This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
New configure bootstrap failure on sparc-sun-solaris2.7
- From: "Kaveh R. Ghazi" <ghazi at caip dot rutgers dot edu>
- To: gcc at gcc dot gnu dot org
- Cc: neroden at gcc dot gnu dot org
- Date: Tue, 6 Jan 2004 08:46:42 -0500 (EST)
- Subject: New configure bootstrap failure on sparc-sun-solaris2.7
Since Jan 3rd, I am unable to bootstrap on sparc-sun-solaris2.7 due to
various target configure problems. The latest one I see is this
during target libiberty configure:
> Running configure in multilib subdir sparcv9
> pwd: /teal/caip5/ghazi/tmpdisk/gcc-testing/34/build/sparc-sun-solaris2.7
> mkdir sparcv9
> expr: Paren problem
> configure: creating cache ./config.cache
> [...]
> checking for cc... cc
> [...]
> checking to probe for byte ordering... cc: Warning: option -2 passed
> to ld
> /usr/ccs/bin/ld: illegal option -- 2
> configure: error: unknown endianess - sorry
> make[1]: *** [configure-target-libiberty] Error 1
> make[1]: Leaving directory `/teal/tmpdisk/ghazi/gcc-testing/34/build'
> make: *** [bootstrap] Error 2
Note that it's using my stage1 $CC for the target library. (I set CC
before I start the bootstrap.) I'm guessing that the complaint about
"option -2" is really because cc/ld gets "-O2". This happens even if
it's done serially.
I also noticed a suspicious message "expr: Paren problem". I don't
know why that happens. I have /usr/ucb in my path first so I get
/usr/ucb/expr if that matters. However I'm sure I got this message
before the latest configure hackery, so it's probably unrelated but
worth investigating.
I won't be able to experiment more until later tonight.
--Kaveh
--
Kaveh R. Ghazi ghazi@caip.rutgers.edu