware: (video memory: 4500kB)
<T0mW> jero: heh, pretty good does not cut it, ;) , pass / fail?
<jero> T0mW: it passed it all
<T0mW> great
<T0mW> :)
nuuhku has joined #linux-sunxi
gianMOD has joined #linux-sunxi
FlibberTGibbet has joined #linux-sunxi
<jero> T0mW: video memory: 4500kB makes sense given my screen size (800x480)
<leviathanch> Turl: do you know Ben personally?
<leviathanch> Turl: I'd love to invite some Allwinner folkfs to the next chaos communication congress in Hamburg in order to make a presentation about the newest community project (this sunxi foss tablet)
<leviathanch> but since I only met Ben once at the entrance it's hard to make contact with him now
<leviathanch> I'd have to go over my other friends at Allwinner
<leviathanch> Turl: and he was like "what the fuck does this guy want here?" kind of reacting
<leviathanch> before lixiang came and explained ^^'
<nove> HKG15-109: 96Boards Software Requirements https://www.youtube.com/watch?v=e8_MatJ_VR0
<nove> leviathanch: talk allwinner in joining this www.96boards.org
<leviathanch> one is going
<leviathanch> uhm... his name
<leviathanch> uh
<leviathanch> do not remember right now how I spell it right
<leviathanch> huxiang something
<leviathanch> anyway
<leviathanch> he is going
<leviathanch> asked me if I had come for the Linaro event as well
<leviathanch> but my visa wouldn't allow me to go to Hong Kong an back
<leviathanch> so I couldn't go there
<leviathanch> but Allwinner is there already so to say
<nove> leviathanch: are you saying that allwinner has plans to make an 96boards
<nove> ?
<leviathanch> hmm
<leviathanch> we are planning to make a fully open source tablet for the community
<leviathanch> we will be done mids of next year
<leviathanch> one step has already begun by using me as kind of a dolmetscher and sharing information
<leviathanch> nove: yes
<leviathanch> now that I've googled it
<leviathanch> yes
<leviathanch> actually... about the board part
<leviathanch> you're just talking with the guy who has gotten the job to design the schematics for it
<leviathanch> half of next year you can expect us to release a fully open source board and stream all the drivers for it into mainline
<leviathanch> they are totally into it, they just only talk Chinese, thats why you never hear something from their side, that's all
<leviathanch> as soon as you talk Chinese to them they are totally cooperative
<leviathanch> and since the board only costs around 20$ since I build it from SEG parts, I can afford giving out some of them for free :-)
<leviathanch> actually I estimate even only 15$
<nove> leviathanch: where this "sunxi foss tablet" idea came from?
gianMOD has quit [Remote host closed the connection]
FreezingAlt has quit [Read error: Connection reset by peer]
FreezingCold has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
gianMOD has joined #linux-sunxi
ssvb has joined #linux-sunxi
<leviathanch> nove: I basically wanted to build an open source tablet
<leviathanch> and then we decided to share all information with you guys
<leviathanch> ^^
<leviathanch> actually it was their idea to do it with all the sunxi members
<leviathanch> in turn they promised to open up as much as they could
<leviathanch> only limiting factors are language barrier (which can be overcome) and licensing with third party contractors
<leviathanch> and since I'm now a part of the Allwinner crew in some way... let's say: *we* decided to do a sunxi foss tablet
<leviathanch> I will send all the relevant maintainers a version of the first board as soon as I've got them ready in Shenzhen
gianMOD has quit [Remote host closed the connection]
khuey|away is now known as khuey_
<leviathanch> nove: is that ok? ^^
<nove> leviathanch: do you know the vivaldi tablet history? that failed but it look that maybe will be tried again, and the firefox os tablet, and jolla, and ubuntu
<nove> leviathanch: what about if all work together?
<leviathanch> nove: I don't only know it, I was involved
<leviathanch> nove: Aaron was at my place multiple times
<leviathanch> nove: and I was at his home
<leviathanch> nove: we hang up together multiple times
<leviathanch> nove: and I layouted the improv board
gianMOD has joined #linux-sunxi
<leviathanch> nove: and there are special reasons why it failed and I told him that I would fail because of this mistakes he did and he wouldn't listen
<leviathanch> nove: and btw, I will use plasma active as desktop environment for the tablet
<leviathanch> nove: he and I discussed this to a beer and some vegan lasagne some time ago and he told me I'd be ok
<leviathanch> nove: he's into the sales share
<leviathanch> I came into the Allwinner office and I guess everyone had a Stephe Jobs impression of me
<leviathanch> because I had multiple mind blowing ideas for them ready
<leviathanch> and they wanted me to stay and work for them right ahead, but I still don't have a work permit for China yet
<nove> leviathanch: the problem with yet another tablet is the software, if will have only android it will not be any difference from what allready exists
<leviathanch> nove: nope, planned is mer-project, with plasma-active as interface
gianMOD has quit [Ping timeout: 264 seconds]
<leviathanch> power vr driver will first be provided by binary rpm
<leviathanch> later on we will upgrade onto the foss power vr driver for rogue
gianMOD has joined #linux-sunxi
<leviathanch> the engineer responsible for the driver at Allwinner just didn't understand the online new saying that the micro code had been published for Rogue already
<ssvb> leviathanch: "the foss power vr driver for rogue", was there something lost in translation?
<leviathanch> so now 上一站: we send all the Allwinner drivers whichs taking all our time forwardporting into mainline, 上二站: we go over providing a driver for the gpu within A64
<leviathanch> no
<leviathanch> ssvb: that's actually which they figured would be a cool idea
<leviathanch> to write a foss driver for the powervr rogue platform within A64
<leviathanch> but first we have to pass stage one
<leviathanch> sending all the drivers mainline so that we have time writing other code
<ssvb> leviathanch: is the imgtec aware of these plans?
<leviathanch> imagine has published the micro code
<leviathanch> so they will not care if we implement a driver based on it
<leviathanch> also imagine has a building around the corner of the Zhuhai office
<leviathanch> so we could go over and ask xD
<leviathanch> but I don't think they will have a problem if we do work for them
<leviathanch> many companies in China have realized: sending code mainline safes time maintaining
<leviathanch> the only issue is the time to send it mainline in the first place
<leviathanch> many companies don't have that manpower
<leviathanch> I imagine img doesn't either
<leviathanch> btw, I've just sent out the first release of the presentation about a new product line of Allwinner
<leviathanch> let's hope my Chinese isn't that bad that it scares the management off
<leviathanch> ^^'
forest has left #linux-sunxi ["Leaving"]
<el3> leviathanch: No problems, just google translate :D
<leviathanch> el3: oh no, that doesn't work at all
<leviathanch> Chinese is soo different to English
<leviathanch> believe me... I now know since I learned the basics
<leviathanch> it's like using google translate to translate prosa into C-code
<el3> leviathanch: hehe I know, I tried on thai. Makes no sense what comes out the other end, and if you translatee it back, it makes even less sence
<leviathanch> el3: exactly :)
FreezingAlt has joined #linux-sunxi
<nove> leviathanch: if that tablet idea goes forward, talk to mozilla, ubuntu and tizen, and any others, maybe they want to be part
orly_owl has joined #linux-sunxi
<nove> leviathanch: and you know, if all together, all winner
<leviathanch> nove: that would be awesome, if they'd provide alternate desktops or help us out with userspace software
<leviathanch> nove: haha, yeah!
<leviathanch> ^^
<el3> T0mW: or anyone who kows this. Can I use a build, that I made with yocto for comercial purpose for free?
<el3> T0mW: sry wrong channel.
<leviathanch> nove: Chinese people are such social and friendly people, together is nearly always a good idea for them, 这是我爱他们的什么
<leviathanch> that's what I love about them ^^
<T0mW> el3: almost all of that stuff is under various licenses: GPLv1, GPLv2, MIT, BSD, etc.
FreezingCold has quit [Ping timeout: 250 seconds]
<el3> T0mW: okey thanks. MIT I know :)
<leviathanch> nove: I just wait. they are just totally stressed. hopefully we can driver by driver send all the code upstream soon, so that they have more time to breath
akaizen has quit [Remote host closed the connection]
<leviathanch> damn, it's one day and I already totally miss China
akaizen has joined #linux-sunxi
<el3> leviathanch: Switzerland is also one of the most expensive countries, along with Denmark and Norway. It is nice to travel a place where you can eat out for very little money, and just spend more on drones and fun stuff :D
<leviathanch> actually, nearly everyone can do it there
akaizen has quit [Ping timeout: 252 seconds]
<leviathanch> food doesn't cost anything
<leviathanch> there
<viccuad> Hi guys, I'm trying to build the mainline scripts ('make scripts') and it gets an error: "scripts/sortextable.c:31:32: fatal error: tools/be_byteshift.h does not exist". The weird thing is that that header exists. Could anybody share a light there? thanks in advance
bonbons has quit [Quit: Leaving]
pmattern has quit [Quit: Genug für heute.]
<Turl> oliv3r: with the current mainline, none of them do spi, the hw itself can do spi
<Turl> jero: yes
<Turl> jero: it should work, it doesn't have any dependency on mali either
<Turl> jero: nor on cedarx
<leviathanch> I'm trying to get hans' mainline patches for A80 to work right now for the cubie4
<leviathanch> together with the Allwinner folks that should be possible
<leviathanch> but till I can design the PCB for the A64 it's still some time
<Turl> leviathanch: I've never met benn in person, but we've chatted on irc
<leviathanch> Turl: ah, ok... I met him ^^
<leviathanch> meeting these people already was worth the journey :D
<Turl> oliv3r: err s/spi/dma/
akaizen has joined #linux-sunxi
naobsd has joined #linux-sunxi
paulk-collins_ has quit [Quit: Quitte]
Froolap has quit [Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/]
FreezingAlt is now known as FreezingCold
naobsd has quit [Quit: naobsd]
tmp100 has quit [Quit: Page closed]
viccuad has quit [Quit: WeeChat 1.0.1]
bengal has quit [Ping timeout: 244 seconds]
gianMOD has quit [Remote host closed the connection]
nove has quit [Quit: nove]
sehraf has quit [Quit: www.miranda-fusion.de ... be part of it...]
FlibberTGibbet has quit [Quit: Leaving]
nuuhku has quit [Quit: Page closed]
F1skr has quit [Quit: WeeChat 1.1.1]
f15h has joined #linux-sunxi
uniqdom has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
f15h has quit [Remote host closed the connection]
<uniqdom> ssvb, I'm using your sunxi-bootsetup-prototype. I have written the sdcard and done the uboot installation, but after the reboot it stays in "Starting kernel ..."
<uniqdom> I'm using the UART port
<ssvb> uniqdom: oh, this explains it, I think the debian installer uses only one console
<uniqdom> ok... i have a HDMI tv here... but no usb keyboard :(
<uniqdom> does your .img have a ssh daemon running?
<ssvb> you can mount the first partition of the sd card and, edit boot.cmd to change the /dev/tty1 to /dev/ttyS0 and re-generate boot.scr
<ssvb> after this the debian installer will use serial console instead of HDMI console
<uniqdom> Now I see the debian-installer in HDMI... I will try that tty change
orly_owl has quit [Quit: leaving]
<ssvb> in fact it probably makes sense to change this image to use the serial console for debian installer
<uniqdom> :)
<ssvb> because the current debian itself is non-workable with hdmi and the serial console is the only proper way to use it (until the ssh daemon is set up)