Re: Schedule for GNU make 3.81

[Paul D. Smith]

%% "Andreas M. Kirchwitz" <address@hidden> writes:

  amk>  fubar:
  amk>          /bin/echo hello

  amk> I've tested this on Solaris 10 3/05 (official release)
  amk> on various SPARC, Intel and AMD systems.

  amk> Solaris 10 doesn't seem to be very popular, but since
  amk> /usr/sfw/bin/gmake (version 3.80) on Solaris shows the
  amk> same behaviour, I couldn't believe that Sun ships a
  amk> non-working gmake, and so I did further investigation.

  amk> Now, I've found out that the environment variable "LC_CTYPE"
  amk> seems to have a big influence.

  amk> If LC_CTYPE is unset or set to "C" or "POSIX",
  amk> it works perfectly fine.

  amk> But if LC_CTYPE is set to "iso_8859_1", then make crashes.

  amk>  LC_CTYPE=iso_8859_1 /usr/local/src/make-3.81beta2/make
  amk>  Abort - core dumped

Hm.  That's a little confusing to me since make doesn't really do
anything with locales except through gettext.


Can you "rm *.o", then re-run make with debugging on and optimization
off, like this:

    make CFLAGS=-g

Then re-run your test, and use the debugger to send me a backtrace of
where it crashes?

Thanks...

-- 
-------------------------------------------------------------------------------
 Paul D. Smith <address@hidden>          Find some GNU make tips at:
 http://www.gnu.org                      http://make.paulandlesley.org
 "Please remain calm...I may be mad, but I am a professional." --Mad Scientist