<drachensun>
techn: That patch works, it seems to have fixed it with multi touch on. My patch still addresses that invert settings are not honored in multi touch mode but that is pretty irrelevant if multi touch works
<drachensun>
so I guess turning off multi touch was just avoiding the real bug
<drachensun>
now I just have to figure out what can actually make use of it
gzamboni has joined #arm-netbook
<hno>
Hopefully u-boot now sets clock to 1GHz by default again. Testers welcome.
<lundman>
as opposed to?
<hno>
lundman?
<lundman>
if its 1GHz now, what was it before
<hno>
384MHz.
<lundman>
neat number
<lundman>
what are the advantages of making it go 1ghz during uboot?
<hno>
some seconds saved boot time.
<lundman>
ah because the boot sequence is executed as 1ghz
<lundman>
that is neat
<lundman>
so, its only a question of, if you leave it on uboot prompt, with the cpu get increasingly hot until it cooks?
<hno>
no
<lundman>
great!
<hno>
hmm.. do not seem to work too well..
<lundman>
hurroo
<hno>
ok. 1GHz disabled for now. Need to look at it closely
tuliom has quit [Quit: Konversation terminated!]
<hno>
or maybe it does work.. that's not the right u-boot for this board /-:
<lundman>
heh
<lundman>
hurrah
<hno>
Looking quite positive.
<lundman>
ok dd clearing an SD card for 1.3beta
<lundman>
although, maybe people dont care about ddimages anymore
<hno>
it gets to the uboot promt just fine, so there is a chance it even works.
<lundman>
might make my own nand writer.. small SD boot kernel that writes whatever image we want to nand
<hno>
lundman, dd images are less evil than non-dd images.
<hno>
but yes, an installation procedure not depending on Allwinner encrypted images would be nicer.
<hno>
phoenixcard images contain the exact same encrypted file, with just some fluff unpacked around it to boot.
ZaEarl has quit [Remote host closed the connection]
<lundman>
are you sure they are encrypted? encoded or compressed I would agree, but encrypted, whats the point in that
<hno>
the images are compressed and encrypted. I would guess at large due to legacy from earlier products and from being China..
<lundman>
when I dd in the SD card, I didn7t get the feeling that was encrypted
<lundman>
guess I didnt look hard
<hno>
the sd card contains the exact same firmware file embedded in the hidden data area.
<lundman>
"hidden"
<hno>
before first partition.
<hno>
boot0, boot1, boot(FAT at fixed location, not recorded in MBR), firmware, visible FAT filesystem on the rest.
<lundman>
seems reasonable. if you dont have the last partition, windows would ask if you want to format it :)
ZaEarl has joined #arm-netbook
<Turl>
lundman: I care about dd images :P
<Turl>
'your custom nand writer' might as well be CWM on a SDCard + a zip :P
<lundman>
SD card with uboot and std linux kernel, initrd script that mounts nand and writes an image to it. pretty much
<lundman>
but there is a lot of partitions and crap, so sounds like a hassle
<lundman>
maybe if I made it zfs
dyoung has quit [Ping timeout: 260 seconds]
<Turl>
why do you love zfs so much :P
dyoung-away has joined #arm-netbook
dyoung-away is now known as dyoung
dyoung has quit [Changing host]
dyoung has joined #arm-netbook
<lundman>
No more partitions, ever. All filesystems are more like creating a directory. tiny, fast, dynamic, change its size, compression, anytime you want
mSquare has joined #arm-netbook
RITRedbeard_ has joined #arm-netbook
<lundman>
ok phoenixcard finished writing
tekzilla has quit [Ping timeout: 246 seconds]
tekzilla has joined #arm-netbook
<lundman>
-rw-r--r-- 1 lundman lundman 306438847 Oct 10 12:12 Mele_HTPC_0929_BETA_V1.3.ddimg.bz2
<lundman>
lets boot then
<lundman>
same garish boot animation :)
<lundman>
4.0.4 / 3.0.8+. Same ad/malware homescreen
<lundman>
ugh, nobody run this.
drachensun has quit [Quit: Leaving]
alcides has quit [Quit: There's no place like 127.0.0.1]
mSquare has left #arm-netbook [#arm-netbook]
L84Supper has left #arm-netbook ["puff of smoke"]
rellla has joined #arm-netbook
Quarx has joined #arm-netbook
mSquare has joined #arm-netbook
<rellla>
sucessfully did a native compile of xbmca10 on my mele (debian unstable armel). xbmc also started, but cpu is up to 95% when scrolling through xbmc.
<rellla>
i'm now trying to compile it with A10HWR=1 like empatzero mentioned. i forgot this the last day....
<rellla>
i did a little protocol of the compile and should provide it in the wiki....
pawel5870 has joined #arm-netbook
sspiff has joined #arm-netbook
rsalveti has quit [Ping timeout: 246 seconds]
rsalveti has joined #arm-netbook
<lundman>
time rsync -a --progress ../linux-3.4.6 /zfs-testpool/
<lundman>
woo
* mnemoc
failed the mind-reading course
<lundman>
oh just compiled zfs RC11, and testing it. I can rsync the whole kernel tree, which is my method of testing I suppose
<lundman>
on the mele and cubox
<mnemoc>
now try compiling it ;-)
<mnemoc>
it's much more intense
<lundman>
root@cubox:/testpool/linux-3.4.6# make
<lundman>
as you command:)
<mnemoc>
no `time` ?
<lundman>
heh
<lundman>
suppose I could clean it, thebn make. but it is actually
<rm>
(meant Cubieboard, but it seems to be out of stock)
<mnemoc>
i also don't understand why luke likes to do these things so complicated, but it's the first sign of possible movement in like a year...
<RaYmAn>
I might have missed something, but if you were to back such a project, how would you actually use it? If it's just the card
<mnemoc>
initially nothing... the CB does everything and more as standalone board
<Turl>
mnemoc: sun5i doesn't have wemac does it?
<mnemoc>
but in theory his "associates" have laptops made already
<mnemoc>
Turl: right
<RaYmAn>
okay - I'm just looking at the crowd funding proposal and it just seems rather confusing
<mnemoc>
the same associates who don't want to make the schematics of their card public
<mnemoc>
Turl: sun3i and sun4i only
* Turl
gets confused by all the talk on arm-netbook lately
<TomNL>
lundman where to upload for you to mirror?
<RaYmAn>
Turl: should we stop? :P
<Turl>
I meant the ML
<lundman>
you have a login?
<lundman>
msg me a password
<mnemoc>
RaYmAn: the other option is simply forget the eoma68 project ever existed.... and change the $topic of the channel
<RaYmAn>
:/
<mnemoc>
RaYmAn: no one will make boards/devices before there are cards, and if no one commits to the cards because there are no devices, it's dead.
<rellla>
TomNL: did you do a cross-compile?
<rm>
a simple outputs board getting USB and HDMI out would probably be like $5
<RaYmAn>
mnemoc: I still feel like I'm missing something
<rm>
not to budget/design one AND expect people to crowdfund "just the board" (useless to them)
<RaYmAn>
mnemoc: is the board usable as a dev board by itself?
<rm>
is bizzare
<mnemoc>
RaYmAn: yes
<RaYmAn>
yeah, exactly what rm is saying
<rm>
<RaYmAn> the card is useless by itself, isn't it? E.g. you need a base board?
<rm>
<mnemoc> sure
<TomNL>
rella: yes cross-compiled
<rm>
so mnemoc make up your mind :S
<mnemoc>
the card is by itself an standalone a board, but it's real value over the CB is with devices to plug it into
<mnemoc>
s/ a //
<ibot>
mnemoc meant: the card is by itself an standaloneboard, but it's real value over the CB is with devices to plug it into
<rm>
there are none
<rm>
and it comes with nothing to plug it into
HeHoPMaJIeH has quit [Read error: Connection reset by peer]
<rm>
so you can look at it for 10 minutes and then put it on a shelf
<TomNL>
lundman: uploading file now
<mnemoc>
rm: sadly true
<rm>
meanwhile
<rm>
<rm> a simple outputs board getting USB and HDMI out would probably be like $5
<mnemoc>
usb and hdmi already have connectors, on the external side
<rm>
why not also make ^ this and have it come with the board in the first place
<rm>
I still believe/hope you misunderstand something
<rm>
and it will in fact come with such a board
<TomNL>
rella: if you want you can find cross compiled version on lundman's ftp
<rm>
otherwise that'd be too silly
<RaYmAn>
mnemoc: I know it's not you doing etc, but really, this is the kind of info that's missing :P
<rellla>
TomNL: what's your host and dest system? the last time i got stuck with the armel-devs. i'm too much newbie to that cross-thing ... is there any good help/ howto in the net?
<RaYmAn>
lkcl: ^^
<rellla>
TonNL: ok thanks, but i've to wait another few hours, i think ;-)
<TomNL>
host is precise...i followed hints in xbmca10 readme.
<TomNL>
basicly just apt-get the linaro armel gcc...
<rm>
One Does Not Simply apt-get the linaro gcc
<TomNL>
apt-get build-dep xbmc (Do this running on mele/mk802)....then copy whole rootfs from SD to folder on host
<rm>
unless probably they are on ubuntu
<mnemoc>
rm: on ubuntu you do :p
<TomNL>
rella: it's not very neat...better is to cross build all libs too..but i didnt manage to get this to work very good
<TomNL>
i tried native...took 6 hours..than errored out because of some memory problem
<rm>
hm, so power, ethernet, usb and sata are on the board itself?
<rm>
not that bad then
<rellla>
doing apt-get build-dep xbmc on the mele i haven't thought about. this should do the trick
<rellla>
thanks for the hint
<mnemoc>
rm: no no... those are two things in that picture, the card. with pcmcia connector on one side and uSD, uUSB, uHDMI and 3.5mm audio jack on the other
<rz2k>
hno: u-boot with mele_a1000 config works on my 1.6 mele that previously loaded up only thru default u-boot and then soft reset
<rm>
yes, from what I see > power, ethernet, usb and sata are on the EOMA card
<mnemoc>
rm: and the "mini engineering board" (which lkcl should make at the same time, as part of the offer)
<rm>
and do not require an extra board to use
<rz2k>
hno: tested it with couple of reboots and cold starts, thanks for your awesome work!
<rellla>
TomNL: i solved the memory problem by creating a swap file during compile
<lundman>
oh guess I should add .gz to the tester
<TomNL>
rella: ah ok :)
<mnemoc>
rm: sata and ethernet are on the pcmcia header, so connectors go in the missing i/o board
<TomNL>
lundman: you test for what?
<rm>
ok
<lundman>
validity
<TomNL>
ah ok hehe
<lundman>
:)
tuliom has joined #arm-netbook
<rellla>
TomNL: Put another "l" in my name, so I get highlighted ;-)
<TomNL>
rellla: ok
<lundman>
ｒｅｌｌｌａ: ok
<rellla>
haha
tzafrir has joined #arm-netbook
<mnemoc>
rm: mail sent
<TomNL>
do we have any update on armhf cedar libs from hipboi>
<TomNL>
>=?
<lundman>
you wish
<TomNL>
i guess i do:P
<lundman>
me too
<mnemoc>
we all do
<lkcl>
mnemoc: que? que? oh. yes. good point.
<rellla>
how can i disable this assembler warnings: "Warning: swp{b} use is deprecated for this architecture"? it trashes my screen :-(
sspiff has quit [Remote host closed the connection]
drachensun has joined #arm-netbook
AndChat|128225 has joined #arm-netbook
wingrime has quit [Ping timeout: 240 seconds]
<rellla>
github.com/hno/uboot-allwinner
<rellla>
is it possible to boot directly from sata already or use sd-card as "/boot" an sata-disk as "/" in hno's tree?
<rellla>
or is that one more recent: github.com/a10linux/uboot-multi-allwinner?
<drachensun>
rellla: nice work on the xbmc wiki
<drachensun>
I really needed that dependency list, I wasted a lot of time with packages last time I messed with it
<rellla>
last row in my terminal: LD xmbc.bin :-) <- hope this works after 6h compiling ...
<drachensun>
wow 6 hours, oh well
<drachensun>
I guess I should have played with the over clocking first
<mnemoc>
rellla: the kernel and script.bin need to be either on uSD or NAND. rootfs can be on sata
<lundman>
did we not add sata to uboot
<lundman>
hmm that may have been something I was looking at
valhalla has joined #arm-netbook
mSquare has left #arm-netbook [#arm-netbook]
wingrime has joined #arm-netbook
sspiff has joined #arm-netbook
<wingrime>
gingerbread roms faster than ics or jb?
<mnemoc>
wingrime: this isn't an android channel
<oliv3r>
bah, I did nothing on the datasheet, and it's hometime in 15 minutes :( i was lazy today :p
<mnemoc>
oliv3r: ordered an a10 device yet?
<oliv3r>
lol nope
<mnemoc>
:|
<oliv3r>
but! it's my birthday tomorrow, so maybe i'll have some extra birthday money to spend :)
<oliv3r>
my pivos should be arriving soon, i'll probably sell it in a few weeks ;( (took 8 weeks to get here)
<oliv3r>
(I ordered the pivos because xbmc does nativly work on it righ tnow, and the source, while questionable, is openly available)
<wingrime>
why in ft5x driver in resume and suspend irq_enable and irq_disable comented out and resume doing over reset?
<mnemoc>
rm: seems to be a lost cause after all... :-/
<rm>
what
<mnemoc>
rm: the crowdfunding of eoma68 cards/devices
<rm>
"crowdfunding" as in "let's get 15 people sign up in a wiki"? :D
revident has joined #arm-netbook
<mnemoc>
rm: i think he did the list because he doesn't believe there will be support
<mnemoc>
and after over a year in the same state reading the same answer just made me give up
avernos has quit [Ping timeout: 246 seconds]
<mnemoc>
rm: you were right. I naively misunderstood some recent mails like there was intention to open the project
<rm>
I was?...
<rm>
I don't remember stating anything in particular :)
<mnemoc>
nothing in particular, only that I misunderstood something. I just figured out what was that something
<mnemoc>
like when oracles talk ;-)
<hno>
lundman, no SATA or NAND in u-boot.
lerc has joined #arm-netbook
specing_ has joined #arm-netbook
zoobab has joined #arm-netbook
RaYmAn_ has joined #arm-netbook
<hno>
Why do people think that "uboot-multi-allwinner" is something to look at? That's the ICS u-boot with only slightly modified default config.
avernos has joined #arm-netbook
<hno>
no SPL support, no MMC support, and most piecies following Allwinner standard code quality.
<mnemoc>
because it's advertised in a popular forum
pawel5870 has quit [*.net *.split]
specing has quit [*.net *.split]
lerc_ has quit [*.net *.split]
RaYmAn has quit [*.net *.split]
zoobab_ has quit [*.net *.split]
WarheadsSE has quit [*.net *.split]
WarheadsSE has joined #arm-netbook
sspiff has quit [Remote host closed the connection]
<hno>
hm. guess a number of u-boot tickets can be closed now,
<mnemoc>
hno: i haven't tested the latest uboot on the a13 or mele yet, but this morning my cubie halted on spl's mmc detection after the second reboot
<mnemoc>
(it used to be on the first ;-)
drachensun has quit [Quit: Leaving]
drachensun has joined #arm-netbook
<rellla>
ok. xbmc is runnings smooth. how can i check, if hw-accel is used?
sspiff has joined #arm-netbook
sspiff has joined #arm-netbook
sspiff has quit [Changing host]
<rellla>
framerate at 1280x1024@60Hz is around 35-40fps
<rellla>
in system info: GPU: Mali-400 MP
<rellla>
OpenGL vendor: ARM
<rellla>
OpenGL version: OpenGL ES 2.0
<rm>
rellla, is this on Android or under a GNU/Linux distro?
<rellla>
debian sid armel
<rm>
interesting
<mnemoc>
also check cpu usage
<mnemoc>
and cpu freq
<rellla>
cpu-usage, what xbmc says ~ 80-90%
<rellla>
top: ~83 us, ~6 sys, ~10 id
<rellla>
cpu freq?
<mnemoc>
/proc/cpuinfo
<mnemoc>
but it smells like it's all sw accelerated
<mnemoc>
rellla: you only need to show the bogomips value
<mnemoc>
400MHz is still low
<rellla>
298.17
<rellla>
varies between 300-400
<mnemoc>
playing an hd video fluently? it's good. vpu is been used
<popolon>
thx
<rellla>
not that fluently. only 15fps.
furan has joined #arm-netbook
<rellla>
could this be, because i use a vga-display and xbmc at 1280x1024@60Hz?
<mnemoc>
nope
wingrime has quit [Ping timeout: 245 seconds]
<rellla>
when setting A10HWR=1 i get ~60fps in gui
<rellla>
when playing the hd-movie, xbmc plays with ~24fps, but nothing can be seen.
<mnemoc>
gui is NOT rendered by the VPU
<rellla>
i get this warning: [DISP_WRN] file:drivers/video/sunxi/disp/de_bsp_sun4i/de/disp_layer.c, line 124: not supported image0 pixel sequence: 216 in img_sw_para_to_reg
<mnemoc>
i've never used xbmc, but mali400 only cares about 3D
<mnemoc>
(GL)
<TomNL>
so you dont see anything on screen when playing back video?
<rellla>
i see the osd-sysinfo, a black area which should possibly be the video? and the rest of the warning 10times a second at the left,right,upper and down borders
<TomNL>
rellla: i hope to be able to try tonight
<TomNL>
rellla: what linux are you on?
<rellla>
TomNL: a debian unstable - armel
<TomNL>
in the xbmc log, it shows its using hw playback with hw layer?
<TomNL>
maybe he is incompatible with newer kernel?
<mnemoc>
nah, nothing has changed that could affect it
<rellla>
maybe. i'm using amery's default allwinner branch?!
<rellla>
amery must know it ;-)
<mnemoc>
amery has never tried to play a video on an A10 device except on stock android
<rellla>
;-)
<rellla>
btw: warning runs down screen, even when i kill xbmc
<Turl>
mnemoc: kexec is really convenient :)
<Turl>
you should try it :P
<mnemoc>
almost all my sunxi time is devoted to vi+make+rsync+reboot+booted?+loop
<Turl>
mnemoc: you can cut it to vi+make+rsync+kexec+goto vi
<mnemoc>
nah, it's not the same
<Turl>
indeed, it's better :) if your kernel crashes on boot the device reboots normally after the panic without having to pull the sdcard to copy a fixed kernel or anything
<mnemoc>
that's why I keep two uImage files in the card ;-)
<Turl>
heh
aexl has joined #arm-netbook
<aexl>
hello
<rm>
hi
<aexl>
hno: is it possibe to get u-boot outpoot with micro sd sniffer? i only get kernel output ...
<aexl>
output
<mnemoc>
yes, modifying the uboot in the nand capable branch of hno's tree
<mnemoc>
lichee-dev iirc
<aexl>
which branch? couldn't find it in the wiki.
<aexl>
:D
<mnemoc>
it's unmaintained, that's why
<mnemoc>
the priority is still to get the SPL working well
<mnemoc>
aexl: in hno's allwinner-info there is some stuff that is supposed to let you have jtag/uart/mmc in the same uSD
<mnemoc>
jtag controlled
wingrime has quit [Remote host closed the connection]
<mnemoc>
voodoo
<aexl>
thx
<aexl>
he made a comment about "tristating the JTAG probe" whatever that means
<aexl>
in the ML
<mnemoc>
yes, that's the voodoo
<mnemoc>
look at allwinner-info repo
<aexl>
jtag folder?
<mnemoc>
clone it an unleash your curiosity ;-)
<aexl>
first i'd have to develop some understanding of c or python or coding in general ... ...
<mnemoc>
most stuff there is information, not code
<cat1>
hmm.. is this something i should worry about:
<afgdsgh>
btw, android system is WoW few years ago when when was android first start i didnt think it will be so much popular because it required to good work like ios strong hardware, but at this moment all devices(dvd players/blue-ray players/tv/ etc..) will have android system instead of its own firmware
<afgdsgh>
mnemoc hm isnt that board buggy and week as raspberry pi?
<zoobab>
got some a10 netbook with android, would like to install debian on it
<hno>
drachensun, random, or every time when trying to compile thatr file?
<hno>
zoobab, for now it's easier to install debian on an sdcard, leaving the internal storage alone.
<Turl>
mnemoc: :)
<rellla>
drachensun: hm... maybe i forgot st in the wiki? cmake?
<drachensun>
I did have to add cmake
<drachensun>
I was going to add that to wiki once I finished my build, to see if anything else came up
<drachensun>
but yeah, I already did that
<drachensun>
that error is just so strange to me I'm not even sure what to try
<drachensun>
is there a premade taglib I can get?
<hno>
drachensun, do you get the same error each timew you try building thsat file?
<drachensun>
hno: yup
<hno>
how much memory + swap do you have?
<drachensun>
hno: it doesnt vary though, sometimes 5% done, sometimes 1%
<drachensun>
well first I was building with my 1G of ram going, 860m that shows or so
<rellla>
drachensun: mine, but tomorrow. im not in front of mele
<rellla>
1G which device?
<drachensun>
now I am running with 512 and have about 460 with the swap
<hno>
how much swap?
<drachensun>
393M swap
<drachensun>
1G tablet, its a generic oem model
<hno>
so very little memory for building stuff.
<rellla>
on the mele i had less
<drachensun>
well I had 700M free before and had the same thing happen, with all the ram enabled
<drachensun>
I can make the swap bigger though I guess
<hno>
you say it crashes at the same file but different %. What do you mean?
<techn>
hno: yearlier boot crashed before X (with a10), now it boots to X.. No, I dont have UART.. yet.. I have ordered one.. Haven't tried to boot android with other u-boot than which is on nand
gimli has quit [Quit: Verlassend]
<drachensun>
as its building [ 1%] Building CXX object taglib/CMakeFiles/tag.dir/mpeg/id3v1/id3v1genres.cpp.o
<techn>
hno: I'm using correct script.bin.. but where I can find u-boot params?
<drachensun>
that [ %] increments
<drachensun>
Its gotten as high as 5%
<drachensun>
usually 1% though
<hno>
but always crashes on /xbmc/xbmca10/tools/a10/depends/taglib/taglib-1.8/taglib/mpeg/id3v1/id3v1genres.cpp:219:1: internal compiler error: in create_fix_barrier, at config/arm/arm.c:12079?
<drachensun>
rellla: Ok thanks, I'll try that tomorrow if this is still stuck
<drachensun>
I don't think so
<drachensun>
this time it was ....
<mnemoc>
memory corruption maybe?
<drachensun>
from /xbmc/xbmca10/tools/a10/depends/taglib/taglib-1.8/taglib/mpeg/id3v1/id3v1genres.cpp:26:
<techn>
hno: after testing more.. with correct power button timing boot success rate can be improved.. I also noticed that mid a13 tablet has samekind problem, but not that bad
<drachensun>
yeah, its gotta be something on my HW, its too random
<drachensun>
I just wonder why is coming up now, the rest of the system built fine
<hno>
drachensun, so then it's random., which can be triggered by bad hardware or running out of memory.
<hno>
or bad hardware configuration.
<drachensun>
is there a way I can force it to use the swap?
<mnemoc>
g++ is very skilled triggering those problems