|
From: Philippe W. <phi...@sk...> - 2017-09-19 20:52:04
|
It looks like all ppc32 is broken on gcc110.
I do not know exactly when it was broken, but it was sometime between
valgrind-3.14.0.SVN-16452M-vex-3398 and the current git version
valgrind-3.14.0.GIT-b9df4c8dec-20170916
any ppc32 under valgrind directly asserts with the below assert.
It might be linked with the work on the v2 and/or v3 regalloc
Philippe
./vg-in-place -q ./none/tests/ppc32/test_fx
vex: priv/host_generic_regs.c:114 (RRegUniverse__check_is_sane): Assertion `hregIndex(reg) == i' failed.
vex storage: T total 0 bytes allocated
vex storage: P total 0 bytes allocated
valgrind: the 'impossible' happened:
LibVEX called failure_exit().
host stacktrace:
==44115== at 0x58076B74: show_sched_status_wrk (m_libcassert.c:355)
==44115== by 0x58076CC7: report_and_quit (m_libcassert.c:426)
==44115== by 0x58076F3F: panic (m_libcassert.c:502)
==44115== by 0x58076F3F: vgPlain_core_panic_at (m_libcassert.c:507)
==44115== by 0x58076F73: vgPlain_core_panic (m_libcassert.c:512)
==44115== by 0x58098BDB: failure_exit (m_translate.c:740)
==44115== by 0x5816EBEF: vex_assert_fail (main_util.c:247)
==44115== by 0x581D521B: RRegUniverse__check_is_sane (host_generic_regs.c:114)
==44115== by 0x581B53DF: getRRegUniverse_PPC (host_ppc_defs.c:152)
==44115== by 0x5816DA3B: libvex_BackEnd (main_main.c:895)
==44115== by 0x5816DA3B: LibVEX_Translate (main_main.c:1198)
==44115== by 0x5809B2C7: vgPlain_translate (m_translate.c:1794)
==44115== by 0x580DD7E7: handle_tt_miss (scheduler.c:1056)
==44115== by 0x580DD7E7: vgPlain_scheduler (scheduler.c:1417)
==44115== by 0x580F2A7F: thread_wrapper (syswrap-linux.c:103)
==44115== by 0x580F2A7F: run_a_thread_NORETURN (syswrap-linux.c:156)
sched status:
running_tid=1
Thread 1: status = VgTs_Runnable (lwpid 44115)
==44115== at 0x401B204: _start (in /usr/lib/ld-2.17.so)
|