Will DE0-Nano image work on a DE10-Nano board?

[mugginsac]

I just wondered if the machinekit image for a DE0-Nano-SOC board will run on a DE10-Nano?

Alan

[Charles Steinkuehler]

On 7/24/2017 1:18 PM, mugginsac wrote:
> I just wondered if the machinekit image for a DE0-Nano-SOC board will run on a
> DE10-Nano?

No.

You'll need to import the pinout and HPS (Qsys) files from one of the
DE10 example projects (or otherwise modify things for the DE10 vs.
DE0) and recompile.

Is anyone using or looking to use a DE10? I don't (yet) have a DE10,
but I can try to setup a project and get bitfiles created if there's
enough interest and someone is willing to test.

--
Charles Steinkuehler
cha...@steinkuehler.net

[Bas de Bruijn]

Funny, was just wondering about this myself today. I'd be willing to purchase a board and test if you start on such a thing.

[mugginsac]

 
Bas and Charles,

I think I will go ahead and order the DE10-Nano instead of the DE0-Nano also.

Alan

[Charles Steinkuehler]

On 7/26/2017 10:31 AM, mugginsac wrote:
>
> Bas and Charles,
>
> I think I will go ahead and order the DE10-Nano instead of the DE0-Nano also.

I'll try to carve out some time to get an FPGA project setup for the
DE10-nano, unless someone else feels up to the task.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

I placed the order for a DE10-Nano.
Alan

[mugginsac]

Charles,

OK, I got my DE10-nano. So let me ask another question, will the DE0-Nano-Soc MachineKit image boot on the DE10-Nano board (as a starting point for doing the implementation)? Or do I need to start with the DE10-Nano image from Terasic?

Alan

[Charles Steinkuehler]

The DE0-Nano image will /probably/ boot on the DE10, but the FPGA will
not be programmed. If the DE10 doesn't boot, the issue is almost
certainly the kernel and/or device tree. If you use the kernel and
device tree from the DE10 image on a DE0 image you ought to be able to
boot with the DE0 uSD.

Basically, it looks like the DE10 is the DE0 with a larger FPGA and an
HDMI output...

...and all of the Cyclone-V SoC+FPGA parts share the same ARM hard
processor side, so the kernel/device-tree will be mostly identical for
things like Ethernet, uSD, flash, SDRAM, etc. as long as the boards
are not too dramatically different.

> --
> website: http://www.machinekit.io blog: http://blog.machinekit.io github:
> https://github.com/machinekit
> ---
> You received this message because you are subscribed to the Google Groups
> "Machinekit" group.
> To unsubscribe from this group and stop receiving emails from it, send an email
> to machinekit+...@googlegroups.com
> <mailto:machinekit+...@googlegroups.com>.
> Visit this group at https://groups.google.com/group/machinekit.
> For more options, visit https://groups.google.com/d/optout.
>


--
Charles Steinkuehler
cha...@steinkuehler.net

[Condit Alan]

Charles,

The DE10 won’t boot from the DE0-Nano-SOC MachineKit image. I need to get an adapter for USB female to Micro male (or USB OTG) so I can plug in my powered USB hub or at the very least a USB keyboard with built in touch pad.  

I guess I will try to set up so I can compile on the DE10-nano. The only linux box I have in the house is an i386 and the current cross compiler only runs 64bit mode.

I may be able to use the 386 box to run linuxcnc in the shop and bring the D525MW inside to set up a cross compile environment but the D525MW only has a 32gb SSD at the moment.

Alan

You received this message because you are subscribed to a topic in the Google Groups "Machinekit" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/machinekit/fleGyyOwDgE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to machinekit+...@googlegroups.com.

[mugginsac]

OK, I got a new 1TB hard disk for my D525MW board, installed Debian Stretch and the associated development software. I have ordered three of the DB25 boards that Charles designed.

I tried to follow the instructions from Robert Nelson's socfpga-kernel-dev page on eewiki. I got down to the stuff on u-boot and blew up on the "make ARCH=arm CROSS_COMPILE=${CC} socfpga_de0_nano_soc_defconfig" command.

So I have to figure out how to build a RT kernel for the DE10-nano. So I guess the first step is to figure out what the differences are between the kernel for DE0-Nano-SOC and the kernel for DE10-Nano. Also what do I need to download to my machine to start the process.

Alan

[Charles Steinkuehler]

On 8/4/2017 6:51 PM, mugginsac wrote:
>
> So I have to figure out how to build a RT kernel for the DE10-nano. So I guess
> the first step is to figure out what the differences are between the kernel for
> DE0-Nano-SOC and the kernel for DE10-Nano. Also what do I need to download to my
> machine to start the process.

Unlike x86, there are no "strange" filesystem requirements for ARM
booting (that's all handled by U-Boot), so you can just rsync any
appropriate root filesystem onto your uSD (or eMMC, or whatever).

If you don't specifically need to recompile U-Boot or the kernel, you
can simply use these as-is from any example image for the DE10 and
just replace the rootfs. This may be helpful to get started, even if
the kernel isn't patched with PREEMPT_RT (you can update the kernel
later once you have a working system).

If you want to build from source, there are lots of differing
instructions which makes things confusing. There are Angstrom bitbake
recipes, Altera custom Makefiles, and various other instructions
floating around. I usually build a kernel following the "Building
Kernel & U-Boot Separately From Git Trees", which is page 4 of the
"Compiling Linux" documentation for the GSRD. One example (make sure
to click over to #4):

https://rocketboards.org/foswiki/Documentation/GSRD160CompilingLinuxArrowSoCKitEdition

--
Charles Steinkuehler
cha...@steinkuehler.net

[Charles Steinkuehler]

Since there seems to be interest, I've gone ahead and ordered a
DE10-Nano so I can do real-world tests.

I don't have lots of free time right now, but I ought to at least be
able to help get a bootable uSD card image generated (if no one beats
me to it) and migrate the DE0 project for the hostmot2 VHDL code to
the DE10.

--
Charles Steinkuehler
cha...@steinkuehler.net

[Michael Brown]

I would like to insert some notes from my own experiments with the DE10 Nano:

I you use the terasic supplied reference project as base you will need to customize the device tree --> dtb

as the qsys system uses different addresses / bridges for the fpga ip cores.

they can boot with identical u-boot and root file system. (without hdmi, piece of cake...)

If you want to use the hdmi as local xserver xorg based display things get a bit more intrigant:

To be able to use the alt_vip fb core + alt_vipfb.c driver (which keeps pooping in and out of the kernel, and also has been obsoleted after quartus v.16.1).

You have to program the fpga from uboot with at least the alt_vip framebuffer qsys core and also have it in the devicetree so the the kernel is ale to pick it up and initialize it when it boot. (also remember to make sure the bridges are offhanded by uboot before starting the kernel, if you keep getting a blank screen after boot and /dev/fb0 is there).

THe intrigant part is then that when machinekit reloads the fpga chances are the framebuffer may go haywire, so the most elegant solution would be to create a partitioned 2 rev design with one containing only the qsys system + a rev 2 containing also the mesa hdl hostmot2 stuff behind a bridge.

This can the be loaded via a partial reconfiguration .dtbo much like the ones currently used except that is does not reconfigure the whole fpga but only adds / replaces the Mk Hostmot2 mesa cores.

---

I hope this is clear and understandable....

---

I am atm tied up in finishing adding commits for a cramps based atlas soc version with functional adc and some more to MK, before I can turn more focus to DE10.NANO, for now I can only point towards my own socfpga build-scripts on the-snowwhite@github which also contains all the elements needed for generating a stretch image with hdmi and 4.9-rt kernet framebuffer rootfs, u-boot etc initially made to create the first working mksocfpga demo. 

Best Wishes Michael

[Michael Brown]

BTW

It may be possible to boot configuring the fpga from u-boot with framebuffer and hm2 cores and then start machinekit hm2_soc_ol driver without firmware=xxx.dtbo for easy DE10-Nano fb startup. (uio driver then needs a uio entty in dts also to load).

[Charles Steinkuehler]

Yes, it would be much easier to load a single FPGA image via U-Boot
rather than deal with partial reconfiguration.

I was planning on just ignoring the HDMI to start. :)

[mugginsac]

Charles,

I understand how ignoring HDMI to start may be easier, however, the reason that I bought the DE10 instead of the DE0 is the HDMI. :-(

Alan

[mugginsac]

I got a USB OTG cable from Digikey and I am now able to boot the DE10-Nano with an HDMI monitor, keyboard and mouse.

[Michael Brown]

Great

Then in the HW divison all you need to do is add the uio-generic / hm2-uio ip core in qsys at address 0x0004_0000.

And then the hm2 qip (include) file wire things up and add the config stuff.

If you are working in verilog / systemverilog I made an intermediary upgrade of the original DE0_NANO_GHRD project adding the new firmware id stuff and made docker compilable before morphing it into an experimental CRAMPS I/O based version:

I suggest using that commit as template for adding your new (renamed) DE10-NANO project to MK mksocfpga @ github:

https://github.com/the-snowwhite/mksocfpga/tree/fca882b8bc25ff2e2e81b4a0979ee2c6bf7c041e/HW/QuartusProjects/DE0_Nano_SoC_Cramps

[mugginsac]

I seem to have hit a roadblock, maybe it is just my blockhead. The linux system on the DE10-nano is Angstrom. It doesn't have apt-get, sudo or much of anything else. I was going to try and copy the appropriate files from the DE10-Nano Angstrom system to the DE0-Nano-SOC Debian system but I can't seem to find the corresponding folders or files on the two systems.

[Charles Steinkuehler]

The main pieces that are system specific are the boot loader, the
kernel, and the device tree. If you start with a working DE10 uSD
image you can leave the boot-loader (partition type = A2) and the FAT
partition (kernel and device tree) alone and use rsync (or whatever)
to copy a Debian rootfs to the uSD, overwriting Angstrom.

That's probably the easiest way to get Machinekit working with the
DE10 until we get some official images created.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

Charles,

That worked. I now have the DE10-Nano booting into Debian Stretch console.
I just formatted the Angstrom partitition, expanded the partition to fill the uSD card, made an ext4 partition out of it, and then followed Robert Nelson's instructions to expand the minfs tar into the directory.

Thanks,
Alan
 

[Michael Brown]

Interresting I have not heard that armhf Stretch compatibility is fully up and working yet so I have been sticking to building jessie images via my recently updated build script(s).

https://github.com/the-snowwhite/soc-image-buildscripts

please share when you get MK armhf and Debian Stretch to play along nicely....
----
I can not atm spin up and share an on line test image as I now are busy moving out of the city... :-)

BEst Wishes Michael Brown

[mugginsac]

Now I am trying to created a bootable image following Robert Nelson's instructions. (Whoa is me!) Where does the .rbf file go? In the rootfs or in the A2 partition?

The u-boot-with-spl.sfp goes in the A2 partition, does it pick up the kernel from the rootfs partition?

Alan

On Monday, July 24, 2017 at 11:18:00 AM UTC-7, mugginsac wrote:

[mugginsac]

Robert Nelson's instructions don't show creating a FAT partition, just an A2  partition and an EXT4 partition. If I am reading his instructions correctly he is writing u-boot-with-spl.sfp in the A2 partition then creating ext4 file system in the 2nd partition. Then he makes creates the rootfs in that ext4 partition. The "boot" directory is in the "rootfs" directory with a dtbs directory,  an extlinux directory and the kernel image in the boot directory. I am not seeing any boot activity so I think I must be doing something wrong.

I know the board still works because I can still boot the DE10 image where I replaced the angstrom filesystem with Debian Stretch.

[Charles Steinkuehler]

The A2 partition is required (by the SoC boot ROM) in order to load
the boot loader. Any other requirements depend on your boot loader
(probably U-Boot), the options you have compiled in, and any settings
or environment you have configured. The "reference" U-Boot from
RocketBoards looks for a FAT boot partition, but that certainly isn't
mandatory...U-Boot can read the kernel and device tree from any
supported device or filesystem (including things like nfs and tftp).

If your system is not booting, it is *VERY* helpful to have a serial
port connected. The DE0 and DE10 have an FTDI serial to USB converter
built-in, so you don't even need a special cable. Press a key to stop
U-Boot from automatically booting and crawl through the configuration.
Or if you're not seeing U-Boot messages, you probably have your A2
partition messed up (ie: U-Boot was not built and/or installed properly).

[mugginsac]

Charles,

I installed putty on my D525MW linux box, and I plugged a USB A to mini cable from the USB port on the desktop to the UartToUSB connector on the DE10. However, I don't know what pseudo serial port to watch for the Nano? Can you tell me how to figure out what /dev/tty???? device to use for putty’s serial connection to the DE10-Nano? I don't have a windows machine.

Thanks in advance,
Alan
 

[mugginsac]

By the way it with with the USB cable connected to the D525MW and running lsusb it shows up on bus 2 as device 32 and is correctly identified as Altera (09fb:6810).
But I don't know how to get a /dev/ttyUSB? or some other serial device handle.

Alan

[schoo...@btinternet.com]

On 18/08/17 05:06, mugginsac wrote:

By the way it with with the USB cable connected to the D525MW and running lsusb it shows up on bus 2 as device 32 and is correctly identified as Altera (09fb:6810).
But I don't know how to get a /dev/ttyUSB? or some other serial device handle.

cat /sys/class/tty/ttyUSB*
will usually bring up the ports that exist (ie are in use with a device plugged into them)

The DE0-Nano usually just came up as /dev/ttyUSB0 or ttyUSB1 AFAIR

Certainly no exotic ports like some of the Arduinos

regards

Alan

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.

To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.

[mugginsac]

OK, I can now get a /dev/ttyUSB0. I had to add altera vendor id 0x09fb and DE10 product id 0x6810 to ftdi_sio. I have to plug in the USB cable from the DE10 to get ftdi_sio to load. Then by the time I get it plugged in and putty loaded there is nothing coming over the console. If I try to start putty before plugging in the DE10 I get no /dev/ttyUSB0 found.

Alan

[Charles Steinkuehler]

If the DE10 is like the DE0, you should be able to leave the USB cable
plugged in, which will keep the FTDI chip powered so you have a USB
serial connection even without 5V power applied to the DE10. Then
apply power to the DE10 and you can see the boot messages.

Otherwise if it's like the early SoC boards (like the SoCKIT), the
FTDI chip "disappears" if you don't have system power applied so to
see all the boot messages you have to hold the HPS in reset while
applying power, get the USB port to show up and connect it to a
terminal program (using the one hand that isn't pressing the reset
button), then release the HPS reset. Lots of fun (not!). :)

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

Well it appears that the not so fun situation is the case. When I unplug power from the DE10-Nano the ftdi chip disappears. I also tried plugging DE10 cable into a powered hub instead of directly into the desktop box but same result.

How about if I plug in the DE10, get the ttyUSB0 and then press and release the HPS reset?

Alan

[Charles Steinkuehler]

That works, but at the risk of corrupting any mounted filesystems if
Linux has already booted. If you can't get Linux booted, resetting
the HPS to watch the serial boot messages is fine, there's no
potential to corrupt filesystems if they aren't mounted! :)

--
Charles Steinkuehler
cha...@steinkuehler.net

[Charles Steinkuehler]

I just looked over the schematic for the DE10-Nano and pulled my DE10
out of the box for testing. If I connect just a USB cable to the
board, a new USB serial port shows up on my workstation. If I connect
to this port (115200 baud, 8n1) and apply power to the board, I get
the expected boot messages (see below) when using the uSD card
supplied with the board.

Are you perhaps connecting to the USB-Blaster II port (J13, next to
the HDMI connector) and not the USB-UART port (J4, next to the
Ethernet port)?

<boot messages>
U-Boot SPL 2017.03-rc2 (Mar 30 2017 - 19:07:16)
/data/de10-nano/release-build-2017.03.31/build/tmp-angstrom-glibc/work/de10_nano-angstrom-linux-gnueabi/u-boot-socfpga/v2017.03+gitAUTOINC+d03450606b-r0/git/drivers/ddr/altera/sequencer.c:
Preparing to start memory calibration
/data/de10-nano/release-build-2017.03.31/build/tmp-angstrom-glibc/work/de10_nano-angstrom-linux-gnueabi/u-boot-socfpga/v2017.03+gitAUTOINC+d03450606b-r0/git/drivers/ddr/altera/sequencer.c:
CALIBRATION PASSED
/data/de10-nano/release-build-2017.03.31/build/tmp-angstrom-glibc/work/de10_nano-angstrom-linux-gnueabi/u-boot-socfpga/v2017.03+gitAUTOINC+d03450606b-r0/git/drivers/ddr/altera/sequencer.c:
Calibration complete
Trying to boot from MMC1


U-Boot 2017.03-rc2 (Mar 30 2017 - 19:07:16 -0700)

CPU: Altera SoCFPGA Platform
FPGA: Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT: SD/MMC Internal Transceiver (3.0V)
Watchdog enabled
I2C: ready
DRAM: 1 GiB
MMC: dwmmc0@ff704000: 0
*** Warning - bad CRC, using default environment

In: serial
Out: serial
Err: serial
Model: Terasic DE10-Nano
Net:
Error: ethernet@ff702000 address not set.
No ethernet found.
Hit any key to stop autoboot: 0
=>

</boot messages>

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

OK, my face is red. Yes I was plugging into the wrong port. It is running u-boot-with-spl but it is looking for an ext2 partition instead of an ext4. So after a while it starts looking for a USB device.

[Charles Steinkuehler]

With the three different USB ports, it's easy to do. I've done this
myself, which is one reason I know what the symptoms are! ;-)

Now that you can see the boot messages, hopefully you will be able to
get the board to boot properly. If not, post details about the next
problem you run into.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

I rebuilt the uSD with an ext2 filesystem.

At first I got a bunch of messages related to the initialization. Then a bunch of characters seem to scroll down the right side of the "terminal" screen in a single column.

Then after awhile, I got a login prompt in minicom, but I couldn't get it to accept enough characters to be able to login.

[mugginsac]

Taking a break from watching the progress of the solar eclipse.

What I am trying to do is get to the point of building a system that will run with HDMI before I start trying to add MachineKit and the hm2 stuff.
At the moment the build doesn't seem to run HDMI. I thought I was using the same dtbs and rbf that the original DE10-Nano system used.  However, since it doesn't seem to boot with my monitor, keyboard and mouse like the original system maybe I didn't get the correct files.

[mugginsac]

Here is what I am getting from trying to boot the uSD I built. Is it possible that I have a bad uSD?

U-Boot SPL 2017.05 (Aug 20 2017 - 08:57:37)

drivers/ddr/altera/sequencer.c: Preparing to start memory calibration

drivers/ddr/altera/sequencer.c: CALIBRATION PASSED

drivers/ddr/altera/sequencer.c: Calibration complete
Trying to boot from MMC1

U-Boot 2017.05 (Aug 20 2017 - 08:57:37 -0700)

CPU:   Altera SoCFPGA Platform
FPGA:  Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT:  SD/MMC Internal Transceiver (3.0V)
       Watchdog enabled
I2C:   ready
DRAM:  1 GiB
MMC:   dwmmc0@ff704000: 0
*** Warning - bad CRC, using default environment

In:    serial
Out:   serial
Err:   serial
Model: Terasic DE10-Nano
Net:  
Error: ethernet@ff702000 address not set.
No ethernet found.
Hit any key to stop autoboot:  0

switch to partitions #0, OK
mmc0 is current device
Unknown command 'part' - try 'help'
Failed to mount ext2 filesystem...
** Unrecognized filesystem type **
a

[Charles Steinkuehler]

On 8/21/2017 4:56 PM, mugginsac wrote:
> Here is what I am getting from trying to boot the uSD I built. Is it
> possible that I have a bad uSD?

Your uSD is probably fine.

There's a CRC error because you don't (yet) have a saved U-Boot
environment on the A2 partition. Based on the output you are getting,
I suspect your uSD setup does not match the default environment
compiled into U-Boot. I recommend dropping into the U-Boot shell
(press a command when prompted to stop the autoboot) and poke around
to see what U-Boot is expecting.

You can probably fix this by either updating your uSD image to match
what U-Boot expects (filesystems, partitions, kernel/dts filenames,
etc) or by modifying the U-Boot environment to match your uSD and
storing the updated environment on the A2 partition.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

Charles,

What terminal software are you using?
Is there any special setup required other than 115,200 baud and 8N1 (data,parity, and stop bits)?
I can't get putty to work with the DE10 and I sometimes get those weird single column scroll down the right hand side with minicom.

Alan

[Charles Steinkuehler]

I'm using Putty on my Windows 7 box:

115,200 baud
8n1
no parity
no flow control

You might play with the terminal settings. I usually use VT100 or
Linux, but I typically don't see any issues with either until I try
using full-screen programs (like vi). Most anything should be OK for
interacting with the U-Boot terminal prompt.

--
Charles Steinkuehler
cha...@steinkuehler.net

[Robert Nelson]

add:

CONFIG_CMD_PART=y

this was missed in v2017.05-ish time frame...

diff --git a/configs/socfpga_de10_nano_defconfig
b/configs/socfpga_de10_nano_defconfig
index 26addc3c4d..fc2dcc3951 100644
--- a/configs/socfpga_de10_nano_defconfig
+++ b/configs/socfpga_de10_nano_defconfig
@@ -25,6 +25,7 @@ CONFIG_CMD_DFU=y
CONFIG_CMD_GPIO=y
CONFIG_CMD_I2C=y
CONFIG_CMD_MMC=y
+CONFIG_CMD_PART=y
CONFIG_CMD_SPI=y
CONFIG_CMD_USB=y
CONFIG_CMD_USB_MASS_STORAGE=y

Regards,

--
Robert Nelson
https://rcn-ee.com/

[mugginsac]

Robert,

Thanks. At least it loaded right up as shown on the serial terminal. Now it is at least looking for the ext4 filesystem before loading the ext2 filesystem.
Something that I read suggested that the journal-ing stuff on ext4 could shorten the life of the uSD card but that the ext4 has higher performance. Should I go ahead and change it back to an ext4 filesystem?

As far as I can tell it isn't activating the HDMI. When I hook up to the monitor and keyboard (like with the other uSD) I don't see any action.
Also I don't see any action on the leds like the original uSD system. Is it likely that it is failing to load the .rbf file?

I don't even know where it expects to find it. But I don't see any message to indicate that it is looking for it and failing to find it.

Thanks again,
Alan

[mugginsac]

I still have problems with the serial terminal.

It seems to me like U-boot is starting up and receiving input that is putting it into command mode. Then it is like the process that outputs the text from the DE10 is being fed into the  console input and confusing the heck out of the command processor. It keeps reporting command not understood or not recognized and then advising to try help. This seems to continue until minicom, putty or screen are overwhelmed and abort or terminate. It is almost like the command processor is still running but u-boot is restarting and outputting the startup messages again.

Does any of this make any sense?

Occasionally, it will hit a state where the boot process will complete to where it will read extlinux.conf and load the kernel.

[Robert Nelson]

+CONFIG_AUTOBOOT_KEYED=y
+CONFIG_AUTOBOOT_PROMPT="Press SPACE to abort autoboot in %d seconds\n"
+CONFIG_AUTOBOOT_DELAY_STR="d"
+CONFIG_AUTOBOOT_STOP_STR=" "

Then the only way to enter command mode, is the "space" key...

[mugginsac]

OK that made some progress. It still looks like some command processor is running at the same time as the login process.

It printed "U-Boot SPL 2017.05-dirty (Aug 23 2017 - 15:20:51)
drivers/ddr/alt" on the serial console and then went quiet. Next thing I saw was that it had apparently loaded the kernel and was waiting for a user name.

However, some process was feeding characters that were being consumed by the login process. Every so often it would hit something that corresponded to something the command processor recognized and display a bunch of stuff on the screen again, like a help menu. Look at the file I captured and see if it helps understand what is going on. I don't know what process is putting up the various help menus.

Thanks,
Alan

[Charles Steinkuehler]

It almost sounds like you've got some sort of serial echo going on,
but typically the echo setting is opposite of what you're seeing (ie:
if echo is on, the putty/minicom Tx characters are echoed back to the
putty/minicom screen, you're seeing what might be DE10 Tx data echoed
back to the DE10 Rx).

My DE10 certainly doesn't behave the way you're describing. I wonder
if there could possibly be a short between the UART Tx and Rx lines on
the DE10?

Do you have a 'scope to check the physical Rx/Tx lines? If not, you
may be able to use the RXD/TXD LEDs (next to the UART USB connector)
to diagnose. You should only see the RXD light if you actually type
something into the terminal window. If you see RXD along with TXD
without typing anything, that's your problem...then you just need to
figure out why it's happening.

Note that the LEDs are driven by the FTDI USB chip and don't directly
indicate the presence of traffic on the physical UART Rx/Tx lines to
the SoC. So just because you don't see the RXD LED turn on doesn't
mean that the Rx/Tx lines couldn't be shorted. You'd need to probe
the HPS_UART_RX line (pin 30 on the U12 FTDI chip) with a 'scope to be
sure what's going on.

What happens if you try to boot with the uSD image that shipped with
the DE10 while you have the USB UART connected?

--
Charles Steinkuehler
cha...@steinkuehler.net

[Robert Nelson]

"Something" else is echoing data into your serial device node from
your development pc.

What program are you using to access the serial terminal?

[mugginsac]

When I boot the disk that has the original boot loader. I get the following:

 UART clock 100000 KHz

CLOCK: MMC clock 50000 KHz

CLOCK: QSPI clock 3125 KHz

RESET: COLD

SDRAM: Initializing MMR registers
SDRAM: Calibrating PHY

SEQ.C: Preparing to start memory calibration



U-Boot SPL 2013.01.01 (Nov 30 2016 - 14:20:53)

BOARD : Altera SOCFPGA Cyclone V Board

CLOCK: EOSC1 clock 25000 KHz

CLOCK: EOSC2 clock 25000 KHz

CLOCK: F2S_SDR_REF clock 0 KHz

CLOCK: F2S_PER_REF clock 0 KHz

CLOCK: MPU clock 800 MHz

CLOCK: DDR clock 400 MHz

CLOCK:?O?WR⑪?

------------------------------------
It stopped with some of unicode characters or control characters. I assume that when I stop seeing output on the serial console that I am at boot prompt because that is about the amount of time it normally takes to reach the u-boot window to choose normal kernel or safe mode.

By the way when I type characters I just see the RX LED. When the DE10 starts from either uSD I just see the TX led while it is processing the initial messages. On the one I am trying to build, when it goes crazy, I see both the RX and TX either flashing at the same time or alternating (at 115200 it is really too fast to tell for sure).

[Charles Steinkuehler]

On 8/24/2017 6:23 PM, mugginsac wrote:
>
> By the way when I type characters I just see the RX LED. When the
> DE10 starts from either uSD I just see the TX led while it is
> processing the initial messages.

That is as expected.

> On the one I am trying to build, when it goes crazy, I see both the
> RX and TX either flashing at the same time or alternating (at
> 115200 it is really too fast to tell for sure).

I think if the Rx LED is flashing, your terminal program is sending
characters. If there was some sort of short between the Rx/Tx lines,
at the serial UART level (ie: the SoC pins), you wouldn't see the Rx
light from the FTDI chip.

...so it sounds like you have something messed up in your custom uSD
setup. I would recommend trying to replace one bit at a time (Root
OS, Linux Kernel, U-Boot boot-loader) from a known good working uSD image.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

Is there a way to get the DE10 to use uenv.txt for boot environment info with the new u-boot?

[mugginsac]

I understand CONFIG_DEFAULT_FDT_FILE but what does CONFIG_DEFAULT_DEVICE_TREE indicate?
CONFIG_DEFAULT_DEVICE_TREE="socfpga_cyclone5_de10_nano"
CONFIG_DEFAULT_FDT_FILE="socfpga_cyclone5_de10_nano.dtb"

The DE10_NANO_SoC_GHRD (Golden Hardware Reference Design) from the DE10-Nano CD uses soc_system.dtb for the device tree binary.

So if I change CONFIG_DEFAULT_FDT_FILE to reference "soc_system.dtb", do I also need to change CONFIG_DEFAULT_DEVICE_TREE?

Where does the system expect to find the .rbf file (like "soc_system.rbf")?

[mugginsac]

I am in the process of downgrading from Stretch to Jessie. I couldn't get Quartus to run on Stretch. It needs libpng12.0 and the notes I found about it and Stretch said that Stretch won't allow you to load libpng12.0 that you have to go to libpng14 or libpng16. Libpng16 is preinstalled on Stretch but Quartus didn't like it.

[Michael Brown]

What Quartus version ?

I have 15.1,16.1 and 17.0 running in Stretch only problem I have had is that it crashes when setting the license variable on first run.

My strategy is to use the native (mostly qt) lib versions instead of the ones in the quartus/linux64  folder renaming  like this:

mib@debian9-ws:~/intelFPGA/15.1/quartus/linux64$ ls -l *bak
-r-xr-xr-x 1 mib mib  3239776 Feb  4  2016 libQtCore.so.4-bak
-r-xr-xr-x 1 mib mib  3239776 Feb  4  2016 libQtCore.so-bak
-r-xr-xr-x 1 mib mib  4265448 Feb  4  2016 libQtDeclarative.so.4-bak
-r-xr-xr-x 1 mib mib  4265448 Feb  4  2016 libQtDeclarative.so-bak
-r-xr-xr-x 1 mib mib 12061688 Feb  4  2016 libQtGui.so.4-bak
-r-xr-xr-x 1 mib mib 12061688 Feb  4  2016 libQtGui.so-bak
-r-xr-xr-x 1 mib mib  1466320 Feb  4  2016 libQtNetwork.so.4-bak
-r-xr-xr-x 1 mib mib  1466320 Feb  4  2016 libQtNetwork.so-bak
-r-xr-xr-x 1 mib mib  2796040 Feb  4  2016 libQtScript.so.4-bak
-r-xr-xr-x 1 mib mib  2796040 Feb  4  2016 libQtScript.so-bak
-r-xr-xr-x 1 mib mib   274992 Feb  4  2016 libQtSql.so.4-bak
-r-xr-xr-x 1 mib mib   274992 Feb  4  2016 libQtSql.so-bak
-r-xr-xr-x 1 mib mib   387216 Feb  4  2016 libQtSvg.so.4-bak
-r-xr-xr-x 1 mib mib   387216 Feb  4  2016 libQtSvg.so.bak
-r-xr-xr-x 1 mib mib   160384 Feb  4  2016 libQtTest.so.4-bak
-r-xr-xr-x 1 mib mib   160384 Feb  4  2016 libQtTest.so-bak
-r-xr-xr-x 1 mib mib 26492952 Feb  4  2016 libQtWebKit.so.4-bak
-r-xr-xr-x 1 mib mib 26492952 Feb  4  2016 libQtWebKit.so-bak
-r-xr-xr-x 1 mib mib  5293336 Feb  4  2016 libQtXmlPatterns.so.4-bak
-r-xr-xr-x 1 mib mib  5293336 Feb  4  2016 libQtXmlPatterns.so-bak
-r-xr-xr-x 1 mib mib   309360 Feb  4  2016 libQtXml.so.4-bak
-r-xr-xr-x 1 mib mib   309360 Feb  4  2016 libQtXml.so-bak
-rwxr-xr-x 1 mib mib  1677716 Feb 22  2016 libsys_cpt.so-bak


mib@debian9-ws:~/intelFPGA/17.0/quartus/linux64$ ls -l *bak
-r-xr-xr-x 1 mib mib  3239776 Feb 12  2015 libQtCore.so.4-bak
-r-xr-xr-x 1 mib mib  3239776 Feb 12  2015 libQtCore.so-bak
-r-xr-xr-x 1 mib mib   570312 Feb 12  2015 libQtDBus.so.4-bak
-r-xr-xr-x 1 mib mib   570312 Feb 12  2015 libQtDBus.so-bak
-r-xr-xr-x 1 mib mib  4265448 Feb 12  2015 libQtDeclarative.so.4-bak
-r-xr-xr-x 1 mib mib  4265448 Feb 12  2015 libQtDeclarative.so-bak
-r-xr-xr-x 1 mib mib 12061688 Feb 12  2015 libQtGui.so.4-bak
-r-xr-xr-x 1 mib mib 12061688 Feb 12  2015 libQtGui.so-bak
-r-xr-xr-x 1 mib mib  1466320 Feb 12  2015 libQtNetwork.so.4-bak
-r-xr-xr-x 1 mib mib  1466320 Feb 12  2015 libQtNetwork.so-bak
-r-xr-xr-x 1 mib mib  2796040 Feb 12  2015 libQtScript.so.4-bak
-r-xr-xr-x 1 mib mib  2796040 Feb 12  2015 libQtScript.so-bak
-r-xr-xr-x 1 mib mib   274992 Feb 12  2015 libQtSql.so.4-bak
-r-xr-xr-x 1 mib mib   274992 Feb 12  2015 libQtSql.so-bak
-r-xr-xr-x 1 mib mib   387216 Feb 12  2015 libQtSvg.so.4-bak
-r-xr-xr-x 1 mib mib   387216 Feb 12  2015 libQtSvg.so-bak
-r-xr-xr-x 1 mib mib   160384 Feb 12  2015 libQtTest.so.4-bak
-r-xr-xr-x 1 mib mib   160384 Feb 12  2015 libQtTest.so-bak
-r-xr-xr-x 1 mib mib 26492952 Feb 12  2015 libQtWebKit.so.4-bak
-r-xr-xr-x 1 mib mib 26492952 Feb 12  2015 libQtWebKit.so-bak
-r-xr-xr-x 1 mib mib  5293336 Feb 12  2015 libQtXmlPatterns.so.4-bak
-r-xr-xr-x 1 mib mib  5293336 Feb 12  2015 libQtXmlPatterns.so-bak
-r-xr-xr-x 1 mib mib   309360 Feb 12  2015 libQtXml.so.4-bak
-r-xr-xr-x 1 mib mib   309360 Feb 12  2015 libQtXml.so-bak
-rwxr-xr-x 1 mib mib  6439682 Apr 26 21:08 libstdc++.so.6-bak

[mugginsac]

Thanks Michael for the heads up on the libraries. I also had to install a bunch of QT packages, but,
I got Quartus 17.1 installed and working.

[mugginsac]

I am trying to follow Robert Nelson's instructions on the eewiki to build a stretch installation for the DE10-Nano with a DE10-Nano-DB25 and 1206 size parts.

I got u-boot built from the github repository but it still won't boot. My guess is that the FPGA is not properly initialized. So my question is this:

Does the soc_system.rbf need to get loaded at boot time? If yes, where is it supposed to be located for u-boot to find it?

[Charles Steinkuehler]

Not necessarily, it depends on how you have the system configured.

If you are setup to boot off the uSD card, you do not ever need to
program the FPGA, although you _can_ program the FPGA either from Linux
at run-time, via u-boot, or via an external flash device.

If you want u-Boot to program the FPGA, the rbf file typically resides
in the FAT partition along with the kernel and device-tree files, but it
can live elsewhere as well.

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

Charles,

Robert’s instructions only create two partitions, an A2 and an EXT4.
The “A2” partition is where  u-boot-with-spl.sfp gets written.
The ext4 partition is for linux. The kernel resides in the “boot” directory along with a directory for the extlinux.conf file, and a directory for the device tree files.

I don’t see a .rbf file copied anywhere.

Alan

[Charles Steinkuehler]

The FAT partition is not necessary, but that's how the Altera created
images are setup. If you want to use U-Boot to program the FPGA, put
the rbf file in the /boot directory along with the kernel. You'll
probably also have to update the U-Boot script, but it may be setup to
program the FPGA if it finds a magically named rbf file (probably
soc_system.rbf, but I haven't worked much with Robert's images for the
SoC+FPGA parts yet).

--
Charles Steinkuehler
cha...@steinkuehler.net

[mugginsac]

I finally got the USB console working under stretch.
I didn't have the dtb file that it was looking (socfpga_cyclone5_de10_nano.dtb) for in the boot/dtbs/4.9.68-socfpga-r3/ directory.
After I fixed that it is either loading the kernel or trying to load the kernel. It is getting a crc error and then starting over to try and load the kernel.

Any help interpreting what it is telling me would be greatly appreciated. The log is below:

=~=~=~=~=~=~=~=~=~=~=~= PuTTY log 2018.02.06 15:47:15 =~=~=~=~=~=~=~=~=~=~=~=

U-Boot SPL 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14)
drivers/ddr/altera/sequencer.c: Preparing to start memory calibration

drivers/ddr/altera/sequencer.c: CALIBRATION PASSED
drivers/ddr/altera/sequencer.c: Calibration complete
Trying to boot from MMC1

U-Boot 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14 -0800)

CPU:   Altera SoCFPGA Platform
FPGA:  Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT:  SD/MMC Internal Transceiver (3.0V)
       Watchdog enabled
I2C:   ready
DRAM:  1 GiB
MMC:   dwmmc0@ff704000: 0
*** Warning - bad CRC, using default environment

In:    serial
Out:   serial
Err:   serial
Model: Terasic DE10-Nano
Net:  
Error: ethernet@ff702000 address not set.
No ethernet found.

Hit any key to stop autoboot:  2 1 0

switch to partitions #0, OK
mmc0 is current device

Scanning mmc 0:2...
Found /boot/extlinux/extlinux.conf
Retrieving file: /boot/extlinux/extlinux.conf
179 bytes read in 21 ms (7.8 KiB/s)
1:    Linux 4.9.68-socfpga-r3
Retrieving file: /boot/vmlinuz-4.9.68-socfpga-r3
5455176 bytes read in 380 ms (13.7 MiB/s)
append: root=/dev/mmcblk0p2 ro rootfstype=ext4 rootwait quiet
Retrieving file: /boot/dtbs/4.9.68-socfpga-r3/socfpga_cyclone5_de10_nano.dtb
14997 bytes read in 45 ms (325.2 KiB/s)
## Flattened Device Tree blob at 02000000
   Booting using the fdt blob at 0x2000000
   reserving fdt memory region: addr=0 size=1000
   Loading Device Tree to 03ff9000, end 03fffa94 ... OK

Starting kernel ...


U-Boot SPL 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14)
drivers/ddr/altera/sequencer.c: Preparing to start memory calibration

drivers/ddr/altera/sequencer.c: CALIBRATION PASSED
drivers/ddr/altera/sequencer.c: Calibration complete
Trying to boot from MMC1

U-Boot 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14 -0800)

CPU:   Altera SoCFPGA Platform
FPGA:  Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT:  SD/MMC Internal Transceiver (3.0V)
       Watchdog enabled
I2C:   ready
DRAM:  1 GiB
MMC:   dwmmc0@ff704000: 0
*** Warning - bad CRC, using default environment

In:    serial
Out:   serial
Err:   serial
Model: Terasic DE10-Nano
Net:  
Error: ethernet@ff702000 address not set.
No ethernet found.

Hit any key to stop autoboot:  2 0

[schoo...@btinternet.com]

Read the gist that Michael did for the DE0

There is a link to it in the blogs.

You have to set the NIC MAC address etc and then restart

--
website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit
---
You received this message because you are subscribed to the Google Groups "Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email to machinekit+...@googlegroups.com.
Visit this group at https://groups.google.com/group/machinekit.
For more options, visit https://groups.google.com/d/optout.

[schoo...@btinternet.com]

On 07/02/18 07:03, 'schoo...@btinternet.com' via Machinekit wrote:

Read the gist that Michael did for the DE0

Specifically here
https://gist.github.com/mhaberler/89a813dc70688e35d8848e8e467a1337

where you will see exactly the same error message you report, because the MAC address is not set.

[mugginsac]

Thanks Schooner and Michael,

I followed the instructions through the setenv default -a; setenv ethaddr ba:d0:4a:9c:4e:ce; saveenv; reset.

Now I get a little farther. Do I need to remove the 'quiet' in extlinux.conf following the 'rootwait' so that I can see what is going on in the starting kernel?
=~=~=~=~=~=~=~=~=~=~=~= PuTTY log 2018.02.07 20:16:46 =~=~=~=~=~=~=~=~=~=~=~=

U-Boot SPL 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14)
drivers/ddr/altera/sequencer.c: Preparing to start memory calibration
drivers/ddr/altera/sequencer.c: CALIBRATION PASSED
drivers/ddr/altera/sequencer.c: Calibration complete
Trying to boot from MMC1

U-Boot 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14 -0800)

CPU:   Altera SoCFPGA Platform
FPGA:  Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT:  SD/MMC Internal Transceiver (3.0V)
       Watchdog enabled
I2C:   ready
DRAM:  1 GiB
MMC:   dwmmc0@ff704000: 0

In:    serial
Out:   serial
Err:   serial
Model: Terasic DE10-Nano

Net:   eth0: ethernet@ff702000
Hit any key to stop autoboot:  0

switch to partitions #0, OK
mmc0 is current device
Scanning mmc 0:2...
Found /boot/extlinux/extlinux.conf
Retrieving file: /boot/extlinux/extlinux.conf
179 bytes read in 21 ms (7.8 KiB/s)
1:    Linux 4.9.68-socfpga-r3
Retrieving file: /boot/vmlinuz-4.9.68-socfpga-r3
5455176 bytes read in 380 ms (13.7 MiB/s)
append: root=/dev/mmcblk0p2 ro rootfstype=ext4 rootwait quiet
Retrieving file: /boot/dtbs/4.9.68-socfpga-r3/socfpga_cyclone5_de10_nano.dtb
14997 bytes read in 45 ms (325.2 KiB/s)
## Flattened Device Tree blob at 02000000
   Booting using the fdt blob at 0x2000000
   reserving fdt memory region: addr=0 size=1000
   Loading Device Tree to 03ff9000, end 03fffa94 ... OK

Starting kernel ...

<My note: At this point it appears to start the process over >

U-Boot SPL 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14)
drivers/ddr/altera/sequencer.c: Preparing to start memory calibration
drivers/ddr/altera/sequencer.c: CALIBRATION PASSED
drivers/ddr/altera/sequencer.c: Calibration complete
Trying to boot from MMC1


U-Boot 2018.01-rc1-dirty (Jan 25 2018 - 15:56:14 -0800)

CPU:   Altera SoCFPGA Platform
FPGA:  Altera Cyclone V, SE/A6 or SX/C6 or ST/D6, version 0x0
BOOT:  SD/MMC Internal Transceiver (3.0V)
       Watchdog enabled
I2C:   ready
DRAM:  1 GiB
MMC:   dwmmc0@ff704000: 0

In:    serial
Out:   serial
Err:   serial
Model: Terasic DE10-Nano

Net:   eth0: ethernet@ff702000

Hit any key to stop autoboot:  2 1 0

=>

[schoo...@btinternet.com]

First thing to check is probably that your dip switches are correctly set.

Seem to recall my DE0 board arrived with them in FPGA programming mode positions