<libv>
Turl: any reason why you feel that it should be excluded from the sunxi party?
<libv>
dl9pf actually has been busy packaging sunxi-mali for suse for us
<Turl>
libv: nope, none at all
<Turl>
it's just an uncommon choice of distro
<libv>
there was a time when things were different.
<libv>
but redhat competition shitthrowing, the bank called novell stupidity, and canonical making a distribution for grandmothers, that all means that suse is not too popular anymore
<drachensun>
I've run suse on A10
<drachensun>
I like it, XFCe runs faster for some reason
<drachensun>
however when I revisited my suse image after a few months it had stopped working with the current kernel
<drachensun>
just a black screen but I can see on the debug port it boots up
<Turl>
mripard: http://sprunge.us/gNiO I probably won't be able to test it until monday, but it should be ok
egbert has quit [Disconnected by services]
egbert_ has joined #linux-sunxi
Dave77 has joined #linux-sunxi
Dave77 has quit []
drachensun has quit [Ping timeout: 252 seconds]
drachensun has joined #linux-sunxi
calris has quit [Ping timeout: 258 seconds]
ganbold__ has quit [Read error: Connection reset by peer]
ganbold__ has joined #linux-sunxi
calris has joined #linux-sunxi
theOzzieRat has quit [Ping timeout: 252 seconds]
theOzzieRat has joined #linux-sunxi
ZaEarl has quit [Ping timeout: 264 seconds]
hramrach has quit [Ping timeout: 276 seconds]
tavish has joined #linux-sunxi
sky770 has joined #linux-sunxi
ZaEarl has joined #linux-sunxi
torqu3e has quit [Read error: Connection reset by peer]
mdfe has joined #linux-sunxi
ZaEarl has quit [Ping timeout: 264 seconds]
Zenton has joined #linux-sunxi
torqu3e has joined #linux-sunxi
wingrime has joined #linux-sunxi
plan_b has joined #linux-sunxi
shineworld has joined #linux-sunxi
<shineworld>
morning
<shineworld>
what mean achronim SPL in u-boot git ?
<wingrime>
uho: why not make SPL retry when MMC fails not simply wait reset, I mean with timeout
<shineworld>
SPL stay for ?
<wingrime>
shineworld if it fails read mmc sector it just want you for reset
tavish has quit [Read error: Connection reset by peer]
<shineworld>
uhm, I'm reading that: Boards with (SPL) include SPL loader to boot directly from SD-card.
<shineworld>
so I only want do understand what is SPL (static program loader ? some program layer ? ... )
<wingrime>
yep
<wingrime>
BROM: read sectors with SPL
<wingrime>
but SPL, can't do it sometimes
<shineworld>
Ok I will search fro BROM doc :I)
<shineworld>
:)
<oliv3r>
shineworld not much known is about the BROM, except that it's hardcoded (ROM) into the SoC itself. The boot order in witch it tries to boot is documented on the wiki; and hno has binary (dissasembled?) dumps in his github
tavish has joined #linux-sunxi
<wingrime>
I have press powerbutton long 5-6 times for boot
<libv>
before, it never bothered me, it apparently automatically did the right thing
<libv>
but now it does get in the way, and the boot.scr apparently gets ignored
ZaEarl has joined #linux-sunxi
shineworld has joined #linux-sunxi
<libv>
ah, ok, so this is not related
bfree has quit [Ping timeout: 252 seconds]
shineworld has quit [Remote host closed the connection]
<mnemoc>
libv: that warning only means that there are no saved env, and that the default will be used
<mnemoc>
s/are/is/
rz2k has joined #linux-sunxi
bfree has joined #linux-sunxi
bfree has quit [Remote host closed the connection]
bfree has joined #linux-sunxi
<Turl>
shineworld: second program loader iirc
torqu3e has quit [Read error: Connection reset by peer]
bfree has quit [Ping timeout: 252 seconds]
torqu3e has joined #linux-sunxi
ganbold___ has joined #linux-sunxi
ganbold__ has quit [Remote host closed the connection]
shineworld has joined #linux-sunxi
bfree has joined #linux-sunxi
ganbold__ has joined #linux-sunxi
ganbold___ has quit [Remote host closed the connection]
ganbold__ has quit [Remote host closed the connection]
ganbold___ has joined #linux-sunxi
ganbold___ has quit [Remote host closed the connection]
ganbold__ has joined #linux-sunxi
torqu3e has quit [Quit: torqu3e]
torqu3e has joined #linux-sunxi
<ganbold__>
mnemoc: there is still no info on CAN, right?
<mnemoc>
not that I'm aware of
<shineworld>
CAN like CAN BUS ?
<ganbold__>
yes
<xenoxaos>
i CAN has?
<mnemoc>
but be aware CAN_TX and CAN_RX conflict with SATA
<xenoxaos>
:P
<shineworld>
not all use sata... can bus is interesting
<mnemoc>
fully agree
<mnemoc>
I only mean that if you plan for CAN, discard SATA
<shineworld>
I've miss that feature and found now it googling CAN BUS Allwinner A10
<ganbold__>
and does IR work in linux?
<mnemoc>
yes
<ganbold__>
cool
<ganbold__>
those are 2 interesting thing to try :)
<ganbold__>
and hdmi cec work? that is other interesting thing too :)
<mnemoc>
kernel side, yes. userspace.... still fuzzy
<ganbold__>
nice :)
<shineworld>
incredible A10 data-sheet don't say nothing about CAN BUS so is what I never listed that before today
<shineworld>
have you a better chip doc than draft in linux-sunxi ?
<mnemoc>
the "user manual" in dl.linux-sunxi.org/A10 it's the most extensive .pdf we have
<mnemoc>
but no, as it was implicit in the initial question from ganbold__, we don't have documentation about the CAN controller beside the pins used and base address of the register
<shineworld>
it is a pity....
<shineworld>
I'm aware that to use a chip features just a draft is useless... Usually my SuperH or freescale chips goes with manuals very heavy, till 1200 pages sometimes
<mnemoc>
that won't EVER happen on the chinese world
<mnemoc>
the datasheet of the A31 is even shorter than the A10's and A13's
<mnemoc>
just 40p
<mnemoc>
for them the way to make product design and manufacturing easy is not by providing documentation, but by providing reference designs and ready to use (and so GPL violating) SDKs
<ganbold__>
and make even one kernel work for all devices (using fex to change configs)
<ganbold__>
and every tablet/phone makers just copy
<shineworld>
strange policy
<mnemoc>
"fast to market" is the policy
<mnemoc>
and understanding the chip is a waste of time for that
<shineworld>
so you don't release thousand pages about internal registers to configure the CPU but you permit a basic use of them with .bin .fex technology
<shineworld>
a sort of HAL
<mnemoc>
yes
<shineworld>
what I haven't yet understood is "who read script.bin" and configure the CPU registers
<mnemoc>
each driver reads it's section from the .bin on init
<shineworld>
so is a kernel driver (with sources available) that read script.bin and configure the CPU for board ?
<mnemoc>
yes
<mnemoc>
or fail to load if the feature is disabled
<shineworld>
interesting I will go to read them
<shineworld>
I've saw there is also a disassembled version of brom in net
<mnemoc>
it's very practical and effective for their aim
<mnemoc>
shineworld: in hno's github there is a repo called allwinner-info
<shineworld>
yeas, what I mean ;
<shineworld>
I'm reading it too
<mnemoc>
the brom only pokes around from where to boot (mmc, nand, usb), and implement FEL mode
<shineworld>
I'm a lot pissed with question of boot because I wasn't able to run somethings ... some already images works, some others not... so I'm trying to understand the question
<mnemoc>
"images" include the dram initialization part
<mnemoc>
and that is board specific
<shineworld>
I've changed them but don't works
<mnemoc>
so a failure by definition
<shineworld>
yes... or I badly create the image... I don't know I need to understand the entire phase to guess what I mistake
<mnemoc>
shineworld: start small. not from a fully featured "image"
shineworld has quit [Remote host closed the connection]
<ganbold__>
very cool
<mnemoc>
they made a very nice work exponsing all the features possible, and not killing anything by picking the wrong gpio
ganbold___ has joined #linux-sunxi
<mnemoc>
your nose keeps growing...
ganbold__ has quit [Remote host closed the connection]
ganbold___ is now known as ganbold__
paulk_ has joined #linux-sunxi
paulk_ has quit [Read error: Connection reset by peer]
merbzt has joined #linux-sunxi
rellla has quit [Ping timeout: 264 seconds]
uro has quit [Ping timeout: 246 seconds]
uro has joined #linux-sunxi
<oliv3r>
nose keeps growing? pinokio? hwo's lieing?
<oliv3r>
hramrach: i'm trying to follow your debian guide, but i don't understand 2 steps, first, you install binfmt support and then you copy a x86-64 statically compiled qemu into the pure arm chroot. but when you actually do the chroot, you don't call qemu. I don't get that. why not use qemu from outside?
<oliv3r>
now if you would do something like qemu ./arm-deb/bin/bash or something ( noidea) then i'd get that and understand
<oliv3r>
but right now, it just says 'chroot arm-deb/bin/bash' which shoudln't work afaik
torqu3e has quit [Ping timeout: 245 seconds]
<ssvb_>
oliv3r: binfmt eliminates the need to call qemu directly
<Turl>
"Mer just provides the equivalent of the MeeGo core. The former Meego user interfaces and hardware adaptation are to be done by various other projects and by hardware manufacturers, which will be able to build their products on top of the Mer core."
<Turl>
sounds like yet another yocto/openembedded/buildroot
<oliv3r>
ssvb_: I was hoping that ROOT=/tmp/qemu-static/ USE="static" emerge qemu; but that failed misserably :S
<ssvb_>
oliv3r: it was last updated on May 11, 2012 so may be slightly out of date, but still explains the general idea :) And if qemu ebuild has problems with static USE flag, then it's likely a bug in qemu ebuild
bsdfox has joined #linux-sunxi
<oliv3r>
ssvb_: I still have to read it, but I run qemu-dynamically (with active VM's) on the box too, so i could, install static one, copy it, reinstall dynamic one I guess :)
<ssvb_>
oliv3r: sure, you just need a single static qemu-arm binary compiled for x86, it does not matter how exactly you get one
<libv>
Turl: it in any case is supposed to be a ready-to-use for-tablets distribution
<libv>
as far as i understood mdfe we could have automatic weekly builds for all devices supported by us, built by obs, and thrown straight to linux-sunxi.org
<oliv3r>
ssvb_: oh! a packaged called qemu-user! that is new :)
<oliv3r>
ssvb_: hmm, it wants to rebuild glib and zlib with the static uselfag set
<oliv3r>
so i don't know if that useflag will build both the regular dynamic AND the static one besdies it (I would expect so)
<ssvb_>
oliv3r: btw, there is #gentoo-embedded channel here, which seems to be a bit more appropriate for this stuff
<oliv3r>
ssvb_: true true :)
<oliv3r>
i did actually post it on the gentoo-forums; i'll add my comments there
<libv>
mdfe: patch pushed
<mdfe>
:)
anunnaki has quit [Ping timeout: 245 seconds]
<oliv3r>
ssvb_: crap, i think i'm being bit by gentoo #454984 :( i'll rip a qemu binary from somewhere else in the meantime :)
<oliv3r>
(comment #2 is my own)
<hramrach>
oliv3r: binfmt support calls qemu automagically so you don't need to
<hramrach>
oliv3r: if you want to use the guide for gentoo there are severeal links to detailed chroot instructions, too
<hramrach>
which explain that bit
<oliv3r>
hramrach: I wrote the gentoo guide on linux-sunxi :p
<oliv3r>
anyway, i think qemu-user build may be broken atm; thank you turl, i'll get that and rip out qemu for now
<oliv3r>
hansg on your fedora image, in the instructions, you speak of first connecting to HDMI or DVI, I have a tablet which is connected to LCD by default (i guess I could connect it to hdmi via the mini-hdmi dongle If i get a cable for it). why is that and will it not work with lcd?
shineworld has quit [Quit: Leaving]
<hramrach>
oliv3r: if you have the .bin file for your tablet it will work with the lcd
slapin has joined #linux-sunxi
<slapin>
hi, all
<slapin>
so small amount of people here comparing to arm-netbook, community fragmentation it seems...
paulk-desktop has quit [Quit: Ex-Chat]
<slapin>
do anybody have allwinner a13 kicad symbol? eagle will do, too
<techn_>
slapin: did you notice NFC driver on linux-sunxi ML
<slapin>
this is always so with OpenSource - if this needs to be done, it will be done with whatever persone comes with solution first regardless on how much effort you really put in this, but it is good for community as things evolve and are gitting done regardless of someone's personal interests.
* slapin
is jealous
<Turl>
slapin: it's not fully functional though
<Turl>
slapin: OOB stuff is broken because DMA autocalculates it or sth
<techn_>
seems that a20 tablets are sub 100$ and tv-boxes sub 50$ :/
<slapin>
Turl: OOB can be made working, I spent enought time with this. Not DMA - controller itself, ECC is not mandatory, and DMA is not mandatory either.
<Turl>
techn_: I saw a zatab A20 linked from the wiki
<slapin>
I just need to recover my memory about this stuff
<libv>
oh wow, 2GB DDR built in
<libv>
but still 800x480 :)
torqu3e has quit [Quit: torqu3e]
<slapin>
libv: 3GB DDR on A20?!
<slapin>
isn't it limited processor?
<slapin>
too bad I haven't got any of olimex boards where I could solder proper chips for experiments :(