<wingrime>
oliv3r: MACC_VE_MMCREQ_WNUM ..... cedarx can read directly from MMC????
<wingrime>
or ???
FR^2 has joined #linux-sunxi
awafaa has joined #linux-sunxi
jemk has joined #linux-sunxi
jeremb_ has joined #linux-sunxi
notmart has joined #linux-sunxi
notmart has quit [Changing host]
notmart has joined #linux-sunxi
<oliv3r>
wingrime: no clue
<oliv3r>
wingrime: i wouldn't be supprised, mmc has it's own dma does it not? streaming directly from mmc might make things faster/lower CPU demanding?
<Turl_>
you know you could've sold the iphone and bought like over 9000 batteries right? :p
Turl_ is now known as Turl
* mnemoc
will try to avoid making a joke about Turl showing his "tail"
<oliv3r>
Turl: a broken iphone without a screen or even a PCB?
<oliv3r>
Turl: i highly doubt it
<oliv3r>
the 3gs is the old (3+ year) company phone, we had a few in the cabinate for spare parts
<oliv3r>
well it was just spare parts now :)
<oliv3r>
but you think i could get 40.000 for a 16GB used iPHone 3gs?
<oliv3r>
i have my doubts ;)
<Turl>
:p
<Turl>
probably 100$ though
<Turl>
wait, no pcb?
<Turl>
that's not an iphone lol
<wingrime>
oliv3r: can you add default y/m to sata?
eebrah has quit [Ping timeout: 265 seconds]
<oliv3r>
Turl: why do you think i used the battery ;)
<oliv3r>
wingrime: sata is broken :p
<oliv3r>
i have to debug it
<Turl>
A20 SOM from Olimex has GbE \o/
<Turl>
and a 2Mp camera :o
rellla2 has joined #linux-sunxi
<wingrime>
Turl: but the same....
<wingrime>
Tsvetan: have you any ideas about camera expansion modules?
<wingrime>
err expandion ?
<wingrime>
not sure I writed it correctly
<Turl>
expansion is correct
jinzo has joined #linux-sunxi
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
<wingrime>
Turl: thanks
<wingrime>
oliv3r: PM on sun7i even not builds
<oliv3r>
wingrime: i noticed; i disabled it for now
cubear has quit [Quit: Leaving]
tinti has joined #linux-sunxi
tinti has quit [Ping timeout: 248 seconds]
<focus>
Is there big differences with cubieboard1 and 2? I thought only 2nd CPU is difference. The script.fex I use same, the kernel i compile for sunxi4, but machine halts on 2nd line with "DRAM:" as the only printed word
<focus>
assume DRAM parameters are wrong?
<focus>
script.fex came from working cubieboard2
<oliv3r>
focus: '2nd cpu'? it's a whole new SoC
<mnemoc>
dram is not configured from .fex
<oliv3r>
but yes, most of the IP is the smae
<rm>
it's not sun4i
<oliv3r>
the IRQ controller is completly different though
<focus>
uuugghhhh... so i need a working script.fex from a cubie1 image i assume
<focus>
if fex isn't where dram is configured, where is it set up?
<mnemoc>
u-boot's spl
<mnemoc>
which doesn't read the .fex
<mnemoc>
`fexc` can dump a myboard.c file for u-boot from the .fex
<mnemoc>
but the file is not read in boot time
<focus>
hmm.. sorry mnemoc having a mental block - it says in fex guide: SD-Ram is usually configured via livesuit when flashing. Livesuit probes the hardware or knows about the hardware and its configuration and configures the SoC accordingly. This luxury is not available from Linux and thus sdram parameters have to be set up by hand.
eebrah|noPC has joined #linux-sunxi
wingrime has quit [Ping timeout: 260 seconds]
<mnemoc>
focus: livesuit injects the data in nand's boot0/boot1
<focus>
ok i got to read up "uboot spl" and configure something there for dram
<oliv3r>
focus: livesuit reads the memory timings or calculates them, then programs them into boot0/boot1
<focus>
ok
<oliv3r>
we extract those parameters from the registers (a10-meminfo) and code them into the SPL
<mnemoc>
uboot spl is used when booting from uSD
<focus>
ok
ykchavan has joined #linux-sunxi
<oliv3r>
the fex is just updated/corrected so that it's properly in git sunxi-boards
<focus>
so the actual parameters are not something done by human hands - it has to be worked out by livesuit and then we grab it from some data structure and put it into uboot spl
<focus>
sheesh!
<oliv3r>
basically yes
eebrah|noPC has quit [Ping timeout: 252 seconds]
<focus>
once the parameters are worked out, i assume it does not change, unless the board changes and track lengths for dram changes?
<oliv3r>
how/why livesuit obtains memory timings, is unknown
<oliv3r>
basically, yes
<oliv3r>
or if you want to overclock it ;)
<mnemoc>
some details can be hand tuned, but most are a mistery
<focus>
i want to experiment having different lengths of line and compensate for line lengths using capacitors in 0.1pf steps - it now means i need livesuit to work out the parameters and then grabbing them and then putting them into "uboot spl" and then booting linux before it can be known what the effects were
<focus>
long winded!!
<focus>
i'm designing new boards and prefer to use the A20. kicad can be persuaded to make tracks of a fixed size using the microwave track building option - but in places if you want dense packing, the lengths have to be slightly different - hence idea of using capacitors to slow down shorter tracks
<mnemoc>
btw, your som seems perfect to test if we really can have dual ethernet on A20
<focus>
The alternative way of checking is to put the memory chips intentionally further away and then use these capacitors to even out line lengths
<oliv3r>
focus: just boot your kernel directly from the livesuit injected parameters?
<oliv3r>
have a rootfs with your test apps in it; write that rootfs using livesuit
<focus>
actually it mentions ethernet as GMAC and EMAC ? - does that mean it has 2 ethernets?
kaspter has quit [Ping timeout: 260 seconds]
<focus>
mnemoc: yes perfect for prototyping work to bring out the extra internal hardware if other boards haven't done it yet
<focus>
mnemoc: i help luke as much as i can - nearly finished i believe - so get on wit SoMs soon
<mnemoc>
cool
<Turl>
sunxi-devel updated :)
gzamboni has quit [Read error: Connection reset by peer]
<focus>
mnemoc: first task is to make a simple board with AXP209, then another with DRAM, and then another with ethernet+hdmi+usb
rellla2 has quit [Remote host closed the connection]
<focus>
and then allow the som board to be wired manually to these boards
<mnemoc>
the most advanced thing I can do in EE is to connect wire jumpers :p
<mnemoc>
but sounds promising :)
<focus>
mnemoc: i also notice an opportunity to build compact som board - many cn pcb makers are offering 0.1mm tracks as standard - it means you can wire 2 wires between the ball pads of bga, which means the board can be very compact - so may be i remake the som boards with these thinner tracks and leave it as a module that others copy and paste to into their product designs
<focus>
mnemoc : its a start! :)
<focus>
jumpers
<focus>
oliv3r: i have to find a way of going there - will ping you when i get there
forcev is now known as FunkyPenguin
FunkyPenguin has quit [Changing host]
FunkyPenguin has joined #linux-sunxi
Soru has quit [Ping timeout: 268 seconds]
Gerwin_J has joined #linux-sunxi
gzamboni has joined #linux-sunxi
Tsvetan2 has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
Tsvetan has quit [Ping timeout: 268 seconds]
<oliv3r>
Tsvetan2: next b oard, bring out the battery temperature sensor from the AXP and connect some leds to the charge led outputs of the axp ;)
\\Mr_C\\ has quit []
rz2k has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
wingrime has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
Soru has joined #linux-sunxi
Guest69871 has joined #linux-sunxi
<Guest69871>
Howdy folks... anyone around who might be able to give me some tips regarding the sunxi-codec alsa module?
<oliv3r>
Guest69871: ask your question; dn't ask to ask
<Guest69871>
Righto... I'm trying to access the stereo line-in ports on an Olimex A20 board via the GPIO header, but following the wiki page it says to set ADC Input Mux to line-in, however my sunxi-codec driver doesn't appear to have the ADC Input Mux as a control
<Guest69871>
so I'm stuck with the on-board mono microphone input via the 3.5mm socket
<Guest69871>
running 3.4.61+, compiled yesterday from the git sources
<oliv3r>
ok, and you went over all FEX settings?
<oliv3r>
doesn't the olimex have a header for the second microphone connection?
<Guest69871>
is the output of my amixer controls, etc.
<oliv3r>
i'm missing the mux control
<Guest69871>
page 185 of the A20 user manual describes the register, which seems to vaguely line up with what's in sunxi-codec.c but thats starting to get out of my range of technical comfort zone...
rellla2 has quit [Remote host closed the connection]
<WarheadsSE>
if you are building it from a native toolchain on an arm box, try just taking off the full chain name
bamvor_ has joined #linux-sunxi
<WarheadsSE>
CROSS_COMPILE=
bamvor has quit [Ping timeout: 268 seconds]
bamvor_ has quit [Ping timeout: 248 seconds]
bamvor_ has joined #linux-sunxi
panda84kde has joined #linux-sunxi
eebrah_ is now known as eebrah
eebrah is now known as Guest82503
Guest82503 is now known as eebrah_
eebrah_ is now known as eebrah|FTS
HeHoPMaJIeH has quit [Remote host closed the connection]
leowt has quit [Quit: leowt]
leowt has joined #linux-sunxi
n01 has quit [Quit: leaving]
boss has joined #linux-sunxi
mcbrick has joined #linux-sunxi
<steev>
oliv3r: what do you do your development with?
<steev>
wondering what's mangling the patch you sent about ahci
Guest69871 has left #linux-sunxi [#linux-sunxi]
_enrico_ has quit [Quit: Bye]
atiti has joined #linux-sunxi
boss has quit [Quit: Copywight 2007 Elmer Fudd. All wights wesewved.]
FR^2 has quit [Quit: Connection reset by peer]
notmart has quit [Quit: notmart terminated!]
<Turl>
steev: the patch is attached, how can it be mangled?
<steev>
Turl: i got dos line endings on every line
panda84kde has quit [Quit: Konversation terminated!]
<Turl>
steev: I downloaded it and it's got unix ones here
<steev>
odd
<steev>
guess it's gmail
<Turl>
I also use gmail
<Turl>
but through thunderbird
<steev>
+^M$
<steev>
definitely see dos line endings from the gmail web interface
<steev>
i really need to get around to re-setting up mutt
<Turl>
steev: yep, downloading from Gmail web interface gets the file with DOS line endings
<Turl>
steev: luckily it's easy to fix it
<steev>
yep
<steev>
dos2unix and then it's just a few whitespace lines
* Turl
resaves the file with nano
<oliv3r>
steev: how is it mangled; i used vim ;) but I did attach the file via thunderbird; so that may have mangled it
<oliv3r>
and since we use google-groups; i blame that.
<steev>
oliv3r: nah, looks like it's just google being shit
<oliv3r>
i have to make myself setup postfix soon
wingrime has joined #linux-sunxi
<oliv3r>
boog oogle
<oliv3r>
i will TRY to make myself start with postfix tomorrow
<oliv3r>
my gf got a tablet; i thought it would be a10 based; turns out, it's a13 based. so I finally have some a13 hardware to test things with :D
massi_ has quit [Remote host closed the connection]
<oliv3r>
atsampson: how's your replicant/CM stuff coming along :)
jinzo has joined #linux-sunxi
<Turl>
oliv3r: breaking gf's tablet is not a smart move :p
<Turl>
oliv3r: it was the gmail client that downloads the file with dos line endings, the email itself is fine
<oliv3r>
pff, boot from SD
<oliv3r>
good, go on and troubleshoot ya'll ;)
<oliv3r>
omg this is so nasty: CODEC_SINGLE("Mic Input Mux",SUNXI_DAC_ACTL,9,15,0),//from bit 9 to bit 12.Mic（麦克风）输入
<oliv3r>
so bit 9-15 control the mic1, mic2, left and right input sources as mute/not-muted
<oliv3r>
so they make it a slider, with 16 values
<oliv3r>
i need to learn about alsa and fix that shit properly
<oliv3r>
on the upside, i think i FINALLY have utf-8 working properly everywhere
<oliv3r>
as i actually copy/pasted chinese symbols there
<Turl>
yep I see chinese symbols
<mnemoc>
re
<Turl>
wb mnemoc
<mnemoc>
thanks Turl :)
<Turl>
oliv3r: chinese says (Microphone) input
<mnemoc>
Turl: what's the LOADADDR= thing we have to use? I want to setup nightly builds of each defconfig of each branch
<Turl>
mnemoc: the load address thing that gets passed to mkimage
<oliv3r>
Turl: yeah, it's mic1 Left source, mic1 right source, mic2 left source, mic2 right source
<oliv3r>
you use that 'slider' to choose which configuration you want
<Turl>
mnemoc: LOADADDR=0x40008000 for all sunxis I know of
<oliv3r>
so level '0' is mute all mics
<mnemoc>
Turl: cool. thanks
<oliv3r>
level 15 is enable all mics; it's kinda silly :)
<Turl>
np
<Turl>
:p
<oliv3r>
and i just did a whole lot of renaming and cleaning; just to realize, we want to have 3.4 code to be as close to the original as possible, to make diffing easier
<oliv3r>
:(
<mnemoc>
Turl: and your rootfs in 3.10?
<oliv3r>
the rootfs works on 3.4 too
<mnemoc>
embedded initramfs I mean
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
<Turl>
do you want to embed an initramfs on all nightly builds? :)
<hramrach>
hello
<hramrach>
how is the thing with CT schematic going?
<mnemoc>
Turl: if you can document the process of "sideloading" the initramfs you can save us a lot of space :p
<Turl>
mnemoc: I was never able to get "uInitramfs" working, but I think someone else here did
<mnemoc>
Turl: I'm building defconfigs, and sunxi_deffconfig wants your rootfs
<Turl>
mnemoc: there's a copy of it on the server btw
<Turl>
I'll see if I can get it uInitramfs'ed
<mnemoc>
I'll try doing an "smart" nightly, only pushing if something actually changed
<Turl>
+1
<Turl>
mnemoc: are you hand coding it all, or using some CI system?
<Turl>
(eg Jenkins)
<mnemoc>
shell scripting
<mnemoc>
+ cron
<oliv3r>
why WHY in the world is in the A20 user manual, the Microphone volume register blanked :S
<Turl>
oliv3r: patents >:D
<Turl>
j/k
<oliv3r>
what's the point? i'm not allowed to know that bits 23-26 are volume left/right?
<mnemoc>
oliv3r: your are allowed to discover it. they are not allowed to tell you
zetaneta has joined #linux-sunxi
<steev>
oliv3r: you expect it to be documented properly?
<Turl>
bbl
leowt has quit [Quit: leowt]
<atsampson>
oliv3r: start of term intervened (I work at a university) ;-)
<atsampson>
the basic stuff all appears to work; the last thing I tried was getting the camera to do something, though (it doesn't work on linux-sunxi either at the moment)
<oliv3r>
mnemoc: luckly, it's right there in the code :p
<oliv3r>
steev: it was purposly removed :p i wonder why
rellla has joined #linux-sunxi
<oliv3r>
steev: the only reason I can imagine, is because they added 2 new registers on sun7i that controls the mic in details
<oliv3r>
or more specifically
<oliv3r>
atsampson: i think the camera is tied somehow together with cedarX or requires it to do realtie encodings
<oliv3r>
atsampson: but a build without cedarX and if it's absolutly required without camera, is still very good
zetaneta has quit [Ping timeout: 248 seconds]
<oliv3r>
there, little patch sent; test and merge mnemoc ;)
<oliv3r>
CODEC_SINGLE("Mic Input Mux",SUNXI_DAC_ACTL,9,15,0),//from bit 9 to bit 12.Mic（麦克风）输入
<oliv3r>
whoops
<oliv3r>
for the first time in 20 years of me using irssi i see utf-8
<oliv3r>
so simple yet so ... joyfull
<oliv3r>
well by far not 20 lol
* mnemoc
has logs from 1993 :p
<mnemoc>
efnet...
<oliv3r>
oh wow
<oliv3r>
only publically logged stuff
<oliv3r>
form #linux probably
<oliv3r>
one name pops into my mind; nunya i think
<mnemoc>
nah, ircII logs
fredy has quit [Excess Flood]
<oliv3r>
BitchX
<mnemoc>
my own
<mnemoc>
yup :p
<mnemoc>
oh, bitchx even release 1.2 this year!
fredy has joined #linux-sunxi
<steev>
do we need to leave the chinese comments in?
<oliv3r>
yep
<oliv3r>
or translate them
<oliv3r>
steev: but we wanna change actual code as much as possible
<oliv3r>
little*
<oliv3r>
so merging is easier later
<steev>
i guess that makes sense
<oliv3r>
takes a lot of effort to do so :)
<oliv3r>
i wanna change stuff :)
wingrime has quit [Ping timeout: 264 seconds]
wingrime has joined #linux-sunxi
<steev>
stupid question but
<steev>
is there an interface BACK to allwinner? so that things that are fixed get back into their internal work?
jinzo has quit [Ping timeout: 265 seconds]
focus_it has joined #linux-sunxi
<oliv3r>
steev: nein
<oliv3r>
steev: allwinner does not care
<oliv3r>
or atleast that's our impression
<steev>
has anyone asked? heh
<mnemoc>
oliv3r: do you have that sound-codec commit in a public branch from where to cherry-pick?
<steev>
freescale was like that for a very long time
<mnemoc>
git am and mails with utf-8 text don't get along well
eebrah|FTS has quit [Ping timeout: 265 seconds]
<oliv3r>
mnemoc: not yet; but can do
<oliv3r>
sun7i_codec
<oliv3r>
250b0f044ae9aff2f4b475a2f12a45225ec9cc6d
atiti has quit [Ping timeout: 248 seconds]
<oliv3r>
but i'm sure some would like to review it
<steev>
i stop when there is unicode characters
<steev>
and incorrect commenting
<steev>
i get not wanting to stray too far from allwinner's shit code, but if you're not gonna do it right, why bother at all?
<steev>
if llvm and clang would hurry up and finish i'd actually look at the sata stuff at least
<oliv3r>
steev: i'd love to clean up stuff; make it all nice and well
<oliv3r>
but the whole 'drop -> our code merge' is where it makes sense to keep it as close to AW as it is
<oliv3r>
unless we repalce it
<steev>
oliv3r: do it in a different branch and only pull in the changes that are necessary? how often do "we" actually get code drops from aw?
<oliv3r>
occasionally
<oliv3r>
but 3.4 should be 'maintanance and bug fixing only'
akaizen has quit [Remote host closed the connection]
<oliv3r>
3.10 we wanna do 'our own thing'
<oliv3r>
and mainline is for new clean drivers
<steev>
is 3.10 coming from them too?
<oliv3r>
nope
<oliv3r>
from us
<steev>
but 3.10 because it's an LTS?
<oliv3r>
aye
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
Black_Horseman has joined #linux-sunxi
<oliv3r>
gotta start somewhere
<oliv3r>
3.0 and 3.4 where LTS, so 3.10 makes sense in that regard
akaizen has joined #linux-sunxi
akaizen has quit [Remote host closed the connection]
<wingrime>
oliv3r: are you sure aw removed pata from a20?
<wingrime>
4
wingrime has quit [Quit: leaving]
aalm1 is now known as aalm
wingrime has joined #linux-sunxi
<wingrime>
oliv3r: I found *some* headers for sunii
<wingrime>
oliv3r: aka f10,f15
<oliv3r>
wingrime: very sure, they removed the mux and put emac in its place
ykchavan has quit [Read error: No route to host]
<wingrime>
oliv3r: f***k still no answer from benn
<oliv3r>
vacation :p
ykchavan has joined #linux-sunxi
<oliv3r>
chinese holiday
<wingrime>
oliv3r: oh,hell
<wingrime>
oliv3r: can be emac and wemac work together?
<wingrime>
oliv3r: ce be it muxed to different pins
<wingrime>
*can
<wingrime>
oliv3r: if thats possible, we will have very nice router
ykchavan has quit [Client Quit]
<oliv3r>
wingrime: emac and gmac in a20
<wingrime>
oliv3r: yes, but can it be muxed to different pins?
<wingrime>
oliv3r: I want 2 ethernet cards on same board?
<wingrime>
(!)
<oliv3r>
yes
<oliv3r>
should be possible
<oliv3r>
emac on PH (previous ATA pins) and gmac on PA
paulk-collins has quit [Quit: Ex-Chat]
<wingrime>
Tsvetan2: ^^^^^^^
<wingrime>
Tsvetan2: why do you still not plan/develop board with two ethernets??
<wingrime>
Tsvetan2: thats awesome, I saw many comments about people want it....
<oliv3r>
wingrime: he said a20 SOM can do it
<oliv3r>
Tsvetan2: also did gbit test; 330 mbit with current kernel
<wingrime>
oliv3r: cool
<wingrime>
oliv3r: badly
<wingrime>
oliv3r: but, realy, som20 big....
<wingrime>
oliv3r: something small ... better
<oliv3r>
yeah well ... yeah :)
<wingrime>
oliv3r: a20-som realy big
<wingrime>
oliv3r: camera too short and stange connector
popolon has quit [Remote host closed the connection]
<focus_it>
i downloaded a cubuntu didstro for uSD card - it has two partitions, the first partition contains uEnv.txt - it doesn't have the boot.cmd / boot.scr in there
<focus_it>
so i assume it boots of uEnv.txt - set the drive in there as /dev/sda, it picks it up but after a lot of operations, it just hangs with waiting for /dev/sda1
ganbold_ has quit [Remote host closed the connection]
<focus_it>
strange because cubieboad2 boots from boot.cmd with root=/dev/sda1 fine
ss__ has quit [Ping timeout: 250 seconds]
<focus_it>
first time i'm trying to transfer boot to self powered external sata for cubieboard1 (works ok with cubieboard2)
rellla has quit [Remote host closed the connection]
Black_Horseman has quit [Quit: Zwi se logou mou!!!]