[Shr-Devel] Problems with own-build shr-lite-image

[Ingvaldur Sigurjonsson]

Arigead wrote:
> Ghislain MARY wrote:
>   
>> Hi everyone,
>>
>> I am trying to build the shr-lite-image from shr-unstable. The image is
>> successfully built but when trying it on my freerunner I have some
>> problems. The image is booting correctly until x is started.
>> Here is the error I get from the x.log:
>>
>> <<<< Enlightenment Error >>>>
>> Enlightenment found Evas can't create a buffer canvas. Please check
>> Evas has Software Buffer engine support.
>>
>> I have checked the content of this image against the last one from
>> build.shr-project.org and saw that a lot of stuff is missing from the
>> image I built. I tried with the full unstable image and the lite testing
>> image and I have the same behaviour.
>>
>> So my question is, do you have local modifications or a specific local
>> configuration to build the images of build.shr-project.org or I am doing
>> something wrong (I am following
>> http://wiki.shr-project.org/trac/wiki/Building%20SHR)?
>>
>> Thanks,
>> Ghislain
>>
>>     
>
> Thanks a million for asking Ghislain, I hope you get an answer to this
> question. I've been trying to build this for half a week and was just
> about to ask a question about this. I thought it must be my local setup
> and have tried to build on two different machines with the same result.
>
> I was hoping to help out with SHR as much as I could and thought the
> first step might be building SHR before starting into specific packages.
> Still working on the first step though.
>
> Please if somebody has any info as to how to build SHR so that it will
> boot in the Free Runner that would be gratefully received.
>   
Hi

  I've been building 'shr-unstable-image' at home for some time now and 
it's running fine on my phone. I had similar problems as you describe in 
beginning of august. I dont do any local local-builds though (i.e. no 
'require local-builds.inc' in my shr-unstable/conf/local.conf).

  From what I recall somebody was trying to get rid of all 
'*-dev'-packages from the image but that lead to a lot of missing 
components that were automatically pulled into the image by the 
'*-dev'-packages. It's a neccessary task to clean up that mess but 
certainly not simple, kudos to those who try.

  Ainulindale suggested on IRC (#openmoko-cdevel)  the following which 
lead to a lot of missing packages were built and problems resolved 
(thank you again :-) ).

    (11:42:51 PM) Ainulindale:make update (twice to be sure)
    (11:42:51 PM) Ainulindale: then
                               cd shr-unstable
    (11:42:59 PM) Ainulindale: rm -rf tmp/stamps/*/task-shr*
    (11:43:05 PM) Ainulindale: rm -rf tmp/stamps/*/shr-splash*
    (11:43:18 PM) Ainulindale: rm -rf tmp/deploy/glibc/ipk/*/shr-splash*
    (11:43:23 PM) Ainulindale: rm -rf tmp/deploy/glibc/ipk/*/task-shr*
    (11:43:30 PM) Ainulindale: and bitbake shr-lite-image

  So I made a script like:

    #!/bin/bash

    cd shr-unstable
    rm -rf tmp/stamps/*/task-shr*
    rm -rf tmp/stamps/*/shr-splash*
    rm -rf tmp/deploy/glibc/ipk/*/shr-splash*
    rm -rf tmp/deploy/glibc/ipk/*/task-shr*
    . ./setup-env
    bitbake shr-lite-image

  Try that and see if the missing parts fall into the puzzle.

  Regarding the missing engines in E make sure you have these packages 
installed:

    root at om-gta02 ~ $ opkg list_installed | grep engine
    etk-engines-evas - 2:0.1.0.042+svnr41040-r8 -
    etk-engines-fb - 2:0.1.0.042+svnr41040-r8 -
    etk-engines-software-x11 - 2:0.1.0.042+svnr41040-r8 -
    etk-engines-software-x11-16 - 2:0.1.0.042+svnr41040-r8 -
    etk-engines-x11 - 2:0.1.0.042+svnr41040-r8 -
    evas-engine-buffer - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-fb - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-software-16 - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-software-16-x11 - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-software-generic - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-software-x11 - 2:0.9.9.050+svnr41040-r2.5 -
    evas-engine-xrender-x11 - 2:0.9.9.050+svnr41040-r2.5 -

  Seeing a little more from /tmp/x.log would have been more describing.

  Good luck and please dont give up hope :-)

Regards
- Ingi