<mrnuke>
I must have mixed up fex with "more stuff than mainline" somehow
<Turl>
mrnuke: mnemoc is working on a dt/fex hybrid too but that's 3.10
<Turl>
anyway :)
<mrnuke>
don't care about fex TBO. I just want a working dt config
<mrnuke>
(so I can test coreboot :) )
<Turl>
coreboot can boot dt directly? :o
<mrnuke>
via uboot
<Turl>
then you can boot fex too? :)
<mrnuke>
theoretically yes, but fex==FUD in my mind
<Turl>
dunno how can fex==FUD but w/e
<mrnuke>
I don't know how to work with fex (uncertainty), I don't know if I can get it to boot (doubt). and I'm afraid I'd be wasting time with it (fear)
<Turl>
haha
<Turl>
it's just a compilable ".ini" file if you ever dealt with those
<Turl>
bbl
<mrnuke>
I need fex2bin and other tools, right?
<Turl>
yeah, those are the (de)compiler
<Turl>
(well, it's a single tool, fexc, the others are just convenience aliases)
<mrnuke>
fuck that. I'd probably fuck up compilation of fexc or something
<Turl>
booting simply consists of reading the binary version into a hardcoded address in memory, then the kernel is booted as with any legacy arm system
<Turl>
apparently you can get a free sample read with their android app if you can read german :)
<rm>
price also matters, I take it the CT is not the cheapest one
<libv>
it's the best and most popular
<libv>
in my nokia days i'd buy them all the time so i had something to do on the plane, my gf once bought me a subscription, and if i was not travelling, they'd just lie around and i'd never get to them
<libv>
but most of my (ex)SuSE colleagues have a subscription
<rm>
of these, the beagleboard may turn out to be more popular than CT
<libv>
the beagleboard has been around for longer
<rm>
best = no argument here
<Turl>
rm: and if they're comparing the pi.. well, not much to say on popularity :)
<libv>
Turl: but it's nice that such a widely spread magazine does go beyond the rpi
<libv>
Turl: it will get us one or two more useful folks in here
[7] has quit [Disconnected by services]
TheSeven has joined #linux-sunxi
bsdfox has quit [Read error: Operation timed out]
bsdfox has joined #linux-sunxi
rz2k has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 272 seconds]
FreezingCold has joined #linux-sunxi
t3st3r has joined #linux-sunxi
wolfy has joined #linux-sunxi
wolfy has left #linux-sunxi [#linux-sunxi]
HeHoPMaJIeH has joined #linux-sunxi
atiti has joined #linux-sunxi
panda84kde has joined #linux-sunxi
vector80 has joined #linux-sunxi
vector80 has quit [Client Quit]
ssvb has quit [Quit: Leaving]
nikrou has joined #linux-sunxi
Quarx has joined #linux-sunxi
nikrou has left #linux-sunxi [#linux-sunxi]
vector80 has joined #linux-sunxi
chenam has quit [Quit: 离开]
<oliv3r>
mornin\
<vector80>
Hi oliver, good morning
<vector80>
Hi, I feel something really really wrong about Allwinner android partitions.
<oliv3r>
Fdisk under RHEL 6 or Oracle Linux 6 has a DOS compatibility mode. It can be turned off (and recommended by the man page) by using the -c parameter.
<oliv3r>
As far as I understand, the warning "Partition 1 does not end on cylinder boundary" can be ignored since most hard drives after 1996 use Logical Block Addressing (LBA) and so the physical sector characteristics of a drive are no longer a relevant concern when aligning disk partitions.
<vector80>
very good info
<vector80>
but this causes one problem
<vector80>
I want to use debian SDcard and NFS, to be able to dd if the android image
<vector80>
but because of 1st partition,
<vector80>
dd gtives "file too big" error
<vector80>
and I can not grab the image from my eMMC
<vector80>
can you recommend a solution for this?
<oliv3r>
then your partition is to small
<vector80>
No, as I wrote, I am using NFS
<vector80>
NFS have at least 22Gbyte of space
<vector80>
When I write my debian image to eMMC, I can easily grab its image to NFS, it works perfectly.
rz2k has quit [Read error: Connection reset by peer]
<oliv3r>
wait, let me re-read :)
<vector80>
But android partitions can't.
<oliv3r>
though this can not cause a problem :)
FR^2 has joined #linux-sunxi
<vector80>
oliv3r: my problem is, once I write sunxi uboot-kernel-rootfs to my eMMC, how can I recover it to be able to use android ... because you know, android image have boot0 and boot1, and my board have recovery button to enter FEL mode... But once I write debi,an image to eMMC, I must know how to enter to FEL mode...
<vector80>
I feel I think I need one special uboot to enter to FEL mode directly
<vector80>
Once I have seen such thing in wiki, but I can't find it now...
<oliv3r>
we have an MMC image, that forces fel mode
<oliv3r>
lets do a little asm talk
<oliv3r>
FEL mode + BROM are stored at 0x4000 in the memory space of the SOC; it's like flash, but can not ever be changed, the BROM/FEL
<oliv3r>
when the SoC turns on, it will always start executing code at 0x4000 first, that's just how it's wired
<vector80>
up to this point, it is clear
<oliv3r>
the BROM will run and check if you push the FEL button; it it is pushed, it jumps to FEL mode (which sits at 0x0002 (i forgot exactly, details are on the wiki)
<oliv3r>
what this mmc image does; is the BROM will boot from MMC, as usual
<oliv3r>
the first instruction in the FEL image is 'jmp(0x0002)';
<oliv3r>
hence, it'll execute fel mode immediatly
<vector80>
here is the feature request ( is it already available? ): current interra3_emmc.bin, can it include boot1 code, I mean, under sunxi uboot, if I press my RECOVERY button, can it enter recovery mode ?
<ccube>
oliv3r, i am a bit confused, because a stock debian does not make any problems on that board
<oliv3r>
vector80: aye; it's a micro-bootloader in a sense :)
<vector80>
oliv3r: you are no:1 !!! I wrote that file to SDcard, and it entered to FEL mode directly
<vector80>
and now writing... to eMMC or SDcard ???
<oliv3r>
ccube: ah, a10s; i know there where some issues in the past; but i assume they where resolved (booting mostly i think) sounds like there's errors in the board
<oliv3r>
ccube: you can try downclocking the memory to 360 MHz
maz_ has quit [Ping timeout: 272 seconds]
<oliv3r>
vector80: wel you are now in FEL mode; so you can only use fel mode via USB from the pc
<ccube>
oliv3r, thx i will try that
<oliv3r>
vector80: but why do you want to enter FEL mode; and we should try resolving your dd issue first :p since dd from/to mmc should work without trouble; but dd is evil; (dd is awesome) loopback mount the image, copy the content
<vector80>
oliv3r: Firstly I think we solved one issue
<vector80>
1st issue was, how to go back from debian to android by using that fel.sdboot file
<vector80>
and it seems working,
<oliv3r>
ah; well it still strange; FEL button works on all devices afaik
<vector80>
PhoenixSuit said it succesfully wrote the image
<oliv3r>
vector80: cool :)
<vector80>
Let me double sure about that
dlan has joined #linux-sunxi
<vector80>
oliv3r: Writing different android image... just to make sure...
<oliv3r>
sure :)
<oliv3r>
though even for android; you don't need pheonix suit :p
<oliv3r>
though we don't have proper android images yet :(
<vector80>
I don't need phoenix Suit? How to write android image to my EMMC then ?
<vector80>
BTW: test is succesfull... I wrote different image, and double confirmed it... Thank you soo much again...
<oliv3r>
mount -o loop android.img; cp -ar :p
<vector80>
Pls tell me, is there an other way to be able to write android image to eMMC ?
<oliv3r>
boot your master SD debian image, and copy the android files over :)
<slapin_>
mripard: ping
<oliv3r>
i mean you CAN use fel mode
<oliv3r>
and sunxi currently doesn't have the tools yet to 'fel-mount'
<oliv3r>
but quite possible though
<slapin_>
mripard: do anybody actively working on DMA support?
<oliv3r>
vector80: if your really board, write a fel program, that exposes the emmc over usb as storage device
<slapin_>
vector80: some firmwares support android fastboot gadget
<vector80>
oliv3r: You mean I must loop mount android image, and and overwrite eMMC ?
<oliv3r>
slapin_: once you've done MTD stuff; we can start working on u-boot fastboot :)
<oliv3r>
vector80: you could if you wanted too yeah
<slapin_>
oliv3r: grrrrrr
<oliv3r>
vector80: but that's all scriptable
<oliv3r>
slapin_: :D
<oliv3r>
slapin_: but our u-boot doesn't do fastboot yet does it
<vector80>
ok, to be able to do this, I must be able to create android.img
<slapin_>
oliv3r: what is the status with bbrezillon's driver?
<vector80>
but currently, I can't create android.img
<oliv3r>
vector80: yeah i don't think we have 'official sunxi android'
<vector80>
oliv3r: you don't need it
<mripard>
slapin_: the plan has always been to merge SPI with PIO, to later use it to debug DMA
<oliv3r>
vector80: well, in the BSP you can do ./configure interra3_android
<oliv3r>
and then make :D
<vector80>
It is highly application based
<slapin_>
oliv3r: leechie sdk u-boot does
<oliv3r>
but i don't think the android build works yet
<oliv3r>
slapin_: really, interesting; though to be fair, i think it's just a u-boot option isn't it
<mripard>
I have SPI working since yesterday, so I guess I'll work on DMA in a few days
<vector80>
for example, we are currently modified official android SDK a lot
<oliv3r>
vector80: but you can use the allwinner sdk to build android, an 'image' that phoenix suit builds, isn't all magical :)
<bbrezillon>
slapin_: I've added HW ECC (I'll post a new series with HW ECC support soon)
<slapin_>
oliv3r: I don't think it was already integrated, but I have not checked lately
<vector80>
modified launcher, modified default home app, modifying SystemUI
<mripard>
but I'll be working on the A31 probably, so feel free to start working on another SoC
<vector80>
Even we need to make OTA with android
<vector80>
According to official sources from allwinner, nobody did that before
<vector80>
And we have to do :(
<vector80>
Anyway, lets go back to dd if issue
<vector80>
I am booting over TFTP-NFS now
<bbrezillon>
I'm still experiencing weird things: when I write on the 4th page of a block I get a lot of bitflips on the 1st one (too much to be corrected by the ECC hw or soft)
<vector80>
wait
<slapin_>
bbrezillon: how do you handle different random seeds and boot blocks?
<bbrezillon>
randomizer is not yet implemented, thought I tested it a bit with a fixed seed (0x4a80)
<slapin_>
bbrezillon: I think we can fight this one, can you post the full patch set on the list?
<bbrezillon>
sure
<slapin_>
I'll have tons of time tomorrow
<bbrezillon>
I'll do it today
<slapin_>
bbrezillon: by full, I mean full state, not only ECC-related
<JohnDoe_71Rus>
oliv3r: BSP does not give the correct android kernel
<slapin_>
bbrezillon: does it work without DMA?
<bbrezillon>
no
<slapin_>
hno: hi, do we have sunxi in mainline u-boot yet?
<slapin_>
bbrezillon: too bad, but how do you test it then as there is no DMA support in mainline yet?
<bbrezillon>
I'm not using DMA at all
<slapin_>
hno: I want to post u-boot related stuff for review
<wigyori>
morning
<bbrezillon>
I managed to get this working without DMA access
<slapin_>
bbrezillon: ah, that is perfect
<slapin_>
bbrezillon: I did it too, that is good
<slapin_>
bbrezillon: and hno did, for that matter
<vector80>
oliv3r: Booting NFS is ok now
<slapin_>
bbrezillon: can't wait to see your work (and add randomizer on top)
<slapin_>
vector80: have you managed to run Android off NFS?
<vector80>
slapin_: I have done that on IMX6 before but never tried on AW
panda84kde has quit [Ping timeout: 272 seconds]
panda84kde has joined #linux-sunxi
<slapin_>
vector80: well, I failed a month ago and was told that it is not possible due to how android treats network, etc, so I stopped, but I have a real need for that as there is device with 10G ethernet, 1G RAM and 2MB NAND chip inside (IPTV box left from unsuccessful project) which I want to use for kid's movies, so I'd like to know of possible solutions to problems
<vector80>
slapin_: it was very hard to do on imx6 too
<vector80>
slapin_: You are experiencing problems after kernel , right?
<vector80>
I mean, kernel is decompressing and starting device drivers,
<slapin_>
vector80: yeah, kernel boots fine off bootp
<vector80>
but when it comes to init, many things fail and it collapses, right?
<slapin_>
vector80: basically, yes
<vector80>
right
<vector80>
I expeienced same thing on IMX6 too
<vector80>
There was a very good PDF file
<vector80>
it explains how to boot android over TFTP+NFS on IMX6
<vector80>
I have did things on that PDF one by one
<oliv3r>
slapin_: we aren't mainlined yet, but we are on par with mainline; patches need big cleanup sthough
<vector80>
there were still issues but finally I managed to do it succesfully, it take about 3-4 days
<vector80>
But you people are much more experienced than me, you have good knowledge about linux, may be you can do it in hours
<vector80>
let me try to find it for you
<slapin_>
oliv3r: so, you mean no SoC in u-boot yet?
<vector80>
slapin_:
<vector80>
slapin_: I found my notes
<oliv3r>
vector80: :p
<vector80>
How can I give it to you?
<oliv3r>
slapin_: no sunxi soc no; it's all on our github
<vector80>
Do you have email ?
<oliv3r>
slapin_: but hno keeps it in sync with mainline
<vector80>
slapin_: I think you may do it on aw too
<vector80>
I just can't understand why aw did soo many good and soo many stupid things...
<slapin_>
vector80: thanks a lot!
<slapin_>
oliv3r: not even the smallest, tiniest SoC? too bad :( then I can't upstream NAND driver... that needs to be fixed, and fast... hno, are you here?
sgo11 has joined #linux-sunxi
<ccube>
oliv3r, still probleatic with 360mhz. when cold booting the board i am getting a kernel oops with sometimes stuck, sometimes automated reboot. after the automated reboot the system comes up (with the previously posted stacktrace in boot log)
<sgo11>
Currently, my ubuntu is working fine. but in linux-sunxi wiki, I saw this sentence "you should also eventually install cpufrequtils to set up an acceptable CPU-behaviour". Do I need to worry about that? How to determine if my CPU is working fine? thanks.
<bbrezillon>
slapin_: NFC v2 series sent
shineworld has joined #linux-sunxi
hno has quit [Ping timeout: 252 seconds]
hno has joined #linux-sunxi
<shineworld>
FYI: if someone is interested to use this kernel + android with AOA I've catch what don't works fine.... When an android device enter in Android Accessory Mode (AOA) and usb switch mode the accessory Interface is created and works BUT the VID:PID of device is wrong and not 0x18d1:0x2D00|0x2D01 but 0x18d1:0x005. At same time a lsusb -v -s X.X of device get an error in recovering same data: http://paste.debian.net/76338/
<shineworld>
At this point my knowledge of usb / init is so little that I can't figure out where but is
<shineworld>
*bug
kivutar has joined #linux-sunxi
<slapin_>
bbrezillon: thanks!
Tsvetan has joined #linux-sunxi
notmart has joined #linux-sunxi
newleaves has quit [Quit: Leaving]
<bbrezillon>
slapin_: which board are you gonna use for your tests ?
<slapin_>
bbrezillon: cubie, and some set of noname sticks
<bbrezillon>
could you test writing on the first 5 pages of a block, and see if you got bitflips on the first page ?
<slapin_>
bbrezillon: I'll do a lot of tests, including this one
<bbrezillon>
(I mean too much bitflips for ECC gracefully correct it)
<vector80>
oliv3r: hi, now, lets go back to my 1st issue, how can I dd my eMMC , any idea ?
hansg has quit [Quit: Leaving]
ZetaNeta has quit [Ping timeout: 272 seconds]
agath_at_work has quit [Ping timeout: 252 seconds]
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
jinzo has joined #linux-sunxi
rellla has joined #linux-sunxi
agath_at_work has joined #linux-sunxi
<sgo11>
where can I find A20 cpu frequency spec? All the pages I found do not mention it.
fredy has quit [Excess Flood]
Quarx has quit [Read error: Connection reset by peer]
Quarx has joined #linux-sunxi
shineworld has quit [Quit: Leaving]
fredy has joined #linux-sunxi
sgo11 has quit [Quit: Leaving]
cajg has joined #linux-sunxi
<bbrezillon>
slapin_,hno,rz2k: do you know the exact meaning of the NFC_REG_TIMING_CFG register ? I have to configure tADL (T19 in the datasheet), but don't where it fits in the NFC_REG_TIMING_CFG register
foubarre has joined #linux-sunxi
<foubarre>
wens: Hi. Are you around?
sspiff has quit [Quit: Konversation terminated!]
hipboi has quit [Quit: Leaving]
enrico_ has joined #linux-sunxi
t3st3r has quit [Remote host closed the connection]
FDCX has joined #linux-sunxi
discopig has joined #linux-sunxi
t3st3r has joined #linux-sunxi
<slapin_>
bbrezillon: AW configure this register to 0xFF and does nothing else.
<bbrezillon>
slapin_: that's what I do too, but the datasheet says T16 to T19 (tWB, tWHR, tRHW and tADL) should be configured in the NFC_TIMING_CFG register
<bbrezillon>
:)
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<wens>
foubarre: what's up?
<slapin_>
bbrezillon: experiments will show up...
pacopad has joined #linux-sunxi
rz2k has joined #linux-sunxi
samul has joined #linux-sunxi
arete74 has quit [Quit: leaving]
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
jinzo has quit [Remote host closed the connection]
<oliv3r>
Nobody ever let me open openstreetmap again
<oliv3r>
i just spent 2 1/2 hours editing shit
<slapin_>
oliv3r: what shit?
Black_Horseman has quit [Ping timeout: 265 seconds]
<slapin_>
bbrezillon: IIRC, you mentioned some IP block which was similar to our NFC, do you remember which one?
<bbrezillon>
slapin_: I was wrong
<bbrezillon>
:(
<bbrezillon>
(this was the cadence IP, used by Altera FPGA)
HeHoPMaJIeH has quit [Remote host closed the connection]
bbrezillon has quit [Read error: Operation timed out]
<oliv3r>
slapin_: my home, it's surroundings etc
nove has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 272 seconds]
ganbold_ has joined #linux-sunxi
arete74 has joined #linux-sunxi
sgo11 has joined #linux-sunxi
<sgo11>
hi, I have "[mmc-err] smc 0 err, cmd 52, RTO" and "[mmc-err] smc 0 err, cmd 5, RTO" errors at boot time. is this OK? I saw other people had it too. thanks.
<oliv3r>
check oyur usb sections in your script.bin
<hno>
bbrezillon, no. But should be easy to find with the help of a logic probe.
<ccube>
oliv3r, disabling usb results in not getting that message, but also no further output (freeze)
kivutar has quit [Quit: Ex-Chat]
<hno>
sgo11, it's ok to have some MMC errors at boot or when you insert a card.
techn__ has joined #linux-sunxi
<oliv3r>
hno: how do we compile fel-sdboot? i get errors about missing lds
techn_ has quit [Ping timeout: 265 seconds]
rz2k has joined #linux-sunxi
<slapin_>
hno: I think these timings are static, so it will be interesting experience in obscene expressionism, but some timings might be detected this way...
<sgo11>
oliv3r, hno, got it, thanks a lot.
<ccube>
thats weird. is there a minimal kernelconfig and script.fex existing for a10s board with most features turned off? only the really needed parts activated?
ddc has joined #linux-sunxi
<oliv3r>
ccube: double check you serial connection? maybe it stops outputting? i think that error is simply the last thing you get and it locks anyway
<oliv3r>
ccube: put used=0 whever you need
<ccube>
oliv3r, which are the neccesary parts in kernel config for serial output. CONFIG_VT=y CONFIG_VT_CONSOLE=y CONFIG_SW_DEBUG_UART=0
<oliv3r>
ccube: that looks like a 3.3 kernel?
<oliv3r>
but i don't recall; default sunxi kernels ahve it all one if im' not mistaken; but make sure you pass the proper bootargs
<ccube>
ah maybe the defconfig is the problem
<ddc>
I'm having a hard time getting csi to work with the ov7670 does any have any luck .
<ccube>
i have one config with too much stuff activated, where the serial is working, and an defconfig from openembedded sunxi layer, where it is not working, and i dont see the responsible part...
<oliv3r>
ddc: jon smirl on the mailing list is trying it too
<ccube>
the last line in serial, ia can see is: <6>Trying to unpack rootfs image as initramfs...
<ddc>
But still no joy
foubarre has quit [Ping timeout: 246 seconds]
<hno>
oliv3r, what fel-sdboot?
<oliv3r>
boot to fel mode from SD card
<hno>
ah, that.
<hno>
you likely need to use a raw-metal toolchain for that.
<hno>
or make a new lds script for it.
<oliv3r>
well there is no lds files anywhere in sunxi-tools
ddc has quit [Ping timeout: 252 seconds]
<oliv3r>
the file in /bin/ works, but yeah,
<hno>
There is somewhat arcane build instructions in the c file, building an object file and then dumping the code segment out of there.
<hno>
never even trying to link it.
<oliv3r>
ah ok
<oliv3r>
well precompiled works :p
<hno>
Tried and it compiled without any messages at all using gcc-4.7 linux gnueabihf toolchain. Haven't looked if the binary contains something meaningful yet.
<hno>
likely does.
hansg has joined #linux-sunxi
<oliv3r>
i just did make fel-sdboot.sunxi CROSS_COMPILE=xxx
<oliv3r>
but it complains about a missing .lds file
<oliv3r>
and i don't see any reference in the make file to build the lds file
<oliv3r>
ls: cannot access fel-sdboot.lds: No such file or directory
<oliv3r>
Current branch master is up to date.
<slapin_>
hno: please read backlog
bbrezillon_ has joined #linux-sunxi
bbrezillon_ has quit [Client Quit]
bbrezillon has quit [Ping timeout: 272 seconds]
bbrezillon has joined #linux-sunxi
sgo11 has quit [Quit: Leaving]
<bbrezillon>
hno: I don't have any logic probe (and I'm not sure these probes can handle such high speed busses, though we can underclock the NFC clock), do you ?
hansg has quit [Quit: Leaving]
<slapin_>
bbrezillon: I have and already did NAND dumps, yes, clock can be lowered
nikrou has joined #linux-sunxi
<nikrou>
Hi all
<bbrezillon>
slapin_: okay, then I'd be interrested in these results (using my driver, you can underclock the NFC using the dt timings definition or by statically configuring the clk_rate field in the sunxi_nand_chip_init_timings function)
kivutar has joined #linux-sunxi
ddc has joined #linux-sunxi
<arokux>
big article about arm boards in german c't including AW-based pcduino, olinuxino micro and cubietruck. linux-sunxi is mentioned in one sentence as: Sunxi-Projekt
<arokux>
hno: you are still lurking?
<WarheadsSE>
arokux: I've seen two people mention this article, non with a link
<nikrou>
But I cannot find a valid rom with a nandj partition pointing to sdcard
ddc has quit [Ping timeout: 252 seconds]
<nikrou>
Two questions : 1) do you think it could work ? 2) do you know where to find that kind of rom ?
<nikrou>
Thanks in advance
<oliv3r>
anybody tried backporting the brcmfmac stuff to 3.4?
<nikrou>
not me ! :-)
samul has quit [Quit: ERC Version 5.3 (IRC client for Emacs)]
<nikrou>
oliv3r, some days ago, you proposed me to install fedora on sdcard to boot my tablet blocked in recovery mode. Do you think I could restore it from the dump of all my nands ?
<nikrou>
I successfully boot my device with debian on sdcard but after I don't know what to do to restore my android system on nands !
<nikrou>
Any idea ? Is it possible ?
<oliv3r>
nikrou: technically; sure
<oliv3r>
nikrou: but first you probably want to see if you can loop-mount your images
<WarheadsSE>
arokux: looks like they overlooked ath Arch runs on all those..
<WarheadsSE>
as usual..
<libv>
ah, nikrou
<libv>
with the rebadged inet_86vz
<arokux>
WarheadsSE: no, there is a mentioning of Arch, in the list of the distros, I think, don't have an article right now
<libv>
nikrou: i see that you've created an account
<libv>
nikrou: did you manage to take a picture of the backside of your device?
<libv>
nikrou: also, you had access to some android configuration files where you got the board identification data from
<WarheadsSE>
arokux: only mentions it on the Pi
<arokux>
oh, ok, well, c't is a phoronix like magazine as for me. i just read some articles it because it lies around
<oliv3r>
nikrou: do you have access to your tablet when booting from fedora/debian? e.g. ssh, keyboard, something?
jemk has joined #linux-sunxi
rz2k has quit [Remote host closed the connection]
<WarheadsSE>
arokux: mm
<WarheadsSE>
relating it to phoronix (moronix) makes suddenly more sense as to why massive details are missed.
<libv>
arokux: it really is quite good in comparison
<libv>
arokux: as i stated many hours ago, several of my friends/colleagues swear by it, and in my nokia days i made sure i had 2-3 copies saved up for planerides
<wens>
oliv3r: you could try copying the SDIO IDs and drive strength stuff
<hno>
and it's easy to run the nand controller at low clock if needed, and might be while measuring as the probes do disturb the bus a bit.
<libv>
WarheadsSE/arokux: this will get a signification portion of german it folk to look into the allwinner boards (which are 3 out of 5)
<libv>
and it will get us a few good folk as a result
<oliv3r>
libv: well i'll be damned
<oliv3r>
Puneet B helped jonsmirl fix his issue
<libv>
oliv3r: hell did freeze over last week
<oliv3r>
if we get 1 new dev from this; a 'stick around one' then that's a win
<arokux>
libv: sure
<wens>
oliv3r: you have the patch Arend posted?
<oliv3r>
wens: yeah
<libv>
(the us town called hell)
<oliv3r>
libv: and that mail was from last week; coincident?
<oliv3r>
and you never replied to my /pm
<nikrou>
Sorry I was back for minutes
<nikrou>
So I take more pictures
<libv>
ah, i indeed didn't
printallthething has joined #linux-sunxi
<hno>
oliv3r, fel-boot was renamed to fel-sdboot, and that file was missed. Just rename it.
<nikrou>
I confirm I have inet 86vz Zeng gc 2013-07-08
<oliv3r>
nikrou: you REALLY should connect serial output and see if boot0/boot1 and even u-boot try to boot; i don't think your boot0/boot1 are broken
<oliv3r>
hno: ah! ok thanks; will do and test, submit, push
<hno>
arokux, always lurking. Not always catching up with backlog.
<nikrou>
Oliv: I need to retrieve them
<nikrou>
I spoke about boot0/boot1
<arokux>
hno: will you pull in some usb patches for u-boot? any plans to mainline u-boot-sunxi?
<nikrou>
(16:09:03) oliv3r: nikrou: do you have access to your tablet when booting from fedora/debian? e.g. ssh, keyboard, something?
<nikrou>
Sure I have an ssh access
<hno>
arokux, oliv3r is the main sunxi u-boot patch monkey at the moment ;)
<oliv3r>
nikrou: i don't remember :)
<oliv3r>
nikrou: if you have ssh access, can you do fdisk -l /dev/nand? do you see partitions at all?
<hno>
yes there is plans, just need to figure out how to make 36-hour days...
<nikrou>
I see no partition
<nikrou>
only nandg created by CWM
<oliv3r>
hno: yes! we need you to have more sunxi-time :p
<oliv3r>
nikrou: that's ok
<oliv3r>
nikrou: try to get nand_part running on the tablet
<hno>
allwinner partition table is not recognised by fdisk. You need nand-part from sunxi-tools to work with those.
<oliv3r>
nikrou: its in sunxi-tools; you will hvae to cross-compile it (shouldn't be a problem; i can send you a binary if you can't manage it or you can check the /bin dir)
<nikrou>
I've got an error when try nand-part don't remember it
<nikrou>
I will give it another try.
<mripard>
oliv3r: so, you're the one I should ping for my A31 uboot patches ? :)
<oliv3r>
nikrou: you may need to compile a sunxi kernel; theres special patches there to work with nandpart
<nikrou>
Need to reinstall my sdcard and boot my device
<oliv3r>
mripard: don't expect heavy technical reviews! :p
<oliv3r>
mripard: but i can try to do my best :)
<oliv3r>
mripard: AND i can't test
<nikrou>
I already had a custom kernel
<nikrou>
I have to had all related nand stuff I imagine
lioka has quit [Quit: leaving]
<hno>
oliv3r, job of patch monkey is to do a final review on style cleanness (no other unrelated stuff), verify that there is sufficient community testing/review and judge if there might be a risk of regressions in other areas, then merge if passing tests.
<hno>
no one can master all technical areas. I certainly don't.
<hno>
note, tests is those listed. Not functionality testing (community peer review takes care of that).
<oliv3r>
testing will be tricky i'm affraid, a31 has an extremly small community
arete74_ has joined #linux-sunxi
arete74 has quit [Read error: Connection reset by peer]
<hno>
oliv3r, then you have to rely on test results by the developer submitting the patch. But seriously, there is only a single person in the community using A31?
<hno>
arokux, does 4.4 require new GPU stuff not in current SDK or some kernel stuff not in our 3.4 kernel? If not then you should just be able build it and run..
<arokux>
no idea
<hno>
then try.
<hno>
libv, you want those nand patches in lichee-dev?
Quarx has quit []
<hno>
libv, first patch is a mix of fat & nand debug changss. Needs split.
cajg has joined #linux-sunxi
slopez has quit [Quit: I'll be back]
FreezingCold has joined #linux-sunxi
pacopad has quit [Quit: pacopad]
FreezingCold has quit [Read error: Operation timed out]
paulk-collins has joined #linux-sunxi
Seppoz has quit [Ping timeout: 260 seconds]
Seppoz has joined #linux-sunxi
panda84kde has quit [Quit: Konversation terminated!]
wingrime has quit [Ping timeout: 252 seconds]
AreaScout has joined #linux-sunxi
bbrezillon has quit [Quit: Ex-Chat]
cajg has quit [Ping timeout: 252 seconds]
ssvb has joined #linux-sunxi
_massi_ has quit [Remote host closed the connection]
vpelletier has joined #linux-sunxi
<vpelletier>
hi. I have a WITS Colombus A31 and would like to boot linux on it. I have the development kit's dvds with kernel, u-boot, buildroot source (and a few more folders), and I've quick-read a few provided pdfs, but I cannot find a description of how I can get it to boot on something else than pre-flashed content
<vpelletier>
is there a standard way of booting from SD, for example ? (there is no eMMC on board)
<hramrach>
vpelletier: see the linux-sunxi.org wiki
<hramrach>
maybe their u-boot does not do mmc
<vpelletier>
well, the in-flash u-boot has a 0 timeout
<hramrach>
sunxi u-boot does mmc but not nand
<vpelletier>
I do see "press any key to interrupt", but it doesn't react
<mripard>
vpelletier: you need to hold the button next to the four component connectors while powering it to boot on SD
<vpelletier>
ah, so that's the "u-boot" button
<hramrach>
it no longer autoboots?
<vpelletier>
so my SD image is doe wrong
<vpelletier>
yes
<vpelletier>
at least, no output on UART0, while I get some when not pressing that button
<libv>
hno: i could just as well drop them, as they were only there to help me find why it failed to do anything
<libv>
and i thought they might be useful for others who will undoubtedly run into the same issue
<hno>
vpelletier, if it's the dev board then there should be a boot selector jumper.
pwhalen has quit [Read error: Operation timed out]
<vpelletier>
hno: there is a "u-boot" button next to the component connectors, as mripard mentioned, which indeed affects boot (ie, nothing output on UART0 when held on power-on with no SD inserted)
<vpelletier>
I'm reading the SD creation page right now
bsdfox has quit [Ping timeout: 260 seconds]
FR^2 has quit [Quit: Connection reset by peer]
<hno>
vpelletier, usually that button enters FEL mode. But maybe it also probes for SD on A31, don't remember.
<hno>
Right, it does.
<hno>
thanks mripard
<vpelletier>
if I understood ARM boot sequence properly, SPL is resposible for setting up the hardware before loading the next loader. is this right ?
<vpelletier>
(ram timings, ...)
bsdfox has joined #linux-sunxi
<hno>
vpelletier, yes.
<vpelletier>
which means that having the right SPL is critical for the board to work, and I do not seem to find one in the WITS-provided tarball (containing code and a few binaries)
printallthething has quit [Quit: Leaving]
<hno>
WITS do not use u-boot SPL. They use Allwinner boot0/boot1.
kivutar has quit [Quit: Ex-Chat]
<mripard>
vpelletier: could you make an iso image of the dvd, and upload it somewhere?
<vpelletier>
I did find a sys_config.fex (plain text) file describing what I understand as pins, voltages, frequencies, etc which I believe belong to that SPL
<mripard>
there was none shipped with mine....
<hno>
vpelletier, sys_config.fex is Allwinners "device-tree" source file for their custom configuration system.
<hno>
mripard, Eva might give you a download link if you ask.
<vpelletier>
hno: so something else exists that SPL would be based upon ?
<hno>
SPL means the u-boot SPL developed boot loader.
pwhalen has joined #linux-sunxi
pwhalen has quit [Changing host]
pwhalen has joined #linux-sunxi
<hno>
community developer boot loader.
<vpelletier>
ah, when I said "based", I meant that SPL needs to know some values about the board so it can initialise it correctly, not where the source comes from
atsampson has quit [Quit: kernel upgrade]
<vpelletier>
ie, some board-specific config file
<hno>
vpelletier, it also needs A31 support. We only have sun[457]i support in SPL yet.
<hno>
a31 is sun6i.
<libv>
there really is noone who cares for sun6i
<libv>
let's hope a23 gets a development board out soon
<vpelletier>
libv: any reason why sun6i would be uninteresting ?
<libv>
vpelletier: for me, because of pvr
<libv>
vpelletier: for others, i do not know
<hno>
pvr is the main setback for most I think.
<libv>
a80 is then going to be ignored just as much :(
<vpelletier>
what is the powervr status ? no graphics at all, or no accelerated 3D ?
<Turl>
A80, if with sata and usb3, could have a place in NASlike devices
<libv>
vpelletier: i will not touch it with a 10' pole
<hno>
libv, any reason why you look at a23 instead of a20?
<libv>
hno: i wanted to buy an a13 tablet a month ago
<libv>
hno: and was quite surprised when i opened up the thing up
<hno>
libv, right..
<libv>
hno: it is higher clocked than the a20
<hno>
libv, given prior track record it should be quite similar to a20. Have you got console access working?
<rz2k>
hno: have you read the u-boot source for A23?
<Turl>
there's a new traffic tab there if you got push access
FreezingCold has joined #linux-sunxi
FR^2 has joined #linux-sunxi
<oliv3r>
hno: mripard has the a31 evb from wits; drachesun had it, but not expecting anything there; hansg has the mele a31 but was hugely dissapointed :)
shineworld has joined #linux-sunxi
atsampson has joined #linux-sunxi
<oliv3r>
arokux: paulk from the replicant project said he was going to port replicant to sunxi; since we have no need for blobs; replicant is sufficient
<oliv3r>
hramrach: vpelletier a31's BROM has different boot order; it checks pins, see the BROM page i think on the wiki
<ccube>
is there a RTC on a10s board?
<oliv3r>
rz2k: have you seen the u-boot sources? afaik they odn't do memory init there,s o marginally interesting
<oliv3r>
not sure if a10s has the rtc built in, sun4i does, sun5i doesn't; a10s might
<oliv3r>
nikrou: no; fi you have the sunxi 3.4.6* kernel; those patches should be there; if you are using 3.3 or older kernel; i don't think it has them
<oliv3r>
but i never had to use nand_part yet
<ccube>
so "hwclock: can't open '/dev/misc/rtc': No such file or directory" are expected and i should disable thin in my kernel
<oliv3r>
i always use mmc
<nikrou>
I cross compiled nand-part for arm. No problem.
<nikrou>
Ok for kernel
<oliv3r>
ccube: a10s doesn't have its own rtc then! :)
<oliv3r>
nikrou: what kernel do you have?
<ccube>
oliv3r, thx
<oliv3r>
try running nand_part then! :)
<nikrou>
I have a 3.4.75
<nikrou>
what is nand-device ?
<nikrou>
/dev/nand ?
<oliv3r>
or /dev/block/nand
<nikrou>
non nand in /dev
<nikrou>
nor /dev/block
<nikrou>
:-(
enrico_ has joined #linux-sunxi
<oliv3r>
nikrou: then i don't know; someone else here might know the details
<oliv3r>
mnemoc: hno or the likes
<hramrach>
oliv3r: that's nice
<oliv3r>
hramrach: i Know i am!
arokux2 has joined #linux-sunxi
<hramrach>
a pair of dip switches could give you boot device selection
<arokux2>
ugh.. after distro upgrade I cannot run lots of programs if there are too much files in the working dir
<oliv3r>
hramrach: it allows you to always boot from a certain media; excluding the others; so it's usefull i usppose
<oliv3r>
nikrou: i don't think your rom is corrupt :)
<oliv3r>
nove: that it is :)
<nikrou>
But I don't have a rom
<hno>
I haven't seen the a23 u-boot sources. Might be interesting for spotting if there is any general differences in hardware, but do not contain anything needed for SPL.
<nikrou>
I hope I can build one following a tutorial by modifing an existing one with as much as possible similarities
<oliv3r>
hno: we have the sources, but it's not hosted yet
pwhalen has quit [Read error: Operation timed out]
pacopad has quit [Quit: pacopad]
Seppoz has quit [Remote host closed the connection]
parabyte_ has joined #linux-sunxi
parabyte_ has quit [Client Quit]
<oliv3r>
hno: we have a big tarball @ dl.linux-sunxi/stulluk
<oliv3r>
or was it vector?
<oliv3r>
i forgot
pwhalen has joined #linux-sunxi
notmart has quit [Ping timeout: 252 seconds]
<hno>
nikrou, the nand driver failed to recognise your nand chip I think.
<oliv3r>
hno: yeah; mnemoc's laptop broke this week; so he can't do anything from home atm
<oliv3r>
he should get his new laptop one of these days
<hno>
found one in stulluk.
<oliv3r>
but u-boot should be cloneable
<oliv3r>
stulluk IS vector80
<oliv3r>
er branchable and pushable
<oliv3r>
but with my book; fosdem; and #linux-sunxi distraction i've been a little busy
nikrou has quit [Ping timeout: 245 seconds]
<hno>
copied it to /SDK/ for a start,
<hno>
Hmm.. readme says sun6i??
<hno>
but pack says chips/sun8iw3p1
<hno>
so yet another generation?
arokux2 has quit [Remote host closed the connection]
<hno>
right, it is sun8i
arokux2 has joined #linux-sunxi
<hno>
wonder what they do when they reach 10..
<libv>
sue us for taking the name?
<hno>
no boot0/boot1 sources in this sdk :(
<libv>
grmbl
<hno>
was hoping the a20 sdk had set a new standard on that
<oliv3r>
hno: they started to 'unify'
<oliv3r>
lkcl: ^ gpl violation again!
<hramrach>
hello
<oliv3r>
hi
Black_Horseman has joined #linux-sunxi
Black_Horseman has quit [Changing host]
Black_Horseman has joined #linux-sunxi
<hno>
oliv3r, not a single sdk released without gpl violtions.
foubarre has joined #linux-sunxi
<oliv3r>
hno: but they are improving!
<oliv3r>
but that's a nice quote :)
FreezingCold has joined #linux-sunxi
<hno>
heh, my monitor have segfaulted, First thougt my stationary computer had crashed with odd crashed text mode graphics displayed with broken font, but it's actually the monitor.
hramrach has quit [Ping timeout: 240 seconds]
agath_at_work has quit [Remote host closed the connection]
agath_at_work has joined #linux-sunxi
Black_Horseman has quit [Remote host closed the connection]
<hno>
no git history n u-boot tree.
<oliv3r>
:(
Black_Horseman has joined #linux-sunxi
<oliv3r>
hno: disassemble, connect lvds to cubie
<hno>
there is a sun9i_defconfig in buildroot.
<oliv3r>
i don't recall that one
<oliv3r>
a23 is sun8i; no clu ewhat sun9i is supopsed to be
<hno>
from sun9i... they are going for large address space?
<hno>
looks like both sun8i and sun9i have openrisk core for deep suspend mode. No dram code in standby code, only message passing to something and lots of SUNXI_R_--- I/O blocks.
arokux2 has quit [Remote host closed the connection]
<hno>
time for bed. More digging later.
<oliv3r>
yeah a23 has an 'arisc' remote_proc, wingrime found that they likley currently use it to play back audio
<oliv3r>
in sleep
<oliv3r>
could be usefull for audio applications in general
<oliv3r>
sun9i is quite likly to be the a80
<oliv3r>
so should show hints of powervr
<lkcl>
oliv3r: oh bloody 'ell not another one :)
<lkcl>
oliv3r: well, with that manager now gone - the one who was deliberately causing GPL violations - it should be easier to sort out.
<oliv3r>
lkcl: we've had it for a little bit; but yeah, libnand, no boot0/1
<oliv3r>
lkcl: oh really? do tell
<oliv3r>
this is news!
<lkcl>
could you email me what's involved?
<oliv3r>
same as always
<lkcl>
oliv3r: it shouldn't be!
<lkcl>
this happened ages ago (months back)
<oliv3r>
BROM/boot0/boot1/libnand/cedarX (though we got better cedrus now anyway)
<lkcl>
oliv3r: email me what's missing / involved, i'll remind them ok?
<oliv3r>
sure
AreaScout has quit []
arokux2 has joined #linux-sunxi
sulky has quit [Ping timeout: 252 seconds]
nove has quit [Quit: nove]
sulky has joined #linux-sunxi
tinti has joined #linux-sunxi
FDCX_ has joined #linux-sunxi
rm___ has joined #linux-sunxi
FDCX has quit [Ping timeout: 252 seconds]
<arokux2>
:$ my nfs setup does not work any more
<arokux2>
and i do not have an idea why...
arokux2 has quit [Remote host closed the connection]
zumbi_ has joined #linux-sunxi
plaes_ has joined #linux-sunxi
arokux2 has joined #linux-sunxi
TheSeven has quit [Disconnected by services]
on1x- has joined #linux-sunxi
[7] has joined #linux-sunxi
mripard_ has joined #linux-sunxi
rm has quit [Ping timeout: 245 seconds]
plaes has quit [Ping timeout: 245 seconds]
on1x has quit [Ping timeout: 245 seconds]
zumbi has quit [Ping timeout: 245 seconds]
ganbold_ has quit [Ping timeout: 245 seconds]
mripard has quit [Ping timeout: 245 seconds]
ganbold_ has joined #linux-sunxi
<oliv3r>
arokux2: only from sunxi, or from your pc too?
rm___ is now known as rm
<arokux2>
my pc is sever, u-boot-sunxi - client. everything worked :)
<oliv3r>
i ment a 2nd client to test
<oliv3r>
so you know if it's sunxi or the server
<arokux2>
ah
log0ff_ has quit [Ping timeout: 272 seconds]
<arokux2>
thought myself about this, but no other pc atm around :(
paulk-collins has quit [Remote host closed the connection]
specing has quit [Remote host closed the connection]
arokux2 has quit [Remote host closed the connection]
arokux2 has joined #linux-sunxi
<arokux2>
oliv3r: huh, downgrading a package solved the issue
penghb has quit [Quit: Leaving.]
popolon has joined #linux-sunxi
popolon has joined #linux-sunxi
arokux2 has quit [Remote host closed the connection]
enrico_ has quit [Quit: Bye]
Night-Shade has quit [Ping timeout: 272 seconds]
ganbold_ has quit [Remote host closed the connection]