00:09
<
gb_master >
so, I generated boot.img by taking the line from the google groups thread
00:09
<
gb_master >
mkbootimg --kernel /tmp/kernel.img --ramdisk /dev/null --cmdline "console=ttyFIQ0 console=tty0 root=/dev/mmcblk0p1 rootfstype=ext4 init=/sbin/init" -o /tmp/boot.img
00:10
<
gb_master >
but there's nothing about the partitions in the cmdline
00:11
<
naobsd >
gb_master: I don't know that kernel image uses cmdline from mkbootimg
00:12
<
gb_master >
well, I don't know
00:12
<
naobsd >
gb_master: but anyway what you need is partition specified by root=
00:13
<
naobsd >
gb_master: if you made 1st partition on SD and did mkfs.ext4, it should be ok
00:13
<
gb_master >
totally
00:16
<
naobsd >
--kernel /tmp/kernel.img <- what's this?
00:16
<
naobsd >
is there any reason not to use boot.img in that package?
00:17
<
gb_master >
I don't know
00:17
<
gb_master >
first post of 27th October
00:19
<
naobsd >
but I guess boot.img doesn't boot if "boot.img (kernel.img plus a null initrd via mkbootimg)" is true :(
00:19
<
naobsd >
mkbootimg uses (z)Image, not kernel.img
00:20
<
naobsd >
so I guess his system is booted with kernel.img
00:20
<
gb_master >
ok, so I can re-do the boot.img by using zImage
00:20
<
naobsd >
you have working zImage?
00:21
<
gb_master >
I compiled one, yes
00:21
<
gb_master >
in the same way he did
00:21
<
naobsd >
(I'm sorry, actually I didn't try his package)
00:21
<
gb_master >
don't worry
00:21
<
naobsd >
gb_master: I see, try it and see result
00:21
<
gb_master >
me neither... I'm just redoing his steps
00:22
<
naobsd >
please don't forget you're trying 2 new things, sd boot and mainline kernel. don't mix them
00:23
<
gb_master >
can't I use sd_miniroot and kernel+boot from mainline?
00:23
<
naobsd >
2 independent things
00:23
<
naobsd >
sd_miniroot?
00:24
<
gb_master >
sdboot_rk3188.img
00:24
<
gb_master >
(sorry)
00:24
<
naobsd >
it's just for "sd boot"
00:24
<
naobsd >
no relation to mainline(or non-) kernel
00:24
<
gb_master >
so I write sdboot+parameters+mainlinekernel+mainlineboot
00:25
<
naobsd >
try it, you can get result
00:26
<
naobsd >
e.g. if you can see kernel message and error, sd boot worked, problem is in kernel or rootfs
00:27
<
gb_master >
I wish I could... I don't have UART
00:27
<
gb_master >
I mean, I have the stuff and things, but not here
00:27
<
naobsd >
how do you access it?
00:28
<
gb_master >
that's why I go on with backups (for now)
00:28
<
naobsd >
but it's hard to see "result" :(
00:28
<
naobsd >
I think LED blinks if kernel is booted
00:28
<
naobsd >
^if you're using my modified dts
00:29
<
gb_master >
rk3188-radxarock-full.dtb
00:30
<
naobsd >
snapshot-20140928 include my change
00:31
<
gb_master >
I kept ramdisk "/dev/null"
00:31
<
gb_master >
and I recreated boot.img with zImage
00:32
<
naobsd >
ah I found "cat arch/arm/boot/zImage arch/arm/boot/dts/rk3188-radxarock-full.dtb > /tmp/kernel.img "
00:33
<
naobsd >
sorry, his kernel.img is fine
00:33
<
naobsd >
did you append dtb to zImage?
00:33
<
gb_master >
I'm going to
00:33
<
gb_master >
I create the boot.img off the kernel.img, instead of zImage?
00:33
<
naobsd >
and his procedure doesn't touch cmdline, it should be defined in kernel config
00:34
<
naobsd >
kernel.img sounds it has RK header, not good name :(
00:34
<
naobsd >
what you need is zImage+dtb file for mkbootimg input
00:34
<
naobsd >
he named it kernel.img
00:34
<
gb_master >
ok done
00:35
<
gb_master >
cmdline defined in the menuconfig?
00:36
<
naobsd >
you have to use boot.img. his "kernel.img" doesn't have RK header, it cannot boot if you write it to NAND/SD as is
00:36
<
naobsd >
somewhere in menuconfig == in rk30_rk31_defconfig/.config
00:36
<
gb_master >
console=ttyS2,115200 earlyprintk init=/sbin/init root=/dev/mmcblk0p1 rootwait
00:36
<
gb_master >
I think
00:37
<
naobsd >
it will be used, seems fine for you too
00:37
<
gb_master >
it's time for some dd
00:38
<
naobsd >
can you paste parameter (CMDLINE)?
00:38
<
gb_master >
I set both seek=1 skip=1 for sdboot, right?
00:38
<
gb_master >
well, in the kernel, as I said there was " console=ttyS2,115200 earlyprintk init=/sbin/init root=/dev/mmcblk0p1 rootwait" defined
00:38
<
naobsd >
what I asked is parameter
00:38
<
gb_master >
in parameters I wrote "console=ttyFIQ0,115200 console=tty0 root=/dev/mmcblk0p1 rw rootfstype=ext4 mac_addr=12:21:34:43:56:65 init=/sbin/init"
00:39
<
naobsd >
well, no, a line start with CMDLINE in parameter file
00:39
<
gb_master >
CMDLINE:console=ttyFIQ0,115200 console=tty0 root=/dev/mmcblk0p1 rw rootfstype=ext4 mac_addr=12:21:34:43:56:65 init=/sbin/init
00:39
<
naobsd >
you erased mtdpart?
00:40
<
naobsd >
partition info in parameter is required to read kernel etc by loader
00:41
<
naobsd >
(other part is ignored if you're using kernel config with fixed command line)
00:41
<
naobsd >
what I said is, try as is, then add/update some things
00:41
<
naobsd >
you're trying 2 new things
00:41
<
gb_master >
CMDLINE:console=console=ttyFIQ0,115200 console=tty0 root=/dev/mmcblk0p1 rw rootfstype=ext4 mac_addr=12:21:34:43:56:65 init=/sbin/init initrd=0x62000000,0x00800000 mtdparts=rk29xxnand:0x00002000@0x00002000(misc),0x00008000@0x00004000(kernel),0x00008000@0x0000c000(boot)
00:43
<
gb_master >
so, if I should not use my kernel.img, what do I use?
00:43
<
naobsd >
boot.img has kernel, it's enough
00:43
<
naobsd >
you may use ramdisk-only ramdisk, then you need kernel.img (<- with RK header)
00:44
<
naobsd >
^ramdisk-only boot.img
00:44
<
gb_master >
so I skip the dd kernel.img line
00:46
<
gb_master >
nope, doesn't boot
00:46
<
naobsd >
you need serial console ;)
00:46
<
naobsd >
how about LED?
00:46
<
gb_master >
I'll get that on Xmas
00:46
<
gb_master >
when I get back home
00:47
<
naobsd >
how about boot.img from google groups?
00:47
<
gb_master >
I can flash that
00:48
<
naobsd >
I'm not sure but LED may not be blinked if kernel panics very soon...
00:49
<
naobsd >
my snapshot is old
00:49
<
naobsd >
it worked but it should be updated :(
00:49
<
gb_master >
you have newer ones?
00:49
<
naobsd >
but you can try latest mainline at any time ;)
00:50
<
gb_master >
I'd like to have some new kernel before :)
00:50
<
gb_master >
but, after, I totally will
00:50
<
gb_master >
btw, what's the status? just the video stuff doesn't work?
00:51
<
naobsd >
not sure for now
00:51
<
gb_master >
and 3.10? is it OK?
00:51
<
naobsd >
personally 3.10 is not ok
00:52
<
gb_master >
just 3.0 is
00:52
<
naobsd >
good for android
00:52
<
gb_master >
less good for other stuff, maybe
00:52
<
naobsd >
RK kernel has a lot of bugs
00:53
<
gb_master >
any news about a new release from rockchip about kernel stuff?
00:53
<
naobsd >
new kernel? 3.10 or mainline for rk3188? no
00:53
<
gb_master >
it's all up to the community
00:53
<
naobsd >
3.0 is latest for rk3188, RK still update some
00:54
<
naobsd >
rockchip community is very weak
00:54
<
naobsd >
ah, there are many mainline developers
00:54
<
naobsd >
but not for 3.0/3.10/etc
00:55
<
gb_master >
I'm writing the boot.img from the group
00:55
<
naobsd >
RR is dev board, but most user is not developer
00:56
<
naobsd >
(I know some amount of people doing his own product/project with RR and get good result)
00:57
<
naobsd >
but not so much(very few) "community developer"
00:57
<
gb_master >
yes, I know
00:57
<
gb_master >
so, it's booting
00:57
<
gb_master >
which is GREAT
00:57
<
gb_master >
but I can't ssh
00:57
<
naobsd >
gb_master: weird, but you got working example
00:58
<
naobsd >
I think eth0 should work with my snapshot
00:58
<
gb_master >
when I type the password, I get "connection to 192.168.x.x closed"
00:58
<
naobsd >
it seems working ;)
00:58
<
gb_master >
but if I can't ssh
00:59
<
naobsd >
ethernet support has some bugs
00:59
<
naobsd >
but I think ssh worked
00:59
<
naobsd >
I think you can see logs on SD card :)
01:00
<
naobsd >
which rootfs are you using? /dev/* may not be populated properly for ssh
01:01
<
naobsd >
anyway, it shouldn't be difficult
01:01
<
gb_master >
mmm MAC address changes everytime
01:01
<
gb_master >
I put the line in parameter.img
01:02
<
naobsd >
because there is no MAC address is stored on RR board
01:02
<
naobsd >
ah, mainline kernel doesn't support it
01:02
<
gb_master >
yes, but why it's not using the one I wrote in the parameter.img?
01:02
<
naobsd >
I think it can be specified in dts
01:03
<
naobsd >
mainline kernel doesn't use the one in the parameter
01:03
<
gb_master >
ok ok, now I got it
01:03
<
gb_master >
well, I think I have to go to sleep now
01:04
<
gb_master >
but I did great things tonight thanks to you guys
01:04
<
naobsd >
gb_master: good night
01:04
<
gb_master >
thanks again
01:05
gb_master has quit [Remote host closed the connection]
01:15
mrcan has joined #radxa
01:15
mrcan has joined #radxa
01:15
mrcan has quit [Changing host]
01:17
mrcan_ has quit [Ping timeout: 256 seconds]
02:03
naobsd has quit [Quit: Page closed]
03:07
naobsd has joined #radxa
04:11
Astralix1 has joined #radxa
04:13
Astralix has quit [Ping timeout: 250 seconds]
05:13
Bludot has quit [Quit: Connection closed for inactivity]
05:41
naobsd has quit [Quit: Page closed]
07:14
AstralixNB has joined #radxa
08:20
naobsd has joined #radxa
08:36
hramrach_ has quit [Ping timeout: 250 seconds]
08:42
hramrach_ has joined #radxa
11:00
naobsd has quit [Quit: Page closed]
11:38
bengal has joined #radxa
13:28
Bludot has joined #radxa
14:17
mrcan has quit [Quit: pai]
16:03
Bludot has quit [Quit: Connection closed for inactivity]
16:10
AstralixNB has left #radxa [#radxa]
16:39
bengal has quit [Ping timeout: 265 seconds]
17:07
orangecms has joined #radxa
17:08
bengal has joined #radxa
17:25
Bludot has joined #radxa
18:43
mrcan has joined #radxa
19:43
Bludot has quit [Quit: Connection closed for inactivity]
19:54
gb_master has joined #radxa
20:24
gb_master has quit [Quit: Leaving]
20:28
mrcan has quit [Read error: Connection reset by peer]
20:41
akaizen has quit [Remote host closed the connection]
20:42
akaizen has joined #radxa
20:46
akaizen has quit [Ping timeout: 244 seconds]
22:04
Bludot has joined #radxa
22:07
orangecms has quit [Quit: Leaving]
22:28
bengal has quit [Quit: Leaving]
22:39
akaizen has joined #radxa
22:52
naobsd has joined #radxa
23:17
akaizen_ has joined #radxa
23:20
akaizen has quit [Ping timeout: 244 seconds]