On Saturday 10 February 2018 20:14:15 Erik Christiansen wrote:

> On 10.02.18 15:20, Gene Heskett wrote:
> > On Thursday 08 February 2018 14:24:25 Gene Heskett wrote:
> > > Greetings all;
> > >
> > > I have not managed to figure out how to install the realtime
> > > kernel I have built on the rockchip yet...  Its a u-boot system.
> >
> > Ping on the above? Anybody any smarter than I am?
>
> Few of us as energetic, or any smarter, Gene.

Thanks for the flowers, but lets face it, oldtimers has come to visit, so 
somebody might have to remember to water them. My green thumb has always 
had stains of black & blue on it... Energetic seems to have 2 hours a 
day limits too, what with the distractions of being a caregiver too.  So 
this is always the 2nd violin in this orchestral ensemble.

> I wuzn't going to say 
> anything, as it's over a decade since I played with u-boot, making my
> few pages of notes more historical than currently relevant. Syncing my
> understanding with your issue, U-boot could be used different ways,
> including:
>
>    1. as a standalone bootloader residing in flash & relocating itself
>    to RAM on startup automatically => "FLASH version"

Which I'm guessing is whats going on now. Thats a WAG, though, no way 
I'll put an S in front of it at this stage. ;-)

> When I did it, all that was needed was a kernel image and ramdisk at
> known flash addresses. But how is u-boot configured on the rockchip?

DamnedifIknow, Eric, but studying what is there will hopefully open the 
roadmap to the right page.  There are 5 "partitions" on the sdisk that 
do not appear to be "mounted"

df reports
rock64@rock64:~$ df
Filesystem     1K-blocks     Used Available Use% Mounted on
udev             2006536        0   2006536   0% /dev
tmpfs             401960    15648    386312   4% /run
/dev/mmcblk1p7   7926212  3187208   4389760  43% / <-this is on the sdisk
tmpfs            2009784        0   2009784   0% /dev/shm
tmpfs               5120        4      5116   1% /run/lock
tmpfs            2009784        0   2009784   0% /sys/fs/cgroup
/dev/mmcblk1p6    102182    44384     57798  44% /boot/efi <-so is this
/dev/sda3      951091568 13261272 889447696   2% /media/slash <-spinning 
rust, a terabyte on a usb3 port. FAST! 
it's also where the realtime kernel is built, v4-14-15-13rt, as is an 
arm64 version of linuxcnc built from an October 2017 git clone. I need 
to do another clone after saving the .config someplace safe. Actually, 
what I ought to do is blow it away, and reclone it every night, build 
it, and see if Seb is interested in useing my builds since he wants to 
spend a aged ton of money on a board and cpu combo that has zero 
relationship to a rock64. He'd at least have an arm64 build that others 
could pull and play with.
tmpfs             401956        0    401956   0% /run/user/1000

> The bumpf at:  http://rockchip.wikidot.com/linux-user-guide#toc20
> seems cryptic and of limited usefulness, but this seems to be the
> goods: http://opensource.rock-chips.com/wiki_Boot_option
>
That does indeed looks like the real goods. But my big printer hasn't 
been really taken for a walk recently so I won't get a good copy till 
the third try. After the first, it was obvious it needed some nozzle 
cleaning, took 2 passes to restore the black, and the one pass on the 
color seems to be trying to tell me the yellow is nearly done.

Then 2nd pass I put it back in portrait but left it in short edge binding 
duplex, which puts even/odd upside down, so now I've reset that to long 
edge binding which should make the 3rd pass good. But in doing test 
prints after the cleaning, I found there is one hell of a difference in 
the two sides of the glossy paper in tray 1, ink puddled all over, 
turned the whole stack of paper over for the next test page, just 
detectably glossier, and its perfect. Whodathunkit.

> There the whole boot sequence is laid out, with kernel and rootfs,
> which can be on various media, at your choice. (4 different boot
> flows) Does the method for one of those media look good to you? (The
> load offsets for kernel and rootfs seem to be 0x8000 and 0x40000
> respectively, if my quick scan is near the mark.) It would be fun to
> see what can be done with the rockchip.

Now I need to see about a roadmap thru all that, that works, including 
booting from the 16GiB eMMC thats been sitting there, plugged in but 
doing squat all this time. From a first glance, there be dragons on that 
path. :) But in my time I have killed a few of those, and although its 
getting harder to "walk on water" these days, hopefully I can do it 
again.

If I am successfull, we'll have a new $44 platform to run linuxcnc on. 
One that hopefully will not throw away keyboard and mouse events from 
its own keyboard/mouse like the pi does. The pi runs the machine quite 
well, once the gcode is written.

And my printer managed to scramble the output page order, inserting a 
couple pages in the middle of the stack, so I had to read the web page 
and put it back in order before I stapled it. Sleep beckons now as its 
23:53 local, so serious study will start at about the middle of cup #2 
tomorrow. Thanks muchly for the leg work and link, Erik. It is 
appreciated.

Take care.

-- 
Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to