build failure for 2.6.11-mm2

Tom,

Do you have any ideas on what changes occurred in the makefiles to cause the uImage related failures:

http://l4x.org/k/?d=2173

- kumar

On Thu, Mar 10, 2005 at 02:10:04AM -0600, Kumar Gala wrote:

Do you have any ideas on what changes occurred in the makefiles to 
cause the uImage related failures:

http://l4x.org/k/?d=2173

This looks like mkimage is just not present on build host. I don't 
think it's included anywhere in kernel tree, so it must be installed 
separately.

--
Eugene

right, but if you look back in the pervious -mm builds that past, they also did not have mkImage but still succeeded.

It appear some change is now report the lack of mkImage as a build error.  If this is desired behavior, should we maybe look at having a copy of its source in the kernel tree?

- kumar

On Mar 10, 2005, at 2:26 AM, Eugene Surovegin wrote:

On Thu, Mar 10, 2005 at 02:10:04AM -0600, Kumar Gala wrote:
 > Do you have any ideas on what changes occurred in the makefiles to

quoted

cause the uImage related failures:

 >

quoted

http://l4x.org/k/?d=2173

This looks like mkimage is just not present on build host. I don't
think it's included anywhere in kernel tree, so it must be installed
separately.

-- 
 Eugene

On Thu, Mar 10, 2005 at 08:06:04AM -0600, Kumar Gala wrote:

right, but if you look back in the pervious -mm builds that past, they 
also did not have mkImage but still succeeded.

It appear some change is now report the lack of mkImage as a build 
error.  If this is desired behavior, should we maybe look at having a 
copy of its source in the kernel tree?

I think what's happened, but I haven't verified is that Andrew took an
incorrect patch to make 'mkuboot.sh', which invokes mkimage to produce a
uImage, fail if mkimage isn't found.  This isn't right.  The problem the
patch was trying to fix is that mkuboot.sh will say that it made a
uImage, even if mkimage isn't found.

-- 
Tom Rini
http://gate.crashing.org/~trini/