EGCS support

[Mark Keates <markk at dendrite dot co dot uk>]

Does this mailing list cover EGCS questions too?

The link at http://gcc.gnu.org/ml/gcc/1999-02/msg00657.html reads:
"This is the mail archive of the egcs@egcs.cygnus.com mailing list for the
EGCS project. 
See the EGCS home page  (http://gcc.gnu.org/) for more information. "...
which has no references to EGCS, hence the post here.

I'm building a cross m68k-unknown-linux gcc on an i586-pc-cygwin32 host
(egcs-1.1.2).
The 'make cross' process runs upto a point of configuring
m68k-unknown-linux/libiberty.
(see the attached log file). I have some questions on the results I get from
xgcc:
1) (See specs attachment) I'm suspicious of *startfile, what derives this
and how is it interpreted?
2) (See print-search-dirs attachment) Is something corrupting the search
paths? 
   Can this be fixed or doesn't it matter as they refer to the build area?

I have made changes to the gcc/Makefile (listed in diffs attachment).
I'm guessing that because of the linux requirements I need to have glibc
built/installed 
for the target? Hence the addition of -Dinhibit-libc to the LIBGCC2 flags.
Bad idea?

I have been able to successfully build an cross m68k-coff by including 
newlib after setting up a symbolic link and using the configuration option
"--with-newlib".
This gets around the chicken/egg issue, so any ideas of how to get glibc
built
for the m68k-unknown-linux? I've used:
../glibc-2.2.2/configure --verbose --target=m68k-unknown-linux --prefix=/gnu
but this returns that *** The GNU C library is currently not available for
this platform.

Thanks,

Mark

 <<config.log>>  <<print-search-dirs>>  <<specs>>  <<diffs>>

config.log

print-search-dirs

specs

diffs