Makefile bug?

[Jeffrey A Law]

  In message < 9709280405.AA04032@negril.msrce.howard.edu >you write:
  > Tracing through the Makefiles I see from the top egcs-970924 directory
  > that the ranlib enviroment variable is passed to the gcc Makefile through
  > RANLIB_FOR_TARGET and RANLIB in both BASE_FLAGS_TO_PASS and EXTRA_GCC_FLAGS
I was just starting to poke at the RANLIB_FOR_TARGET vs RANLIB stuff
when I had to switch gears and deal with EOQ stuff.  So odds are,
we're on the right track.

Basically I couldn't reconcile RANLIB vs RANLIB_FOR_TARGET since
the _value_ of the latter is the correct thing to use inside gcc
since it only builds target libraries, but everything seemed to
be using RANLIB.

I was a little confused about things until I found the magic
EXTRA_TARGET_FLAGS and EXTRA_GCC_FLAGS in the toplevel Makefile.in
Basically it sets RANLIB to RANLIB_FOR_TARGET (similarly for other
target things).  Those flags are then passed down to the gcc
make.  So, the values we need should be in RANLIB/RANLIB_TEST
instead of the _FOR_TARGET variants.

So, it seems to me that the right thing is to pass down RANLIB
and RANLIB_TEST like y'all have done (and AR).  I'll make this
change now so hopefully this problem will go away with the next
snapshot!

  > have fixed the build error I kept on getting.  2) How about a make
  > bootstrap-lean that has the three rm's mentioned above?
It's been mentioned as desirable by several folks; just waiting
on someone to implement it :-)

jeff