akaizen has quit [Remote host closed the connection]
cnxsoft has joined #linux-rockchip
ganbold has joined #linux-rockchip
_vaibhav_ has joined #linux-rockchip
xcasex has quit [Ping timeout: 244 seconds]
xcasex has joined #linux-rockchip
GriefNorth has joined #linux-rockchip
* bashintosh
has found out why kernel v3.14 won't boot on Firefly/CX929 - there is no support for PMIC ACT8846. Chromebooks use RK808.
xcasex has quit [Ping timeout: 264 seconds]
xcasex has joined #linux-rockchip
<lautriv>
karlp, the amount. machine has 2G but when booting linux from SD it shows just 1G, hence kernels fault.
<lautriv>
bashintosh, the appended dtb is vanilla only since android is using old kernels and has no such option. if appended, there is a config-option telling the kernel about it.
AndChat|536769 has joined #linux-rockchip
ckeepax has quit [Ping timeout: 252 seconds]
markm has joined #linux-rockchip
_vaibhav_ has quit [Quit: Leaving.]
Luke-Jr has quit [Excess Flood]
Luke-Jr has joined #linux-rockchip
Luke-Jr has quit [Excess Flood]
naobsd has joined #linux-rockchip
Luke-Jr has joined #linux-rockchip
<bashintosh>
lautriv: "CONFIG_BUILD_ARM_APPENDED_DTB_IMAGE" and "CONFIG_ARM_APPENDED_DTB" are what I have found - thanks, that clarified the doubt I had regarding the dtb! :)
<bashintosh>
naobsd: thanks as well :) Trying to get v3.14 to work on Firefly or CX929 boards... More work to do than expected but that's the fun part ;)
ckeepax has joined #linux-rockchip
<lautriv>
bashintosh, maybe you know a more recent site instead linux-rockchip.info regarding linux on such boxen ?
AndChat|536769 has quit [Ping timeout: 246 seconds]
<bashintosh>
lautriv: not really :( I refer to that site a fair bit as well.
<lautriv>
bashintosh, several 404 make it rather pointless for my situation :(
<bashintosh>
lautriv: are you working on mainline kernel? Will get there soon too..
<lautriv>
bashintosh, i work on mainline for another arm , trying to get one single 3188 running linux where i'm banned to androshit. thingie boots but what i collected does only 1G RAM (from 2G) and lacks on WiFi (no source for rtl8723).
<bashintosh>
lautriv: 1G where there are 2G mmh.. Quick guesses, which you might have checked already: kernel boot parameters? DTS?
cristian_c has joined #linux-rockchip
<lautriv>
bashintosh, android 3.0.36 does not use/have any dts. i read somewhere to edit config.h but there is no RAM defined, i assume the ddr-init extracted from loader is responsible but sargs tools to extract that vanished.
<cristian_c>
hello
<cristian_c>
I've backupped partitions by flash tool
<cristian_c>
now, I've got files with .backup extension
<cristian_c>
How can I open them?
<cristian_c>
Any ideas?
<lautriv>
cristian_c, partitions from NAND ?
<cristian_c>
lautriv: yeah
<lautriv>
cristian_c, loopmount if FS is known.
<cristian_c>
lautriv: I think it easy detect filesystem type
<cristian_c>
it's
<cristian_c>
*to detect
<cristian_c>
lautriv: does it work also with .backup extension?
<lautriv>
cristian_c, linux doesn't care about extensions but magic.
<cristian_c>
some partitions have not filsystem type, but I'm interested only to partition having it :)
<cristian_c>
lautriv: ok, thanks
<lautriv>
cristian_c, some FS are meant to be intentionally not supported.
<cristian_c>
yeah
<cristian_c>
the /system partition, for example, ha ext4 fs
<cristian_c>
*has
<lautriv>
anyone extracted FlashData.bin from a loader supporting 2G RAM ?
<cristian_c>
uhm, I've got 2 GB ram in my device, but I've never extracted files by bootloader
<cristian_c>
from
<lautriv>
cristian_c, you run android.
<cristian_c>
lautriv: yes
<lautriv>
HA ! repo still exists, just link on linux-rockchip.info broken :(
<lautriv>
preparing on single machine with a rockchip consumes more space for tools than the OS o.O
<sjoerd>
lautriv: note there is some rtl8723 support in mainline though, depending on which variant you have
<lautriv>
sjoerd, mainline but not android. box is 3188T
<sjoerd>
oh i don't really care about android
<lautriv>
sjoerd, the only thing i care about android is to get rid of it but mainline doesn't support 3188 graphics.
<sjoerd>
ah yes that requires some more work indeed
<lautriv>
sjoerd, do you mean some work as in "there is a solution around" or rather "someone may patch it" ?
apritzel has joined #linux-rockchip
<naobsd>
lautriv: generally same bootloader is used on all RK3188 devices
<naobsd>
lautriv: if you are using mainline, memory size should be described in dts
<lautriv>
naobsd, i'm bound to 3.0.36 from android. if i boot from nand, i have 2G but not if i boot linux from SD so anything on that SD must init wrong. since the android-kernel contains no word of any RAM specific definition, i assumed the loader part.
<Inode>
is there support available for RK3026 based devices?
<naobsd>
lautriv: which loader are you using on SD?
<lautriv>
naobsd, image was generated by dd (linuxium) i can't tell.
<naobsd>
Inode: what kind of "support" do you expect? here is developers/users community, not commercial support forum
<naobsd>
lautriv: all I can say is something unclear shouldn't be used
<sjoerd>
lautriv: someone needs to write the code
<Inode>
naobsd: is there u-boot and linux kernel source available that features support for RK3026 SoC?
<Inode>
if i was interested in commercial support, i would attempt to contact the manufacturer :)
<naobsd>
sjoerd: single RK loader can support both 1GB/2GB model. I've never seen RK loader _only for xGB_
<naobsd>
lautriv: you can't access serial console? loader (accurately DDR init blob) should print some about RAM
<sjoerd>
naobsd: wrong person :)
apritzel has left #linux-rockchip [#linux-rockchip]
<sjoerd>
Or wrong topic
<sjoerd>
I was talking about drm support on RK3188 not the ram issue
apritzel has joined #linux-rockchip
<lautriv>
naobsd, i'm about to extract the right one from my update.img but all that tools/code is a decade behind or incomplete. sarg talking about extracting with rk29kitchen and then pull it further, that rk29kitchen tells me just "ROM files not found"
<naobsd>
ah sorry
<sjoerd>
no worries
<naobsd>
oh, rk29kitchen, one more unclear thing...
<naobsd>
Inode: as far as I know there is no RK3026 specific SDK/source tree on the public net
<naobsd>
Inode: but some SDK/source tree may have some code for it
<naobsd>
lautriv: are you really sure that kernel on NAND and on SD are 100% same?
<lautriv>
Inode, you may upload a highres picture of that board.
<lautriv>
naobsd, i'm sure they're not and even worse the kernel on that image does not match it's own modules. i pulled several sources for 3188 from github and try to build my own but like mentioned the kernel has neither some config for RAMsize nor does it use any dtb.
<naobsd>
lautriv: I think if you use kernel in NAND for SD, you'll get 2GB RAM on SD too
cnxsoft1 has joined #linux-rockchip
<naobsd>
lautriv: I have no idea why DDR init blob reports only 1GB. it might be better to use more recent loader binary. but, anyway, loader is same, kernel is different, all I can guess is kernel makes difference
<lautriv>
naobsd, that would require more details e.g. partitioning and extraction of initramfs but since all rockchip are messed up with rc4 it is hard to extract anything.
cnxsoft has quit [Ping timeout: 268 seconds]
cnxsoft1 is now known as cnxsoft
<naobsd>
kernel DDR init code might be different. I have no idea because what kernel source you use is unclear too.
<naobsd>
lautriv: sorry, what are you talking? you can dump kernel.img from NAND. you can use it for SD w/o any hard work.
wb|biketron has joined #linux-rockchip
<naobsd>
I really have no idea about "since all rockchip are messed up with rc4"
<naobsd>
kernel binary (things like zImage) can be extracted easily from kernel.img and from boot.img
<Inode>
lautriv: that will mean obtaining a decent camera first
<naobsd>
but no need to extract it for using kernel.img in NAND for SD, just copy it
<lautriv>
naobsd, the RC4 is referring to "Creating a bootable SD" from linux-rockchip.info and i built actually from sources which aloksinha2001 put on github. i can't just copy from nand because no info about start/end.
<naobsd>
lautriv: start/size is written in CMDLINE in parameter very very very clearly. you can dump it from NAND. use it for SD.
<lautriv>
but i had to change content of embedded initramfs anyway.
<naobsd>
lautriv it's in boot partition or boot.img, not in kernel
<naobsd>
lautriv: and what I'm talking (you asked, don't you???) is about RAM size, not about initramfs customization
<lautriv>
naobsd, i may be trapped by wrong information but all i found is telling initramfs is embedded.
BorgCuba has joined #linux-rockchip
<naobsd>
lautriv: it's technically possible to embed it in kernel. but I don't think it should be considered now
<naobsd>
anyway
<naobsd>
I think we are talking about RAM size
<naobsd>
or am I wrong???
<naobsd>
why initramfs need to be mentioned?
<lautriv>
naobsd, i'm aware how it can be done, the question is rather how THEY did it. yes, we talk about ram-size but preparing a proper boot needs both informations anyway, regardless if combined or separate initramfs i have to provide the right modules and early init.
<naobsd>
what I said is just copy images
<naobsd>
I can't say no more, sorry.
<lautriv>
0x00006000@0x00004000(kernel) , will see if androshit has some dd.
<naobsd>
you should read /proc/mtd
<lautriv>
no entry
<naobsd>
your device really use NAND (not eMMC)? is rk...nand.ko loaded???
<lautriv>
nope
<naobsd>
no?
<lautriv>
rk29_ipp, mali and ump are the only appearing in lsmod
<naobsd>
you mentioned "NAND"
<naobsd>
how did you know about it?
<naobsd>
please understand, you don't tell detail, so I cannot know detail, I cannot tell right answer even when I have it...
<sjoerd>
mmind00: working wireless on the rock2 \o/
<lautriv>
sure, but such el-chapo thingie is used to NAND and i have that partitioning parameters which would be pointless for emmc
<mmind00>
yay ... what did you change since the last try?
<sjoerd>
mmind00: Turns out i used the wrong clock name in pwrseq, causing the 32k clock to never actually be enabled
<sjoerd>
In that case the sdio interface works fine, but the main wireless core fails to start up causing timeout errors...
<sjoerd>
:/
<naobsd>
in general (in RK products), partition info in parameter is used even if it uses eMMC :(
<naobsd>
but if you are sure it's pointless, then, sorry, I really have no idea any more.
<naobsd>
I have to do my work
<naobsd>
sjoerd: congrat! :)
<sjoerd>
naobsd: thanks :)
<lautriv>
naobsd, grabbed datasheet, thingie is indeed equipped with emmc, exact partnumber yields no match but family is emmc 8G, probably 200MHz ( or above since last dihits are higher )
<lautriv>
*digits
<lautriv>
maybe someone found exact match for THGBM4G6D2HBAIC
<lautriv>
(Toshiba)
<lautriv>
bbl
<naobsd>
lautriv: all I can say is it must be partitioned, otherwise android cannot work. but if you are sure partition info is pointless, I really have no idea how it works. sorry.
<naobsd>
probably all my info is pointless for your device
<lautriv>
naobsd, you got that part wrong. i said the nand-map is useless since emmc is usually partitionen for real aka, not list with some offests.
<naobsd>
all I know is "usually" doesn't help in RK world, but all my word will be pointless too.
rperier_ has joined #linux-rockchip
<naobsd>
you know a lot what I don't know
<lautriv>
that is what me makes me angry ..... there should not be any difference.
<naobsd>
I'm not talking should be or should not be at all
rperier has quit [Ping timeout: 246 seconds]
<naobsd>
all I can talk is the fact in front of me
<lautriv>
naobsd, point is, the only info is that long line about sizes and offsets, no real partition and no entry in /proc/mtd.
<lautriv>
... there is indeed something in /proc/partitions but no output by fdisk -l
<naobsd>
as I said it must be partitioned.
<naobsd>
there is no partition info for fdisk
<naobsd>
well
<naobsd>
if you want only about "it should be", I can't help you, sorry.
<naobsd>
s/want/care/
<naobsd>
I have to go out soon
<lautriv>
naobsd, regardless of the right or wrong way, partition for kernel which would be /dev/nlock/mmcblk0p2 is not mountable, hence i would need to dd anyway, and if so i should not grab the whole, no ?
<lautriv>
*/block/
naobsd has quit [Remote host closed the connection]
Inode has quit [Quit: ]
Inode has joined #linux-rockchip
<lautriv>
anyone used to linuxium's stuff ?
<lautriv>
nvm, found.
<cristian_c>
I've used marvin scrpts
<sjoerd>
mmind00: did you play with bluetooth by chance on your veyron?
<mmind00>
sjoerd: nope, I don't think I even own some bt device to talk to :-)
<sjoerd>
you don't have a phone or other computers ? :)
<mmind00>
sjoerd: you're right ... as I never consciously use bluetooth, I really forgot about the phone ;-)
cristian_c has quit [Quit: Bye]
<mmind00>
sjoerd: are you facing any specific problem?
<sjoerd>
No just starting to give it a go
<sjoerd>
i've only worekd with usb attached (either dongles or internally) bluetooth before
<sjoerd>
mmind00: btw what's with the ``ttyS0 - failed to request DMA'' errors?
<mmind00>
sjoerd: the 8250 serial driver tries to get the dma channels ... we currently don't carry these dma properties [patch from rperier_ already pending], due to the broken pl330 and the still pending on the lists fixup from one Rockchip guy
<mmind00>
sjoerd: and I don't think they are acutal errors ... the dma channels are optional, so I guess that counts more as hint/info :-)
<sjoerd>
That was obvious from having a working uart :)
<sjoerd>
but they annoy me every so often
<mmind00>
sjoerd: hehe ... the dmaengine maintainer seems to be quite occupied, hopefully he'll get to it soon
GriefNorth has quit [Quit: WeeChat 1.3]
Bludot has quit [Quit: Connection closed for inactivity]
cnxsoft has quit [Remote host closed the connection]
<lautriv_>
when i copy a parameter.img from emmc to SD, i get E:Invaid tag(0x00000000)! but checked the content several times ?
rubensm has quit [Ping timeout: 246 seconds]
rubensm has joined #linux-rockchip
Bludot has joined #linux-rockchip
BorgCuba has quit [Quit: leaving]
ssvb has quit [Ping timeout: 240 seconds]
gb_master has joined #linux-rockchip
JohnDoe_71Rus has joined #linux-rockchip
markm has quit [Ping timeout: 240 seconds]
markm has joined #linux-rockchip
ssvb has joined #linux-rockchip
markm has quit [Ping timeout: 240 seconds]
akaizen has joined #linux-rockchip
jas-hacks has joined #linux-rockchip
maz has quit [Quit: Leaving]
markm has joined #linux-rockchip
apritzel has quit [Ping timeout: 264 seconds]
tlwoerner has quit [Quit: Leaving]
<sjoerd>
mmind00: do you happen to know of any know issues with the uart driver when it comes to changing speed
ChanServ has quit [*.net *.split]
<sjoerd>
One of the fun things i'm hitting with bluetooth is that once the kernel tries to change the speed on both sides, the uart side doesn't seem to behave properly and the chips stop talking to eachother
<mmind00>
sjoerd: is it a fractional rate?
<sjoerd>
It's even when swithing to the same rate
<sjoerd>
I havent looked at what happens with the clock setup tbh
<sjoerd>
It took me quite some time to root cause this (that and figure out how all this is meant to hang together currently)
ChanServ has joined #linux-rockchip
<mmind00>
sjoerd: some hours ago, I just send a ping on the fractional divider series ... so that might be one issue
<sjoerd>
I saw :)
<sjoerd>
Thanks for that
akaizen_ has joined #linux-rockchip
rubensm has quit [Ping timeout: 240 seconds]
akaizen has quit [Ping timeout: 246 seconds]
markm_ has joined #linux-rockchip
markm_ has quit [Client Quit]
rubensm has joined #linux-rockchip
xcasex has quit [Ping timeout: 246 seconds]
gb_master has quit [Remote host closed the connection]