<div dir="ltr">Thanks, Javi...<br><br>For my setup that doesn't seem to be sufficient.<br>Deleting the kernel files fixes the problem with old versions, but I still need to rebuild gumstix-kernel, then task-base-gumstix, then the image.<br>
<br>Perhaps there is something wrong with my (or gumstix) setup?<br><br>Lewis<br><br><br><div class="gmail_quote">On Tue, Aug 26, 2008 at 2:48 AM, Javi Roman <span dir="ltr"><<a href="mailto:javiroman@kernel-labs.org">javiroman@kernel-labs.org</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div><div></div><div class="Wj3C7c">On Tue, Aug 26, 2008 at 5:53 AM, Lewis Girod <<a href="mailto:girod@nms.csail.mit.edu">girod@nms.csail.mit.edu</a>> wrote:<br>
> Hi,<br>
><br>
> I am using bitbake to build images for a Gumstix.<br>
><br>
> Their setup has a .conf file that sets preferred versions for things like<br>
> the kernel; I used this to choose a different version than standard, and<br>
> this worked. Later, I wanted to switch back to the original version, but<br>
> although the right version built, it was still pulling in modules from the<br>
> other version.<br>
><br>
> Based on info on the web, etc, I tried forcing a rebuild of various tasks<br>
> and the main image, to no avail, e.g.<br>
><br>
> bitbake -f -c clean task-base-gumstix<br>
> bitbake -f -c clean gumstix-basic-image<br>
><br>
> bitbake -f -c rebuild task-base-gumstix<br>
> bitbake -f -c rebuild gumstix-basic-image<br>
><br>
> I ended up just using the sledgehammer and deleting the tmp directory... :(<br>
><br>
> My question is this: what is the correct principle to apply when rebuilding<br>
> parts of an image?<br>
> It seems that rebuilding at the top level won't necessarily rebuild stuff<br>
> that has changed.<br>
> How can I know what to rebuild? Is there a way to rebuild everything above<br>
> the toolchain (kernel, userspace, image)?<br>
> I don't have a good understanding of how this works so my solutions so far<br>
> have been ad-hoc and trial and error.<br>
> If someone could help me get the right model in mind, that would be great!<br>
><br>
> BTW, openembedded is a really nice system.. a bit of a learning curve though<br>
> :)<br>
><br>
> Lewis<br>
><br>
<br>
</div></div>The rootfs process pulls the latest ipk files from the<br>
tmp/deploy/.../ipk/.. directory to create the rootfs. You can clean<br>
manually kernel and modules ipk files from deploy directory, and then<br>
rebuild your rootfs.<br>
<font color="#888888"><br>
<br>
--<br>
Javi Roman<br>
<br>
_______________________________________________<br>
Openembedded-users mailing list<br>
<a href="mailto:Openembedded-users@lists.openembedded.org">Openembedded-users@lists.openembedded.org</a><br>
<a href="http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-users" target="_blank">http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-users</a><br>
</font></blockquote></div><br></div>