mnemoc changed the topic of #arm-netbook to: EOMA: Embedded Open Modular Architecture - Don't ask to ask. Just ask! - http://elinux.org/Embedded_Open_Modular_Architecture/EOMA-68 - ML arm-netbook@lists.phcomp.co.uk - Logs http://ibot.rikers.org/%23arm-netbook or http://irclog.whitequark.org/arm-netbook/ - http://rhombus-tech.net/
<mnemoc> hno: can I add http://sprunge.us/QZTg?diff to lichee-dev? to fix O= support. ? that ugly thing with $(shell mkdir) is uses in several other places
freakazoid0223 has quit [Read error: Connection reset by peer]
tinti_ has joined #arm-netbook
tinti_ has quit [Read error: Connection reset by peer]
freakazoid0223 has joined #arm-netbook
popolon has quit [Quit: Quitte]
slash_random has quit [Ping timeout: 240 seconds]
slash_random has joined #arm-netbook
stefanro has joined #arm-netbook
tuliom has quit [Quit: Konversation terminated!]
stefanro1 has quit [Ping timeout: 265 seconds]
avernos has quit [Remote host closed the connection]
vinifm has joined #arm-netbook
bsdfox has joined #arm-netbook
bsdfox has quit [Changing host]
bsdfox has joined #arm-netbook
<bsdfox> anyone had crashes with wemac on sunxi-3.0 / mele a2000?
<hno> mnemoc, feel free to massage lichee-dev to your liking.
<hno> U-Boot SPL 2013.01-rc1-00092-g63cdc30-dirty (Nov 06 2012 - 04:16:08)
<hno> Board: Cubieboard
<hno> DRAM: 1024MB
<hno> spl: mmc init failed: err - -1
<hno> SUNXI SD/MMC: 0
<Turl> bsdfox: none that I know of
<hno> Thats SPL booted from SDC2 B)
avernos has joined #arm-netbook
avernos has quit [Changing host]
avernos has joined #arm-netbook
<bsdfox> is there a way to do that with git or should I just manually do it?
* bsdfox noob
<Turl> `git revert 6a7a8bd1adced6d0da860b004eb4967dae2a3d00'
<bsdfox> glad I asked.. that's pretty slick :)
<Turl> mnemoc: what happened to stage/ ? :)
<Turl> hno: nice :) why is mmc init failing though?
avernos has quit [Remote host closed the connection]
* hno is almost happy. SDC2 soldering worked on first attempt, but forgot to add some easy way to block the NAND before glueing the uSD adapter ontop of the NAND and now it's hard to reach.
<hno> Turl, because I am using an SPL made for SDC0 and have no card there.
<Turl> oh :)
<Turl> bsdfox: if that revert works, please fill an issue on https://github.com/linux-sunxi/linux-sunxi/issues and attach/link your device's script.bin/fex
<hno> Apart from forgetting to add something to disable the NAND (need to use a metal piece now) it works perfect. The SD and NAND chip happily coexists on the same wires. Can't access both concurrently obviously but that's no issue.
<bsdfox> Turl, looks like it's working
<bsdfox> Turl, where is my script.fex located? I've got script.bin
<Turl> bsdfox: script.fex is the output of `bin2fex' with your script.bin as input
<Turl> you can get the tool from sunxi-tools repo
<bsdfox> is there a way to boot an image not name uImage ? I'm wondering if I can keep a safe uImage on my SD card so I don't need to keep pulling it out and walking across my house to toss it in the machine my cross compiler is on
<Turl> bsdfox: yeah, you just need a proper uboot environment to load something named something else :)
benjamin__ has quit [Ping timeout: 255 seconds]
<bsdfox> Turl, any way to do it from the sun4i# prompt I get when I "Hit any key to stop autoboot"?
<bsdfox> I'd guess ext2load or something
<Turl> bsdfox: run help
<Turl> there's a command do print the env
<Turl> and there's another one to set env variables
<Turl> find where uImage is mentioned and reset those variables accordingly
<Turl> then run boot :)
<bsdfox> :)
<hno> bsdfox, stop at the prompt. Then "setenv kernel nameofyourkernel" and "boot"
<hno> assuming you are using my u-boot.
<bsdfox> works :)
<bsdfox> now I can be lazy and test things
QingPei has joined #arm-netbook
<Turl> hno: hm
<Turl> hno: doesn't cubieboard have a reset button already?
<Turl> under microUSB connection
freakazoid0223 has quit [Quit: Leaving]
ZaEarl has quit [Read error: Operation timed out]
ZaEarl has joined #arm-netbook
avernos has joined #arm-netbook
avernos has quit [Changing host]
avernos has joined #arm-netbook
ZaEarl has quit [Read error: Operation timed out]
ZaEarl has joined #arm-netbook
slash_random has quit [Ping timeout: 240 seconds]
slash_random has joined #arm-netbook
hno has quit [Ping timeout: 240 seconds]
hno has joined #arm-netbook
drachensun has quit [Quit: Leaving]
vinifm has quit [Quit: Ex-Chat]
bsdfox has quit [Ping timeout: 265 seconds]
cheng has joined #arm-netbook
cnxsoft has joined #arm-netbook
freakazoid0223 has joined #arm-netbook
<cnxsoft> I'm trying XBMC arnhf, and it can start but I got some problems with video playback ("not supported image0 pixel" message + black screen) that might be fixed with stage/sunxi-3.0, or so I read. The problem is that this branch is gone... Which branch should I use with xbmc armhf.
<cnxsoft> s/arnhf/armhf/
<ibot> cnxsoft meant: I'm trying XBMC armhf, and it can start but I got some problems with video playback ("not supported image0 pixel" message + black screen) that might be fixed with stage/sunxi-3.0, or so I read. The problem is that this branch is gone... Which branch shoul...
cheng has quit [Quit: Leaving]
slash_random has quit [Ping timeout: 240 seconds]
ZaEarl has quit [Read error: Operation timed out]
ZaEarl has joined #arm-netbook
QingPei has quit [Quit: Leaving.]
QingPei has joined #arm-netbook
jquip has joined #arm-netbook
eFfeM has joined #arm-netbook
gzamboni has quit [Read error: Operation timed out]
cat_x301 has quit [Ping timeout: 268 seconds]
rellla has joined #arm-netbook
rellla2 has joined #arm-netbook
rellla2 has quit [Remote host closed the connection]
rellla2 has joined #arm-netbook
rellla has quit [Ping timeout: 260 seconds]
mSquare has joined #arm-netbook
ZaEarl has quit [Read error: Operation timed out]
rellla2 is now known as rellla
<rellla> cnxsoft: sunxi-3.0 is what i use. changes of stage should be in master
<cnxsoft> rellla: Do you mean the change in stage/sunxi-3.0 have been merged to the master branch (sunxi-3.0) ?
<rellla> think so. imho it uses mali r3p0. it worked for me a few commits ago, but after merging from stage.
<rellla> newest is not tested.
<cnxsoft> Maps: Do you still have the "not supported image0 pixel" issue for XBMC you had at the end of November. If it's working now, what was the issue? Just the USB performance?
<cnxsoft> s/Maps/Maqs
<cnxsoft> s/Maps/Maqs/
<ibot> cnxsoft meant: s/Maqs/Maqs
<Maqs> i'll have to check that, but for that i need to be at home :-) will do it tomorrow
slapin has quit [Ping timeout: 276 seconds]
<cnxsoft> Thanks
<rellla> anybody had success with using ir on a mele? should it work together with lirc?
slapin has joined #arm-netbook
<mnemoc> Turl: as no one cared about stage branches I switched to tags instead
<mnemoc> Turl: the PHY_POWER change was pushed intentionally, knowing it will cause troubles, but a null emac_power doesn't have the same effect on my cubie
<mnemoc> Turl: the mali thing was unexpected
arnd_ has joined #arm-netbook
<hno> Turl, thats only FEL.
arnd_ has quit [Ping timeout: 256 seconds]
slapin has quit [Ping timeout: 240 seconds]
pawel5870 has joined #arm-netbook
slapin has joined #arm-netbook
ZaEarl has joined #arm-netbook
jquip has quit [Read error: Connection reset by peer]
arnd_ has joined #arm-netbook
Quarx has joined #arm-netbook
Quarx has quit []
ZaEarl has quit [Read error: Operation timed out]
ZaEarl has joined #arm-netbook
cheng has joined #arm-netbook
mSquare has quit [Quit: Leaving.]
tzafrir_laptop has joined #arm-netbook
jquip has joined #arm-netbook
popolon has joined #arm-netbook
<mnemoc> kind of long the page name :p
<mnemoc> may I rename it to LIRC ?
<rellla> +1
<mnemoc> also, please try to add categories, like [[Category:IR]] [[Category:Software]] [[Category:Tutorial]] :)
<mnemoc> helps to find stuff
<mnemoc> rename + cat done
<mnemoc> thanks btw ;-)
<rellla> it's not tested in production use ;-) but irw gives reasonable output at least
<mnemoc> good enough to make it happy day :)
cheng has quit [Quit: Leaving]
<jquip> Does the yaffs2 work fine on nand? Has anyone tried that?
<mnemoc> jquip: you mean on sunxi? we don't have an mtd driver yet
<jquip> yes on sunxi.. :) no mtd driver.. okie...
<mnemoc> but slapin is working on it
<slapin_> hi, all!
<jquip> that was.. ahem enthusiastic!
<mnemoc> :)
<jquip> thanks mnemoc :) . slapin: You have a repo somewhere for the mtd? If you show me the ropes, i can try to help out a bit..
<slapin_> do anybody knows how to fix minicom to transfer terminal size to console? I'm tired of wrecked busybox shell output :(
<slapin_> jquip: none
<slapin_> jquip: I'm currently documenting the controller
<slapin_> jquip: no code yet
<mnemoc> eval `resize` ?
<slapin_> jquip: I won't use nand_sunxi, only as documentation.
<slapin_> mnemoc: minicom. not ssh.
<jquip> you mean youre building from scratch??
<mnemoc> the current driver is way too different, refactoring is not possible
<mnemoc> well... maybe
<slapin_> mnemoc: how to transfer the result of resize using minicom?
<mnemoc> slapin_: i forgot minicom emulates a tty.... I use microcom
<slapin_> jquip: well, it is not that hard as soon, as I understand how it works.
<slapin_> mnemoc: regardless, microcom doesn't know its size either.
<mnemoc> it doesn't need to
<slapin_> mnemoc: how come? you have 100x40 terminal, and userspace assumes it is 80x25
<slapin_> mnemoc: everything will be broken
<mnemoc> resize will ioctl the real tty
<mnemoc> and then the exported vars will tell apps the size
<mnemoc> eval `resize` over microcom just gave me a lovely 156x20 htop
<slapin_> mnemoc: how the ttyS will know the size of your pty?
<slapin_> weird :/
<mnemoc> strace resize ? :p
<mnemoc> /dev/tty is magic
<jquip> slapin_ Do lemme know when documentation is sort of finishing! would love to chip in on the work if I am up to it... Can't seem to get my head around the kernel code yet... too big and changing quite fast.. and I don't know where to start to help... Some documentation would be a good thing..
<slapin_> jquip: you might look at http://ossfans.org/nand_registers.ps all suggestions and ideas are welcome
<slapin_> mnemoc: I don't have xterm nor any graphics on target :(
<mnemoc> slapin_: so? /dev/tty doesn't have anything to do with graphic interfaces
<slapin_> mnemoc: I mean I don't have resize tool on target
Almamuetya11 has joined #arm-netbook
<mnemoc> slapin_: ah, ok
<mnemoc> I think busybox has an applet for it too
<mnemoc> yes, it does
ZaEarl_ has joined #arm-netbook
ZaEarl has quit [Read error: Connection reset by peer]
gzamboni has joined #arm-netbook
merbanan1 has joined #arm-netbook
merbanan1 has quit [Remote host closed the connection]
<slapin_> mnemoc: rebuilding busybox is not an option at this time :( will have to script minicom somehow...
<mnemoc> it's just an ioctrl() to /dev/tty
<mnemoc> TIOCGWINSZ
<slapin_> stty -a shows cols 80 rows 25
<slapin_> so I don't know if this one will work...
<slapin_> mnemoc: thanks, will think about all this...
Gujs has quit [Ping timeout: 268 seconds]
Gujs has joined #arm-netbook
<hno> slapin, stty cols NN rows NN
<mnemoc> hno: the doubt is why `stty` sees 80x24 but `resize` is able to get real size of your terminal
<mnemoc> using microcom inside a term
<hno> resize sends an escape code querying the terminal program on what window size it has and then sets the tty. stty sets the tty.
<mnemoc> aha!
<mnemoc> write(3, "\0337\33[r\33[999;999H\33[6n", 19); read(3, "\33[20;156R", 4096); write(3, "\0338", 2)
<mnemoc> so that is the magic part
<mnemoc> and then the ioctl to set it
<rz2k> uboot-allwinner/sunxi-current
<mnemoc> hno: and http://dpaste.com/828028/plain/ from the A10s' SDK
<hno> rz2k, yes, those need to be fixed.
slapin has quit [Ping timeout: 276 seconds]
<hno> mnemoc, interesting, so there is both A12 and A10s.
<mnemoc> yup
<hno> and to confuse more A10s is a sun5i, not usn4i.
<mnemoc> :)
slapin has joined #arm-netbook
jquip has quit [Read error: Connection reset by peer]
RITRedbeard has quit [Read error: Connection reset by peer]
RITRedbeard has joined #arm-netbook
ZaEarl_ has quit [Ping timeout: 260 seconds]
tzafrir_laptop has quit [Ping timeout: 248 seconds]
tzafrir_laptop has joined #arm-netbook
QingPei has quit [Ping timeout: 265 seconds]
<rz2k> where libcedarxalloc.a comes from?
<Turl> rz2k: allwinner
<rz2k> interesting
<rz2k> looks like there is libcedarxalloc for armhf, right?
jquip has joined #arm-netbook
* slapin_ hates when people change things they don't quite understand :(
<rz2k> s/is/is no/
<ibot> rz2k meant: looks like there is no libcedarxalloc for armhf, right?
<rz2k> or libcedarv.a = libcedarxalloc.a?
QingPei has joined #arm-netbook
<mnemoc> no, uses a different allocator now
<mnemoc> it's in the sample code iirc
vinifm has joined #arm-netbook
<mnemoc> so one less piece of closes software
<mnemoc> closed*
<rz2k> awesome, so wills wrote his vlc decoder based on old software :/
avernos has quit [Remote host closed the connection]
QingPei has left #arm-netbook [#arm-netbook]
cat_x301 has joined #arm-netbook
<slapin_> hno: which branch it is best to base NAND work on, so to not have Allwinner NAND driver stuff floating around, and have u-boot booting from mmc?
* slapin_ hates green leds, this one is extremely annoying, need to deactivate this one somehow, a person, which invented testing by LEDs had sick sence of taste :(
<slapin_> have anybody tried using koen's narcissus to make OE images locally?
xxiao has quit [Remote host closed the connection]
<Turl> slapin_: `sunxi' branch I'm guessing
<Turl> slapin_: blue and red leds are even more annoying :)
<WarheadsSE> especially REALLY bright blue LEDs
* WarheadsSE looks at Mele
<Turl> and at night
<Turl> >.<
<Turl> you come into the room and there's all these bright lights
<orly_owl> put some tape over it
<WarheadsSE> at least you can tone them down a bit with some settings
<Turl> at least they help you not hit them by accident
<Workboot> Anyone seen an A15 in action, yet?
* Workboot is torn between Nit6x and Arndale.
<Workboot> I have blu-tack stuck into the recesses where the blue LEDs are in one of my headphone amps. It's spread just thin enough that I can see the light without being blinded by it.
hg_5 has joined #arm-netbook
vinifm has quit [Quit: Ex-Chat]
<hg_5> hello, what tablet you prefer ,7 inch with dual core and supports hspa+ 900mhz
<mnemoc> n7 :p
<hg_5> n7 is?
<hg_5> i will be use this tablet in car also
<hg_5> to navigation
<hg_5> and sometimes in home
<hg_5> so you are sure it supports hspa+ 900mhz ?
<mnemoc> NETWORK
<mnemoc> Unlocked GSM/UMTS/HSPA+
<mnemoc> 3G (850, 900, 1700, 1900, 2100 MHz)
<mnemoc> GSM/EDGE/GPRS (850, 900, 1800, 1900 MHz)
<mnemoc> HSPA+ 21
<hg_5> there are 2 versions of nexus 7 with 3g and without 3g?
<mnemoc> yes
<mnemoc> also variants of the wifi-only with different storage. but this is kind of way off topic
<hg_5> hm do you know any cheaper?
<hg_5> version with 3g is way too overpriced
<mnemoc> cheap tablets usually don't have 3g
slash_random has joined #arm-netbook
<Turl> 3G modems cost quite a bit :)
* slapin_ found goot metal duct tape to cover the damn thing and now sane again
<slapin_> s/goot/good/
<slapin_> ibot: ?
cnxsoft has quit [Quit: Leaving]
lkcl has quit [Ping timeout: 264 seconds]
<hg_5> hm or i could go cheap 7 tablet + mifi 3g modem
<hg_5> but what tablet would you prefer then without 3g ?
<rm> mnemoc, one model in 100 has 3G
<rm> and chances are it's on an MTK chipset
<mnemoc> s/mnemoc/hg_4/
<rm> ah, you were talking about Nexus
<rm> I meant generally on the market
<libv> mali-libs is a real mess. r3p1/armhf has 10 copies of libMali.so
lkcl has joined #arm-netbook
<libv> but no worries, i am closer and closer to pushing a huge patchset out for it :)
<mnemoc> copies or symlinks?
<libv> copies.
Quarx has joined #arm-netbook
mSquare has joined #arm-netbook
<libv> patchset will clean all of it up, and add some Makefiles
<libv> make ABI=armhf VERSION=r3p0 x11
<hno> slapin, sunxi-current is completely void of any NAND driver. Only lichee-* branches have NAND driver.
<libv> on native hw, with mali modules loaded, you can just do make x11
<libv> with mali modules loaded, and gcc installed that is
<libv> will come with headers for framebuffer use as well, not tested on x11
<mnemoc> :D
<hno> but need to fix some things in -current to have it build again..
<libv> egl test will need to be ported to x11 too, but should only be a few lines
<Turl> libv: why is gcc needed? :P
<libv> Turl: -dumpmachine
<libv> Turl: i found no saner way to autodetect armel/armhf
<Turl> uname -a not explicit enough?
<libv> Turl: does armv7l signify hf?
<Turl> hm, nope, it means armv7l
<Turl> :P
<libv> Turl: or could that also happen on armel?
<Turl> yeah armel will print the same
<libv> Turl: so no, uname -a will not be useful in any way.
<libv> i could of course do a full autoconf/automake
<libv> or people could just provide ABI= at the make command line.
<libv> like the first line i pasted here
<Turl> libv: do you have an armel system handy?
<libv> same for the version, on native hw, with gcc, libc installed, and with the mali module loaded, we can autodetect it
<libv> Turl: nope
<Turl> IS_ARMHF=$(ls /lib/ld-linux*|grep -c armhf)
<Turl> works for me on armhf, haven't tried on armel :P
<libv> Turl: are you on a debian based installation, or on something else?
<Turl> libv: Arch Linux ARM
<libv> Turl: what does gcc -dumpmachine tell you?
<mnemoc> it's saner to just assume it's armhf unless explicitly told differently
<mnemoc> magic tends to go wrong
<libv> mnemoc: it's even saner to just give up
mSquare has quit [Ping timeout: 246 seconds]
<mnemoc> :(
<Turl> mnemoc: is the nand rootfs armel?
<libv> gcc gives out a nice tuple
cat_x301 has quit [Read error: Connection reset by peer]
<mnemoc> Turl: can't know what you put in there
rzk|2 has joined #arm-netbook
<Turl> mnemoc: whatever comes from tom
<Turl> I didn't touch nand
<libv> mnemoc: i do not want to overwrite things in /usr/lib/ on an x86 machine for instance
mSquare has joined #arm-netbook
rz2k has quit [Quit: ZNC - http://znc.in]
<Turl> libv: $ gcc -dumpmachine
<Turl> -bash: gcc: command not found
<Turl> :)
<libv> Turl: ABI=armhf for you then
<Turl> libv: can you add IS_ARMHF=$(ls /lib/ld-linux*|grep -c armhf) as fallback magic when there's no gcc?
<libv> i think so, but i will do so as a separate patch and give you credit for that one
<Turl> fair enough
<libv> now i am cleaning up the tree and providing Makefiles for each version
rzk|2 is now known as rz2k
<libv> mnemoc: i am not too happy with the name mali-libs though, we are more and more converging towards sunxi-mali-driver or so
<libv> since it now will do installation and includes a test utility
mSquare has quit [Ping timeout: 255 seconds]
<libv> this is what the top level looks like with me now: include lib Makefile README test version
<libv> with lib having the different mali versions.
<rz2k> libv: you have the right to rename repos
<rz2k> disp team at linux-sunxi can manage cedarx-libs, mali-libs and x11 stuff
<libv> ah, ok
<libv> rz2k: so you have x11 set up fully with mali?
<rz2k> with drivers from ARM yeah
<rz2k> libs from Tom
<rz2k> check Mali400 at wiki
<libv> ok, once the patches go out, with the gl headers, could you give the hello triangle test a try under x11
<rz2k> yes, ofc
mSquare has joined #arm-netbook
<libv> it will need to get its native window from X/DRI/whatever, and that bit needs to be built conditionally
rellla has quit [Quit: Nettalk6 - www.ntalk.de]
mSquare has quit [Ping timeout: 255 seconds]
<hno> slapin, sunxi-current should build fine again.
kop has quit [Ping timeout: 260 seconds]
pawel5870 has quit [Ping timeout: 245 seconds]
sspiff has quit [Ping timeout: 245 seconds]
arnd_ has quit [Ping timeout: 244 seconds]
kop has joined #arm-netbook
<WarheadsSE> rz2k: got them happy on Arch?
Scepterr has quit [Remote host closed the connection]
<rz2k> WarheadsSE: nope, segfaults from libfb.so
<techn__> libfb.so?
<rz2k> that is fbdev in 1.13.x
<rz2k> our driver relies on fbdev or libfb for direct output to framebuffer
<rz2k> check the xorg logs, it always loads fbdev
<rz2k> in 1.13 after first initialization (DRI2 stuff, getting resolutions of the screen from mali_lcd) it tries to access framebuffer and segfaults in libfb.so
<WarheadsSE> yeah, kinda what I expected, since the api change
<rz2k> before that for some unknown reason I had message from libdricore9.9.9 (strange name) that _glapi_* stuff is missing, even with libglapi present in system
<rz2k> only way to fix that is track down the fbdev changes and fix the driver?
<rz2k> or documentation for X is already updated?
gimli has joined #arm-netbook
<WarheadsSE> That's pretty much it. Have to patch for the differences.
<rz2k> also interesting thing that I've got from Arch is that 1.13.x fbdev and 1.13 xserver is much, much faster than linaro's default ones
<WarheadsSE> :) good to hear
<rz2k> I had no stuttering moving windows and such 2D stuff, like if it was accelerated
<WarheadsSE> I had some issues very early on, but not much
<WarheadsSE> that was while it was 1.12 as well
<WarheadsSE> I never quite got the mali drivers working with it
<WarheadsSE> And when I did, I got some very odd video output.
pawel5870 has joined #arm-netbook
vinifm has joined #arm-netbook
pawel5870 has quit [Ping timeout: 260 seconds]
rellla has joined #arm-netbook
<libv> now for patchset cleanup :)
<Turl> WarheadsSE: AL(ARM?) servers are unusually slow today, anything going on? :P
<Turl> barely getting stuff at 20K/s
<Turl> WarheadsSE: fuu, crond went again and started eating all my cpu :(
<rz2k> yeah, I was frustrated by cron renewing man-db each day too.
<Turl> rz2k: yeah but it's not man-db
<Turl> it's crond itself eating it all :(
<WarheadsSE> i think thats a check upsteam
<WarheadsSE> server still slow, or is it crond
<Turl> still slow
<Turl> upgrading crashed or something and now arch doesn't even let me log in and loading libs fails with errors :(
<Turl> time to reinstall I guess
<Turl> WarheadsSE: got that updated systemd-able rootfs up yet? :)
<WarheadsSE> oh, no,
<WarheadsSE> the conversion is simple though
<Turl> yeah but the upgrade crashed :P
<WarheadsSE> roops
<WarheadsSE> did you bork glibc?
<Turl> yeah
<Turl> on boot I see "/bin/sh: error while loading shared libraries: /usr/lib/libgcc_s.so.1: file too short"
<WarheadsSE> Ah
<WarheadsSE> ouch
<RaYmAn> It seems kind of wrong that an app in /bin should depend on /usr..But I guess I'm thinking of the old way of doing linux.
<Turl> I think /bin doesn't exist anymore RaYmAn
<Turl> RaYmAn: it's a symlink to /usr/bin iirc
<Turl> /lib -> usr/lib /lib64 -> usr/lib64 /sbin -> usr/sbin /bin -> usr/bin
<WarheadsSE> ^
<Turl> RaYmAn: https://fedoraproject.org/wiki/Features/UsrMove#You_are_doing_it_wrong.21_.2Fbin_and_.2Fsbin_are_there_to_rescue_a_broken_.2Fusr.21
<RaYmAn> erk.
<Turl> in any case, if stuff explodes and your system doesn't work you're way more likely to fix it from a working system on a liveusb/cd/dvd/by netbooting
<RaYmAn> right, that kind of argument :P
<jinzo> Turl, depends if it's still running or not
<jinzo> or if it has to keep running
pawel5870 has joined #arm-netbook
MMlosh has joined #arm-netbook
pawel5870 has quit [Ping timeout: 252 seconds]
lerc has quit [Remote host closed the connection]
rellla has quit [Remote host closed the connection]
Quarx has quit [Ping timeout: 252 seconds]
pawel5870 has joined #arm-netbook
arnd_ has joined #arm-netbook
pawel5870 has quit [Ping timeout: 276 seconds]
<vgrade> libv hi
<vgrade> what are you brewing here?
cheng has joined #arm-netbook
slash_random has quit [Ping timeout: 240 seconds]
<rm> what's with all those people building kernels from a Versatile .config
bsdfox has joined #arm-netbook
bsdfox has quit [Changing host]
bsdfox has joined #arm-netbook
<mnemoc> doh, didn't notice that detail
<mnemoc> he said that he was using sun4i_defconfig in the first most :|
pawel5870 has joined #arm-netbook
slash_random has joined #arm-netbook
hg_5_ has joined #arm-netbook
<rm> ah I remember where I saw this before https://github.com/linux-sunxi/linux-sunxi/issues/96
hg_5_ has quit [Client Quit]
<rm> and it actually was the same person
<rm> > arch/arm/mach-versatile/built-in.o: In function `versatile_init_early':
<Turl> what's mach-versatile? o.O
<RaYmAn> it's an ARM dev board (from ARM)
<rm> it's what you end up compiling your kernel for, if you derp up your .config
<rm> (apparently.)
hg_5 has quit [Ping timeout: 252 seconds]
tzafrir_laptop has quit [Ping timeout: 248 seconds]
jquip has quit [Ping timeout: 268 seconds]
<traeak> so where's the aarch64 procs? (hehe)
<Turl> traeak: I have one here in my desk crunching some numbers
<rm> > the Cortex-A53 processor will efficiently run your connected life.
ZaEarl_ has joined #arm-netbook
slapin has quit [Ping timeout: 276 seconds]
* rm misread "will effectively ruin your connected life."
<Turl> rm hahaha
<traeak> Turl: gibbme :-p
<Turl> traeak: :P
<traeak> was looking at some of that yesterday, a tad
rsalveti_ has joined #arm-netbook
<rm> the worst part about the new architecture is the official name
<jelly-home> big.LITTLE?
<rm> I wish they just went with "arm64"
<jelly-home> aarm64?
<rm> no the Aaaaaaaaarch
<traeak> hehe
<jelly-home> aaha.
<jelly-home> they probably had a Dutchie person on board
<Turl> hopefully userspace is arm64 or something saner :)
pawel5870 has quit [Ping timeout: 240 seconds]
<rm> Debian went with "arm64" afaik
rsalveti has quit [Ping timeout: 246 seconds]
rsalveti_ is now known as rsalveti
<mnemoc> the kernel too. upstrea didn't accept the aarch64 crap
<rm> but now there is this confusion and a mess, forever
<mnemoc> arm64 and full stop
<rm> "what do I use as ARCH= in this particular tool"
<rm> aarch64 or arm64
<mnemoc> :)
<Turl> ARCH=native
<Turl> :P
<WarheadsSE> :p
<mnemoc> o.o
<WarheadsSE> le fuw
<Turl> cool stuff :O
<Turl> NetBSD 6.99.14 (RUMP_ROAST) #0: Sat Nov 3 18:28:54 CET 2012 pooka@T61:/home/pooka/src/nbsd/src/sys/rump/librump/rumpkern
jquip has joined #arm-netbook
eFfeM has quit [Quit: Leaving.]
cheng has quit [Quit: Leaving]
<Turl> rm russian email on cubieboard list :P
<rm> that must be from the sunxi git
<mnemoc> eh?
<rm> or... what do you mean
<mnemoc> what's the relation between a russian asking for exclusivity distributing the cubieboard in .ru with the sunxi git?
<rm> I didn't write anything to the cubieboard list, so if you see anything from me, then it must be cc'ed by Github, etc
<Turl> rm a russian company mailed cb list about selling cb on russia
<rm> oh I see now
<rm> mnemoc, I am not on that list actually
<rm> should subscribe
<mnemoc> indeed
<RaYmAn> it sounds so much like spam mail, lol
<Turl> yeah, but it isn't :P
<RaYmAn> I know it's probably just english issues, but :P
<Turl> I'm surprised they left the subject line empty though
<techn__> I wonder why he send that offer to ML :p
<mnemoc> that tells you a bit about their capabilities :<
<rm> well what I can say, "hopefully not"
Gujs has quit [Ping timeout: 276 seconds]
<rm> to the exclusivity part
<mnemoc> unless they are willing to pre-pay for a bunch 10k+ pieces I see no reason why would tom grant exclusivity to anyone
<jelly-home> indeed
<mnemoc> rz2k :)
<rz2k> mnemoc: 50% chances for another reseller who wants monopoly
Gujs has joined #arm-netbook
<jelly-home> smells scammy
slapin has joined #arm-netbook
<mnemoc> slapin: found this in the A10s SDK ... the "boot0" part in the path seems interesting/weird http://sprunge.us/BcGX?diff
<mnemoc> looks like a simplified driver
<mnemoc> interestingly they also have a "libnand.a" in modules/
<rz2k> vlc works somewhat okay on armel
<jelly-home> in X?
<rz2k> I was able to view some dvd rip movies that I've found on external hdd
<rz2k> without stuttering and etc
<rz2k> jelly-home: yes, with patches from wills adding cedar support
<Turl> lundman: ping, can you do copy-on-write on zfs?
<Turl> rz2k: what's the cpu usage for vlc when governor=performance?
<rz2k> 60% in without upscale and 100% with upscale to fullhd (thanks to mali400 drivers that cant do 2D), still working much faster than libavcodec
<rz2k> I might compile something wrong by the way
<rz2k> when wills will fix his libcedarx to armhf, then I will do normal benchmarks
bsdfox has quit [Ping timeout: 240 seconds]
RITRedbeard_ has quit [Ping timeout: 240 seconds]
jquip has quit [Read error: Connection reset by peer]
slapin has quit [Ping timeout: 276 seconds]
tinti has quit [Quit: Leaving]
pawel5870 has joined #arm-netbook
bsdfox has joined #arm-netbook
bsdfox has quit [Changing host]
bsdfox has joined #arm-netbook
<techn__> rz2k: I think vlc still uses libavcodec
<techn__> like most/all players
<techn__> hmm.. rz2k is there possibility to use qt as output.. it should do accelerated gles scaling/csc
<libv> vgrade: getting a proper linux system going for lima work, what does it look like :)
<libv> what else does it look like even
techn__ is now known as techn
lerc has joined #arm-netbook
pawel5870 has quit [Ping timeout: 240 seconds]
Sternennebel has joined #arm-netbook
<techn> rz2k: yep.. I saw those.. Still wondering why no one has done libva plugin for cedarx :/
focus_well has quit [Ping timeout: 246 seconds]
<techn> would save a lot of work
<bsdfox> if I want to test out some code amery posted to https://github.com/linux-sunxi/linux-sunxi/commit/09946bae6c414e62c066c57169100c62cd303500 how would I apply that patch with git?
<bsdfox> git cherry-pick 09946bae6 doesn't work
<techn> also if some one could verify that qt's or sdl's gles2 acceleration works.. then we would have hw accelerated scaling and csc :/
vinifm has quit [Quit: Saindo]
<bsdfox> nm I think git pull allowed me to do that
<rz2k> techn: qt's acceleration is tied to openvg
<rz2k> we had libs with openvg, but they were missing symbols
<techn> rz2k: there is nowaday gles2 acceleration also
<techn> I couldn't find csc matrixe's.. but they atleast were there
<rz2k> I dont think linaro's qt is updated though.
freakazoid0223 has quit [Read error: Connection reset by peer]
<jlj> rz2k: is will's code interface code to a blob or has he been ale to reverse engineer cedarx ?
<rz2k> he used cedarx-libs for armel
<rz2k> reverse engeneered stuff died this summer, not enough people capable of doing that.
<libv> s/capable of/steadfast enough/
<libv> +ffor
<jlj> rz2k: okay, does it seem to work well?
<rz2k> seems to work fine, I need to make it work on armhf or wait for wills to update his work.
<jlj> rz2k: cool
RITRedbeard has quit [Read error: Connection reset by peer]
RITRedbeard has joined #arm-netbook
Almamuetya11 has quit [Ping timeout: 260 seconds]
benjamin__ has joined #arm-netbook
vinifm has joined #arm-netbook
<vinifm> is there a uart driver to a13?
<mnemoc> sure
<mnemoc> it's an 8250
focus_it has joined #arm-netbook
<slapin_> hno: is basing on sunxi branch of u-boot ok atm?
* slapin_ wants answers
<slapin_> do anybody knows of some hackable tablet with GSM/3G and GPS?
<slapin_> GNSS is cool but not really mandatory
<mnemoc> slapin_: saw the diff? it's an small raw nand driver with dma and dram initialization
<rz2k> slapin_: allwinner tablet + external bluetooth usb with gps attached + usb gsm modem
<rz2k> also renesas ev2 tablets that I've hacked recently have kernel sources and they are available with both GSM and GPS embedded.
<rz2k> but they use old processor and nothing above 2.3 android is available.
<slapin_> rz2k: no strigs should be attached, all should be inside, since this will be used by untrained peronell...
<rz2k> (gsm is on mini pci-e, gps is on uart inside )
<slapin_> rz2k: any links? we don't care for android version at the moment
<rz2k> sorry I have to leave for an hour, will post later.
<slapin_> rz2k: thanks
<rz2k> google renesas emma ev2
graffiti has joined #arm-netbook
<slapin_> mnemoc: which diff?
<mnemoc> slapin_: http://sprunge.us/BcGX?diff <--- the boot0 part
<RaYmAn> partial boot0 sourcecode?
<mnemoc> looks like so
<mnemoc> it was in the sun5i_nand driver in the A10s SDK
<slapin_> what is difference between AmLogic and AllWinner SoCs? they all reference Softwinner...
<mnemoc> AmLogic references softwinner???
<slapin_> mnemoc: yes :/
<mnemoc> softwinner is part of allwinner..... amlogic is a totally different company
<mnemoc> but they might have bought some IP from them...
<rm> softwinner is a schizophrenic dual personality of allwinner :p
<rm> or "we randomly won't use the real company name here, because I dunno why"
<vinifm> mnemoc: 8250_sunxi.c?
<mnemoc> vinifm: yes
<vinifm> by Russel King, cool :)
<mnemoc> nah. they just take other drivers as template
<slapin_> wim: it is copied from somewhere
tinti has joined #arm-netbook
tinti has quit [Client Quit]
<vinifm> this can be done with gpio_lib?
<mnemoc> there is no gpiolib support for sunxi
<mnemoc> and by the time we have gpiolib we won't need a sunxi uart driver
<mnemoc> because there is a generic 8250 for DT
focus_well has joined #arm-netbook
<vinifm> i mean, take a gpiolib template and adapt
<slapin_> gpiolib drivers are very easy to implement
rellla has joined #arm-netbook
graffiti has quit [Ping timeout: 245 seconds]
<rz2k> slapin_: https://github.com/Renesas-EMEV2 board photo (I have version without gsm/gps) http://cs406126.userapi.com/v406126573/2833/L5RFXY2KGCM.jpg main problem is u-boot and kernel dying when compiled with anything newer than gcc 4.3 and tons of other little issues, like end of life of the processor
<Turl> rz2k: that's a huge book on C++ :P
<rz2k> thats stroustroup
<rz2k> :p
<rz2k> slapin_: amlogic have closed u-boot afaik, so no luck there.
<rz2k> also softwinner is allwinner before they've produced sun3i.
<rz2k> so they've might been working for other chinese companies
<mnemoc> don't forget boxchip :)
<Turl> softwinner + hardware = allwinner
<rz2k> yeah
<Turl> makes sense :P
freakazoid0223 has joined #arm-netbook
<mnemoc> Turl: the cubie has a power button and a fel button
<mnemoc> Turl: if you power it off without unplugging then you turn it on pressing the button
<mnemoc> or hard power-off by pressing the power button 10s
<Turl> where is this button? o.O I'm not seeing it still
<Turl> I only see one under microUSB
<Turl> libv: that 'clean up binaries' is a rather... huge patch >.<
rz2k has quit [Read error: Connection reset by peer]
<libv> Turl: someone messed up with symlinks and stuck 10 copies of libMali.so in there
tzafrir_laptop has joined #arm-netbook
<libv> now symlinks are created by the Makefiles
<slapin_> mnemoc, hno: do you know why spl is not built by my new u-boot checkout?
<slapin_> make ARCH=arm CROSS_COMPILE=arm-angstrom-linux-gnueabi- sun4i_config
<slapin_> make ARCH=arm CROSS_COMPILE=arm-angstrom-linux-gnueabi- all
<slapin_> I play with mk802 clone of clone w/o axp209, do need to set something somewhere in boards.cfg?
<mnemoc> slapin_: no idea what are you building there
<mnemoc> it doesn't look like our u-boot
vinifm has quit [Quit: Saindo]
<slapin_> mnemoc: I use github.com/hno/uboot-allwinner repo
<mnemoc> what branch? anyhow, it would be `make ARCH=arm CROSS_COMPILE=arm-angstrom-linux-gnueabi- yourboard`
<mnemoc> if it's 512M try using mele_a1000
<mnemoc> there is no support for mk802
<slapin_> mnemoc: u-boot README tells use board_config and it works fine, I use master branch and I had to add SPL to define flags in board.cfg and it works.
<slapin_> mnemoc: it seems I can use sunxi branch for stable work, and master is bleeding edge... or so
<slapin_> hno: ping
tuliom has joined #arm-netbook
<slapin_> mnemoc: I see, DRAM initialization is added to SPL, and mk802 is not supported, so I need to hack some SDRAM config data...
<mnemoc> slapin_: the mele a1000 dram config is very likely to match your mk802
<mnemoc> slapin_: sunxi-current, not master
<mnemoc> slapin_: fexc (from sunxi-tools) can generate a dram.c for you from your script.bin
<mnemoc> those old devices had complete [dram_para] there. newer devices rely in livesuit capability of probing the dram
<slapin_> probing thing is weird...
ol1ver has quit [Quit: ol1ver]
<slapin_> my board has 2 chips hence it seems 32bit interface...
<oliv3r> speaking of a1000; is there any reason to prefer a a2000? gonna order a a1000 next week from cubies shop
<mnemoc> oliv3r: nicer case :)
<oliv3r> 10 USD price difference for a case? then i'll go for the a1000 :p
<mnemoc> oliv3r: there are 1GB variants of the A1000/A2000 .... ask him if he can get them
<oliv3r> ah, ok deffoo
<oliv3r> 1GB over case :p
<oliv3r> brb
<rellla> i have a a2000 and a100. a100 case imho is much cooler. if you do not need sata
<oliv3r> a100 or a1000?
<rellla> a100
<lundman> turl: it does copy on write, yes
<mnemoc> a100 is sata-less
oliv3r has quit [Quit: Leaving.]
<mnemoc> same pcb but that connector is not there
<rellla> a2000 will get server, a100 near tv ;-)
<slapin_> not having sata is pity
<mnemoc> it's a cheaper alternative for people who doesn't want sata
<mnemoc> mele really makes knows to reuse a single pcb :p
oliv3r has joined #arm-netbook
<mnemoc> s/makes//
<ibot> mnemoc meant: mele really knows to reuse a single pcb :p
<slapin_> any links to Cubie shop?
<slapin_> mnemoc: thanks!
<mnemoc> tw
<slapin_> latest u-boot seems to work now, how patches are generally submitted? via github pull requests or something?
<mnemoc> linux-sunxi ml
<mnemoc> but hno takes pull requests too
<oliv3r> mailin hipboi now about 1GB mele :)
arnd_ has quit [Ping timeout: 244 seconds]
<libv> rpi people are complete tossers.
* libv saw the commit history of the quake3 rpi port
<slapin_> libv: ?!
<libv> git init, cp, git commit
<libv> not git svn clone
<slapin_> ah
<slapin_> git svn clone is so slow...
<libv> it's automatic.
<Turl> mnemoc: import complete ;)
<libv> you can do something else for 20 minutes
<mnemoc> Turl: and you haven't see the A10S SDK code yet :p
<slapin_> do anybody knows how to solder external antenna to dreamplug? wifi is weird here now...
<Turl> libv: these might interest you "video: sun4i: import changes"
<libv> Turl: for the display code, it seems that they stuck a lot of the sun5i changes in the sun4i tree
<Turl> and "includes: disp: update headers"
<mnemoc> I have sun5i disp updates in the A10s SDK too
<mnemoc> hopefully they are consistent with yours
<mnemoc> Turl: not sure if we want to reserve memory for the GPS before we have the sources for a driver
<Turl> mnemoc: they fail at reserving it anyway :)
<libv> Turl: disp changes seem like an easy merge job
<Turl> mnemoc: just linked it so you can laugh :P
<libv> can the a10s arch be identified in some way?
<mnemoc> libv: I assume the chip-id is in the BROM as with the A10 and A13
<libv> is this a different mach?
<mnemoc> sun5i, same as A13
mikey_w has quit [Ping timeout: 246 seconds]
<mnemoc> but with hdmi and ethernet
mikey_w has joined #arm-netbook
<libv> but the disp changes are to the sun4i code
<mnemoc> libv: they have a function to detect the differnt sun5i (6 chips), but it's inside a built-in.o
<mnemoc> libv: Turl imported a newer SDK for sun4i, yes
<libv> objdump -d
<mnemoc> yes, that will be next after importing the diffs in code
rellla has quit [Quit: rellla]
<mnemoc> append .patch to see it as plain text
ccssnet has quit [Ping timeout: 245 seconds]
<slapin_> hno: ping
rmeyerriecks has joined #arm-netbook
<libv> rmeyerriecks: one of the prerequisites is, you probably should be using an allwinner based device, and it really helps if you can attach serial
<libv> with other mali based socs you are pretty much on your own
<mnemoc> slapin_: want you give me your script.fex for the bsp?
<mnemoc> s/want/can/
<ibot> mnemoc meant: slapin_: can you give me your script.fex for the bsp?
gimli has quit [Quit: Verlassend]
tinti has joined #arm-netbook