Fix signal handlers regression (RH BZ 1542149, Pedro Alves).

Add %patch directive to the commit below. (and bump version)
Applying the patch for last commit.
2018-02-06 22:41:37 +01:00 · 2018-01-25 14:33:21 -05:00 · 2018-01-25 14:28:45 -05:00 · 2018-01-24 15:00:35 -05:00 · 2017-12-06 21:09:32 +01:00 · 2017-12-06 12:45:27 +01:00
5 changed files with 1037 additions and 3 deletions
--- a/gdb-random-libstdcpp-prettyprinters-fail.patch
+++ b/gdb-random-libstdcpp-prettyprinters-fail.patch
@ -0,0 +1,229 @@
+From: Pedro Alves <palves at redhat dot com>
+To: gdb-patches at sourceware dot org
+Subject: [PATCH] Fix GCC PR83906 - [8 Regression] Random FAIL: libstdc++-prettyprinters/80276.cc whatis p4
+Date: Wed, 24 Jan 2018 17:27:22 +0000
+Message-Id: <20180124172722.31553-1-palves@redhat.com>
+
+GCC PR83906 [1] is about a GCC/libstdc++ GDB/Python type printer
+testcase failing randomly, as shown by running (in libstdc++'s
+testsuite):
+
+ make check RUNTESTFLAGS=prettyprinters.exp=80276.cc
+
+in a loop.  Sometimes you get this:
+
+ FAIL: libstdc++-prettyprinters/80276.cc whatis p4
+
+I.e., this:
+ type = std::unique_ptr<std::vector<std::unique_ptr<std::list<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >>[]>>[99]>
+
+instead of this:
+ type = std::unique_ptr<std::vector<std::unique_ptr<std::list<std::string>[]>>[99]>
+
+Jonathan Wakely tracked it on the printer side to this bit in
+libstdc++'s type printer:
+
+            if self.type_obj == type_obj:
+                return strip_inline_namespaces(self.name)
+
+This assumes the two types resolve to the same gdb.Type but some times
+the comparison unexpectedly fails.
+
+Running the testcase manually under Valgrind finds the problem in GDB:
+
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ ==6118== Conditional jump or move depends on uninitialised value(s)
+ ==6118==    at 0x4C35CB0: bcmp (vg_replace_strmem.c:1100)
+ ==6118==    by 0x6F773A: check_types_equal(type*, type*, VEC_type_equality_entry_d**) (gdbtypes.c:3515)
+ ==6118==    by 0x6F7B00: check_types_worklist(VEC_type_equality_entry_d**, bcache*) (gdbtypes.c:3618)
+ ==6118==    by 0x6F7C03: types_deeply_equal(type*, type*) (gdbtypes.c:3655)
+ ==6118==    by 0x4D5B06: typy_richcompare(_object*, _object*, int) (py-type.c:1007)
+ ==6118==    by 0x63D7E6C: PyObject_RichCompare (object.c:961)
+ ==6118==    by 0x646EAEC: PyEval_EvalFrameEx (ceval.c:4960)
+ ==6118==    by 0x646DC08: PyEval_EvalFrameEx (ceval.c:4519)
+ ==6118==    by 0x646DC08: PyEval_EvalFrameEx (ceval.c:4519)
+ ==6118==    by 0x646DC08: PyEval_EvalFrameEx (ceval.c:4519)
+ ==6118==    by 0x646DC08: PyEval_EvalFrameEx (ceval.c:4519)
+ ==6118==    by 0x646DC08: PyEval_EvalFrameEx (ceval.c:4519)
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+That "bcmp" call is really a memcmp call in check_types_equal.  The
+problem is that gdb is memcmp'ing two objects that are equal in value:
+
+ (top-gdb) p *TYPE_RANGE_DATA (type1)
+ $1 = {low = {kind = PROP_CONST, data = {const_val = 0, baton = 0x0}}, high = {kind = PROP_CONST, data = {const_val = 15, baton = 0xf}}, flag_upper_bound_is_count = 0,
+   flag_bound_evaluated = 0}
+ (top-gdb) p *TYPE_RANGE_DATA (type2)
+ $2 = {low = {kind = PROP_CONST, data = {const_val = 0, baton = 0x0}}, high = {kind = PROP_CONST, data = {const_val = 15, baton = 0xf}}, flag_upper_bound_is_count = 0,
+   flag_bound_evaluated = 0}
+
+but differ in padding.  Notice the 4-byte hole:
+
+  (top-gdb) ptype /o range_bounds
+  /* offset    |  size */  type = struct range_bounds {
+  /*    0      |    16 */    struct dynamic_prop {
+  /*    0      |     4 */        dynamic_prop_kind kind;
+  /* XXX  4-byte hole  */
+  /*    8      |     8 */        union dynamic_prop_data {
+  /*                 8 */            LONGEST const_val;
+  /*                 8 */            void *baton;
+
+				     /* total size (bytes):    8 */
+				 } data;
+
+which is filled with garbage:
+
+  (top-gdb) x /40bx TYPE_RANGE_DATA (type1)
+  0x2fa7ea0:      0x01    0x00    0x00    0x00    0x43    0x01    0x00    0x00
+						  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+  0x2fa7ea8:      0x00    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+  0x2fa7eb0:      0x01    0x00    0x00    0x00    0xfe    0x7f    0x00    0x00
+  0x2fa7eb8:      0x0f    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+  0x2fa7ec0:      0x00    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+  (top-gdb) x /40bx TYPE_RANGE_DATA (type2)
+  0x20379b0:      0x01    0x00    0x00    0x00    0xfe    0x7f    0x00    0x00
+						  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+  0x20379b8:      0x00    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+  0x20379c0:      0x01    0x00    0x00    0x00    0xfe    0x7f    0x00    0x00
+  0x20379c8:      0x0f    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+  0x20379d0:      0x00    0x00    0x00    0x00    0x00    0x00    0x00    0x00
+
+  (top-gdb) p memcmp (TYPE_RANGE_DATA (type1), TYPE_RANGE_DATA (type2), sizeof (*TYPE_RANGE_DATA (type1)))
+  $3 = -187
+
+In some cases objects of type range_bounds are memset when allocated,
+but then their dynamic_prop low/high fields are copied over from some
+template dynamic_prop object that wasn't memset.  E.g.,
+create_static_range_type's low/high locals are left with garbage in
+the padding, and then that padding is copied over to the range_bounds
+object's low/high fields.
+
+At first, I considered making sure to always memset range_bounds
+objects, thinking that maybe type objects are being put in some bcache
+instance somewhere.  But then I hacked bcache/bcache_full to poison
+non-pod types, and made dynamic_prop a non-pod, and GDB still
+compiled.
+
+So given that, it seems safest to not assume padding will always be
+memset, and instead treat them as regular value types, implementing
+(in)equality operators and using those instead of memcmp.
+
+This fixes the random FAILs in GCC's testcase.
+
+[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83906
+
+gdb/ChangeLog:
+2018-01-24  Pedro Alves  <palves@redhat.com>
+
+	GCC PR libstdc++/83906
+	* gdbtypes.c (operator==(const dynamic_prop &,
+	const dynamic_prop &)): New.
+	(operator==(const range_bounds &, const range_bounds &)): New.
+	(check_types_equal): Use them instead of memcmp.
+	* gdbtypes.h (operator==(const dynamic_prop &,
+	const dynamic_prop &)): Declare.
+	(operator!=(const dynamic_prop &, const dynamic_prop &)): Declare.
+	(operator==(const range_bounds &, const range_bounds &)): Declare.
+	(operator!=(const range_bounds &, const range_bounds &)): Declare.
+---
+ gdb/gdbtypes.c | 41 +++++++++++++++++++++++++++++++++++++++--
+ gdb/gdbtypes.h | 20 ++++++++++++++++++++
+ 2 files changed, 59 insertions(+), 2 deletions(-)
+
+Index: gdb-8.0.1/gdb/gdbtypes.c
+===================================================================
+--- gdb-8.0.1.orig/gdb/gdbtypes.c
+++ gdb-8.0.1/gdb/gdbtypes.c
+@@ -855,6 +855,44 @@ allocate_stub_method (struct type *type)
+   return mtype;
+ }
+ 
+/* See gdbtypes.h.  */
+
+bool
+operator== (const dynamic_prop &l, const dynamic_prop &r)
+{
+  if (l.kind != r.kind)
+    return false;
+
+  switch (l.kind)
+    {
+    case PROP_UNDEFINED:
+      return true;
+    case PROP_CONST:
+      return l.data.const_val == r.data.const_val;
+    case PROP_ADDR_OFFSET:
+    case PROP_LOCEXPR:
+    case PROP_LOCLIST:
+      return l.data.baton == r.data.baton;
+    }
+
+  gdb_assert_not_reached ("unhandled dynamic_prop kind");
+}
+
+/* See gdbtypes.h.  */
+
+bool
+operator== (const range_bounds &l, const range_bounds &r)
+{
+#define FIELD_EQ(FIELD) (l.FIELD == r.FIELD)
+
+  return (FIELD_EQ (low)
+	  && FIELD_EQ (high)
+	  && FIELD_EQ (flag_upper_bound_is_count)
+	  && FIELD_EQ (flag_bound_evaluated));
+
+#undef FIELD_EQ
+}
+
+ /* Create a range type with a dynamic range from LOW_BOUND to
+    HIGH_BOUND, inclusive.  See create_range_type for further details. */
+ 
+@@ -3466,8 +3504,7 @@ check_types_equal (struct type *type1, s
+ 
+   if (TYPE_CODE (type1) == TYPE_CODE_RANGE)
+     {
+-      if (memcmp (TYPE_RANGE_DATA (type1), TYPE_RANGE_DATA (type2),
+-		  sizeof (*TYPE_RANGE_DATA (type1))) != 0)
+      if (*TYPE_RANGE_DATA (type1) != *TYPE_RANGE_DATA (type2))
+ 	return 0;
+     }
+   else
+Index: gdb-8.0.1/gdb/gdbtypes.h
+===================================================================
+--- gdb-8.0.1.orig/gdb/gdbtypes.h
+++ gdb-8.0.1/gdb/gdbtypes.h
+@@ -408,6 +408,16 @@ struct dynamic_prop
+   union dynamic_prop_data data;
+ };
+ 
+/* Compare two dynamic_prop objects for equality.  dynamic_prop
+   instances are equal iff they have the same type and storage.  */
+extern bool operator== (const dynamic_prop &l, const dynamic_prop &r);
+
+/* Compare two dynamic_prop objects for inequality.  */
+static inline bool operator!= (const dynamic_prop &l, const dynamic_prop &r)
+{
+  return !(l == r);
+}
+
+ /* * Define a type's dynamic property node kind.  */
+ enum dynamic_prop_node_kind
+ {
+@@ -568,6 +578,16 @@ struct range_bounds
+   int flag_bound_evaluated : 1;
+ };
+ 
+/* Compare two range_bounds objects for equality.  Simply does
+   memberwise comparison.  */
+extern bool operator== (const range_bounds &l, const range_bounds &r);
+
+/* Compare two range_bounds objects for inequality.  */
+static inline bool operator!= (const range_bounds &l, const range_bounds &r)
+{
+  return !(l == r);
+}
+
+ union type_specific
+ {
+   /* * CPLUS_STUFF is for TYPE_CODE_STRUCT.  It is initialized to
--- a/gdb-rhbz1228556-bpt-inlined-func-name-1of2.patch
+++ b/gdb-rhbz1228556-bpt-inlined-func-name-1of2.patch
@ -130,7 +130,7 @@ Index: gdb-8.0.1/gdb/symtab.h
   struct program_space *pspace;
 
   struct symtab *symtab;
-+  struct symbol *symbol = NULL;
+  struct symbol *symbol;
   struct obj_section *section;
   /* Line number.  Line numbers start at 1 and proceed through symtab->nlines.
      0 is never a valid line number; it is used to indicate that line number
--- a/gdb-rhbz1542149-spawn-default-signal-handlers-regression.patch
+++ b/gdb-rhbz1542149-spawn-default-signal-handlers-regression.patch
@ -0,0 +1,320 @@
+commit e379cee61f3890e535e995828e8846b020ef2a32
+Author: Pedro Alves <palves@redhat.com>
+Date:   Fri Jan 5 18:26:18 2018 +0000
+
+    Fix regression: cannot start with LD_PRELOAD=libSegFault.so (PR gdb/18653#c7)
+    
+    At https://sourceware.org/bugzilla/show_bug.cgi?id=18653#c7, Andrew
+    reports that the fix for PR gdb/18653 made GDB useless if you preload
+    libSegFault.so, because GDB internal-errors on startup:
+    
+     $ LD_PRELOAD=libSegFault.so gdb
+     src/gdb/common/signals-state-save-restore.c:64: internal-error: unexpected signal handler
+     A problem internal to GDB has been detected,
+     further debugging may prove unreliable.
+     Aborted (core dumped)
+     $
+    
+    The internal error comes from the code saving the signal dispositions
+    inherited from gdb's parent:
+    
+     (top-gdb) bt
+     #0  0x000000000056b001 in internal_error(char const*, int, char const*, ...) (file=0xaf5f38 "src/gdb/common/signals-state-save-restore.c", line=64, fmt=0xaf5f18 "unexpected signal handler") at src/gdb/common/errors.c:54
+     #1  0x00000000005752c9 in save_original_signals_state() () at src/gdb/common/signals-state-save-restore.c:64
+     #2  0x00000000007425de in captured_main_1(captured_main_args*) (context=0x7fffffffd860)
+         at src/gdb/main.c:509
+     #3  0x0000000000743622 in captured_main(void*) (data=0x7fffffffd860) at src/gdb/main.c:1145
+     During symbol reading, cannot get low and high bounds for subprogram DIE at 24065.
+     #4  0x00000000007436f9 in gdb_main(captured_main_args*) (args=0x7fffffffd860) at src/gdb/main.c:1171
+     #5  0x0000000000413acd in main(int, char**) (argc=1, argv=0x7fffffffd968) at src/gdb/gdb.c:32
+    
+    This commit downgrades the internal error to a warning.  You'll get
+    instead:
+    
+    ~~~
+     $ LD_PRELOAD=libSegFault.so gdb
+     warning: Found custom handler for signal 11 (Segmentation fault) preinstalled.
+     Some signal dispositions inherited from the environment (SIG_DFL/SIG_IGN)
+     won't be propagated to spawned programs.
+     GNU gdb (GDB) 8.0.50.20171213-git
+     Copyright (C) 2017 Free Software Foundation, Inc.
+     License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
+     This is free software: you are free to change and redistribute it.
+     There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
+     and "show warranty" for details.
+     This GDB was configured as "x86_64-pc-linux-gnu".
+     Type "show configuration" for configuration details.
+     For bug reporting instructions, please see:
+     <http://www.gnu.org/software/gdb/bugs/>.
+     Find the GDB manual and other documentation resources online at:
+     <http://www.gnu.org/software/gdb/documentation/>.
+     For help, type "help".
+     Type "apropos word" to search for commands related to "word"...
+     (gdb)
+    ~~~
+    
+    This also moves the location where save_original_signals_state is
+    called a bit further below (to after option processing), so that "-q"
+    disables the warning:
+    
+    ~~~
+     $ LD_PRELOAD=libSegFault.so gdb -q
+     (gdb)
+    ~~~
+    
+    New testcase included.
+    
+    gdb/ChangeLog:
+    2018-01-05  Pedro Alves  <palves@redhat.com>
+    
+            PR gdb/18653
+            * common/signals-state-save-restore.c
+            (save_original_signals_state): New parameter 'quiet'.  Warn if we
+            find a custom handler preinstalled, instead of internal erroring.
+            But only warn if !quiet.
+            * common/signals-state-save-restore.h
+            (save_original_signals_state): New parameter 'quiet'.
+            * main.c (captured_main_1): Move save_original_signals_state call
+            after option handling, and pass QUIET.
+    
+    gdb/gdbserver/ChangeLog:
+    2018-01-05  Pedro Alves  <palves@redhat.com>
+    
+            PR gdb/18653
+            * server.c (captured_main): Pass quiet=false to
+            save_original_signals_state.
+    
+    gdb/testsuite/ChangeLog:
+    2018-01-05  Pedro Alves  <palves@redhat.com>
+    
+            PR gdb/18653
+            * gdb.base/libsegfault.exp: New.
+
+### a/gdb/ChangeLog
+### b/gdb/ChangeLog
+## -1,3 +1,15 @@
+2018-01-05  Pedro Alves  <palves@redhat.com>
+
+	PR gdb/18653
+	* common/signals-state-save-restore.c
+	(save_original_signals_state): New parameter 'quiet'.  Warn if we
+	find a custom handler preinstalled, instead of internal erroring.
+	But only warn if !quiet.
+	* common/signals-state-save-restore.h
+	(save_original_signals_state): New parameter 'quiet'.
+	* main.c (captured_main_1): Move save_original_signals_state call
+	after option handling, and pass QUIET.
+
+ 2018-01-05  Pedro Alves  <palves@redhat.com>
+ 
+ 	* spu-tdep.c (spu_catch_start): Pass
+--- a/gdb/common/signals-state-save-restore.c
+++ b/gdb/common/signals-state-save-restore.c
+@@ -35,7 +35,7 @@ static sigset_t original_signal_mask;
+ /* See signals-state-save-restore.h.   */
+ 
+ void
+-save_original_signals_state (void)
+save_original_signals_state (bool quiet)
+ {
+ #ifdef HAVE_SIGACTION
+   int i;
+@@ -45,6 +45,8 @@ save_original_signals_state (void)
+   if (res == -1)
+     perror_with_name (("sigprocmask"));
+ 
+  bool found_preinstalled = false;
+
+   for (i = 1; i < NSIG; i++)
+     {
+       struct sigaction *oldact = &original_signal_actions[i];
+@@ -59,9 +61,31 @@ save_original_signals_state (void)
+ 	perror_with_name (("sigaction"));
+ 
+       /* If we find a custom signal handler already installed, then
+-	 this function was called too late.  */
+-      if (oldact->sa_handler != SIG_DFL && oldact->sa_handler != SIG_IGN)
+-	internal_error (__FILE__, __LINE__, _("unexpected signal handler"));
+	 this function was called too late.  This is a warning instead
+	 of an internal error because this can also happen if you
+	 LD_PRELOAD a library that installs a signal handler early via
+	 __attribute__((constructor)), like libSegFault.so.  */
+      if (!quiet
+	  && oldact->sa_handler != SIG_DFL
+	  && oldact->sa_handler != SIG_IGN)
+	{
+	  found_preinstalled = true;
+
+	  /* Use raw fprintf here because we're being called in early
+	     startup, because GDB's filtered streams are are
+	     created.  */
+	  fprintf (stderr,
+		   _("warning: Found custom handler for signal "
+		     "%d (%s) preinstalled.\n"), i,
+		   strsignal (i));
+	}
+    }
+
+  if (found_preinstalled)
+    {
+      fprintf (stderr, _("\
+Some signal dispositions inherited from the environment (SIG_DFL/SIG_IGN)\n\
+won't be propagated to spawned programs.\n"));
+     }
+ #endif
+ }
+--- a/gdb/common/signals-state-save-restore.h
+++ b/gdb/common/signals-state-save-restore.h
+@@ -28,9 +28,10 @@
+    back to what was originally inherited from gdb/gdbserver's parent,
+    just before execing the target program to debug.  */
+ 
+-/* Save the signal state of all signals.  */
+/* Save the signal state of all signals.  If !QUIET, warn if we detect
+   a custom signal handler preinstalled.  */
+ 
+-extern void save_original_signals_state (void);
+extern void save_original_signals_state (bool quiet);
+ 
+ /* Restore the signal state of all signals.  */
+ 
+### a/gdb/gdbserver/ChangeLog
+### b/gdb/gdbserver/ChangeLog
+## -1,3 +1,9 @@
+2018-01-05  Pedro Alves  <palves@redhat.com>
+
+	PR gdb/18653
+	* server.c (captured_main): Pass quiet=false to
+	save_original_signals_state.
+
+ 2018-01-01  Joel Brobecker  <brobecker@adacore.com>
+ 
+ 	* gdbreplay.c (gdbreplay_version): Update copyright year in
+--- a/gdb/gdbserver/server.c
+++ b/gdb/gdbserver/server.c
+@@ -3712,7 +3712,7 @@ captured_main (int argc, char *argv[])
+      opened by remote_prepare.  */
+   notice_open_fds ();
+ 
+-  save_original_signals_state ();
+  save_original_signals_state (false);
+ 
+   /* We need to know whether the remote connection is stdio before
+      starting the inferior.  Inferiors created in this scenario have
+--- a/gdb/main.c
+++ b/gdb/main.c
+@@ -506,7 +506,6 @@ captured_main_1 (struct captured_main_args *context)
+ 
+   bfd_init ();
+   notice_open_fds ();
+-  save_original_signals_state ();
+ 
+   saved_command_line = (char *) xstrdup ("");
+ 
+@@ -849,6 +848,8 @@ captured_main_1 (struct captured_main_args *context)
+       quiet = 1;
+   }
+ 
+  save_original_signals_state (quiet);
+
+   /* Try to set up an alternate signal stack for SIGSEGV handlers.  */
+   setup_alternate_signal_stack ();
+ 
+### a/gdb/testsuite/ChangeLog
+### b/gdb/testsuite/ChangeLog
+## -1,3 +1,8 @@
+2018-01-05  Pedro Alves  <palves@redhat.com>
+
+	PR gdb/18653
+	* gdb.base/libsegfault.exp: New.
+
+ 2018-01-05  Joel Brobecker  <brobecker@adacore.com>
+ 
+ 	PR gdb/22670
+--- /dev/null
+++ b/gdb/testsuite/gdb.base/libsegfault.exp
+@@ -0,0 +1,84 @@
+# Copyright 2017-2018 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+# This file is part of the gdb testsuite.
+
+# Test that GDB tolerates being started with libSegFault.so preloaded
+# with LD_PRELOAD, and that GDB warns about a custom SIGSEGV custom
+# handler.  See PR gdb/18653
+# <https://sourceware.org/bugzilla/show_bug.cgi?id=18653#c7>.
+
+# We cannot expect remote hosts to see environment variables set on
+# the local machine.
+if { [is_remote host] } {
+    unsupported "can't set environment variables on remote host"
+    return -1
+}
+
+# Spawn GDB with LIB preloaded with LD_PRELOAD.  CMDLINE_OPTS are
+# command line options passed to GDB.
+
+proc gdb_spawn_with_ld_preload {lib cmdline_opts} {
+    global env
+
+    save_vars { env(LD_PRELOAD) } {
+	if { ![info exists env(LD_PRELOAD) ]
+	     || $env(LD_PRELOAD) == "" } {
+	    set env(LD_PRELOAD) "$lib"
+	} else {
+	    append env(LD_PRELOAD) ":$lib"
+	}
+
+	gdb_spawn_with_cmdline_opts $cmdline_opts
+    }
+}
+
+proc test_libsegfault {} {
+    global gdb_prompt
+
+    set libsegfault "libSegFault.so"
+
+    # When started normally, if libSegFault.so is preloaded, GDB
+    # should warn about not being able to propagate the signal
+    # disposition of SIGSEGV.
+    gdb_exit
+    gdb_spawn_with_ld_preload $libsegfault ""
+
+    set test "gdb emits custom handler warning"
+    gdb_test_multiple "" $test {
+	-re "cannot be preloaded.*\r\n$gdb_prompt $" {
+	    # Glibc 2.22 outputs:
+	    # ERROR: ld.so: object 'libSegFault.so' from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.
+	    untested "cannot preload libSegFault.so"
+	    return
+	}
+	-re "Found custom handler.*won't be propagated.*\r\n$gdb_prompt $" {
+	    pass $test
+	}
+    }
+
+    # "-q" should disable the warning, though.
+    gdb_exit
+    gdb_spawn_with_ld_preload $libsegfault "-q"
+
+    set test "quiet suppresses custom handler warning"
+    gdb_test_multiple "" $test {
+	-re "^$gdb_prompt $" {
+	    pass $test
+	}
+    }
+}
+
+test_libsegfault
--- a/gdb-upstream.patch
+++ b/gdb-upstream.patch
@ -74,3 +74,461 @@ Date:   Wed Aug 9 05:01:55 2017 -0700
 	  || attr->form == DW_FORM_GNU_strp_alt)
 	str = DW_STRING (attr);
       else
+
+
+
+commit f24b864960e61f9a91f8c168c1afe12a6676ad7a
+Author: Walfred Tedeschi <walfred.tedeschi@intel.com>
+Date:   Mon Oct 16 08:59:38 2017 +0200
+
+    PR22137: gdbserver crashes on host with pkru register.
+    
+    This patch adds missing backslash on a makefile and regenerate the
+    files created via the xml files.  Those were not in sync with the xml file.
+    
+    gdb/ChangeLog:
+    
+    2017-10-16  Walfred Tedeschi  <walfred.tedeschi@intel.com>
+    
+            * features/Makefile (i386-avx-mpx-avx512-pku.dat): Add backslash
+            at the end of the line.
+            * regformats/i386/amd64-avx-mpx-avx512-pku-linux.dat: Regenerate.
+            * regformats/i386/amd64-avx-mpx-avx512-pku.dat: Regenerate.
+
+### a/gdb/ChangeLog
+### b/gdb/ChangeLog
+## -1,3 +1,10 @@
+2017-10-16  Walfred Tedeschi  <walfred.tedeschi@intel.com>
+
+	* features/Makefile (i386-avx-mpx-avx512-pku.dat): Add backslash 
+	at the end of the line.
+	* regformats/i386/amd64-avx-mpx-avx512-pku-linux.dat: Regenerate.
+	* regformats/i386/amd64-avx-mpx-avx512-pku.dat: Regenerate.
+
+ 2017-09-07  Joel Brobecker  <brobecker@adacore.com>
+ 
+ 	* version.in: Set GDB version number to 8.0.1.DATE-git.
+--- a/gdb/features/Makefile
+++ b/gdb/features/Makefile
+@@ -281,7 +281,7 @@ $(outdir)/i386/i386-avx-avx512.dat: i386/32bit-core.xml i386/32bit-avx.xml \
+ 			       i386/32bit-avx512.xml
+ $(outdir)/i386/i386-avx-avx512-linux.dat: i386/32bit-core.xml i386/32bit-avx.xml \
+ 			       i386/32bit-linux.xml i386/32bit-avx512.xml
+-$(outdir)/i386/i386-avx-mpx-avx512-pku.dat: i386/32bit-core.xml
+$(outdir)/i386/i386-avx-mpx-avx512-pku.dat: i386/32bit-core.xml \
+ 			       i386/32bit-avx.xml i386/32bit-mpx.xml i386/32bit-avx512.xml \
+ 			       i386/32bit-pkeys.xml
+ $(outdir)/i386/i386-avx-mpx-avx512-pku-linux.dat: i386/32bit-core.xml \
+--- a/gdb/regformats/i386/amd64-avx-mpx-avx512-pku-linux.dat
+++ b/gdb/regformats/i386/amd64-avx-mpx-avx512-pku-linux.dat
+@@ -157,3 +157,4 @@ expedite:rbp,rsp,rip
+ 256:zmm29h
+ 256:zmm30h
+ 256:zmm31h
+32:pkru
+--- a/gdb/regformats/i386/amd64-avx-mpx-avx512-pku.dat
+++ b/gdb/regformats/i386/amd64-avx-mpx-avx512-pku.dat
+@@ -60,9 +60,6 @@ expedite:rbp,rsp,rip
+ 128:xmm14
+ 128:xmm15
+ 32:mxcsr
+-64:orig_rax
+-64:fs_base
+-64:gs_base
+ 128:ymm0h
+ 128:ymm1h
+ 128:ymm2h
+
+
+
+commit 50a1fdd59c1777672a9be0e81fe2301c2a115fce
+Author: Pedro Alves <palves@redhat.com>
+Date:   Mon Dec 4 15:59:20 2017 +0000
+
+    Fix displaced-stepping RIP-relative VEX-encoded instructions (AVX) (PR gdb/22499)
+    
+    PR gdb/22499 is about a latent bug exposed by the switch to "maint set
+    target-non-stop on" by default on x86-64 GNU/Linux, a while ago.  With
+    that on, GDB is also preferring to use displaced-stepping by default.
+    
+    The testcase in the bug is failing because GDB ends up incorrectly
+    displaced-stepping over a RIP-relative VEX-encoded instruction, like
+    this:
+    
+     0x00000000004007f5 <+15>:    c5 fb 10 05 8b 01 00 00 vmovsd 0x18b(%rip),%xmm0        # 0x400988
+    
+    While RIP-relative instructions need adjustment when relocated to the
+    scratch pad, GDB ends up just copying VEX-encoded instructions to the
+    scratch pad unmodified, with the end result that the inferior ends up
+    executing an instruction that fetches/writes memory from the wrong
+    address...
+    
+    This patch teaches GDB about the VEX-encoding prefixes, fixing the
+    problem, and adds a testcase that fails without the GDB fix.
+    
+    I think we may need a similar treatment for EVEX-encoded instructions,
+    but I didn't address that simply because I couldn't find any
+    EVEX-encoded RIP-relative instruction in the gas testsuite.  In any
+    case, this commit is forward progress as-is already.
+    
+    gdb/ChangeLog:
+    2017-12-04  Pedro Alves  <palves@redhat.com>
+    
+            PR gdb/22499
+            * amd64-tdep.c (amd64_insn::rex_offset): Rename to...
+            (amd64_insn::enc_prefix_offset): ... this, and tweak comment.
+            (vex2_prefix_p, vex3_prefix_p): New functions.
+            (amd64_get_insn_details): Adjust to rename.  Also skip VEX2 and
+            VEX3 prefixes.
+            (fixup_riprel): Set VEX3.!B.
+    
+    gdb/testsuite/ChangeLog:
+    2017-12-04  Pedro Alves  <palves@redhat.com>
+    
+            PR gdb/22499
+            * gdb.arch/amd64-disp-step-avx.S: New file.
+            * gdb.arch/amd64-disp-step-avx.exp: New file.
+
+### a/gdb/ChangeLog
+### b/gdb/ChangeLog
+## -1,3 +1,13 @@
+2017-12-04  Pedro Alves  <palves@redhat.com>
+
+	PR gdb/22499
+	* amd64-tdep.c (amd64_insn::rex_offset): Rename to...
+	(amd64_insn::enc_prefix_offset): ... this, and tweak comment.
+	(vex2_prefix_p, vex3_prefix_p): New functions.
+	(amd64_get_insn_details): Adjust to rename.  Also skip VEX2 and
+	VEX3 prefixes.
+	(fixup_riprel): Set VEX3.!B.
+
+ 2017-12-03  Simon Marchi  <simon.marchi@ericsson.com>
+ 
+ 	* target.h (mem_region_vector): Remove.
+--- a/gdb/amd64-tdep.c
+++ b/gdb/amd64-tdep.c
+@@ -1037,8 +1037,9 @@ struct amd64_insn
+ {
+   /* The number of opcode bytes.  */
+   int opcode_len;
+-  /* The offset of the rex prefix or -1 if not present.  */
+-  int rex_offset;
+  /* The offset of the REX/VEX instruction encoding prefix or -1 if
+     not present.  */
+  int enc_prefix_offset;
+   /* The offset to the first opcode byte.  */
+   int opcode_offset;
+   /* The offset to the modrm byte or -1 if not present.  */
+@@ -1124,6 +1125,22 @@ rex_prefix_p (gdb_byte pfx)
+   return REX_PREFIX_P (pfx);
+ }
+ 
+/* True if PFX is the start of the 2-byte VEX prefix.  */
+
+static bool
+vex2_prefix_p (gdb_byte pfx)
+{
+  return pfx == 0xc5;
+}
+
+/* True if PFX is the start of the 3-byte VEX prefix.  */
+
+static bool
+vex3_prefix_p (gdb_byte pfx)
+{
+  return pfx == 0xc4;
+}
+
+ /* Skip the legacy instruction prefixes in INSN.
+    We assume INSN is properly sentineled so we don't have to worry
+    about falling off the end of the buffer.  */
+@@ -1242,19 +1259,30 @@ amd64_get_insn_details (gdb_byte *insn, struct amd64_insn *details)
+   details->raw_insn = insn;
+ 
+   details->opcode_len = -1;
+-  details->rex_offset = -1;
+  details->enc_prefix_offset = -1;
+   details->opcode_offset = -1;
+   details->modrm_offset = -1;
+ 
+   /* Skip legacy instruction prefixes.  */
+   insn = amd64_skip_prefixes (insn);
+ 
+-  /* Skip REX instruction prefix.  */
+  /* Skip REX/VEX instruction encoding prefixes.  */
+   if (rex_prefix_p (*insn))
+     {
+-      details->rex_offset = insn - start;
+      details->enc_prefix_offset = insn - start;
+       ++insn;
+     }
+  else if (vex2_prefix_p (*insn))
+    {
+      /* Don't record the offset in this case because this prefix has
+	 no REX.B equivalent.  */
+      insn += 2;
+    }
+  else if (vex3_prefix_p (*insn))
+    {
+      details->enc_prefix_offset = insn - start;
+      insn += 3;
+    }
+ 
+   details->opcode_offset = insn - start;
+ 
+@@ -1329,10 +1357,22 @@ fixup_riprel (struct gdbarch *gdbarch, amd64_displaced_step_closure *dsc,
+   arch_tmp_regno = amd64_get_unused_input_int_reg (insn_details);
+   tmp_regno = amd64_arch_reg_to_regnum (arch_tmp_regno);
+ 
+-  /* REX.B should be unset as we were using rip-relative addressing,
+-     but ensure it's unset anyway, tmp_regno is not r8-r15.  */
+-  if (insn_details->rex_offset != -1)
+-    dsc->insn_buf[insn_details->rex_offset] &= ~REX_B;
+  /* Position of the not-B bit in the 3-byte VEX prefix (in byte 1).  */
+  static constexpr gdb_byte VEX3_NOT_B = 0x20;
+
+  /* REX.B should be unset (VEX.!B set) as we were using rip-relative
+     addressing, but ensure it's unset (set for VEX) anyway, tmp_regno
+     is not r8-r15.  */
+  if (insn_details->enc_prefix_offset != -1)
+    {
+      gdb_byte *pfx = &dsc->insn_buf[insn_details->enc_prefix_offset];
+      if (rex_prefix_p (pfx[0]))
+	pfx[0] &= ~REX_B;
+      else if (vex3_prefix_p (pfx[0]))
+	pfx[1] |= VEX3_NOT_B;
+      else
+	gdb_assert_not_reached ("unhandled prefix");
+    }
+ 
+   regcache_cooked_read_unsigned (regs, tmp_regno, &orig_value);
+   dsc->tmp_regno = tmp_regno;
+### a/gdb/testsuite/ChangeLog
+### b/gdb/testsuite/ChangeLog
+## -1,3 +1,9 @@
+2017-12-04  Pedro Alves  <palves@redhat.com>
+
+	PR gdb/22499
+	* gdb.arch/amd64-disp-step-avx.S: New file.
+	* gdb.arch/amd64-disp-step-avx.exp: New file.
+
+ 2017-12-03  Pedro Alves  <palves@redhat.com>
+ 
+ 	* gdb.threads/process-dies-while-detaching.c: Include <errno.h>
+--- /dev/null
+++ b/gdb/testsuite/gdb.arch/amd64-disp-step-avx.S
+@@ -0,0 +1,70 @@
+/* Copyright 2009-2017 Free Software Foundation, Inc.
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+   This file is part of the gdb testsuite.
+
+   Test displaced stepping over VEX-encoded RIP-relative AVX
+   instructions.  */
+
+	.text
+
+	.global main
+main:
+	nop
+
+/***********************************************/
+
+/* Test a VEX2-encoded RIP-relative instruction.  */
+
+	.global test_rip_vex2
+test_rip_vex2:
+	vmovsd ro_var(%rip),%xmm0
+	.global test_rip_vex2
+test_rip_vex2_end:
+	nop
+
+/* Test a VEX3-encoded RIP-relative instruction.  */
+
+	.global test_rip_vex3
+test_rip_vex3:
+	vextractf128 $0x0,%ymm0,var128(%rip)
+	.global test_rip_vex3
+test_rip_vex3_end:
+	nop
+
+	/* skip over test data */
+	jmp done
+
+/* RIP-relative ro-data for VEX2 test above.  */
+
+ro_var:
+	.8byte 0x1122334455667788
+	.8byte 0x8877665544332211
+
+/***********************************************/
+
+/* All done.  */
+
+done:
+	mov $0,%rdi
+	call exit
+	hlt
+
+/* RIP-relative data for VEX3 test above.  */
+
+.data
+var128:
+	.8byte 0xaa55aa55aa55aa55
+	.8byte 0x55aa55aa55aa55aa
+--- /dev/null
+++ b/gdb/testsuite/gdb.arch/amd64-disp-step-avx.exp
+@@ -0,0 +1,141 @@
+# Copyright 2009-2017 Free Software Foundation, Inc.
+
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+# This file is part of the gdb testsuite.
+
+# Test displaced stepping over VEX-encoded RIP-relative AVX
+# instructions.
+
+if { ![istarget x86_64-*-* ] || ![is_lp64_target] } {
+    verbose "Skipping x86_64 displaced stepping tests."
+    return
+}
+
+standard_testfile .S
+
+set additional_flags "-Wa,-g"
+
+if { [prepare_for_testing "failed to prepare" ${testfile} ${srcfile} \
+	  [list debug $additional_flags]] } {
+    return -1
+}
+
+# Get things started.
+
+gdb_test "set displaced-stepping on" ""
+gdb_test "show displaced-stepping" ".* displaced stepping .* is on.*"
+
+if ![runto_main] then {
+    fail "can't run to main"
+    return 0
+}
+
+# GDB picks a spare register from this list to hold the RIP-relative
+# address.
+set rip_regs { "rax" "rbx" "rcx" "rdx" "rbp" "rsi" "rdi" }
+
+# Assign VAL to all the RIP_REGS.
+
+proc set_regs { val } {
+    global gdb_prompt
+    global rip_regs
+
+    foreach reg ${rip_regs} {
+	gdb_test_no_output "set \$${reg} = ${val}"
+    }
+}
+
+# Verify all RIP_REGS print as HEX_VAL_RE in hex.
+
+proc verify_regs { hex_val_re } {
+    global rip_regs
+
+    foreach reg ${rip_regs} {
+	gdb_test "p /x \$${reg}" " = ${hex_val_re}" "${reg} expected value"
+    }
+}
+
+# Set a break at FUNC, which starts with a RIP-relative instruction
+# that we want to displaced-step over, and then continue over the
+# breakpoint, forcing a displaced-stepping sequence.
+
+proc disp_step_func { func } {
+    global srcfile
+
+    set test_start_label "${func}"
+    set test_end_label "${func}_end"
+
+    gdb_test "break ${test_start_label}" \
+	"Breakpoint.*at.* file .*$srcfile, line.*" \
+	"break ${test_start_label}"
+    gdb_test "break ${test_end_label}" \
+	"Breakpoint.*at.* file .*$srcfile, line.*" \
+	"break ${test_end_label}"
+
+    gdb_test "continue" \
+	"Continuing.*Breakpoint.*, ${test_start_label} ().*" \
+	"continue to ${test_start_label}"
+
+    # GDB picks a spare register to hold the RIP-relative address.
+    # Ensure the spare register value is restored properly (rax-rdi,
+    # sans rsp).
+    set value "0xdeadbeefd3adb33f"
+    set_regs $value
+
+    gdb_test "continue" \
+	"Continuing.*Breakpoint.*, ${test_end_label} ().*" \
+	"continue to ${test_end_label}"
+
+    verify_regs $value
+}
+
+# Test a VEX2-encoded RIP-relative instruction.
+with_test_prefix "vex2" {
+    # This case writes to the 'xmm0' register.  Confirm the register's
+    # value is what we believe it is before the AVX instruction runs.
+    # Fedora: 0* for missing: https://sourceware.org/bugzilla/show_bug.cgi?id=16225
+    gdb_test "p /x \$xmm0.uint128" " = 0x00*" \
+	"xmm0 has expected value before"
+
+    disp_step_func "test_rip_vex2"
+
+    # Confirm the instruction's expected side effects.  It should have
+    # modified xmm0.
+    # Fedora: 0* for missing: https://sourceware.org/bugzilla/show_bug.cgi?id=16225
+    gdb_test "p /x \$xmm0.uint128" " = 0x0*1122334455667788" \
+	"xmm0 has expected value after"
+}
+
+# Test a VEX3-encoded RIP-relative instruction.
+with_test_prefix "vex3" {
+    # This case writes to the 'var128' variable.  Confirm the
+    # variable's value is what we believe it is before the AVX
+    # instruction runs.
+    gdb_test "p /x (unsigned long long \[2\]) var128" \
+	" = \\{0xaa55aa55aa55aa55, 0x55aa55aa55aa55aa\\}" \
+	"var128 has expected value before"
+
+    # Run the AVX instruction.
+    disp_step_func "test_rip_vex3"
+
+    # Confirm the instruction's expected side effects.  It should have
+    # modifed the 'var128' variable.
+    gdb_test "p /x (unsigned long long \[2\]) var128" \
+	" = \\{0x1122334455667788, 0x0\\}" \
+	"var128 has expected value after"
+}
+
+# Done, run program to exit.
+gdb_continue_to_end "amd64-disp-step-avx"
--- a/gdb.spec
+++ b/gdb.spec
@ -20,13 +20,13 @@ Name: %{?scl_prefix}gdb
 # Freeze it when GDB gets branched
 %global snapsrc    20170420
 # See timestamp of source gnulib installed into gdb/gnulib/ .
-%global snapgnulib 20150822
+%global snapgnulib 20161115
 %global tarname gdb-%{version}
 Version: 8.0.1

 # The release always contains a leading reserved number, start it at 1.
 # `upstream' is not a part of `name' to stay fully rpm dependencies compatible for the testing.
-Release: 31%{?dist}
+Release: 36%{?dist}

 License: GPLv3+ and GPLv3+ with exceptions and GPLv2+ and GPLv2+ with exceptions and GPL+ and LGPLv2+ and LGPLv3+ and BSD and Public Domain and GFDL
 Group: Development/Debuggers
@ -737,6 +737,13 @@ Patch1259: gdb-rhbz1498758-5of5.patch
 Patch1261: gdb-rhbz1228556-bpt-inlined-func-name-1of2.patch
 Patch1262: gdb-rhbz1228556-bpt-inlined-func-name-2of2.patch

+# Fix random 'FAIL: libstdc++-prettyprinters/80276.cc whatis p4' (GCC
+# PR83906) (Pedro Alves).
+Patch1263: gdb-random-libstdcpp-prettyprinters-fail.patch
+
+# Fix signal handlers regression (RH BZ 1542149, Pedro Alves).
+Patch1270: gdb-rhbz1542149-spawn-default-signal-handlers-regression.patch
+
 %if 0%{!?rhel:1} || 0%{?rhel} > 6
 # RL_STATE_FEDORA_GDB would not be found for:
 # Patch642: gdb-readline62-ask-more-rh.patch
@ -1149,6 +1156,8 @@ done
 %patch1259 -p1
 %patch1261 -p1
 %patch1262 -p1
+%patch1263 -p1
+%patch1270 -p1

 %patch1075 -p1
 %if 0%{?rhel:1} && 0%{?rhel} <= 7
@ -1731,6 +1740,24 @@ then
 fi

 %changelog
+* Tue Feb  6 2018 Jan Kratochvil <jan.kratochvil@redhat.com> - 8.0.1-36.fc26
+- Fix signal handlers regression (RH BZ 1542149, Pedro Alves).
+
+* Thu Jan 25 2018 Sergio Durigan Junior <sergiodj@redhat.com> - 8.0.1-35.fc26
+- Add %patch directive to the commit below.
+
+* Wed Jan 24 2018 Sergio Durigan Junior <sergiodj@fedoraproject.org> - 8.0.1-34.fc26
+- Fix random 'FAIL: libstdc++-prettyprinters/80276.cc whatis p4' (GCC PR83906)
+  (Pedro Alves).
+
+* Wed Dec  6 2017 Jan Kratochvil <jan.kratochvil@redhat.com> - 8.0.1-33.fc26
+- [rhel7] Fix C++ compiler compatibility.
+
+* Tue Dec  5 2017 Jan Kratochvil <jan.kratochvil@redhat.com> - 8.0.1-32.fc26
+- Backport upstream fix for Intel PKRU (Walfred Tedeschi).
+- Backport upstream fix AVX instr. single-stepping (RH BZ 1515209, Pedro Alves).
+- Fix snapgnulib bundle number.
+
 * Sat Dec  2 2017 Jan Kratochvil <jan.kratochvil@redhat.com> - 8.0.1-31.fc26
 - [testsuite] Fix BuildRequires for non-x86* arches.
Author	SHA1	Message	Date
Jan Kratochvil	a85ad4966c	Fix signal handlers regression (RH BZ 1542149, Pedro Alves).	2018-02-06 22:41:37 +01:00
Sergio Durigan Junior	259e224402	Add %patch directive to the commit below. (and bump version)	2018-01-25 14:33:21 -05:00
Sergio Durigan Junior	60f38d2ebd	Applying the patch for last commit.	2018-01-25 14:28:45 -05:00
Sergio Durigan Junior	81b9cf010a	Fix random 'FAIL: libstdc++-prettyprinters/80276.cc whatis p4' (GCC PR83906) (Pedro Alves).	2018-01-24 15:00:35 -05:00
Jan Kratochvil	cc58fc0cba	[rhel7] Fix C++ compiler compatibility.	2017-12-06 21:09:32 +01:00
Jan Kratochvil	081d76f3ce	Backport upstream fix for Intel PKRU (Walfred Tedeschi). Backport upstream fix AVX instr. single-stepping (RH BZ 1515209, Pedro Alves). Fix snapgnulib bundle number.	2017-12-06 12:45:27 +01:00