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
paulk-collins_ has quit [Quit: Leaving]
Ziyuan has joined #linux-rockchip
vagrantc has quit [Quit: leaving]
ZeldaStar has joined #linux-rockchip
ZeldaStar has quit [Remote host closed the connection]
XingZheng has joined #linux-rockchip
Ziyuan has quit [Ping timeout: 268 seconds]
Hyperion has joined #linux-rockchip
XingZheng is now known as Xing
cnxsoft has joined #linux-rockchip
Hyperion is now known as hyperion
Aussie_matt has quit [Read error: Connection reset by peer]
Xing has quit [Ping timeout: 265 seconds]
LongChair1 has joined #linux-rockchip
LongChair1 is now known as LongChair
LongChair has quit [Remote host closed the connection]
<geekerlw> stdint, I used the newest mpp and gst-rk, when I looped a video, sometimes it stoped working, the debug log is http://paste.fedoraproject.org/514901/67513148/
<stdint> geekerlw, don't send me too much log
<stdint> I just need vpudec part
<geekerlw> stdint, with --gst-debug-level=vpudec:5 ?
<stdint> yes
<geekerlw> stdint, I found the more cpu occupancy rate, the more probability the vpu decoder process dead
<wzyy2> I have found that too
<geekerlw> stdint, sometimes it looped more than two hours, sometimes it looped less than 10 minutes
<stdint> geekerlw, which player do you use?
<wzyy2> If I go to the drag window, it's easy to stop working.
<wzyy2> If I drag window...
<geekerlw> stdint, I tested with my simple test program and the xfce parole
<stdint> if you still writing a player yourselves, don't ask me
<geekerlw> booth have this problem
<wzyy2> which sink do you use?
<stdint> use gst-launch
<geekerlw> stdint, no, I'll use the parole instead, the program is only to test the plugin
<wzyy2> eglglessink?
<stdint> I wonder how do you choose my plugin at parole
<geekerlw> stdint, I just found parole yesterday, when I start playing, it seems found your plugin automatic
<geekerlw> wzyy2, the eglglessink have the problem also
<stdint> without log, without truth
<wzyy2> The problem may not caused by gstreamer or mpp, because I haven't got it recently
<wzyy2> we will update a complete environment and then you can test in it..
<stdint> just send me a correct log first or I can't say anything
<geekerlw> wzyy2, another problem , the desktop may stop working with the xserver-xorg 1.18.4 and libmali-1.5-2 , and the old 1.18.21-4 worked more well
<geekerlw> stdint, I'll test more times, and I'll send you the correct log
<geekerlw> happy New Year you all !
<stdint> no need to care about X anymore, branch is nearly over in debian
<stdint> try wayland or waiting the porting for X 1.19
<stdint> but I still welcome the report about branch 1.18, it is not too difficult, I would have a look
<geekerlw> wayland is a good choose
<wzyy2> Debian stretch is updated too frequently.....We can not ensure the stability....
<wzyy2> How long will it become the stable distribution?.
<stdint> the testing become stable costs one and half year
<stdint> but stretch have not become testing I think
geekerlw has quit [Read error: Connection reset by peer]
geekerlw has joined #linux-rockchip
geekerlw has quit [Read error: Connection reset by peer]
geekerlw has joined #linux-rockchip
hyperion has quit [Ping timeout: 264 seconds]
_whitelogger has joined #linux-rockchip
ganbold has quit [Quit: This computer has gone to sleep]
ganbold has joined #linux-rockchip
geekerlw has left #linux-rockchip ["Leaving"]
geekerlw has joined #linux-rockchip
geekerlw has quit [Client Quit]
geekerlw has joined #linux-rockchip
geekerlw has quit [Quit: Leaving]
geekerlw has joined #linux-rockchip
<LongChair> morning, did anyone played with the mali fbdev drivers already ?
libv has quit [Ping timeout: 260 seconds]
LongChair has quit [Quit: LongChair]
libv has joined #linux-rockchip
LongChair has joined #linux-rockchip
XingZheng has joined #linux-rockchip
<geekerlw> stdint, I got the log http://paste.fedoraproject.org/514931/48308085/
<stdint> vpudec gstvpudec.c:101:gst_vpudec_start:<vpudec34> Starting
<stdint> of course it won't be stop
<stdint> it is start again
<geekerlw> but the video stop playing
<geekerlw> It stoped at the last frame
<stdint> you mean the "vpudec gstvpudec.c:489:gst_vpudec_handle_frame:<vpudec34> Handling frame 22" is the last?
<stdint> then no buffer is push to vpudec
<stdint> once the vpudec received a buffer it would print that message
<stdint> if the video start again and you could see the video output
<stdint> the problem looks that the upstream is dead
<geekerlw> no , the last frame seems 239, the log is incomplete
<stdint> please send me the correct log
<stdint> nothing different
<stdint> no error from vpudec, no more buffer from upstream
<stdint> blame the upstream
<stdint> the upstream element
<geekerlw> ok, I'll check my environment and test again
cnxsoft has quit [Ping timeout: 256 seconds]
<stdint> geekerlw, you may enable the debug of vpubufferpool
<stdint> if you didn't find any problem there, report that to #gstreamer
<geekerlw> ok, I'll have a try
cnxsoft has joined #linux-rockchip
cnxsoft has quit [Read error: Connection reset by peer]
cnxsoft has joined #linux-rockchip
muvlon has quit [Ping timeout: 240 seconds]
zzarr has quit [Remote host closed the connection]
XingZheng has quit [Ping timeout: 256 seconds]
muvlon has joined #linux-rockchip
lkcl has joined #linux-rockchip
mac-l1 has joined #linux-rockchip
Omegamoon has joined #linux-rockchip
Aussie_matt has joined #linux-rockchip
zzarr has joined #linux-rockchip
paulk-collins has joined #linux-rockchip
Aussie_matt has quit [Remote host closed the connection]
Aussie_matt has joined #linux-rockchip
<mmind00> diego71: not sure if you made any progress on your cpufreq issue, but I just ran my firefly with some 4.10-rc1-based kernel build from multi_v7_defconfig and get working cpufreq ... dmesg shows cpufreq moving away from the unlisted 500MHz to 600MHz as well
<mmind00> diego71: maybe you can mount debugfs again and look at /debug/regulator/regulator_summary and look for vdd_cpu if it is present?
<mmind00> which should have a child named cpu0
jkstrick has joined #linux-rockchip
Aussie_matt has quit [Remote host closed the connection]
cnxsoft has quit [Quit: cnxsoft]
muvlon has quit [Ping timeout: 240 seconds]
libv_ has joined #linux-rockchip
libv has quit [Ping timeout: 246 seconds]
libv has joined #linux-rockchip
libv_ has quit [Ping timeout: 264 seconds]
libv_ has joined #linux-rockchip
libv has quit [Ping timeout: 246 seconds]
<diego71> mmind00: yes, i've tested the linux-next one, with multi_v7_defconfig, and they looks fine (except, i had to patch mmc driver)
<diego71> mmind00: thanks anyway
<mmind00> diego71: ok, glad to hear its working for you again
<diego71> and there is till the problem with the usb
<diego71> *still
<diego71> (I was told there is a fix around, that should arrive in mainline...)
libv has joined #linux-rockchip
libv_ has quit [Ping timeout: 268 seconds]
vagrantc has joined #linux-rockchip
vagrantc has quit [Quit: leaving]
libv_ has joined #linux-rockchip
libv has quit [Ping timeout: 248 seconds]
libv has joined #linux-rockchip
libv_ has quit [Ping timeout: 258 seconds]
libv has quit [Ping timeout: 264 seconds]
libv has joined #linux-rockchip
wzyy2 has quit [Ping timeout: 260 seconds]
vagrantc has joined #linux-rockchip
<phh> LongChair: mali fbdev driver...? you mean fbdev-x11 driver with mali support? it's obsolete compared to either rockchip's X or even wayland
<LongChair> I meant a video driver that doesn't require a window manager
<LongChair> not fbdev-x11, just fbdev
<phh> oh right of course, sorry
<LongChair> np :)
Omegamoon has quit [Ping timeout: 264 seconds]
Omegamoon has joined #linux-rockchip
libv has quit [Ping timeout: 248 seconds]
dlezcano has quit [Ping timeout: 258 seconds]
libv has joined #linux-rockchip
dlezcano has joined #linux-rockchip
mac-l1 has quit [Ping timeout: 260 seconds]
vagrantc_ has joined #linux-rockchip
vagrantc has quit [Ping timeout: 256 seconds]
vagrantc_ is now known as vagrantc
<diego71> vagrantc: fyi, the firefly, with kernel 4.10-rc1 is much faster than the 4.8 from debian
<diego71> now it uses 1.6Ghz, instead of 500Mhz of the old kernel
<vagrantc> good news
<vagrantc> wonder what changed
<vagrantc> i've got one board running 4.9-rc8, too
<vagrantc> debian's likely going to be shipping 4.9 for the upcoming release
<diego71> vagrantc: how it works with 4.9?
<vagrantc> oh, wait, i'm running 4.7 and 4.8 ...
<vagrantc> i should try 4.9
<diego71> looks like the debian kernel (4.8) doesn't support cpufreq for the rockchip
<diego71> (I don't understand if it is a bug, or just missing some feature needed for cpufreq)
<vagrantc> well, if it means it runs at 500MHz, i'd call it a bug worth backporting a fix for :)
<diego71> I suspect that 500Mhz is the standard clock at boot. With 4.10-rc1, I got this message:
<diego71> [ 0.662501] cpufreq: cpufreq_online: CPU0: Running at unlisted freq: 500000 KHz
<diego71> [ 0.662741] cpufreq: cpufreq_online: CPU0: Unlisted initial frequency changed to: 600000 KHz
<vagrantc> cpufreq-info still not showing anything on 4.9-rc8
<vagrantc> i don't get any cpufreq info out of dmesg
<diego71> I can try to bisect from 4.9 to 4.10-rc1 to see where cpufreq start to work
<vagrantc> that'd be awsome
<vagrantc> it might also just be modules we need to enable in debian
<diego71> for 4.10, i've just used multi_v7_defconfig
<vagrantc> i've always thought my firefly boards were a bit slow
<vagrantc> so it's not surprising to hear cpufreq not working might make them run at a slow clock speed
<vagrantc> multi_v7_defconfig is like looking for a needle in a haystack in which someone's already sifted through maybe 10% of the haystack
<vagrantc> ok, maybe not that bad :)
<diego71> The idea is to test 4.9, with multi_v7_defconfig. If it works, is something with the debian kernel (maybe something different in the config).
<diego71> if not, it is something fixed between 4.9 and 4.10-rc1
<vagrantc> well, between 4.9-rc8 and...
<vagrantc> but yes, bisecting would be good
<diego71> right ...
<vagrantc> i'm also looking at the device-tree to see if there's an obvious driver we've missed in the config or something
<mmind00> vagrantc: also maybe look into debugfs regulator/regulator_summary ... I could fathom that the vdd_cpu is not present
<vagrantc> i swear i got cpufreq on the veyron-speedy, though ...
<vagrantc> i'm going to boot that and see
<mmind00> vagrantc: pmic setup is vastly different between veyron (rk808) and firefly (act8846 + syr-something)
<vagrantc> ah, so it's basically meaningless with the veyron
<vagrantc> for the firefly
<vagrantc> needed to boot it anyways :)
<mmind00> vagrantc: firefly's regulator-drivers are in multi_v7_defconfig since september 2015, but I'm not sure if that's also the case for your debian kernel-config
<vagrantc> paulk-collins: was it you that mentioned something about a usb bug causing problems with usb-wifi adapters?
<vagrantc> mmind00: which are they?
<mmind00> vagrantc: CONFIG_REGULATOR_ACT8865=y and CONFIG_REGULATOR_ACT8865=y
<paulk-collins> vagrantc, yeah
<mmind00> vagrantc: whops CONFIG_REGULATOR_FAN53555=y
<vagrantc> paulk-collins: any news on that?
<paulk-collins> vagrantc, with ath9k_htc
<paulk-collins> didn't have time to try the diff I was sent yet
<paulk-collins> I hope I'll get to it soon
<paulk-collins> I have to prepare exams these days, mostly
<vagrantc> mmind00: CONFIG_REGULATOR_FAN53555=m in the debian configs
<vagrantc> mmind00: and CONFIG_REGULATOR_ACT8865=m
<vagrantc> modules, of course
<diego71> vagrantc: so we should have loaded this modules to make cpufreq works... I suppose
<vagrantc> lsmod | grep act: act8865_regulator 11502 19
<vagrantc> not sure what module FAN53555 produces
<vagrantc> ok...
<vagrantc> manually loading the fan53555 module made cpufreq work on 4.9-rc8
<vagrantc> not sure why it didn't load automatically
<vagrantc> ditto for 4.8.x
<vagrantc> alright, well, that ought to get some more performance out of the archive
<diego71> good, I don't have to compile another kernel :)
<vagrantc> presuming they were running at a slower clock speed
<diego71> vagrantc: very likely...
<vagrantc> diego71: thanks for nudging the issue!
<vagrantc> paulk-collins: if you could forward me the patch, i might get a chance to try it
<diego71> vagrantc: you're welcome :)
<diego71> uhm, with flash-kernel, can I tell to boot with a older kernel?
<diego71> (without removing the current one)
<diego71> root
<diego71> ops
<vagrantc> flash-kernel --force VERSION
<vagrantc> you can also interrupt u-boot and editenv boot_scripts
<vagrantc> and select which boot scripts to try
<diego71> vagrantc: thanks
<diego71> confirmed: modprobe fan53555, fixes the problem
<diego71> I've also run a quick benchmark, and now it is faster
jkstrick has quit [Ping timeout: 246 seconds]
<vagrantc> yay
<diego71> I have to set "on demand governor", because performance seems to be the default (I notice higher current at power source)
<diego71> but yay indeed :)
<diego71> I can go to sleep happy... Thanks to everyone.
<vagrantc> diego71: another good catch
<vagrantc> diego71: "service cpufrequtils restart" set it to ondemand for me
<vagrantc> suspect it would also work at boot now that i have all the modules configured to load properly
paulk-collins has quit [Quit: Leaving]
Omegamoon has left #linux-rockchip [#linux-rockchip]