clone 718104 -1 reassign -1 debhelper block 718104 by -1 block 718074 by -1 block 718068 by -1 found -1 9.20130630 thanks
On Sat, Jul 27, 2013 at 09:32:53PM +0200, David Suárez wrote: > Source: bsdiff > Version: 4.3-14 > Severity: serious > Tags: jessie sid > User: debian...@lists.debian.org > Usertags: qa-ftbfs-20130726 qa-ftbfs > Justification: FTBFS on amd64 > > Hi, > > During a rebuild of all packages in sid, your package failed to build on > amd64. > > Relevant part (hopefully): > > fakeroot debian/rules clean > > dh clean > > dh_testdir > > dh_auto_clean > > make[1]: Entering directory `/«PKGBUILDDIR»' > > Makefile:13: *** missing separator. Stop. > > make[1]: Leaving directory `/«PKGBUILDDIR»' > > dh_auto_clean: make -j1 distclean returned exit code 2 > > make: *** [clean] Error 2 > > The full build log is available from: > http://aws-logs.debian.net/ftbfs-logs/2013/07/26/bsdiff_4.3-14_unstable.log > > A list of current common problems and possible solutions is available at > http://wiki.debian.org/qa.debian.org/FTBFS . You're welcome to contribute! > > About the archive rebuild: The rebuild was done on EC2 VM instances from > Amazon Web Services, using a clean, minimal and up-to-date chroot. Every > failed build was retried once to eliminate random failures. I looked into this one. The first observation is, that there is a successful build of the same version using the same version of make. https://buildd.debian.org/status/fetch.php?pkg=bsdiff&arch=i386&ver=4.3-14&stamp=1337838907 So what is different here? One difference is the debhelper version in use. Working 9.20120523 vs broken 9.20130630. A quick check to rebuild with the old debhelper shows, that this is indeed the case. Versions up to 9.20130604 (inclusive) do not exhibit the broken behaviour. The most likely cause appears to be the fix of #713257. Note that in all of these cases, the Makefiles are bsd-ish Makefiles, that use some form of .if (note the leading dot). These Makefiles do not normally work with GNU make, but did work earlier, because they were only partly evaluated. It is not yet entirely clear to me what the best solution to bsd-ish Makefiles is. A different solution is to ask GNU make to support bsd-ish conditionals and includes. Helmut -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
