From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mga01.intel.com ([192.55.52.88]) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1SWrZm-0005h8-5c for openembedded-core@lists.openembedded.org; Tue, 22 May 2012 18:04:46 +0200 Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by fmsmga101.fm.intel.com with ESMTP; 22 May 2012 08:53:30 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.71,315,1320652800"; d="scan'208";a="155718988" Received: from unknown (HELO [10.255.12.14]) ([10.255.12.14]) by fmsmga001.fm.intel.com with ESMTP; 22 May 2012 08:53:29 -0700 From: Tom Zanussi To: Patches and discussions about the oe-core layer In-Reply-To: References: <4FB50005.7000807@gherzan.ro> Date: Tue, 22 May 2012 10:53:47 -0500 Message-ID: <1337702027.32464.22.camel@elmorro> Mime-Version: 1.0 X-Mailer: Evolution 2.28.3 Subject: Re: Linux-yocto 3.0 crashes at do_patch... X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.11 Precedence: list Reply-To: Patches and discussions about the oe-core layer List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 22 May 2012 16:04:47 -0000 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit FWIW, I just did a fresh minimal beagleboard build without any workaround and didn't see any problems building. This was using the proxy setup that had previously failed for me with the zero-length 3.0 kernel tarball. Tom On Tue, 2012-05-22 at 13:02 +0300, Andrei Gherzan wrote: > Even after using your workaround i can still reproduce this issue. > Yocto linux still fails. > > | ERROR. Could not find an excutable target for yocto/standard/base > | ERROR. Could not locate meta series for yocto/standard/beagleboard > | ERROR. Could not modify yocto/standard/base > NOTE: package linux-yocto-3.0.24+git1 > +34e0d2b4b4e9778b31f9ea99ca43f0dc71a7ee23_1 > +6b4bf6173b0bd2d1619a8218bac66ebc4681dd35-r4: task do_patch: Failed > ERROR: Task 3 > (/media/HDD/yocto/poky/meta/recipes-kernel/linux/linux-yocto_3.0.bb, > do_patch) failed with exit code '1' > > > On Thu, May 17, 2012 at 11:48 PM, Flanagan, Elizabeth > wrote: > On Thu, May 17, 2012 at 7:59 AM, Andrei Gherzan > wrote: > > Today i just gave another try.... same result. > > Responding here as well as the yocto list: > > A few people have been having this issue and after some > discussion we > think we have a root cause. Let me first explain what > happened. > > The autobuilder infrastructure shares a common DL_DIR in order > to > optimize the builds and provide the source mirror. This DL_DIR > is > rsynced to the publicaly available source mirror on a regular > basis. > During our infrastructure upgrade the autobuilder attempted to > pull > the tar archive for linux-yocto-3.0 > (http://downloads.yoctoproject.org/mirror/sources/git2_git.yoctoproject.org.linux-yocto-3.0.tar.gz) > but due to network issues related to our recent infrastructure > upgrade > the file ended up as a 0 byte file. This file was rsynced the > the > source mirror. A few days ago this was mentioned as an issue. > Michael > Halstead corrected the network issue and I manually removed > the file. > The autobuilder repulled the archive correctly and repopulated > the > source mirror with it. > > Things should have been ok at this point. However, we were > still > getting reports of people sporadically running into a 0 byte > file. Not > all the time, but some of the time. Michael Halstead and I > just tailed > out the apache logs with Tom Zanussi wgetting the problematic > file, > watching for failures. No failures were seen in the error logs > however > Tom was still getting sporadic 0 byte files. Which means that > his > request wasn't even hitting the server. I was not able to > replicate > this. The only difference we could see is that I generally do > not go > through a proxy whereas Tom was. > > Our theory at this point is that all of the people who are > hitting > this issue are going through a proxy and that their proxy > setup has > cached that 0 byte file and refuses to regenerate it. There > isn't a > lot we can do from our end about that other than suggest the > following > work around. If you run into odd fetch issues with *just* the > linux-yocto-3.0 tar archive, clear out your DL_DIR of > git2_git.yoctoproject.org.linux-yocto-3.0.tar.gz and set the > following: > > PREMIRRORS ?= "\ > bzr://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > cvs://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > git://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > hg://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > osc://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > p4://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > svk://.*/.* http://autobuilder.yoctoproject.org/sources/ \n > \ > svn://.*/.* http://autobuilder.yoctoproject.org/sources/ \n" > > MIRRORS =+ "\ > ftp://.*/.* http://autobuilder.yoctoproject.org/sources/ > \n \ > http://.*/.* http://autobuilder.yoctoproject.org/sources/ > \n \ > https://.*/.* http://autobuilder.yoctoproject.org/sources/ > \n" > > This will force the fetch to go to the autobuilder host of > sources and > should work around any proxy caching issues. Then call your > local IT > folks and ask them to clear out your proxy cache. If you run > into any > fetch issues not related to this file, please let me know. > > -b > > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core > > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core