<quitte>
I'll change the environment back to mmc for now. having it change between nand and mmc is annoying. doesn't help that every now and then it breaks on nand
<ddc>
bbrezillon : thank you for that. you just made my day
<quitte>
bbrezillon: have you seen my paste of 8 hours ago?
<ddc>
quitte: I don't have CT to test with but I'm using the same kernel
<ddc>
I could be an issue with the dt or with your mmc
<ddc>
you can try nfs
<quitte>
xmodem didn't give me any trouble at any time. I'll try a 3.13 dtb
<quitte>
04:14 < quitte_> [ 0.763193] Could not find valid ONFI parameter page; aborting
<quitte>
04:15 < quitte_> it's exactly at this point that the boot will stop if u-boot has NAND enabled. I'm entirely out of ideas what might be causing this
<quitte>
04:22 < quitte_> it doesn't get to the last of those lines in that case
<bbrezillon>
quitte: are you sure this bug is related to the NAND driver
<quitte>
since it's raining outside going to college and downloading pre-V4 is not an option today
<bbrezillon>
quitte: did you try to boot the same kernel without the NAND driver ?
<quitte>
bbrezillon: yes. I did a lot of double checking
<quitte>
yes I did. the same kernel runs just fine without nand in u-boot
<quitte>
bbrezillon: you have a cubietruck. can you send me a known working uimage and dts combination?
jukivil1 is now known as jukivili
<quitte>
and the dts
<bbrezillon>
quitte: ddc one is not working ?
<quitte>
i don't get any serial output. not even with the 3.13 dtb
<quitte>
bbrezillon: i now have u-boot such that i could triple check and know that nand support is the only change while the environment stays the same
<bbrezillon>
quitte: try to enable earlyprintk (CONFIG_EALRYPRINTK in your .config and earlyprintk keywoed in your bootargs)
<quitte>
i can't find it in menuconfig (3.13) it is in the bootargs
<bbrezillon>
quitte: when you search for a config option in menuconfig you can use "/'config-option-name'"
<bbrezillon>
quitte: in this specific case this gives /EARLY_PRINTK
<bbrezillon>
quitte: my bag this is not EARLYPRINTK but EARLY_PRINTK
<bbrezillon>
quitte: s/my bag/my bad/ :D
paulk-aldrin has joined #linux-sunxi
<quitte>
okay found it. i can't believe i never knew about searching
FreezingCold has quit [Ping timeout: 240 seconds]
Gerwin_J has joined #linux-sunxi
<quitte>
oh! zip download on github is downloading a snapshot. not the whole thing. that is great. i do have the pre-V4 source after all
<ddc>
quittee: can you set u=boot env to fdt_high=ffffffff
afaerber has joined #linux-sunxi
<quitte>
ddc: instead of 0xffffffff ?
<ddc>
quittee: yes
<ddc>
and save ur envs
FreezingCold has joined #linux-sunxi
<quitte>
bbrezillon: with early printk i now get a useful kernel panic where it's unable to mount /.
<bbrezillon>
quitte: that's better
<bbrezillon>
quitte: have you enabled the MMC driver
<quitte>
ddc: it was not the 0x
<quitte>
bbrezillon: it's enabled in both uboot and the kernel,yes
<bbrezillon>
quitte: and FS support for the FS type you're trying to boot on ?
<bbrezillon>
quitte: can you paste the boot logs ?
<quitte>
bbrezillon: this is another problem - but I guess I can figure out part of it from here. after all it worked with nand-less uboot already
<quitte>
the rootfs is squashfs and supposedly the free space on that mtd partition will be used as a jffs2 overlay
<quitte>
still having trouble with using that automatically created partition , though
<quitte>
i'm not sure how mtdblock is supposed to behave ... but i was getting io errors when trying to mark that partitioin for jffs2
<quitte_>
it happens - it seems - if i do a nand read in u-boot
<quitte_>
this happened every time with nand enabled uboot since it read env from nand
<quitte_>
as an explanation for the paste: "run mmcwrt" will boot just fine. "run thisbreaks" will just stop . this is with the earlyprintk enabled kernel
Gerwin_J has quit [Quit: Gerwin_J]
<wens>
damn, my delicate uart cable fell off :(
libcg_ has joined #linux-sunxi
libcg has quit [Read error: Connection reset by peer]
naobsd has joined #linux-sunxi
boycottg00gle has joined #linux-sunxi
quitte_ has quit [Read error: Connection reset by peer]
quitte has quit [Read error: Connection reset by peer]
quitte has joined #linux-sunxi
quitte_ has joined #linux-sunxi
<quitte>
bbrezillon: exactly the same with your kernel
<paulk-aldrin>
patches are probably awaiting moderation on the mailing list
<libv>
paulk-aldrin: add a quick one line blurp as to how this device has a 3g module :)
<libv>
and then yes, it is
<paulk-aldrin>
in which category?
<libv>
and i just oked your patches
<libv>
paulk-aldrin: at the very top of the page
<libv>
this device is special in that it should be very cheap and have a 3G modem
<libv>
i'll push your patches up right now, as they look good
<paulk-aldrin>
well the 3g modem wasn't sold with the device
<paulk-aldrin>
not on DX
<paulk-aldrin>
on other shops, it is
<libv>
ooh
<libv>
so this modem is available commonly as such a module?
<paulk-aldrin>
libcg_, yeah
<paulk-aldrin>
it sells with a regular USB adapter
<paulk-aldrin>
so you can use it on any computer
<libv>
then there definitely needs to be a 3G page for this modem
<libv>
which talks about this modules format
<libv>
and the ainol_aw1 should link to that when describing that it has this sort of module
<libv>
also, the infobox should change
<libv>
it's the first time i have seen such a thing though :)
<libv>
you do know how to pick strange hw :)
<paulk-aldrin>
well, AFAIK, only 2 ainol tablets can handle it
<paulk-aldrin>
and they only made that modem for the connector
bengal has quit [Ping timeout: 255 seconds]
<quitte>
bbrezillon: it took a while to find the place. there is no sunxi_nfc_rst() here. however it doesn't look like that helped. I'm doing a full kernel rebuild right now just to be sure the change gets used....
<libv>
still, a separate page is preferred, and it might get other people to add to a 3g/cellular modem page
deasy has joined #linux-sunxi
<paulk-aldrin>
alright
<libv>
thanks :)
<bbrezillon>
quitte: just use 'git apply' to apply the patch
<libcg_>
paulk-aldrin, dat fail :p
<quitte>
bbrezillon: this patch won't apply against V3 because of the very first line not matching
<quitte>
so i wrote the changes in manually
<wens>
libv: i seem to have shorted out my rx and tx lines :(
paulk-aldrin has quit [Read error: Connection reset by peer]
<libv>
wens: ouch
<libv>
wens: do you know how that happened?
<wens>
libv: one of the lines fell off, i had to re-solder it
<wens>
now i get loads of giberish from the uart
<libv>
i guess you already verified your soldering work?
<wens>
it worked the first time, when i hadn't put the case back together
<deepe>
Hello, I tried initialize SPI on A10-olinuxino-Lime, firstly I tested it with spi-dev on user space and it works great, but when I try just simple kernel module, which will send 8bits to the SPI
<deepe>
I got kernel panic
<quitte>
bbrezillon: the patch doesn't help. "%s:%i random enabled\n", __func__, __LINE__ is not printed on the screen when it works, either
<deepe>
Does anybody experience with SPI communication, with 3.4 linux-sunxi kernel?
xavia has joined #linux-sunxi
<libv>
again, you must've opened the case again to verify
<quitte>
can i find the boot0 binary somewhere on the web?
rafaelMOD has quit [Quit: Saindo]
<libv>
quitte: iirc, there was a disassembled one on our wiki
<quitte>
yes there is. but while waiting i wanted to try writing it to the spl partition.
HeHoPMaJIeH has quit [Quit: Konversation terminated!]
netlynx has joined #linux-sunxi
<wens>
crap, the pin fell off the soc
<wens>
so much for trying :(
<libv>
oh, this was on the a13 device where you soldered directly to the pin as the pads broke?
<libv>
now you broke off the pin?
<libv>
yikes
<wens>
libv: that's right
quitte has quit [Read error: Connection reset by peer]
FR^2 has quit [Quit: Connection reset by peer]
<libv>
:(
<Turl>
time to fel-boot and use uart on the sd port
libcg_ has quit [Ping timeout: 260 seconds]
<wens>
Turl: a13 doesn't have uart on the sd port
pwhalen_ is now known as pwhalen
pwhalen has quit [Changing host]
pwhalen has joined #linux-sunxi
<wens>
anyway, i'll submit whatever DT i had, but that's about as far as i can go on this one
<Turl>
meh :(
<quitte_>
bbrezillon: that was it! thanks. made my day,too.
bengal has joined #linux-sunxi
<libv>
wens: that's too bad, as it is really difficult to get exactly the same a13 device again :(
<wens>
libv: it is, and they're likely not made anymore...
<quitte_>
what package is that?
<bbrezillon>
quitte_: new things to fix for me :-)
<quitte_>
bbrezillon: the block of hex values in the dts is a random seed? is that seed similar to the spl one? or is that just used for randomization and could be replaced at any time by different values
<bbrezillon>
quitte_: the one defined on boot0 and boot0-recovery is the one used by the ROM code to detect a valid boot0 parition
ddc has joined #linux-sunxi
<bbrezillon>
quitte_: it has to be this one if you want to boot from NAND
<quitte_>
yes. which in theory should enable me to flash boot0 if i had it
<bbrezillon>
yep
<quitte_>
tomorrow's goal: try to have everything but tage 2 uboot running from nand ;)
<bbrezillon>
for the other partitions, you can change the random seeds, if you want to access the NAND from the SPL, u-boot and the linux kernel, they have to be the same every where
<quitte_>
Isn't that randomization good enough to reenable jffs2?
<bbrezillon>
quitte_: BTW, I noticed that yuq driver is not enabling the randomizer except for the nfc_read_page1k function (not sure what they're used for)
<quitte_>
after all as far as my googling tells me it was specifically disabled for MFC
<quitte_>
the 1k functions are used for writing the spl
<quitte_>
at least that's what I get from all the scripts he put in the environment
<quitte_>
in config/boards/sunxi-common.h
<quitte_>
s/mfc/mlc
<bbrezillon>
quitte_: then you'll have to disable the randomizer for the other partitions if you want to be able to read/write from u-boot
<wens>
heh, at least tx still works, so i have a readonly console
<wens>
i can probably use wifi to work on this one
<quitte_>
bbrezillon: I don't care about that kind of cooperation at the moment. I don't want my basic system, kernel and dtb easily changeable in a rw way. that extra trust that the system is as when i installed it is a lot of why i like openwrt and live systems
<bbrezillon>
quitte_: but you want to boot from NAND, and thus u-boot will have to read from a NAND paritition ;-)
<quitte_>
yes. but i can write that from u-boot itself
dack has joined #linux-sunxi
<bbrezillon>
quitte_: yes but your system will most likely use a NAND FS
<quitte_>
rootfstype=squashfs,jffs2 ;)
<quitte_>
well squashfs,tmpfs at the moment
<bbrezillon>
quitte_: or you'll have to store your kernel in a NAND paritition read/written by u-boot and your rootfs in another partition
<quitte_>
i have 5 partitions right now: spl;uboot;env;dtd;kernel;rootfs
<quitte_>
the kernel makes 2 out of rootfs if it finds a squashfs on it. rootfs, the whole thing and rootfs_data: everything after the squashfs
<bbrezillon>
quitte_: okay, it should work, as long as you don't touch u-boot, env, dtb and kernel parititions in Linux
<quitte_>
right. but i can easily mark them ro in the dts iirc
<bbrezillon>
quitte_: yep
<quitte_>
well then. i guess i'll use archaic phoenixtool to put boot0 back in place to play with it ;) thanks again
<quitte_>
yes I have. but i think this only makes sense without randomization
<quitte_>
?
<bbrezillon>
quitte_: I'm not sure (but I'm not a JFFS2 expert)
ddc has quit [Ping timeout: 272 seconds]
<quitte_>
but you could tell me what c->mtd->something to check for randomization?
<bbrezillon>
quitte_: anyway, with such a big NAND, I highly recommend that you use UBI/UBIFS just for scalability
<quitte_>
yes I read that jffs2 slows down linearily with size. otoh i really want the truck to act as a fileserver already,still.
<bbrezillon>
quitte_: the randomization stuff is just a proposal I made, there's now randomizer support in the mainline kernel, and I haven't exported this information to mtd users yet
<bbrezillon>
quitte_: and what's the problem with UBI/UBIFS ?
<quitte_>
whatever magic openwrt doesn to create the unionfs in preinit doesn't use it
boycottg00gle has quit [Remote host closed the connection]
ddc has joined #linux-sunxi
ecelis has quit [Quit: Lost terminal]
ecelis has joined #linux-sunxi
bonbons has joined #linux-sunxi
avsm has joined #linux-sunxi
<ddc>
quittee:In openwrt I noticed that the current way things are done is very specific for JFFS2 and there is no abstraction which allows different filesystems for the overlay.
<quitte_>
ddc: thanks. won't even look any time soon then
<libv>
i hope you made pictures before you soldered things on
<JannF>
yes, well documented
<libv>
but not on the wiki yet :p
<JannF>
they are on my DSLR as I had no time the past weeks
<JannF>
do you have a build environment set up?
<libv>
why?
<libv>
JannF: if you are putting linux on your device, you really should go through the motions yourself
<libv>
it's not that hard
<libv>
and we do have rather excellent documentation
<JannF>
ok
<libv>
JannF: but please first spend some time on that wiki page
<JannF>
that's my problem, time...
<libv>
JannF: this way, when you run into trouble, you can immediately document any issues you encounter
<libv>
heh
<JannF>
yeah I see
<libv>
this saves time in the long run
<libv>
JannF: actually, before you touch that page, let me paste in the latest device page example
ddc has quit [Ping timeout: 272 seconds]
<JannF>
libv ok
quitte has joined #linux-sunxi
bertrik has joined #linux-sunxi
_massi has quit [Remote host closed the connection]
<libv>
JannF: done
ganbold__ has quit [Read error: Connection reset by peer]
merbanan has quit [Ping timeout: 246 seconds]
ddc has joined #linux-sunxi
merbanan has joined #linux-sunxi
ganbold_ has joined #linux-sunxi
vbmithr_ has quit [Ping timeout: 246 seconds]
Renard has joined #linux-sunxi
vbmithr has joined #linux-sunxi
Nyuutwo has quit [Ping timeout: 246 seconds]
Nyuutwo has joined #linux-sunxi
<wens>
Turl: you beat me to the reply!
avsm has quit [Quit: Leaving.]
<JannF>
libv I've got a problem converting the script.bin into an script.fex
ddc has quit [Quit: ddc]
<JannF>
libv, nevermind, just my own fault
<Turl>
wens: :)
<Turl>
wens: yours is way better anyway :p
rz2k has quit [Read error: Connection reset by peer]
<Turl>
mripard_: is there something going on with your lxr? it has looked utterly broken all day (it still works though, it's just like if css were missing)
<JannF>
libv why doesn't bin2fex output any script.fex?
<quitte__>
rafaelMOD: are you sure that g_ether matches your stick? is that what a desktop loads?
<libv>
quitte__: g_ether is for running a small ethernet like network between a host pc and a device
<libv>
like adb
<quitte__>
oh
<quitte__>
so basically just a slip connection on steroids
<libv>
it is pretty neat, especially since you can then happily do development on trains and such, where wifi is decidedly flaky
<rafaelMOD>
libv: not working on A10 for me, did you enable any other menuconfig option for OTG?
<libv>
rafaelMOD: if i didn't document it on the wiki, then i probably didn't
<libv>
rafaelMOD: i'd state that otg is also not working on cubieboard1
<quitte__>
libv: while doing all those xmodem transfers i sure wished the whole usart part would disappear. having minicom work directly with a usb connection to u-boot would be pretty neat. preferrably without the hassle of setting up networking.
<libv>
quitte__: *shrug*
<libv>
having uart is very nice
<libv>
take that from a man who spent 2 years hacking lima on just android devices and no kernel source
<libv>
compared to that, you better be very thankful that you can usually have a working uart on our hw
akaizen_ has quit [Remote host closed the connection]
akaizen has joined #linux-sunxi
<quitte__>
i am. still - it seems very obvious to me to use a usb host and a usb client as a serial connection at full speed. but for some reason that doesn't exist. instead on uCs it is common to implement a usb to serial adapter in software ...
<Turl>
quitte__: there's g_serial too iirc
akaizen has quit [Ping timeout: 240 seconds]
<quitte__>
Turl: interesting. if that has no artificial speed limit it might be exactly what i was hoping for
<libv>
what? how does the speed of uart hamper you in any way?
<libv>
just wait a second longer for the dmesg to syphon through
<quitte__>
try loading a kernel or worse a rootfs over loadx
<libv>
quitte__: my feeling is that you're just whining about uart because you believe the grass is so much greener on the other side
<quitte__>
nope. I actually was too lazy to setup tftp and therefore had to wait about 7 minutes when transferring a kernel
<libv>
quitte__: we don't need to
<libv>
we have fel
<libv>
quitte__: how about you hitting our wiki and reading up on usbboot?
<quitte__>
i know how to do that. but i needed to transfer kernels to u-boot, not just run them
<quitte__>
hmm actually if the memory content survives resets that could work
<quitte__>
okay. I don't see at which point I get to uboots shell using the usb connection
<libv>
you don't.
<libv>
you get to that through the uart
<libv>
but you can load all the big bits you need over usbboot.
Faisal has joined #linux-sunxi
mmarker1 has joined #linux-sunxi
mmarker has quit [Read error: Connection reset by peer]
ogi has quit [Remote host closed the connection]
FR^2 has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
oliv3r has joined #linux-sunxi
<oliv3r>
l
<oliv3r>
lo
<Turl>
hiya
deasy has joined #linux-sunxi
<libv>
mnemoc: why on earth did allwinner not provide a signature for script.bin :(
<libv>
i'll code up a sanity check though
<libv>
section count and version digits smaller than 0x100 is a start
bengal has quit [Ping timeout: 240 seconds]
<libv>
then checking section offsets
atsampson has joined #linux-sunxi
<Turl>
libv: doesn't it have one?
akaizen has joined #linux-sunxi
<libv>
apparently not
<Turl>
maybe I'm confusing it with the boot0 one
<Turl>
the eGON thing
<libv>
otherwise jannf wouldn't have had the issues with the garbage he felled from his device
<libv>
fel
<libv>
n 1: to use FEL to pull device specific information from allwinner based devices.
<Turl>
heh
<Turl>
yeah, now that I think of it, it doesn't surprise me
<Turl>
the kernel just explodes if you don't load the fex
<libv>
ok, so the same sanity check (apart from comparing offset to size - perhaps compare that to some sane value) should also hit the sunxi-3.4 kernel
<rafaelMOD>
libv: I was able to modprobe g_ether.
<rafaelMOD>
Needed to enable OTG in .config.
<rafaelMOD>
Device Drivers→USB support→[*] SUNXI USB2.0 Dual Role Controller Support
<rafaelMOD>
and in OTG and related infrastructure: CONFIG_USB_OTG_UTILS=y / CONFIG_NOP_USB_XCEIV=y / CONFIG_USB_SW_SUNXI_USB=y / CONFIG_USB_SW_SUNXI_USB_MANAGER=y / CONFIG_USB_SW_SUNXI_USB0_OTG=y
avsm has joined #linux-sunxi
avsm has quit [Client Quit]
<paulk-aldrin>
libv, thanks for filling in the modem page, I got serious connection issues after we talked…
bonbons has quit [Quit: Leaving]
mmarker1 has quit [Read error: Connection reset by peer]
mmarker has joined #linux-sunxi
quitte has joined #linux-sunxi
quitte__ has quit [Read error: Connection reset by peer]
dack has quit [Remote host closed the connection]
ninolein has quit [Ping timeout: 260 seconds]
ninolein has joined #linux-sunxi
awafaa_ is now known as awafaa
traeak has joined #linux-sunxi
<traeak>
anyone stay alive in here?
souther has quit [Ping timeout: 240 seconds]
imcsk8 has joined #linux-sunxi
<libv>
rafaelMOD: throw that in the wiki :)
<libv>
paulk-aldrin: nah, thank you for ndhing your second device :)
linkmauve1 has quit [Ping timeout: 260 seconds]
quitte has quit [Ping timeout: 260 seconds]
oliv3r has quit [Ping timeout: 260 seconds]
buZz has quit [Ping timeout: 260 seconds]
orly_owl has quit [Ping timeout: 260 seconds]
deasy has quit [Ping timeout: 260 seconds]
buZz has joined #linux-sunxi
oliv3r has joined #linux-sunxi
montjoie has quit [Ping timeout: 260 seconds]
astr has quit [Ping timeout: 260 seconds]
montjoie[home] has quit [Ping timeout: 260 seconds]
bengal has quit [Read error: Connection reset by peer]
<rafaelMOD>
libv: Ok!
leviathanch2 has quit [Ping timeout: 272 seconds]
FR^2 has quit [Quit: Leaving]
FreezingAlt has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 246 seconds]
souther has joined #linux-sunxi
<Turl>
traeak: no, we're all zombies or sunxi devices with turing-test-approved software
<libv>
Turl: pfff
<libv>
traeak: you can tell by the way i use my walk...
<traeak>
hehe :-p
<traeak>
crap i wonder if i got overheat problems
akaizen has quit [Remote host closed the connection]
<traeak>
unplugged, let it sit 5 mins and it booted up okay
akaizen has joined #linux-sunxi
quitte has joined #linux-sunxi
<Turl>
that could be bad soldering
paulk-aldrin has quit [Quit: Ex-Chat]
<rafaelMOD>
libv: Didn't worked on cubieboard2, still working on it.
<rafaelMOD>
Only working on cubieboard1.
MSameer has quit [Ping timeout: 255 seconds]
<traeak>
forthe heat you mean? i really haven't had this thing running for at least a year
<traeak>
justwanted to test something
MSameer has joined #linux-sunxi
<libv>
rafaelMOD: a20 has broken otg
xavia has quit [Remote host closed the connection]
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
<rafaelMOD>
libv: driver broken or hardware broken?
Gerwin_J has joined #linux-sunxi
<Turl>
rafaelMOD: driver
<rafaelMOD>
Really? We (me and ricardocrudo) managed to make it work. Its a little unstable. It lost the connection twice but now is working, we PING the host.
ricardocrudo has quit [Remote host closed the connection]
quitte has quit [Read error: Connection reset by peer]
Nazcafan has quit [Quit: Quitte]
<rafaelMOD>
libv: I eddited the wiki but i don't know if it worked. I saved but I don't have permission.
<libv>
why don't you have permission?
bertrik has quit [Remote host closed the connection]
<libv>
rafaelMOD: i made you people now
<libv>
but that shouldn't change anything with respect to editing wiki pages
<libv>
unless of course you are stupidly trying to edit one of very few pages.