[Bug other/11412] New: boehm-gc testing problems

[gcc-bugzilla at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11412

           Summary: boehm-gc testing problems
           Product: gcc
           Version: 3.4
            Status: UNCONFIRMED
          Severity: enhancement
          Priority: P3
         Component: other
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: ro at techfak dot uni-bielefeld dot de
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: sparc-sun-solaris2.8
  GCC host triplet: sparc-sun-solaris2.8
GCC target triplet: sparc-sun-solaris2.8

I noticed a couple of problems with make check in boehm-gc:

* While make check produces output in a format appropriate for the testing
  summary produced for make mail-report.log, the output isn't included, so
  boehm-gc failures are easily overlooked.

* make check isn't multilib-aware: 

  running make check only tests the default multilib

* running make check in a non-default multilib may fail since libgcc_s.so
  isn't found

* make check (i.e. gctest) produces only a single PASS/FAIL result

Environment:
System: SunOS tikal 5.8 Generic_108528-19 sun4u sparc SUNW,Sun-Blade-1000
Architecture: sun4

	
host: sparc-sun-solaris2.8
build: sparc-sun-solaris2.8
target: sparc-sun-solaris2.8
configured with: /vol/gnu/src/gcc/gcc-dist/configure --prefix=/vol/gcc --with-local-prefix=/vol/gcc --disable-nls

How-To-Repeat:
Run make check in sparc-sun-solaris2.8/boehm-gc.  make check only emits one
PASS result, even when both sparc and sparcv9 multilibs are present.

Running make check in sparc-sun-solaris2.8/sparcv9/boehm-gc fails:

[...]
ld.so.1: gctest: fatal: ../../../gcc/libgcc_s.so.1: wrong ELF class: ELFCLASS32
/bin/ksh: 10890 Killed
FAIL: gctest

Running gctest with the correct LD_LIBRARY_PATH works, though:

> LD_LIBRARY_PATH=../../../gcc/sparcv9 ./gctest 
Completed 3 tests
Allocated 5192693 collectable objects
Allocated 306 uncollectable objects
Allocated 3716286 atomic objects
Allocated 32763 stubborn objects
Finalized 6631/6631 objects - finalization is probably ok
Total number of bytes allocated is 232476824
Final heap size is 20561920 bytes
Collector appears to work
Completed 72 collections
------- Additional Comments From ro at techfak dot uni-bielefeld dot de  2003-07-02 18:54 -------
Fix:
It would probably be useful to integrate gctest into the dejagnu framework,
which gives us automatic reporting by mail-report.log.  Besides, it might
be useful to split the single monolithic test into several smaller ones to
get individual test results instead of the single PASS/FAIL.