When Cygwin gets fussy about forking a new process

Things like

     3 [main] bash 2216 child_info_fork::abort: data segment start: parent(0x26D000) !=     child(0x38D000)
-bash: fork: retry: Resource temporarily unavailable

It appears to be a recent Red Hat mishap. One of those bugs that happen because of the common belief that if you don’t play around with your software all the time, it will be considered dead.

Workaround: Close all Cygwin programs (and windows of course), find the ash.exe in the /bin directory, and double click it to open a primitive command window.

Go

$ /bin/rebaseall

And the problem is over with. I won’t bother understanding what went wrong, since the explanation is most like not very inspiring in cases like this. More about rebaseall here.