wdtz
-
Posts
49 -
Joined
-
Last visited
Content Type
Forums
Store
Crowdfunding
Applications
Events
Raffles
Community Map
Posts posted by wdtz
-
-
4 hours ago, mb16 said:
adb root worked out of the box, devices like this one seem to be rooted by default?
Clearly,, you have not read whole thread,, about 5 pages back, but differently
in adb# ,, ls -l /sbin/su; ls -l /system/xbin/su ,, notice that they are not the same?
(google) verity off; mount -o remount,rw /system; cd /system/xbin; mv su su.bak; cp /sbin/su .
and then root will work in android
FWIW
-
4 hours ago, nathanwz said:
can anyone give me a hint what the problem might be
I have one of these, in the simple shell ls -l /dev/block will show at best mmcblk1 (the emmc),
(almost) never the sd card
I think it's something to do with the regulator,, pwm does not seem to work,,
the android dtb seem to use simpleReg vrs rk805-regulator
I did a "hack" bypassing gpio, barely works, lots of errors in dmesg
(regulator errors,, boots and runs from sd)
The other method is to copy rootfs to a usb stick, rename the partitions (e2label)
-
16 hours ago, armar said:
dding just the fw files and the config with the old 8723bs.ko does not work either (modprobes...but no wifi).
Dmesg , if all is well, will give a message, ie (dmesg |grep firmw)
usb 1-1.4.3: rtl8xxxu: Loading firmware rtlwifi/rtl8192cufw_TMSC.bin
for a usb dongle
-
17 hours ago, armar said:
What do the firmware files for wifi look like? And how does the module know where to pick it up from?
Well, it varies, alas. The firmware files can be bin or fw or other. Usually from /lib/firmware or a folder there BUT
can be in /usr/somewhere, this is not common
I don't know how the ko knows where to find the firmware
Firmware are for the wifi chip, are version and OS independent
17 hours ago, armar said:is a mountpoint /vendor which is on /dev/mmcblk2p17 . Is this the android directory?
It is 1 of them, there are 9 or 10 in total plus uboot and miniLoader and ?
Maybe I'm telling you something you already know, / in Android is only in mem,
loaded from the boot part, a gz cpio. Most of the folders in / (android ) are links to
the appropriate partition
17 hours ago, armar said:/etc/firmware/rtl8723bs_fw ./etc/firmware/rtl8723bs_config./etc/firmware/rtl8723bs_VQ0_config /etc/firmware/rtl8723bs_VQ0_fw
I would guess that these are the ones you need, try copy them to lib/firmware/rtlwifi
-
6 hours ago, mg-to said:
Am I correct?
As I said, usually an android dtb won't work for linux, but it is a guide on how to change
an almost working dtb
For wifi,, 3 things have to be "right". A reference in the dtb, a module, and the firmware
You will really have to get dtc (device tree compiler) and learn how to use it,
it is not that hard.. Harder is making the "right" device tree
IMO,, you really should have a "perfect" boot with sdcard (everything works, no glitches)
before thinking about overwriting emmc
----more--
For wifi the reference in the dtb is not at all important, only if you
want it to work during boot. I have seen wrong wifi chip referenced in dtb,
modprobe right module (if firmware is there) and the wifi works
This assumes it is talking to the right gpio
There is a discussion about that in this thread a few pages previous
-
5 hours ago, mg-to said:
DEV_EMMC=/dev/mmcblk0
No, mmcblk0 is almost certainly your sdcard, just from the partitions on it
It's really common for android to have boot0, boot1, rpmb on the emmc, that is likely where it is (blk2)
If you have checked your evb.dtb by turning it into a dts, check dwmmc@ff500000, and @ff520000
for bus-width = < 0x04 >; or 0x08 and the aliases { ,, then you will know for sure
You do know this will wipe out your android? Before you do that, it would be good to
extract the android dtb because it is sometimes hard to get the right image for your hardware,
there are so many slightly different images
-
7 hours ago, balbes150 said:
The easiest way to get the desired dtb is to unpack the native firmware file
There are dozens of different tv boxes, easy to get wrong image, usually does not "work out"
in output/Image (specifically my box, h96max+-ssv, from the vendor, ssv is wifi)
ls -l *.img
-rw-rw-r-- 1 d d 1048576 Nov 11 12:35 baseparameter.img
-rw-rw-r-- 1 d d 1708536 Nov 11 12:35 boot.img
-rw-rw-r-- 1 d d 19036180 Nov 11 12:35 kernel.img
-rw-rw-r-- 1 d d 49152 Nov 11 12:35 misc.img
-rw-rw-r-- 1 d d 168060 Nov 11 12:36 oem.img
-rw-rw-r-- 1 d d 8971308 Nov 11 12:35 recovery.img
-rw-rw-r-- 1 d d 5592576 Nov 11 12:35 resource.img
-rw-rw-r-- 1 d d 1547079912 Nov 11 12:36 system.img
-rw-rw-r-- 1 d d 4194304 Nov 11 12:35 trust.img
-rw-rw-r-- 1 d d 4194304 Nov 11 12:35 uboot.img
-rw-rw-r-- 1 d d 163962992 Nov 11 12:36 vendor.imgand in output
-rw-rw-r-- 1 d d 162126 Nov 11 12:33 boot.bin
drwxrwxr-x 4 d d 4096 Nov 30 01:54 Image/
-rw-rw-r-- 1 d d 162126 Nov 11 12:35 MiniLoaderAll.bin
-rw-rw-r-- 1 d d 724 Nov 11 12:35 package-file
-rw-rw-r-- 1 d d 981 Nov 11 12:35 parameter.txtStill have to extract dtb from resource.img
-
14 hours ago, mb16 said:
Could I mount /dev/mmcblk2boot<x> somehow?
Ah, android.. No, if you run fdisk -l on your mmcblk2 file you will see 3 4mB partitions,,
where is the rest?? 15 or 30 gB missing, right?
The rest is in a mtd filesystem,, under the stanza "chosen" you may see something like this
bootargs .........
mtdparts=rk29xxnand:0x00002000@0x00002000(uboot),0x00002000@0x00004000(trust),\
0x00002000@0x00006000(misc),0x00000800@0x00008000(baseparameter),0x00008000@0x00008800(resource),\
0x00010000@0x00010800(kernel)...and so on (#'s are blocks/sectors,,,, length@offset (+12M))
You want the first 70k of resource, starts 2k in, magic is 0xd00dfeed, about 64k long. The rest is 2 png or jpg,
the splash screen, 1st is for 720p, 2nd (larger) for 1080p
I don't know how to mount mtd filesystems, AND I have had no luck
converting android dtb to linux dtb,, nothing that I have found on google either
Usually it just doesn't boot, fails very early on kernel load (serial console)
-----
but usually with these tv boxes you have root in adb,, have you tried?
If so, cd to proc, somewhere in there is the device tree, use busybox tar and write it to sdcard,
and then to your computer (on an ext fs), then expand and use dtc -fs on the expanded directory,,
then you will have the android dtb
Or,, easier, if you have adb root,,,
find which partition resource is , /dev/mmcblk2p??? ,,
dd 70k to internal sdcard (I think this is ext fs,, vfat will lose permissions etc)
then to computer, chop leading 2k, trim excess from end (hexeditor and dd)
BTW the reason adb has root but not android...
the su in /system/xbin is no good,, the one in /sbin is the right one
Use verity off to make /system rw, then you can mv xbin/su xbin/su.bak
and /sbin/su -> xbin and then get su in android
-
Err, I have a mt7601u,, 148f:7601,, I don't think it is so good,, in iwconfig I see
Tx excessive retries:35031 Invalid misc:1837 Missed beacon:0
(this is with uptime 9d 13h)
-
A bit late, but I have narrowed down the problem, for at least my TV box
It is a problem with the regulators both linux and android have problems (at least their dtbs')
linux uses rk805,, android pwm/simple
with 805,, 00 1e 01 ,, sdio doesn't work, so no mmc (gpio, offset, function)
sdmmc-regulator
with pwm 02 04 01 and 02 05 01 half fails (there is power to cpu and not cooked),, no gpu and debugfs
vdd-center, vdd-log
That is why the liantong had mmc but no vid, uses pwm, same as any other android dtb
I did luck into a blodge, works for armbian but not Balbes' TV
(cut out gpio0, go directly to pwm, not sure what to do with pinctrl-0, sorta works)
Clearly,, a dtb is a looong way from a bios,,, it would be really helpful if
as much as possible,, if phandle , if they corresponded to real hardware, had the same value as linux devices,,ALWAYS
Standards are a wonderful thing
--later--
A report on new wifi-dev, No sd found
for rk3328-8d4-v1.1 ,, h96Max+
of the 5 dtbs, both liantong and evb-android crashed
z28 no mmc,,,. roc-cc emmc, no sd,,,, evb emmc, noticed sd but no p1,p2 no usb2 (phy-supply)
--a day later--
wanted to try my "blodged" dtb,,,,,Unable to read file my.dtb
but uboot ls saw it OK, had been lower-cased, so change extlinux.conf, ,
a uboot artifact, had NOT been lower cased,,, then tried one of the originals
Unable to read file rk3328-roc-cc.dtb,, spelling correct,, uboot ls sees it
?????,, nothing on google
-
And now, after trying about 20 dtb's, only the liantong (all 3) see mmc0 and mmc2 and the partitions
(and despite running dtdiff until my eyes are crossed I don't see anything significantly different in dwmmc)
Unfortunatly they all stop, about 4 seconds into the kernel load with
rockchip-dmc dmc vop bandwidth to dmc rate (all 3 liantong)
I don't think this is something tuneable in a dtb
The post above, those are spurious error msgs,, but about 1/2 of all the others have (early,,8x)
rockchip_mmc_get_phase invalid clock rate
all those apparently dangling clocks point to the clock name, I have no idea what is in that
I don't have enough hd space or a powerful enough box to compile, alas
--edit--
the 2nd and 3rd liantong are essentially the same, only fastest cpu removed
-
Bables,, after viewing about 20 dmesg logs, I see some things in common
(re no mmc,, ls -l /dev/block)
The 1 and only 1 showing any mmc was the legacy dev with roc-cc ,,Ubuntu_bionic_dev_4.4.161_desktop,,
and I see rpmb, boot0 and boot1,, that's all
I think lacking (in init) is rkpart for the rest of mmcblk0 (the emmc)
Most dtb lack (in aliases) mmc0 = "/dwmmc@ff500000"; and so on
This is an error in all? most anyway
I don't know how severe this is (all? most)
rockchip-dmc dmc: Cannot get the regulator "center" (missing stanza?)
dwmmc_rockchip ff510000.dwmmc: No vmmc regulator found (also ff500000 and ff520000)
dwmmc_rockchip ff510000.dwmmc: No vqmmc regulator foundand minor error
dwmmc_rockchip ff500000.dwmmc: num-slots property not found, assuming 1 slot is available
and probably not related
rk-vcodec ff360000.rkvdec: vcodec regulator not ready, retry
many of the uboot are older, and for roc-cc
--edit--
and yet, when I try with normal armbian (console=both) there are no errors in any dwmmc lines,
but no mmcblk devices or partitions are discovered (except by uboot, kernel, initrd and dtb get loaded OK)
-
I must say that I am most un-impressed with the dtb system
I have been struggling with it for days, with almost NO success
at first I thought I was a fuck-up none of my modified dtb's seemed to work,
now,, after getting a serial connection (which is very little help, except for scroll back),
I see that it is not me,,
** Unable to read file /dtb/rockchip/an-untouched-dtb **
and then the dtb',s every one different, pin#'s ,clocks, everything
I even tried making a dtb -fs from a running android, (tarred /proc/device-tree) lots of errors in android dmesg,
a wonder that it booted, and then edit androidisms out,, that didn't work either
What a pain,, I'm starting to think this is more trouble than it is worth
--edit--
It could find the dtb, previous 2 lines retrieving file,, reading file
oops the retrieving and reading lines were in error, the file didn't exist
Too many different versions, easy to lose track
-
Bables for your next update,, this is from android kernel config
CONFIG_SSV6051=m
CONFIG_SSV6051_P2P=y
CONFIG_SSV6051_SDIO=ybecause my H96Max+ has a SV6051P chip
thanks
-
Sorry, looking at dmesg (android) firmware is in /vendor/etc
Yesterday,, another non-booting dtb, grr
-
nytoc,, by the way, most firmware is distro and version agnostic ,,
byte-order certainly matters, but probably not arch
Sometimes it is in non-default locations and doesn't load,
sometimes the package manager forces unnecessary restrictions
sometimes ethernet has firmware
--more--
Since you have a rooted android?? ,,
you already have the firmware,, /vendor/lib/firmware/some-directory-with-8723-in-it's-name
-
>Did you check if any mmcblk0\1\2 appeared ?
I don't understand what you mean here?
This was under rooted android,, so if I wanted to I could have dd'd them somewhere
Under your distro, I have never seen any mmc devices, bad dtb I think
>You need to try the tools to extract the firmware file. Maybe then you can get the dtb.
You mean fdtdump? Initially, did not work, had to strip leading 2k to magic d00dfeed
So far, all the dts that I have looked at are a MESS, pointing to non-existant clocks,
phandle confusions, missing lines etc. It's a miracle that they boot
Like there is 100 years of cruft in them, hard to believe these are newish devices
if you can get a boot,, dmesg> to-something-to-study,, lots to fix
(the initrd has dmesg in it, but unless it recognizes a usb stick you nowhere to write to)
Nytoc, is the wifi loading the firmware?, dmesg | grep firmware Balbes's distro doesn't seem to come with firmware, apt-get
-
>- microSD + USB: I can reach the Kodi interface but I don't have any network support
Not even ethernet?
If your android is rooted, then,, su; lsmod ,, you should be able to tell which is wifi module
OR,, take it apart and look at chip markings, the antenna will go to wifi chip (for me sv6051p)
I assume you have both ethernet and wifi in android?
-
I was also going to write about too high iozone numbers, but not as high as Bernie_O
I see up to 1.6M with larger blocks, evo, evo plus, evo+, 32 & 64GB
Interesting, numbers are about equal for 2 year old evo+ and new evo plus
The sandisk 64 GB A1 is also about equal,,, BUT, but the 32 GB is shit (sandisk ultra A1)
1/6 the speed? less? I greatly regret buying 2 of them
I also tested with the iozone parameter -s 500M, numbers were about the same
really should only see 500 -600K, so something is wrong
this is with a U1 or maybe U3 reader attached to usb3
Here 32 GB sandisk ultra A1 ,, ---slow, slow, slow
102400 4 1811 1745 7607 7610 7054 1384
102400 16 3327 4794 23309 23288 21946 4240
102400 128 17653 14829 90395 90565 67828 10891
102400 512 16577 15942 93699 93687 86111 13926
102400 1024 18142 15978 94203 94270 90049 14355
102400 16384 17585 15736 94533 94624 94274 20876and comparison sandisk ultra A1 64GB ,, notice the "too high" numbers
102400 4 239202 281839 301375 296697 38806 38378
102400 16 562704 667356 708478 734044 144208 142922
102400 128 979088 1420801 1688792 1694830 790470 723220
102400 512 953029 1377012 1514590 1580078 1209057 1128835
102400 1024 975497 1362280 1555382 1569586 1377268 1218438
102400 16384 918773 1259275 1379822 1381468 1333371 1214032iozone -e -I -a -s 100M -r 4k -r 16k -r 128k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2
above 64GB with Bernie's dd
time sh -c "dd if=/dev/zero of=testfile bs=100k count=50k && sync"
51200+0 records in
51200+0 records out
5242880000 bytes (5.2 GB, 4.9 GiB) copied, 153.192 s, 34.2 MB/s -
I had to do it twice, my regular email provider does not let you leave their site,
I always have to open a new tab. So, my 1st registration attempt failed, had to use
my gmail account instead. All in all, a bit of a pain
-
Hi Balbes
Like nytoc, during boot it could not find rootfs
This is a H96Max+ ,, board labeled RK3328_8D4_V1.1
It has sv6501P wifi
About 1/2 the dtb don't even start to boot,
evb-android, both rock64, roc-cc, box-liantong
The rest, evb, box, box-z28, box-trn9, rockbox,
all have no mmc (ls /dev/block}
They will all boot with rootfs on usb stick
obviously uboot can read mmc, since uinitrd gets loaded (and kernel)
Once or twice, booted only with usb, I'm not sure which, maybe
box or rockbox
The next update, please compile ssv6051 module
I tried to strip the dtb out of resource.img, doesn't boot (mmcblk2p5)
Maybe there was more in that 8M partition, the dtb <60K
The android on it (8.1) is 4.4.140 (and rooted, that's how I got resource.img)
thanks for the work
--more--
There was no way I could make it go to maskrom with the button,
maybe there is a trick,, also could not do adb with usb,, lsusb never showed it
Had to use wireless adb
Armbian for TV box rk3328
in Rockchip CPU Boxes
Posted
Err,, adb is not "android",, android is the graphical shell,,
hypothetical example,, if you had firefox installed,, typing firefox in adb would NOT startup firefox
But, for sure , adb is very powerful, despite being a considerable pain
More on point, adb uses the su in /sbin,, android uses the su in /system/xbin,, it is no good