rz2k changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - Don't ask to ask. Just ask and wait! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi | FOSDEM talks - http://dl.linux-sunxi.org/users/nove/sunxi_at_fosdem2014/
FreezingAlt has quit [Ping timeout: 240 seconds]
popolon has quit [Quit: Quitte]
pwhalen has quit [Ping timeout: 264 seconds]
leviathanch2 has joined #linux-sunxi
kivutar has quit [Quit: Ex-Chat]
pwhalen has joined #linux-sunxi
rz2k has quit []
moofree_ has quit [Ping timeout: 240 seconds]
moofree has joined #linux-sunxi
newleaves_ has joined #linux-sunxi
hero100 has joined #linux-sunxi
setkeh1 has joined #linux-sunxi
hero100 has quit [Ping timeout: 264 seconds]
Gerwin_J has joined #linux-sunxi
deffrag has quit [Ping timeout: 240 seconds]
deffrag has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
TheSeven has quit [Remote host closed the connection]
<wens_> great, got a tablet claiming to be A23 but is actually an atm7021A :/
wens_ is now known as wens
bsdfox has quit [Ping timeout: 244 seconds]
bsdfox has joined #linux-sunxi
TheSeven has joined #linux-sunxi
geecko has quit [Ping timeout: 240 seconds]
egbert has quit [Ping timeout: 264 seconds]
egbert has joined #linux-sunxi
FreezingAlt has joined #linux-sunxi
ssvb has joined #linux-sunxi
<Turl> wens: :/
<Turl> wens: what's atm?
<Turl> ah, actions
<Turl> they're also "Q88", confusing
<wens> Yeah, but the seller explicitly mentioned A23, so I'm asking for a replacement
bsdfox has quit [Ping timeout: 240 seconds]
leviathanch2 has quit [Ping timeout: 240 seconds]
bsdfox has joined #linux-sunxi
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 240 seconds]
pacopad has joined #linux-sunxi
pacopad_ has quit [Ping timeout: 244 seconds]
FreezingAlt has quit [Read error: Operation timed out]
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 265 seconds]
pacopad_ is now known as pacopad
pacopad_ has joined #linux-sunxi
pacopad has quit [Ping timeout: 240 seconds]
pacopad_ has quit [Ping timeout: 244 seconds]
pacopad has joined #linux-sunxi
pwhalen has quit [Ping timeout: 240 seconds]
pwhalen has joined #linux-sunxi
pwhalen has quit [Changing host]
pwhalen has joined #linux-sunxi
leviathanch has quit [Ping timeout: 245 seconds]
leviathanch has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
penghb has quit [Ping timeout: 252 seconds]
pseudomind has quit [Ping timeout: 240 seconds]
ijc_ is now known as ijc
penghb has joined #linux-sunxi
arky has joined #linux-sunxi
pacopad has quit [Quit: pacopad]
xeros has quit [Ping timeout: 265 seconds]
idella4 has left #linux-sunxi ["Once you know what it is you want to be true, instinct is a very useful device for enabling you to know that it is"]
xeros has joined #linux-sunxi
linkmauve1 has left #linux-sunxi [#linux-sunxi]
sehraf has joined #linux-sunxi
arky has quit [Quit: Leaving]
bsdfox has quit [Ping timeout: 240 seconds]
diego_r has joined #linux-sunxi
mdp has quit [Excess Flood]
bsdfox has joined #linux-sunxi
mdp has joined #linux-sunxi
arky has joined #linux-sunxi
FDCX has quit [Remote host closed the connection]
arky has quit [Quit: Leaving]
FDCX has joined #linux-sunxi
_massi_ has joined #linux-sunxi
bsdfox has quit [Ping timeout: 240 seconds]
penghb has quit [Ping timeout: 240 seconds]
FR^2 has joined #linux-sunxi
penghb has joined #linux-sunxi
<nazcafan> let's say I try out the linux-next kernel on my cubietruck ... there is no way I can interact with it unless I install ssh, right? I see it does not have the video devices on the device tree ... Unless maybe I can use this strange usb-serial adapter? Could this work?
<mnemoc> nazcafan: prefer sunxi-next
<mnemoc> but no display is supported up there yet anyway. libv is working on a kms driver
<nazcafan> mnemoc: sorry, I really meant sunxi-next
<nazcafan> is the usb-serial adapter supported?
<gzamboni> nazcafan, yes, you can use the ttl serial console
<mnemoc> serial console is the first thing people adds :)
<gzamboni> i dont know the current state of the sd/nand, so you will probably need to use nfs or initramfs
notmart has joined #linux-sunxi
notmart has quit [Changing host]
notmart has joined #linux-sunxi
<wens> sd / sata works with sunxi-next
<gzamboni> nazcafan, --^
<wens> hans' usb phy driver hasn't been merged into mainline yet :/
newleaves_ has quit [Read error: Connection reset by peer]
<nazcafan> but already in sunxi-next?
<wens> yup
Faisal has joined #linux-sunxi
<lkcl> memfault: yyyep :)
<mnemoc> nazcafan: sunxi-devel include all work intended for mainline, sunxi-next includes the work *accepted* by mainline
HeHoPMaJIeH has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
geecko has joined #linux-sunxi
<wens> nazcafan: oh crap, I misread it. but sata is definitely accepted.
mkutsevol has joined #linux-sunxi
vector80 has joined #linux-sunxi
<vector80> Hi there, after a long time, I hope you all are fine.
<vector80> My device is an A20 based board ( as you may know), but it incorporates a SANDISK eMMC, with a size of 7456Mbytes
<vector80> I can run both android and Debian on it, I have ready uboot, kernel,modules, rootfs...etc
<vector80> I finished an online update system for android builds, by using OTApackage method.
<vector80> I would like to do similar way for Debian also
<vector80> My debian rootfs is about 2.2Gbyte size,
<vector80> What can you recommend for online updating the whole system?
<vector80> dump and recover ?
<vector80> dd if ?
<vector80> Firstly, I am thinking such partitioning: 32Mbyte for script.bin and uImage, 3Gbyte for debian rootfs, and remaining for userspace apps and for downloading/storing space big image files.
<vector80> I have 1Gbyte DDR
<vector80> I appreciate all your recommendations
deffrag has quit [Ping timeout: 244 seconds]
<oliv3r> vector80: what's wrong with apt-get update?
<oliv3r> set up your own repository :)
<oliv3r> so only approved packages get upgraded
<oliv3r> vector80: if i look at openwrt, they have the entire rootfs in 1 squashfs fs
<oliv3r> and RO use that
<vector80> Hi oliv3r
<vector80> The problem is like this
<vector80> For example, if I would like to update kernel and its modules
<vector80> What would be the best way ?
<vector80> uImage file is in the 1st parttition, but *.ko files are inside the 2nd partition
<vector80> And also, the products should have some kind of "update.sh" running in the backgroud, so they would check my server periodically, and if there is an update available, they would update themselves
deffrag has joined #linux-sunxi
<juanfont> vector80, why not rsync?
<vector80> yes, I also thought about that
deasy has joined #linux-sunxi
<vector80> Looks like rsync is a good idea
<vector80> When I update kernel and modules via rsync, does it cause any issue on the system ?
<vector80> but in order to use rsync, I should tar.gz whole files, right ?
penghb has quit [Ping timeout: 240 seconds]
penghb has joined #linux-sunxi
Faisal has quit [Ping timeout: 240 seconds]
<oliv3r> vector80: quite easy really
<oliv3r> vector80: you are running debian, so by default, you should add (or even replace) the update servers (/etc/apt*) with your own servers. You should host all of the update packages
<oliv3r> this ensures that only tested configurations exist
<oliv3r> and only packages you have approved
<oliv3r> if someone wants to add the default debian stuff, that's fine, but then they are on their own naturally
<oliv3r> as for updating the kernel and its modules
<oliv3r> easy, you create your own kernel package that includes both
<oliv3r> the question basically becomes 'how does debian update the kernel package'
<oliv3r> bootloader is more difficult, you can still do a package + a script (script that writes bootloader via dd) and copies a new script.bin)
<oliv3r> with 'a script' meaning the post sccript inside the package
<oliv3r> with more difficult i mean, if people mess that up, they need a recovery SD card to fix things, something you should offer too
<oliv3r> 'here, download this image, dd it to an SD card and it will automatically pop up a menu that gives you the option to repair the bootloader, repair the kernel + script.bin)
<oliv3r> so rsync I think is kinda of a bad idea :p
<oliv3r> and as far as I know; going from the ubuntu/debian/fedora ways, They always ask you to automatically install security updates, so there should be a way to have apt update itself automatically doesn't it
<oliv3r> and since update updates itself from YOUR repositories, your fien
<oliv3r> you technically become a mirror :p
<oliv3r> but with only approved packages that you approve
<oliv3r> (and remember to host the source/patches for modified packages)
Faisal has joined #linux-sunxi
sehraf has quit [Ping timeout: 245 seconds]
penghb has quit [Ping timeout: 240 seconds]
penghb has joined #linux-sunxi
sehraf has joined #linux-sunxi
cajg has quit [Ping timeout: 264 seconds]
cajg has joined #linux-sunxi
<vector80> oliv3r: Thank you very much for detailed explanation
<vector80> but your explanation incorporates a creating an APT repository
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<vector80> I am thinking of multiple scenarios,
<vector80> Such as like this
<vector80> I provide my products to customer with 32+3G+3.5G partitioning with debian
<vector80> and if the customer wants fedora in future,
<vector80> what should I do ?
<vector80> In moth case, I feel I would need dd :(
<vector80> moth -> both
bertrik has joined #linux-sunxi
penghb has quit [Ping timeout: 265 seconds]
<vector80> And actually, I have just a few executables running in the background, for these products... how to say, they are headless systems
<vector80> They will be used for home automation
<vector80> They are hidden inside people's houses,
<vector80> they are only connected to KNX network, some GPIO relay circuit and internet..
Faisal has quit [Remote host closed the connection]
<vector80> Operators will reach them via SSH if required, but mostly they need to be running / updating by themselves.
<vector80> In my previous job, you know, I was developing Linux HD STBs, and that time, my Flash size was just 32Mbyte and DRAM size was 256Mbytes... it was soo easy to update those partitions via "flashwrite"... But now, eMMC size is 8G, but DRAM is 1G, so I must find a better way
penghb has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
woprr has quit [Ping timeout: 265 seconds]
Net147 has joined #linux-sunxi
Wizzup has quit [Read error: Connection reset by peer]
woprr has joined #linux-sunxi
<vector80> Can I use u-boot to update IMG file over root partition ?
Net147 has quit [Quit: HydraIRC -> http://www.hydrairc.com <- IRC with a difference]
bsdfox has joined #linux-sunxi
_massi_ has quit [Remote host closed the connection]
Wizzup has joined #linux-sunxi
_massi_ has joined #linux-sunxi
<Turl> vector80: you could do something like android does
leviathanch has quit [Remote host closed the connection]
<wens> just hacked up an hwmon driver for axp20x
<wens> trying to figure out why it doesn't work :/
<Turl> wens: :)
<Turl> wens: temp readings?
<wens> Turl: temp, voltage, current :)
<Turl> wens: is it a hw problem or something else?
<wens> though the current reading will only be accurate on the olinuxino boards
<Turl> there was one to read temps on linux-sunxi iirc
<wens> just found the error, stupid logic error
<Turl> if (1==2) do_stuff()
<wens> if (!ret) goto err
<ccaione> wens: cool :)
<nedko> wens: there is already hack for this
<Turl> goto error*
<Turl> :p
<nedko> Turl: apple or gnu style? :)
<Turl> sorry, goto fail*
<Turl> I fail today :p
<wens> nedko: oh really? I must have missed it
<wens> oh I see it, for 3.4
<Turl> -ENOCAFE
<ccaione> -ENOCOCAINE
<nedko> wens: by LABBE Corentin
leviathanch has joined #linux-sunxi
<ccaione> jesus, amlogic sucks shit
<nedko> i was also able to see the voltages
<nedko> but not through lm-sensors
<Turl> ccaione: s/amlogic/sunxi/
<wens> nedko: yeah I know, again that's for 3.4
<ccaione> Turl: I'm working on amlogic at $work$
<nedko> yup
<wens> regmap_bulk_read gives bogus readings :/
<Turl> ccaione: zero chance of replacing those to sunxis?
<ccaione> Turl: I don't think so, I'd like though
<Turl> wens: make sure your buffer is sane
nove has joined #linux-sunxi
<wens> I missed the part that the buffers should be native register size
<wens> I was giving it 3 ints
<wens> still the AXP registers are funny, if they'd put the higher 4 bits in the high register, everything would be easier
<ccaione> wens: what do you mean?
<ccaione> I have no problem using regmap_read/write/update
<wens> ccaione: I'm using bulk_read to get 3 consecutive registers
<wens> I thought the values would be spread out in 3 ints, but it's actually 3 bytes
<ccaione> yep, 8bit access
<rellla> does anybody have experience with deinterlacing? with motion adaptive what is meant by "Current frame tile flag buffer address" https://github.com/linux-sunxi/linux-sunxi/blob/sunxi-3.4/drivers/video/sunxi/disp/de_fe.h#L1097 and "tile flag line-stride" https://github.com/linux-sunxi/linux-sunxi/blob/sunxi-3.4/drivers/video/sunxi/disp/de_fe.h#L1170 ?
<wens> internal: +178.2 C :X
shineworld has joined #linux-sunxi
<Turl> wens: better not touch that board :p
shineworld has left #linux-sunxi [#linux-sunxi]
<Turl> wens: from what I recall from the list, the temps axp gave were not absolute
<wens> yeah I noticed I forgot the offset
<Montjoie> wens, if you need help for the hwmon:)
<wens> I got the basic hack version done
<wens> but it's missing stuff, like fuel gauge, DT, GPIO ADCs
<wens> not to mention the code style isn't good
<nedko> i'm missing something
<nedko> wens: what you are doing more than the existing code has?
<nedko> like the git commit linked by Turl
popolon has joined #linux-sunxi
popolon has joined #linux-sunxi
popolon has quit [Changing host]
<wens> nedko: I'm working on mainline, not 3.4
<wens> anyway I didn't know that patch existed, thought it would be fun to do it
Tsvetan has quit [Quit: Leaving]
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
nedko has quit [Quit: Changing server]
nedko has joined #linux-sunxi
nedko has quit [Changing host]
nedko has joined #linux-sunxi
<focus> wow wow and WOW again!! :)
nedko has quit [Client Quit]
<focus> that must change a lot of things soon
nedko has joined #linux-sunxi
nedko has quit [Changing host]
nedko has joined #linux-sunxi
<WarheadsSE> focus: I don't see it as them "joining" Linaro but joining Linaro in terms of co-running this Digital Home group
<Wizzup> you think?
<focus> WarheadSE: have to admit I don't understand that last bit
<WarheadsSE> They are cooperating with Linaro in an industry group
<WarheadsSE> Not actually _joining_ linaro in an work
<focus> still i bet they begin to get more coperation and software that works with fewer problems?
<focus> last time i checked the dual core A20 gone down to $7
<focus> that is relatively speaking thundering performance for a lot of gadgets waiting to be built
<focus> as far as i can judge everyone is going 8 core
<focus> they be needing even more supported software that works out the box i guess
<WarheadsSE> well, 4x4 core, really
<focus> (after 2 cores, the benefit is very much nil :( )
<focus> (the benefits seem to come from extra graphics cores)
<WarheadsSE> Depends on your work loads
<WarheadsSE> For most _consumer_ uses, sure.
<WarheadsSE> For fools like me that make/play with work-horse class devices, they make sense.
<focus> The $7 A20 makes sense for me now if it be used as embedded CPU to drive LCD and the remaining 300+ pins for IO :)
<focus> high end PICs are about $7!!
<focus> having to add $2 of RAM and $2 of power management pushes it up a little - but considering the alternatives, not too bad
<WarheadsSE> Yeah, for digital signage, 2 cores is enough. Generally the same for phones/tablets/media consumption devices
aaribaud has joined #linux-sunxi
<aaribaud> Bonjour Maxime
<aaribaud> mripard: hi -- and excuse my French
<aaribaud> mripard: asking here in case others that mripard could answer: what is the status of A31 re U-Boot?
<aaribaud> sorry, u-boot+linux
FR^2 has quit [Quit: Connection reset by peer]
jemk has joined #linux-sunxi
vector80 has quit [Remote host closed the connection]
jemk has quit [Remote host closed the connection]
jemk has joined #linux-sunxi
jemk has quit [Remote host closed the connection]
jemk has joined #linux-sunxi
<rellla> or is this documented somewhere?
<jemk> rellla: the physical addresses of the of the image to display
<jemk> rgb only needs addr[0], our yuv addr[0] = luma addr[1] = chroma
<jemk> and yv12 needs all three
<rellla> ok. and same for video?
<jemk> guess so
<rellla> ok, so and last, what is os->vs->plane_size?
rz2k has joined #linux-sunxi
setkeh1 has quit [Ping timeout: 244 seconds]
<jemk> the size of the chroma plane, i should rename that, left over from the first hack version
<jemk> s/chroma/luma/
<rellla> jemk: ok. now it's more clear for me. i only have to find out, what "line-stride" and "flag buffer address" is within motion adaptive deinterlacer
<rellla> and find someone to fix g2d ;)
setkeh1 has joined #linux-sunxi
<jemk> that's some extra data the video engine can generate to help deinterlacer, it looks like it contains some data from the motion vectors of the compressed video
<jemk> but i didn't figure out yet how to properly set things up to generate them
mkutsevol has left #linux-sunxi [#linux-sunxi]
<jemk> at least the blob can do something like this if you set use_maf = 1
<rellla> who wants to use the blob? :p
<jemk> me, but only to trace it ;)
<rellla> we can set maf_valid=1 but then you must set flag_addr, too.
<rellla> the user manual tells us, its "Current frame tile flag buffer address" and "tile flag line-stride"
<jemk> this is set by the blob
HeHoPMaJIeH has quit [Read error: Connection reset by peer]
tomboy64 has quit [Ping timeout: 264 seconds]
uwe_ has quit [Ping timeout: 264 seconds]
uwe_ has joined #linux-sunxi
tomboy64 has joined #linux-sunxi
mkutsevol has joined #linux-sunxi
kuldeepdhaka has joined #linux-sunxi
polavto has joined #linux-sunxi
_massi_ has quit [Remote host closed the connection]
polavto has quit [Remote host closed the connection]
notmart has quit [Excess Flood]
notmart has joined #linux-sunxi
notmart has quit [Changing host]
notmart has joined #linux-sunxi
FreezingAlt has joined #linux-sunxi
jemk has quit [Remote host closed the connection]
diego_r has quit [Read error: No route to host]
diego_r has joined #linux-sunxi
diego_r has quit [Client Quit]
diego_r has joined #linux-sunxi
notmart has quit [Quit: notmart terminated!]
paulk-collins has joined #linux-sunxi
aaribaud has quit [Ping timeout: 240 seconds]
wigyori has quit [Remote host closed the connection]
FreezingAlt is now known as FreezingCold
<oliv3r> vector80: in case you backread (and i haven't yet); if you upgrade the entire backend to fedora; well yeah; that's a massive change that'll require re-installing. 'imaging' is STILL a horrible idea! ask google as to why. That said, if you do a full reinstall like that, doing it via ssh is going to be very hard. You'll have to do it via an SD card. It's relativily easy via an SD card and a script. script will do what you allready did; format, mount, in
<oliv3r> vector80: the only way the imaging thing will work somewhat reliably, is if all units will always be 100% identical, which I tell you know will never happen. Soon 16gb flash will be cheaper then 8gb or it is decided due to supply problems to switch to a nand based solution; your dd thing falls flat in the water
<oliv3r> btw, in case i hadn't mentioned it yet; looks like the temperature sensor in a23 is re-using the temp framework of sun4i-ts; it's at the same address :) supprise
<oliv3r> ccaione: (me backreading) amlogic shit? really? i know lkcl refuses to use it as it's GPL unfriendly and doesn't care (more or less)
<ccaione> oliv3r: dunno about licences (yet) but their BSP il really shitty (code wise)
hawi has joined #linux-sunxi
aaribaud has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
<paulk-collins> I read on the wiki that A31 has u-boot and kernel source, is it new or nothing has changed?
bertrik has quit [Quit: No Ping reply in 180 seconds.]
bertrik has joined #linux-sunxi
bertrik has quit [Changing host]
bertrik has joined #linux-sunxi
geecko has quit [Ping timeout: 265 seconds]
pwhalen has quit [Ping timeout: 264 seconds]
kivutar has joined #linux-sunxi
netlynx has quit [Quit: Leaving]
pwhalen has joined #linux-sunxi
pwhalen has quit [Changing host]
pwhalen has joined #linux-sunxi
sehraf has quit [Ping timeout: 245 seconds]
sehraf has joined #linux-sunxi
deffrag has quit [Ping timeout: 252 seconds]
deffrag has joined #linux-sunxi
<lkcl> oliv3r: it's more that amlogic fucked us. we had a client, we bought samples, we put the samples in front of the client. the client asked for the source code, we asked for the source code.
<lkcl> amlogic refused, and we lost the client.
<lkcl> so it's real simple: don't design products using amlogic processors.
<lkcl> ccaione: i recommend you stop, right now, and find an alternative processor or product.
<ccaione> lkcl: I wish I could. Today I asked for source code, let's see what happens
<ccaione> we are working with geniatech that uses amlogic unfortunately
paulk-collins has quit [Quit: Ex-Chat]
geecko has joined #linux-sunxi
bertrik has quit [Remote host closed the connection]
kuldeepdhaka has quit [Quit: gn]
zumbi has quit [Ping timeout: 264 seconds]
hawi has quit [Ping timeout: 252 seconds]
rz2k has quit []
ganbold_ has quit [Remote host closed the connection]
deasy has joined #linux-sunxi
nove has quit [Quit: nove]
<Turl> aaribaud: the community uboot does not have A31 support so far
<Turl> aaribaud: maxime sent some patches for some basics, but last I heard it still didn0't support SPL building
<Turl> so you still need to use AW bootloaders and then chainload uboot
kivutar has quit [Quit: Ex-Chat]
<Turl> aaribaud: as for mainline linux, there is some basic A31 support on there
Wizzup has quit [Quit: Change me]
geecko has quit [Read error: Operation timed out]
diego_r has quit [Ping timeout: 240 seconds]
Wizzup has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
Black_Horseman has joined #linux-sunxi