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
victhor__ has joined #linux-sunxi
victhor__ has quit [Ping timeout: 246 seconds]
victhor__ has joined #linux-sunxi
vagrantc has quit [Quit: leaving]
Andy-D has quit [Ping timeout: 256 seconds]
victhor_ has joined #linux-sunxi
victhor__ has quit [Ping timeout: 244 seconds]
phipli has quit [Ping timeout: 244 seconds]
libv_ has joined #linux-sunxi
victhor__ has joined #linux-sunxi
victhor_ has quit [Ping timeout: 244 seconds]
libv has quit [Ping timeout: 248 seconds]
libv has joined #linux-sunxi
libv_ has quit [Ping timeout: 250 seconds]
libv_ has joined #linux-sunxi
victhor_ has joined #linux-sunxi
libv has quit [Ping timeout: 260 seconds]
victhor__ has quit [Ping timeout: 240 seconds]
victhor__ has joined #linux-sunxi
victhor_ has quit [Ping timeout: 246 seconds]
victhor_ has joined #linux-sunxi
libv has joined #linux-sunxi
victhor__ has quit [Ping timeout: 246 seconds]
libv_ has quit [Ping timeout: 250 seconds]
victhor_ has quit [Ping timeout: 245 seconds]
libv_ has joined #linux-sunxi
libv has quit [Ping timeout: 250 seconds]
victhor_ has joined #linux-sunxi
yann has quit [Ping timeout: 260 seconds]
victhor_ has quit [Ping timeout: 250 seconds]
terra854 has joined #linux-sunxi
Ntemis has quit [Remote host closed the connection]
Ixnus has joined #linux-sunxi
<Ixnus> Please reduce the off-topic, the log is nightmare to read lately.
<Ixnus> For all the many people who read it to stay up to date with sunxi mainlining...
Ixnus has quit [Quit: Page closed]
<apritzel> alright, my OPi PC 2 boots now upstream U-Boot from SPI, including ATF
perr has quit [Ping timeout: 245 seconds]
<apritzel> only problem: it's that nasty libdram SPL, so I can't share an image that others could just flash onto it :-(
libv has joined #linux-sunxi
libv_ has quit [Ping timeout: 256 seconds]
Colani1200 has joined #linux-sunxi
<wens> apritzel: so i found out the GIC address programmed into the A80 CBAR was incorrect
<apritzel> wens: haha
<wens> apritzel: after fixing it in u-boot, it seems all interrupts jump somewhere else and no longer return to the kernel
<apritzel> wens: this CBAR is such a simple thing, but seemingly everyone seems to mess it up
libv_ has joined #linux-sunxi
kaspter has joined #linux-sunxi
Colani1210 has quit [Ping timeout: 250 seconds]
<wens> apritzel: it's set as part of synthesis?
<wens> apritzel: the value they had was valid for the other socs... looks like they changed the memory layout but forgot to update CBAR
<wens> makes me wonder what else they forgot
<apritzel> wens: yes, I think Samsung got it wrong on the Arndale board as well
<wens> no wonder the #define in u-boot
libv has quit [Ping timeout: 265 seconds]
<apritzel> wens: yeah, I needed to add the config symbol to overwrite it: http://git.denx.de/?p=u-boot.git;a=commitdiff;h=16212b594f385bd594d5d316bf11b13c1186e3d7
kaspter has quit [Ping timeout: 260 seconds]
<wens> oh, that was you!
<apritzel> yeah, my early sins, I was totally clueless at this time - and still am, actually, but don't tell anyone ;-)
libv has joined #linux-sunxi
<apritzel> wens: do you have a repo with your current code somewhere?
<apritzel> wens: maybe I can take a look tomorr^Wtoday
<apritzel> bedtime here for now
<wens> it's pretty much standard power sequencing for allwinner
libv_ has quit [Ping timeout: 248 seconds]
<wens> was working until i fixed the gic address, but before that the processor cores would jump back to secure mode, which isn't good either :/
apritzel has quit [Ping timeout: 256 seconds]
alexxei has quit [Ping timeout: 244 seconds]
zzeroo has quit [Ping timeout: 252 seconds]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
ninolein has quit [Ping timeout: 260 seconds]
ninolein has joined #linux-sunxi
whaf has quit [Quit: Leaving.]
terra854 has quit [Quit: Connection closed for inactivity]
whaf has joined #linux-sunxi
terra854 has joined #linux-sunxi
zzeroo has joined #linux-sunxi
pg12 has quit [Ping timeout: 245 seconds]
whaf has quit [Ping timeout: 260 seconds]
pg12 has joined #linux-sunxi
vagrantc has joined #linux-sunxi
paulk-collins has quit [Ping timeout: 260 seconds]
<robogoat> I don't know who has the H5, but is anyone looking at booting it into secure mode?
<robogoat> I don't have one yet, but would be interested if that were available.
motlib has joined #linux-sunxi
[7] has quit [Ping timeout: 260 seconds]
JohnDoe_71Rus has joined #linux-sunxi
TheSeven has joined #linux-sunxi
jernej has joined #linux-sunxi
whaf has joined #linux-sunxi
jernej has quit [Ping timeout: 265 seconds]
IgorPec has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
kaspter has joined #linux-sunxi
whaf has quit [Quit: Leaving.]
reinforce has joined #linux-sunxi
<KotCzarny> beeble: you are scaring me.
jernej has joined #linux-sunxi
zzeroo has quit [Quit: WeeChat 1.3]
zzeroo has joined #linux-sunxi
mpmc has quit [Ping timeout: 252 seconds]
jernej has quit [Ping timeout: 258 seconds]
mpmc has joined #linux-sunxi
perr has joined #linux-sunxi
perr has quit [Changing host]
perr has joined #linux-sunxi
lkcl has quit [Ping timeout: 244 seconds]
IgorPec has quit [Ping timeout: 256 seconds]
premoboss has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
<wens> beeble: does emmc need pull-ups as well?
premoboss has quit [Ping timeout: 246 seconds]
f0xx has joined #linux-sunxi
vagrantc has quit [Ping timeout: 268 seconds]
Putti has quit [Ping timeout: 252 seconds]
lkcl has joined #linux-sunxi
IgorPec has joined #linux-sunxi
<KotCzarny> ~sdkvm
<KotCzarny> oops. wrong chan
massi has joined #linux-sunxi
whaf has joined #linux-sunxi
perr has quit [Quit: Leaving]
whaf has quit [Ping timeout: 245 seconds]
whaf has joined #linux-sunxi
whaf has quit [Client Quit]
whaf has joined #linux-sunxi
fkluknav has joined #linux-sunxi
<beeble> wens: emmc do need pull ups too. there sre pull ups included in the emmc so it can be sufficient to just have those. but i would activate the pull ups on the cpu outputs too just to be sure. the increase in current consumption is neglectable and having stronger pull ups can improve bus conditions
paulk-collins has quit [Remote host closed the connection]
<wens> thanks, a grep through sunxi-boards show all of them enable it
<wens> mripard: any reason to explicitly reset the tcon at component bind time?
<wens> mripard: i'm just looking through the drm driver
<wens> mripard: i enabled the display pipeline by default on sun6i, but with tcon disabled
<wens> and now it will kick out simplefb, try to install a new framebuffer, and fail :|
leviathanch has joined #linux-sunxi
terra854 has quit [Quit: Connection closed for inactivity]
<mripard> wens: the two choices we have are basically either resetting it or reading its state from the hardware and constructing the DRM state from that
<mripard> resetting is easier
<mripard> and removing the reset will not help anyway
<mripard> you don't have a way to switch to the new display atomically
premoboss has joined #linux-sunxi
<mripard> so you'll always have situations like this
florianH has joined #linux-sunxi
perr has joined #linux-sunxi
<wens> mripard: ok, so blank slate i guess
<wens> makes sense
matthias_bgg has joined #linux-sunxi
<jski> opinions on the pi pc plus H3?
<jski> sorry the plus 2E
<KotCzarny> good one, but you might want to hold on and wait for newer version based on h5
<KotCzarny> also, when you ask for an opinion, please state your desired use case
<jski> just the reliabilty of the board. the feature set looks good
<KotCzarny> the usual xunlong quality, quite good for the moneys, big board also helps with heat dissipation
motlib has quit [Ping timeout: 260 seconds]
whaf has quit [Ping timeout: 260 seconds]
whaf has joined #linux-sunxi
orly_owl has quit [Quit: leaving]
<wens> mripard: another question, are we supposed to create our framebuffer before or after drm_dev_register?
yann|work has joined #linux-sunxi
massi has quit [Ping timeout: 248 seconds]
Putti has joined #linux-sunxi
massi has joined #linux-sunxi
yann-kaelig has joined #linux-sunxi
<mripard> "drivers must perform all initialization before calling drm_dev_register()"
<mripard> in the drm_dev_register kerneldoc
<wens> so the vc4 driver is doing it wrong
<wens> anyway, was mostly curious
perr has quit [Quit: Leaving]
bugzc has quit [Ping timeout: 260 seconds]
Putti has quit [Ping timeout: 248 seconds]
Pepe has joined #linux-sunxi
fvogt has joined #linux-sunxi
<MoeIcenowy> mripard: still cannot find any reason for my sun4i-drm issue on sun8iw5?
kaspter has quit [Ping timeout: 260 seconds]
<mripard> MoeIcenowy: I already told you where to look at, I can't do more than that without the hardware
<MoeIcenowy> ah-oh
<MoeIcenowy> do you have any general q8 tablet?
<mripard> no
<KotCzarny> make a ssh account for him and let him play?
<MoeIcenowy> but he cannot see what appeared on the screen...
<KotCzarny> camera + webserver
<MoeIcenowy> ah-oh
<MoeIcenowy> I have no webcam :-(
<KotCzarny> no cams in phones/tablets?
<MoeIcenowy> mripard: can you list the points that is possible to be responsible for the issue?
<KotCzarny> i bet you have plethora
<MoeIcenowy> I can test them on by one
<MoeIcenowy> one by one *
<mripard> MoeIcenowy: can you tell me what the issue is again?
<MoeIcenowy> mripard: After restart the TCON by switch back from Xorg to fbcon, the screen is blank, and TCON cannot receive vblank interrupt
<mripard> then try to see if you have an interrupt in the first place
<mripard> if you have, if it's properly reporting the vblank event
<mripard> and if it does, if the disable sequence is working properly
mzki has joined #linux-sunxi
alexxei has joined #linux-sunxi
<MoeIcenowy> mripard: thx... got it
<MoeIcenowy> for question one, according to drm debug info, the first start of TCON have interrupt reports
<mripard> I meant when you disable it
<mripard> just put in printk in the interrupt handler
<mripard> the drm debug info is output when you have reported the vblank event
<mripard> so it doesn't help for 1 and 2
<MoeIcenowy> or check the irq count in sysfs?
popolon has joined #linux-sunxi
<mripard> you're not that fast
IgorPec has quit [Quit: Nettalk6 - www.ntalk.de]
Ntemis has joined #linux-sunxi
pg12 has quit [Ping timeout: 258 seconds]
mhlavink has quit [Ping timeout: 260 seconds]
yann|work has quit [Ping timeout: 256 seconds]
pg12 has joined #linux-sunxi
sW` has quit [Ping timeout: 260 seconds]
sW` has joined #linux-sunxi
perr has joined #linux-sunxi
gzamboni has quit [Ping timeout: 248 seconds]
mhlavink has joined #linux-sunxi
fvogt has quit [Ping timeout: 258 seconds]
whaf has quit [Quit: Leaving.]
fvogt has joined #linux-sunxi
fvogt has quit [Ping timeout: 240 seconds]
fvogt has joined #linux-sunxi
dlan has quit [Ping timeout: 256 seconds]
fkluknav has quit [Quit: Leaving]
dlan has joined #linux-sunxi
diego71 has quit [Ping timeout: 268 seconds]
fkluknav has joined #linux-sunxi
diego71 has joined #linux-sunxi
alexxei has quit [Ping timeout: 260 seconds]
alexxei has joined #linux-sunxi
Gerwin_J has quit [Read error: Connection reset by peer]
massi has quit [Ping timeout: 252 seconds]
mhlavink has quit [Ping timeout: 245 seconds]
matthias_bgg has quit [Quit: Leaving]
massi has joined #linux-sunxi
mhlavink has joined #linux-sunxi
Christos_ has joined #linux-sunxi
Christos_ has quit [Client Quit]
<jelle> oh hmm drive by dump :P. It's already here though https://linux-sunxi.org/H5
yann|work has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
<MoeIcenowy> mripard: seems that one VBLANK IRQ is received
<MoeIcenowy> I mean, entered sun4i_tcon_handler
<MoeIcenowy> and, theortically it executed drm_crtc_handle_vblank
Gerwin_J has quit [Ping timeout: 264 seconds]
tkaiser has joined #linux-sunxi
<tkaiser> Hehe, seems the voltage regulation stuff with H5 is really done by 'axpdummy' driver: https://github.com/OrangePiLibra/OrangePi_H5SDK/commit/039ab08730f38bcff328ffd5bac59fb64aece327
<jelle> shouldn't we added H5 to the table http://linux-sunxi.org/Mainlining_Effort ?
<wens> jelle: add it when there's actual wip stuff on it?
<KotCzarny> jelle: was it booted already at all with mainline?
<MoeIcenowy> wens: apritzel is WIP on it
<jelle> KotCzarny: no I tried aprtizel's u-boot yesterday
<KotCzarny> then no is the answer i guess
<wens> MoeIcenowy: thought that was u-boot
<KotCzarny> unless you ewant to add all grays
<jelle> wens: yes u-boot
<MoeIcenowy> he has also bootable kernel
<jelle> oh ofcourse this is the kernel matrix
<MoeIcenowy> although without any new driver
<wens> then there's really nothing?
<jelle> yup
<KotCzarny> its a hybrid between h3/a64, so it will definitely work one day
<jelle> wens: well I would have added 'NO' everywhere
paulk-collins has joined #linux-sunxi
<jelle> let me atleast add apritzel's branch to the orange pi pc 2 wiki
<MoeIcenowy> mripard: drm_crtc_handle_vblank do returns
<MoeIcenowy> but still "vblank wait time out"
<MoeIcenowy> but this time only one irq is generated, compared with the irq flood in the first TCON start
<wens> jelle: you can do that for any of the unsupported allwinner socs, it doesn't help
<mripard> MoeIcenowy: I'm sorry, I don't have the time to deal with this today
<jelle> wens: how does it not help?
<jelle> wens: oh wait sorry
<wens> jelle: add it when something is actually WiP or finished
<jelle> wens: yes I will, sorry I thought you replied on my last message
<KotCzarny> for now it will be 'uboot: WIP'
<KotCzarny> ;)
<jelle> KotCzarny: there is no u-boot column ;-)
<wens> u-boot is a separate page, which kind of needs some love
<wens> bbl
<KotCzarny> maybe add similar table of support in uboot?
<KotCzarny> if you really want to add anything
<KotCzarny> :)
<jelle> uboot needs a maintainer :S
<KotCzarny> network/usb/emmc/sd etc
<MoeIcenowy> mripard: ok. I will dig with it.
<KotCzarny> though in case of uboot its more about per-board, than per soc
<MoeIcenowy> we need at least per soc driver support ;-)
<jelle> hmm H3 eth in u-boot is broken hmm
alexxei has quit [Read error: Connection reset by peer]
Putti has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
alexxei has joined #linux-sunxi
<NiteHawk> jellle: is it? i tried v2016.11 on opipc and at least got an IP via dhcp
<KotCzarny> is there any h3 device with gigabit ?
fl__0 has joined #linux-sunxi
<jelle> NiteHawk: will try it tonight and update the u-boot page
<KotCzarny> bpi m2+/m3
<jelle> NiteHawk: since I believe it worked
<KotCzarny> and opipc2/opi+/opi+2/opi+2e
fl_0 has quit [Ping timeout: 248 seconds]
lemonzest has joined #linux-sunxi
Gerwin_J has quit [Ping timeout: 264 seconds]
sunxi_fan has left #linux-sunxi [#linux-sunxi]
gumblex has joined #linux-sunxi
fl__0 is now known as fl_0
gumblex has quit [Ping timeout: 245 seconds]
mhlavink has quit [Ping timeout: 258 seconds]
mhlavink has joined #linux-sunxi
nikre has joined #linux-sunxi
<nikre> is there an h5 with sata2 and gbit?
<jelle> nikre: from the datasheet it shows that sata is via usb
<jelle> so you wouldn't want that :p
<tkaiser> jelle: Where is SATA mentioned?
<nikre> jelle, is the gbit utilized fully?
<nikre> on opi pc for example
<nikre> pc2
<jelle> oh fail
<jelle> nikre: haven't tested that yet
<jelle> tkaiser: I'm confused it seems
<nikre> tkaiser, do u know?
<tkaiser> nikre: I'm just writing a script to test through all TX/RX combinations. In RX direction we already get above 900 Mbits/sec
<tkaiser> nikre: And I'm pretty confident to exceed the 400 Mbits/sec I got in the other direction
<jelle> was there a h2 datasheet?
<tkaiser> nikre: Do the math: Every test will take 3 minutes and I test through 32*32 combinations ;)
<nikre> tkaiser, doesnt that imply as fast hdd communication?
<tkaiser> nikre: No, why should it? OPi PC 2 can also boot from fast iSCSI storage. Bootloader in SPI flash, everything else on another machine
<nikre> is that pxe boot?
<tkaiser> nikre: But there's also btrfs transparent file compression and stuff like btrfs mirrors
<jelle> > USB-SATA SATAII 3.0GBps
<tkaiser> s/mirror/stripe/
<nikre> so is sata2 implementable but not implemented on opi pc2?
<tkaiser> nikre: There is no SATA with H5 and regarding SPI NOR flash search in our wiki for exactly that
<jelle> nikre: USB SATA
<jelle> you don't ever want that :P
<tkaiser> jelle: You want that but only with good USB-to-SATA bridges that feature SMART, TRIM and UASP
<nikre> why would they write 3.0Gbps when it is limited to usb2 ?
<nikre> or is it not limited?
<jelle> nikre: no idea
<jelle> nikre: ask allwinner :P
<nikre> also GBps is gigabyte ps, should be Gbps
<nikre> looks more like xunlong's mistake
<jelle> might be too
<tkaiser> nikre: Because the SATA part of an USB to SATA bridge talks SATA. And now let's please stop this off-topic trash talk so IRC log doesn't look like the last three days!!
<nikre> what is trash about this
<jelle> tkaiser: ++
whaf has joined #linux-sunxi
<KotCzarny> nikre: marketing
<KotCzarny> dont trust anything in the specs
<KotCzarny> or marketing info
<KotCzarny> trust real use cases
<tkaiser> KotCzarny: BS, the specs are correct in this case. But anyway, let's stop flooding this *developer channel* with BS like that!
Gerwin_J has joined #linux-sunxi
<KotCzarny> tkaiser: sure, sata 3.0Gbps suggests something else
matthias_bgg has joined #linux-sunxi
<KotCzarny> its just not saying what's on the other side of the wire
<nikre> tkaiser, implied 3.0gbps reachable i guess but i'm still confused
<tkaiser> nikre: SATA II, that's just a mode/dialect. A20 SoC and R40 now speak also SATA II. And are pretty slow. So what?!
<nikre> and words like trash!! BS!! should be less frequent if you are into keeping the logs clean. tyvm.
<tkaiser> nikre: Yes, will stop here, no more info for you (about negotiation and stuff like that)
<nikre> the answer for "So what?!" is pretty obvious.
perr has quit [Quit: Leaving]
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<willmore> Does anyone 3d print their own cases? If so, does anyone have a favorite design for the Opi PC 2 (same as the PC, I would guess).
Pepe has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
libv_ has joined #linux-sunxi
afaerber has joined #linux-sunxi
gumblex has joined #linux-sunxi
libv has quit [Ping timeout: 258 seconds]
libv has joined #linux-sunxi
libv_ has quit [Ping timeout: 264 seconds]
afaerber has quit [Ping timeout: 245 seconds]
afaerber has joined #linux-sunxi
libv_ has joined #linux-sunxi
sunxi_fan has joined #linux-sunxi
<nikre> willmore, i had printed a case for opi pc. there used to be a single 3d model at that time and i don't recommend using that 3d model
<nikre> this one not recommended http://www.thingiverse.com/thing:1068457
libv has quit [Ping timeout: 264 seconds]
cnxsoft has quit [Read error: Connection reset by peer]
cnxsoft has joined #linux-sunxi
<naobsd> hmm
<naobsd> is cdc ethernet not supported on mainline u-boot (sunxi musb)?
jstein__ has joined #linux-sunxi
jstein is now known as Guest18891
<KotCzarny> its amazing that hdmi is supported at all
<KotCzarny> ahm, you talk about usb
jstein__ is now known as jstein
<silviop> i'm using kernel 4.9 on A33 , there is a guide to add rtl8703AS support ? semms that is not on mainline kernel
libv_ is now known as libv
<mripard> naobsd: it is, but it's slightly hackish
<mripard> and I haven't had the time to test it yet
<mripard> let me push the branch
Guest18891 has quit [Ping timeout: 245 seconds]
cnxsoft has quit [Quit: cnxsoft]
<naobsd> if cdc ether and netconsole work, it's nice for NES classic ;)
<naobsd> I'm trying to prepare initial patch for u-boot
<NiteHawk> they should have named that "BPi-Mk Ultra" :D
<mripard> naobsd: ^
<mripard> that won't go upstream as is though
<MoeIcenowy> silviop: follow the guide of 8723bs
<MoeIcenowy> Hans de Goede had a tutorial
<silviop> i have problem with
<silviop> #ifndef CONFIG_WIRELESS_EXT
<silviop> #error CONFIG_WIRELESS_EXT needs to be enabled for this driver to work
<silviop> #endif
<silviop> in menuconfig is not configured , but it depends of toshiba vebdor that depend from pcie , and my tablet does not have that
<naobsd> mripard: thanks
<MoeIcenowy> silviop: you can try to set it
<MoeIcenowy> I don't think it will depend on pcie
<MoeIcenowy> I have also an A33 tablet w/ 8723BS, which works ;-)
<silviop> make -j4 ARCH=arm CROSS_COMPILE=arm-linux-gnueabihf- KSRC=../linux-4.9-rc5/
<silviop> on rtl8723bs
<MoeIcenowy> silviop: set it in the kernel
<MoeIcenowy> then rebuild the kernel
<MoeIcenowy> then build rtl8723bs
<silviop> no i cannot set
<MoeIcenowy> silviop: no you can set it
<silviop> No depends from :
terra854 has joined #linux-sunxi
<MoeIcenowy> silviop: just set it
<MoeIcenowy> do not care its dependencies
<silviop> is setted from a big expression that override .config
<silviop> i enable all i thnk shuld be right
<MoeIcenowy> without it you cannot just use rtl8723bs
<silviop> but it remain unsetted
<silviop> probably you have link random enable another device that enable expression
<MoeIcenowy> mripard: sorry, I wrongly anaylzed the info from the dmesg
<MoeIcenowy> the last IRQ is before TCON shutdown
<MoeIcenowy> after TCON restart there's no IRQ
Andy-D has joined #linux-sunxi
<silviop> Staging + rt8188eu seems enable WIRELESS_EXT !!!
mzki has quit [Ping timeout: 245 seconds]
<silviop> It needed to be documented
<silviop> and now dtb ??
<silviop> where can i take a dtb example for my q8?
mzki has joined #linux-sunxi
deskwizard has quit [Read error: Connection reset by peer]
<montjoie> update sun8i-ce with A64 support (A80/A83T is coming)
<MoeIcenowy> silviop: I think current upstream q8 dts have already enabled sdc1
<MoeIcenowy> check whether there's a sdio card recognized
<MoeIcenowy> mripard: the TCON failed to be restarted at all
<tkaiser> longsleep: Do you remember whether you did test through all possible TX/RX delay combinations with Pine64+ back then?
<longsleep> tkaiser: i did not test all, only until it started failing and the default values
<tkaiser> longsleep: Ok, I do now a brute-force attempt and suggested the same to an BPi-M64 owner (there GbE sucks but that's no wonder given that they've overtaken Pine64 settings blindly)
<mripard> MoeIcenowy: that would explain why it doesn't work :)
<MoeIcenowy> mripard: I checked the regmap, and the GCTL register is 0
<MoeIcenowy> which indicates the TCON is not enabled
<mripard> indeed
<wens> i think i sent a patch for that just today :)
<wens> hmm, probably not
deskwizard has joined #linux-sunxi
<MoeIcenowy> mripard: but this time it's 0x80000000, and the vblank still fails...
<mripard> what about the interrupt mask /
<MoeIcenowy> 0xc0000000, as expected
whaf has quit [Quit: Leaving.]
<MoeIcenowy> wens: where's your patch?
Ntemis has quit [Remote host closed the connection]
hojnikb has joined #linux-sunxi
<hojnikb> brute forcing gbe isn' that bad of an idea
<hojnikb> it would take like 2.5days to get thru all the combinations
<hojnikb> definitely doable
<hojnikb> i can do it, when my board arrives
Pepe has joined #linux-sunxi
<hojnikb> just give me a working image and script
<tkaiser> hojnikb: Already running, should be finished on Friday.
<hojnikb> great
<hojnikb> shame on xunlong for not doing this
<hojnikb> this is really basic stuff
<tkaiser> hojnikb: Can we please stop bashing/trolling here :)
<hojnikb> i'm not trolling :)
<hojnikb> just keeping this channel alive xD
Pepes has joined #linux-sunxi
<KotCzarny> hmm, one cute idea might be slapping socket onto opipc2 and make it into spi burner machine
Pepe has quit [Client Quit]
<hojnikb> you can get a spi flasher for like 4$
<MoeIcenowy> You'd better use the socket on opi0
Pepes is now known as Pepe
<MoeIcenowy> hojnikb: one of my friends told me that dedicated flasher performs much worse than SBCs
<KotCzarny> opi0 costs ~8usd
<hojnikb> MoeIcenowy: i don't know, i had a pretty good experience with CH341A
<hojnikb> and it costs like 2$ now
<MoeIcenowy> So the problem is that the TCON just do not like sending vblank interrupts now?!
premoboss has quit [Ping timeout: 248 seconds]
<mripard> MoeIcenowy: you can try to compare the registers between the time it works and the time it doesn't
<MoeIcenowy> mripard: ok
<mripard> you can use the regmap debgufs interface to do that easily
reinforce has quit [Quit: Leaving.]
matthias_bgg has quit [Quit: Leaving]
<wens> MoeIcenowy: its unrelated
<MoeIcenowy> there's only one difference I found:
<MoeIcenowy> 0x044 (DCLK_CFG): 0xf0000007 when working, 0xf0000006 when not
<mripard> hmmm, that shouldn't make any difference
<mripard> at least at the interrupt level
<MoeIcenowy> yes
<hramrach> hello
<hramrach> I am considering building a BT audio DAC. For prototype something like Orange Pi or chip might work
<hramrach> but I have no idea what to use for the DAC
<hramrach> There is ton of BT chips out there so that side should not be that mcuh of a problem
* jelle recently bought a FC-215
<jelle> $4 USB DAC
<hramrach> that might work
<hramrach> I have an USB sound card lying somewhere
<jelle> but it has a silly usb B connector ;-)
<hramrach> I have some AXAGO stuff with A connector
matthias_bgg has joined #linux-sunxi
<hramrach> probably AXAGON ADA-15 mini hq
<hramrach> I will try it I guess
<jelle> haven't tested if I can power this thing from an orange pi yet
<hramrach> probably can
gumblex has quit [Quit: Bye.]
<hramrach> hm, google says fc215 is feed-through coax connector
<jelle> hramrach: I'm using the jack connector ;-)
<jelle> hramrach: there are probably other cheap alternatives on aliexpress
whaf has joined #linux-sunxi
d[^_^]b has quit [Ping timeout: 250 seconds]
<hramrach> potential problem with USB is CPU usage and latency of the chip itself. Given the other side is going to be BT it is only getting worse
<hramrach> I would hope something like i2s would put less load on the CPU but have no idea where to get decent i2s DAC board
whaf has quit [Quit: Leaving.]
alexxei has quit [Ping timeout: 260 seconds]
<KotCzarny> what is nice and well supported i2s module supported by mainline/sunxi ?
fkluknav has quit [Ping timeout: 240 seconds]
<hramrach> yes, I would like to know that as well
<MoeIcenowy> maybe my TCON is too lazy to send vblank interrupts :-(
<MoeIcenowy> mripard: can TCON work without vblank?
lamer14793112227 has joined #linux-sunxi
hojnikb has quit [Quit: Page closed]
<mripard> no
tkaiser has quit [Ping timeout: 258 seconds]
<fvogt> Quick question: Where can I find the most recent working version of a linux source tree with working USB for Pine64?
<fvogt> I found several ones based on 4.7, 4.9-rc2 and so on on github, but which one is the "best"?
<MoeIcenowy> my github
<miasma> jelle: fc-215? didn't find anything with that from ebay/ali/google
<hramrach> no idea what good these would be, though
<fvogt> MoeIcenowy: https://github.com/Icenowy/linux branch ice-a64-v6.1? I tried that one, it doesn't boot, while apritzel/linux does fine (same config)
jernej has joined #linux-sunxi
<MoeIcenowy> fvogt: could you give me some log about your boot failure?
<fvogt> MoeIcenowy: No output whatsoever, booted by GRUB2 over EFI
<fvogt> I'm not sure what to set to get some kind of earlyprintk, I couldn't find any info on that
<MoeIcenowy> fvogt: are you sure you are really using the same config and kernel cmdline?
<MoeIcenowy> My branch merges only some USB-related fixes
<fvogt> MoeIcenowy: Yes, but I had to disable ARMv8.1 atomics as they are broken in the 4.9rc2 kernel with GCC < 6.0
<fvogt> Could you maybe just upload a known working config of yours?
<MoeIcenowy> it's a really fat one... will you care this?
<MoeIcenowy> oh I cannot find the config now...
<MoeIcenowy> (recently I'm working on H3/A33, not A64.
<maz> fvogt: what makes you think v8.1 atomics are broken?
fkluknav has joined #linux-sunxi
<fvogt> MoeIcenowy: Big is not an issue, my current build is close to allmodconfig
reinforce has joined #linux-sunxi
<MoeIcenowy> https://pastebin.anthonos.org/view/5f4de27c it's my 4.7.0 one
<maz> fvogt: https://patchwork.kernel.org/patch/9411345/ (which is -rc5)
<fvogt> MoeIcenowy: That should work just fine with your ice-a64-v6.1 branch, which is at 4.9-rc2 IIC?
<MoeIcenowy> I think so
<maz> fvogt: (and the brokenness is not really related to the atomics at all)
<fvogt> maz: Thanks, I'll enable it again and apply the patch
<maz> fvogt: you'd be much better off fast-forwarding to -rc5.
<fvogt> maz: I'll try this first, as each modification might just cause bugs
<maz> fvogt: well, if you find regressions at -rc5, we definitely want to hear about them.
<maz> fvogt: staying at -rc2 is just asking for trouble.
<fvogt> maz: My last built kernel does exactly nothing after booting...
<maz> fvogt: have you tried a hammer? ;-)
<fvogt> maz: Where should I aim it? My head, the kernel building machine or the A64 SoC? ;-)
<maz> fvogt: definitely the A64. unless you're trying self-hosted debugging...
<fvogt> maz: Done, now I've got four A16s instead, yay
<fvogt> Hm, "arch/arm64/crypto/crc32-arm64.c:1:0: error: unknown feature modifier 'crc'" - that's a new one to me
<miasma> jelle: i see. that should be quite ok. i have the mini version with only 3.5mm jack. didn't think it would have this name 'fc-215' :D
alexxei has joined #linux-sunxi
Andy-D has quit [Ping timeout: 252 seconds]
<maz> fvogt: how old is your toolchain?
<hramrach> hm, the H5 chip even looks usable
<hramrach> but only USB2
<hramrach> always something missing
<lamer14793112227> hramrach: Add features, add costs ;)
<hramrach> make a more useful device
<fvogt> maz: I used gcc 4.8.5 for that build, I rebased it to rc5 now and will retry with gcc-6 instead (which I should've done in first place)
<hramrach> I have several useless boards already
lamer14793112227 has quit [Quit: jIRCii - http://www.oldschoolirc.com]
tkaiser has joined #linux-sunxi
<hramrach> USB3 seems to be becoming the universal high-speed peripherial bus
<hramrach> so they put it on A80 with its carappy graphics ... and remove it from h5 again
ornitorrincos has quit [Read error: Connection reset by peer]
<maz> fvogt: not a bad idea. it is also unlikely that your compiler has support for some of the A53 errata...
<hramrach> yes, and broken ARM cores, heh
<fvogt> maz: Those fixes were backported, but there are probably other bugs that weren't fixed...
premoboss has joined #linux-sunxi
whaf has joined #linux-sunxi
<nikre> hramrach, is there a board with usb 3.0 or sataII other than odroid-xu3?
<KotCzarny> espressobin
<nikre> ty KotCzarny
<KotCzarny> but you have missed the campaign (dont know if you can still order it via kickstarter)
<nikre> price is good too
<KotCzarny> but if you are interested in data moving around drives, marvel is better than anything else
gzamboni has joined #linux-sunxi
<tkaiser> nikre: USB 3.0 and SATA are nothing fixed, implementations differ a lot. Also the use case is important, databases love random IO and then an USB 3.0 implementation without UASP can be slower than a boring H5 since Allwinner SoCs are able to benefit from UASP even when only USB2 capable.
<KotCzarny> tkaiser: did you preorder espressobin?
<tkaiser> KotCzarny: Nope, have already another Marvell board here: Clearfog Pro
<tkaiser> nikre: In case you want to get into details: https://forum.armbian.com/index.php/topic/1925-some-storage-benchmarks-on-sbcs/ (there is i.MX6 quad also and depending on the use case a combination of fast eMMC and slow USB HDD can already outperform 'SATA II')
<nikre> is there a test that includes odroid-xu3?
<tkaiser> nikre: I don't have this board since fans suck. But XU3/XU4 with UAS enabled kernel simply rock
<nikre> ty
<tkaiser> nikre: In case you want to read that article, there's of course a link to results. But I think no one likes walls of text ;)
<terra854> Well, I would prefer a ARM64 version of the XU4
<nikre> i guess it deserves the money given. how come are there not other boards that use the same SoC?
<tkaiser> terra854: Since 64 is more than 32, true?
<terra854> Well yeah. I prefer 64-bit processors
<terra854> 32-bit is outdated
<fvogt> I built "4.9.0-rc5-aosc+" with gcc 6.2 now, result: Nothing at all :-/
<KotCzarny> in which way outdated?
<nikre> could xunlong decide to use a samsung soc as it wishes?
Ntemis has joined #linux-sunxi
<tkaiser> longsleep: Your getrandom hint did it: https://github.com/OrangePiLibra/OrangePi_H5SDK/issues/3
<premoboss> got a nanopi neo v1.0. happy work wit armbian legacy kernel 3.4.112. got also twice nanopi neo V1.1, still not used, but they seems to be an enanchement under heat dissipation viewpoint, afaik, they are les hotter. it is righth? later i see ther eare also nanopi neo v 1.2 but i dont know what difference vs v1.1.
<premoboss> someone can clarify me this point?
<fvogt> Argh, I found the issue: u-boot changed the name of the dtb file in version 2011.6, so it passed its internal DT to the kernel
<fvogt> *2016.11
<fvogt> But that only works for 4.7 and not 4.9, apparently
<tkaiser> premoboss: there's a wiki no one ever looks into: http://linux-sunxi.org/FriendlyARM_NanoPi_NEO_%26_AIR#Voltage_regulators_.2F_heat
<tkaiser> premoboss: It's fixed with PCB rev 1.1 and maybe improved with 1.2
<premoboss> i go to see link
matthias_bgg has quit [Quit: Leaving]
<premoboss> to make v1.0 no HOT i activate all tricks by H3control tool to reduce consume (turn off 3 cpu, slow down ram, ecc). not H3 is not so terrible hot.
<premoboss> that link is good, they sould place a reference into http://wiki.friendlyarm.com/wiki/index.php/NanoPi_NEO
<tkaiser> premoboss: Turning off CPU cores doesn't help with idle consumption, only peak consumption under full load affected. And it's a wiki so waiting for your edits
<premoboss> can no registered people editing wiki? if yes i can cotribute as well.
<tkaiser> premoboss: Speaking of FA or our wiki?
<premoboss> uhm, you are rignt, i am mixing to different sites.
<premoboss> i mean, can i contribute to Friendly arm site even if i am not registered user of it?
<fvogt> USB works now on my pine64, thanks for helping!
<tkaiser> premoboss: No idea, I already waste too much time in our wiki to check others ;) But FA people listen to suggestions.
<premoboss> ok
<premoboss> i uderstand it is necessary to avoid missuse of wiki access, but i dislike to register myself to sites.
jernej has quit [Ping timeout: 244 seconds]
fvogt has quit [Quit: Konversation terminated!]
simosx has joined #linux-sunxi
diego_r has quit [Ping timeout: 252 seconds]
petr has quit [Ping timeout: 250 seconds]
petr has joined #linux-sunxi
Nacho has quit [Remote host closed the connection]
massi has quit [Quit: Leaving]
BigFellow has joined #linux-sunxi
phipli has joined #linux-sunxi
Nacho__ has joined #linux-sunxi
popolon has quit [Quit: WeeChat 1.4]
cptG has joined #linux-sunxi
The_Loko has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 260 seconds]
cptG_ has quit [Ping timeout: 240 seconds]
mossroy has joined #linux-sunxi
afaerber has quit [Quit: Ex-Chat]
Nacho__ has quit [Remote host closed the connection]
Nacho has joined #linux-sunxi
simosx has quit [Remote host closed the connection]
simosx has joined #linux-sunxi
nikre has quit [Quit: Leaving]
phipli has quit [Ping timeout: 240 seconds]
berenm has joined #linux-sunxi
premoboss has quit [Ping timeout: 260 seconds]
scream has joined #linux-sunxi
vagrantc has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
HeavyMetal has joined #linux-sunxi
HeavyMetal has joined #linux-sunxi
HeavyMetal has quit [Changing host]
<tkaiser> terra854: This should enable OPi PC 2 to be 'switched on' by power key. The process with Allwinner BSP kernel is that a 'poweroff' sets a flag, then the board reboots and BSP u-boot is sitting around and waiting for events (a 'key' like power button, WiFi, BT, IR, whatever).
<tkaiser> tkaiser: Should work with A64 BSP kernel too
<KotCzarny> does it work on a20/h3 too?
BigFellow has quit [Quit: BigFellow]
<tkaiser> With H3 it works but then say good-bye to mainline u-boot and welcome 2011.09 ;)
<KotCzarny> heh
<tkaiser> No idea about A20 though, it's both a BSP kernel flag and u-boot support is also needed of course
<KotCzarny> would be nice to port to mainline uboot
<tkaiser> KotCzarny: Go! ARISC stuff you know ;)
<KotCzarny> let me check the drawer with arisc softwares, right, got it! ;)
<KotCzarny> android build system is funny: 13984 root 20 0 5299412 1.741g 9552 D 7.5 88.7 1:26.64 ld.gold
<KotCzarny> 1.7GB and rising
Nacho has quit [Remote host closed the connection]
Nacho has joined #linux-sunxi
<tkaiser> terra854: You could give it a try with longsleep's kernel and if it works send a PR to longsleep.
smooker has joined #linux-sunxi
phipli has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
JohnDoe_71Rus has quit [Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/]
<miasma> tkaiser: does it use much power while waiting for events in the u-boot ?
yann|work has quit [Ping timeout: 256 seconds]
<tkaiser> miasma: Did no measurements now but when I discovered that with H3 in March it was pretty low but also my powermeter not precise enough. Will test that later maybe. But with H5 then.
<miasma> tkaiser: so it requires teh bsp kernel? does not work with armbian?
<miasma> i could measure it later this month
<KotCzarny> amazing. seems android did build. and its 360M, going to check tomorrow if it works
<tkaiser> miasma: With Armbian's H3 legacy kernel it should (still) work. Back in March I tested around with a desktop image (maybe the first time in my live I connected a SBC to a display) and was curious what happens when you choose 'suspend' in the UI. Well that happened and a press on the power button woke up the whole system from 'syspend to RAM'
<miasma> nice
<tkaiser> Tried it then with an A20 device and sleeping worked but resuming not (no key mapping defined in fex file was the reason IIRC)
<miasma> if it's really low power, i could use it in my car pc :)
<tkaiser> miasma: In fact that's how all those OTT boxes implement 'power off', they enter low-power state and then a Cortex-M or in our case the OpenRISC core idles around and waits for events. Eg IR remote
<miasma> yes, that's a sane way to do it
<tkaiser> Or Bluetooth. On Opi Zero WoWLAN should be possible too.
<tkaiser> Then the Cortex-A7 are sleeping and only the M0 in XR819 and OpenRISC in H2+ are active.
jernej has joined #linux-sunxi
<miasma> i'll get back to that once i finish my build scripts that update the opi image via usb. i'm hoping i could do full system update without removing the sd card
<KotCzarny> that feature would be much more useful with a20 (which has real way for battery backup)
<miasma> i realized the arm cross-compilers that arch linux provide work out of the box for building sunxi stuff so i updated the instructions for people who want to run unsupported sunxi boards with arch
<tkaiser> KotCzarny: As already said: sleeping works so resuming should be possible too. But input sources/events have to be defined first. And I've no idea whether in A20 also an OpenRISC core lives or how the SoC will achieve energy savings there
<KotCzarny> no openrisc core in a20
<miasma> tkaiser: does the openrisc core also have some wire to the network chip so it can do WoL
<tkaiser> miasma: Don't think that works since the 'network chip' on those SBC is always just used as PHY and the MAC lives inside the SoC.
<miasma> right
<tkaiser> With XR819 Wi-Fi it seems to be different since the chip has an own M0 core
<miasma> it's not such a problem. i
<miasma> i've set up some esp8266 controllers for pcs to remote wake/shutdown them :)
<miasma> should the wiki also list the new banana pi board with R40 SoC
<tkaiser> miasma: And OPi Zero with some extreme settings idles below 0.5W with Ethernet active :)
<KotCzarny> if you have confirmed information, why not.
<miasma> well, there's the product page
<tkaiser> KotCzarny: Confirmed information means: buy one and test through, their product page is usually partially only 'copy&paste gone wrong'
<KotCzarny> tkaiser: it should be directed to miasma, not to me ;)
<KotCzarny> on a side note, android's fex from h2 sdk sets dram to 576
<tkaiser> KotCzarny: And configures dvfs broken, so why trusting in at all?
<KotCzarny> tkaiser, just a note, not going to use that fex anyway
<tkaiser> KotCzarny: Ok, but I knew already and chose 408 MHz instead for Armbian :p
<KotCzarny> i wonder if h2 might have troubles with dram speeds
<tkaiser> Heat problems when clocked higher, just like the other H3 devices with single bank DRAM config: NanoPi NEO/Air.
<KotCzarny> fun.
<tkaiser> tkaiser: But I did no extensive testing since if people are really that dumb and buy a $7 device to complain about 'low performance' later then they're already lost
<beeble> tkaiser: yiu can implement WOL without mac support. there are phys that can be set to look for the magic pattern itself and generate an interrupt
netlynx has quit [Ping timeout: 260 seconds]
<beeble> *you
<KotCzarny> beeble, but it lives in soc which is powered off?
netlynx has joined #linux-sunxi
<KotCzarny> or receives some standby power?
<miasma> low power devices isn't preventing the rpi crowd from building clusters of rpis
<miasma> *low performance
<beeble> KotCzarny: the phy has of course to be powered on (talking about external phys here)
<KotCzarny> crapsters (crap clusters)
<beeble> not sure about the realtek ones, don't use them in our designs
netlynx has quit [Client Quit]
tkaiser has quit [Ping timeout: 256 seconds]
<miasma> you could probably assume that if the board specs don't list WoL, they didn't even consider that when designing the board
plm has joined #linux-sunxi
<plm> Hi all
tkaiser has joined #linux-sunxi
smooker has quit [Remote host closed the connection]
terra854 has quit [Quit: Connection closed for inactivity]
<tkaiser> beeble: How much are those PHYs more expensive compared to RealDreck?
IgorPec has joined #linux-sunxi
<beeble> tkaiser: don't have volume pricing for realtek gigabit phys as i never used them. but you get one that supports it for a dollar in volume
<miasma> the RTL8211E that some opi devices use advertises WoL
<miasma> i got the impression that it's implemented in the PHY
<miasma> "The PMEB pin needs to be connected with a 4.7k-ohm resistor and pulled up to 3.3V or 5V. When the Wakeup Frame or a Magic Packet is sent to the PHY, the PMEB pin will be set lo w to notify the system to wake up."
<beeble> checked the datasheet
<beeble> yes, has the same feature. phy looks for the packet and sends out an interrupt
f0xx has quit [Ping timeout: 268 seconds]
Gerwin_J has quit [Quit: Gerwin_J]
leviathanch has quit [Remote host closed the connection]
HeavyMetal has quit [Ping timeout: 246 seconds]
fkluknav has quit [Ping timeout: 260 seconds]
HeavyMetal has joined #linux-sunxi
HeavyMetal has joined #linux-sunxi
HeavyMetal has quit [Changing host]
IgorPec has quit [Quit: Nettalk6 - www.ntalk.de]
The_Loko has quit [Ping timeout: 260 seconds]
IgorPec has joined #linux-sunxi
Mr__Anderson has joined #linux-sunxi
The_Loko has joined #linux-sunxi
HeavyMetal has quit [Ping timeout: 245 seconds]
HeavyMetal has joined #linux-sunxi
HeavyMetal has quit [Changing host]
HeavyMetal has joined #linux-sunxi
nashpa has quit [Quit: Going away]
<naobsd> I'll submit this version soon
<naobsd> btw, I noticed NES (US) ver. should have LED... right?
<naobsd> Famicom(JP) ver. doesn't have it :(
yann|work has joined #linux-sunxi
reinforce has quit [Quit: Leaving.]
mossroy has quit [Quit: Leaving]
Andy-D has joined #linux-sunxi
apritzel has joined #linux-sunxi
paulk-minnie has joined #linux-sunxi
gzamboni has quit [Ping timeout: 265 seconds]
The_Loko has quit [Ping timeout: 240 seconds]
The_Loko has joined #linux-sunxi
yann|work has quit [Ping timeout: 260 seconds]
simosx has quit [Quit: Yakkety Bye!]
Pepe has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]
afaerber has joined #linux-sunxi
* apritzel thinks he sticks with software, though soldering a 128MBit SPI flash on the OPi Zero just worked ...
<tkaiser> apritzel: And what about PoE? ;)
mmmm has joined #linux-sunxi
mmmm has quit [Client Quit]
scream has quit [Remote host closed the connection]
calhemp has joined #linux-sunxi
<calhemp> hi hramrach, I don't know much about DAC and I2C, but
<apritzel> tkaiser: I have turned off the soldering iron already ;-)
<apritzel> tkaiser: but FEL mode on the Zero is flaky, since you need to use the bus power from your host to power it
<calhemp> I follow www.volumio.org , and commity has some threads about i2c and Usb dacs compatibility,
<miasma> i2s?
<agraf> apritzel: why? just wire the power and data lines differently :)
<calhemp> if It can help, with your search of a good DAC
<tkaiser> apritzel: An USB3 port should do it. And you should also be able to use pins 1/2 on the 13 pin header oder 2/6 on the other one
<apritzel> sure there are ways around (like GPIO), just out of the box connecting a microB cable may not work
<apritzel> one cable did, another not
<miasma> calhemp: i2c is a low speed bus, different from i2s. it can only do 100 kbps on some boards and cd audio is 1400 kbps
<apritzel> it seems to be worse than supplying power from a "proper" charger
nashpa has joined #linux-sunxi
<calhemp> sorry, my mistake, links is for I2S devices not I2C as write before :(
<tkaiser> apritzel: I have a couple of Micro USB cables here that are 20AWG rated. Really helps. And every Mac from 2012 or later provides up to 1A on the USB ports.
<miasma> calhemp: no prob :)
<calhemp> ;)
<miasma> apritzel: a charger might provide a bit more than 5V since it doesn't know the lenght of the cable. computer ports might offer exactly 5.0V
<miasma> length even
<apritzel> miasma: yeah, I know
<apritzel> I think my charger says 5.2V on the tin
<miasma> yep
<tkaiser> apritzel: Since I cramped the Zero into a tiny enclosure I learned a new way to enter FEL mode. With inserted SD card I simply zero out the SPL header, then reboot ;)
<miasma> would the sunxi tools also work on mips? i was thinking of powering my broken opi pc using my openwrt box
calhemp has quit [Quit: ChatZilla 0.9.93 [Firefox 49.0.2/20161019084923]]
<apritzel> tkaiser: I have an even better way: http://pastebin.com/EheedEMG
<tkaiser> miasma: Should work, I use it all the time on x86 ;)
<tkaiser> apritzel: MMC0 is... ?
<apritzel> sd card
<apritzel> I wrote an extended version of uart0_helloworld.sunxi to the SD card
<tkaiser> apritzel: So that's a special SD card image to enter FEL mode?
<apritzel> yes
<apritzel> I am tempted to extend this to actually ask for this
<apritzel> tkaiser: and fun thing: I wrote this to the SPI flash as well
<tkaiser> apritzel: But my use case is different: I want to flash a new OS image to the SD card that is inside the enclosure. So I zero out the SPL header on the present SD card, reboot and double click the OS X sunxi flash tool ;) https://github.com/ThomasKaiser/sunxi-armbian-flasher-osx
jernej_ has joined #linux-sunxi
<apritzel> tkaiser: is this graphical? Or just a command line script?
<apritzel> tkaiser: and what do you do once in FEL mode?
<tkaiser> apritzel: GUI app. But just an application wrapper showing CLI output in a nice window. But monkeys can deal with it :)
<apritzel> tkaiser: load U-Boot? tftp image, write to SD?
<apritzel> tkaiser: I actually have a user for that, so thx for the link!
<tkaiser> apritzel: Then the payload is a sun8i legacy kernel that exposes either SD card or eMMC if present as USB mass storage gadget.
<apritzel> tkaiser: oh, that works much better with mainline U-Boot
<tkaiser> apritzel: So next step is opening the monkey friendly burn tool (Etcher) and then write the image to SD card or eMMC.
<apritzel> it supports mass storage, Android fastboot and USB DFU
<tkaiser> apritzel: It's mainline u-boot + legacy kernel (3.4.113)
<miasma> does the mainline u-boot support booting from the usb otg as an option
<apritzel> miasma: define booting
<miasma> loading the kernel
<apritzel> miasma: somehow you have to get U-Boot loaded and started
<apritzel> but then you can switch to mass storage mode for instance
jernej has quit [Ping timeout: 260 seconds]
The_Loko has quit [Quit: Leaving]
<apritzel> telling U-Boot which block device to export
<apritzel> for instance SD, or even better eMMC
<miasma> so the scenario is: i have the u-boot on sd and want to start a software update via usb otg only if the pc at the other end requests that
<tkaiser> apritzel: But performance is currently a bit low IIRC? Speaking about u-boot now...
<apritzel> tkaiser: I think ssvb mentioned about 3MB/s from U-Boot
<apritzel> so yes, slower than Linux
<apritzel> miasma: once U-Boot starts you have to enter this mass storage mode
<apritzel> miasma: which either can be done automatically via a boot script or typed on the command line
<miasma> apritzel: e.g. on arduino you can boot normally when powered with usb, but if the host requests an update, it can receive the new firmware
<miasma> currently i need to press the button to enter this update mode with opi
<apritzel> miasma: well, you can implement something like that U-Boot enters OTG and checks if someone is listening
<miasma> right. so if it detects the host, does it have a module for downloading stuff from usb?
<apritzel> I think this is all host driven at the moment
<miasma> i guess the easiest might be to detect host in u-boot and then boot a failsafe kernel from sd
<apritzel> right now you type that command in U-Boot, then your host sees a bog standard mass storage device
<miasma> ok
<miasma> i'll go study the u-boot scripting a bit
<miasma> i can also netboot, but it slows down normal boot
<apritzel> miasma: or you can enable the fastboot feature, which lets you flash it like a phone
nashpa has quit [Quit: Going away]
<apritzel> miasma: I think you can even start some code or reset the board
<miasma> cool.yea i'll study that a bit
nashpa has joined #linux-sunxi
phipli has quit [Ping timeout: 256 seconds]
vagrantc has quit [Ping timeout: 250 seconds]
lynxis has quit [Ping timeout: 250 seconds]
tkaiser has quit [Quit: jIRCii - http://www.oldschoolirc.com]
Mr__Anderson has quit [Remote host closed the connection]
plm has quit [Quit: leaving]
jstein has quit [Remote host closed the connection]
yann-kaelig has quit [Quit: Leaving]