Black_Horseman has quit [Ping timeout: 246 seconds]
<wens>
didn't we go over this yesterday? :p
* orly_owl
scrolls
HeHoPMaJIeH has quit [Quit: Leaving]
<orly_owl>
k
<ccaione>
lol
<orly_owl>
sun-x-i i guess
smotocel69_ has quit [Remote host closed the connection]
Rhys79 has joined #linux-sunxi
Rhys79 is now known as rhys79
xavia has joined #linux-sunxi
<rhys79>
Hello
rhys79 is now known as Rhys79
<Rhys79>
Anybody active in here? Have some technical questions I was hoping someone might be able to help me with.
<mru>
see topic, just ask and wait
<Rhys79>
Ahh, yep, helps if I read *smh*....
<Rhys79>
I have some LVDS related questions. First, is there support for configuring an LVDS display using EDID in Linux-Sunxi, and if so, what I/O lines would I attach the EDID lines to on an AllWinner A20?
<Rhys79>
Second, is there support for backlight control with LVDS, or will I need to write a program myself to interface the backlight controls using GPIO?
<mru>
edid is an i2c interface
<Rhys79>
Ok, that is a starting point :)
<Rhys79>
That explains why EDID is a two wire interface, CLK and DATA. I haven't used I2C yet, I'll have to read up on that. Now I know what IO to use for the EDID lines.
<Rhys79>
You wouldn't happen to know if the EDID data can be read out without powering the main display processor? I could at least get the configuration data off the panel that way so I can manually configure it. There's instructions for that on the Linux-Sunxi Wiki.
<mru>
that would depend on the panel
<mru>
for TVs and desktop monitors it's usually possible
<mru>
an lcd panel for embedding is another story
<Rhys79>
I can't find a datasheet for the LG LP154W01-TLAD panel I have, and the timings are different from one version to the next for the similar ones I can find datasheets for.
<mru>
does it even have an edid interface?
<Rhys79>
The other LP154W01 datasheets I could find indicate it does support EDID
<Rhys79>
Pins 4,6,7 on the FI-X30M connector on the display
<Rhys79>
Seems to be part of the standard pinout for LVDS display panels, and every LG datasheet I've looked at indicates they do have an EDID ROM on the panel.
<mru>
so hook it up to any i2c port on the board and try reading it
<mru>
there's got to be some userspace tool that'll dump data from an arbitrary i2c address
<Rhys79>
I'll give it a shot when I have a chance. I have a pile of different dev boards that have I2C on them.
<mru>
no need for a full display driver
<mru>
or wire it into a vga port on a pc :)
<mru>
making sure voltages are compatible of course
<libv>
Rhys79: attach it to twi2
<Rhys79>
I can write a small program for one of my TI Launchpads to scan the I2C bus, dump the EDID ROM and spew it to the serial console.
<libv>
that is, if you are not using a full and proper vga connector
<libv>
Rhys79: there is a pwm which is used for backlights
<Rhys79>
libv: Ok, will that allow Linux-Sunxi to automatically configure the panel, instead of having to manually enter the timings in the UBOOT config?
<libv>
Rhys79: not yet
<libv>
Rhys79: but cubietruck has vga i2c attached to twi2
<libv>
it's the only known device to link up ddc so far
<libv>
and my sunxi-kms driver will be making use of this
<starchild>
do you guys recommend cubieboard2 for a dvb-s2 set-top box?
<libv>
Rhys79: for the time being, you will need to edit the fex
<libv>
also...
<libv>
there is no way for us to find out over edid what type of a connection a panel has
<libv>
base edid says just "digital" or "analog"
<Rhys79>
libv: That sucks.... What's missing at this point to make it work. I just got my iTeaduino Plus A20 last week, so I'm still getting familiar with everything and figuring out what I can and can't do currently and what I may need to work on.
<Rhys79>
libv: I don't mind having to manually tell UBOOT what display port to use, but I would like to be able to tell it to use LVDS, and then pull all of the data that you would normally need a datasheet to find out from EDID.
<Rhys79>
libv: Then I can use any random LVDS panel I strip out of a laptop assuming I have the proper cable and the panel supports EDID.
<libv>
Rhys79: you will end up being on your own with that
<Rhys79>
libv: Without having to spend hours searching the 'net for a datasheet
<libv>
so it is not likely that i will check for that in my kms driver
<libv>
Rhys79: then just attach it to twi2 per default, and if needs be, pull it manually
<Rhys79>
libv: I don't mind doing the leg work if it's within my skill level. Usually I just need someone to point me in the right direction and say have at it :)
<libv>
Rhys79: well, do some legwork, and document the iteaduino properly first.
<Rhys79>
libv: WIll do, it's sitting right in front of me :) Pretty slick little board. I went with it over the Cubieboard2 because of the processing core being a plug-in module. My intentions for it are to create my own baseboard for 3D printer use to create an all-in-one 3D printer controller that doesn't require a secondary processor for stepper control.
<libv>
Rhys79: the chances of this baseboard being useful are pretty small
<mru>
I'm sceptical towards baseboard + cpu module designs
<libv>
Rhys79: it only works for a10 and a20
<mru>
most of them end up only ever supporting a single cpu module anyway
<libv>
and you already have the more powerful version
<libv>
but then, if you wish to design your own baseboard, then yes, this is the better option i guess
<libv>
alternatively, you could just use the olimex cad files
<Rhys79>
I believe iTead Studio is also working on an A31 plug-in module as well.
<Rhys79>
I can't see any particular reason why an i.MX6 module couldn't be designed as well to fit the same form factor.
<libv>
i am not sure whether they can make that one would be 100% compatible
<libv>
Rhys79: because if you do, you have to limit the exported functions to the shared functionality
<libv>
a10 and a20 is easy, as they are pin-compatible
<libv>
everything else, and you will be slicing and dicing
<libv>
now given that they are not the most clueful a company, and that they did their original design on a10 and a20...
<libv>
Rhys79: how long has the a31 been out btw?
<libv>
how long has the iteaduino been out?
<Rhys79>
True. It serves my purposes though. I don't need every last IO from the chip. The existing modules have far more IO than I actually need.
<libv>
how long have they been promising the a31 iteaduino?
<Rhys79>
The iTeaduino has been out for a year or so, maybe a little longer. The A31 module I've seen mentioned but no promise of it being produced. I'm not really concerned about it though. The A20 should be able to handle what I want to do with it.
<libv>
Rhys79: then you would've been just as happy with a cubietruck or an a20-olinuxino-micro (OSHW!)
<libv>
Rhys79: but never mind, this way we get the device finally documented for our project :)
srxa has joined #linux-sunxi
<libv>
Rhys79: as the resident graphics driver developer, you will need to keep on hacking .fex/devicetrees for this panel stuff to work
<libv>
it's going to be easier in the long run
<libv>
and if the lcds have ddc, then just hook them to twi2, and pull in the ddc manually
<libv>
err s/the ddc/the edid/
xavia has left #linux-sunxi ["PART #interlock :PART #arduino :PART #wnydev :PART #buffalolab :PART #docker :PART #symfony :PART #virtuvia :QUIT :Leaving."]
<Rhys79>
Actually, the Cubieboard and olimexino have a crap ton of stuff on it I don't need on them, that's why I went with the A20 core. I want to be able to design a baseboard that has exactly what I want on it and non of the extra stuff. I won't need video out (other than LVDS), no SATA, no ethernet, but I do need five stepper drivers, wireless, bluetooth, and a bunch of screw terminals.
<Rhys79>
The A20 core has most all of the IO dropped onto two rows of headers. Plop that in the middle of the baseboard and I'm good to go. If iTead ever does release an A31 core using the same footprint, it would be a drop in replacement for a faster processor.
<libv>
Rhys79: the good thing about olimex, is that it is oshw, and you can do whatever you want with that
<libv>
also, a31 is not the most popular SoC, and if iteaduino is in any way smart, it will be looking at a newer generation already
<libv>
a80 is supposed to happen soon
<Rhys79>
Most of iTeads stuff is OSHW. The baseboard files are all available, and the A10/A20 core schematic is available. The only thing they haven't posted is the core PCB design.
<libv>
the first sentence is incorrect, the latter two sentences are correct
<libv>
but yes, if it suits your needs, why not
Gerwin_J has quit [Ping timeout: 252 seconds]
<Rhys79>
libv: Ok, I'll rephrase that, most everything I care about from iTead is OSHW.....
<libv>
:)
<Rhys79>
I'm generally vendor neutral. I don't really care who makes something, as long as it does what I want it to without being forced to pay for a bunch of additional garbage I don't need :)
<Rhys79>
Preferably without having to make expensive one-off creations.
<Rhys79>
I am part of the hackerspace started by the guys that own SeeMeCNC. Hence my interest in putting together a complete all-in-one 3D printer controller design. Something that can go from STL to print in a single package.
<Rhys79>
The A20 core makes for the most efficient design option.
<libv>
nice.
<libv>
the a13 tablet i just got does not seem to share uart pads with the sd-card
* libv
closes the device up again for the final pictures.
<Rhys79>
Having a 3D printer manufacturer as a friend comes in handy :) All sorts of fun tools I can go play with when I need them....
<F1skr>
Has anyone tried to boot the latest mainline (after ahci-sunxi support) with a sata disk as root? Mine hangs at "Waiting for root device /dev/sda4.." on a cubieboard1
<Rhys79>
For all practicle purposes the A20 core is identical to the A10 core. I'll see if I can get the Wiki updated for you guys though. Then I'll start looking into the graphics driver settings.
<Rhys79>
Anyone know if BSP supports the iTeaduino Plus? I have manually built a uBoot image, but didn't have any luck with BSP building an HW-Pack
xinj has quit [Read error: Connection reset by peer]
xavia has joined #linux-sunxi
xinj has joined #linux-sunxi
souther has quit [Ping timeout: 246 seconds]
dolgosheev has joined #linux-sunxi
srxa has quit [Remote host closed the connection]
<libv>
i never have used the bsp, hwpacks, or images
dolgosheev has quit [Remote host closed the connection]
Skaag has quit [Ping timeout: 245 seconds]
xinj has quit [Ping timeout: 255 seconds]
<Rhys79>
Ok, I was planning on building my own anyways. Just was curious for the Wiki.
prasannapete has quit [Quit: prasannapete]
<Rhys79>
Just made a bunch of edits to the iTeaduino Plus A10 wiki page. Really, the A10 and A20 should share a page, as for all practical purposes, they are identical except which processor is on the core board.
<Rhys79>
Not sure how to do that though....
souther has joined #linux-sunxi
<libv>
seems like you will have to find out on your own then, i cannot do everything and document everything on my own.
lioka has quit [Ping timeout: 246 seconds]
lioka has joined #linux-sunxi
<libv>
Rhys79: please use the device page example.
<libv>
as otherwise that's going to cost me another hour, again.
lioka has quit [Changing host]
lioka has joined #linux-sunxi
<libv>
ah, and the reason why you do not have a bsp is because you haven't worked through the new device howto.
<libv>
tldr for you?
rta has quit [Ping timeout: 246 seconds]
<Rhys79>
tldr???
geecko_ has quit [Read error: Operation timed out]
<Rhys79>
Sorry, don't know that acronym.
<Rhys79>
I'll work through the new device walkthrough as soon as I have time.
FreezingCold has joined #linux-sunxi
<Rhys79>
I started on it but keep getting interrupted....
geecko_ has joined #linux-sunxi
<Rhys79>
Gotta run for now, TTYL. Thank you again for the assistance!
nove has joined #linux-sunxi
Rhys79 has quit [Quit: Page closed]
rta has joined #linux-sunxi
arky has quit [Quit: Leaving]
arky has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 240 seconds]
arky has quit [Client Quit]
<juanfont>
nove, ping
<nove>
starchild: vdr is somewhat working, but rellla knows more
<nove>
juanfont: pang
arky has joined #linux-sunxi
<juanfont>
hi :) i've been messing with the cedarv binaries you shared some days ago
<juanfont>
did you try them?
<nove>
juanfont: no time
<nove>
juanfont: are they working? and is really armhf?
<juanfont>
well, the armhf version is, in fact, armhf... but it is compiled using the ION memory allocator
<juanfont>
instead of using the sunxi one. so... useless. i think.
<juanfont>
looking to old libcedarv code, it seems that in order to use the sunxi allocator you just have to edit a #define option. where did you get those binaries?
<nove>
juanfont: which binaries? be a bit more clear
<nove>
starchild: well not the best log link to point,
<nove>
starchild: this vdpau thing is not the end result that we are aiming, it just is something to experiment and vdpau was chosen because is the api with bigger players support
<starchild>
heh vdr osd looks like the star trek deck :-D
<starchild>
any timeframe for when one can use an A20 device on debian as a set-top box?
Gerwin_J has joined #linux-sunxi
<nove>
starchild: is hard to give a date, but as you saw in the video, there are people already trying this use
Fusing2 has quit [Read error: Connection reset by peer]
<nabblet>
lol
FreezingCold has quit [Ping timeout: 240 seconds]
<mnemoc>
oliv3r: it's good that there is still some innocense in you
<nabblet>
although my BS-Meter was off scale when i read "magical Windows SD card cache accelerat"
Fusing has quit [Ping timeout: 252 seconds]
<mnemoc>
innocence*
<oliv3r>
hehe
<oliv3r>
well at first i thought it was a windows themed distro
<mnemoc>
even the real windows deprectated the "metro" theme :p
jemk has joined #linux-sunxi
Rhys79 has joined #linux-sunxi
<Rhys79>
Having an issue with creating a livesuit image. The script is looking for an ~/sunxi-bsp/allwinner-tools/livesuit/a20 folder that does not exist. There are only a10 and a13 folders there. Any idea where to acquire this folder from?
KevinEG has quit []
<Rhys79>
Just checked the github repo, and it's not there either. I'm guessing we can't build livesuit images for A20 devices?
<mnemoc>
iirc the open source pack/unpack worked for a20 just fine
akaizen has joined #linux-sunxi
<nabblet>
lucky guess - but can't the A10 image be used for A20?
arky has quit [Ping timeout: 240 seconds]
<nabblet>
live image that is
<Rhys79>
Hmm, the one A10 image I tried to use would not boot on the A20. Regardless, I don't want a stock image, I'm trying to use the linux-sunxi BSP tools to build my own. However, when it gets to the point of "Make boot0 boot1" it fails because it is looking for a directory that doesn't exist.
<megal0maniac_afk>
Rhys79: Have you successfully flashed an image using livesuit?
<megal0maniac_afk>
(To that board)
<Rhys79>
nope, been running off of SD
<Rhys79>
iTeaduino Plus A20, doesn't come with anything loaded on the NAND and have not tried flashing anything to it yet. Just trying to build an image.
deasy has quit [Remote host closed the connection]
<Rhys79>
I guess I could try loading a CB2 image to it. Those supposedly work.
<megal0maniac_afk>
Ah, I see. I've got the same (core) board, and LiveSuit isn't playing nicely
<megal0maniac_afk>
Rhys79: Not sure about the NAND images, but SD images yes
physis has joined #linux-sunxi
<Rhys79>
Well, I'm going to grab one and try it :) I did plug the board in with LiveSuit running and it threw an error because I hadn't selected an image yet. So, it at least sees the device....
<megal0maniac_afk>
Mine constantly says "Invalid image file" regardless of the image file, and has "Failed to jump to update mode" in the program window. Not much in the way of debug output :/
Fusing has joined #linux-sunxi
<Rhys79>
downloading CB2 lubuntu now, we'll see how it goes :)
<Rhys79>
Yep, image file invalid
megal0maniac_afk has quit [Ping timeout: 246 seconds]
notmart has quit [Quit: notmart terminated!]
Fusing has quit [Ping timeout: 255 seconds]
<Rhys79>
Haha, got phoenixsuit to recognize it and loading firmware now!
megal0maniac_afk has joined #linux-sunxi
<Rhys79>
megal0maniac_afk: Haha, got phoenixsuit to recognize it and loading firmware now!
<megal0maniac_afk>
Rhys79: Just saw the log, great!
arky has joined #linux-sunxi
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
<ssvb>
mnemoc: how much time do we have left until promoting stage/sunxi-3.4 to sunxi-3.4?
<mnemoc>
ssvb: as much as you want :p
<mnemoc>
ssvb: or until tonight if the current is considered good
<mnemoc>
ssvb: should I merge/tag/jump-to-next-stable ?
<ssvb>
jemk: basically, they only protect an assignment to a variable instead of ensuring that /dev/cedar_dev is only opened exclusively by a single user
<mru>
ssvb: semaphore down/up around writing to devp->irq_flag?
<mru>
that's wrong on so many levels
Rhys79 has quit [Quit: Page closed]
<nove>
ssvb: there is example code from allwinner that shows simultaneous encode/decoding, and the device is open two times
<ssvb>
mnemoc: yes, it has been a long time since the last release
<jemk>
ssvb: hm, does it even do anything there?
<mnemoc>
ssvb: agree.
<ssvb>
jemk: if we agree that it is fine to have only a single user at a time for the cedar driver, then the fix is trivial
<Wizzup>
ssvb: Maybe settle on it being single user for the time being
<jemk>
ssvb: i think many blob-based players open cedar_dev multiple times, once in the blob and once in the wrapper lib to allocate memory
<jemk>
for me it would be ok xD
<ssvb>
nove: has anybody tested/confirmed that it works with the current linux-sunxi kernel?
netlynx has quit [Remote host closed the connection]
<jemk>
nove: but that's in one program, right? alternating is no problem, as long as you have full control about both parts and ensure they don't work at the same time
<mnemoc>
ssvb: sunxi-3.4.79-bfs ? :p
<nove>
ssvb: that example were garbage, armel binaries, i only made it to decode/encode 1 frames it would stop after that
<ssvb>
jemk: in any case, I will just test libvecore based vlc & xbmc to see if they break :)
<ssvb>
mnemoc: I have it just for benchmarking purposes, it happens to speed up gtkperf quite a lot :)
<oliv3r>
do the 3.4 kernels of ours have the GPIO stuff enabled by default?
<oliv3r>
gpio via sysfs i mean
<mnemoc>
ssvb: :D
nabblet has quit [Ping timeout: 252 seconds]
<ssvb>
mnemoc: but from what I have read, bfs is not very compatible with systemd :(
\\Mr_C\\ has joined #linux-sunxi
<mnemoc>
die systemd die
<mru>
systemd is not very compatible with anything but possibly itself
<mru>
and even that is doubtful
<ssvb>
mnemoc: damn pottering
<mnemoc>
*g*
<libv>
mnemoc: i heard the following via via
<libv>
kai and lennart at the latest redhat event...
<nove>
ssvb: enforcing one user should be okay, but new binaries, how knows what they will do, and that could complicate tracing
<libv>
talking in a small group about the advantages of kdbus...
<libv>
and kai throwing in "yeah, with kdbus people will have no other option than to use systemd!"
<jemk>
ssvb: btw, if you have some time, set .tpr3=0x00072222 and see how far you can overclock
<libv>
(double paraphrase there)
<jemk>
on CT
<libv>
world domination...
<mru>
libv: that's been their goal for a long time
<libv>
and building an empire to last a thousand years.
<mru>
slowly squeeze people out of other options
<libv>
they started by annexing austria^Wudev
<mru>
lennart is on record saying he looks forward to the day udev doesn't work without systemd
<ssvb>
deasy: which spec? in any case, you can get more than 2GB/s speed for memset on A20, it would be simply impossible with only 480MT/s and 32-bit memory interface
<deasy>
ssvb, the ram spec
<Nazcafan>
I noticed that the cubietruck wifi and bluetooth modules have landed in sunxi-devel kernel, are there equivalent commits for the 3.4 "standard" branch?
<deasy>
if it cans run at ddr 960
<F1skr>
Nazcafan: have you gotten the wifi to work with sunxi-devel?
<megal0maniac_afk>
I give up. Again.
<ssvb>
deasy: oh, the used memory modules are typically even rated as ddr3-1333 :)
<deasy>
i have a cubieboard 1 :)
<deasy>
i prefer follow spec
<deasy>
if the MC have spec for 400 max i set 400
<deasy>
(sorry for my wrong english)
<deasy>
has*
<Nazcafan>
F1skr, not yet, need to rebuild first, why? is it not working at all?
kivutar has quit [Quit: Ex-Chat]
<F1skr>
Nazcafan: I have not been able to get it to work yet, I load the module but I get no wlan device.
jemk has quit [Remote host closed the connection]
bertrik_ is now known as bertrik
<Nazcafan>
I won't have time to try it out tonight, but maybe tomorrow evening, I'll let you know how lucky I got
leviathanch_ has quit [Remote host closed the connection]
leviathanch has joined #linux-sunxi
bgal has quit [Ping timeout: 240 seconds]
<bertrik>
is there some special extended register set in the axp21X that gets enabled by writing to register 0xFF ?
Nazcafan has quit [Read error: Connection reset by peer]
xavia has left #linux-sunxi ["PART #interlock :PART #arduino :PART #wnydev :PART #buffalolab :PART #docker :PART #symfony :PART #virtuvia :QUIT :Leaving."]
Nazcafan has joined #linux-sunxi
t3st3r has quit [Remote host closed the connection]
leviathanch has quit [Read error: No route to host]
leviathanch has joined #linux-sunxi
mturquette has quit [Changing host]
mturquette has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 255 seconds]
linkmauve1 has joined #linux-sunxi
book` has quit [Quit: ERC Version 5.3 (IRC client for Emacs)]
book` has joined #linux-sunxi
akaizen has quit [Remote host closed the connection]
bertrik has quit [Remote host closed the connection]
<physis>
Guys, some one have XBMC with HDMI audio working fine using HDMI? I found this issue (https://github.com/rellla/xbmca10/issues/18) from rellla (I'm using A20) but still not working even recompiled with this changes.