public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
* PR10687 Printing c++ class with static array (of same type) can produce  infinite output
@ 2010-04-14  2:14 Chris Moller
  2010-04-20 18:27 ` Tom Tromey
  2010-06-14 16:12 ` Ulrich Weigand
  0 siblings, 2 replies; 4+ messages in thread
From: Chris Moller @ 2010-04-14  2:14 UTC (permalink / raw)
  To: gdb-patches

[-- Attachment #1: Type: text/plain, Size: 332 bytes --]

Added yet another recursion detector (the third, so far as I noticed) to 
cp-valprint.c.  This one detects recursing static arrays which, although 
it comes unstuck in the same place as the 9067 recursing static structs, 
comes unstuck in a different way and thus needs a slightly different 
mechanism to detect.

No regressions.



[-- Attachment #2: pr10687.patch --]
[-- Type: text/x-patch, Size: 7922 bytes --]

? testsuite/gdb.cp/pr10687
Index: ChangeLog
===================================================================
RCS file: /cvs/src/src/gdb/ChangeLog,v
retrieving revision 1.11610
diff -u -r1.11610 ChangeLog
--- ChangeLog	13 Apr 2010 21:07:15 -0000	1.11610
+++ ChangeLog	14 Apr 2010 01:42:03 -0000
@@ -1,3 +1,10 @@
+2010-04-13  Chris Moller  <cmoller@redhat.com>
+
+	* cp-valprint.c (global): Adding new static array recursion
+	detection obstack.
+	(cp_print_value_fields, cp_print_static_field): Added new static
+	array recursion detection code.
+
 2010-04-13  Mark Kettenis  <kettenis@gnu.org>
 
 	* i386-linux-tdep.c (i386_linux_regset_sections): Remove extended
Index: cp-valprint.c
===================================================================
RCS file: /cvs/src/src/gdb/cp-valprint.c,v
retrieving revision 1.64
diff -u -r1.64 cp-valprint.c
--- cp-valprint.c	8 Feb 2010 18:04:16 -0000	1.64
+++ cp-valprint.c	14 Apr 2010 01:42:04 -0000
@@ -71,6 +71,7 @@
 
 static struct obstack dont_print_vb_obstack;
 static struct obstack dont_print_statmem_obstack;
+static struct obstack dont_print_stat_array_obstack;
 
 extern void _initialize_cp_valprint (void);
 
@@ -155,12 +156,17 @@
 {
   int i, len, n_baseclasses;
   int fields_seen = 0;
+  static int last_set_recurse = -1;
 
   CHECK_TYPEDEF (type);
   
-  if (recurse == 0
-      && obstack_object_size (&dont_print_statmem_obstack) > 0)
-    obstack_free (&dont_print_statmem_obstack, NULL);
+  if (recurse == 0)
+    {
+      if (obstack_object_size (&dont_print_statmem_obstack) > 0)
+	obstack_free (&dont_print_statmem_obstack, NULL);
+      if (obstack_object_size (&dont_print_stat_array_obstack) > 0)
+	obstack_free (&dont_print_stat_array_obstack, NULL);
+    }
 
   fprintf_filtered (stream, "{");
   len = TYPE_NFIELDS (type);
@@ -181,12 +187,20 @@
   else
     {
       void *statmem_obstack_top = NULL;
+      void *stat_array_obstack_top = NULL;
       
       if (dont_print_statmem == 0)
 	{
 	  /* Set the current printed-statics stack top.  */
 	  statmem_obstack_top
 	    = obstack_next_free (&dont_print_statmem_obstack);
+
+	  if (last_set_recurse != recurse)
+	    {
+	      stat_array_obstack_top
+		= obstack_next_free (&dont_print_stat_array_obstack);
+	      last_set_recurse = recurse;
+	    }
 	}
 
       for (i = n_baseclasses; i < len; i++)
@@ -307,9 +321,16 @@
 
       if (dont_print_statmem == 0)
 	{
-	  /* In effect, a pop of the printed-statics stack.  */
 	  if (obstack_object_size (&dont_print_statmem_obstack) > 0) 
 	    obstack_free (&dont_print_statmem_obstack, statmem_obstack_top);
+
+	  if (last_set_recurse != recurse)
+	    {
+	      if (obstack_object_size (&dont_print_stat_array_obstack) > 0) 
+		obstack_free (&dont_print_stat_array_obstack,
+			      stat_array_obstack_top);
+	      last_set_recurse = -1;
+	    }
 	}
 
       if (options->pretty)
@@ -508,6 +529,7 @@
 		       const struct value_print_options *options)
 {
   struct value_print_options opts;
+  
   if (TYPE_CODE (type) == TYPE_CODE_STRUCT)
     {
       CORE_ADDR *first_dont_print;
@@ -531,10 +553,12 @@
 	}
 
       addr = value_address (val);
+
       obstack_grow (&dont_print_statmem_obstack, (char *) &addr,
 		    sizeof (CORE_ADDR));
 
       CHECK_TYPEDEF (type);
+
       cp_print_value_fields (type, value_enclosing_type (val),
 			     value_contents_all (val),
 			     value_embedded_offset (val), addr,
@@ -542,6 +566,32 @@
       return;
     }
 
+  if (TYPE_CODE (type) == TYPE_CODE_ARRAY)
+    {
+      struct type **first_dont_print;
+      int i;
+      struct type * target_type = TYPE_TARGET_TYPE (type);
+
+      first_dont_print
+	= (struct type **) obstack_base (&dont_print_stat_array_obstack);
+      i = obstack_object_size (&dont_print_stat_array_obstack)
+	/ sizeof (CORE_ADDR);
+
+      while (--i >= 0)
+	{
+	  if (target_type == first_dont_print[i])
+	    {
+	      fputs_filtered ("<same as static member of an already"
+			      " seen type>",
+			      stream);
+	      return;
+	    }
+	}
+
+      obstack_grow (&dont_print_stat_array_obstack, (char *) &target_type,
+		    sizeof (struct type *));
+    }
+
   opts = *options;
   opts.deref_ref = 0;
   val_print (type, value_contents_all (val), 
@@ -672,6 +722,7 @@
 			   show_objectprint,
 			   &setprintlist, &showprintlist);
 
+  obstack_begin (&dont_print_stat_array_obstack, 32 * sizeof (CORE_ADDR));
   obstack_begin (&dont_print_statmem_obstack, 32 * sizeof (CORE_ADDR));
   obstack_begin (&dont_print_vb_obstack, 32 * sizeof (struct type *));
 }
Index: testsuite/ChangeLog
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/ChangeLog,v
retrieving revision 1.2231
diff -u -r1.2231 ChangeLog
--- testsuite/ChangeLog	12 Apr 2010 09:49:34 -0000	1.2231
+++ testsuite/ChangeLog	14 Apr 2010 01:42:21 -0000
@@ -1,3 +1,9 @@
+2010-04-13  Chris Moller  <cmoller@mredhat.com>
+
+	* gdb.cp/Makefile.in  (EXECUTABLES): Added pr10687
+	* gdb.cp/pr10687.cc: New file.
+	* gdb.cp/pr10687.exp: New file.
+
 2010-04-12  Phil Muldoon  <pmuldoon@redhat.com>
 
 	* gdb.python/py-breakpoint.c: Make result global.
Index: testsuite/gdb.cp/Makefile.in
===================================================================
RCS file: /cvs/src/src/gdb/testsuite/gdb.cp/Makefile.in,v
retrieving revision 1.9
diff -u -r1.9 Makefile.in
--- testsuite/gdb.cp/Makefile.in	8 Feb 2010 18:27:53 -0000	1.9
+++ testsuite/gdb.cp/Makefile.in	14 Apr 2010 01:42:21 -0000
@@ -5,7 +5,7 @@
 	derivation inherit local member-ptr method misc \
         overload ovldbreak ref-typ ref-typ2 templates userdef virtfunc namespace \
 	ref-types ref-params method2 pr9594 gdb2495 virtfunc2 pr9067 \
-	pr1072
+	pr1072 pr10687
 
 all info install-info dvi install uninstall installcheck check:
 	@echo "Nothing to be done for $@..."
Index: testsuite/gdb.cp/pr10687.cc
===================================================================
RCS file: testsuite/gdb.cp/pr10687.cc
diff -N testsuite/gdb.cp/pr10687.cc
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.cp/pr10687.cc	14 Apr 2010 01:42:22 -0000
@@ -0,0 +1,24 @@
+class vec2 
+{
+  public:
+    vec2() { _v[0] = _v[1] = 0; }
+    vec2(int x, int y) { _v[0] = x; _v[1] = y; }
+    static vec2 axis[2];
+    static vec2 axis6[6];
+  private:
+    int _v[2];
+};
+
+vec2 vec2::axis[2] = { vec2(1,0), vec2(0,1) };
+vec2 vec2::axis6[6] = { 
+  vec2(1,0), vec2(0,1),
+  vec2(2,0), vec2(0,2),
+  vec2(3,0), vec2(0,3) 
+};
+
+int main(int argc, char*argv[])
+{
+  vec2 a;
+
+  return 0;  // marker
+}
Index: testsuite/gdb.cp/pr10687.exp
===================================================================
RCS file: testsuite/gdb.cp/pr10687.exp
diff -N testsuite/gdb.cp/pr10687.exp
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.cp/pr10687.exp	14 Apr 2010 01:42:22 -0000
@@ -0,0 +1,31 @@
+#Copyright 2010 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/>.
+
+set testfile pr10687
+set srcfile ${testfile}.cc
+if [prepare_for_testing $testfile.exp $testfile $srcfile {debug c++}] {
+    return -1
+}
+
+if ![runto_main] then {
+    fail "Can't run to main"
+    return
+}
+
+gdb_breakpoint [gdb_get_line_number "marker"]
+gdb_continue_to_breakpoint "marker"
+
+gdb_test "p a" "{static axis = {{static axis = <same as static member of an already.*"
+

[-- Attachment #3: pr10687-gdb-sum.diff --]
[-- Type: text/x-patch, Size: 986 bytes --]

--- virgin-gdb.sum	2010-04-13 20:28:27.100326020 -0400
+++ patched-gdb.sum	2010-04-13 21:24:49.447325737 -0400
@@ -1,4 +1,4 @@
-Test Run By moller on Tue Apr 13 20:19:36 2010
+Test Run By moller on Tue Apr 13 21:16:27 2010
 Native configuration is i686-pc-linux-gnu
 
 		=== gdb tests ===
@@ -12264,6 +12264,9 @@
 Running ../../../src/gdb/testsuite/gdb.cp/pr-574.exp ...
 PASS: gdb.cp/pr-574.exp: continue to breakpoint: end of constructors
 PASS: gdb.cp/pr-574.exp: PR gdb/574
+Running ../../../src/gdb/testsuite/gdb.cp/pr10687.exp ...
+PASS: gdb.cp/pr10687.exp: continue to breakpoint: marker
+PASS: gdb.cp/pr10687.exp: p a
 Running ../../../src/gdb/testsuite/gdb.cp/pr10728.exp ...
 PASS: gdb.cp/pr10728.exp: continue to breakpoint: marker 1
 PASS: gdb.cp/pr10728.exp: print x->y2 - x->y1
@@ -16698,7 +16701,7 @@
 
 		=== gdb Summary ===
 
-# of expected passes		15878
+# of expected passes		15880
 # of unexpected failures	17
 # of expected failures		44
 # of untested testcases		4

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: PR10687 Printing c++ class with static array (of same type) can produce  infinite output
  2010-04-14  2:14 PR10687 Printing c++ class with static array (of same type) can produce infinite output Chris Moller
@ 2010-04-20 18:27 ` Tom Tromey
  2010-04-20 20:24   ` Chris Moller
  2010-06-14 16:12 ` Ulrich Weigand
  1 sibling, 1 reply; 4+ messages in thread
From: Tom Tromey @ 2010-04-20 18:27 UTC (permalink / raw)
  To: Chris Moller; +Cc: gdb-patches

>>>>> "Chris" == Chris Moller <cmoller@redhat.com> writes:

Chris> Added yet another recursion detector (the third, so far as I noticed)
Chris> to cp-valprint.c.  This one detects recursing static arrays which,
Chris> although it comes unstuck in the same place as the 9067 recursing
Chris> static structs, comes unstuck in a different way and thus needs a
Chris> slightly different mechanism to detect.

Thanks.

A couple nits and then this is ok.

Chris>        addr = value_address (val);
Chris> +
Chris>        obstack_grow (&dont_print_statmem_obstack, (char *) &addr,
Chris>  		    sizeof (CORE_ADDR));
 
Chris>        CHECK_TYPEDEF (type);
Chris> +
Chris>        cp_print_value_fields (type, value_enclosing_type (val),
Chris>  			     value_contents_all (val),
Chris>  			     value_embedded_offset (val), addr,

These two newline additions seem gratuitous.

Chris> +      struct type * target_type = TYPE_TARGET_TYPE (type);

No space after the "*" here.

This is ok with those changes.  Thanks.

Tom

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: PR10687 Printing c++ class with static array (of same type) can  produce  infinite output
  2010-04-20 18:27 ` Tom Tromey
@ 2010-04-20 20:24   ` Chris Moller
  0 siblings, 0 replies; 4+ messages in thread
From: Chris Moller @ 2010-04-20 20:24 UTC (permalink / raw)
  To: tromey; +Cc: gdb-patches

On 04/20/10 14:27, Tom Tromey wrote:
>>>>>> "Chris" == Chris Moller<cmoller@redhat.com>  writes:
>>>>>>              
>
> Chris>  Added yet another recursion detector (the third, so far as I noticed)
> Chris>  to cp-valprint.c.  This one detects recursing static arrays which,
> Chris>  although it comes unstuck in the same place as the 9067 recursing
> Chris>  static structs, comes unstuck in a different way and thus needs a
> Chris>  slightly different mechanism to detect.
>
> Thanks.
>
> A couple nits and then this is ok.
>
> Chris>         addr = value_address (val);
> Chris>  +
> Chris>         obstack_grow (&dont_print_statmem_obstack, (char *)&addr,
> Chris>   		sizeof (CORE_ADDR));
>
> Chris>         CHECK_TYPEDEF (type);
> Chris>  +
> Chris>         cp_print_value_fields (type, value_enclosing_type (val),
> Chris>   			value_contents_all (val),
> Chris>   			value_embedded_offset (val), addr,
>
> These two newline additions seem gratuitous.
>
> Chris>  +      struct type * target_type = TYPE_TARGET_TYPE (type);
>
> No space after the "*" here.
>
> This is ok with those changes.  Thanks.
>    

Nits fixed and patch committed.

Chris

> Tom
>    

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: PR10687 Printing c++ class with static array (of same type) can produce  infinite output
  2010-04-14  2:14 PR10687 Printing c++ class with static array (of same type) can produce infinite output Chris Moller
  2010-04-20 18:27 ` Tom Tromey
@ 2010-06-14 16:12 ` Ulrich Weigand
  1 sibling, 0 replies; 4+ messages in thread
From: Ulrich Weigand @ 2010-06-14 16:12 UTC (permalink / raw)
  To: Chris Moller; +Cc: gdb-patches

Chris Moller wrote:

> +  if (TYPE_CODE (type) == TYPE_CODE_ARRAY)
> +    {
> +      struct type **first_dont_print;
> +      int i;
> +      struct type * target_type = TYPE_TARGET_TYPE (type);
> +
> +      first_dont_print
> +	= (struct type **) obstack_base (&dont_print_stat_array_obstack);
> +      i = obstack_object_size (&dont_print_stat_array_obstack)
> +	/ sizeof (CORE_ADDR);
> +
> +      while (--i >= 0)
> +	{
> +	  if (target_type == first_dont_print[i])
> +	    {
> +	      fputs_filtered ("<same as static member of an already"
> +			      " seen type>",
> +			      stream);
> +	      return;
> +	    }
> +	}
> +
> +      obstack_grow (&dont_print_stat_array_obstack, (char *) &target_type,
> +		    sizeof (struct type *));
> +    }

This is broken on platforms where sizeof (CORE_ADDR) != sizeof (struct type *).
As the objects stored into the dont_print_stat_array_obstack are in fact of
type "struct type *", the size computations also need to use this type's size.

I'm seeing this problem on a spu-elf GDB built as 32-bit PowerPC executable.

Fixed by the patch below.

Tested on spu-elf, committed to mainline.

Bye,
Ulrich


ChangeLog:

	* cp-valprint.c (cp_print_static_field): Members of
	dont_print_stat_array_obstack are of type "struct type *".
	(_initialize_cp_valprint): Likewise.

Index: gdb/cp-valprint.c
===================================================================
RCS file: /cvs/src/src/gdb/cp-valprint.c,v
retrieving revision 1.69
diff -u -p -r1.69 cp-valprint.c
--- gdb/cp-valprint.c	11 Jun 2010 15:36:04 -0000	1.69
+++ gdb/cp-valprint.c	14 Jun 2010 16:03:53 -0000
@@ -615,7 +615,7 @@ cp_print_static_field (struct type *type
       first_dont_print
 	= (struct type **) obstack_base (&dont_print_stat_array_obstack);
       i = obstack_object_size (&dont_print_stat_array_obstack)
-	/ sizeof (CORE_ADDR);
+	/ sizeof (struct type *);
 
       while (--i >= 0)
 	{
@@ -764,7 +764,7 @@ Show printing of object's derived type b
 			   show_objectprint,
 			   &setprintlist, &showprintlist);
 
-  obstack_begin (&dont_print_stat_array_obstack, 32 * sizeof (CORE_ADDR));
+  obstack_begin (&dont_print_stat_array_obstack, 32 * sizeof (struct type *));
   obstack_begin (&dont_print_statmem_obstack, 32 * sizeof (CORE_ADDR));
   obstack_begin (&dont_print_vb_obstack, 32 * sizeof (struct type *));
 }

-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2010-06-14 16:12 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2010-04-14  2:14 PR10687 Printing c++ class with static array (of same type) can produce infinite output Chris Moller
2010-04-20 18:27 ` Tom Tromey
2010-04-20 20:24   ` Chris Moller
2010-06-14 16:12 ` Ulrich Weigand

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).