Re: ld test failure on ARM GNU/Linux

[Catherine Moore <clm at cygnus dot com>]

: > Then the testsuite does not match the test case, and selective.exp
: > should be fixed.  And the test will presumably start failing when
: > using gcc 2.95.2 on i386 GNU/Linux.
: 

  This patch looks reasonable to me.  This test (selective 4) seems to have been
  broken from the beginning.  Unfortunately, it passed on other platforms so
  it's possible that we have a --gc-sections bug for some other target.

  Catherine

: Not if you simply remove the test for absense of foo__1B
: Like so.
: 
: --- ld/testsuite/ld-selective/selective.exp~	Sat Dec 11 12:07:22 1999
: +++ ld/testsuite/ld-selective/selective.exp	Tue Apr 11 11:35:18 2000
: @@ -136,20 +136,14 @@ if ![ld_simple_link $ld tmpdir/3.x "$ldf
:      if ![ld_nm $nm tmpdir/3.x] {
:         unresolved $test4
:      } else {
: -       if {[info exists nm_output(foo__1B)]} {
: -          send_log "foo__1B == $nm_output(foo__1B)\n"
: -          verbose "foo__1B == $nm_output(foo__1B)"
: -          fail $test4
: -       } else {
: -           if {[ info exists nm_output(bar__1A)]} {
: -              send_log "bar__1A== $nm_output(_bar__1A)\n"
: -              verbose "bar__1A == $nm_output(_bar__1A)"
: -              fail $test4
: -           } else {
: -               pass $test4
: -             }
: -          }
: -      }
: +	if {[ info exists nm_output(bar__1A)]} {
: +	    send_log "bar__1A== $nm_output(_bar__1A)\n"
: +	    verbose "bar__1A == $nm_output(_bar__1A)"
: +	    fail $test4
: +	} else {
: +	    pass $test4
: +	}
: +    }
:  }
:  
:  if { ![ld_compile "$CC $cflags $cxxflags" $srcdir/$subdir/4.cc tmpdir/4.o]} {
: