In scripts/lib/bsp/kernel.py, find_bblayers() fails in case bitbake_env_cmd doesn't output BBLAYERS. This is due to a bad Python formatted string on line 66. To reproduce: 1)In your build directory, set up an unexisting layer in conf/bblayers.conf (that will make 'bitbake -e' fail) 2)Call yocto-bsp create with any target arch. Actual results: The script fails with a TypeError. Expected results: The script fails with an error message describing the cause. Here's a proposed patch: --- /home/bjauny/yocto/poky-dora/scripts/lib/bsp/kernel.py 2013-11-05 16:13:38.250129173 +0100 +++ /home/bjauny/yocto/poky-dora/scripts/lib/bsp/kernel.py.new 2013-11-05 16:13:17.042128497 +0100 @@ -62,7 +62,7 @@ break if not bblayers: - print "Couldn't find BBLAYERS in 'bitbake -e' output, exiting." % \ + print "Couldn't find BBLAYERS in %s output, exiting." % \ bitbake_env_cmd sys.exit(1) Seen on dora head revision.
Created attachment 1608 [details] Proposed patch Attached proposed patch.
Hi, I've verified the problem and the patch looks good - feel free to submit it to the poky mailing list, or if you'd rather, I can do that (and give you credit).
Hi Tom, I'm not used to the mailing list process. Should I just post a new thread describing this bug or do I also need to commit my patch? Sorry if these are noob questions but I'm brand new to this, even if I used yocto intensively the last months. (In reply to comment #2) > Hi, I've verified the problem and the patch looks good - feel free to submit > it to the poky mailing list, or if you'd rather, I can do that (and give you > credit).
(In reply to comment #3) > Hi Tom, > I'm not used to the mailing list process. Should I just post a new thread > describing this bug or do I also need to commit my patch? Sorry if these are > noob questions but I'm brand new to this, even if I used yocto intensively > the last months. > > (In reply to comment #2) > > Hi, I've verified the problem and the patch looks good - feel free to submit > > it to the poky mailing list, or if you'd rather, I can do that (and give you > > credit). You can just submit the patch to the poky mailing list - the patch itself should describe teh problem and what you did to fix it. Here are some basic instructions on how you can do that - let me know if you need help with any of it. https://wiki.yoctoproject.org/wiki/Poky_Contributions If you don't want to go to the trouble of getting poky-contrib access, since this is such a simple patch, you could just send the patch in e-mail only e.g. make your change in git, 'tit commit -s' the change, then do 'git format-patch -1' and send that patch to the list.
(In reply to comment #4) > (In reply to comment #3) > > Hi Tom, > > I'm not used to the mailing list process. Should I just post a new thread > > describing this bug or do I also need to commit my patch? Sorry if these are > > noob questions but I'm brand new to this, even if I used yocto intensively > > the last months. > > > > (In reply to comment #2) > > > Hi, I've verified the problem and the patch looks good - feel free to submit > > > it to the poky mailing list, or if you'd rather, I can do that (and give you > > > credit). > > You can just submit the patch to the poky mailing list - the patch itself > should describe teh problem and what you did to fix it. Here are some basic > instructions on how you can do that - let me know if you need help with any > of it. > > https://wiki.yoctoproject.org/wiki/Poky_Contributions > > If you don't want to go to the trouble of getting poky-contrib access, since > this is such a simple patch, you could just send the patch in e-mail only > e.g. make your change in git, 'tit commit -s' the change, then do 'git > format-patch -1' and send that patch to the list. Just did your latest suggestion. Screwed up a little bit my mail, but I'll correct that. Thanks!
Fixed by poky/master commit: d60b4ff3517487bd111c6d7e1410882280229377