Hi, On Fr, 2016-05-27 at 21:42 +0100, André Draszik wrote: > > On Fri, May 27, 2016 at 4:21 PM, Mark Hatle <mark.ha...@windriver.com> > wrote: > > On 5/27/16 9:42 AM, André Draszik wrote: > > > On Fr, 2016-05-27 at 09:21 -0500, Mark Hatle wrote: > > > > On 5/27/16 5:54 AM, André Draszik wrote: > > > > > This doesn't work for me (at least on krogoth): > > > > > > > > > > ERROR: poky/meta/recipes-core/images/build-appliance- > > > > > image_14.0.0.bb: No IMAGE_CMD defined for IMAGE_FSTYPES entry > > > > > 'debugfs_vmdk' - possibly invalid type name or missing support > > > > > class > > > > > ERROR: Failed to parse recipe: poky/meta/recipes- > > > > > core/images/build- > > > > > appliance-image_14.0.0.bb > > > > > > > > > > I *instead* am using the attached patch successfully. > > > > > > > > > > Is anybody else facing the same issue? > > > > > > > > Ross's patch should have fixed this. It ensures that > > > > IMAGE_FSTYPES_DEBUGFS is > > > > always set to 'something'. > > > > > > Yes, and now bitbake is looking for an IMAGE_CMD for debugfs_vmdk > > > according > > > to the error message. Note that I didn't specify vmdk anywhere, my > > > IMAGE_FSTYPES is set to .tar.gz only. > > > > > > > You should check if the patch has been applied to the "krogoth" > > > > branch. > > > > > > I did apply it locally. For me, it only works if i revert Ross' patch > > > and > > > apply my patch instead. > > > > > > > Use bitbake -e <image>. View the IMAGE_FSTYPES and > > IMAGE_FSTYPES_DEBUGFS. Do > > you see vmdk in the former? If so, look at the comments above it and it > > shoudl > > show you where it came from. > > > > My guess is you have an image or BSP that is hard coding a specific > > 'vmdk'. > > > > I agree there is a problem here, some filesystem types do not make sense > > for > > debugfs type. I'm not sure how to necessarily reconcile this behavior, > > but vmdk > > doesn't seem reasonable to me here.. > > > > But first figure out where it's coming from. > > > > Setting IMAGE_FSTYPES_DEBUGFS = "tar.gz" in your local.conf should be > > enough to > > set the value and avoid the behavior. If it's not, again look at the > > bitbake -e > > output and see what it's saying it's doing.
> Nothing is pulling vmdk into my builds. vmdk does not end up in > my IMAGE_FSTYPES or IMAGE_FSTYPES_DEBUGFS. Again, the output from bitbake > -e is fine when reverting Ross' patch and applying the patch I posted > earlier instead. With Ross' patch (only), bitbake doesn't even finish > parsing due to the exception being thrown (so bitbake -e output is empty > apart from the exception). > > I know I can set IMAGE_FSTYPES_DEBUGFS explicitly to work around the > issue, but that's not it is intended to work. Do you have any more suggestions how to debug this? I am still in the exact same position. I need to revert Ross' patch and apply the patch I sent before. Is anybody successfully using this feature? Cheers, Andre' -- _______________________________________________ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core