<naobsd>
levd: by the way, firefly 4GB ver. uses same board as 2GB (production) ver.?
<Tony_>
naobsd, where I can fond the miniroot's source ?
<Tony_>
I can only get images now.
<naobsd>
Tony_: main parts are script, you can see them.
<naobsd>
rest of binary files are picked from ubuntu 14.04 rootfs
<naobsd>
anyway that's outdated a little
<naobsd>
I'm using a little different version
<Tony_>
Could you share your miniroot's files ?
<naobsd>
I think I need to do some cleanups
<naobsd>
I always ignore some errors ;)
<naobsd>
I'm doing some mainline things
<naobsd>
I'll update my miniroot for it
<Tony_>
okay, thank you.
<naobsd>
I want to finish dts work 1st...
<naobsd>
for firefly and rayeager
<levd>
naobsd: yes, firefly 4GB uses same board as 2GB.
<naobsd>
levd: thanks, then only 1 line in .dts needs to be modified ;)
<naobsd>
I'm not sure we can specify value of 4GB in dts and kernel can run w/o modification
<naobsd>
oh I should ask it...
<naobsd>
levd: are you(or someone in firefly) already trying mainline Linux?
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<levd>
naobsd: we tried the mainline Linux, but got stuck. We're not familiar with mainline kernel yet, a lot of differences from SDK kernel.
<naobsd>
levd: I see, I'll prepare it and ask to test on 4GB board :)
<levd>
Good
<levd>
The same firmware can be used in both version. The DDR and eMMC size are auto detected.
_andrew_ has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<_andrew_>
naobsd: Hi, do you have a working (or working enough to boot to /bin/sh over serial console from the eMMC) device tree for the Firefly and current mainline kernel that I'd be able to try?
<_andrew_>
Currently fiddling with one based on rk3288-r89.dts from the devel/workbench branch of https://github.com/mmind/linux-rockchip.git but it's getting stuck at the eMMC with
<_andrew_>
dwmmc_rockchip ff0f0000.dwmmc: No vmmc regulator found
<_andrew_>
dwmmc_rockchip ff0f0000.dwmmc: No vqmmc regulator found
<naobsd>
_andrew_: no
<naobsd>
_andrew_: I cannot understand why you said "give me" just after joined here. if you saw log, you should know answer.
<naobsd>
good news, gyroscope doesn't exist on rayeager, "cannot find" problem is solved ;)
<naobsd>
let delete lines from dts and kernel config... ;)
RayFlower has quit [Read error: Connection reset by peer]
<_andrew_>
naobsd: Well, I didn't just say "give me", I asked politely. It seemed, yes, from the log, that you are working on creating a Firefly device tree file. Sorry if I read that wrong?
RayFlower has joined #linux-rockchip
<naobsd>
"are working on creating a Firefly device tree file"
<naobsd>
sorry, who said "working version is ready"?
<_andrew_>
naobsd: No one.
<naobsd>
usually we share a lot of things
<naobsd>
because we are talking frequently
<naobsd>
I may be wrong, but, I surprised
<naobsd>
suddenly appeared and said "give me" very politely
<_andrew_>
That sentence you quoted, doesn't imply a "working version" if that's what you thought. It means I got the impression that you are in the process of creating one.
<naobsd>
"in the process of creating" yes
<_andrew_>
Right. That's what I was asking, if it was even in some minimally working state? From your reply, I guess not...
<Tony_>
naobsd, how to covert boot.img to file tree ?
<Tony_>
cpio ?
<ganbold__>
naobsd: seems expensive
<naobsd>
Tony_: it's cpio+gz with 8bytes header and 4bytes trailer
<naobsd>
oh, popmetal has GPS/NFS
<naobsd>
maybe one of AP6xxx
<naobsd>
NFS -> NFC
<_andrew_>
Obviously English isn't the natively spoken language around here, so some communications issues. No hard feelings I hope...
_andrew_ has quit [Quit: I'll be back]
<ganbold__>
naobsd: yeah, NFC+GPS
<naobsd>
as like as rayeager board, there are a lot of parts. pin headers are small(1.27mm pitch? not sure) but board size is relatively large (compared to firefly)
<naobsd>
both rayeager and popmetal are 15cm x 9.?cm
Tony_ has quit [Remote host closed the connection]
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 260 seconds]
Tony_ has joined #linux-rockchip
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 255 seconds]
apritzel has joined #linux-rockchip
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 255 seconds]
levd1 has quit [Remote host closed the connection]
levd has joined #linux-rockchip
levd1 has joined #linux-rockchip
wildea01 has joined #linux-rockchip
levd has quit [Ping timeout: 264 seconds]
levd has joined #linux-rockchip
<Tony_>
naobsd, actually, I haven't upload non-working(modified) boot.img alone.
<Tony_>
before.
naobsd has joined #linux-rockchip
levd1 has quit [Ping timeout: 244 seconds]
Tony_ has quit [Remote host closed the connection]
levd has quit [Ping timeout: 265 seconds]
<naobsd>
mm
naobsd has quit [Ping timeout: 246 seconds]
naobsd has joined #linux-rockchip
RayFlower has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
antoinemaillard has quit [Quit: antoinemaillard]
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
naobsd has quit [Quit: Page closed]
antoinemaillard has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
bengal has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
Omegamoon has joined #linux-rockchip
<AstralixNB>
hey omegamoon :)
<Omegamoon>
Hi there astralix
<AstralixNB>
everything fine?
<Omegamoon>
Busy as always
<AstralixNB>
same here
<Omegamoon>
Struggling to get the rtl8723bs bluetooth working currently
<AstralixNB>
Which board?
<Omegamoon>
Mk809iii
<AstralixNB>
... mainline kernel or which one?
<Omegamoon>
No, this is on the 3.0.x branch
<AstralixNB>
ah... so is it ON and communicating?
bengal has quit [Ping timeout: 244 seconds]
<Omegamoon>
Yes, it is
<Omegamoon>
But it is giving trouble downloading the firmware
<AstralixNB>
Ah, we have this at least ... wait a second...
naobsd has joined #linux-rockchip
<Omegamoon>
Would be nice if this is a 'known issue' which has been solved before
<AstralixNB>
We struggled at the very same point some Weeks ago with an older KK and an Omni release. But after an update of the AOSP and Omni sources, the problem was just solved
<Omegamoon>
Are the omni sources freely available?
<AstralixNB>
We got stuck at the point where all chipsets load the firmware blob, except this RTL. Even it was handled the same and got the right rights.
<AstralixNB>
Omni is available at their site
<Omegamoon>
Could you point me to the known working source?
<Astralix>
First section, the link to the arctablet loader collection has some trailing characters, so the link ends at 404
<Astralix>
Important Note No 1:
<field^Mop>
whoo, thx! i noticed that kind of problem with some links and corrected them. obviously, i did not verify every single link, so thx for that!
<Astralix>
Same with the to radxa
<Astralix>
%7C at the end
<Astralix>
Ok, for the Loader....
<Astralix>
1st the Loader binary itself must roughly match the kernel driver (rknand_ko.ko)
<Astralix>
2nd the Loader must match the CPU
<Astralix>
3rd the kernel module must match the kernel release 3.0.8+ or 3.0.36+
<field^Mop>
Astralix: saw it, all the links hat the pipe sign. my fault
ferric has joined #linux-rockchip
<Astralix>
just tell when you're ready
<Astralix>
After Important Note 1 I would add Note 2 telling:
<field^Mop>
Astralix: ok, links in header should be fixed now.
<field^Mop>
next?
<Astralix>
Yust read above :)
<Astralix>
The Loader initializes DRAM and Flash memory. By trying a different Loader version it may happen, that the chips on the board are not covered by the Loader tried.
<Astralix>
So alway ensure that you have a backup or at least an original firmware image of your device.
<Astralix>
Important Note 3
<Astralix>
is obsolete...
<Astralix>
May be you can add to note 1 that in original rockchip kernels, the rknand_ko.ko are saved as
<Astralix>
rknand_ko.ko.3.0.8+ and rknand_ko.ko.3.0.36+
<Astralix>
The kernel is modified to test the following sequence:
<field^Mop>
Astralix: lol, sure? double .ko.o?
<field^Mop>
+k
<Astralix>
yes
<field^Mop>
ok
<field^Mop>
strange
<Astralix>
If there is a modprobe rknand_ko.ko the kernel tries loading this module. If it fails, it attaches its own kernel version to the name and tries again.
<field^Mop>
ok
<field^Mop>
i'll add those lines and you tell me whether thats what you mean, ok?
<Astralix>
With that you can build an android, that runs on 3066 with kernel 3.0.8+ as well as on 3188 with kernel 3.0.36+ without changing anything in the Android
<Astralix>
Just add it in so the page keeps your style. If something is wrong, I can edit it for minor corrections.
<field^Mop>
t8
<field^Mop>
r8
<field^Mop>
Astralix: what about the rkxxxxnand.ko module? same _ko.ko stuff?
<Astralix>
With loader one, there where two modules per chip for each kernel version.
<Astralix>
With the current loader two version, there are only two modules, one for each kernel version.
<field^Mop>
okaaaayy.. that important note will evolve ..
<field^Mop>
how could we name a new section replacing the two notes? something about bin bootloader prerequisites?
<field^Mop>
binary dependency.. matching..
<Astralix>
What's important to know before wressling with the Loaders?
<Astralix>
:)
<Astralix>
"Important details about the Loader"
<field^Mop>
:)
<Astralix>
If you covered that, there are some technical traps in the text about the "how it works"
<field^Mop>
Astralix: note aside: what you told me about the idb topic, that might rightuously go to in that bin bootloader section, too?
<field^Mop>
ok, give me some minute
<field^Mop>
Astralix: clarification: ver 2 loader came with one kernel module per kernel version?
<field^Mop>
Astralix: (as there are two kernel versions distributed (..8 and ..36+) it's two modules available?)
<Astralix>
yes: Loader 1: one module per SOC and per kernel, later it changed to one module per kernel version only
<field^Mop>
Astralix: r8. thx
<Astralix>
The loader package supplied by RK to any SDK is always a Loader per SOC and a module per kernel version
<field^Mop>
Astralix: so, ver2 loader: one kernel module per soc _and_ per kernel version, obviously.
<field^Mop>
Astralix: which one is it? rknand or rkxxxxnand?
<Astralix>
no... Loader Version 2 has one Loader per SOC and one module per kernel
<Astralix>
So if you think of 3066 and 3188 at Loader 1
<Astralix>
you have RK3066Loader(L)_v1.xx.img and RK3188Loader(L)_v1.xx.img
<Astralix>
In Loader 2 this is the same for the Loader image, but there are only two modules, one generic for kernel 3.0.8+ and one for 3.0.36+ but for all SOCs
<field^Mop>
ok, maybe a grasped it this time. you'll notice.. ;) thx!
<Astralix>
hasn't there been an online service, where you can edit a single document from multiple parties at the same time?
<field^Mop>
Astralix: sure, but not for mediawiki, afaik
<Astralix>
However
<Astralix>
Tell me when you have a first version
ferric has quit [Ping timeout: 256 seconds]
<field^Mop>
yep
ferric has joined #linux-rockchip
gb_master has quit [Quit: Leaving]
ferric has quit [Ping timeout: 264 seconds]
ferric has joined #linux-rockchip
<Astralix>
So, ready for some Loaer deatils?
<Astralix>
details?
<field^Mop>
Astralix: saved the unfinished work, have to go, will continue later!
<field^Mop>
that did help me a lot to get to know the details. excellent.
<field^Mop>
Astralix: tomorrow, loader details? btw, it is this exact section that still needs work to be done. as of now, it is rather a paraphrased and slightly cleaned up irc log.