[PATCH 0/3] AMD64 Displaced Stepping Fix

[Andrew Burgess <aburgess@redhat.com>]

I started looking at PR gdb/22921, which turns out to be a bug in our
displaced stepping implementation.

Patch #1 adds some extra debug output that I found super useful.

Patch #2 is a minor cleanup.

Patch #3 fixes the amd64 bug, but does change the displaced stepping
logic a little (specifically how we fixup after displaced stepping),
so this is what needs to most reviewing.

All feedback welcome.

Thanks,
Andrew

---

Andrew Burgess (3):
  gdb: more debug output for displaced stepping
  gdb: remove gdbarch_displaced_step_fixup_p
  gdb: fix reg corruption from displaced stepping on amd64

 gdb/aarch64-tdep.c                            |  19 ++-
 gdb/aarch64-tdep.h                            |   3 +-
 gdb/amd64-tdep.c                              |  28 ++---
 gdb/amd64-tdep.h                              |   2 +-
 gdb/arm-tdep.c                                |  27 ++++-
 gdb/arm-tdep.h                                |   3 +-
 gdb/displaced-stepping.c                      | 113 ++++++++++++------
 gdb/displaced-stepping.h                      |   3 +
 gdb/gdbarch-components.py                     |  26 ++--
 gdb/gdbarch-gen.h                             |  25 ++--
 gdb/gdbarch.c                                 |  18 +--
 gdb/i386-tdep.c                               |  28 ++---
 gdb/i386-tdep.h                               |   2 +-
 gdb/infrun.c                                  |  35 ------
 gdb/infrun.h                                  |   3 -
 gdb/rs6000-tdep.c                             |  15 ++-
 gdb/s390-tdep.c                               |  16 ++-
 .../gdb.arch/amd64-disp-step-signal.c         |  30 +++++
 gdb/testsuite/gdb.arch/amd64-disp-step.S      |  15 +++
 gdb/testsuite/gdb.arch/amd64-disp-step.exp    | 106 +++++++++++++---
 20 files changed, 348 insertions(+), 169 deletions(-)
 create mode 100644 gdb/testsuite/gdb.arch/amd64-disp-step-signal.c


base-commit: 5e39600a691e3ba76acf6ab94edb24844c2e82b7
-- 
2.25.4