results for egcs-971031 gcc/g++/g77 testsuite on mips-sgi-irix6.2

results for egcs-971031 gcc/g++/g77 testsuite on mips-sgi-irix6.2

[Jeffrey A Law]

--
Sender: jsd@tigger.scc.uni-weimar.de
Message-ID: <345D1F9D.41C6@informatik.uni-weimar.de>
Date: Mon, 03 Nov 1997 01:49:33 +0100
From: Jan Springer <jan.springer@informatik.uni-weimar.de>
Organization: [atelier, virtual [i-group] (B11R2.34)
X-Mailer: Mozilla 3.01SC-SGI (X11; I; IRIX 6.2 IP22)
MIME-Version: 1.0
To: egcs@cygnus.com
Subject: results for egcs-971031 gcc/g++/g77 testsuite on mips-sgi-irix6.2
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

xxxconfig:
- -------

xxxconfigure: --enable-multilib --enable-shared --enable-threads
- --enable-haifa
BOOT_CFLAGS: -O4 -g
CFLAGS: -O -g (see notes)
CXXFLAGS: -O -g (see notes)

method:
- -------

gmake bootstrap[-lean]

results:


Native configuration is mips-sgi-irix6.2

                === gcc tests ===

FAIL: gcc.c-torture/compile/920928-2.c,  -O1  
FAIL: gcc.c-torture/execute/complex-5.c execution,  -O0 
FAIL: gcc.c-torture/execute/complex-5.c execution,  -O1 
FAIL: gcc.c-torture/execute/complex-5.c execution,  -O2 
FAIL: gcc.c-torture/execute/complex-5.c execution,  -O2
- -fomit-frame-pointer -finline-functions 

                === gcc Summary ===

# of expected passes            4878
# of unexpected failures        5
# of expected failures          5
# of unsupported tests          7
/usr/local/src/egcs-971031/gcc/xgcc version egcs-2.90.15 971031
(gcc2-970802 experimental)

                === g++ tests ===

FAIL: g++.benjamin/warn01.C (test for excess errors)
XPASS: g++.jason/destruct3.C - (test for bogus messages, line 38)
FAIL: g++.jason/rfg10.C enum overflow (test for errors, line 9)
FAIL: g++.jason/rfg10.C (test for excess errors)
FAIL: g++.jason/template31.C (test for excess errors)
XPASS: g++.mike/dyncast1.C  Execution test
XPASS: g++.mike/dyncast2.C  Execution test
FAIL: g++.mike/p9732b.C  Execution test

                === g++ Summary ===

# of expected passes            3372
# of unexpected failures        5
# of unexpected successes       3
# of expected failures          83
# of untested testcases         6
/usr/local/src/egcs-971031/gcc/testsuite/../xgcc version egcs-2.90.15
971031 (gcc2-970802 experimental)

                === g77 tests ===

                === g77 Summary ===

# of expected passes            108
/usr/local/src/egcs-971031/gcc/g77 version egcs-2.90.15 971031
(gcc2-970802 experimental)

notes:
- ------

  - dwarf2out.c patch applied
    [ http://www.cygnus.com/ml/egcs/1997-Nov/0009.html ]

  - w/ the above mentioned config the compilation aborts due to
    virtual memory exhaustion while compiling libio/outfloat.c
    (default was -O2 -g; w/ -O -g this went ok)

  - gmake bootstrap will not propagate CFLAGS or CXXFLAGS to the
    library built; gmake all-target-lib* will

  - while building libstdc++.2.8.0.so the linker complains not to
    find a file named -lm; kicked -lm fom the SHDEPS var in the
    Makefile to complete the linking

  - when configuring w/o --enable-multilib the compilation aborts
    by attempting to combine object files from the libiberty in n32
    format to the libio in 64 format; hence --enable-multilib should
    be should be default on mips-sgi-irix*

    [side note: sgi prefers an naming scheme for the various object
     formats; so i've arranged my systems to have

	/usr/local/lib
	/usr/local/lib32
	/usr/local/lib64

     debugging/profiling libs go into debug/profile subdirectories
     beneath their appropriate object format directory. this seems
     to me a bit more rational than to find directories name "mabi=64".
     the sgi ld (the one w/ the ido) uses different variables for the
     various library paths, i.e.:

	LD_LIBRARY_PATH=/lib:/usr/lib:/usr/lib/abi:/usr/local/lib:.
	LD_LIBRARYN32_PATH=/lib32:/usr/lib32:/usr/local/lib32:.
	LD_LIBRARY64_PATH=/lib64:/usr/lib64:/usr/local/lib64:.

    ]

  - configuring w/ --enable-threads is ok while
    --enable-threads=pthread or sth. similar isn't known to config;
    shouldn't this be possible? (irix 6.3 and above have pthread
    support in the base system; irix 6.2 needs some patches,
    though it works)

  - this testrun is the first ever that was able to build and run
    w/ the haifa scheduler enabled on irix6.2 (at least for me)

	j.

- -- 
+---------------------------------------+-----------------------+
| jan springer                          | this                  |
| student of applied computer science   |    space              |
| bauhaus-university weimar, germany    |        reserved       |
+---------------------------------------+               for     |
| jan.springer@informatik.uni-weimar.de |                 quote |
+---------------------------------------+-----------------------+

------- End of Forwarded Message