Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Keith Seitz <keiths@redhat.com>
To: Joel Brobecker <brobecker@adacore.com>
Cc: "gdb-patches@sourceware.org ml" <gdb-patches@sourceware.org>
Subject: Re: [RFA] Fix gdb/15827 (crash w/corrupt DWARF)
Date: Fri, 21 Mar 2014 17:26:00 -0000	[thread overview]
Message-ID: <532C7624.4080704@redhat.com> (raw)
In-Reply-To: <20140321171453.GJ4282@adacore.com>

[-- Attachment #1: Type: text/plain, Size: 1509 bytes --]

Hi, Joel!

Thank you for having a look at this.

On 03/21/2014 10:14 AM, Joel Brobecker wrote:
> Use "(void)" instead of "()". There was a recent policy clarification
> regarding the CS to be using with testcases, and basically we decided
> to try to follow the GCS as much as we reasonably could.

Cut-n-paste-o. Fixed.


>> +# If we get here and gdb hasn't crashed, the tests pass.
>> +pass "corrupt DWARF"
>
> That's just me but I usually do a "print 1" test, just to make sure
> that even if the testing framework did not detect the GDB process
> dying, the "print 1" test definitely will. Not important on most,
> if not all platforms, but ISTR some odd platforms where this helped.
> That's just a suggestion, you don't have to follow it.

Actually, I think that's a very good idea (which did not occur to me). 
My big hesitation with this is that this failure gets reported as 
UNRESOLVED. While I try to be studious about checking 
XFAIL/UNRESOLVED/ERROR, I sometimes overlook these in favor of a raw 
PASS/FAIL check in gdb.sum.

I've attached a revision with those two changes (ChangeLog remains 
unchanged).

Keith

ChangeLog
2014-03-20  Keith Seitz  <keiths@redhat.com>

	PR gdb/15827
	* dwarf2read.c (skip_one_die): Check that all relative-offset
	sibling DIEs fall within range of the current reader's buffer.
	(read_partial_die): Likewise.

testsuite/ChangeLog
2014-03-20  Keith Seitz  <keiths@redhat.com>

	PR gdb/15827
	* gdb.dwarf2/corrupt.c: New file.
	* gdb.dwarf2/corrupt.exp: New file.


[-- Attachment #2: 15827-2.patch --]
[-- Type: text/x-patch, Size: 4017 bytes --]

diff --git a/gdb/dwarf2read.c b/gdb/dwarf2read.c
index 705dc2d..c30b1b3 100644
--- a/gdb/dwarf2read.c
+++ b/gdb/dwarf2read.c
@@ -7103,6 +7103,8 @@ skip_one_die (const struct die_reader_specs *reader, const gdb_byte *info_ptr,
 	      if (sibling_ptr < info_ptr)
 		complaint (&symfile_complaints,
 			   _("DW_AT_sibling points backwards"));
+	      else if (sibling_ptr > reader->buffer_end)
+		dwarf2_section_buffer_overflow_complaint (reader->die_section);
 	      else
 		return sibling_ptr;
 	    }
@@ -15416,6 +15418,8 @@ read_partial_die (const struct die_reader_specs *reader,
 	      if (sibling_ptr < info_ptr)
 		complaint (&symfile_complaints,
 			   _("DW_AT_sibling points backwards"));
+	      else if (sibling_ptr > reader->buffer_end)
+		dwarf2_section_buffer_overflow_complaint (reader->die_section);
 	      else
 		part_die->sibling = sibling_ptr;
 	    }
diff --git a/gdb/testsuite/gdb.dwarf2/corrupt.c b/gdb/testsuite/gdb.dwarf2/corrupt.c
new file mode 100644
index 0000000..bcd5fd8
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/corrupt.c
@@ -0,0 +1,24 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   Copyright 2014 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/>.  */
+
+/* Dummy main function.  */
+
+int
+main (void)
+{
+  return 0;
+}
diff --git a/gdb/testsuite/gdb.dwarf2/corrupt.exp b/gdb/testsuite/gdb.dwarf2/corrupt.exp
new file mode 100644
index 0000000..048ae0c
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/corrupt.exp
@@ -0,0 +1,77 @@
+# Copyright 2014 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/>.
+
+# Test corrupt DWARF input
+# PR gdb/15827
+
+load_lib dwarf.exp
+
+if {![dwarf2_support]} {
+    return 0
+}
+
+standard_testfile corrupt.c corrupt.S
+
+# Make the DWARF used for the test.
+#
+# Here we put DW_AT_sibling DIEs into the output which
+# point off into la-la land.  The whole purpose is to simulate
+# corrupt DWARF information and make sure that GDB can handle it
+# without crashing.
+
+set asm_file [standard_output_file $srcfile2]
+Dwarf::assemble $asm_file {
+    cu {} {
+	compile_unit {} {
+	    declare_labels int_label
+
+	    int_label: base_type {
+		{byte_size 4}
+		{name "int"}
+	    }
+
+	    enumeration_type {
+		{name "ENUM"}
+		{byte_size 4}
+	    } {
+		enumerator {
+		    {name "A"}
+		    {const_value 0}
+		}
+		enumerator {
+		    {name "B"}
+		    {const_value 1}
+		    {sibling 12345678 DW_FORM_ref4}
+		} {
+		    base_type {
+			{byte_size 1}
+			{name "char"}
+		    }
+		}
+		array_type {
+		    {type :$int_label}
+		    {sibling 12345678 DW_FORM_ref4}
+		}
+	    }
+	}
+    }
+}
+
+if {[prepare_for_testing $testfile.exp $testfile \
+	 [list $srcfile $asm_file] {nodebug}]} {
+    return -1
+}
+
+gdb_test "print 1" "= 1" "recover from corrupt DWARF"

  reply	other threads:[~2014-03-21 17:26 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-21 16:48 Keith Seitz
2014-03-21 17:14 ` Joel Brobecker
2014-03-21 17:26   ` Keith Seitz [this message]
2014-03-21 17:54     ` Joel Brobecker
2014-04-16 21:50       ` Keith Seitz
2014-03-21 19:27   ` Doug Evans

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=532C7624.4080704@redhat.com \
    --to=keiths@redhat.com \
    --cc=brobecker@adacore.com \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox