<PietreLinux>
u-boot denx with q8 800_480defconfig
<PietreLinux>
and kernel mainline
jernej has quit [Ping timeout: 252 seconds]
<lordlod>
PietreLinux: you probably need to target a board, rather than a chip. Different systems have the power, video etc. connected in different ways
<lordlod>
PietreLinux: Having a look at the way armbian pulls it together will probably help you a lot, it basically does what you are after https://github.com/igorpecovnik/lib
<PietreLinux>
I talk to igor occasionally, it is good person, but I do not mean that, I'm talking about an operating system for tablet,
<PietreLinux>
for allwinner tablet
<PietreLinux>
kernel sunxi not optimiced for tablet, only optimized for a dev board
<PietreLinux>
led,sata,ethernet, very modifications in the kernel for run kernel in tablet
<PietreLinux>
kernel sunxi only for dev board
tkaiser has joined #linux-sunxi
taCoUser5916_FGF has joined #linux-sunxi
taCoUser5916 has quit [Read error: Connection reset by peer]
tkaiser has quit [Ping timeout: 252 seconds]
popolon has quit [Quit: WeeChat 1.4]
leio has joined #linux-sunxi
dh1tw has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Rondom has quit [Remote host closed the connection]
Rondom has joined #linux-sunxi
Colani1200 has joined #linux-sunxi
Colani1210 has quit [Ping timeout: 260 seconds]
florianH has quit [Quit: Connection closed for inactivity]
<lordlod>
PietreLinux: sorry I misunderstood, where are you having problems?
<PietreLinux>
Problems??
<PietreLinux>
no problems
<PietreLinux>
only ilusssions
<PietreLinux>
for creating a operating system for tablet
<PietreLinux>
I having problems?
gzamboni has quit [Ping timeout: 260 seconds]
<wens>
apritzel: yay! patches!
<apritzel>
wens: I know, finally
<apritzel>
so let the 32 vs 64 battle begin ;-)
tkaiser has joined #linux-sunxi
gzamboni has joined #linux-sunxi
<wens>
i still have 30 patches to send out :|
<apritzel>
some patches are left on my side as well
<apritzel>
but that's something for tomorrow ...
tkaiser has quit [Ping timeout: 268 seconds]
cnxsoft has joined #linux-sunxi
apritzel has quit [Ping timeout: 244 seconds]
PietreLinux has quit [Ping timeout: 260 seconds]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
perr has joined #linux-sunxi
perr has quit [Changing host]
perr has joined #linux-sunxi
tkaiser has joined #linux-sunxi
tkaiser has quit [Ping timeout: 256 seconds]
ninolein has quit [Ping timeout: 260 seconds]
ninolein has joined #linux-sunxi
Rondom has quit [Remote host closed the connection]
Rondom has joined #linux-sunxi
<wens>
argh, i keep adding stuff for the audio codec, for some minor differences at the board level :/
PietreLinux has joined #linux-sunxi
<PietreLinux>
That is sunxi, is a nonprofit organization, a subsidiary of allwinnner ??? what ??
<wens>
PietreLinux: its just a bunch of volunteers, not affiliated with allwinner
<PietreLinux>
I will make tablets running ubuntu desktop, touch, camera, accelerometer, everything, and I will launch a tablet with that system, but I need people to help me in the decision-making process
perr has quit [Ping timeout: 250 seconds]
<PietreLinux>
you have created this, I am morally bound to say
<miasma>
how do you create the tablets? do you design the hardware
<PietreLinux>
hardware and is designed by our friends allwinner us we put another software
<PietreLinux>
not android
<PietreLinux>
only Linux compiled with libc
<miasma>
i see
<miasma>
doesn't sound too bad, but there might be some issues with graphics and open source drivers
<miasma>
our friends at allwinner & arm holdings don't want people to use their hardware
Axl_z has joined #linux-sunxi
<PietreLinux>
apple did not want that their devices were hacked, but in court the reason that I modify the kernel was given device
<dr1337>
wens: any updates on H3 audio codec?
<PietreLinux>
with my orange pi plus 2 no problem in jack 3.5
tkaiser has joined #linux-sunxi
<PietreLinux>
allwinner my friends and arm is not who do not want Linux on their devices, they do not want to develop anything for it, but I made in blender fluid dynamics in a orange pi plus 2
<wens>
dr1337: in progress
<wens>
can't figure out why someone would add a gpio to control headphone mute :/
tkaiser has quit [Ping timeout: 260 seconds]
<PietreLinux>
Although for reasons I do not know you are a very closed community, only in English and with the knowledge and drabs, my amigps of Brazil, Indonesia, India, France, Peru, Argentina, Mexico, would like a tablet with linux inside, no matter if the great sound does not work or that you can not see video in 4K
<PietreLinux>
the dev board are fine, but were designed to test apps on android, we now have the opportunity to create tablets with linux
mzki has quit [Ping timeout: 268 seconds]
<PietreLinux>
I am determined to make a tablet for the common market, a33 perhaps, and no matter what it may cause the sound to work, no matter what it takes to make the mali 400 work in Linux, a greeting to all
mzki has joined #linux-sunxi
PietreLinux has quit [Quit: Page closed]
Axl_ has joined #linux-sunxi
Axl_z has quit [Ping timeout: 256 seconds]
Axl_ has quit [Client Quit]
Axl_ has joined #linux-sunxi
pg12_ has quit [Ping timeout: 260 seconds]
pg12 has joined #linux-sunxi
tkaiser has joined #linux-sunxi
tkaiser has quit [Ping timeout: 260 seconds]
foxx_ has joined #linux-sunxi
lkcl has quit [Ping timeout: 250 seconds]
Putti has quit [Ping timeout: 260 seconds]
Axl has joined #linux-sunxi
Axl is now known as Guest40318
Axl_ has quit [Ping timeout: 244 seconds]
[7] has quit [Ping timeout: 245 seconds]
TheSeven has joined #linux-sunxi
IgorPec has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
gianMOD has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
NetForHack has joined #linux-sunxi
<NetForHack>
Hi all.
Guest40318 has quit [Read error: Connection reset by peer]
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
<NetForHack>
What opinion have this chat about Orange Pi Zero ?
JohnDoe71rus has joined #linux-sunxi
reinforce has joined #linux-sunxi
jernej has joined #linux-sunxi
NetForHack has quit [Read error: Connection reset by peer]
tsuggs has quit [Remote host closed the connection]
tkaiser has joined #linux-sunxi
cnxsoft has quit [Read error: Connection reset by peer]
cnxsoft has joined #linux-sunxi
tkaiser has quit [Ping timeout: 245 seconds]
Putti has joined #linux-sunxi
Putti has quit [Ping timeout: 256 seconds]
Nuck-TH has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
jernej has quit [Ping timeout: 265 seconds]
Putti has joined #linux-sunxi
<Nuck-TH>
huh, search range got more or less narrow - 4.4.3 boots fine from "bad" card
<Nuck-TH>
althrough u-boot fails to detect it from time to time
tkaiser has joined #linux-sunxi
<terra854>
Hey guys, any updates on the A64 mainline effort?
<plaes>
terra854: yes
<plaes>
just updated wiki
tkaiser has quit [Ping timeout: 252 seconds]
<Nuck-TH>
what is GR8 SoC?
Putti has quit [Ping timeout: 250 seconds]
JohnDoe71rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
JohnDoe_71Rus has joined #linux-sunxi
<plaes>
Nextthing's custom SoC based on A13
<Nuck-TH>
great, 4.6.2 doesnt boot
<plaes>
what device?
<Nuck-TH>
cubietruck. i'm trying to guess in what kernel version boot from certain "bad" sdcards was broken
florianH has joined #linux-sunxi
dr1337 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<wens>
if you use the fixed function instead of configfs driven gadgets, you shouldn't need that?
<plaes>
wens: don't you want to use SUN4I_CODEC_ADC_ACTL_UNKxxx for those?
<beeble>
assuming your kernel config has gadget configured via configfs on CONFIG_USB_CONFIGFS
<wens>
plaes: it doesn't really help
<plaes>
then you can later grep for _UNK stuff
<wens>
beeble: nah, mine is fixed function
<wens>
beeble: and it works fine on the a23 :/
<wens>
but gives "VBUS_ERROR in a_wait_vrise (80, <SessEnd), retry #3, port1 00080108" on the A31
<beeble>
wens: i can give it a try later (if i don't forget)
apritzel has joined #linux-sunxi
petr has quit [Ping timeout: 260 seconds]
petr has joined #linux-sunxi
hp197 has quit [Remote host closed the connection]
kaspter has joined #linux-sunxi
tkaiser has joined #linux-sunxi
tkaiser has quit [Ping timeout: 260 seconds]
repka has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
popolon has joined #linux-sunxi
gianMOD has joined #linux-sunxi
<oliv3r>
olimex has released a new lime2 board revision with a changed PHY. They changed the troublesome RTL8211CL with an RTL8211E. Obviously we want a single u-boot for all lime2 boards, as it would get quickly to confusing
<oliv3r>
the problem is, the 8211E on the lime2 needs TX_DELAY=2 rathr then TX_DELAY=0, which is used on all other lime2's
<oliv3r>
is there some generic way to fix this?
<oliv3r>
my first (noteably wrong) method would be to do check a board.c for lime2, and if the PHY is rtl8211e set TX_DELAY=2
<oliv3r>
I think the quick fix may be to just always use Tx_delay, which reduces the performance slightly on older boards ...
<oliv3r>
hm we do already have board/sunxi/gmac.c where this delay is set ...
<oliv3r>
so it's not a PHY thing, but a configuration of the PHY on the MAC side
<apritzel>
oliv3r: anyway a slight decrease in network performance sounds feasible for U-Boot
<wens>
apritzel: bad design decision :(
<apritzel>
wens: ah, that's the GMAC
<wens>
apritzel: it wasn't for the older ones
kaspter has quit [Ping timeout: 260 seconds]
pcduino is now known as ski
ski has left #linux-sunxi [#linux-sunxi]
IgorPec has joined #linux-sunxi
Nuck-TH has joined #linux-sunxi
tkaiser has joined #linux-sunxi
<oliv3r>
apritzel: i don't think the kernel fully reinits the mac
<oliv3r>
hence why we do it in u-boot
<oliv3r>
i on' tknow how big the inpact is yet however
<wens>
oliv3r: the delay settings is in the clock controller, not the gmac
<mripard>
apritzel: having it in the DT wouldn't change anything, it would just move the problem
tkaiser has quit [Ping timeout: 260 seconds]
<oliv3r>
inu= u-boot?
<apritzel>
mripard: sure, I was just wondering if a separate DT would solve this magically
<oliv3r>
wens: i see it applied in boards/sunxi/gmac
<mripard>
wens: do you know if some autodetection of the delay could work ?
<wens>
oliv3r: i meant the registers
<wens>
mripard: if you put the phy into loopback, then maybe?
jski has joined #linux-sunxi
kaspter has joined #linux-sunxi
<oliv3r>
sounds reasonable, send data out the mac, over the rgmii, loop it in the mac, and it goes back; devide by /2 profit
<oliv3r>
but currently, we have an tx delay, not an rx delay, though i guess that doesn't matter over the wire, we never know when something arrives, ina loopback setup it might matter; then again, these traces generally are parallel
<oliv3r>
wens: ah indeed, ccm->gmac_clk; so it's the mac clock stuff; do we change that from the u-boot init in linux atm?
<wens>
linux doesn't touch it
IgorPec has quit [Ping timeout: 260 seconds]
gianMOD has quit [Remote host closed the connection]
<jski>
is yocto project the way to go now?
<plaes>
to go where? ;)
<NiteHawk>
way to go insane? :D
<jski>
a new linux dev we hired keeps raving about it but am not convinced
repka has quit [Quit: Leaving]
lkcl has quit [Ping timeout: 252 seconds]
<plaes>
depends what you want to do...
lkcl has joined #linux-sunxi
yann-kaelig has joined #linux-sunxi
<jski>
build a production friendly minimal system with basic functionality. in the past have always just built u-boot, kernel and used gentoo but was considering other distro's
<buZz>
jski: just stick to buildroot , no need to hipster up
<buZz>
buildroot2 is now also coming around
<jski>
I can't see the point either
<buZz>
re: yocto
tkaiser has joined #linux-sunxi
<buZz>
i bet your new dev just doesnt know what's around
<plaes>
or debian
<jski>
I use arch on pc but will have problems with systemd if on not mainline kernel
<buZz>
i use devuan
<buZz>
cause i dont like poetering
<jski>
gentoo has always been good in the past because I can rebuild all the source so very useful in the future
<jski>
either that or I have to make our own repo for full system upgrades
<plaes>
jski: why do you need mainline for?
<buZz>
:D
<buZz>
emerge -E world, daily
<jski>
don't need.. just need :P
<jski>
have used gentoo on our production boards for years now and has been handy just rebuilding in an arm chroot, should work on 3.4 kernel...
<beeble>
yocto isn't that bad but no real reason to switch if you have already another build system in-house
<jski>
I guess there is no need to strip the system down to busybox size these days when we have so much storage on embedded devices
<jski>
a minimal rootfs plus packaged minus man pages is good enough
<beeble>
busybox is nice for having a single binary. makes building a initramfs for examples easy
<Wizzup>
we do the same with gentoo @ work
<Wizzup>
but we stripped a bit more than just manpages
<Wizzup>
jski: do you make thosw images into a squashfs?
<NiteHawk>
i was about to ask, what's missing when using armv7a?
gianMOD has quit [Ping timeout: 260 seconds]
<tkaiser>
jski: That's crazy, Gentoo users still fry their H3 boards relying on overvolted overclocker settings.
<jski>
oh I see armv8 only adds optional 64bit
<jski>
is they why they still have arm64 in experimental? :P
<jski>
that why* :/
<apritzel>
jski: both 64-bit and 32-bit are optional in ARMv8 ;-)
<apritzel>
of course not both at the same time
<jski>
yes did not realise that :)
gianMOD has joined #linux-sunxi
<apritzel>
but you can have an implementation that only runs AArch64 (Cavium) or only AArch32 (Cortex-A32), while many do both (Cortex-A53, for instance)
<tkaiser>
jski: One more time: If you follow the instructions from Gentoo wiki your H3 devices will overheat and this might also have an impact on longevity.
apritzel has quit [Read error: Connection reset by peer]
bugzc has quit [Ping timeout: 256 seconds]
<jski>
tkaiser: I was not going to use their instructions
gianMOD has quit [Remote host closed the connection]
<jski>
thought I had to use arm64 and did not realise I could use armv7
gianMOD has joined #linux-sunxi
<tkaiser>
jski: But these instructions are wrong anyway. The dvfs settings are dangerous.
<ssvb>
tkaiser: where exactly do they suggest bad dvfs settings?
lamer14781880248 has joined #linux-sunxi
<jski>
ssvb: I have a armv7 running now on existing product was just going to try booting it then the upgrade package works :)
<ssvb>
either way, gentoo is mostly all about the userland, and one is free to use any kernel with it :-)
<ssvb>
and the systemd cancer is still optional there
<jski>
yes
<jski>
prefered until it is mainline stable
tkaiser has quit [Ping timeout: 244 seconds]
<jski>
ssvb: orange pi running stable?
<jski>
have ordered the pi pc plus and plus 2
<ssvb>
it depends on what you are running on it
<jski>
web server and mysql
apritzel has joined #linux-sunxi
<ssvb>
btw, the Orange Pi PC is a 32-bit board, it does not support aarch64
<lamer14781880248>
jski: Plus 2 is the worst choice, way too expensive, less IO bandwidth and ultra slow USB-to-SATA bridge. Plus 2E is the far better choice.
<ssvb>
there will be probably Allwinner H5 based Orange Pi boards soon, and they will be 64-bit
<lamer14781880248>
ssvb: Orange Pi PC 2 with H5 will start to sell soon.
<lamer14781880248>
In case you want to get your hands on one simply drop me a note, Steven wants to send out samples.
<tkaiser>
jski: GL830 is ultra slow, 15 MB/s sequential write, 30 MB/s sequential read, no UASP and also good for data corruption.
<jski>
Then I will just use the A20 for sata
kaspter has quit [Remote host closed the connection]
<apritzel>
tkaiser: no schanetics this time? ;-)
<jski>
was going to try use the Orange Pi PC Plus as a product
<tkaiser>
apritzel: Yeah, realized it only after uploading :)
<nove>
ssvb, about yesterday qemu + tracing, this is what was done http://dl.linux-sunxi.org/users/nove/dtv.png, but i am also in agreement in the question of time as it would be much more useful if the tracing was from real hardware
<tkaiser>
jski: If you look forward I would try to get my hands on a R40 device, the only one known today is called 'Banana Pi M2 Ultra'. R40 is more or less a H3 but lacking the 3rd USB host port and replacing it with native SATA.
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
lkcl has joined #linux-sunxi
rah has quit [Ping timeout: 250 seconds]
gianMOD has quit [Remote host closed the connection]
KotCzarny has joined #linux-sunxi
taCoUser5916_FGF has quit [Quit: hAvE yOu mOOEd tOdAY]
nikre has joined #linux-sunxi
jernej has joined #linux-sunxi
<ssvb>
tkaiser: thanks for the Orange Pi PC 2 schematics! Do you happen to know if the SPI NOR flash is always populated there?
<nikre>
will there be faster harddisk r/w speeds or same?
<tkaiser>
ssvb: No idea, at least on OPi Zero it's optional (not populated) ATM. PC 2 sales should start soon and then we'll know.
<tkaiser>
nikre: Are you refering to R40?
<nikre>
the new opi 64bit
<KotCzarny>
64bit is overrated
<tkaiser>
nikre: You mean H5? Given that's more ore less a H3 with different CPU cores performance should be the same as with H3 now.
<nikre>
what is the catch about it? 10bit 4k?
<tkaiser>
nikre: It's deca-core!!1!11!! Since deca-core Mali450. Also it's said to have better video decoding support and other stuff I don't care about.
<tkaiser>
Sorry, hexa-core Mali450
<KotCzarny>
mali 450 should be at least +50 better than mali 400, right?
<ssvb>
tkaiser: A53 is a little bit faster than A7 though
<apritzel>
KotCzarny: 50 more GPL violations, probably
<tkaiser>
ssvb: Sure, but this won't affect USB disk access :)
<KotCzarny>
ssvb: how git is little bit? in rough percent
<KotCzarny>
i wonder how much faster arm cores will have to get to make firefox run comfortably
<ssvb>
tkaiser: thanks! linpack is a nice example of a 2x neon performance advantage
<KotCzarny>
umkay, time to get back to my lovely bpi-r1
<tkaiser>
KotCzarny: if you calculate prime numbers for a living (as does sysbench --cpu) then H5/A64 is 15 times faster than H3.
<KotCzarny>
anyone doing computations for a living wouldnt use arm
<KotCzarny>
not yet
<tkaiser>
KotCzarny: Firefox is a database hammering your storage. Get ODROID-C2 with super fast eMMC and it flies.
<KotCzarny>
tkaiser, i shall try moving .profile to tmpfs and check
<KotCzarny>
for me even opening new window is a bit painful on h3
<KotCzarny>
which is a pity, because comparably (per watt used) system (thinkpad x40/x60) is much snappier
<tkaiser>
KotCzarny: Simply browse and run iotop. It's just insane. You have to tweak Firefox or Chromium settings regarding IO when running from SD cards.
gianMOD has joined #linux-sunxi
chomwitt has quit [Ping timeout: 260 seconds]
gianMOD has quit [Ping timeout: 256 seconds]
vagrantc has joined #linux-sunxi
chomwitt has joined #linux-sunxi
popolon has quit [Quit: WeeChat 1.4]
rah has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
Xal1u5 has joined #linux-sunxi
leviathanch has quit [Remote host closed the connection]
hp197 has joined #linux-sunxi
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
The_Loko has joined #linux-sunxi
|Jeroen| has joined #linux-sunxi
Da_Coynul has joined #linux-sunxi
gianMOD has joined #linux-sunxi
chomwitt has quit [Ping timeout: 250 seconds]
gianMOD has quit [Read error: Connection reset by peer]
gianMOD has joined #linux-sunxi
Da_Coynul has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
vagrantc has quit [Quit: leaving]
Putti has joined #linux-sunxi
KotCzarny has quit [Ping timeout: 260 seconds]
gianMOD has quit [Remote host closed the connection]
KotCzarny has joined #linux-sunxi
<KotCzarny>
hoo-raah. my bpi-r1 is now working with 4.9.0-rc3
<KotCzarny>
(using old b53 patch though, but fully functional)
KotCzarny has quit [Quit: brb]
|Jeroen| has quit [Quit: dada]
IgorPec has quit [Ping timeout: 268 seconds]
<nove>
50 more "unspeakable combination of two words" when the context of the discussion is mali450 ???
* nove
tries to joke
apritzel has quit [Read error: Connection reset by peer]
terra854 has quit [Quit: Connection closed for inactivity]
nove has quit [Quit: nove]
apritzel has joined #linux-sunxi
Nuck-TH has quit [Ping timeout: 260 seconds]
florianH has quit [Quit: Connection closed for inactivity]
KotCzarny has joined #linux-sunxi
Da_Coynul has joined #linux-sunxi
Gerwin_J has quit [Ping timeout: 264 seconds]
p_rossak has quit [Read error: Connection reset by peer]
netlynx has quit [Quit: Ex-Chat]
apritzel has quit [Ping timeout: 250 seconds]
gianMOD has joined #linux-sunxi
dh1tw has joined #linux-sunxi
p_rossak has joined #linux-sunxi
p_rossak has quit [Remote host closed the connection]
<apritzel>
AneoX: which kernel? mainline or 3.ancient Allwinner?
The_Loko has quit [Remote host closed the connection]
<AneoX>
apritzel: with any, problem in u-boot, not in kernel
<apritzel>
I was just wondering about the MMC driver in the kernel which worked
gianMOD has joined #linux-sunxi
<apritzel>
it would be a different story if it was only the BSP driver
jstein_ has joined #linux-sunxi
<apritzel>
in fact I have the opposite problem: works in U-Boot, fails in the kernel
<AneoX>
Boot kernel from sd on sdc0 lines, sunxi-kernel init emmc on sdc2 well, in 4 bit mode
whaf has joined #linux-sunxi
whaf has quit [Client Quit]
chomwitt has joined #linux-sunxi
jstein_ is now known as jstein
gianMOD has quit [Ping timeout: 260 seconds]
<apritzel>
AneoX: can you check the actual clock rate in the kernel?
<apritzel>
/sys/kernel/debug/clk/clk_summary
<AneoX>
in cat /sys/kernel/debug/mmc0/ios: clock:..52000000 Hz timing spec:.1 (mmc high-speed)
<apritzel>
AneoX: and you are sure it's SDC2? The two A20 boards with eMMC support in U-Boot use SDC3, apparently
<apritzel>
maybe there is something hardcoded to that port for the A20?
Mr__Anderson has joined #linux-sunxi
<AneoX>
i have no success with sdc3 lines, dont know why. Just switch to sdc2, and it works
<apritzel>
so you are connected to PC6 - PC11? Just wanted to check
<AneoX>
yes
<apritzel>
the kernel reads the info from the device tree, but U-Boot has this hardcoded
<apritzel>
so I was wondering if anyone ever tried an eMMC on SDC2 on an A20
<AneoX>
i use sunxi kernel with script.bin, my mainline u-boot not working with emmc, it fails to switch 4-bit mode and fails to read data
<AneoX>
which u-boot can boot from emmc on A20?
<apritzel>
Apparently mainline U-Boot should support eMMC boot on two Olimex boards
<apritzel>
but they use SDC3
<AneoX>
i have SD on sdc0, emmc on sdc2. If boot from SD, kernel well work with both flash.But if tries to boot from emmc, u-boot fails. I gave tried emmc on sdc0 and sdc2, not success. U-boot works if i try SD on sdc0 or sdc2, both works
Da_Coynul has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<AneoX>
Tried on 3 boards, A20-marsboard, Cubieboard2 and my custom
* apritzel
wonders if U-Boot sets up the right clocks
<AneoX>
u-boot seems set 52000000 Hz, but fails to switch to 4-bit width bus and read data. If boot to u-biit from sd, mmc list shows both cards, sd and emmc. mmc info shows correct values for emmc, but mmc part fails on emmc and works on sd. U-boot return mmc_load_image_raw_sector: mmc block read error when tried to boot from emmc
<apritzel>
mmh, the clocks seem to be right
<apritzel>
so have you tried the latest mainline kernel? 4.9-rc1
<apritzel>
or -rc3
<apritzel>
there were some MMC changes in that release for the A20
<AneoX>
no, i have not problem with kernel, mine works well if u-boot boot kernel)
<AneoX>
do you tried to boot from emmc on a20?
<apritzel>
AneoX: I don't have an A20 board with eMMC
<AneoX>
(
<apritzel>
only A64
<apritzel>
so you changed the .fex file for the sunxi kernel?
<AneoX>
seems like this bug in u-boot on sun7i driver
<apritzel>
I was wondering if this was just a config issue in U-Boot
<AneoX>
in fex i have just [mmc2_para] sdc_used = 1
<AneoX>
i cant find this issue in U-Boot for a month(
Da_Coynul has joined #linux-sunxi
<apritzel>
mmh, the clock setup seems to be correct
p_rossak has quit [Ping timeout: 260 seconds]
p_rossak has joined #linux-sunxi
afaerber has quit [Quit: Ex-Chat]
<apritzel>
AneoX: you could hack sunxi_mmc.c:sunxi_mmc_init() to comment the host_caps = MMC_MODE_4BIT line
<apritzel>
that should set 1 bit mode in U-Boot, if I get the code correctly
<AneoX>
have tried
<AneoX>
board freezed
<AneoX>
last message from mmc trace mmc 0, cmd 16(0x80000150), arg 0x00000200 mmc resp 0x00000900 mmc 0, cmd 18(0x80003352), arg 0x00000050 trans data 452608 bytes
dr1337 has joined #linux-sunxi
whaf has joined #linux-sunxi
<apritzel>
AneoX: do you have a clue which voltage the kernel uses?
Mr__Anderson has quit [Remote host closed the connection]
<AneoX>
vdd:..7 (1.65 - 1.95 V)
<apritzel>
AneoX: oh, so it's a 1.8V eMMC?
<apritzel>
U-Boot hardcodes the voltage to 3.3 V
<AneoX>
no, emmc powered by 3.3v
<AneoX>
vdd:..7 (1.65 - 1.95 V) it from /sys/kernel/debug/mmc0/ios
<AneoX>
dont know which voltage in u-boot
<apritzel>
which looks like the kernel negotiates 1.8V with the eMMC
<apritzel>
also shouldn't it be .../mmc2/ios
<apritzel>
?
<AneoX>
kernel give emmc higher priority, looks like
<apritzel>
can you dump the whole file? for both interfaces?
<AneoX>
cat /sys/kernel/debug/mmc0/ios clock:..52000000 Hz vdd:..7 (1.65 - 1.95 V) bus mode:.2 (push-pull) chip select:.0 (don't care) power mode:.2 (on) bus width:.2 (4 bits) timing spec:.1 (mmc high-speed)
<AneoX>
cat /sys/kernel/debug/mmc1/ios clock:..50000000 Hz vdd:..16 (2.8 ~ 2.9 V) bus mode:.2 (push-pull) chip select:.0 (don't care) power mode:.2 (on) bus width:.2 (4 bits) timing spec:.2 (sd high-speed)
<AneoX>
emmc on mmc0 and /dev/mmcblk0, but connected to sdc2 lines
lerc has quit [Remote host closed the connection]
<apritzel>
yeah, I see
paulk-collins has quit [Quit: Leaving]
<apritzel>
AneoX: you could try to #define DEBUG in sunxi_mmc.c and board/sunxi/board.c
<apritzel>
and see what that tells you about the configuration
lerc has joined #linux-sunxi
<apritzel>
AneoX: I think it's worth to chase this voltage setting in U-Boot
<apritzel>
the eMMC works on the A64 in U-Boot, but my eMMC reports "vdd: 21 (3.3 ~ 3.4 V)"
<apritzel>
which matches U-Boot's config
<AneoX>
A20-OLinuXino-Lime_defconfig
<AneoX>
i use
<apritzel>
have you tried to use A20-OLinuXino_MICRO_defconfig as a base?
<apritzel>
this has eMMC, but on SDC3
jmcneill has joined #linux-sunxi
<apritzel>
and it defines a CD_PIN for that, which is a bit odd
Da_Coynul has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<jmcneill>
Has anybody looked at audio w/ AC100 codecs? Curious to know how what property you're using for the codec node to find it
<AneoX>
CD_PIN i clean and CONFIG_MMC_SUNXI_SLOT_EXTRA set 2
<AneoX>
CONFIG_MMC_SUNXI_SLOT_EXTRA=3 from micro config, will not effect
<AneoX>
sunxi/board.c has no debug output errors, just checked
<apritzel>
AneoX: try to OR in MMC_VDD_165_195 to the voltages listing in sunxi_mmc.c
<apritzel>
like it's done in drivers/mmc/dw_mmc.c
p_rossak has quit [Remote host closed the connection]