Black_Horseman has quit [Ping timeout: 272 seconds]
HeavyMetal has joined #linux-sunxi
libv has quit [Ping timeout: 264 seconds]
libv has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 256 seconds]
khuey is now known as khuey|away
HeavyMetal has joined #linux-sunxi
awe00 has quit [Ping timeout: 252 seconds]
libv_ has joined #linux-sunxi
libv has quit [Ping timeout: 264 seconds]
libv_ has quit [Ping timeout: 264 seconds]
libv has joined #linux-sunxi
naobsd has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 256 seconds]
HeavyMetal has joined #linux-sunxi
simosx has quit [Quit: Leaving]
HeavyMetal has quit [Ping timeout: 265 seconds]
lucaswang has joined #linux-sunxi
DagoRed has joined #linux-sunxi
HeavyMetal has joined #linux-sunxi
HeavyMetal has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 256 seconds]
libv has quit [Ping timeout: 264 seconds]
HeavyMetal has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
kaspter has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
jemk has quit [Ping timeout: 256 seconds]
jemk has joined #linux-sunxi
TheQuestionmark has quit [Ping timeout: 252 seconds]
TheQuestionmark has joined #linux-sunxi
kaspter has quit [Remote host closed the connection]
kaspter has joined #linux-sunxi
lucaswang has quit [Ping timeout: 245 seconds]
cnxsoft1 has joined #linux-sunxi
cnxsoft has quit [Read error: Connection reset by peer]
naobsd has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
cnxsoft1 has quit [Ping timeout: 252 seconds]
p1u3sch1 has joined #linux-sunxi
p1u3sch1_ has quit [Ping timeout: 244 seconds]
bsdfox has quit [Ping timeout: 246 seconds]
bsdfox has joined #linux-sunxi
TheSeven has quit [Ping timeout: 256 seconds]
TheSeven has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
kurain has joined #linux-sunxi
<kurain>
hello all
<kurain>
Recently, I am debugging owncloud on A31 soc, but I met one problem, mariadb and apache2 needed by owncloud needs transparent_hugepage support, which need LPAE support for kernel
<kurain>
some one have ever tested this feature?
konradoo77 has quit [Ping timeout: 245 seconds]
reinforce has joined #linux-sunxi
cubeast has joined #linux-sunxi
souther has quit [Ping timeout: 256 seconds]
<wens>
LPAE is fine, just enable it
lucaswang has joined #linux-sunxi
souther has joined #linux-sunxi
Andy-D has joined #linux-sunxi
domidumont has joined #linux-sunxi
domidumont has quit [Remote host closed the connection]
premoboss has quit [Read error: Connection reset by peer]
domidumont has joined #linux-sunxi
HeHoPMaJIeH has joined #linux-sunxi
<kurain>
hey, wens
<kurain>
but what I want is to enable transparent_hugepage, which needs LPAE support
<kurain>
after transparent_hugepage patching, the kernel hangs after loaded
<kurain>
also one problem is that if we can enable both LPAE and high memory features.
<kurain>
?
gianMOD has joined #linux-sunxi
<wens>
does LPAE alone work for you?
gianMOD has quit [Remote host closed the connection]
reinforce has quit [Quit: Leaving.]
<wens>
could someone test if a10-lime is stable @ 960 MHz?
iamfrankenstein has joined #linux-sunxi
_massi has joined #linux-sunxi
sehraf has joined #linux-sunxi
Akagi201 has joined #linux-sunxi
<kurain>
nop
<kurain>
it will hang after loaded
ricardocrudo has joined #linux-sunxi
cnxsoft has quit [Remote host closed the connection]
cnxsoft has joined #linux-sunxi
<maz>
kurain: at least on A20, THP+LPAE work perfectly, I use that all the time.
<kurain>
maz, what version of kernel are you using, I am working on the kernel provided by allwinner
<kurain>
not by the community
<maz>
kurain: I only use mainline, nothing else. which version is allwinner's kernel?
<kurain>
the linux-3.3 version
<kurain>
although it has configure for LPAE, but without THP
<maz>
kurain: very unlikely to work. we fixed a number of critical bugs after that.
<maz>
kurain: THP support was added much later.
<maz>
kurain: what prevents you from using a mainline kernel?
<kurain>
hmmm, the display of A31, which is based on PowerVR 544
<wens>
ssvb: seems like a20 @ 960 MHz / 1.4V is dangerous too
<maz>
kurain: OK. what you described looked a lot more like a server workload, so I imagined you didn't really need a display.
<kurain>
hmm, good advice
<wens>
if you're not doing anything fancy gpu/display wise, simplefb should be enough
<kurain>
I have ever tested mainline kernel, maybe I will have another try
<kurain>
thx, wens and maz
<maz>
kurain: no worries.
FR^2 has joined #linux-sunxi
<kurain>
maz, you use the kernel directly from Linus's repository? or the next branch of linux-sunxi
<ssvb>
maz: what about non-LPAE THP?
<ssvb>
wens: was it the result of your tests on a20?
<maz>
ssvb: it's been tried (Steve capper wrote some patches), but the complexity of the page tabl format makes is... hmmm... interesting.
<wens>
ssvb: mine looks stable, but i haven't kept it running for extended periods of time
<maz>
kurain: only Linus' repository.
<kurain>
ok, fine, maz, thx, I will try.
<maz>
kurain: at least for my A20 boards, there is everything I need. things might be different for your A31 though.
<ssvb>
maz: yes, I have tested these patches and they help really a lot on Cortex-A8, too bad that there is a resistance against taking them in use
<maz>
ssvb: there is resistence because you can break them pretty easily.
<maz>
ssvb: getting the required atomicity of updates with the split HW/SW PTEs is a nightmare.
<ssvb>
maz: one can break many things pretty easily, but this is not a good reason to avoid implementing something useful
<maz>
ssvb: if you can break it from userspace, this is definitely useless.
<ssvb>
maz: what about just having an option to use larger pages instead of the i386 style 4KiB?
<wens>
hugetlbfs?
<maz>
ssvb: the only option for classic page tables would be 64k.
<ssvb>
maz: yes, and it looks like a rather good choice
<maz>
ssvb: you clearly haven't thought it through... ;-)
<maz>
ssvb: I run with 64k pages on a bunch of ARMv8 platforms, and that's a massive waste of RAM.
<maz>
ssvb: so on small machines such as most of 32bit platforms, it is terrible.
<ssvb>
well, i386 was a "small" machine, with something like 2 or 4 MiB of RAM :-)
<maz>
ssvb: the sweet spot would be something like 8 to 16k, but that won't happen on ARMv7.
<maz>
ssvb: yes, and it never used 64k pages for a good reason.
gianMOD has joined #linux-sunxi
<ssvb>
but now we have something like 256 times more RAM than i386 had :-)
<maz>
ssvb: which is still very little.
<maz>
ssvb: frankly, try and run some very basic workloads on an ARMv8 platform.
<ssvb>
where is the most of RAM wasted with oversized pages?
<maz>
ssvb: you can use QEMU for that
<ssvb>
is it stack? heap? or something else?
<maz>
ssvb: you waste memory everywhere, from the process stacks to the page cashe, to the minimal alignment of data structures...
<ssvb>
how are the data structures related here?
<maz>
ssvb: read again. alignment.
lucaswang has quit [Ping timeout: 256 seconds]
<ssvb>
why would the userland code rely on the page size for alignment?
<ssvb>
except for some very specialized code
<maz>
ssvb: for example, a hackbench run that takes about 600M on a 4k page size, and about 3G with 64k pages.
gianMOD has quit [Ping timeout: 246 seconds]
<ssvb>
this sounds like its memory allocator just sucks
<ssvb>
has anyone looked into *why* this is happening?
<maz>
ssvb: I'll let you fix it then...
<maz>
ssvb: I'm sure you'll have a brilliant solution for us, looking forward to reviewing your patches! :)
<ssvb>
maz: I would prefer to be paid for implementing a brilliant solution :-)
Andy-D has quit [Ping timeout: 252 seconds]
<ssvb>
maz: is this hackbench thing just a kind of fork bomb? how are these artificial results relevant to real workloads?
naobsd has quit [Quit: naobsd]
<ssvb>
maz: okay, looks like I need to revisit this THP thing again and check what's up there :-)
<ssvb>
wens: hugetlbfs is kind of awkward to use, transparent huge pages are better (and they can be enabled/disabled for memory buffers via madvise flags MADV_HUGEPAGE/MADV_NOHUGEPAGE)
<ssvb>
I was not impressed with automatic transparent huge pages, the performance impact is a hit and miss
<kurain>
maz, I want to use THP because when I use owncloud with opensuse, it shows that mariadb needs this feature, or it will not work
<kurain>
that's the reason
<kurain>
and I am using the loftq board produced by Mixtile, 2GiB memory and 8Gib emmc
<kurain>
also external sata support, I need a personal cloud storage on it
<ssvb>
kurain: just move to the mainline kernel, this will save you from a lot of troubles
<kurain>
ok
<ssvb>
what kind of a31 device is that?
awe00 has joined #linux-sunxi
leviathancn has joined #linux-sunxi
<ricardocrudo>
has the A20 internal temperature sensor?
<ssvb>
but it does not look accurate or complete yet
<ssvb>
oh, this was in fact you :)
<kurain>
ssvb, I created it
<kurain>
;)
<ssvb>
does it mean that you have tested the mainline kernel with the sun6i-a31-hummingbird.dtb on this board?
<kurain>
yes
<kurain>
it works
<kurain>
also the mainline uboot works for loftq with the legacy kernel
<wens>
best create your own dts though
<ssvb>
is it 100% identical to the humminboard?
<kurain>
nop
<kurain>
the main difference is wifl&bt module, rtc, display, emmc, and external sata, usb host
<kurain>
well, many many, I think, the best thing I think is that it has on NXP jn516p chip for zigbee
<ssvb>
then you should really create your own dts file, just as wens said
<kurain>
NXP Jn5168
<kurain>
yep, I am on my way, ;)
<Wizzup>
m/win 28
<Wizzup>
oops. sorry.
ssvb has quit [Quit: Leaving]
hipboi has joined #linux-sunxi
hipboi_ has quit [Ping timeout: 245 seconds]
awe00 has quit [Ping timeout: 264 seconds]
awe00 has joined #linux-sunxi
gzhechu has joined #linux-sunxi
gianMOD has joined #linux-sunxi
gianMOD has quit []
leviathancn has quit [Ping timeout: 256 seconds]
naobsd has joined #linux-sunxi
Akagi201 has quit [Ping timeout: 246 seconds]
hno has quit [Quit: Coyote finally caught me]
hno has joined #linux-sunxi
kaspter has quit [Ping timeout: 256 seconds]
naobsd has quit [Ping timeout: 245 seconds]
naobsd has joined #linux-sunxi
FreezingAlt has quit [Read error: Connection reset by peer]
FreezingAlt has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
MY123 has joined #linux-sunxi
Andy-D has joined #linux-sunxi
naobsd has joined #linux-sunxi
Froolag has joined #linux-sunxi
froolap has quit [Read error: Connection reset by peer]
Froolag has quit [Client Quit]
leviathancn has joined #linux-sunxi
froolap has joined #linux-sunxi
kaspter has joined #linux-sunxi
Andy-D has quit [Ping timeout: 252 seconds]
sillysnowflake has joined #linux-sunxi
<sillysnowflake>
hi. anyone who has the banana pi, what type is the power connector for the hard disk drive?
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
premoboss has joined #linux-sunxi
ssvb has joined #linux-sunxi
selfbg has joined #linux-sunxi
tomcheng76 has quit [Read error: Connection reset by peer]
tomcheng76 has joined #linux-sunxi
<NiteHawk>
sillysnowflake: i don't know the exact naming / part number of that specific connector type. it's a simple 2-pin socket that takes the same plugs that cubieboard/cubietruck use. see e.g. http://linux-sunxi.org/File:Cubietruck-7_Frontside_RevA.JPG (banana pi supplies only +5V)
<NiteHawk>
ready-made cables are available from the usual suspects (e.g. amazon and ebay)
cnxsoft has quit [Quit: cnxsoft]
cubeast has quit [Quit: Leaving]
<atsampson>
NiteHawk: "JST XH" (or a clone of it), I think
<MY123>
Does the A20 have a 16650 compatible UART?
selfbg has quit [Quit: Leaving]
<premoboss>
MY123, at least there is one serial port
<MY123>
premoboss, is it NS16650 compatible?
<MY123>
(I need this)
<premoboss>
it run and 0-3.3V, 5Volt NO
<MY123>
premoboss, no voltage at all , but UART IP type
<premoboss>
i dont understnd meaning of "UART IP type"
<ssvb>
MY123: use the source!
<MY123>
ssvb, For this project, I can't look to GPL sources
<MY123>
(license issues)
<MY123>
ssvb, I prefer saying now that it would be propretary
<ganbold_>
it is from synopsis designware
<MY123>
ganbold_, UART from Synopsis?
<ganbold_>
yes
premoboss has quit [Remote host closed the connection]
<MY123>
ganbold_, so impossible except with ARM VT and a crappy emulation
naobsd has quit [Quit: naobsd]
<MY123>
[I'm trying to port Windows RT to sunxi after the successful port on the RPi2(only with ramdisk,no uSD driver) and that specifically needs a NS16650 for debugging]
naobsd has joined #linux-sunxi
<MY123>
and the ALT5 includes that on the platform
<ssvb>
MY123: have you tried to look into the A20 manual? it says "Compliant with industry-standard 16550 UARTs" about the UART
<ssvb>
and has the hardware registers description
<MY123>
ssvb, where is the manual? is it in Chinese like the one I have right now and that I can't read?
<ssvb>
have you been living under a rock?
<ssvb>
sorry, no offense intended
<ssvb>
the linux-sunxi wiki has a lot of information and all the necessary references to the SoC documentation
<MY123>
ssvb, did you try run the Diehard tests with 20GB of random data?
<ssvb>
TestU01 is supposed to be a newer and much better test
<ssvb>
I suspect that you are using the PRNG wrong and getting bogus data from it
<MY123>
ssvb, using 3.4.75
<MY123>
(for Mali r3p2)
hipboi has joined #linux-sunxi
kill_-9_1 has joined #linux-sunxi
MY123 has quit [Ping timeout: 246 seconds]
khuey|away is now known as khuey
bonbons has joined #linux-sunxi
naobsd has quit [Remote host closed the connection]
naobsd has joined #linux-sunxi
naobsd has quit [Client Quit]
<sillysnowflake>
isn't intel hasswel true randomness because physics? how can it be "low randomness"?
naobsd has joined #linux-sunxi
kill_-9_1 is now known as MY123
<MY123>
sillysnowflake, it is backdoored
<MY123>
not true randomness at all
<sillysnowflake>
do you have proof that it is backdoored?
<MY123>
sillysnowflake, it fails miserably at diehard
<sillysnowflake>
do you have links?
* ssvb
believes that this just shows that MY123 is using the diehard test incorrectly :-)
<MY123>
ssvb, I'm just using /dev/random
<ssvb>
it is not a rocket science to implement a reasonably good PRNG in hardware, one can just use the output of a symmetric cipher
<ssvb>
and the crypto accelerator already implements plenty of them
<ssvb>
MY123: are you trying to test /dev/random output with diehard?
<MY123>
ssvb, yeah
<ssvb>
can you share the code?
<sillysnowflake>
i don't wanna be all pro-intel (better not trust something unless proven it is secure by 3rd parties), but wasn't haswell actually using quantum tunneling for randomness?
<montjoie[home]>
MY123: if you can share your driver for the prng
<montjoie[home]>
for comparing methods to configure it
dev1990 has joined #linux-sunxi
<montjoie[home]>
but for the moment my driver pass the rngtest tool
MY123 has quit [Ping timeout: 272 seconds]
<khuey>
anybody know what device tree incantations are needed to enable nand flash on an olimex board?
<khuey>
all of the kernels I've seen that support it are ancient
<khuey>
I guess I could dig through the source to figure out what they do
MY123 has joined #linux-sunxi
awe00 has quit [Ping timeout: 245 seconds]
paulk-collins has quit [Remote host closed the connection]
paulk-collins has joined #linux-sunxi
simosx has joined #linux-sunxi
rsevero has joined #linux-sunxi
<rsevero>
Hi, I'm trying to use a USB videograbber on a Banna Pi. As far as I can tell, the river isn't working properly despite recognizing the device and avconv starting to capture the file. It hangs after a second or 2. How should I proceed to fix this issue?
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
hipboi has quit [Ping timeout: 244 seconds]
hipboi_ has joined #linux-sunxi
ricardocrudo has quit [Remote host closed the connection]
leviathancn has quit [Ping timeout: 252 seconds]
leviathancn2 has joined #linux-sunxi
_massi has quit [Quit: Leaving]
paulk-collins has quit [Remote host closed the connection]
paulk-collins has joined #linux-sunxi
Night-Shade has joined #linux-sunxi
iamfrankenstein has quit [Ping timeout: 256 seconds]
leviathancn2 has quit [Ping timeout: 265 seconds]
f15h has joined #linux-sunxi
hipboi has joined #linux-sunxi
joedj has quit [Ping timeout: 250 seconds]
hipboi_ has quit [Ping timeout: 265 seconds]
joedj has joined #linux-sunxi
ZbooNet has joined #linux-sunxi
awe00 has joined #linux-sunxi
khuey is now known as khuey|away
Netlynx has quit [Quit: Leaving]
premoboss has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
MY123 has quit [Quit: Leaving]
Seppoz_nowork has quit [Remote host closed the connection]