Re: gcc-4.0.0-20050410 relase candidate test results

[Bryce McKinlay <mckinlay at redhat dot com>]

Same results results for powerpc-unknown-linux-gnu:

FAIL: Array_3 -O3 execution - bytecode->native test

               === libjava Summary ===

# of expected passes            3070
# of unexpected failures        1
# of expected failures          14
# of untested testcases         19


ia64-unknown-linux-gnu fails all native compilation tests due to a 
linker warning:

$ gcj Hello.java --main=Hello
/usr/bin/ld: warning: libunwind.so.1, needed by 
/home/mckinlay/gcc-ia64/lib/gcc/ia64-unknown-linux-gnu/4.0.0/../../../libgcj.so, 
may conflict with libunwind.so.7

Sure enough, libgcj.so seems to be linked against two different unwind 
libraries:

$ ldd ~/gcc-ia64/lib/libgcj.so.6.0.0
       libpthread.so.0 => /lib/tls/libpthread.so.0 (0x2000000001fe4000)
       libdl.so.2 => /lib/libdl.so.2 (0x2000000002014000)
       libgcc_s.so.1 => /home/mckinlay/gcc-ia64/lib/libgcc_s.so.1 
(0x2000000002030000)
       libc.so.6.1 => /lib/tls/libc.so.6.1 (0x2000000002050000)
       libunwind.so.1 => /usr/lib/libunwind.so.1 (0x20000000022bc000)
       /lib/ld-linux-ia64.so.2 => /lib/ld-linux-ia64.so.2 
(0x2000000000000000)
       libunwind.so.7 => /home/mckinlay/gcc-ia64/lib/libunwind.so.7 
(0x2000000002308000)

I wonder what causes this?

Bryce



Andrew Haley wrote:

> x86_64-unknown-linux-gnu passes smoke tests:
>
> FAIL: Array_3 -O3 -m32 execution - bytecode->native test
>
>                === libjava Summary ===
>
> # of expected passes            3070
> # of unexpected failures        1
> # of expected failures          14
> # of untested testcases         19
>
> FAIL: Array_3 -O3 execution - bytecode->native test
>
>                === libjava Summary ===
>
> # of expected passes            3070
> # of unexpected failures        1
> # of expected failures          14
> # of untested testcases         19
> runtest completed at Mon Apr 11 11:56:42 2005
>
> Sigh.  I shoulda made Array_3 an XFAIL.
>
> Andrew.
>