[GDBserver] Fix gdb.threads/siginfo-threads.exp, gdb.base/siginfo-obj.exp regressions.

[GDBserver] Fix gdb.threads/siginfo-threads.exp, gdb.base/siginfo-obj.exp regressions.

This fixes the regressions reported at
<http://sourceware.org/ml/gdb-patches/2013-06/msg00280.html>:

 $ runtest-gdbserver gdb.base/siginfo-obj.exp gdb.base/siginfo-thread.exp gdb.threads/siginfo-threads.exp
 Running ./gdb.base/siginfo-thread.exp ...
 FAIL: gdb.base/siginfo-thread.exp: p ssi_addr
 Running ./gdb.threads/siginfo-threads.exp ...
 FAIL: gdb.threads/siginfo-threads.exp: signal 0 si_pid
 FAIL: gdb.threads/siginfo-threads.exp: signal 1 si_pid
 FAIL: gdb.threads/siginfo-threads.exp: signal 2 si_pid
 FAIL: gdb.threads/siginfo-threads.exp: signal 3 si_pid
 Running ./gdb.base/siginfo-obj.exp ...
 FAIL: gdb.base/siginfo-obj.exp: p ssi_addr
 FAIL: gdb.base/siginfo-obj.exp: p ssi_addr

The multi-arch patch made GDBserver do the the wrong siginfo layout
conversion, because most uses of `linux_is_elf64' were removed, and it
ended up never set.  A global really is the wrong thing to use as
elf64-ness is a per-process property; `linux_is_elf64' was just
accidentally left behind.

Tested on x86_64 Fedora 17.

gdb/gdbserver/
2013-06-12  Pedro Alves  <palves@redhat.com>

	* linux-x86-low.c (linux_is_elf64): Delete global.
	(x86_siginfo_fixup): Replace reference to `linux_is_elf64' global
	with local linux_pid_exe_is_elf_64_file use.