ChanServ changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - Don't ask to ask. Just ask! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi
Tartarus has quit [Read error: Connection reset by peer]
tinti has joined #linux-sunxi
rz2k has quit []
rellla has quit [Quit: Nettalk6 - www.ntalk.de]
delusr_ has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
anunnaki has quit [Remote host closed the connection]
christopher has joined #linux-sunxi
torqu3e has quit [Quit: torqu3e]
christopher is now known as anunnaki
hipboi|cubie has quit [Ping timeout: 256 seconds]
hipboi|cubie has joined #linux-sunxi
servili007 has joined #linux-sunxi
BJfreeman has quit [Quit: had a good time]
delusr_ has quit [Quit: ChatZilla 0.9.90 [Firefox 20.0/20130329030848]]
anunnaki has quit [Remote host closed the connection]
christopher has joined #linux-sunxi
christopher is now known as anunnaki
anunnaki has quit [Remote host closed the connection]
<Turl> hardware md5 is 20x slower than doing it on the cpu
<Turl> I suppose DMA would speed it up considerably, sending 32 bits at a time via a fifo is time consuming
drachensun has quit [*.net *.split]
zumbi has quit [*.net *.split]
zumbi has joined #linux-sunxi
zumbi is now known as Guest43339
Yaku321 has joined #linux-sunxi
drachensun has joined #linux-sunxi
servili007 has quit [Read error: Connection reset by peer]
vakkov has joined #linux-sunxi
Quarx has joined #linux-sunxi
rellla has joined #linux-sunxi
paulk-desktop has joined #linux-sunxi
shineworld has joined #linux-sunxi
Quarx has quit []
mdfe has joined #linux-sunxi
hipboi has joined #linux-sunxi
_BJFreeman has joined #linux-sunxi
_BJFreeman is now known as BJfreeman
shineworld has quit [Quit: Leaving]
delusr_ has joined #linux-sunxi
n01 has joined #linux-sunxi
rellla has quit [Remote host closed the connection]
rellla has joined #linux-sunxi
BJfreeman has quit [Quit: had a good time]
<Yaku321> are there applications that can benefit through hardwareaccelleration somehow like cuda, just on the mali in this case ?
<Yaku321> i am not talking of mediaplayers :P
<Yaku321> if i use the a10 headless i would have the gpu free, so letting the mali run for seti@home and alike
<oliv3r> Yaku321: I don't think mali is that far yet, but I guess eventually lima may have some sortof work in this, but ibet you'll have to write it
<oliv3r> in theory of course, whynot
<lunra> The Mali 400 on the A10 processor seems to only have one core and is clocked at 320MHz, it is questionable as to whether it would be worth the effort
ganbold___ has quit [Remote host closed the connection]
<hramrach> it will chew the stuff, just slow
<hramrach> not sure how the units are scheduled and if he schedule does not time out before mali would proces it
sanka has joined #linux-sunxi
shineworld has joined #linux-sunxi
vakkov has quit [Ping timeout: 258 seconds]
hipboi has quit [Quit: Leaving]
vakkov has joined #linux-sunxi
tinti has quit [Quit: Leaving]
n01 has quit [Ping timeout: 240 seconds]
Yakuzza has joined #linux-sunxi
Yaku321 has quit [Ping timeout: 246 seconds]
ganbold_ has joined #linux-sunxi
Undertasker has joined #linux-sunxi
hipboi|cubie has quit [Ping timeout: 256 seconds]
hipboi|cubie has joined #linux-sunxi
Yakuzza is now known as Yaku321
Yaku321 has quit [Ping timeout: 256 seconds]
Yaku321 has joined #linux-sunxi
mdfe has quit [Remote host closed the connection]
<rellla> hi all
<rellla> hdmi.ko aren't installed with modules_install here with sunxi-3.0
<rellla> anyone else with his issue?
<mnemoc> check .config
<ssvb> rellla: maybe it is statically compiled
<mnemoc> it can be disabled or built-in
<rellla> CONFIG_LYCHEE_HDMI_SUN4I=y
gzamboni has quit [Ping timeout: 260 seconds]
<Turl> they've all been made builting recently
servili007 has joined #linux-sunxi
<rellla> ok. found it in menuconfig. then i should test it ;)
_BJFreeman has joined #linux-sunxi
_BJFreeman is now known as BJfreeman
<Turl> s/builting/built-in/
<Turl> that happens when you're not fully awake :)
Yaku321 has quit [Ping timeout: 258 seconds]
rellla2 has joined #linux-sunxi
<rellla> so, it's working. xbmc runs on it, like it was never changed. may i have some fear to try stage/sunxi-3.0?
Yaku321 has joined #linux-sunxi
rellla2 has quit [Remote host closed the connection]
rz2k has joined #linux-sunxi
<servili007> Anyone had any issues making recent linaro builds on newer toolchains, or is it just me
<servili007> Rather, everything goes fine on the build side, just getting x crashes
tinti has joined #linux-sunxi
ganbold__ has joined #linux-sunxi
ganbold_ has quit [Remote host closed the connection]
<slapin_nb> hi, all!
<slapin_nb> hipboi|cubie: hi! are you here?
Yaku321 has quit [Ping timeout: 276 seconds]
servili007_ has joined #linux-sunxi
servili007 has quit [Ping timeout: 258 seconds]
servili007_ is now known as servili007
<ssvb> rellla: in this comment willswang says that VC1 and RMVB don't work - https://github.com/willswang/libcedarx/pull/1#issuecomment-14989658
<ssvb> I have added a link to this sample to http://linux-sunxi.org/CedarXVideoRenderingChart
Yaku321 has joined #linux-sunxi
<Turl> servili007: I've had segfaulting init (pid 1) when using linaro on buildroot, I'd skip their toolchains if you're having trouble
<Turl> sigill (illegal instruction) rather than segfault
<servili007> Turl: roger that
<shineworld> /cpu/procinfo Serial is : 0000000000000000 only for 3.0 or was implemented in last kernel ?
<rellla> after 2 hours letting xbmc be itself, i come back and screen is black :( gui-sounds are hearable.
<rellla> restarting xbmc doesn't solve the problem. is there a kind of power managment via hdmi?
<rellla> ...rebooting...
<rellla> ... and xbmc-gui is back on top. mow testing your sample, ssvb
<rellla> same issues here, ssvb. artifacts - mainly if screen is bright.
<rellla> and hramrach's sample crashes, or it's very short ;)
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
<servili007> Getting the same thing - Plymouthd reporting 'Assertion `terminal != ((void *)0)' failed.'
<ssvb> rellla: thanks for testing and adding the results to the wiki page, looks like xbmc and vlc mostly have the same issues
<ssvb> rellla: I'm trying to play with x264 encoding settings to see if there is an obvious culprit
<rellla> yw. i'm investigating why my screen gets black all the time *grmbl*
<rellla> i'll read some code tonight. maybe there are some differences in implementation...
<ssvb> this video plays just fine in android
<ssvb> I wonder if the problem could be related to handling multiple reference frames
<rellla> am i wrong, when i say, that the linux- and android-cedarx libs base on the same source?
<ssvb> I optimistically assume that they had no reasons to intentionally cripple the linux libs
<rellla> so if some file works on android, the binary-lib in linux shouldn't be the problem!? being an optimist.
<ssvb> there are many h.264/avc profiles with different features as can be seen in the table there - http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Profiles
<ssvb> it would be interesting to check which of them might be causing problems
<rellla> lol
<rellla> i opened same table ;)
<rellla> same sample with different profiles baseline/main/high would be helpful
<ssvb> yeah, that's what I'm trying to do
<ssvb> also the section "Decoded picture buffering" explains that the decoder may use multiple previously decoded frames for prediction of the future frames
<ssvb> have a look at "Figure 9 Architecture of FBM module" in https://github.com/linux-sunxi/cedarx-libs/blob/master/doc/Libve%20User%20Guide-2011.7.6.pdf
<ssvb> it shows the transition of the frame status, and I guess that the misuse of cedarx lib API could potentially mess it up
n01 has joined #linux-sunxi
<ssvb> anything else is most likely out of our control and is real cedarx blob issue :)
gzamboni has joined #linux-sunxi
<hramrach> rellla: the sample is about 5s
<hramrach> you should check logs to see if it crashes something
<rellla> same sample with different profiles baseline/main/high would be helpful
<rellla> sry
Undertasker has left #linux-sunxi [#linux-sunxi]
<rellla> i will, after i figured out, why me screen gets blank. i think, it has something to do with the kernel...
<hramrach> rellla: there 3 Y shaped colums that come out of the left and don't even come into view in vlc
<hramrach> something like a stopped wind power plant
<rellla> not so here.
<hramrach> if you can't see that you probably have the FPE
<rellla> i can see the movie, it stops and goes back to xbmc-gui
<hramrach> that's what it's supposed to do?
<rellla> yeah, but as it was so short, it looked like xbmc stopped playback
ganbold__ has quit [Remote host closed the connection]
ganbold__ has joined #linux-sunxi
Yaku321 has quit []
<ssvb> rellla, hramrach: it's a weightp=2 issue, I have added some information to http://linux-sunxi.org/CedarXVideoRenderingChart
<ssvb> so indeed we have a broken blob in linux :(
<ssvb> hramrach: about your sample which causes crashes, I probably have missed a link to it
<ssvb> thanks
<rellla> hramrach, ssvb have you both got compiled vlc successfully?
<rellla> i get a libmad error when trying to crosscompile
<hramrach> rellla: I built natively
<hramrach> crosscompiling something with this many dependencies ..
<hramrach> ssvb: who do you tell that something has weightp=2 ?
<ssvb> rellla: I compiled vlc natively, having a fast dual-core ARM Cortex-A15 device sharing the same NFS root makes me lazy :)
<rellla> time to burn my sd-card ;)
<rellla> or take some time to jump into flags and makefiles ...
<ssvb> hramrach: you can get this information using mediainfo - https://gist.github.com/ssvb/5427109
<ssvb> hramrach: because x264 is kind enough to provide some metadata string with all these options (but it might be not true for the other encoders)
<hramrach> rellla: you can do emulated build, too
<rellla> sure. yeah i'll chroot into it ;)
<rellla> too late. too sad. too bad news from ssvb.
<hramrach> seems vlc gets more prone to that issue over time but some videos cause it more than others
<hramrach> and it has weightp=2 so it's probably no surprice it has artiacts
<hramrach> this sub group is a gold mine. every single encode I try is broken ;-)
Undertasker has joined #linux-sunxi
shineworld has quit [Quit: Leaving]
rz2k has quit []
<ssvb> hramrach: maybe I need to figure out how to snatch the libve.so blob from the android rootfs in nand and check if it can be loaded in linux
<ssvb> http://git.videolan.org/?p=x264.git;a=commit;h=87de2346225721e8ca68a1b59bc87133fc598a42 suggests that the artifacts are likely to be triggered by fade-in or fade-out scenes
<hramrach> which the video I posted has loads
<hramrach> but also cutscenes seem to trigger the artifacts which is not fun
<ssvb> maybe somebody needs to politely contact allwinner and ask for an updated blob?
<ssvb> but for the start even properly working vp8/webm acceleration is good enough for me :)
vakkov has quit [Ping timeout: 245 seconds]
<ssvb> it would be interesting to try hardware accelerating youtube html5 videos in browsers
rellla has quit [Remote host closed the connection]
tinti has quit [Quit: Leaving]
paulk-desktop has quit [Quit: Ex-Chat]
delusr_ has quit [Ping timeout: 248 seconds]
delusr_ has joined #linux-sunxi
vicenteH has quit [Read error: No route to host]
<mnemoc> libv: mdfe's video of mer using mali with libhybris is amazingly fluid
<libv> mdfe or mbrook?
<libv> url?
<libv> not mdfe then :)
<libv> but using mdfe his obs work very directly :)
<mnemoc> :)
<libv> mnemoc: i should be ready for a20 now btw
tinti has joined #linux-sunxi
<libv> limare now supports multiple PPs
servili007 has quit [Read error: Connection reset by peer]
<mnemoc> neat
<mnemoc> charbax has a video of a mele with a20 and 1g externalllu identical to the a2000
<mnemoc> they said retail price is going to be $60
<libv> any news on how high the mali can be clocked?
<libv> and where is the memory at?
<libv> and cpu
<libv> seems to be a31 :(
<mnemoc> they showed both, a31 ($110) and a20 ($60)
<libv> ah, indeed
<mnemoc> weird considering the a31 doesn't have sata
<mnemoc> faked over usb?
delusr_ has quit [Remote host closed the connection]
delusr_ has joined #linux-sunxi
Undertasker has left #linux-sunxi [#linux-sunxi]
<mnemoc> going to take a nap. bbl
<libv> gnight
sanka has quit [Quit: Leaving]
n01 has quit [Ping timeout: 264 seconds]
tinti has quit [Quit: Leaving]
rz2k has joined #linux-sunxi