Re: multiple outputs rule

[Ralf Wildenhues]

Hello Bruno,

Thank you for the bug report.

* Bruno Haible wrote on Thu, Dec 14, 2006 at 03:04:20PM CET:
> 2) When the first branch of the "if mkdir ..." is executed, the data.lock
> directory is never removed. It should be removed after the recursive $(MAKE)
> invocation, preserving the return code.

Right, but:

>      ## This code is being executed by the first process.
>                  rm -f data.stamp; \
>                  $(MAKE) $(AM_MAKEFLAGS) data.stamp; \
>                  result=$$?; rm -rf data.lock data.stamp; exit $$result; \

Do you really want to remove the data.stamp file here?  That causes
unnecessary rebuilds.

Is there any reason why both the trap and the above code don't just use
  rmdir data.lock

instead of 'rm -rf data.lock'?  Is that to allow us to succeed the
second time in case data.lock was an unrelated user file before?
If not, do we anticipate missing permissions on '.', and if yes, why?

Cheers,
Ralf