leviathanch2 has quit [Quit: No Ping reply in 180 seconds.]
leviathanch2 has joined #linux-sunxi
<ccaione>
wingrime2: pong
<wingrime2>
ccaione: on what branch I should base my mainline patch
<wingrime2>
ccaione: dts get modification on every our patch
<wingrime2>
ccaione: on what base I send it
leviathanch2 has quit [Ping timeout: 240 seconds]
Night-Shade has joined #linux-sunxi
hipboi has quit [Ping timeout: 240 seconds]
netlynx has joined #linux-sunxi
<wingrime2>
Turl: I find issue, I just forget add pinctrl to dts
<wingrime2>
Turl: so, irrecord --device=/dev/lirc0 MyRemote looks works
akaizen has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
maksimlin_away is now known as maksimlin
akaizen has quit [Ping timeout: 265 seconds]
maksimlin has quit [Quit: ChatZilla 0.9.90.1 [Firefox 28.0/20140317233623]]
leviathanch2 has quit [Remote host closed the connection]
leviathanch2 has joined #linux-sunxi
hipboi has joined #linux-sunxi
<sappel>
good morning! I need urgent help, I'm totally stuck right now. I just need to recompile a kernel with enabled hidraw for my cubieboards. I have a running sd card image with lubuntu...
<sappel>
so if I'm trying to load a new kernel (I already cross compiled it), do I need all the u-boot + boot.cmd/scr + script.bin stuff?
<sappel>
or should I be fine copying the new uImage to /boot partition + copy modules to /root?
deffrag_ has quit [Ping timeout: 265 seconds]
deffrag has quit [Ping timeout: 255 seconds]
hipboi has quit [Ping timeout: 240 seconds]
<sappel>
hm, obviously not. although with the new kernel the system takes ages to boot...including some ply-terminal problem. might be the reason for that..
bonbons has joined #linux-sunxi
leviathanch2 has quit [Ping timeout: 240 seconds]
bbrezillon has joined #linux-sunxi
linkmauve1 has left #linux-sunxi [#linux-sunxi]
hipboi has joined #linux-sunxi
<mripard>
leviathanch: don't put the addresses, interrupts, clock indices in the documentation
<montjoie[home]>
I have that in uEnv.txt uenvcmd=fatload mmc 0 0x46000000 uImage; fatload mmc 0 0x49000000 sun7i-a20-cubieboard2.dtb; bootm 0x46000000 - 0x49000000
smotocel69 has quit [Quit: Leaving]
<montjoie[home]>
and just before fdt_high=ffffffff
<mripard>
that doesn't seem to be the case
<mripard>
the loaded kernel is at 4800000000
cabania has quit [Ping timeout: 265 seconds]
<wingrime2>
mripard: clk_set_rate we have implemented?
<mripard>
yep
<wingrime2>
mripard: it need Hz value?
<plaes>
wingrime2: what board do you use to test irda?
<wingrime2>
cb2
<mripard>
wingrime2: yes
<wingrime2>
plaes: driver now only for RC , not irda, for irda I have no HW
<smotocel69>
oliv3r linux-3.4.79 have full cubietruck support?
bertrik has quit [Read error: Connection reset by peer]
labrador has joined #linux-sunxi
<srxa>
is there any chance to enable secondary display with EGL with sunxi-mali drivers? ((EGLNativeDisplayType)1)
<wingrime2>
mripard: lirc works but, I only get it in "raw mode"
<labrador>
hello guys.. is gadgetfs is working for you ?
<mripard>
wingrime2: I'm not sure about what the raw mode is, but cool if it works :)
Seppoz has quit [Remote host closed the connection]
Seppoz has joined #linux-sunxi
<wingrime2>
oliv3r: lirc works
<ccaione>
wingrime2: pong?
<wingrime2>
ccaione: nothing )
<ccaione>
:) ok
Seppoz has quit [Remote host closed the connection]
kuldeepdhaka has quit [Ping timeout: 265 seconds]
kuldeepdhaka has joined #linux-sunxi
<ssvb>
srxa: do you mean GLES with two monitors?
<srxa>
eglfs with two monitors
Seppoz has joined #linux-sunxi
<ssvb>
srxa: as far as I know, mali binary drivers do not support eglfs flavour of egl
kuldeepdhaka has quit [Ping timeout: 265 seconds]
<ssvb>
srxa: Qt5 needs to be patched even to support a single monitor with mali framebuffer drivers
<srxa>
i have qt5.3 working with eglfs on primary monitor, i can run another application using linuxfb:fb=/dev/fb1, but i would like to have eglfs on both monitors
Seppoz has quit [Remote host closed the connection]
<ssvb>
hmm, does qt5.3 work out of the box with the mali framebuffer drivers now?
<srxa>
no
<srxa>
still i need to modifi native window
<ssvb>
I see
<srxa>
but there is qeglfshooks for imx6..and example how to enable two eglfs displays
<srxa>
with fbGetDisplayByIndex
<srxa>
which is vendor specific call
<ssvb>
in any case, "strings libMali.so | grep /dev/fb" (framebuffer mali driver) returns both "/dev/fb0" and "/dev/fb%i", so it *might* be possibly convinced to use /dev/fb1 somehow
<ssvb>
srxa: imx6 uses vivante gpu, right?
<srxa>
they introduce variable QT_QPA_EGLFS_FB for that, but i dont know how to use it for mali
<srxa>
yes
Seppoz has joined #linux-sunxi
<ssvb>
afaik none of this eglfs stuff is mandated by any standard, these are the implementation details of some egl/gles drivers
<ssvb>
you can't expect mali binary drivers to behave the same as the drivers for some other gpu and also support all their nonstandard extensions
Seppoz has quit [Remote host closed the connection]
Seppoz has joined #linux-sunxi
kivutar has joined #linux-sunxi
<srxa>
i know that, but if there is such extension, i would like to find it :)
<ssvb>
you have the blob
<srxa>
yes
<ssvb>
and you can poke it until it works the way you like :)
<srxa>
lol
<ssvb>
btw, it looks like there is a reference to 'getenv' in the blob, so it might honour some environment variables
<montjoie[home]>
hello does uboot can read ext4 fs ? I got some Unrecognized filesystem type with it
<wingrime2>
mripard: I have to change sun7i.dtsi sun4i.dtsi , and all IR reciver enabled devices
<wingrime2>
mripard: I should choose some base for it
<specing>
montjoie[home]: uboot is just for booting, afaik it only reads fat32
<wingrime2>
mripard: I mean, we have bunch non accepted stuff, so, mainline patch should be sended over sunxi-next?
<wingrime2>
mripard: and how it will be resolved
<wingrime2>
mripard: more one question, witch parts of patch should be separated
<wingrime2>
mripard: dts stuff can be in single patch with driver, or not, same with Makefile/Kconfig
<wingrime2>
?
<montjoie[home]>
strange my first cubieboard2 have only ext4 and it works with old u-boot
cabania has joined #linux-sunxi
bbrezillon has quit [Quit: WeeChat 0.3.7]
bbrezillon has joined #linux-sunxi
<mripard>
wingrime2: don't worry about how it will be resolved, it's part of my job :)
sappel has joined #linux-sunxi
<mripard>
but usually, we do one patch per thing you're doing
<mripard>
so one for the driver, one to add the driver in the defconfig, one for the DTSI, one for the pin muxing, and another one for the boards.
hawi has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 265 seconds]
bgal has joined #linux-sunxi
hawi has quit [Client Quit]
bgal has quit [Read error: Connection reset by peer]
bgal has joined #linux-sunxi
<wingrime2>
mripard: aganist sunxi-next
<wingrime2>
?
<mripard>
yes
<wingrime2>
mripard: MODULES [=n] ... how I can enable this ...
<wingrime2>
mripard: I can't find in menuconfig
bgal has quit [Quit: Leaving]
bgal has joined #linux-sunxi
kuldeepdhaka has joined #linux-sunxi
bgal has quit [Client Quit]
techn_ has joined #linux-sunxi
<mripard>
it's at the top level
<mripard>
"Enable Loadable Modules Support"
<wingrime2>
mripard: sorry , search by "/" not find anything
<wingrime2>
mripard: looks like I can't do platform device as module
bgal has joined #linux-sunxi
techn__ has joined #linux-sunxi
orly_owl has quit [Read error: Connection reset by peer]
orly_owl has joined #linux-sunxi
<wingrime2>
hm, can
techn_ has quit [Ping timeout: 264 seconds]
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
<montjoie[home]>
I finaly got booting a sunxi-devel kernel, but now it stop after bootconsole [earlycon0] disabled , http://pastebin.com/W9BhU1jj for boot log
nabblet has joined #linux-sunxi
theskilledworker has joined #linux-sunxi
theskilledworker has joined #linux-sunxi
<montjoie[home]>
it seems that the kernel runs a bit further after but the uart show no nothing, any idea ?
sap has joined #linux-sunxi
<ccaione>
montjoie[home]: which console are you using at login time? (inittab?)
<bgal>
montjoie[home]: it seems that the kernel console is not enabled, you should see something like: console [ttyS0] enabled
<montjoie[home]>
where can I enable it ?
sappel has quit [Ping timeout: 265 seconds]
<montjoie[home]>
ccaione I always use ssh for login before
<montjoie[home]>
it's y beginning with uart
<nabblet>
hi, i was looking at your wiki (want to install gentoo on olimex a20) read about crosscompliling. now (it's my first time with compiling for embedded) is it really necessary to crosscompile? i think the a20 has enough performance?
<montjoie[home]>
nabblet yes gentoo can be used on it
hipboi has quit [Remote host closed the connection]
<Seppoz>
can i dsable automatic ra check in uboot
<nabblet>
montjoie[home]: hm, but do i need to crosscompile. can't i do the whole thing natively as i do it with normal gentoo installs?
<montjoie[home]>
compiling in RAM helps
Night-Shade has quit [Ping timeout: 255 seconds]
<bgal>
montjoie[home]: in theory it's already enabled in commandline
<nabblet>
btw props to this channel for being so active and having the wiki :)
<montjoie[home]>
nabblet the problem is the first kernel creation, I have set a gentoo stage with a arch kernel the first time
<nabblet>
montjoie[home]: ah, and then you used the arch kernel for compiling?
<montjoie[home]>
yes for compiling the real "gentoo" kernel
<nabblet>
montjoie[home]: tbh i am willing to wait some time (although i think the a20 will be sufficiently fast) to get around the crosscompiling pain
<nabblet>
what i read there are many thing i can do wrong :P
<montjoie[home]>
with crossdev it is easy to crosscompile
<nabblet>
montjoie[home]: well the wiki says the catch is to emerge/install/setup crossdev
<nabblet>
montjoie[home]: your trick with the archkernel is nice. didn't think of this option at all!
<srxa>
ssvb: ping
<montjoie[home]>
nabblet I forgot the u-boot, I have get an already built one also
<nabblet>
montjoie[home]: do you (personally) have experience with gentoo @ olimex a20 ?
<montjoie[home]>
nabblet no only cubieboard2
<nabblet>
oh, ok
<nabblet>
would anyone benefit from me writing down my steps (given i succeed)
<nabblet>
?
<nabblet>
i see there is already a section about the olimex a20 in the wiki
<nabblet>
montjoie[home]: ok, so what i need/want is compiled kernel + compiled u-boot + stage 3 ?
<montjoie[home]>
for sunxi-devel does someone know if CONFIG_SUN4I_EMAC=y is sufficient to have the network ?
<nabblet>
and the clue is to get the compiled kernel and the compiled uboot i need crossdev? and stage 3 i just copy over as in normal gentoo installs?
<montjoie[home]>
nabblet yes, but if you use gentoo, crossdev is not a pain
<montjoie[home]>
yes just copy stage3
<montjoie[home]>
and the uImage in /boot
<montjoie[home]>
and all u-boot tricks
<nabblet>
montjoie[home]: i set up a vm with gentoo. i read that (with crossdev) you need to take care about your home arch and your target arch. will that be tricky in a vm?
<montjoie[home]>
It just be slower inside a vm, nothing more as I know
<nabblet>
montjoie[home]: ok. thank you. just wanted to make double-sure i am not starting off in the wrong direction :) I'll go with crossdev on my gentoo vm and see how far this will get me.
<nabblet>
montjoie[home]: oh, and one more question. the olimex a20 can only boot from micro sd-card. now, i have a 4 GB sd-card which will probably suffice for an X-less setup. How much should i take into account for a setup with X (and fluxbox)? do you have some kind of clue?
<montjoie[home]>
no idea
<nabblet>
montjoie[home]: ok, thank you for your kind answers
tgaz has quit [Ping timeout: 246 seconds]
<nabblet>
montjoie[home]: out of interest: why did you chose to use the arch-kernel instead of crossdev?
<montjoie[home]>
it is the first I found, too enthusiast to wait for the compilation to end
<nabblet>
montjoie[home]: ah so you were using crossdev parallely?
<montjoie[home]>
now yes
<montjoie[home]>
I compile all gentoo in tmpfs, but kernel si too large, and nfs is too slow
<nabblet>
i don't understand "to wait for the compilation to end" <- compiling on what machine? your cubieboard?
<montjoie[home]>
no my main machine
<nabblet>
montjoie[home]: what are its specs and how long did you wait? just want to get an idea of how long it might take me
tgaz has joined #linux-sunxi
<montjoie[home]>
a quad core with only one core enabled and undercloked due to hardware problems, not a good example:)
<nabblet>
montjoie[home]: fair enough. still how long does it take (roughly)?
<montjoie[home]>
compiling a kernel is less than a hour, 15min I think
<nabblet>
montjoie[home]: ok. and you were too impatient so you got the arch kernel instead? I think it would take me more time to find a suitable precompiled kernel :P
<nabblet>
+ precom. u-boot
<montjoie[home]>
:)
<nabblet>
the google-foo is strong with this one...
<montjoie[home]>
and time to compile crossdev the first time
<nabblet>
that's what i am afraid of -> a consequence of that is this chat :)
<nabblet>
muha! Nice. Just assembled the case i bought with the olimex. They made an extra spacing for the sata-hdd! smart!
<megal0maniac>
Hi all. I've just gotten my hands on a A20 based box. Apparently uBoot is configured to boot from SD card, but what I've tried so far hasn't worked (single partition ext2 or ext4)
<megal0maniac>
Serial doesn't seem to work, so it looks like I'm blind here. Are there any uBoot defaults that you might suggest I try?
<montjoie[home]>
nabblet I have just built one, 17m for a kernel
Fusing has joined #linux-sunxi
<Skaag>
how do you convert a boot.cmd to boot.scr? and can I just replace them with uEnv.txt?
<Seppoz>
when i try to compile oder uboot i get strip:gen_eth_addr: File format not recognized, anyone had that before?
<libv>
but start with the clean device_page_example, and you can probably copy a thing or two over
<libv>
heh, there isn't any real info on that page.
<megal0maniac>
libv: Yeah, it's a20 core and a different base board so I'd rather start fresh
<megal0maniac>
There's pretty much nothing on that page :P
<libv>
megal0maniac: yeah, definitely start fresh
<brain_>
I still gotta finish up my wiki article once I get some free time.
<libv>
megal0maniac: as for getting the board to boot, you do have serial, right?
<megal0maniac>
libv: I "have" serial, but the line just stays low
<libv>
ah, you said
<megal0maniac>
Checked with a LA as well
<libv>
edit the fex, but first, please spend some time documenting this device, and you can then add all your findings to the resulting page, as you go along
<megal0maniac>
fex?
<megal0maniac>
(I am not very familiar with sunxi platforms)
<brain_>
I have been working on AMLOGIC 8726MXS 1.5 dual arch the past month. got everything working a native port of ubuntu running on flash, nice HD screen on one and another 8" same specs.
<libv>
megal0maniac: then all the more reason why you should work through that howto :)
<megal0maniac>
Have been using Seagate Dockstar up until now
<libv>
megal0maniac: after the initial documenting of the device, you will be referred to the manual build howto
<megal0maniac>
If I had serial, or if I could just boot into arch...
<libv>
megal0maniac: there already is u-boot and sunxi-boards support
<libv>
megal0maniac: and if this iteaduino-plus-a20 page had existed already, you would've immediately found it :)
<libv>
hence me being so pedantic here
<libv>
we _need_ people to document their hardware, and help themselves and others in the best way possible
<megal0maniac>
libv: I'm with you on that. But if all I have access to is un-rooted Android, and no uboot, or serial, or ability to boot from something else, I'm pretty limited in what I can achieve
<libv>
megal0maniac: that's all fine atm
<libv>
megal0maniac: getting info off of the android device is also documented out of that new_device_howto
<libv>
but i do not think you need to do that
<libv>
as we already have the necessary data captured
<libv>
just nothing documented :(
<libv>
so just work through the New_Device_howto
<libv>
it'll take a bit of time and a few photographs, but you will see things coming together quickly
<libv>
and it will all make sense afterwards
<libv>
if not, that's what we're here for
<megal0maniac>
I've started with photos as well
<megal0maniac>
I have two weeks with this thing
<brain_>
libv: I have a question for you concerning the lima driver, i do not know if there is code out yet or not but if so let me know, I would like to help work on it and also was wondering if it is architecture dependent? my AML8726MXS has dual MALI400@333MHZ i believe. I would like to try to work on some acceration. I wrote a GSLX680 Firmware dumper for kernels that don't have a separate .ko, pulled some code from the sunxi driver over
<brain_>
to AML tree and wrote the LCD drivers from scratch. I think i am ready to tackle 3d now lol.
<libv>
it should take just a few hours :)
<libv>
brain_: code hasn't been dumped for months now
<libv>
brain_: check out my fosdem talk, there was nothign new since then
<oliv3r>
wingrime2: aweseme! amazing :) it'll be good to have this driver upstream
<megal0maniac>
libv: And once I've done all of this? (Which I'm happy to do)
<brain_>
libv: :( os there ANY code sitting on github anywere? I am halfway throgh the fosdem talk page.
<libv>
megal0maniac: then you should have also worked through a full manual build, and then you know how things fit together and have a working device booting happily off of SD
<libv>
brain_: it's still the q3a code from last may or so
<libv>
megal0maniac: you can skip step 1 and 3 though
<megal0maniac>
Hmmm.. Okay
<libv>
megal0maniac: it's mostly about building up the wiki page now and getting useful pictures in
<brain_>
ok, what about FBturbo for the GUI acceleration? how portable is that?
<megal0maniac>
I'll start uploading pictures in the meanwhile
<libv>
5,6 is also useless for you
<libv>
just wiki the hell out of it :)
<Skaag>
Turl: just saw your URL from last night. thanks!
<libv>
and then you can immediately whatever things you encounter to the wiki page as well
<libv>
+add
<megal0maniac>
libv: I'm still not seeing how this will help me make any progress, since I'm pretty much at a dead-end, but I shall continue regardless :)
<oliv3r>
the nightly kernel keeps crashing on me :S
nove has joined #linux-sunxi
<ccaione>
mripard: how can I add default attributes for platform devices?
<ccaione>
just filling the .groups doesn't work. maybe using device_create_file()?
nabblet has quit [Quit: leaving]
nabblet has joined #linux-sunxi
<libv>
megal0maniac: how does one get serial on this board?
<mripard>
ccaione: iirc, yes
<libv>
megal0maniac: we will get you further, don't worry
<mripard>
ccaione: look at how it was done for the SID
<mripard>
oliv3r spent an awful lot of time figuring it out
<megal0maniac>
libv: The unit was sent to me as an "engineering sample" for review, but I'd like to go a little bit further than that. So I'll contact Itead for more info, I'm sure they'll be accomodating - they've already given me the device for free :P
<megal0maniac>
libv: It should be on pins 9/10 on the header on the side of the board
nabblet has quit [Client Quit]
<megal0maniac>
No detail about logic level, but I assume 3v3 ttl
<libv>
ah, i see
<libv>
yeah, since it is a20 exported directly
<libv>
pics look good :)
<ccaione>
mripard: ok, even though I preferred shutdown and startup attr as attribute of the input device
<libv>
megal0maniac: make the page iteaduino_plus_a20, and i'll add a redirect from ibox
<megal0maniac>
Thanks :) Backers of the indiegogo campaign get a breakout board which gives SATA and USB-serial interface. I didn't get that :/
<libv>
megal0maniac: someone else can add the info on those then :)
<libv>
megal0maniac: so this is just a 2.54mm header?
<megal0maniac>
It's just a breakout, plus a PL2303 or similar for USB-serial. And yeah, 2.54mm but the pins are skinny so nothing connects nicely :/
nabblet has joined #linux-sunxi
techn__ has quit [Ping timeout: 240 seconds]
<oliv3r>
crap even hans's kernel crashes now that always ran stabily
<oliv3r>
u-boot broken
<libv>
oliv3r: ?
<libv>
oliv3r: did you just break u-boot?
<megal0maniac>
Okay, added pin header to album
<megal0maniac>
Heh :) Way too high-res for what it is
<libv>
well, you better too high than too low.
<libv>
s/you //
<Seppoz>
i want to checkout the 2012 verison of sunxi uboot, how cna i do that?
<nabblet>
as i see it allwinner a20 has hardfloat. correct?
<megal0maniac>
nabblet: Yip
<nabblet>
thx
<Seppoz>
is thee a way how much memory there is in linux instead of uboot
<Turl>
leviathanch: " and as "allwinner,sun5i-a13-mmc" on all other A1x and A2x allwinner chipsets." - be more specific. Some day we'll have A23 support and it may or may not be compatible.
<Turl>
leviathanch: you got a typo on "appart"
<megal0maniac>
Eugh. Can't get into uboot
<ccaione>
Seppoz: what do you mean
<Seppoz>
im having big issues getting uboot to work
<Seppoz>
only one i have working is bin
<Seppoz>
and it has 1gig mem hardcoded
<Seppoz>
but mine only have 512
<Turl>
leviathanch: and I think the base, clock, irq stuff is not necessary at all
<Turl>
leviathanch: grab another binding and have a look at the format, try to follow it
<ccaione>
doesn't uboot use a sort of mem-test to checko for available memory?
<Seppoz>
it wont help if i cnat modify uboot
<Seppoz>
is there a way to do it in the kernel?
<megal0maniac>
libv: The uboot delay is set to 0, so it isn't possible to get in. Any ideas?
<megal0maniac>
I'm going out in a few minutes, but will start with wiki stuff tomorrow
<megal0maniac>
Once I'm in, I don't quite get a shell, but I can execute commands. Although I
<megal0maniac>
assume that's all Android
<Seppoz>
its possible to get in
<Seppoz>
just rename the kernel
<ccaione>
Seppoz: rename the kernel?
Seppoz has quit [Remote host closed the connection]
Seppoz has joined #linux-sunxi
<Turl>
Skaag: you're welcome
<Turl>
ok, finished backreading :)
<ssvb>
srxa: pong
<srxa>
ssvb: it seems that enabling secondary eglfs has nothing to do with mali driver, rather with sunxi disp
<ssvb>
srxa: how so?
<srxa>
mali will render on any EGLDisplay that is defined, and phisical screen is defined with sunxi disp
<ssvb>
srxa: sunxi disp provides you with /dev/fb0 for one display and /dev/fb1 for another one (if configured for dual monitor support)
xavia has joined #linux-sunxi
deasy has joined #linux-sunxi
<ssvb>
srxa: how is EGLDisplay related to anything?
xinj has quit [Ping timeout: 265 seconds]
<labrador>
anyone having uart & rtc wakeup working ?
<srxa>
EGLDisplay first must be initialized, and initialization is from display driver
<srxa>
with linuxfb i have half application on both screens
<ssvb>
what is the resolution of each individual screen?
<srxa>
1280x720
akaizen has quit [Remote host closed the connection]
<ssvb>
this just does not compute, for a combined xinerama framebuffer I would expect 2560x720
* ssvb
has never used this xinerama mode though
<srxa>
but i must set fb0_height in fex i think, to double, because now i have only half of each monitor :)
<srxa>
i will manualy set width for fbo now
<ssvb>
what is your final goal? this xinerama mode in Qt5? or two independent displays both running their own Qt5 applications?
<srxa>
whatever, just to use dual display in accelerated mode
<ssvb>
for the xinerama mode, it might be indeed a problem in the kernel sunxi disp driver
<ssvb>
but this has nothing to do with mali or Qt5
<srxa>
manually seting fb0 dimensions does not help
<srxa>
same info in fbtest
<srxa>
fbset, sorry
<ssvb>
if you can get properly behaving framebuffer with 2560x720 resolution on /dev/fb0 for userland applications with no glitches, then mali might be fine with it too
akaizen has joined #linux-sunxi
<srxa>
ok, i will try acomplish that too
<ssvb>
I think that almost nobody has really used this functionality, so there may be still tons of bugs in this xinerama support
<srxa>
then, i will bw
<srxa>
be first to test everything
<srxa>
:)
<ssvb>
and I can't even imagine what it is supposed to do, for example, for LCD (low resolution) + HDMI (high resolution) :)
<srxa>
no, vga+hdmi on same resolutions
<srxa>
i am developing games for gambling machines, and i am trying to use this platform for that, if i acomplish my goal, it will be idally for that
kuldeepdhaka_ has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 265 seconds]
kuldeepdhaka_ has quit [Max SendQ exceeded]
kuldeepdhaka_ has joined #linux-sunxi
<labrador>
i am trying to build 3.10 but always whatever changes i do to my config it fails.. anyone have succesffuly build 3.10 kernel ?
<montjoie[home]>
for which board ?
<montjoie[home]>
you need to give more info
<labrador>
montjoie[home]: A13 olinuxino
<labrador>
the reason that trying to use 3.10 is cause i cant have wakeup from uart or rtc in the 3.4 kernel
xavia has left #linux-sunxi ["PART #interlock :PART #arduino :PART #wnydev :PART #buffalolab :PART #docker :PART #symfony :PART #virtuvia :QUIT :Leaving."]
<nove>
labrador: from what i am aware, 3.10 is *still* not usable
<labrador>
nove: ok
<labrador>
any idea howto be able to wakeup my board after sleep from uart or rtc0 ?
kuldeepdhaka__ has joined #linux-sunxi
<ccaione>
labrador: on A20 there is an alarm IP. did you check if there is anything similar on A13?
<labrador>
ccaione: alarm IP ?
kuldeepdhaka_ has quit [Ping timeout: 265 seconds]
theskilledworker has quit [Quit: Leaving]
<ccaione>
labrador: it is a kind of alarm able to wake-up the processor
<labrador>
ccaione: how to search if A13 has one ?
<ccaione>
datasheet? :)
<labrador>
ok
<labrador>
thank you
<labrador>
also may i ask.. my olinuxino with a li-po 1400mah battery can hold maximum 1hour with powersave governor
<labrador>
its kinda thursty right ?
nabblet has quit [Quit: leaving]
<specing>
sounds about right
kuldeepdhaka__ has quit [Max SendQ exceeded]
bgal has joined #linux-sunxi
kuldeepdhaka__ has joined #linux-sunxi
<labrador>
is it worth disable audio and video to reserve some power ?
<labrador>
or the difference will be very small that dont woth it ?
techn_ has joined #linux-sunxi
<wingrime2>
labrador: just fix PM and device code
<wingrime2>
labrador: add to pm wakeup source just by Not disabling IRQ from device
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
<labrador>
wingrime2: i am not following.. can you explain farther please ?
<wingrime2>
labrador: .., you can make any device to be wakeup source, still it have IRQ
<wingrime2>
labrador: sleep for sunxi is .. just asm("wfi") - wait for IRQ
<mnemoc>
moin
hramrach has quit [Remote host closed the connection]
<srxa>
i will try, but first i must compile qt with xcb support :)
<ssvb>
srxa: there are lots of issues everywhere, but if you have independent /dev/fb0 and /dev/fb1 framebuffer, then mali framebuffer drivers need special handling for this
<ssvb>
srxa: just for fun, you might try to patch libMali.so binary (framebuffer mali drivers) and replace /dev/fb0 string with /dev/fb1
<ssvb>
srxa: and maybe the "/dev/fb%i" string too
<srxa>
that is first what i will try, just to see if it will work at least on secondary display only
deffrag has joined #linux-sunxi
<labrador>
wingrime2: any document that explains the progress of assign that device as wakeup source?
akaizen has quit [Remote host closed the connection]
<wingrime2>
labrador: pure logic
deasy has joined #linux-sunxi
<wingrime2>
labrador: 1) commentout resume/suspend in source driver (it must have External IRQ)
<Turl>
ccaione: also moin doesn't mean morning :p just hello
<labrador>
can you help me to understand for ex: what i have to do enable the uart interrupt ? (i dont know what file is coresponding for that i am searching for)
<Turl>
labrador: if your uart works then it is already enabled
<ccaione>
thought moin==morning
<labrador>
Turl: my uart works fine but it not wakeup when data is received and systam is in sleep state
<Turl>
labrador: what kernel are you using?
<Turl>
and on what platform?
<labrador>
Turl: sunxi-3.4 olinuxino A13
<Turl>
ccaione: see [[Moin]]
<ccaione>
hooooo
<Turl>
labrador: you may want to have a look at arch/arm/plat-sunxi/pm/ then
<Turl>
I think normally USB, LRADC and a couple other things are set as wake sources, but not uart
xavia has joined #linux-sunxi
<labrador>
Turl: also if i put sytem on sleep and plug a usb device it dont wakeup
<labrador>
system wakesup only when one of the onboard buttons are pressed
<wingrime2>
Turl: earlyer I have no normal way test it as module (no mmc,etc)
<labrador>
Turl: i see that in pm.c there is the wakeup sources.. but how can i add uart for example ?
<Turl>
wingrime2: PC is at __kmalloc_track_callerunxi-ir o.O
<wingrime2>
labrador: for uart you only need add cmdline noconsolesuspend
<Turl>
labrador: see standby/standby.c
<Turl>
wingrime2: I think aw standby code disable interrupts?
<Turl>
but it's easy to try, maybe that works
<wingrime2>
Turl: usualy driver suspend routine do it
<Turl>
wingrime2: yes, but aw quality code :)
<wingrime2>
Turl: but , for some cases, for example LRADC , PM saves regs and . reconfigure it again
<wingrime2>
Turl: and enable irq
<wingrime2>
Turl: yes, allthings belong to suspend blob and drivers resume/suspend , and anything that can generate IRQ can be wakeup source
<Turl>
wingrime2: suspend blob mask all interrupt in standby/standby_int.c it seems
<wingrime2>
Turl: just for make sure ))
<Turl>
:)
<nove>
help me decide here http://linux-sunxi.org/Register-mediawiki-extension, is this worth the trouble that will create? Should be done? I can't decide, this web things are so complicated and with many limitations not allowing to do exactly what one wants.
<labrador>
wingrime2: noconsolesuspend on bootargs not working
<Turl>
nove: table is enough imo
<Turl>
nove: like on user manual
<wingrime2>
labrador: no_console_suspend=1
<wingrime2>
labrador: to cmdline
<labrador>
ok.. let me try it
<mnemoc>
ssvb: sorry... I didn't see it. will apply it now. what about merging stage to sunxi-3.4 and making a version jump? any other outstanding patch for 3.4?
<labrador>
wingrime2: nope.. not working too
<wingrime2>
labrador: than you need unmask uart IRQ in PM
<wingrime2>
too
<Turl>
you will need to add code to enable the uart interrupt on standby/standby.c
<mnemoc>
nove: will you write the extension to hook packetdiag in mediawiki?
<Turl>
that together with no_console_suspend=1 should be enough
<labrador>
wingrime2: how can it be done ?
<labrador>
there is 2 files.. pm.c and standby.c that i have to edit and add sections for uart ?
<srxa>
ssvb: when i grep strings from libMali.so, there is no /dev/fbX...just /dev/mali and /dev/gator/anotate
<Turl>
labrador: yes
<nove>
mnemoc: i could write, but until know only found problems, not solutions
<ssvb>
srxa: where did you take your libMali.so?
<srxa>
r3p2-01rel1
<mnemoc>
nove: packetdiag is intended for long registers, not only one word
<Turl>
mnemoc: isn't it intended for network packets? :P
<mnemoc>
Turl: indeed :p
<mnemoc>
Turl: >4 bytes :p
<nove>
mnemoc: this will take a bit of work, and i am undecided because there are more things in that time would be better spend
<labrador>
Turl: i have to create also standby_uart.h file ?
<mnemoc>
but there is no much difference between the binary header of a network packet and the register of an IP
<srxa>
ssvb: i am using your kernel, 20130913 branch
Night-Shade has joined #linux-sunxi
<mnemoc>
nove: registers described only in tables are hard to digest. a visual representation makes it much easier to understand
<nove>
mnemoc: but as Turl said, only the table is enough, and for that wiki templates should be enough
<ssvb>
srxa: heh, I see
<ssvb>
we really need to add mali r3p2 support to sunxi-3.4 kernel eventually
<mnemoc>
nove: enough != good
<nove>
mnemoc: i agree
<srxa>
yea :)
* nove
likes pretty pictures
* mnemoc
too
<mnemoc>
and even better if they are generated out of simple plain text
<Turl>
labrador: something like this, together with the cmdline option should be more than enough for what you want to do http://sprunge.us/RgjP
<labrador>
Turl: i will give it a try..
* Turl
likes useful pictures
<Turl>
for cat pics I already have a facebook account
<ssvb>
srxa: for r3p2 mali blobs, these references to "/dev/fb0" and "/dev/fb%i" are in libEGL.so
<srxa>
ssvb: ok
<ccaione>
ssvb: do we have r3p2 (api 19) blobs?
<ssvb>
srxa: unlike r3p0, it is not a single blob for everything with just a set of symlinks to it :)
<srxa>
:)
<ssvb>
ccaione: r3p2-01rel1 is no good?
<libv>
our r3p0 binaries also reference /dev/fb0 and /dev/fb%i
<ccaione>
ssvb: I need APIv19 for another project.
<nove>
the solution is to put usefulness and prettiness together, but i am not seeing how
<Turl>
ccaione: *cough*sellout*cough* :)
<Turl>
j/k :p
<libv>
also, i think that MALI_FBDEV is an environment variable that can be used
<nove>
to not waste more time, for now lets stay with table templates and continue on
<mnemoc>
iirc someone from canonical took that job, not greg hk
<ssvb>
does this affect us, android or anybody else?
<mnemoc>
ssvb: if 3.14 becomes next android instead of 3.10 there is much more sunxi stuff in by default
akaizen has joined #linux-sunxi
<mnemoc>
ssvb: and so more problems for aw to hack their crap in
<mnemoc>
and so, better chances of getting them collabirating
itdaniher has quit [Ping timeout: 252 seconds]
<mnemoc>
with libv's sunxi-kms in progress I find hard to justify sunxi-3.10 at all
itdaniher has joined #linux-sunxi
akaizen has quit [Ping timeout: 240 seconds]
<libv>
i suck at progress tbh, i spent the week procrastinating on the wiki
<libv>
all i wanted to do was update the mali binary driver page for the split ump
<nove>
by the way, is common occurrence here that new people wanting to compile a kernel version without knowing what is supported
<libv>
nove: you mean "compile the mainline kernel"
<nove>
would it be good to have a feature matrix, with what is support by what
nedko has joined #linux-sunxi
nedko has quit [Changing host]
nedko has joined #linux-sunxi
<mnemoc>
that's the main point for the stable sunxi- branches... and imo it should match an android supported kernel. sunxi-3.14 could be it
<libv>
nove: what is not supported?
<libv>
a31, a31s, a23
<libv>
other things are mostly bugs, apart from touchscreen code which is often MIA
<libv>
nove: or are you talking about people who wrongly wander into the mainline kernel work without a clue, thinking that that is the one true way, and that everything will magically be better?
<nove>
libv: there is some misunderstanding, like per example getting mainline kernel, and then asking why display don't work
<libv>
nove: you should've been here a few months ago
<nove>
libv: yes, that
<libv>
type in kernel on the wiki and you'd only get to the mainline kernel page
nedko_ has joined #linux-sunxi
<libv>
which only booted on a handful of dev boards (not sure whether that has massively changed yet)
nedko has quit [Read error: Connection reset by peer]
<Turl>
libv: people with things that are not dev boards don't have too much a use for mainline, because there's no display support
<libv>
nove: the issue is that most people believe that this is like an x86 pc
<nove>
and maybe a page in the with a matrix of socs vs kernel versions, and what its state (what works and what doesn't)
<libv>
they are new to the arm world
<libv>
nove: there are plenty of warnings on the wiki
<libv>
and that's the other big issue, people do not search or read wikis
<libv>
nove: the only real solution is device pages.
<libv>
device pages which clearly guide people to the manual build howto, which tells people how to build the sunxi-3.4 kernel
<libv>
device pages is what people do tend to read, especially if they are kept structured and to the point
<libv>
anything else will always result in TLDR
jemk has joined #linux-sunxi
<nove>
libv: and if searching in search engines for the device, getting our wiki in the first result
nedko has joined #linux-sunxi
nedko_ has quit [Read error: Connection reset by peer]
nedko has quit [Changing host]
nedko has joined #linux-sunxi
<libv>
nove: good point
<libv>
i just googled for linux leliktec and got nowhere
<libv>
oh, with the accent correctly, i do get a hit
<Turl>
libv: our wiki has 0 results for leliktec
<libv>
quite quickly as well
lkcl_ has quit [Ping timeout: 265 seconds]
<libv>
Turl: inet k70hc, i brought that one up
bgal has quit [Read error: Connection reset by peer]
<libv>
"LélikTec linux" has the page as the third hit, so good
nedko has quit [Ping timeout: 245 seconds]
<libv>
good
<libv>
some other tests are quite good indeed
<libv>
so device pages
<libv>
device pages
<libv>
device pages
<libv>
device pages
<libv>
device pages
<mnemoc>
:D
<mnemoc>
and distro pages
nedko has joined #linux-sunxi
nedko has quit [Read error: Connection reset by peer]
nedko has joined #linux-sunxi
nedko has quit [Changing host]
nedko has joined #linux-sunxi
xavia has joined #linux-sunxi
<srxa>
ssvb: editing /dev/fb0 to /dev/fb1 and i have eglfs on secondary display
cajg has quit [Ping timeout: 264 seconds]
<ssvb>
srxa: this means that you already can use two displays
<srxa>
yes, now just find way how
<ssvb>
srxa: by just running two programs and using different sets of mali libraries for them (by tweaking LD_LIBRARY_PATH or something like this)
sap has quit [Ping timeout: 265 seconds]
<srxa>
yea, that is the easiest way :)
<ssvb>
but it still makes sense to try using environment variables, libv suggested that MALI_FBDEV might work
<srxa>
hm...i dont think so, i already try that
<srxa>
but there is FRAMEBUFFER variable also
<ssvb>
are you tracking what is getting passed to getenv function?
<srxa>
no
<srxa>
i set variable in qt environement variables for deployment
<ssvb>
the natural way to debug this would be to set a breakpoint on getenv and see what kind of environment variables are used by the mali blob
<ssvb>
you can also ask in the official mali support forums, if the blob is configurable via environment variables, then they surely should know it :)
<srxa>
ssvb: FRAMEBUFFER is solution...
<srxa>
ssvb: it's so simple on the end ...
<ssvb>
:)
akaizen has joined #linux-sunxi
deffrag_ has joined #linux-sunxi
deffrag has quit [Ping timeout: 255 seconds]
<srxa>
ssvb: but now i have 40 fps per screen, instead of 80...
<srxa>
it will be enough
<Turl>
srxa: you still have 1 mali, not 2 :p
akaizen has quit [Remote host closed the connection]
<srxa>
but dual mali :)
akaizen has joined #linux-sunxi
netlynx has quit [Quit: Leaving]
<labrador>
guys is there anyway to put my system to sleep for ex: 1 minute and then wakeup ?
<labrador>
i am asking because there is not rtc0 wakeup in sys
<labrador>
wakealarm or something
Black_Horseman has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
<libv>
srxa: be happy that the mali scales that well
<libv>
you do get exactly half the framerate.
<libv>
srxa: but it happily does so without complaints
<libv>
unlike some competitors
<srxa>
yes, exactly
techn_ has quit [Ping timeout: 255 seconds]
<labrador>
its a13 olinuxino
cajg has joined #linux-sunxi
andoma has quit [Ping timeout: 246 seconds]
andoma has joined #linux-sunxi
swabbles has quit [Ping timeout: 246 seconds]
swabbles has joined #linux-sunxi
captainigloo has quit [Ping timeout: 246 seconds]
rm has quit [Ping timeout: 246 seconds]
captainigloo has joined #linux-sunxi
rm has joined #linux-sunxi
<Turl>
labrador: A13 doesn't have a built in rtc, but olimex (and pretty much anyone) include an i2c one, so yes, it should be possible
<labrador>
but as i see its not currently available right ?
<megal0maniac_afk>
What does one do if the uboot timeout is 0 and you don't have access to setenv in the host system?
<Turl>
replace uboot with one with a different timeout? :)
<megal0maniac_afk>
Eugh
<megal0maniac_afk>
I guess. And if you break uboot? :P
<Turl>
you fix it? :p
<megal0maniac_afk>
Why the hell was my Dockstar easier to hack than this "open development platform"??
<megal0maniac_afk>
Turl: Seriously though, what are the recovery options with A10/20?
<megal0maniac_afk>
Is there JTAG, or..?
<Turl>
megal0maniac_afk: 1) insert sd card 2) ??? 3) system works
<Turl>
sd card is the first storage medium checked to boot from
<Turl>
so the only chance of bricking it is if you completely hose the hardware or so
<megal0maniac_afk>
So booting from SD card is handled by the processor directly, not uboot?
<Turl>
megal0maniac_afk: yes
<Turl>
megal0maniac_afk: you need to install uboot on the sd card
<Turl>
check the BROM page on the wiki
<Turl>
it explains it with more detail
<megal0maniac_afk>
Cool, thanks for clearing that up :)
<megal0maniac_afk>
Another question. This box has 4 regular USB ports. How do you install an image onto the internal flash?
<megal0maniac_afk>
i.e. there is no micro/mini USB connector
<Turl>
with an sd card :P
<megal0maniac_afk>
The SD card truly is the universal solution here... :P
<Turl>
there is a tool that converts those livesuit images to a sd card that you can put in on the box and it boots and flashes the nand
<Turl>
yeah it is
<megal0maniac_afk>
That's cool though, it's versatile and difficult to kill
<Turl>
I'd really suggest you start from there and forget about the nand. Unless you really need the sd slot for something else, that is
<megal0maniac_afk>
Turl: Well, common practice for me is to put something "useful" onto nand in case sd card boot fails. Then I still have something usable
<megal0maniac_afk>
usable != android
<Turl>
ok :)
akaizen has quit [Remote host closed the connection]
<megal0maniac_afk>
So.. practical example. I have lubuntu nand image for A20 which is an .img file. I have an SD card. What now?
<megal0maniac_afk>
Do I dd it or is there a utility or how do I get it onto nand or..?
smotocel69 has quit [Remote host closed the connection]
nove has quit [Quit: nove]
akaizen has joined #linux-sunxi
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
sap has joined #linux-sunxi
<libv>
megal0maniac_afk: all that is on the wiki
<megal0maniac_afk>
As I'm slowly discovering :/
<megal0maniac_afk>
There is a heck of a lot to learn on a new platform. Beginning to realise just how different it is from what I know
<libv>
megal0maniac_afk: there are thousands of different devices out there
jemk has quit [Remote host closed the connection]
<libv>
megal0maniac_afk: it's not just one or two like with the dockstar
<libv>
you only rarely just dd 1 image to an sd card and be done
<Turl>
the exception being fedora :P
<megal0maniac_afk>
Which is okay. It just means I need to spend many hours on this to get comfortable :)
<Turl>
if you have a nand image for your device you usually use phoenixsuit to make a card to flash your device
<libv>
megal0maniac_afk: not as many as you'd think, for just getting something on an sd-card
<libv>
megal0maniac_afk: but you do need to follow the howtos or it'll not work
<libv>
i got quite pissed at a suse guy a year ago, he was playing with a cubieboard, and said "it doesn't boot" every 5 minutes
<megal0maniac_afk>
Turl: Are you sure phoenixsuite does that? It seems like it only talks over adb and provides no other options
<libv>
even though he got told to read the fucking wiki every time
<megal0maniac_afk>
libv: Don't worry, I won't be that guy ;)
<libv>
he just read 1 line and then gave up reading again
sap has quit [Remote host closed the connection]
<Turl>
libv: hahah
<megal0maniac_afk>
When you're messing around with this stuff, "not booting" is the default state. Anything else is a special occasion
sappel has joined #linux-sunxi
<Turl>
megal0maniac_afk: phoenixsuite is like the new version of livesuite as far as I know
<Turl>
and it can do flashing via FEL mode (OTG) or make the cards
<megal0maniac_afk>
I've got it, but it seems pretty stripped down
<Turl>
previously you'd make the cards with phoenixcard I think
cabania has quit [Ping timeout: 255 seconds]
<srxa>
megal0maniac_afk: it is nothing awfully to do...just 20 days ago i've got my device, then, as i was in windows, i try linux for first time(before that i only know $ls and $man and $cd from linux, $nano also :D), after few days i was compiling kernel, then i make Qt5 crosscompiled and work with my device, and today i was messing with hexeditor and binary blobs to make damn thing work as i wish :D
<megal0maniac_afk>
Turl: I'll double check. Maybe I have an outdated version or something