git clone git://github.com/Angstrom-distribution/setup-scripts.git
works fine.
I’m guessing that the right machine name for the BBB is the same as the BBW = beaglebone.
Can anyone confirm that is the right name to be using?
MACHINE=beaglebone ./oebb.sh config beaglebone
quickly dies due to this test in the script:
if ! git help log | grep -q no-abbrev ; then
echo “Your installed version of git is too old, it lacks --no-abbrev. Please install 1.7.6 or newer”
exit 1
fi
That bit of code seems to be using the presence of “no-abbrev” in the git docs as a proxy decision for the installed git version. I’m thinking that will always fail on the BBB as
git help log
(when run on the BBB ) results in:
No manual entry for git-log
Since git --Version tells me that git is 1.7.7, I commented out the “troublesome” test and forged ahead.
MACHINE=beaglebone ./oebb.sh update
tries to run - but fails with
Processing meta-xilinx: Fetching origin
updating to d196fa93c7ff5e080d4c44e2b83aed472f32b2c7
fatal: Could not parse object ‘d196fa93c7ff5e080d4c44e2b83aed472f32b2c7’.
I’d like to know two things:
can the BBB anstrom package be built with the BBB?
If so, Where can I find the right incantations to do so?
From greping the scripts, I have deduced that the name of the BBB is beaglebone-black.
I have tried to build Angstrom from Debian (Linux debian7 3.2.0-4-amd64 #1 SMP Debian 3.2.46-1 x86_64 GNU/Linux) and from Ubuntu (Linux ubuntu 3.5.0-23-generic #35~precise1-Ubuntu SMP Fri Jan 25 17:13:26 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux) and in both cases the process fails on the command:
Processing meta-xilinx: Fetching origin updating to d196fa93c7ff5e080d4c44e2b83aed472f32b2c7 fatal: Could not parse object ‘d196fa93c7ff5e080d4c44e2b83aed472f32b2c7’.
I went back and tried to build Angstrom for the plan beagleboard and got the same result. For what it’s worth, this procedure worked on Monday, but not now.
Note: I was using git 1.7.10.4 (Debian) and 1.7.9.5 (Ubuntu).
So at this point I would suggest that the problem is unrelated to trying to run this on the BBB itself, but may be caused by some fault in the git repository?
…
Using /home/jlf/.oe/environment-angstromv2012.12 to setup needed variables. It is recommended to do ‘. /home/jlf/.oe/environment-angstromv2012.12’ and run ‘bitbake something’ without using ./oebb.sh as wrapper
Executing: MACHINE=beaglebone-black bitbake cloud9-gnome-image
Pseudo is not present but is required, building this first before the main build
ERROR: OE-core’s config sanity checker detected a potential misconfiguration.
Either fix the cause of this error or at your own risk disable the checker (see sanity.conf).
Following is the list of potential problems / advisories:
Please set a valid MACHINE in your local.conf or environment
ERROR: Execution of event handler ‘check_sanity_eventhandler’ failed
I always end up with: Processing meta-xilinx: Fetching origin updating to d196fa93c7ff5e080d4c44e2b83aed472f32b2c7 fatal: Could not parse object ‘d196fa93c7ff5e080d4c44e2b83aed472f32b2c7’.
So you are already farther along than I am. When I solve the fatal error problem, I can look forward to this next obstacle. Have you had any luck trying to build for the beaglebone rather than the beaglebone-black?
Yes, one time with ‘MACHINE=beaglebone’ I got all the way through ‘./oebb.sh bitbake cloud9-gnome-image’. Lots of warnings, but no errors. And the resulting rootfs did not match the one in the downloaded latest “BBB-eMMC-flasher-2013.06.20.img.xz” from http://beagleboard.org/latest-images.
That’s why I posted “[BBB] Am I building the latest Angstrom image right?” asking what MACHINE should be and what image I should be building to match what’s on the latest-images site.
I’ve run into a couple of build issues…haven’t worked on them enough to solve just yet:
makeinfo got upgraded, and this causes a build issue with linaro-gcc (build failure occurs on Debian Jessie & Fedora19)
nodejs failed to build with an error which effectively meant path too long
One thing you might try to ‘fix’ the meta-xilinx issue you’re seeing is remove the ~/.oe/environment* file (or save it somewhere else) and re-run the initial config steps…I did run into a random issue where setup-scripts picked up a different branch than master, and that caused issues. I have no idea quite how it did that.
I’m running into this problem this weekend. Were any of you able to figure out what was going wrong? It definitely seems like it’s something wrong with the repo, or maybe the OpenEmbedded glue that holds it all together?
Would greatly appreciate any insight you all had after (presumably) getting past this issue.
First of all, I have not seen the "git" error. My version of git is 1.7.10.4.
Secondly, I too get the error in the xilinx fetch. I presume it is pegged to a removed commit. God only knows why! Anyway I have just ignored that "fatal" message and continued with:
First of all, I have not seen the "git" error. My version of git is
1.7.10.4.
Secondly, I too get the error in the xilinx fetch. I presume it is pegged
to a removed commit. God only knows why! Anyway I have just ignored that
"fatal" message and continued with:
As best I can tell based on some IRC discussion over the weekend, 2012.12
no longer builds due to changes in dependent repositories. You'll need to
try 2013.06. You should probably check the angstrom developer mailing list
for the history.