<naobsd>
KDDI is one of big 3G/4G network service provider in Japan
Soopaman has quit [Quit: Leaving.]
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
GriefNorth has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
Soopaman has joined #linux-rockchip
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
Astralix1 has joined #linux-rockchip
Astralix has quit [Ping timeout: 250 seconds]
Soopaman has quit [Quit: Leaving.]
Valduare has quit [Quit: Valduare]
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
dlezcano has quit [Ping timeout: 272 seconds]
dlezcano has joined #linux-rockchip
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
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
akaizen has quit [Remote host closed the connection]
akaizen has joined #linux-rockchip
Soopaman has joined #linux-rockchip
akaizen has quit [Ping timeout: 272 seconds]
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
naobsd has quit [Quit: Page closed]
naobsd has joined #linux-rockchip
akaizen has joined #linux-rockchip
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
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
Soopaman1 has joined #linux-rockchip
Soopaman has quit [Read error: Connection reset by peer]
irsol has quit [Ping timeout: 250 seconds]
irsol has joined #linux-rockchip
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
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<hramrach_>
naobsd: people ask for the simplest way possible, of course. Ready-made images for many tasks are available for other platforms
<hramrach_>
comparing/confirming tool output is kind of a wasted effort when the tool is opensource and whoever wrote it already knows what it outputs just did not write the man page or readme for it
Avagetto has joined #linux-rockchip
<hramrach_>
it's multiply wasted when every user of the tool has to do so
RayFlower has quit [Read error: Connection reset by peer]
<naobsd>
where is "image for audio streaming" "image for LAMP server" "image for software developemt" etc etc from Ubuntu.com? they provides only "desktop" and "server"
<naobsd>
same for gentoo, arch, etc etc
RayFlower has joined #linux-rockchip
<naobsd>
I don't say "everyuser must confirm output of tool"
<naobsd>
I said "do rkcrc -k"
<naobsd>
^I just said
<naobsd>
it just work
<naobsd>
I don't understand why people try to do different thing after right way was given
<userx____>
hi guys, I'm interest to get information about bootloader in Rockchip devices, in privacy for rk3188
<naobsd>
what I said is, "if you cannot trust my way, you can confirm yourself"
<naobsd>
userx____: no official document is available. if you can read binary, it's best source of info
wildea01 has joined #linux-rockchip
<userx____>
naobsd: not so 'low level' information :). I just want to find out how it works, what the internal loader is capable to load, where the second load should be loacated, what is that 'parameter' file that should be located at the beginning and what i can configure with it
<naobsd>
userx____: why do you want to know "where loader/parameter is located"? it sounds low level thing
<naobsd>
userx____: you can put parameter at logical 0 sector on NAND, or physical 0x2000(=logical 0) sector on SD/eMMC
<userx____>
naobsd: just interesting what bootrom/external loader are capable to do and are they configurable. I don't want to modify them
<naobsd>
userx____: on RK3188, NAND/eMMC/SD/USB can be used for device to load loader
<userx____>
So, parameter file must be on address 0, bootloader must be on 2000 and kernel at A000 ? Is that obligatory ?
<naobsd>
^NAND/eMMC/SD/USB is known as bootable device, but I'm not sure it's complete or not
<naobsd>
bootloader is not in logical area
<userx____>
naobsd: where RK3188Loader(L)_V2.10.bin should be loaded then ?
<naobsd>
1: CPU loads bootloader from NAND/eMMC/SD/USB, 2: bootloader loads parameter from logical 0, 3: bootloader loads kernel etc from logical addresses described in parameter
<naobsd>
userx____: I think it's low-level thing
<naobsd>
userx____: loader bin is stored at somewhere in NAND, or 92nd sector on eMMC/SD
<naobsd>
userx____: and loaded to at 0x60000000 or around there on RAM
<naobsd>
92 and after on eMMC/SD
<userx____>
naobsd: ok, if my NAND goes broken, and I resolder it where exactly to write this bootloader? How CPU maps physically this NAND to Logical addresses ?
<naobsd>
userx____: you can flash loader bin to empty NAND without information about where NAND controller is mapped
<userx____>
And which booloader reads this parameter file - the internal bootrom or external boot loader ?
<naobsd>
<@naobsd> 1: CPU loads bootloader from NAND/eMMC/SD/USB, 2: bootloader loads parameter from logical 0, 3: bootloader loads kernel etc from logical addresses described in parameter
<userx____>
what is "KERNEL_IMG: 0x60408000" for in parameter file ?
<naobsd>
kernel(any ARM code stored in logical kernel partition) load/entry address to boot
<naobsd>
(kernel can be stored in boot/recovery partition as ANDROID! image too)
<naobsd>
bootloader(RK implementation) load kernel from logical kernel partition described in parameter to KERNEL_IMG address (after removing KRNL/ANDROID! header etc)
<userx____>
naobsd: so if I crap the content of boot partition (addr: 0x00002000) it shall not destroy external boot loader ?
<naobsd>
^if you use non-RK bootloader, it may not be true
<naobsd>
userx____: of course, you cannot break bootloader by writing something to logical partitions/whole logical area
<naobsd>
userx____: you can confirm by flashing garbage to boot partition, you still can talk with bootloader
<userx____>
naobsd: except those 'parameter' file, right ? You are saying that I can't crap external bootloader because it is located on unmapped physical area ?
<naobsd>
^or bootloader try to load/boot kernel from recovery in this case
<naobsd>
userx____: there is a command to erase bootloader in official flash tool from RK, but you cannot break bootloader with command/tool for "flash image to logical partition"
<naobsd>
userx____: you can break/erase parameter too
<naobsd>
1: CPU loads bootloader from NAND/eMMC/SD/USB, 2: bootloader loads parameter from logical 0, 3: bootloader loads kernel etc from logical addresses described in parameter
<userx____>
naobsd: then what is that radxa 'unbrick' for? What it 'Unbrick', because i was thinking that it just reflash external bootloader
<naobsd>
it just fail at 2: and waiting
<naobsd>
userx____: please define "brick" at first. you can make radxa unbootable by erasing kernel/boot/recovery
<naobsd>
and/or parameter
<naobsd>
and/or erasing bootloader,
<naobsd>
userx____: to unbrick, all you need is flashing some or all images you broke
<naobsd>
e.g. if you broke parameter, you need to reflash working parameter only
<naobsd>
no need to understand low-level thing
<userx____>
naobsd: i never reach this 'step' but I read about it into RADXA wiki. And concatenating the infrom that you provide I'm curious how then i can brick RADXA, as I can load data at any logical location with RKAndroidTool
ganbold_ has quit [Remote host closed the connection]
<PaowZ_>
you cannot brick radxa, actually.. as naobsd says, you just have to flash a provided image to get the thing back.
<naobsd>
please define "brick" ;)
<naobsd>
even if no bootloader is flashed
<naobsd>
even if no bootloader is flashed, it's not "brick" because on-chip(in CPU) rom can load loader from USB
<naobsd>
^mask rom mode
<userx____>
but then who decide how to map physical NAND to logical address, how big is that NAND and etc ?
<PaowZ_>
I think he talks about writing data into logical partitions versus writing into physical partitions, if I'm not wrong..
<userx____>
naobsd: bring - the state in which you must connect pin 8 and 9 on the nand...
<PaowZ_>
writing into physical is more likely to cause damages than writing into logical, I'd say..
<naobsd>
userx____: rockchip doesn't open any info about NAND.
<naobsd>
userx____: and that information is not needed to reflash bootloader, mask rom mode and official RK flash tool does everything we need to reflash NAND
<naobsd>
userx____: short 2 NAND pins is needed only when bootloader is flashed in NAND and loaded by CPU but doesn't work correctly
<naobsd>
userx____: if you erase loader in NAND, (and no other bootable media such as SD is supplied), CPU just go mask rom mode, (CPU cannot do any other thing without loader)
<userx____>
naobsd: that's the case when talking for radxa, but what about if you want to design custom hardware with different components for example ? I suppose CPU should know how to setup or whatever
<userx____>
naobsd: or rockchip is not supposed to build custom hardware with it ? I'm not kidding, i just not understand why can't find any info..
<naobsd>
there is supported NAND chip list doc in SDK
<userx____>
in SDK ?!?
<userx____>
what sdk ?
<naobsd>
and you have to connect pins between cpu and other components/connectors/etc properly
<naobsd>
well
<naobsd>
to make custom hardware
<naobsd>
knowledge about hardware is needed
<naobsd>
I don't think it can be provided as easy howto etc
<userx____>
naobsd: i already connected SPI device to some gpio and make it working but with bitbang spi driver
<userx____>
naobsd: that why master spi kernel driver don't works is other story that .... anyway
cristian_c has joined #linux-rockchip
<cristian_c>
Hi
<cristian_c>
naobsd, I don't know how to exactly use cpio command
<naobsd>
please define "custom hardware". "adding SPI device" and "designing whole board such as RR" is totally different
<userx____>
naobsd: by custom hardware i mean designing board. I know very well what the difference is ..
<userx____>
naobsd: For example now i have radxa, but tomorrow i can get any other rockhip board. And this one - I want to use radxa with 3.10 kernel and android on NAND - then why i can't and what the problem with NAND driver is ?
<naobsd>
userx____: there is no official 3.10 kernel for RK3188
<userx____>
btw, i'm not sure that it is a right place to ask about, but that is the only informational source that i found, really
<naobsd>
userx____: no one except RK has answer about NAND
<userx____>
naobsd: and here is not , how to say - 'official' rockchip channel ?
<naobsd>
userx____: of course, here is not official rockchip channel
<userx____>
naobsd: understand.
Soopaman1 has quit [Quit: Leaving.]
<userx____>
naobsd: btw who makes given kernel version for rockchip 'official'?
<naobsd>
userx____: sorry, I cannot understand your question
<naobsd>
userx____: rockchip makes rockchip official kernel
<PaowZ_>
rockchip does it
<naobsd>
^only
<userx____>
naobsd: I mean if i want to send official complaint about SPI who I should target ? :)
<naobsd>
userx____: sorry, I don't know. if there is email address or something, I cannot guarantee they reply your question
<naobsd>
userx____: but, recently rockchip is joined Linux mainline developer community
<userx____>
naobsd: I'm sure they shall not
<naobsd>
oops, "rockchip joined"
<PaowZ_>
you can always give it a try..
<naobsd>
they are doing e.g. submitting patches, answering question, fixing docs, etc
<userx____>
naobsd: where I can see what is the latest supported linux kernel, or there is no such thing ?
<naobsd>
"for Linux mainline"
<naobsd>
and "for Chrome OS"
<karlp>
userx____: spi is one of the things in linux mainling getting worked on, you may want to look at recent patches if you're having spi problems
<naobsd>
I'm not sure they can answer about kernel in their Android SDK
<userx____>
karlp: i was looking but was not able to integrate into current radxa stable
<naobsd>
understand SPI controller from mainline code
<userx____>
and i need to run android, make simple jni bridge to spi device and control it from Android UI
<userx____>
just simple :)
<naobsd>
and fix/reimplement driver for 3.0 :)
<karlp>
that sounds like fun :)
<naobsd>
userx____: did you try/compare SPI driver in radxa-dev-jb/kk branch?
<userx____>
not really
<userx____>
there is 10 kernel repositories
<userx____>
after second i check i decide to go with spi gpio emulation
<naobsd>
not so bad workaround
<naobsd>
btw, I cannot count how many repo/branches on the net, but I "specified" only 2
<naobsd>
I didn't say try all random repos
<userx____>
yeah but i did not expect to search for workarounds .. Even found bug into bitbangxxx.h which means that officially nobody is running SPI :)
<userx____>
naobsd: ok, i shall check it. You just propose that now :)
<naobsd>
I don't know any device which uses SPI ;)
<userx____>
asd7846n
<naobsd>
sorry, I don't know any RK based product which uses SPI
<userx____>
in current i use it as touchscreen controller
<userx____>
naobsd: there is spi support, spi sample drivers, so i suppose at least the guy who made that support
<naobsd>
probably I have accelerometer...
<naobsd>
there are also many bugs in RK kernel :)
<naobsd>
not only SPI
<userx____>
but who can gives guaranty that it is not the same with i2c controller ?
cristian_c has quit [Ping timeout: 272 seconds]
<naobsd>
well
<naobsd>
about this kind of thigns, generally, I don't expect everything works fine from beginning
<naobsd>
not only cheap SoCs from china
<naobsd>
I can see a lot of errata e.g. freescale etc
<naobsd>
sometimes they issues errata weekly ;)
<naobsd>
same for board
<userx____>
naobsd: which is not bad, the support is in progress :)
<naobsd>
often it need hardware patching(jumper etc)
<userx____>
i don;t think that RK are cheap
cristian_c has joined #linux-rockchip
<userx____>
probably Freescale is cheaper, OMAP also
<naobsd>
userx____: cheap. you cannot get weekly errata magazine!
<userx____>
allwander
<userx____>
naobsd: ?
<naobsd>
you cannot get good support = cheap
<userx____>
allwinner actually..
<naobsd>
I love Rockchip, Allwinner, etc etc
<userx____>
naobsd: but all soc are similar price, i can't understand whick one is cheap
<naobsd>
and I think they are cheap ;)
<userx____>
btw Allwinner seems are better supported. I saw some posts where SPI was working :D
<naobsd>
quality is cheap, and it needs extra cost to make it work ;)
<naobsd>
well
<naobsd>
SoC is good
<userx____>
yeah but to not have even driver for NAND ...
<naobsd>
I really like it
<userx____>
as rockchip
<naobsd>
some people blame rockchip is not open
<naobsd>
but I think some other majors also have binary blob
<userx____>
this is ridiculous
<userx____>
and this SDK you said, what is for ?
nighty^ has joined #linux-rockchip
<karlp>
it's what rockchip hands to vendors building a product,
<karlp>
and then the vendor hacks it a bit whenthey change to use a different wifi chip, or a different power supply
<userx____>
are they freely available ?
<karlp>
and then, _mayb_ that vendor releases their version, and hence we have xxx hundred slightly different 3.0.x kernels.
<karlp>
naobsd: what's the links to dl.linux-rockchip.info again? it's not giving a listing ont he root?
<karlp>
userx____: naobsd set up some mirrors for a few of the official rockchip sdk's he's collected
<naobsd>
karlp: I guess there is a rk3066 project in mozilla.org
<naobsd>
I'm not sure at all
GriefNorth has quit [Quit: WeeChat 1.0.1]
Valduare has joined #linux-rockchip
<userx__>
hmm regarding RKNandFlashSupportList Ver2.63_2013_12_15 very limited set of memory chips are tested ... same is for DRR
<userx__>
looks like this soc is design for some big china vendor, with specific hardware requirement :)
<userx__>
and not exactly for mass production
<naobsd>
I think if "limited set of memory chips" are well supplied, it's ok
<naobsd>
^tested with major/popular chips
<naobsd>
and don't forget this is "very poor document" world. I don't believe that document is up-to-date
naobsd has quit [Quit: Page closed]
<Valduare>
hey guys, I have a few mk808 devices and one 802iii
<Valduare>
these are my first arm devices.
<Valduare>
are all arm devices like this where you have to do a rootfs and special boot loader stuff and what not?
<Valduare>
not able to upgrade ubuntu to new releases etc?
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
Avagetto has left #linux-rockchip [#linux-rockchip]
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
benja` is now known as benja
benja is now known as benja`
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
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
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
Soopaman has joined #linux-rockchip
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
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
wildea01 has quit [Quit: leaving]
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
khilman has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
GriefNorth has joined #linux-rockchip
Soopaman has quit [Ping timeout: 244 seconds]
Soopaman has joined #linux-rockchip
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
Astralix1 has quit [Ping timeout: 250 seconds]
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<dlezcano>
mmind00: sorry to bother you
<dlezcano>
mmind00: do you know how I can have the rootfs in the sd card
<mmind00>
dlezcano: hi ... it's hard to "bother" me, so no problem ;-)
<dlezcano>
mmind00: or embed the initrd in the kernel
<mmind00>
dlezcano: I normally always only embedded a small initramfs in the kernel ... using the initramfs source option under "General Setup" in the kernel config
<dlezcano>
mmind00: I tried that but it is always booting the nand rootfs
<mmind00>
is it booting your own kernel or still one in the other partitions?
<dlezcano>
my own kernel
<mmind00>
dlezcano: hmm, what does the kernel commandline say, that the kernel outputs on boot?
<mmind00>
khilman: also I don't know if the coreboot variants behave differently thant the uboot ones, maybe dianders knows
<mmind00>
dlezcano: yep that is correct ... but rockchip boards have 2 or 3 different partitions ("kernel", "boot", "recovery") ... when I get a new board I always fumble some time with these until it magically boots
<dlezcano>
mmind00: do you have any pointer to fix that ?
<mmind00>
dlezcano: sorry, no real "tutorial"; when I got another rk3288 box, I just wrote crap to the boot ("di -b") and recovery ("di -r") parts, flash the regular "di -k" again, and it magically bootet from there from then on
<khilman>
tried different cable, tried plugging directly into host (instead of hub) and still no luck
<dlezcano>
khilman: is it plugged on the other side on an USB2 ?
<dlezcano>
mmind00: how did you write "crap" to boot and recovery ?
<khilman>
dlezcano: the host is usb2, yes
<dlezcano>
mmind00: I mean what kind of "crap"
<dlezcano>
khilman: does the board show "rockusb key pressed" ?
<khilman>
no, it just boots the coreboot then the kernel
<dlezcano>
khilman: s/board/console/
Valduare has quit [Quit: Valduare]
<dlezcano>
khilman: likely the cable connexion is not right
<khilman>
here's a bootlog when holding the vol/recovery button down then pressing reset (while continuing to hold recovery): http://hastebin.com/esadubiqif.coffee
<mmind00>
dlezcano: random bytes that kill the image checksum the bootloader looks for [like taking the first 1000byte from some file] ... I guess there are other people around here, that know the flashing process better then this though
<khilman>
mmind00: ah, so coreboot is probably not checking for the recovery button, but u-boot does?
<mmind00>
khilman: I'd guess so ...
Valduare has joined #linux-rockchip
<mmind00>
khilman: I guess this "depthcharge" coreboot payload should be the part that should handle recovery/flashin/whatever
<khilman>
mmind00: ok, hopefully dianders or another chromeos person can shed some light
Soopaman has quit [Ping timeout: 272 seconds]
Soopaman has joined #linux-rockchip
Valduare has quit [Quit: Valduare]
GriefNorth has quit [Ping timeout: 245 seconds]
Soopaman has quit [Ping timeout: 272 seconds]
Soopaman has joined #linux-rockchip
Valduare has joined #linux-rockchip
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
Astralix has joined #linux-rockchip
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
<amstan>
srao_: ping
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<srao_>
looks like dianders helped him figure it out
<khilman>
mmind00: dianders got me into recovery mode, and upgrade_tool now sees the board
<khilman>
how do I get u-boot on ther now?
<Astralix>
uboot compiled as loader or uboot compiled as loaded by the loader?
<khilman>
Astralix: I'm not sure. :/
<Astralix>
So you built the rockchip uboot sources pulled from somewhere?
<khilman>
this is my first rockchip board, so I'm a bit lost in all the rockchip specific boot stuff
<khilman>
I built uboot from the rkchrome/uboot repo
<khilman>
and it built RK3288Loader_uboot_V2.15.bin
<khilman>
so I"m guessing that's uboot as a loader
<Astralix>
however, try it as a loader so flash it with "ef uboot.bin"
<khilman>
Astralix: use u-boot.bin or the RK3288Loader_uboot_V2.15.bin
<Astralix>
the loader thing.
<khilman>
./upgrade_tool ef ~/tmp/RK3288Loader_uboot_V2.15.bin
<khilman>
Loading loader...
<khilman>
Creating Comm Object failed!
<Astralix>
The true name of your loader file
<Astralix>
Ah, you should use sudo or tweak some files in /etc/udev/rules.d
<khilman>
ah, ok. seems to be working now
<Astralix>
with rd you can reset the device
<khilman>
Loading loader...
<khilman>
Download Boot Start
<Astralix>
should be ready nowe
<khilman>
Astralix: after reboot, it's still booting coreboot, not uboot. hmm
<Astralix>
tried it myself, just to be sure
<Astralix>
after ef it stays with erased flash, so it is in mask-rom
<Astralix>
after that send
<Astralix>
ul [uboot.img]
<Astralix>
Then it boots whatever you flashed there
<khilman>
so upgrade_tool just sits there after "Download Boot Start". Do I wait for it to finish, or ctrl-c it?
<khilman>
sudo ./upgrade_tool ul ~/tmp/RK3288Loader_uboot_V2.15.bin
<khilman>
Loading loader...
<khilman>
Support Type:RK320A Loader ver:2.15 Loader Time:2014-10-03 13:09:59
<khilman>
Download Boot Start
<Astralix>
What upgrade_tool do you have?
<Astralix>
2.16 or 2.21?
<khilman>
Rockusb>V
<khilman>
Upgrade Tool ver 1.16
<Astralix>
this one is old. Could you get the 2.21 version please?
<khilman>
I assume you meant 1.21 above, not 2.21?
<khilman>
it still never returns after "Download Boot Start"
<khilman>
note: apparently my rk3288 EVB is a bit different from others in that its default is booting coreboot from SPI flash
<khilman>
and I'm assuming that upgrade_tool is writing NAND
RayFlower has quit [Read error: Connection reset by peer]
RayFlower has joined #linux-rockchip
<Astralix>
Sorry...
<Astralix>
Ah, I see
<Astralix>
the upgrade_tool doesn't probably know of SPI flash
<Astralix>
if you have an EVB, you should disable the SPI
<Astralix>
Where do you guy get all these rk3288 EVB from?
<khilman>
I got it from rockchips
<khilman>
IIUC, it is supposed to boot from NAND before SPI if there's valid content in the NAND
<Astralix>
No, the loader_tool installs your uboot to the NAND. So you need to get the board to boot by nand only.
<Astralix>
If your uboot fails, you can switch to spi again and back up, try something else
<Astralix>
If you disable boot and killed the nand, rk3xxx stays in mask.rom mode
<Astralix>
the upgrade_tool can upgrade NAND and loader with only mask-rom load active
<Astralix>
And it can only do it with mask-rom active
<Astralix>
For upgrading the loader, the upgrade_tool forces the SOC into mask-rom and waits until it is there, then flashes the loader to the boot section of the nand.
<khilman>
if upgrade_tool is detecting the soc, doesn't that mean its in mask-rom mode?
<Astralix>
If your SPI loader intercepts this, i.e. it always keeps the SOC in loader mode, the upgrade tool never sees mask-rom and therefore fails
<Astralix>
check your serial port, it sends a detection strin in mask-rom