Turl changed the topic of #linux-sunxi to: Allwinner/sunxi /development discussion - did you try looking at our wiki? https://linux-sunxi.org - Don't ask to ask. Just ask and wait! - https://github.com/linux-sunxi/ - Logs at http://irclog.whitequark.org/linux-sunxi
interrobangd has quit [Quit: Leaving]
mosterta has quit [Ping timeout: 276 seconds]
MaDMaLKaV has quit [Ping timeout: 260 seconds]
adj__ has joined #linux-sunxi
adj_ has quit [Ping timeout: 248 seconds]
orly_owl has joined #linux-sunxi
catphish has quit [Quit: Leaving]
MaDMaLKaV has joined #linux-sunxi
MackBoy has quit [Ping timeout: 240 seconds]
avph has quit [Ping timeout: 240 seconds]
MackBoy has joined #linux-sunxi
avph has joined #linux-sunxi
vagrantc has quit [Quit: leaving]
<MoeIcenowy> apritzel: congrats
<MoeIcenowy> in aarch64 mode?
<apritzel> MoeIcenowy: yes
uwe_ has quit [Ping timeout: 240 seconds]
uwe_ has joined #linux-sunxi
apritzel has quit [Ping timeout: 248 seconds]
gzamboni has quit [Read error: Connection reset by peer]
cnxsoft has joined #linux-sunxi
naobsd has joined #linux-sunxi
MaDMaLKaV has quit [Ping timeout: 276 seconds]
MaDMaLKaV has joined #linux-sunxi
<wens> KotCzarny: net-next should have some patches improving dwmac performance in general, iirc it was 7xx mbps -> 8xx mbps
cnxsoft has quit [Read error: Connection reset by peer]
<wens> my mail is still down :/
cnxsoft has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
hipboi_ has quit [Quit: Leaving]
hipboi has joined #linux-sunxi
cnxsoft1 has joined #linux-sunxi
ganbold has quit [Quit: This computer has gone to sleep]
cnxsoft has quit [Ping timeout: 250 seconds]
cnxsoft1 is now known as cnxsoft
p1u3sch1 has quit [Ping timeout: 250 seconds]
p1u3sch1 has joined #linux-sunxi
<wens> mripard: there's some errors in the clock patches you pushed
TheSeven has quit [Ping timeout: 240 seconds]
TheSeven has joined #linux-sunxi
<wens> plaes: thanks, regulators are pretty much done :)
<wens> only thing missing is multi-phase binding for axp806
ganbold has joined #linux-sunxi
IgorPec has joined #linux-sunxi
sunxi_fan1 has joined #linux-sunxi
IgorPec has quit [Ping timeout: 245 seconds]
JohnDoe_71Rus has joined #linux-sunxi
specing has quit [Ping timeout: 248 seconds]
orly_owl has quit [Ping timeout: 264 seconds]
vagrantc has joined #linux-sunxi
IgorPec has joined #linux-sunxi
vagrantc has quit [Ping timeout: 252 seconds]
orly_owl has joined #linux-sunxi
<KotCzarny> wens: great! is similiar thing possible for sata?
vagrantc has joined #linux-sunxi
vagrantc has quit [Ping timeout: 240 seconds]
domidumont has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
domidumont has quit [Remote host closed the connection]
vagrantc has joined #linux-sunxi
domidumont has joined #linux-sunxi
<wens> KotCzarny: no idea, that was performance issues in the code, not hardware
tchiwam_ has joined #linux-sunxi
tchiwam has quit [Ping timeout: 248 seconds]
domidumont has quit [Ping timeout: 246 seconds]
Gerwin_J_ has joined #linux-sunxi
Gerwin_J has quit [Ping timeout: 248 seconds]
Gerwin_J_ is now known as Gerwin_J
domidumont has joined #linux-sunxi
sunxi_fan1 has quit [Ping timeout: 240 seconds]
<wens> anyone know how to use earlycon? earlycon=uart8250,mmio32,0xXXXXXXXX,115200n8 gives me garbage
<mripard> wens: damn, I was tired and screwed the merge :/
reinforce has joined #linux-sunxi
yann|work has quit [Ping timeout: 250 seconds]
vagrantc has quit [Quit: leaving]
ricardocrudo_ has joined #linux-sunxi
<wens> shit happens
ricardocrudo_ has quit [Ping timeout: 240 seconds]
<kadamski> what is a CPUS that is sometimes referenced in the Allwinner User Manuals?
<ssvb> kadamski: it is http://linux-sunxi.org/AR100
<kadamski> oh, makes sense
<kadamski> and I'm wondering, H3 has a second PIO port like some other SoCs
<kadamski> which is refered to as R_PIO in other socs
<kadamski> in H3 documentation it is in the CPUs-Port section
<kadamski> would that CPUs be also somehow related to CPUS?
_massi has joined #linux-sunxi
<kadamski> or it's just a confusing naming?
<ssvb> yes, I believe that Allwinner just expected that the peripherals with the R_* prefix are going to be controlled by the OpenRISC core
hipboi has quit [Quit: This computer has gone to sleep]
<ssvb> this is not mandatory because both ARM and OpenRISC can access them
<ssvb> but some kind of arbitration is necessary to share these peripherals safely and prevent ARM and OpenRISC stepping on each other's toes
ganbold has quit [Quit: This computer has gone to sleep]
<kadamski> makes sense - on OrangePi PC this port is used for CIR, USB_OTG voltage, PWR led, and some i2c voltage regulator etc, all of which could be useful to have access to in suspend mode
<jelle> btw the voltage regulator of the orange pi pc isn't that fancy
<jelle> I've submitted the simple voltage setting driver to u-boot
<kadamski> jelle: you're talking about the SY8106A which is connected to I2C on PL ?
tkaiser has joined #linux-sunxi
<jelle> kadamski: yes the SY8106A
<tkaiser> jelle: Nice! But are the defaults (1008 MHz @ 1.2V) not safe for u-boot and the dvfs stuff is needed later by the kernel?
<jelle> tkaiser: uhh sorry I'm not much of an expert :S
<ssvb> tkaiser: if you have DVFS supported in the kernel and reboot the board, then U-Boot starts with whatever was the last configured voltage
<tkaiser> ssvb: Ah, now I understand. Thx
<tkaiser> So having this in u-boot is a prerequisit to add dvfs stuff to the kernel since otherwise 'reboot impossible' situations might occur?
<kadamski> is AR100 functionality used somehow from the mainline kernel or not yet?
<wens> not yet
<jelle> tkaiser: basically https://gist.github.com/jelly/baf3696e09368d11f3fc if you are interested
<jelle> submitted it to hans, since he helps me around with all this mainlining :-)
<kadamski> and what about the kernel-3.4 ?
<ssvb> jelle: why not just post it to the U-Boot mailing list?
<jelle> but the sY8106A can't power off the board completely
<wens> for the h3 the openrisc is needed more, as a pmu since the external one doesn't do much :(
<ssvb> kadamski: considering the last developments, AR100 is probably not going to be used by the kernel at all
<jelle> ssvb: not sure, this works ok for me
<kadamski> ssvb: what developments are you reffering to?
<ssvb> jelle: well, you can't claim that you "have submitted the simple voltage setting driver to u-boot" until you submit it for real
<jelle> ssvb: that's correct
<ssvb> kadamski: Allwinner A64 sets the OpenRISC core in the bootloader and it is supposed to be used by the firmware
<ssvb> kadamski: the firmware already provides PSCI interface for the use by the kernel (for bringing up secondary cores), and it looks like there will be SCPI interface covering the power management tasks in the future too
<kadamski> ok, so you're talking about newer SoCs, right?
<kadamski> but what about older ones?
<ssvb> I'm talking about the software
<kadamski> oh, and this can be done for H3 too?
<ssvb> most likely yes
gzamboni has joined #linux-sunxi
IgorPec has quit [Quit: Nettalk6 - www.ntalk.de]
<wens> anyone against disabling CONFIG_LEGACY_PTYS?
<wens> mripard: let me know when you fixed up the merge, i redo a merge of sunxi-next
<mripard> it's done
<wens> cool
<kadamski> mripard: continuing our discussion from yesterday, I've found that u-boot configures SMTA so that all peripheras are NOT secured (https://github.com/u-boot/u-boot/blob/master/arch/arm/include/asm/arch-sunxi/tzpc.h), so I should probably still have acess to this register, right?
<ssvb> kadamski: you can still boot the kernel in secure mode for testing purposes by just adding "setenv bootm_boot_mode sec" to boot.cmd
<ssvb> kadamski: that is if you have some doubts about accessing some register and want to experiment with it
<wens> kadamski: it's possible the code is not correct, or some peripherals are "secure", or non-configurable
<kadamski> ssvb: that's usefull, I'll defenity try it
<kadamski> wens: I was trying accessing CPU_CLK_GATING_REG
<wens> or it might be the prcm has something similar to the CCU, a switch to allow non-secure access
<kadamski> I can do that from u-boot console
<wens> u-boot runs secure
<kadamski> but can't do that using /dev/mem & mmap from linux
premoboss has joined #linux-sunxi
<kadamski> but you're talking about some other switch than SMTA?
<wens> kadamski: yeah, see arch/arm/cpu/armv7/sunxi/clock_sun8i_a83t.c:clock_init_safe
<kadamski> oh, you mean CCU_SEC_SWITCH_REG?
<wens> yes
<wens> though it is found on a83t
<wens> i'm saying maybe there's something similar in the PRCM
<wens> we can only guess
<kadamski> ah, that may be it - BUS_SEC bit, from Manual - "Includingg AXI/AHB/APB relevant registers,such as CPUX/AXI Configuration
<kadamski> Register,AHB1/APB1 Configuration Register,APB2 Configuration Register,
<kadamski> AHB2 Configuration Register.
<kadamski> however CPU_CLK_GATING_REG is part of CPUCFG module so maybe it's not that
<kadamski> but that might as well be undocumented
yann|work has joined #linux-sunxi
ganbold has joined #linux-sunxi
ricardocrudo_ has joined #linux-sunxi
apritzel has joined #linux-sunxi
<wens> apritzel: how do you use earlycon?
<apritzel> earlycon=uart,mmio32,0x01c28000
<apritzel> wens: was that what you wanted to know?
<wens> iirc documents say if you dont include the last "options" bit, it doesn't get initialized
<wens> but if i add ,115200n8 , it gives me garbage :/
<apritzel> wens: wrong clock, then?
<wens> maybe a few erroneous in the front screwing up the rest
<wens> not sure
<apritzel> I think in my first DT I gave it the 24MHz fixed clock directly and that worked
<wens> after it boots up it works fine
<wens> just the initial bits are bad
<wens> but that does bad things when one is using a multibyte encoding :(
<apritzel> wens: could it be due to this mysterious clock rate doubling we see on the MMC?
<wens> doubtful
<wens> i'm not using it on the a64
<wens> it's for my a23 tablet, so i can use one image for all my boards
<wens> was previously using earlyprintk
naobsd has quit [Quit: naobsd]
tomboy64 has quit [Quit: Uhh ... gotta go.]
tomboy64 has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
iamfrankenstein has quit [Ping timeout: 248 seconds]
iamfrankenstein has joined #linux-sunxi
IgorPec has joined #linux-sunxi
iamfrankenstein has quit [Ping timeout: 272 seconds]
TheLinuxBug has quit [Ping timeout: 264 seconds]
valkyr1e has quit [Ping timeout: 264 seconds]
book` has quit [Ping timeout: 264 seconds]
pekka30 has quit [Ping timeout: 264 seconds]
wens has quit [Ping timeout: 264 seconds]
tomboy65 has joined #linux-sunxi
tomboy64 has quit [Ping timeout: 264 seconds]
premoboss has quit [Ping timeout: 264 seconds]
mzki has quit [Ping timeout: 264 seconds]
vpeter has quit [Ping timeout: 264 seconds]
vpeter has joined #linux-sunxi
wens has joined #linux-sunxi
mzki has joined #linux-sunxi
TheLinuxBug has joined #linux-sunxi
pietrushnic`away has quit [Ping timeout: 245 seconds]
premoboss has joined #linux-sunxi
tkaiser has quit [Ping timeout: 248 seconds]
book` has joined #linux-sunxi
pekka30 has joined #linux-sunxi
p1u3sch1 has quit [Ping timeout: 245 seconds]
p1u3sch1 has joined #linux-sunxi
valkyr1e has joined #linux-sunxi
caog has joined #linux-sunxi
tchiwam_ has quit [Ping timeout: 240 seconds]
tchiwam has joined #linux-sunxi
cosm has quit [Ping timeout: 248 seconds]
oliv3r_ is now known as oliv3r
cosm has joined #linux-sunxi
lemonzest has joined #linux-sunxi
Dodger78 has joined #linux-sunxi
<Dodger78> hi guys . has anyone analog audio running on Allwinner A10/A20 on mainline ? my cubietruck has the kernel tweaked for it (4.5 / next) but playing doesnt work, mpg123 complains about not able to output buffers
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<wens> check alsamixer
<wens> did you unmute?
<Dodger78> hmm one sec
<Dodger78> where can i see that ? will this cause the mpg123 buffer output error =
<Dodger78> the DACs say off , the power amplifier is adjustable
<wens> unmute some of them
<Dodger78> how can i do it ?
<Dodger78> im on console alsamixer
<wens> left / right to navigate, m to mute/unmute
<wens> or press ? for help...
specing has joined #linux-sunxi
<Dodger78> hmm ok , no error anymore , i will walk to my audio receiver. do i need to setup asound.conf ?
pietrelinux has joined #linux-sunxi
<Dodger78> ok, no sound playing . but no errors anymore
<pietrelinux> Hello everyone, I 'm with a tablet a10 ( inet 3f ) and I 'm looking for a lightweight desktop can anyone tell me if you know some good desktop ?
speakman_ is now known as speakman
<Dodger78> @wens , is my asound config now wrong ?
<Dodger78> hmm now i got Monooutput .... lets check the cable
<Dodger78> @wens , works , thanks alot !
IgorPec has quit [Ping timeout: 256 seconds]
Dodger78 has quit [Quit: Lost terminal]
rubensm has joined #linux-sunxi
tkaiser has joined #linux-sunxi
pietrushnic has joined #linux-sunxi
reinforce has joined #linux-sunxi
iamfrankenstein has joined #linux-sunxi
naobsd has joined #linux-sunxi
cnxsoft has quit [Quit: cnxsoft]
pietrelinux has quit [Quit: Page closed]
catphish has joined #linux-sunxi
cptG has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
<cptG> the uboot-noob asks himself: can I override bootargs from the command line? I tried setting it and entering "boot" but they don't seem to stick.
<WarheadsSE> Does you kernel have them compiled in?
<cptG> WarheadsSE: no, it doesn't, it loads boot.scr and takes whatever is in there - at least I can permanently configure kernel args that way it seems.
<cptG> I got it working by just manually invoking every step provided in boot.txt (load mmc 0:1 0x48000000 /boot/uImage; load mmc 0:1 0x43000000 /boot/script.bin; etc...) and it then took my bootargs. Scary shit for me but it does the job
<cptG> ...still I'm wondering. Should I modify my boot.scr to append an optional bootargs_misc variable to bootargs?
Dodger78 has joined #linux-sunxi
Gerwin_J has quit [Ping timeout: 248 seconds]
reinforce has quit [Ping timeout: 245 seconds]
<Dodger78> hi guys, sorry i got another question: after using latest mainline kernel i get problems with systemd stuck in loop because RTC isnt set. is there the possiblity to avoid that ? best guess is a boot.scr which adjust the hwclock and then reboots normal system, does any1 have a better idea ?
<WarheadsSE> cptG: with boot.scr or uEnv.txt, you aren't writing it to the flash
<WarheadsSE> its not permentantly stuck
premoboss has quit [Ping timeout: 240 seconds]
<Dodger78> oh ok its not an endless loop ? but i heard of waiting times like 30min
<WarheadsSE> Dodger78: system or systemd?
<Dodger78> ehrm ....
<Dodger78> i need to check that , basically its debian jessie
<WarheadsSE> jessie is default systemd, iirc
<Dodger78> think so , didnt change it
<WarheadsSE> Can you get _exact_ error message? I have never seen this behavior on Arch Linux ARM
<WarheadsSE> (not telling you to switch, and I have not magically just verified.)
reinforce has joined #linux-sunxi
<Dodger78> [ 3215.319286] systemd[1]: Time has been changed
<Dodger78> something like this, guess its PID1 like in this post :
<Dodger78> i could only access this system after attaching serial console , networking and system did not start
<Dodger78> (but i didnt wait for more than 5min)
<Dodger78> it only happenes after i remove power from my sys
<Dodger78> when i set the hwclock and reboot with attached power, all is ok
<Dodger78> happened recently, after i switched to linux-next kernel
<Dodger78> 4.4 -> current linux-next
<Dodger78> but 4.5rc1 is also failing
nemunaire_ has joined #linux-sunxi
bmeneg_ has joined #linux-sunxi
<Dodger78> @WarheadsSE , i have seen a post about Archlinux also
<WarheadsSE> humm
<WarheadsSE> I didn't thin that the A10/A20 have rtc's
<Dodger78> is it possible to set some fake- init time, im setting my system time with ntp anyways ?
matthias_bgg_ has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
zumbi_ has joined #linux-sunxi
<WarheadsSE> There are a few methods
Netlynx has joined #linux-sunxi
<Deskwizard> Dodger78: theres one in the link you posted btw...
<Dodger78> with the firstboot ?
<Deskwizard> ya... just drag down, theres a script fix.
<Deskwizard> no idea if it'll work for ya, but to me it sounds like its what you're looking for, but again, what do I know :P~
<Dodger78> the topic is overall funny, since there are alot systems without a RTC / power buffered clock
<WarheadsSE> We have something called fakeclock @ ALARM
<WarheadsSE> basically, on the way down, it records the time. Then when you boot, if the time is well outside of resonable, it sets the clock to the last known
<WarheadsSE> let me find the package..
<Dodger78> ok, somehow this didnt work for me. but anyway, a bootloop is the worst case , because the system is not configurable / accessible anymore
<Dodger78> should they change systemd to not being stuck ?
IgorPec has joined #linux-sunxi
<Dodger78> oh i already got that installed .... :-P but how to properly use it ?
bmeneg has quit [*.net *.split]
matthias_bgg has quit [*.net *.split]
solarnetone has quit [*.net *.split]
zumbi has quit [*.net *.split]
nemunaire_ is now known as nemunaire
<Deskwizard> never been hit with that issues for like 10 different reasons, but I can see how it can be painful
<WarheadsSE> systemctl enable?
<Dodger78> fake-hwclock.service ?
<WarheadsSE> yup
<WarheadsSE> covers everything to enable/get it started
<Dodger78> thx i give it a shot
<WarheadsSE> yup yup
<WarheadsSE> lol @ comments @ systemd issue 1143 ""Unfortunately, discussion on this has petered out on LKML with no clear solution, my patch is still in limbo. It is a real problem, and yes, it is a kernel problem, and it will only be completely solved when 32bit userspace is gone.""
* WarheadsSE facepalm
<Dodger78> :-) i reboot, then i get kicked off irc ( im on that sys ) will report
avph has quit [Ping timeout: 240 seconds]
solarnetone has joined #linux-sunxi
Dodger78 has quit [Ping timeout: 248 seconds]
avph has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
Dodger78 has joined #linux-sunxi
<Dodger78> hi guys , fake-hwclock worked i guess . thx
<Deskwizard> ftw :)
<tkaiser> Dodger78: In case you want to read a bit more about the issue you ran into: http://forum.armbian.com/index.php/topic/624-cubietruck-rtc-problem-with-jessie-kernel-433/?p=4478 ;)
<Dodger78> thx i will read that
<Dodger78> @wens, my cubietruck is outputting mono , i doubt that the jack is defect , am i missing another thing ?
<Dodger78> my receiver only plays one channel, but mpg123 says stereo
<Dodger78> solved it myself , but dunno why , i disabled the mixer DACs left / right in alsamixer rofl
<Dodger78> then the left channel was playing
Dodger78 has quit [Quit: Lost terminal]
aep has quit [Ping timeout: 252 seconds]
libv_ has joined #linux-sunxi
Turl has quit [Ping timeout: 252 seconds]
Turl has joined #linux-sunxi
libv has quit [Ping timeout: 240 seconds]
domidumont has quit [Read error: Connection reset by peer]
aep has joined #linux-sunxi
yann|work has quit [Ping timeout: 240 seconds]
dev1990 has joined #linux-sunxi
catphish has quit [Quit: Leaving]
libv_ is now known as libv
domidumont has joined #linux-sunxi
<KotCzarny> dodger: systemd is evil trojan to linux
vagrantc has joined #linux-sunxi
vagrantc has quit [Ping timeout: 250 seconds]
aep has quit [Ping timeout: 252 seconds]
orly_owl has quit [Ping timeout: 250 seconds]
aep has joined #linux-sunxi
vagrantc has joined #linux-sunxi
apritzel has quit [Ping timeout: 248 seconds]
<Deskwizard> *sigh* kodi, H3, h.265... simple yes or no please :P
<Deskwizard> I just hate amlogic so much :( lol
FDCX has quit [Remote host closed the connection]
caog has quit [Quit: Ex-Chat]
matthias_bgg_ has left #linux-sunxi ["Leaving"]
orly_owl has joined #linux-sunxi
arossdotme has joined #linux-sunxi
naobsd has quit [Read error: Connection reset by peer]
naobsd has joined #linux-sunxi
sunxi_fan1 has joined #linux-sunxi
mzki has quit [Quit: leaving]
_massi has quit [Quit: Leaving]
tomboy65 has quit [Ping timeout: 250 seconds]
jstein has joined #linux-sunxi
tomboy64 has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
naobsd has joined #linux-sunxi
reinforce has quit [Ping timeout: 245 seconds]
IgorPec has quit [Quit: Nettalk6 - www.ntalk.de]
JohnDoe_71Rus has quit [Quit: KVIrc KVIrc Equilibrium 4.2.0, revision: 42021, sources date: 20120701, built on: 2013-10-21 12:25:22 UTC 42021 http://www.kvirc.net/]
vagrantc has quit [Quit: leaving]
cptG has quit [Quit: Konversation terminated!]
jjwerner has joined #linux-sunxi
tomboy64 has quit [Quit: Uhh ... gotta go.]
<jjwerner> hey guys, wens: you said that power regulator is almost done. how did you go around the arisc firmware? I assume you communicate using RSB with the hope that you get rid of the binary blobs in bootloader ?
robogoat has quit [Ping timeout: 250 seconds]
leio has joined #linux-sunxi
robogoat has joined #linux-sunxi
vagrantc has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
interrobangd has joined #linux-sunxi
catphish has joined #linux-sunxi
reinforce has joined #linux-sunxi
cptG has joined #linux-sunxi
interrobangd has quit [Quit: Leaving]
ricardocrudo_ has quit [Remote host closed the connection]
vagrantc has quit [Ping timeout: 240 seconds]
domidumont has quit [Ping timeout: 246 seconds]
<cptG> I have not found anything on the wiki/googlenet - has anyone ever tried kexec'ing 3.4 from mainline?
<zoobab_> nice idea
interrobangd has joined #linux-sunxi
<cptG> yeah, I'd like to experiment with it a little and wanted to test the check if anyone has experience, saving myself some trouble. But I guess the best way to find out is just to try it ;)
<catphish> KotCzarny tkaiser - just tested the A20 Ethernet transmit only, it maxes out the gigabit line when its not simultaneously receiving
<catphish> so it seems it can do 1Gbps tx, and 1Gbps rx, just not at the same time
aep has quit [Ping timeout: 276 seconds]
<KotCzarny> catphish: how about eth tx with sata read?
aep has joined #linux-sunxi
<catphish> KotCzarny: honestly, i would have to spend a couple of days learning how sata works, but i'll see what I can do
<KotCzarny> uhum
<KotCzarny> can you adapt your fixes to linux?
<KotCzarny> another idea, can you check how other interrupts affect max tx
<catphish> KotCzarny: 1) i haven't actually fixed anything, all i've done is seeminly identified a hardware limitation, i haven't overcome it :(
<catphish> KotCzarny: 2) i dont know anything about interrupts yet, sorry
<catphish> unfortunately my understanding so far is highly focused on the ethernet
<KotCzarny> so eth is not full duplex in short?
<catphish> KotCzarny: well its better than half duplex, but not as fast as full duplex, i'd suggest the DMA bandwidth is about 200MB/s
<catphish> not as much as 2Gbps
<KotCzarny> hmm
<catphish> although this in no way explains the slow SATA writes other people have reported :(
<KotCzarny> dma is the bottleneck then?
<catphish> KotCzarny: actually i can't even prove that, it could just be that the GMAC has some other bottleneck
<catphish> i dont understand the silicon well enough to say for sure
<KotCzarny> maybe sata writes get interrupted or happen in too little chunks
<catphish> all i can tell you is that it can do 120MB/s Ethernet TX or RX, but only 200MB/s combined
<catphish> and when combined, its weighted towards RX, 120MB/s RX 80MB/s TX
<KotCzarny> not that bad for home uses
<catphish> i agree, its pretty impressive for the money
<catphish> i was initially disapointed because i was thinking "it can only do 666Mbps", but in reality, what it was doing was 1Gbps + 666Mbps, which is a lot of data
<catphish> KotCzarny: i'm sorry my findings don't immediately help the SATA issues, if i have time i'll do some tests streaming SATA => LAN and LAN => SATA
yann|work has joined #linux-sunxi
<catphish> but i'll have to learn how sata works first, i'm quite a noob at OS development i'm afraid
<KotCzarny> catphish: no worries, what you are doing is a good piece of applied experimentation
<catphish> KotCzarny: my project is to learn about ethernet hardware and build a low cost high performance router, so i'm pretty focused on the ethernet performance, and my code is useful at this point because it has no CPU overhead and so maxes out the ethernet hardware
iamfrankenstein has quit [Ping timeout: 248 seconds]
reinforce has quit [Quit: Leaving.]
mosterta has joined #linux-sunxi
petr_ is now known as petr
solarnetone has quit [Remote host closed the connection]
catphish has quit [Quit: Leaving]
solarnetone has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
iamfrankenstein has joined #linux-sunxi
iamfrankenstein has quit [Client Quit]
jjwerner has quit [Ping timeout: 252 seconds]
ricardocrudo has quit [Remote host closed the connection]
vagrantc has joined #linux-sunxi
paulk-collins has quit [Quit: Quitte]
iamfrankenstein has joined #linux-sunxi
jemk has quit [Quit: Lost terminal]
lamer14547135341 has joined #linux-sunxi
tkaiser has quit [Ping timeout: 272 seconds]
tomboy64 has joined #linux-sunxi
sunxi_fan1 has quit [Ping timeout: 272 seconds]
interrobangd has quit [Quit: Leaving]
tomboy64 has quit [Ping timeout: 272 seconds]
mosterta has quit [Ping timeout: 256 seconds]
buZz has joined #linux-sunxi
buZz is now known as Guest33192
Guest33192 is now known as buZz
lamer14547135341 has quit [Quit: jIRCii - http://www.oldschoolirc.com]
tomboy64 has joined #linux-sunxi