Solaris/x86 testsuite results (egcs-1.1.1-prerelease)
Elgin Lee
ehl@funghi.com
Mon Nov 2 19:37:00 GMT 1998
Platform: Solaris 2.6 (x86) (i386-pc-solaris2.6)
Compiler: gcc version egcs-2.91.58 19981101 (egcs-1.1.1 pre-release)
FYI, I ran the test suite under Solaris 2.6 (x86) with the following
results:
=== gcc Summary ===
# of expected passes 7491
# of unexpected failures 16
# of expected failures 7
# of unsupported tests 11
/opt/ehl/egcs-1.1.1-prerelease/gcc/xgcc version egcs-2.91.58 19981101 (egcs-1.1.1 pre-release)
=== g++ Summary ===
# of expected passes 4220
# of unexpected failures 1
# of unexpected successes 1
# of expected failures 85
# of untested testcases 7
/opt/ehl/egcs-1.1.1-prerelease/gcc/testsuite/../xgcc version egcs-2.91.58 19981101 (egcs-1.1.1 pre-release)
=== g77 Summary ===
# of expected passes 415
/opt/ehl/egcs-1.1.1-prerelease/gcc/g77 g77 version egcs-2.91.58 19981101 (egcs-1.1.1 pre-release) (from FSF-g77 version 0.5.24-19980804)
=== libio Summary ===
# of expected passes 40
=== libstdc++ Summary ===
# of expected passes 30
--------------------
To save bandwidth, I put the compressed gcc.log and g++.log files
for this run at
ftp://ftp.via.net/pub/ehl/egcs-1.1.1-pre-sol2x86/gcc.log.gz
ftp://ftp.via.net/pub/ehl/egcs-1.1.1-pre-sol2x86/g++.log.gz
--------------------
Six of the unexpected gcc failures are errors reported by the Solaris
assembler. For example,
Executing on host: /opt/ehl/egcs-1.1.1-prerelease/gcc/xgcc -B/opt/ehl/egcs-1.1.1-prerelease/gcc/ /opt/ehl/egcs-1.1.1-prerelease/gcc/testsuite/gcc.c-torture/compile/calls.c -w -O0 -c -o /opt/ehl/egcs-1.1.1-prerelease/gcc/testsuite/calls.o
spawn /opt/ehl/egcs-1.1.1-prerelease/gcc/xgcc -B/opt/ehl/egcs-1.1.1-prerelease/gcc/ /opt/ehl/egcs-1.1.1-prerelease/gcc/testsuite/gcc.c-torture/compile/calls.c -w -O0 -c -o /opt/ehl/egcs-1.1.1-prerelease/gcc/testsuite/calls.o
Assembler: calls.c
aline 11 : Illegal PC-relative relocation
compiler exited with status 1
output is:
Assembler: calls.c
aline 11 : Illegal PC-relative relocation
FAIL: gcc.c-torture/compile/calls.c, -O0
For reference, the associated assembler for that case is
.file "calls.c"
.version "01.01"
gcc2_compiled.:
.text
.align 4
.globl f1
.type f1,@function
f1:
pushl %ebp
movl %esp,%ebp
call 0
.L1:
leave
ret
.Lfe1:
.size f1,.Lfe1-f1
.align 4
.globl f2
.type f2,@function
f2:
pushl %ebp
movl %esp,%ebp
call 1000
.L2:
leave
ret
.Lfe2:
.size f2,.Lfe2-f2
.align 4
.globl f3
.type f3,@function
f3:
pushl %ebp
movl %esp,%ebp
call 10000000
.L3:
leave
ret
.Lfe3:
.size f3,.Lfe3-f3
.align 4
.globl f4
.type f4,@function
f4:
pushl %ebp
movl %esp,%ebp
pushl %ebx
movl 8(%ebp),%ebx
call *%ebx
.L4:
movl -4(%ebp),%ebx
leave
ret
.Lfe4:
.size f4,.Lfe4-f4
.align 4
.globl f5
.type f5,@function
f5:
pushl %ebp
movl %esp,%ebp
subl $4,%esp
pushl %ebx
movl $f3,-4(%ebp)
movl -4(%ebp),%ebx
call *%ebx
.L5:
movl -8(%ebp),%ebx
leave
ret
.Lfe5:
.size f5,.Lfe5-f5
.ident "GCC: (GNU) egcs-2.91.58 19981101 (egcs-1.1.1 pre-release)"
where the problematic line appears to be
call 0
Please let me know if you'd like me to provide any further
information.
--Elgin
More information about the Gcc-bugs
mailing list