khuey|away has quit [Read error: Connection reset by peer]
khuey has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
ninolein has joined #linux-sunxi
ninolein_ has quit [Ping timeout: 265 seconds]
physis has quit [Read error: Connection reset by peer]
physis has joined #linux-sunxi
physis has quit [Read error: Connection reset by peer]
physis has joined #linux-sunxi
TheSeven has quit [Ping timeout: 256 seconds]
TheSeven has joined #linux-sunxi
p1u3sch1 has quit [Ping timeout: 264 seconds]
p1u3sch1_ has joined #linux-sunxi
naobsd has joined #linux-sunxi
jawbonegroove has joined #linux-sunxi
<jawbonegroove>
hey, so i'm just an enthusiast and this is my first allwinner/non-legacy rodeo, so to speak. i just picked up an a83t tablet and i'm not sure where to start with it.
<jawbonegroove>
i did the first thing that came to mind (bad idea?) and ran adb reboot bootloader; now it appears to be stuck on the bootloader screen. adb/fastboot report nothing.
<jawbonegroove>
so my questions are, 1) is this normal, 2) what next, 3) why did it do that
naobsd has quit [Ping timeout: 246 seconds]
naobsd has joined #linux-sunxi
sehraf has joined #linux-sunxi
linkmauve1 has quit [Ping timeout: 276 seconds]
naobsd has quit [Ping timeout: 246 seconds]
Gerwin_J has joined #linux-sunxi
vishnup has joined #linux-sunxi
microboy has joined #linux-sunxi
vishnup has quit [Remote host closed the connection]
iflema has quit [Ping timeout: 256 seconds]
<plaes>
jawbonegroove: start with the new device howo on the wiki
<plaes>
or.. "You should have started..."
Gerwin_J has quit [Quit: Gerwin_J]
<jawbonegroove>
plaes: all it says is start a wiki page.
<plaes>
yes
<jawbonegroove>
ok, then will you help?
<plaes>
kind of hard to help when we know nothing about the hardware *you* have
<plaes>
that's why we want people to create device pages
orly_owl has quit [Ping timeout: 265 seconds]
iamfrankenstein has quit [Quit: iamfrankenstein]
iamfrankenstein has joined #linux-sunxi
iamfrankenstein has quit [Client Quit]
jemk has joined #linux-sunxi
Netlynx has joined #linux-sunxi
Netlynx has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
domidumont has joined #linux-sunxi
domidumont has quit [Remote host closed the connection]
domidumont has joined #linux-sunxi
cajg has joined #linux-sunxi
kaspter has joined #linux-sunxi
bonbons has joined #linux-sunxi
p1u3sch1_ has quit [Ping timeout: 250 seconds]
p1u3sch1 has joined #linux-sunxi
bgola has quit [Ping timeout: 252 seconds]
Andy-D has quit [Ping timeout: 264 seconds]
Andy-D has joined #linux-sunxi
diego71 has quit [Ping timeout: 246 seconds]
diego71 has joined #linux-sunxi
mauro_ has joined #linux-sunxi
orly_owl has joined #linux-sunxi
<microboy>
I have an optimusboard with pcduino8 ubuntu installed, ffmpeg compiled with cedrus264 but when I try to transcode "Cannot allocate frame" error appears.
microboy has quit [Ping timeout: 246 seconds]
mauro__ has joined #linux-sunxi
mauro_ has quit [Quit: Leaving]
mauro__ has quit [Client Quit]
mauro_ has joined #linux-sunxi
mauro_ has quit [Client Quit]
paulk-collins has joined #linux-sunxi
kaspter has quit [Remote host closed the connection]
mauro_ has joined #linux-sunxi
mauro_ has quit [Client Quit]
mauro_ has joined #linux-sunxi
kaspter has joined #linux-sunxi
lemonzest has joined #linux-sunxi
<mauro_>
Hello! I'd like to run a system with linux kernel 4.0.1 with only one core on my Cubietruck. I have used maxcpus=1 in bootargs and during the boot, only 1 core is brought up. But, when i log into the system (ubuntu) and type cat /proc/cpuinfo it shows me 2 cores online.
<mauro_>
is it actually using one core or have I done something wrong?
<NiteHawk>
mauro_: i assume you're using a recent/current mainline u-boot?
<mauro_>
2015.01
<NiteHawk>
a20 smp initialization for a 4.0 kernel works via PSCI. maybe that's a separate code path that doesn't necessarily respect maxcpus=1...
<mauro_>
uhm.. so shall I try to downgrade the kernel?
<NiteHawk>
i'm AFK for a while, will get back to you later...
<mauro_>
the log says clearly that only one CPU has been brought up
<mauro_>
ok
paulk-collins has quit [Ping timeout: 255 seconds]
<mauro_>
NiteHawk: i have tried with the bootm_boot_mode=sec, but it still reports two cores online. http://pastebin.com/bueBrrPF
paulk-collins has joined #linux-sunxi
jinzo has joined #linux-sunxi
reinforce has joined #linux-sunxi
iamfrankenstein has joined #linux-sunxi
paulk-collins has quit [Ping timeout: 252 seconds]
orly_owl has quit [Read error: Connection reset by peer]
orly_owl has joined #linux-sunxi
diego71 has quit [Ping timeout: 264 seconds]
diego71 has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
<NiteHawk>
mauro_: hmm, no idea there. i'll try to reproduce that with a slightly older kernel (4.0-rc6)
JohnDoe_71Rus has quit [Quit: KVIrc Insomnia 4.0.2, revision: 402, sources date: 20100627, built on: 2010-08-02 17:44:33 UTC 402 http://www.kvirc.net/]
<mauro_>
yes, it's the same as mine when i tried both of them too
<NiteHawk>
(i'm on banana pi, which is A20 too)
<mauro_>
yes i saw that :>
<mauro_>
uhm... going into /sys/devices/system/cpu/ i can only see cpu0
<mauro_>
maybe it is referred to the whole processor and not to the cores?
<NiteHawk>
no, i don't think so. i've got bot /sys/devices/system/cpu/cpu0 and /sys/devices/system/cpu/cpu1
<NiteHawk>
...both - did you double-check your dmesg output?
<mauro_>
yes, it says only one CPU has been brought up
<mauro_>
so, maybe is just lscpu which is not correct?
linkmauve1 has joined #linux-sunxi
<NiteHawk>
then i think lscpu (or /proc/cpuinfo) is misleading in your case
<mauro_>
i'll haave a try again with the normal boot in a moment
<NiteHawk>
btw, out of curiosity: is there a particular reason why you'd only want one core?
<mauro_>
yes, i am trying to compare the behavior of a system on the bare metal, Virtualized with Xen and than analyzing the benefits of VCPU affinity over the PCPUs
<NiteHawk>
ah, i see. i'll have to edit my boot.scr a bit to check the "sec" bootm_boot_mode - gimme a sec
<mauro_>
i'm putting some CPU-bound benchmarks in the "system" and using some scripts to get infos from top, sar, etc...
<mauro_>
ok sure
<mauro_>
Oh sorry, I made a mistake: I had both cpu0 and cpu1 in /sys/devices/system/cpu/
<mauro_>
i was looking at the virtual machine terminal :p
<NiteHawk>
so the info is consistent and i'm actually restricted to a single core there
<NiteHawk>
this is with u-boot 2015.04
<mauro_>
anyway i have tried to bring it down on the fly with # echo 0 > /sys/devices/system/cpu/cpu1/online
<mauro_>
but the whole MPSoC restarted
<mauro_>
after 5 seconds
<NiteHawk>
i think that feature isn't (fully) supported yet
<mauro_>
uhm.. i might try with the updated Uboot version
<mauro_>
yeah
<mauro_>
i'll try it right now
<NiteHawk>
mauro_: http://pastebin.com/UW2VULn0 - my kernel seems to respect the maxcpus=1 parameter event with bootm_boot_mode=nonsec
<NiteHawk>
s/event/even/
<NiteHawk>
...and even "echo 1 > /sys/devices/system/cpu/cpu1/online" works!
iamfrankenstein has quit [Quit: iamfrankenstein]
<mauro_>
Uhm... that is how it should be: the dmesg part is the same, but lscpu shows it's actyally not
<mauro_>
uhm...
<NiteHawk>
dmesg is different too: the PSCI interface is present (which is only available in HYP mode), but nevertheless only one core gets started
<NiteHawk>
(bootm_boot_mode=sec would cause "SVC mode", and shows no PSCI)
<mauro_>
yes, when i tried the sec mode i has a different output too
diego71 has quit [Ping timeout: 265 seconds]
<mauro_>
still, both the cpus went up
diego71 has joined #linux-sunxi
<NiteHawk>
weird. maybe the smp code is still pretty much in flux - i might have to try with a more recent kernel
<NiteHawk>
so far my versions seems to behave quite as expected (contrary to yours)
<mauro_>
ok i have tried with uboot 2015.07, but the problem still remains
<mauro_>
could it be the kernel at this point?
<NiteHawk>
possible. afaik the smp (via PSCI) code for sunxi has only been added recently...
<mauro_>
uhm... well i am using 4.0.1
<mauro_>
i'll give it a try
<NiteHawk>
i might check the latest stable (4.0.5)
RSpliet has quit [Read error: Connection reset by peer]
RSpliet has joined #linux-sunxi
<mauro_>
brb while compiling
<mauro_>
tried with 4.0.5, it is still the same
<mauro_>
NiteHawk: it is strange... the booter clearly starts only the booting CPU and leave the other offline
awe00 has joined #linux-sunxi
Andy-D has quit [Ping timeout: 252 seconds]
Net147 has joined #linux-sunxi
ricardocrudo has quit [Ping timeout: 256 seconds]
diego71 has quit [Ping timeout: 244 seconds]
diego71 has joined #linux-sunxi
<mauro_>
NiteHawk: I removed the cpu1 definition from the kernel's device tree. It seems now i have only one cpu.
ricardocrudo has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
<NiteHawk>
hmm... maybe a workaround, but that seems a rather crude solution
ricardocrudo has quit [Ping timeout: 246 seconds]
<mauro_>
I agree, but it should do the job for now.
<mauro_>
NiteHawk: anyway, thanks a lot for the help!
<NiteHawk>
if i got you right, you see two cpus brought up even in SVC mode (bootm_boot_mode=sec), and setting /sys/devices/system/cpu/cpu?/online fails for you?
<mauro_>
yes
<NiteHawk>
strange. maybe i'll try 4.0.5 myself later and report back the result
<NiteHawk>
and: you're welcome :)
<mauro_>
I have just tried with it and failed the same way
<mauro_>
after I've finished the tests i am on, i'll try to figure it out better too :>
dd has joined #linux-sunxi
Net147 has quit [Read error: Connection reset by peer]
JohnDoe_71Rus has quit [Ping timeout: 276 seconds]
patap has joined #linux-sunxi
<patap>
dd: AFAIK cedurus should't work/haven't been tested at all on a80
<patap>
it's bit different ip core
<jemk>
dd: a31 and newer have a slightly different VE kernel driver and hardware, cedrus won't work without modifications.
<jemk>
dd: dor example memory needs to be allocated with sunxi_mem or ion now
<jemk>
*for
<dd>
thank you very much to both of you patap and jemk
<jemk>
btw, which ffmpeg do you use?
<dd>
the one linked to git from sunxi
<dd>
now i've tried to install the deb package and the error is a little bit different, instead of "cannot allocate frame" now is "can't allocate VE memory"
<jemk>
dd: the source doesn't match the deb file, libavcodec/cedrus264.c is an old version. Ask stulluk to clean up his git.
<jemk>
dd: to make it work on A80 you need to add sunxi_mem/ion supoort
<jemk>
dd: I don't know of anyone having done this yet
<dd>
It would be wonderful if it worked, A20 is the last working soc with cedrus, and it rocks
vishnup has joined #linux-sunxi
Andy-D has joined #linux-sunxi
kaspter has quit [Remote host closed the connection]
mauro_ has quit [Quit: Leaving]
<wens>
montjoie: you're working on SS for A83?
orly_owl has quit [Quit: leaving]
iamfrankenstein has joined #linux-sunxi
paulk-collins has quit [Ping timeout: 245 seconds]
paulk-collins has joined #linux-sunxi
paulk-collins has quit [Ping timeout: 250 seconds]
RSpliet has quit [Read error: Connection reset by peer]
RSpliet has joined #linux-sunxi
vishnup has quit [Ping timeout: 256 seconds]
dd has quit [Ping timeout: 246 seconds]
vishnup has joined #linux-sunxi
Guest87467 has quit [Excess Flood]
vishnup has quit [Ping timeout: 256 seconds]
fredy has joined #linux-sunxi
vishnup has joined #linux-sunxi
vishnup has quit [Remote host closed the connection]
mauro has joined #linux-sunxi
<NiteHawk>
mauro: my SMP results with kernel 4.0.5 are unchanged. SVC mode brings up only one core, maxcpus=1 is respected in HYP mode, and /sys/devices/system/cpu/cpu1/online works fine for me - http://pastebin.com/tgHjd1P7
domidumont has quit [Remote host closed the connection]
<mauro>
NiteHawk: well then. I think it is is something i changed in the kernel configuration.
<mauro>
I'll have another try with a default config file later in the night
<montjoie>
wens: certainly the driver will be common with A80
techn has joined #linux-sunxi
Netlynx has quit [Remote host closed the connection]
ricardocrudo has joined #linux-sunxi
<pstef>
Would anyone know how to set max core frequency on the optimusboard? This is Linux "3.4.39+". I am able to set max. freq. with cpufreq-set, but that has two issues: 1) has to be done per-core 2) get's "forgotten" as soon as the core goes offline.
jkll has joined #linux-sunxi
jkll has quit [Client Quit]
<Turl>
pstef: sounds like you answered your own question :)
<pstef>
Yeah, but isn't there something more permanent?
paulk-collins has joined #linux-sunxi
domidumont has quit [Ping timeout: 246 seconds]
<Turl>
pstef: well, once you set them it should be good to go? or are you using something to hotplug cores?
<pstef>
I'm telling you, as soon as the core goes offline, the setting is forgotten.
patap has quit [Ping timeout: 246 seconds]
<pstef>
Not using anything to hotplug cores, not knowingly.
joost_dtn has quit [Ping timeout: 276 seconds]
<pstef>
I'm not sure if "offline" is the correct term here; it's more like cores go to sleep and wake up depending on how much CPU time is needed.
<Turl>
pstef: well, you have something hotplugging them then, otherwise they'd stay always on and your problem would be moot :)
<pstef>
I probably could set max freq. on boot, but that would be inflexible.
<pstef>
Well, you know what, thinking about it more it's a nice feature. I don't mean to be picky, but my actual problem is that the board seems to overheat when all cores go full speed - and not that they sleep when they aren't used. Classic X/Y problem.
<pstef>
I think the better move would be setting "performance" governor as the default.
<pstef>
Uh, performance is what's causing the trouble, I mean powersave.
<TheLinuxBug>
do you have heatsink on the chip? I think the even smarter move would be to find a way to cool ths chip instead of trying to limit it
<pstef>
Yeah, I don't have a heatsink on it. It's something I've been thinking about in parallel.
<pstef>
I'm not sure that the observed symptoms are actually caused by overheating, so a heatsink is not the only thing I'm thinking about here. Although overheating is the simplest explanation I can currently see.
paulk-collins has quit [Ping timeout: 276 seconds]
jinzo has quit [Quit: Leaving]
paulk-collins has joined #linux-sunxi
iamfrankenstein1 has joined #linux-sunxi
iamfrankenstein has quit [Ping timeout: 265 seconds]
iamfrankenstein1 has quit [Client Quit]
Andy-D has quit [Read error: Connection reset by peer]
sehraf has quit [Read error: Connection reset by peer]
MackBoy has quit [Ping timeout: 264 seconds]
MackBoy has joined #linux-sunxi
Andy-D has joined #linux-sunxi
lemonzest has quit [Quit: Leaving]
ricardocrudo has quit [Ping timeout: 250 seconds]
paulk-collins has quit [Ping timeout: 258 seconds]