ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion | Wiki at http://linux-rockchip.info | Logs at http://irclog.whitequark.org/linux-rockchip | ML at http://groups.google.com/group/linux-rockchip
vagrantc has quit [Quit: leaving]
mac-l1 has quit [Quit: Page closed]
lkcl has quit [Ping timeout: 240 seconds]
descend has joined #linux-rockchip
lkcl has joined #linux-rockchip
Laibsch has quit [Quit: Leaving.]
cnxsoft has joined #linux-rockchip
muvlon has quit [Remote host closed the connection]
ganbold_ has joined #linux-rockchip
sjg1__ has joined #linux-rockchip
daniels_ has joined #linux-rockchip
fullstop_ has joined #linux-rockchip
dlezcano- has joined #linux-rockchip
diego71_ has joined #linux-rockchip
ganbold has quit [Ping timeout: 255 seconds]
sjg1_ has quit [Ping timeout: 255 seconds]
dlezcano has quit [Ping timeout: 255 seconds]
diego71 has quit [Ping timeout: 255 seconds]
hramrach has quit [Ping timeout: 255 seconds]
daniels has quit [Ping timeout: 255 seconds]
fullstop has quit [Ping timeout: 255 seconds]
mrueg has quit [Ping timeout: 255 seconds]
pozzoni has quit [Ping timeout: 255 seconds]
pozzoni_ has joined #linux-rockchip
sjg1__ is now known as sjg1_
pozzoni_ is now known as pozzoni
mrueg has joined #linux-rockchip
hramrach has joined #linux-rockchip
daniels_ is now known as daniels
muvlon has joined #linux-rockchip
cnxsoft1 has joined #linux-rockchip
cnxsoft has quit [Ping timeout: 252 seconds]
cnxsoft1 is now known as cnxsoft
stdint has quit [Ping timeout: 252 seconds]
stdint has joined #linux-rockchip
beeno has joined #linux-rockchip
muvlon has quit [Ping timeout: 245 seconds]
wzyy2 has joined #linux-rockchip
muvlon has joined #linux-rockchip
nighty has joined #linux-rockchip
Myy has joined #linux-rockchip
rperier has quit [Ping timeout: 245 seconds]
rperier has joined #linux-rockchip
Omegamoon has joined #linux-rockchip
<LongChair> stdint: morning
<stdint> morning
<LongChair> can you tell me more about that task API ? does it make a big difference compared to the regular MPP/MPI API ?
<stdint> just let me have a look on your code
<stdint> if you follow the unit test code, it would be token(task) version
<stdint> but if you look into the vpu api, you would know the parallel version of mpi
<LongChair> yeah i had a look at that one ... that make quite a lot more code
<LongChair> so I am wondering if that makes a real difference as i don't really need to decode 2 videos at the same time
<LongChair> if performance is the same, i'd probably stick with the current code as it looks way simpler
wadim_ has joined #linux-rockchip
<LongChair> so my first question will be, if we only need one decoding at a time, would it make any difference using that API ?
holin has joined #linux-rockchip
<Omegamoon> I am getting these... Kernel panic - not syncing: hung_task: blocked tasks
<Omegamoon> not pretty :-(
<Omegamoon> see here: http://pastebin.com/Jg5kTUPS
<Omegamoon> any idea where to look for the cause of this?
<stdint> LongChair, it is hard to talk about the task
<stdint> just look into vpu_api_legacy to see how the non-token version work
wzyy2 has quit [Read error: Connection reset by peer]
wzyy2 has joined #linux-rockchip
nighty has quit [Quit: Disappears in a puff of smoke]
premoboss has joined #linux-rockchip
wzyy2 has quit [Read error: Connection reset by peer]
andoma has quit [Ping timeout: 240 seconds]
andoma has joined #linux-rockchip
paulk-blaze has joined #linux-rockchip
andoma has quit [Ping timeout: 260 seconds]
paulk-blaze has quit [Ping timeout: 260 seconds]
scelestic has joined #linux-rockchip
andoma has joined #linux-rockchip
paulk-blaze has joined #linux-rockchip
nighty has joined #linux-rockchip
paulk-blaze has quit [Ping timeout: 240 seconds]
paulk-blaze has joined #linux-rockchip
lerc has joined #linux-rockchip
premoboss has quit [Ping timeout: 245 seconds]
paulk-blaze has quit [Remote host closed the connection]
paulk-blaze has joined #linux-rockchip
lkcl has quit [Ping timeout: 260 seconds]
Myy has quit [Quit: Leaving]
paulk-blaze has quit [Quit: Leaving]
mrueg has quit [Ping timeout: 255 seconds]
mrueg has joined #linux-rockchip
paulk-collins has joined #linux-rockchip
zzarr has joined #linux-rockchip
premoboss has joined #linux-rockchip
premoboss has quit [Ping timeout: 240 seconds]
premoboss has joined #linux-rockchip
<LongWork> stdint: i can look at the code, bu i just want to know what's the benefits of using that more complex API vs the simple one
<LongWork> you seemed to mention it was bringing better performance ... is it like significant ?
<LongWork> if we're talking about 5% more then it's probably not worth changing everything
<LongWork> if it's like 50% more, then i need to consider it
cnxsoft has quit [Quit: cnxsoft]
premoboss has quit [Ping timeout: 260 seconds]
premoboss has joined #linux-rockchip
<ayaka> I am not sure about performance as I never compare them
<LongWork> i seem to be able to get 4K@24p about consistently
<LongWork> 4K@30 is already pretty tough ...
irsol has quit [Ping timeout: 260 seconds]
premoboss has quit [Ping timeout: 240 seconds]
holin has quit [Quit: leaving]
JohnDoe_71Rus has joined #linux-rockchip
tlwoerner has quit [Quit: Leaving]
<ayaka> I think the 4K@60fps is not a problem with mpp
<ayaka> if the display screen is just 1920x1080
wadim_ has quit [Remote host closed the connection]
tlwoerner has joined #linux-rockchip
<LongWork> ayaka: this is my current config and i can't hold 60 fps
<LongWork> ayaka: i'd be iterested in you looking at my code, see if you can spot anything fishy
<phh> LongWork: possibly miniarm's clock aren't properly set :s
<LongWork> which clocks ?
<LongWork> i'm using the miniarm dts file in rk repo ...
<phh> don't know, could be RAM or vpu
<phh> can you hdparm -T ?
<phh> this will bench linear ram performance
<LongWork> if you tell me how to do that sure :)
<LongWork> because that doesn't sound familiar :)
<phh> hdparm -T /dev/mmcblk0
premoboss has joined #linux-rockchip
<LongWork> hmmm does hdparam come with any package .. cuz i don't seem to have it
<phh> it comes in hdparm
<phh> please note it's parm not param
<LongWork> installed hdparm package but no hdparm binary there
<LongWork> hold on found it
<LongWork> linaro@linaro-alip:~$ sudo /sbin/hdparm -T /dev/mmcblk0
<LongWork> Timing cached reads: 1740 MB in 2.00 seconds = 870.37 MB/sec
<LongWork> /dev/mmcblk0:
<LongWork> how is that ? :p
<phh> wow
<phh> that sucks
<LongWork> somehow it looks like a good news that it sucks :)
<phh> LongWork: how many RAM chips does the board have?
<phh> oh wait I can check photos of it
<LongWork> 2 gigs
<phh> that's not the question :P
<LongWork> oh ram chips ... no idea
<phh> noone took a photo underneath -_-'
<phh> I don't see any RAM
<ayaka> memory frequency is set in u-boot
<ayaka> not kernel
<LongWork> ah
<LongWork> i more likely use the tinekr image uboot as i only replaced the dts & zImage with a recent one
<LongWork> not sure what they have put there
<LongWork> i suppose there is no way to update only uboot ?
<phh> ok actually my chromebook also has ~ 900MB/s
utente has joined #linux-rockchip
<LongWork> so that would look like a normal bandwith ?
premoboss has quit [Ping timeout: 240 seconds]
utente__ has joined #linux-rockchip
irsol has joined #linux-rockchip
utente has quit [Ping timeout: 260 seconds]
<phh> (3840*2048*2*60)/(1024*1024) = 960
<LongWork> So it looks a bit short
<LongWork> in mpv if i remove framedropping, i can see that it would decode about 30-40 fps (visual estimation), but with framedrop, it's basically dropping 1 frame over 2 .. :/
<LongWork> 24 fps looks ok
<LongWork> 30 fps depends ...
<LongWork> mostly on bitrate
<phh> "a bit short" ? you mean there is a x2 ratio missing
<phh> you need both to read and to write
<LongWork> kinda true
<LongWork> phh : i'm displayng only in 1080p at this point
<LongWork> so write operation requires prbably much less bandwith
<phh> LongWork: ok, but it's the VOP/DRM who does the scaling atm?
<LongWork> probably, i'm using the drm stuff behind, so not sure who is doing the jb
<LongWork> phh: i'd be curious to see what kind of performance you can get out of gstreamer on 4K video
<LongWork> but if i recall what you said you can't get vpu-service on those
<phh> I still haven't got a device where I have a fully working vpu/hevc... I have tried with mainline, but it doesn't work, and it still couldn't boot rockchip's 4.4 on my tvbox...
<phh> LongWork: but testing gstreamer should be easy for you, what's your problem with it?
<LongWork> i didn't build it, and i don't know if getting it from a package will have the recent vpu stuff
<LongWork> is there a package already for the gst vpu thing ?
<LongWork> so that deb pkg is all i need ?
<LongWork> hmmm
<LongWork> sudo dpkg -i gstreamer1.0-rockchip1_1.10-4_armhf.deb
<LongWork> dpkg-deb: error: `gstreamer1.0-rockchip1_1.10-4_armhf.deb' is not a debian format archive
<LongWork> am i doing anything wrong there ?
<phh> you download by right clicking download? :P
<phh> you need to click on the link which links to a webpage
<phh> then download
<LongWork> i seem to have deps problem to install all that
<LongWork> like libstdc++6 to start with
<LongWork> for some reasons it says i have a too old version
<LongWork> librockchip-mpp1 : Depends: libstdc++6 (>= 5) but 4.9.2-10 is to be installed
<LongWork> i'm not sure if i want to break my current devenv
<LongWork> i think it's because i have the defualt gcc installed which comes with 4.9 stuff
<LongWork> yeah looks like build-essential comes with everything in 4.9
<LongWork> and that all the gst / mpp / vpu packages require >= 5
<LongWork> kinda sad
fullstop_ is now known as fullstop
beeno has quit [Ping timeout: 240 seconds]
utente has joined #linux-rockchip
Omegamoon has left #linux-rockchip [#linux-rockchip]
utente__ has quit [Ping timeout: 260 seconds]
JohnDoe_71Rus has quit [Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/]
vagrantc has joined #linux-rockchip
utente has quit [Ping timeout: 256 seconds]
lkcl has joined #linux-rockchip
zzarr has quit [Quit: This computer has gone to sleep]
irsol has quit [Ping timeout: 240 seconds]
paulk-collins has quit [Quit: Leaving]
irsol has joined #linux-rockchip
nighty has quit [Quit: Disappears in a puff of smoke]
vagrantc has quit [Quit: leaving]