kaspter has quit [Read error: Connection reset by peer]
kaspter has joined #linux-sunxi
victhor has quit [Ping timeout: 240 seconds]
camus has joined #linux-sunxi
kaspter has quit [Ping timeout: 240 seconds]
camus is now known as kaspter
buzzmarshall has quit [Remote host closed the connection]
ganbold has joined #linux-sunxi
kaspter has quit [Remote host closed the connection]
kaspter has joined #linux-sunxi
\\Mr_C\\ has quit [Quit: Ping timeout: 999,999,999 years]
vagrantc has joined #linux-sunxi
ganbold has quit [Quit: Leaving]
ganbold has joined #linux-sunxi
apritzel has joined #linux-sunxi
hlauer has joined #linux-sunxi
apritzel has quit [Ping timeout: 260 seconds]
\\Mr_C\\ has joined #linux-sunxi
jstein has joined #linux-sunxi
reinforce has joined #linux-sunxi
<mripard>
smaeul: yeah, I don't really know what's up with the clk maintainers, I'm struggling to get patches reviewed too
<mripard>
I still think this is the best solution, so just keep pinging them?
gsz has joined #linux-sunxi
cmeerw has joined #linux-sunxi
cmeerw has quit [Ping timeout: 260 seconds]
vagrantc has quit [Quit: leaving]
fourkbomb has quit [*.net *.split]
camus has joined #linux-sunxi
jbrown has quit [Remote host closed the connection]
kaspter has quit [Ping timeout: 252 seconds]
camus is now known as kaspter
matthias_bgg has joined #linux-sunxi
kaspter has quit [Ping timeout: 240 seconds]
kaspter has joined #linux-sunxi
fourkbomb has joined #linux-sunxi
netlynx has joined #linux-sunxi
netlynx has joined #linux-sunxi
apritzel has joined #linux-sunxi
prefixcactus has joined #linux-sunxi
<mru>
are you being clock-blocked?
victhor has joined #linux-sunxi
<warpme_>
guys: i'm trying to get 5.12 mainline boot on TanixTX6s (h616). Distro i'm using works ok with OrangePIZero2 so ATF should be ok. I started with uboot (current sunxi custodians repo) + patch like this: https://pastebin.com/8uXaMnwT This gives me uart output like this: https://pastebin.com/aTxAa9f6 It looks like uboot looking on mmc0 while this tvbox seems to have sd card as mmc1. maybe somebody has some ideas
<warpme_>
here?
<apritzel>
warpme_: SD card as mmc1? are you sure? How do you boot?
<apritzel>
buZz: sure it has (eMMC on MMC2, probably), but this boots the Android
<buZz>
doesnt fex/dtb decide which mmc is which device? could just swap it
<apritzel>
buZz: please don't mention fex ever again ;-)
<apritzel>
last time I checked U-Boot had the hardcoded assumption of mmc0 being the SD card (in the SPL code, for instance)
<apritzel>
yes, normally this is covered by the DT, and is no problem for fully compliant software, like the kernel or U-Boot proper
<buZz>
ah
<buZz>
apritzel: :D :D
<apritzel>
warpme_: do they put the WiFi on mmc0 instead?
<apritzel>
warpme_: I guess you are booting via FEL then?
<apritzel>
warpme_: also to preempt libv: please add a Wiki page for that, especially if the TX6s is a special snowflake with its SD card
<apritzel>
warpme_: and do you have a 1GB DRAM version? Or is the DRAM special and we detect only half of it?
indy has quit [Ping timeout: 260 seconds]
<warpme_>
apritzel: i suspect sd card is on mmc1 as to get uboot spl i need to set CONFIG_MMC1_CD_PIN="PF6" in defconfig. with CONFIG_MMC0_CD_PIN="PF6" there is no spl at all. with CONFIG_MMC1_CD_PIN="PF6" uboot tells me https://pastebin.com/aTxAa9f6 re: RAM - boc has 2G ram so it looks uboot detects wrongly only 1G.. re: your branch for h616 - i already using code from it to get nicelly working OPIzero2
<apritzel>
so with MMC0_CD_PIN is looks at PF6, and sees "no card"? But MMC1_CD_PIN would get ignored, and the missing MMC0_CD_PIN let's U-Boot assume non-removable?
specing_ has joined #linux-sunxi
specing has quit [Ping timeout: 240 seconds]
<warpme_>
apritzel: well. idk as i don't have any schematics. i was guessing it is on mmc1 as simple change in CONFIG_MMC1_CD_PIN from MMC0 to MMC1 allows uboot move forward.... but for sure it is just guessing -(
specing_ is now known as specing
jbrown has joined #linux-sunxi
indy has joined #linux-sunxi
fl__0 has joined #linux-sunxi
fl_0 has quit [Ping timeout: 276 seconds]
<warpme_>
apritzel: i don't checked uboot code - but if your statement "MMC1_CD_PIN would get ignored" comes from uboot code - then indeed hypothesis that sd card is on mmc0 + setting CD pin to mmc1 in reality ignores CD pin on mmc0 -> causes uboot assumes sd card is non-removable mmc0 start to make sense. Maybe simple test with: defconfig CONFIG_MMC0_CD_PIN entry removed will help us to move forward?
choozy has joined #linux-sunxi
<warpme_>
apritzel: indeed. commenting MMC0_CD_PIN gives the same result like using MMC1_CD_PIN so sd card seems to be on mmc0 + this box as CD uses different than PF6 pin.
reinforce has quit [Quit: Leaving.]
reinforce has joined #linux-sunxi
jstefanop has joined #linux-sunxi
<apritzel>
warpme_: great! You can just use CONFIG_MMC0_CD_PIN="" in your defconfig then
<apritzel>
warpme_: and does the box have 3 USB ports? The pictures I found online suggest so
camus has joined #linux-sunxi
kaspter has quit [Ping timeout: 260 seconds]
camus is now known as kaspter
Net147 has quit [Ping timeout: 268 seconds]
cnxsoft has quit [Quit: cnxsoft]
<warpme_>
apritzel: yes. it has 3 ports: 1x blue (3.0 i think) + 2x white (2.0 i suppose)
<apritzel>
there is no USB 3.0 controller in the H616
<apritzel>
but selling it as USB 3.0 doesn't seem to be uncommon ;-)
<warpme_>
ah ok - so it looks like box producer had just such component (or this another famous Chinese marketing spin :-) )
<apritzel>
(probaby because the H6 had it, and people just copy&pasted everything from their previous TV box models)
<warpme_>
:-)
<apritzel>
for instance the Tanix TX6 (without the "s")
elros1 has joined #linux-sunxi
<warpme_>
h616 is really nice mediaplayer soc imho
<apritzel>
I don't know about that, but it seems to be cheap
<warpme_>
exactly. OrangePIzero2 is 17Eur board with periph. exactly enough for mediaplayer. so for price-functionality it is killler
<warpme_>
so sad it hasn't yet foss traction it deserves....
Net147 has joined #linux-sunxi
gaston1980 has quit [Ping timeout: 240 seconds]
<warpme_>
apritzel: infact your+jernej code runs really nice on h616. HDMI, USB, Eth working well. To get multimedia player on OPIzero2 missing are: GPU/Audio/VideoDEC.
<warpme_>
CEC as well :-)
<warpme_>
i mean CEC works nicelly as well
<apritzel>
well, that doesn't look too bad, does it? we got our hands on boards only end of last year, so that's pretty fast, I'd say
reinforce has quit [Quit: Leaving.]
Net147 has quit [Ping timeout: 268 seconds]
<warpme_>
yeah. it look really good!
<jernej>
warpme_: VI planes still don't work with my WIP patches, so you would need GPU rendering
<warpme_>
gpu is g31 - well supported by panfrost. probably some entries in DT and it will pop-up. I don't know about DE3.3 - as iirc jernej mentined no yuv support (iirc).
<jernej>
warpme_: regarding sd card issue - you could try adding "broken-cd;" to mmc0 node just to see if it helps
<apritzel>
jernej: broken-cd is ignored by U-Boot proper ;-)
<jernej>
ah, ok
<apritzel>
but just not specifying a CD GPIO should fix that
<jernej>
I have tx6s android dt, let me check what it says regarding cd pin
<warpme_>
jernej: saying "need GPU rendering" - do you mean DMABUF in EGL mode?
<jernej>
I don't have specific method in mind, whatever works with GPU
<jernej>
apritzel, warpme_: it seems CD pin on TX6s is PI10
<warpme_>
h6 t720 in DMABUF in EGL mode works really well for me. comparing cpu load on gs1: DRM planes is 5-10% load while EGL is 15-25%. not bad - giving that EGL mode gives me deinterlacing by GPU.
yann has quit [Ping timeout: 252 seconds]
Net147 has joined #linux-sunxi
<warpme_>
jernej: qll. let me try with PI10
<warpme_>
jernej: in DT i have: cd-gpios = <&pio 5 6 GPIO_ACTIVE_LOW>; how this should be changed?
<jernej>
<&pio 8 10 ...
<warpme_>
PI10 - is it papa leon 10?
<warpme_>
it seems Papa India 10
<libv>
lima. the now too overweight glider pilot in me remembers that L is lima.
<warpme_>
:-)
<libv>
for completeness sake, original name was remali, but arm complained about possible trademark violations
<karlp>
hrm, I thought I remembered leon in one of the older regional ones, but no, there was "liverpool" and "love" and "luis" and "lewis" though...
<libv>
and one of the other codethink employees suggested lima
<libv>
i do not remember who though
<jernej>
apritzel: speaking of CD pin, T95 also seems to use PI10 pin for it
gaston1980 has joined #linux-sunxi
<jernej>
so X96 and T95 are thus not totally compatible
<apritzel>
and I thought we were past non-standard CD pins ...
<apritzel>
because PF6 cries "use me for CD!!!" all over the place
<jernej>
well, it's just a pin, nobody said it must be one or another
<apritzel>
sure, bu just because you *can* doesn't mean you must deviate
<apritzel>
that's not really well understood in the ARM world, though
<jernej>
and that's why DT was born :)
fl__0 has quit [Ping timeout: 245 seconds]
<jernej>
well, used
<warpme_>
apritzel: totally agree. it should be rule of thumb: deviate only if you must!
<apritzel>
jernej: we have: "default "PF6" if MACH_SUN8I_A83T || MACH_SUNXI_H3_H5 || MACH_SUN50I" in U-Boot
fl_0 has joined #linux-sunxi
<warpme_>
hmm - with PI10 in def config + &pio 8 10 in DT i'm getting:
<tuxd3v>
I don't believe uboot can handle the name the kernel atributes to its disks..
<tuxd3v>
maybe via device-tree.. but I don't know how to do that..
<Jin^eLD>
tuxd3v: no, no, u-boot knows your device, i.e. the one you will be booting from
<Jin^eLD>
and then no matter which one it figured out, you give a fixed name like, "rootfs" for your user space
jstefano_ has joined #linux-sunxi
jstefanop has quit [Ping timeout: 246 seconds]
<Jin^eLD>
basically my thought was, you set tha bootargs via u-boot to root=/dev/mmcblkXpY:part-name=rootfs
<tuxd3v>
yes uboot knows, and its pass the partuuid to kernel args, and then lated sometimes the kernel decide that disk 1, is now disk 2 or vice versa..
<Jin^eLD>
and you construct the /dev/mmcblk stuff by a u-boot script since you have to know where you are booting from, and fstab just uses the part-name
<Jin^eLD>
since the part name remains fixed regardless of what device you picked for booting
<tuxd3v>
I wanted to fix the devices
<tuxd3v>
like uboot mmc0 is mmcblk1
<tuxd3v>
fixed names, that could be possible via device t5ree
<tuxd3v>
tree*
<tuxd3v>
ofcourse I can use PARTUUID in /etc/fstab
<Jin^eLD>
I was told it should be possible via device trees indeed, but that's where my knowledge ends :P
<tuxd3v>
plaes, mentioned a workaround
<tuxd3v>
but I liked to have a solution :)
<tuxd3v>
plaes, many thanks for remembering me of that :)