On Thu, Feb 23, 2012 at 2:29 AM, Tassilo Philipp wrote: > Hello ml, > > I recently ran into many problems using rake to build a few of my projects, > whenever rake calls itself recursively (for subprojects, etc. as in a typical > make-style build tool setup). It's very random - sometimes I get a single > error, sometimes nothing, but most of the time it sits there and prints the > following until I kill the build: > > 17 [main] ruby 1028 child_info_fork::abort: address space needed by > 'etc.so' (0x360000) is already occupied > 1 [main] ruby 2076 child_info_fork::abort: address space needed by > 'etc.so' (0x360000) is already occupied > 1 [main] ruby 2128 child_info_fork::abort: address space needed by > 'etc.so' (0x360000) is already occupied > .... > > I checked the mailing list and found all kind of infos about similar problem, > but none of the solutions worked for me. What I - unsuccessfully - tried so > far: > > - using the snapshot of feb 9, 2012 (which contains a fix by Corinna, that > apparently solved a very similar problem pointed out a few days ago) > - using latest snapshot of feb 20, 2012 > - using cygwin 1.7.10 and 1.7.9 > - playing around with rebaseall and peflagsall, with and without rebooting > the machine > - using ruby 1.8.7 as it is available via setup.exe > - using self-built version of llatest release of ruby 1.9.3 > - wiping cygwin off my machine and reinstalling a mint version > - many different combinations of the above > > So well, I guess 'm stuck. Any ideas?
rebaseall ? That is the standard solution for fork problem. http://cygwin.com/faq-nochunks.html#faq.using.fixing-fork-failures It is likely needed any time you install a different cygwin version including snapshots or upgrade any other packages. Regards Marco -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple