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"
next prev parent 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