Difference between revisions of "How to load and run kernel from sdram"

From RidgeRun Developer Connection
Jump to: navigation, search
m (Created page with 'When testing a new kernel configuration, it can be helpful to not always save the kernel to flash; instead simply load and run from SDRAM. You can do this manually by entering t...')
 
m
Line 27: Line 27:
 
</pre>
 
</pre>
  
The filename will be something like <tt>XXXX</tt>.
+
The filename will be something like <tt>kernel.uImage.tfischer.overo</tt>.
  
 
== Configuring uboot to auto-load / auto-boot using tftp ==
 
== Configuring uboot to auto-load / auto-boot using tftp ==
Line 34: Line 34:
  
 
<pre>
 
<pre>
 
+
setenv autoload no
 +
dhcp
 +
setenv autorun no
 +
tftp
 
</pre>
 
</pre>
  
 
[[Category:Whitepaper]]
 
[[Category:Whitepaper]]

Revision as of 22:50, 18 December 2010

When testing a new kernel configuration, it can be helpful to not always save the kernel to flash; instead simply load and run from SDRAM. You can do this manually by entering the u-boot commands shown below.


Configuring the kernel

Using a RidgeRun SDK, you can configure the kernel via

make -C $DEVDIR config

which launches the SDK configuration tool. Select kernel from the main menu and adjust the kernel settings as needed.

Building the kernel image

Using a RidgeRun SDK, you can build the new kernel image via

make -C $DEVDIR kernel

This will automatically copy the new kernel image to /srv/tftp on your desktop computer. You can see the file name being used

ls -ltr /srv/tftp

The filename will be something like kernel.uImage.tfischer.overo.

Configuring uboot to auto-load / auto-boot using tftp

Adjust the follow u-boot commands to match your host computer IP address and /srv/tftp kernel filename.

setenv autoload no
dhcp
setenv autorun no
tftp