<uniqdom>
Hi. Yesterday I have installed Debian in a Cubieboard2 with help from user ssvb, today I was compiling linux-sunxi kernel. I have generated the uImage file. The thing is that I can't find any file named uImage in the Debian sd card (find / -name uImage).
<uniqdom>
I can see a vmlinuz in /boot folder that's the kernel right?
<Turl>
uniqdom: yeah, it may be using a zimage instead of an uimage
<ssvb>
uniqdom: yes, debian is using zimage format, you can also check the boot.scr file
<uniqdom>
I have used this command line to make the uImage: make -j2 ARCH=arm CROSS_COMPILE=arm-linux-gnueabihf- uImage modules, I know should i change uImage by zImage in that line??
<ssvb>
uniqdom: btw, why are you trying to install the linux-sunxi kernel? are you missing some features in the debian kernel?
<uniqdom>
I need to use audio.
<uniqdom>
I haven't tested hdmi, but I have read that it doesn't work. I will try in a second
<uniqdom>
should i change uImage by zImage in that line command??
<ssvb>
uniqdom: more like you should replace the debian stuff with a new boot.scr, script.bin and the kernel, and also install modules to the right place
<ssvb>
heh, the syntax of the new u-boot is a bit different
<ssvb>
I think it should be "load mmc 0:1 <addr> <filename>"
<ssvb>
uniqdom: you only need to compile uImage and script.bin and then tweak boot.cmd and copy modules, re-formatting the sd card is surely not necessary
<uniqdom>
ssvb, I have the uImage and modules compiled. I can't find script.bin and boot.cmd in the sd card nor in the system where i have compiled the kernel
<uniqdom>
in the sdcard i see /boot/boot.scr
<uniqdom>
nvm... i see a chapter called script.bin in the link you gave me
<ssvb>
uniqdom: script.bin is serving the same role as *.dtb files in sunxi-3.4
<ssvb>
uniqdom: you can create a new boot.cmd from scratch, but taking a look at debian boot.scr may be also useful to see what exactly debian needs
<uniqdom>
okay
<ssvb>
I'm not sure if there is any tool to convert boot.scr back into boot.cmd, but that would be surely useful
<uniqdom>
+1
<uniqdom>
your link says: fatload mmc 0 0x43000000 script.bin || ext2load mmc 0 0x43000000 boot/script.bin. How to know if those address are going to work with a CubieBoard?
ricardocrudo has joined #linux-sunxi
<ssvb>
uniqdom: these addresses should work, but you need to use "load mmc 0:1 <addr> <filename>" commands instead of what is written in the wiki
<ssvb>
uniqdom: (unless you want to downgrade the bootloader)
<uniqdom>
no
<uniqdom>
thanks... i think that i'm finishing
akaizen has quit [Remote host closed the connection]
akaizen has joined #linux-sunxi
<ssvb>
uniqdom: oh, and you also need to set the "bootm_boot_mode" variable in boot.cmd to "sec" with the new u-boot
<uniqdom>
ok :) i will
leviathanch has quit [Read error: Connection reset by peer]
<ssvb>
otherwise the old kernel will not boot :)
<ssvb>
too bad that the mainline u-boot coverage in the linux-sunxi wiki sucks :(
akaizen has quit [Ping timeout: 250 seconds]
popolon has quit [Quit: WeeChat 1.1.1]
leviathanch has joined #linux-sunxi
akaizen has joined #linux-sunxi
<leviathanch>
I'm still waiting for my work permit
<leviathanch>
it will certainly take at least two weeks since it's Chinese new year
<leviathanch>
...
ricardocrudo has quit [Remote host closed the connection]
<uniqdom>
ssvb: the kernel load but with some kind of error I think, because I can't get a console.
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
<uniqdom>
ssvb: nvm: It is workin :) with just what you said. there is no login console in ttyUSB0. uname -a: Linux debian-cubie 3.4.103-g9a1cd03 #1 SMP PREEMPT Fri Feb 13 11:14:49 CLST 2015 armv7l GNU/Linux
<uniqdom>
hdmi is working
<uniqdom>
audio is also working... now I need a realtime kernel :D
<uniqdom>
thanks again
<wens>
leviathanch: i want the one for the optimus :)
<leviathanch>
wens: schematics?
<leviathanch>
wens: since we decided to support sunxi at bringing all the Allwinner drivers mainline so that they have more spare time
<leviathanch>
wens: and since we wanna sell more Optimus boards anyway
<leviathanch>
wens: this will not be a problem I think
<leviathanch>
wens: at the moment the few around are very stressed since during Chinese new year every team is only half the size
<leviathanch>
wens: at top
* tyler-baker
just received his cubieboard 4
* leviathanch
is very happy for tyler-baker
<leviathanch>
tyler-baker: where did you buy it?
<tyler-baker>
is there a kernel support anywhere for it yet?
<tyler-baker>
hk
<leviathanch>
tyler-baker: I'm just working on it
<tyler-baker>
aw sweet, hopefully it's mostly just bindings
<leviathanch>
wens: interesting... which u-boot were you using?
<wens>
tyler-baker: ah right, was referring to the optimus
* wens
just got out of bed
<tyler-baker>
:)
<leviathanch>
wens: the cubie is just based on the optimus
<wens>
leviathanch: still using appended dt with allwinner's u-boot
<leviathanch>
so using this one is a good start
<leviathanch>
hmm
<leviathanch>
okey
<tyler-baker>
I just got off a flight from HK so I am delirious :)
<wens>
haven't tried hansg's stuff
<leviathanch>
I ask Allwinner if they could enable ftd
* leviathanch
sends a skype message
<wens>
leviathanch: nah, i'm just sticking to whatever is working :)
<wens>
if i screw up my boot sequence, i have to spend time on it instead of the kernel
<leviathanch>
hmm
<leviathanch>
okey
<leviathanch>
well, boot0 just has to be closed partially at least
<leviathanch>
because otherwise all the concurrents around can see how they have built their chip
<leviathanch>
we really have to find a solution for that
<tyler-baker>
ttyl thanks for the info, very helpful as always :)
<leviathanch>
something which makes them happy and us happy
<leviathanch>
that's why I offered that the guy who wrote boot0 would help writing the mainline u-boot
<leviathanch>
only issue: he is at Chinese hollidays right now
<leviathanch>
he would write the functions
egbert has quit [Disconnected by services]
<leviathanch>
and we would clean it up and push it mainline
egbert has joined #linux-sunxi
<leviathanch>
so he is relieved of upstreaming work but has provided us the code which is too sensitive to be explained too deeply
<leviathanch>
I think that might be a half ways satisfying way to go
<leviathanch>
or we just don't initialize some of the hardware the boot0 would
<leviathanch>
another option
<leviathanch>
wens: what do you think about that?
<leviathanch>
I was just told that boot0 is too sensitive to be released fully... partially: 好
Renard has quit [Quit: Renard]
<wens>
i thought that boot0 just does dram and initial clock/pll stuff
<wens>
leviathanch: though since hansg is working on this, maybe you should write him a mail?
<vishnup>
leviathanch: Yes, DRAM is working
<leviathanch>
vishnup: perfect
<leviathanch>
vishnup: so you did talk to the person who has done the DRAM controller?
<leviathanch>
vishnup: and it was a success?
<vishnup>
I did not get any response from Flord
<leviathanch>
oh
<leviathanch>
maybe because they have new year right now
<vishnup>
No, I've yet told him that DRAM is working
<vishnup>
may be
<leviathanch>
ok
<leviathanch>
just tell him
<leviathanch>
maybe he has some suggestions to make
<leviathanch>
after all, he has built it
<leviathanch>
lixiang came out long after I had upstreamed the mmc driver
<leviathanch>
and fixed some issues
<leviathanch>
the more code we have mainline the more response we can expect
<ssvb>
leviathanch: honestly, the dram code in boot0 really stinks, I would guess that a more reasonable explanation could be that they are too embarrassed to show it ;-)
<leviathanch>
ssvb: that might as well be true
<leviathanch>
ssvb: they do everything under a lot of time pressure
<leviathanch>
so ugly code is very likely
<ssvb>
leviathanch: don't tell them that boot0 is very small and easy to disassemble (if the need arises)
<wens>
hehe
<leviathanch>
hehe
<leviathanch>
as soon as I get my desk there I will see the code anyway ^^
<leviathanch>
so...
<leviathanch>
but I guess you are right, it's most likely just shame
<leviathanch>
anyway
<leviathanch>
4.a.m.... I should sleep
<ssvb>
I'm afraid of the trustzonish things which might make life harder in the future...
<leviathanch>
hmm
<leviathanch>
ssvb: trust zone?
<leviathanch>
in which way?
<ssvb>
arm security features, signed bootloaders and other crap
<leviathanch>
ah
<leviathanch>
yes, we discussed that as well
<leviathanch>
you don't need that on tablets
<leviathanch>
it's to expensive anyway
<leviathanch>
*too
<leviathanch>
and on the other project line I suggested I explained as well why we don't need it
<leviathanch>
so... I don't think they will introduce this crap
<wens>
the a80 has the full set of those though, documented in the manual
* wens
doesn't understand any of it
<leviathanch>
wens: they mean crypto engines
<leviathanch>
I think
<leviathanch>
anyway
<leviathanch>
gotta sleep already
cnxsoft has joined #linux-sunxi
ykchavan has joined #linux-sunxi
gzamboni has quit [Ping timeout: 245 seconds]
zombu2 has quit [Read error: Connection reset by peer]
zombu2 has joined #linux-sunxi
uniqdom has quit [Quit: Ex-Chat]
p1u3sch1_ has joined #linux-sunxi
p1u3sch1 has quit [Ping timeout: 245 seconds]
<ssvb>
Seppoz: hi, are you here?
<ssvb>
diego71: ping
ykchavan has quit [Quit: Leaving]
TheSeven has quit [Ping timeout: 265 seconds]
TheSeven has joined #linux-sunxi
zombu2 has quit [Read error: Connection reset by peer]
zombu2 has joined #linux-sunxi
Andy-D has joined #linux-sunxi
zombu2 has quit [Read error: Connection reset by peer]
cnxsoft1 has joined #linux-sunxi
cnxsoft has quit [Ping timeout: 250 seconds]
zombu2 has joined #linux-sunxi
premoboss has quit [Ping timeout: 264 seconds]
adj_ has quit [Quit: Leaving]
adj_ has joined #linux-sunxi
premoboss has joined #linux-sunxi
Andy-D has quit [Remote host closed the connection]
naobsd has quit [Remote host closed the connection]
naobsd has joined #linux-sunxi
ganbold_ has quit [Ping timeout: 255 seconds]
ninolein has quit [Ping timeout: 246 seconds]
ninolein has joined #linux-sunxi
iamfrankenstein1 has quit [Ping timeout: 245 seconds]
konradoo77 has joined #linux-sunxi
gianMOD has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
gianMOD has quit [Ping timeout: 252 seconds]
arete74 has quit [*.net *.split]
cajg_ has quit [*.net *.split]
gianMOD has joined #linux-sunxi
<a1d3s>
i'll give up .. 3.19-rc5 was latest kernel who worked on bananapro for me .. testet with next-20150213 there network doesnt do anything and the flickering still persist .. always used same config as in 3.19-rc5 and before
gianMOD has quit [Remote host closed the connection]
arete74 has joined #linux-sunxi
cajg_ has joined #linux-sunxi
lerc has quit [Ping timeout: 244 seconds]
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
<el3>
leviathanch: Okey sounds cool. I also might try follow you guide when you finish it.
<el3>
mainline u-boot
<leviathanch>
el3: ok.
<leviathanch>
since it's now part of my university work it will be faster (not just hobby anymore)
<el3>
leviathanch: will it work for pcduino?
<leviathanch>
we at Allwinner have mainlining u-boot for all our chips a top priority
<libv>
leviathanch: oh?
<libv>
leviathanch: how does that show, and where?
<leviathanch>
we are overworked
<leviathanch>
well, they are
<libv>
...
<leviathanch>
I'm the new guy who is intended to take work load from their shoulders
<leviathanch>
libv: and it shows that we now are communicating
<leviathanch>
the only problem is these long holliday season they have just right now
<libv>
no, now you joined and feel that you are communicating on their behalf
<libv>
but i do not like noise, i like actual actions and results
<leviathanch>
hmm, I just repeat what they told me I should communicate to you in a better English they have
<libv>
and the more noise, the more suspicious i get
<leviathanch>
they are willing to help you
<leviathanch>
libv: I get that
<leviathanch>
libv: the will to support mainlining is there
<libv>
see, i dislike statements like "they are willing to help you"
<libv>
i prefer "they actually help"
<Wizzup>
Well, I'm curious to see how it works out if they pledge to help :)
<leviathanch>
libv: well... many of them just do not understand any English so it is impossible for them to communicate with any non-Chinese
nove has joined #linux-sunxi
<libv>
leviathanch: is that their excuse for abusing the gpl?
<leviathanch>
no
<el3>
leviathanch: ok. I actually had problems installing a toolchain that works. The guide on linux-sunxi gets me 4.8. and I get error message, that its known to compile with errors or something. That happens when I try to build mainline kernel. And I cant figure out to install 4.7 compiler
<leviathanch>
libv: their excuse for abusing the GPL is just time pressure
<libv>
leviathanch: which is a weak excuse for breaching the one license their whole business fully depends on
<leviathanch>
libv: we decided two weeks ago that we will drop the in-kernel binary blobs (except power vr)
<libv>
leviathanch: face it, as long as that is going on, whatever you state is completely useless
<libv>
leviathanch: allwinner has already dropped all the in kernel binary blobs
<libv>
literally
<libv>
allwinner does apparently have a very literal understanding of some english
<leviathanch>
libv: they do not have much of an understanding of English at all actually
<leviathanch>
we had chin-english conversations (supported with dictionary)
<libv>
leviathanch: i have heard promises from allwinner before
<libv>
i do not buy any of them anymore.
<libv>
actions talk
<libv>
if you do not have any actions, then stfu
<libv>
because everything else we have heard before
<leviathanch>
libv: yes, just wait for the A64 devboard
<leviathanch>
:-)
<libv>
leviathanch: oh, that is going to make us all forget the massive gpl breaches on previous chipsets?
<libv>
leviathanch: stating to us: "the next soc is going to fix all" is bullshit
<libv>
how many carrots are you going to dangle?
<libv>
and for how many years?
<libv>
the fact is, allwinner has gotten progressively worse with each soc release
<libv>
and claiming "the next big thing is going to be great"... well, i hear that about new free software projects all along
<libv>
and i don't believe those statements either
<leviathanch>
uhm...
<libv>
so how do you think i feel when such statements come out of a chinese chipmaker who is known to massively breach the gpl?
<leviathanch>
wel... I do since we decided this project together
<libv>
leviathanch: again, you are part of allwinner
<libv>
whatever you say, you say because you are on the payroll
<libv>
i have been given no convincing reason to see this otherwise
<libv>
and communicating is not helping with a guy like me.
<leviathanch>
uhm, yeah... you said actions
<libv>
so i suggest that you lay off of the communication
jelly has quit [Ping timeout: 245 seconds]
<libv>
and get stuff actually done.
<leviathanch>
libv: yes, I'm working on the u-boot right now
ssvb has quit [Ping timeout: 245 seconds]
<leio>
yes, lets see results, meanwhile public scolding just potentially kills the motivation and interest of those trying to fix it from the inside
<libv>
leio: this was just a stern reminder of allwinners hard obligations, and that talk is cheap.
<leviathanch>
plaes: yes... I communicate it to them
<el3>
leviathanch: the 2nd link you sent, shows me. Your connection is not private. And a button Back to safety. I just reinstalled ubuntu
<leviathanch>
el3: what?
<el3>
leviathanch: the 2nd link you sent, shows me. Your connection is not private. And a button Back to safety.
<leio>
self-signed SSL cert or some such.
<leviathanch>
ahh
<leviathanch>
okey, I see
<leviathanch>
just accept the cert
<leviathanch>
now I get it
<leviathanch>
just write an email to me if you need docs and support with mainlining
<el3>
Ah okey. got in now
<leviathanch>
plaes: just write an email to me if you need docs and support with mainlining
Akagi201 has quit []
<leviathanch>
plaes: I will then translate the questions into Chinese and pass it on to Allwinner
gianMOD has quit [Ping timeout: 244 seconds]
<leviathanch>
libv: same for you, when you have questions about hardware components, just send me an E-Mail
<leviathanch>
libv: I will translate it into Chinese and pass it on to Allwinner
<libv>
leviathanch: i do not have such questions at this time, i just want to ensure that allwinner meets its obligations.
<libv>
and that people do not get blinded or sidetracked by bullshit promises
<leviathanch>
libv: they promised to support my bachelors thesis (community FOSS tablet)
<a1d3s>
ssvb : after bisect i got these [59343cd7c4809cf7598789e1cd14563780ae4239] Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
ricardocrudo has quit [Ping timeout: 245 seconds]
<leviathanch>
libv: but now that I know how they live I can just urge everyone to be patient about email answers, they *will* answer
<leviathanch>
if they don't after half a week
<leviathanch>
try to use me as a translator
<leviathanch>
sometimes they just don't understand English
<leviathanch>
so they just don't answer because they are ashamed to admit
<leviathanch>
(what I learned during my trip there)
<leviathanch>
I'm working on supporting the cubie4 now in u-boot
cnxsoft has quit [Quit: cnxsoft]
<libv>
ashamed to admit errors does not fix them, it actually makes errors progressively worse.
<leio>
it's how chinese culture works, unfortunately
<libv>
so you are just doing a bachelors thesis?
<libv>
which translates in this case to: relocated to shenzhen, work for (almost) no pay, and for something which is utopic at best?
<libv>
how can i (and others who been along long enough) be expected to believe that allwinner takes any of this serious?
<libv>
s/along/around/
<libv>
so please, lay off the noise
<vishnup>
libv: they must have to pay the cost for breaching the GPL
<vishnup>
like Samsung has paid in past
<libv>
vishnup: like, actually investing the time and money to become fully compliant with the gpl?
<vishnup>
and had best practices to not to breach GPL
<libv>
vishnup: i see none of the sort happening.
<libv>
and no signs that they are earnestly investing in improving their behaviour
<vishnup>
libv: I'm not much familiar with legal terms
<libv>
leviathanch: in case you haven't caught this yet: allwinner has a massive and very deep set credibility issue, and you and your goals being there does nothing to ameliorate those.
<vishnup>
however, I wonder, is there any action taken on Allwinner with respect to GPL?
<libv>
not yet
ssvb has joined #linux-sunxi
<libv>
i told someone who was earnestly going to employ a known organization in legal action against allwinner, to wait many moons ago
<libv>
i should've picked things up at fosdem, but i was too overrun with devroom and attachments
<vishnup>
libv: great, allwinner must know that cost of breaching GPL is much more than investing in following GPL
<libv>
vishnup: that's the issue, there usually is no extra cost as no-one can or will pursue anything beyond full and complete compliance
<el3>
I got 5 pcduinos that I want to use in a comercial project
<plaes>
libv: did you have slides in your fosdem talk?
<libv>
plaes: yeah (well, barely), and a blogpost that i should polish and publish
<plaes>
ok.. I missed the livestream :(
<libv>
yeah, and the videos are still up in the air
<libv>
i really do think last years setup was preferable
<libv>
if only they could've managed the storage better last year
<el3>
vishnup: ok I have read it.
<NiteHawk>
btw: what happened to gpl-violations.org? the site seems down. i've noticed gpl issues with a router/repeater device i bought (not related to sunxi)
<nove>
leviathanch: we, this community is here to help allwinner, but we can't without allwinner, that is why this concerns
netlynx has quit [Ping timeout: 265 seconds]
<nove>
leviathanch: if we and allwinner work together for upstream, the license issues are resolved
<a1d3s>
a1d3s> ssvb : after bisect i got these [59343cd7c4809cf7598789e1cd14563780ae4239] Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
<leviathanch>
nove: that is exactly what we decided two weeks ago
<leviathanch>
nove: and I've got the job to translate between proper Chinese and proper English
<leviathanch>
nove: so that sunxi and Allwinner can communicate
<leviathanch>
nove: since many engineers there do not understand English very well
<leviathanch>
nove: the management doesn't talk a single word English
<ssvb>
a1d3s: do you mean after finishing bisecting? or after telling git the first "good"/"bad" pair of revisions?
<a1d3s>
ssvb, after good/bad part
<nove>
leviathanch: and that is very good
<leviathanch>
nove: no, now I will go and buy me some 馒头 and 烤鸭 in order to overcome my away-from-China melancholy
<ssvb>
a1d3s: now you need to compile this kernel, test it, and either say "git bisect good" (if there is no bug observed) or "git bisect bad" (if there is a bug)
<leviathanch>
ah
<leviathanch>
nove: yes
Akagi201 has joined #linux-sunxi
<a1d3s>
ok i will do so :)
<leviathanch>
nove: I forgot a w
<leviathanch>
damnit
<leviathanch>
ah
viccuad has joined #linux-sunxi
<leviathanch>
low sugar, I gotta buy some food
<leviathanch>
nove: can you write to the sunxi e-mail list, that I'm the official guy to ask, if they wanna know some details and specs from allwinner?
* leviathanch
is going to buy food now
<libv>
leviathanch: i do not buy it, you keep on telling us that we should ask you for details and specs about hw
<a1d3s>
ssvb, i m right if i put http://linux-sunxi.org/LCD config to u-boot/configs/bananapro_defconfig ? i used the mele config and changed the khz and mode xy for 7" display
<libv>
but the big list of big open questions that are blocking us are out there
<ssvb>
leviathanch: you can post this to the mailing list yourself :)
<leviathanch>
ssvb: ok
<libv>
why not solve those first and then come back to us?
<libv>
and stop promising us stuff which we know allwinner has never delivered upon
<leviathanch>
libv: it's just on a wiki, it's not written who wants the email with the PDF
<leviathanch>
libv: you need a pdf, just ask me
<libv>
what?
<a1d3s>
but cant find anywhere ft5x_ts
<nove>
leviathanch: is you, that most write to maillist and tell
<leviathanch>
nove: ok
<libv>
is it not enough that we have written our issues out on both wiki, endless emails, and even an issue tracker on github?
<libv>
i believe we all have "asked" more than enough, even though it is not our task to ask, or even demand, allwinner should just deliver on its own.
<diego71>
leviathanch: read the wiki, see the issue, post a url with the documentation. It looks easy to me....
<libv>
that is how this relationship should work.
netlynx has joined #linux-sunxi
<libv>
but do keep on dangling those carrots
<libv>
there are enough members of this community who believe empty promises and statements still.
<viccuad>
ssvb: sadly, that uses 3.16, not mainline (so not sfb). Do you know if debian arm nightly images with mainline exist? I recall seeing them, but google shows nothing
<ssvb>
viccuad: that's what is offered by the debian devs at the moment
<ssvb>
viccuad: maybe you can ping ijc
vishnup has quit [Ping timeout: 264 seconds]
<ssvb>
viccuad: of course you can build a newer kernel yourself if you want to, but this will be not quite 'official' debian anymore
<ssvb>
uniqdom: moving to 3.4.104 is more reasonable, and you can report problems if you find some
cubear has joined #linux-sunxi
<uniqdom>
thanks, I will :)
<ssvb>
uniqdom: I think it's mostly a matter of promoting the 'stage/sunxi-3.4' branch to 'sunxi-3.4'
<uniqdom>
I will try to compile it first without the RT patch
bonbons has quit [Quit: Leaving]
bonbons has joined #linux-sunxi
montjoie[home] has quit [Remote host closed the connection]
montjoie[home] has joined #linux-sunxi
paulk-aldrin has joined #linux-sunxi
<T0mW>
I've some driver code working with 3.4.102 that I wish to move onto 3.19. I need an example of how to use the external interrupts (PH16) "the right way". My present code hacks at reading / writing into the PIO interrupt register set.
<T0mW>
Any code / docs that I can look at?
uniqdom has quit [Ping timeout: 250 seconds]
gianMOD has joined #linux-sunxi
Zboonet has joined #linux-sunxi
vishnup has joined #linux-sunxi
<el3>
T0mW: Just built my first yocto minimal core image. But some errors while building http://pastebin.com/Ky9yNGq5 And runqemu wont run. No image present, probably not built
<el3>
T0mW: sry for asking here. But people in yocto seem very quiet
<T0mW>
el3: I haven't used hte qemu, usually have the hardware to put the code onto.
<el3>
T0mW: Then maybe I should try build a debian image for pcduino3?
gianMOD has quit [Ping timeout: 265 seconds]
<T0mW>
el3: let me see your conf/local.conf
gianMOD has joined #linux-sunxi
Andy-D has quit [Ping timeout: 252 seconds]
<el3>
T0mW: 2 sec. my ubuntu bugs very much. Cannot see letters
<diego71>
ssvb: it works, and now we are using it to install debian
gianMOD has quit [Ping timeout: 264 seconds]
uniqdom has joined #linux-sunxi
<diego71>
ssvb: ok, debian installer complaint about not having the right kernel modules to install ... :)
<diego71>
and i was wondering if it can work with u-boot
<T0mW>
diego71: that looks like the old TTL parallel inteface (40 pin)?
<T0mW>
diego71: I'm using LVDS on a 7" display
<T0mW>
FWIW, the stock linux-3.19 is missing a device tree (dtsi) definition for anything else but the HDMI, you have to add the definiton for the LCD.
<diego71>
T0mW: yes it use a 40 pin connector
<diego71>
T0mW: how hard is it to add a definition for the lcd?
<ssvb>
diego71: if you have a fex file for it, then it can work
<MY123>
When will a true KMS driver be published?
<a1d3s>
T0mW, how you get it working? 7" lvds
<ssvb>
MY123: how much are you willing to pay for its development?
<MY123>
ssvb: hmm...
vishnup has joined #linux-sunxi
<diego71>
ssvb: dtsi should be only for the uboot right? Is there a tool for converting fex to dts?
<ssvb>
diego71: currently the LCD settings are configured in u-boot defconfigs
<MY123>
ssvb, 20 euros?(as you can always buy a $Raspberry_Pi_2 which already has a DRM/KMS driver)
<ssvb>
diego71: the http://linux-sunxi.org/LCD page provides instructions and also a bunch of automatically converted settings for the fex files, which are present in the sunxi-boards repository
<MY123>
with a proper Mesa
<libv>
MY123: go ahead and buy an rpi
<diego71>
ssvb: ok, i found it. Thanks.
<libv>
MY123: do you know why the rpi has a free driver at all?
akaizen has quit [Ping timeout: 246 seconds]
<MY123>
libv: I think because of BroadcomCM.
<MY123>
(running a modified Android)
<libv>
MY123: in any case, making endless demands and trying to resort to some cheap form of emotional blackmail, that's not going to motivate anyone to write or produce code
<libv>
MY123: no, it is because some people were freeing ARM GPUs, and because the rpi foundation was being seriously showed up by their big statements but by not having anything to back them up
<libv>
when they then did and made another big statement without any content, someone called them out
<MY123>
libv: I know, Oct.2012 code was just shit
<libv>
MY123: oh?
<libv>
MY123: how did you know?
<MY123>
libv: I wrote a wayland-egl abstraction
<libv>
when?
<MY123>
libv: 2014
<libv>
haha.
<libv>
go and read up.
<MY123>
libv: and completly disassembled the blob
<MY123>
( a unstripped version)
<MY123>
libv: There is a VP8 reference decoder, clockman, powerman, a shader compiler,...
<MY123>
and many other things
<ssvb>
MY123: sounds like you are perfectly capable of developing the sunxi KMS driver yourself, just go ahead
<ssvb>
MY123: and I even can give you the 20 euro tip when it's done :)
<MY123>
ssvb: I only wrote USB driver
<MY123>
(s)
gianMOD has joined #linux-sunxi
<MY123>
Never coded everything beyond GPU buffer management
<libv>
MY123: you, on your own, disassembled the videocore blob?
<MY123>
libv, yeah
<libv>
using what disassembler?
viccuad has quit [Quit: WeeChat 1.0.1]
<MY123>
libv: GNU binutils ( github.com/mm120/binutils-vc4 branch vc4)
<libv>
were you actively involved with the videocore REing project?
<MY123>
I tried to code GPU buffer management for RPi Android, but it crashed
<libv>
if so, you would have known better than to try and argue here
<MY123>
after starting the launcher
<MY123>
Never had a complete success :(
premoboss has joined #linux-sunxi
gianMOD has quit [Ping timeout: 264 seconds]
<MY123>
libv: as of VideoCore, there is a whole part of the blob currently managing HDCP, even when it's off
<libv>
MY123: if you like the rpi so much, why are you here?
<T0mW>
ald3s currently, I'm using u-boot-2015.01 and the linux-sunxi 3.4.102 kernel. You need to modify the fex (script.bin) to enable the LVDS
<MY123>
libv: I currently have three running A20 devices
<MY123>
libv: running the mittorn/sun7i-nand-sunxi driver trashed boot0
<MY123>
And I still did not find how to update_boot0 from a ARM computer
<MY123>
(after compiling it)=
<ssvb>
MY123: ask your device vendor?
akaizen has joined #linux-sunxi
<ssvb>
MY123: they might have a PhoenixSuite recovery image
<MY123>
ssvb: which does not want to ( because there is DRM keys inside, it's Kurio)
<ssvb>
MY123: you were just playing with fire trying experimental nand stuff without having a backup plan for recovery
<ssvb>
MY123: normally in this case it's best to leave nand alone and boot gnu/linux from the sd card
<MY123>
ssvb: I should still *have* a leaked Aug. 2014 NAND image in the other side of the English Channel.
<MY123>
(that's a little too far, just for a tablet)
<MY123>
ssvb: I can still access to all the partitions
<MY123>
(even bootloader, I only need boot0/10
<MY123>
*boot0/boot1)
guizamboni has quit [Quit: Leaving]
<MY123>
Let's see if " Booting Android from Sdcard" will work
iamfrankenstein has quit [Ping timeout: 250 seconds]
ricardocrudo has quit [Ping timeout: 264 seconds]
Zboonet has quit [Ping timeout: 264 seconds]
<T0mW>
a1d3s: ping
Zboonet has joined #linux-sunxi
WarheadsSE has left #linux-sunxi ["WeeChat 1.0.1"]
FreezingAlt has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 245 seconds]
Zboonet has quit [Ping timeout: 244 seconds]
vishnup has quit [Remote host closed the connection]
gianMOD has joined #linux-sunxi
Zboonet has joined #linux-sunxi
gianMOD has quit [Ping timeout: 264 seconds]
<mripard_>
T0mW: interrupts-extended = <&pio <bank> <pin> <flags>>; in the dt
gianMOD has joined #linux-sunxi
<mripard_>
and then the usual irq API in your driver
<T0mW>
ok
MY123 has quit [Quit: Leaving]
<T0mW>
mripard_: external interrupt handling of sunxi should be transparent, IOW, no need to bang pins / register bits. Simply IRQ_HANDLED and underlying code with ACK to corresponding pin ctrl regsiter, etc.?
<T0mW>
s/with/will/
gianMOD has quit [Ping timeout: 244 seconds]
premoboss has quit [Remote host closed the connection]
<mripard_>
T0mW: good thing you don't need to then
<T0mW>
mripard_: I've done interrupt code under X86 and it was straight-forward, the linux-sunxi external interupt examples show the driver configuring gpio pins to interrupt mode, then config the GIC masks, etc.
<mripard_>
first, you don't need to configure the GIC mask
<T0mW>
good
<mripard_>
you're not using the GIC in the first place.
<T0mW>
ok
<T0mW>
PIO controller thingy
<mripard_>
then, it depends
<mripard_>
you have two options, either you just declare it as an interrupt, and that's it
<mripard_>
with interrupts or interrupts-extended in the DT
<T0mW>
PIO_INT_STAT_OFFSET, PIO_INT_CTRL_OFFSET and so forth is what I have now.
<T0mW>
mripard_: is this part of hte device tree docs?
<mripard_>
yes.
<mripard_>
google ePAPR
<T0mW>
mripard_: ok, so I can grep down on 'interrupt-extended' into the source.
<mripard_>
if you have an interrupts property, the only thing you need to do is caling request_irq
<T0mW>
mripard_: AH! thanks
<T0mW>
cool
<mripard_>
if you can't use interrupts proprety for some reason
<mripard_>
then you need to declare is as gpio
<mripard_>
and call gpio_to_irq on it
<mripard_>
(an then call request_irq)
<mripard_>
so you don't need to poke anything in the PIO in your driver
<T0mW>
mripard_: yeah, I was going down the gpio_to_irq path a few minutes ago. Okay, will continure reading, just need to get pointed in the right direction.
<mripard_>
(and btw, saying it works like that on x86, it must work like that on ARM is a bit bogus.)
WarheadsSE has joined #linux-sunxi
lerc has joined #linux-sunxi
Renard has quit [Ping timeout: 244 seconds]
viccuad has joined #linux-sunxi
Renard has joined #linux-sunxi
gianMOD has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
gianMOD has quit [Ping timeout: 250 seconds]
nove has quit [Quit: nove]
popolon has joined #linux-sunxi
<uniqdom>
sunxi-v3.4.104 seems to be working fine, at least for A20, cubieboard2
orly_owl has quit [Ping timeout: 256 seconds]
<uniqdom>
and now, 8 patches were rejected trying to apply preempt_rt patch in that kernel version
orly_owl has joined #linux-sunxi
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 264 seconds]
tomcheng76 has quit [Read error: Connection reset by peer]
iamfrankenstein has joined #linux-sunxi
tomcheng76 has joined #linux-sunxi
Zboonet has quit [Ping timeout: 245 seconds]
konradoo77 has quit [Ping timeout: 246 seconds]
Zboonet has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
F1skr has quit [Quit: WeeChat 1.1.1]
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 245 seconds]
Zboonet has quit [Ping timeout: 255 seconds]
FreezingAlt is now known as FreezingCold
topi` has joined #linux-sunxi
<topi`>
there's a new Orange pi + based on the H3 chip... anyone seen these?
<topi`>
it's said to support SATA disks and Gigabit ethernet however the H3 doesn't have these and the ethernet on H3 is just 10/100 so what's the deal?
<NiteHawk>
there are several models. Orange Pi, Orange Pi Mini and Orange Pi Plus. they have different hardware features, and the Plus is the H3 one (with both gigabit ethernet and SATA)
<topi`>
if the GigE is implemented by an auxiliary chip, ditto SATA, how can they provide full bandwidth if they're limited by the USB2.0 bus?
netlynx has quit [Quit: Leaving]
<NiteHawk>
the pcb images show a separate ethernet chip. i've no idea how it is coupled with the cpu
<topi`>
a peek at the schematics would reveal on which bus they're hanging in, but most probably USB2.0
<topi`>
I remember a benchmark of the original Banana PI, which most probably has similar ethernet/sata arrangement, and it showed GigE speeds easily exceeding 400 Mbps ...
<topi`>
well, theoretically, the USB2.0 can provide 480Mbps
Zboonet has joined #linux-sunxi
<NiteHawk>
the BPi has a gigabit PHY that's directly connected to the A20
FreezingAlt has joined #linux-sunxi
<topi`>
so, it's on the AMBA bus?
<topi`>
I don't know which core interconnect the A20 uses, probably CCI-400?
FreezingCold has quit [Ping timeout: 240 seconds]
<topi`>
I think, overall, the new Orange Pi Plus looks promising, a quad A7 at 1.6 GHz is much faster than the new Raspberry Pi 2
steev has quit [Ping timeout: 250 seconds]
<topi`>
the big question is, how well is the new H3 chip supported by linux-sunxi?
konradoo77 has quit [Ping timeout: 240 seconds]
<ssvb>
topi`: it depends on how similar it is to the other Allwinner SoCs
konradoo77 has joined #linux-sunxi
ojn has quit [Read error: Connection reset by peer]