From: Tom de Vries <tdevries@suse.de>
To: Simon Marchi <simon.marchi@polymtl.ca>, gdb-patches@sourceware.org
Subject: Re: [PATCH master + 11][gdb/symtab] Fix zero address complaint for shlib
Date: Sun, 1 Aug 2021 19:59:55 +0200 [thread overview]
Message-ID: <31b389e1-fbfc-cdfa-14dd-5a4a99167430@suse.de> (raw)
In-Reply-To: <8e1b2a6e-dedb-3184-713f-381c8b51be0b@polymtl.ca>
[-- Attachment #1: Type: text/plain, Size: 998 bytes --]
On 7/30/21 7:50 PM, Simon Marchi wrote:
>> Thanks for the review, and the .debug_rnglists addition.
>>
>> I've added the .debug_rnglists addition here as well, and added testing
>> the complaint when doing gdb_load $lib1, in other words, the unrelocated
>> case (which was already passing before this patch).
>
> Thanks, this LGTM.
>
> Note, when merging in master, you can use the Co-Authored-By tag, since
> we don't have a ChangeLog file there.
>
Ack.
I've updated the test-case to test the complaint both during partial
symtab reading and full symtab reading, and re-tested with a variety of
target board, which required additional updates, and ran into a problem
with .gdb_index, filed as PR28159 - "[gdb/symtab] gdb generates
.gdb_index with only empty hashtable entries, then fails to recognize it
as empty".
The test-case now contains an kfail for this. I'll submit a patch shortly.
I'm planning to look over the commit message once more tomorrow, and
then commit.
Thanks,
- Tom
[-- Attachment #2: 0001-gdb-symtab-Fix-zero-address-complaint-for-shlib.patch --]
[-- Type: text/x-patch, Size: 13416 bytes --]
[gdb/symtab] Fix zero address complaint for shlib
In PR28004 the following warning / Internal error is reported:
...
$ gdb -q -batch \
-iex "set sysroot $(pwd -P)/repro" \
./repro/gdb \
./repro/core \
-ex bt
...
Program terminated with signal SIGABRT, Aborted.
#0 0x00007ff8fe8e5d22 in raise () from repro/usr/lib/libc.so.6
[Current thread is 1 (LWP 1762498)]
#1 0x00007ff8fe8cf862 in abort () from repro/usr/lib/libc.so.6
warning: (Internal error: pc 0x7ff8feb2c21d in read in psymtab, \
but not in symtab.)
warning: (Internal error: pc 0x7ff8feb2c218 in read in psymtab, \
but not in symtab.)
...
#2 0x00007ff8feb2c21e in __gnu_debug::_Error_formatter::_M_error() const \
[clone .cold] (warning: (Internal error: pc 0x7ff8feb2c21d in read in \
psymtab, but not in symtab.)
) from repro/usr/lib/libstdc++.so.6
...
The warning is about the following:
- in find_pc_sect_compunit_symtab we try to find the address
(0x7ff8feb2c218 / 0x7ff8feb2c21d) in the symtabs.
- that fails, so we try again in the partial symtabs.
- we find a matching partial symtab
- however, the partial symtab has a matching symtab, so
we should have found a matching symtab in the first step.
The addresses are:
...
(gdb) info sym 0x7ff8feb2c218
__gnu_debug::_Error_formatter::_M_error() const [clone .cold] in \
section .text of repro/usr/lib/libstdc++.so.6
(gdb) info sym 0x7ff8feb2c21d
__gnu_debug::_Error_formatter::_M_error() const [clone .cold] + 5 in \
section .text of repro/usr/lib/libstdc++.so.6
...
which correspond to unrelocated addresses 0x9c218 and 0x9c21d:
...
$ nm -C repro/usr/lib/libstdc++.so.6.0.29 | grep 000000000009c218
000000000009c218 t __gnu_debug::_Error_formatter::_M_error() const \
[clone .cold]
...
The unrelocated addresses can be found in the partial symbols addresss map:
...
(gdb) set $map = (addrmap_fixed *)m_partial_symtabs->psymtabs_addrmap
(gdb) p /x $map->transitions[24]
$47 = {addr = 0x9989c, value = 0x231d1b0}
(gdb) p /x $map->transitions[25]
$48 = {addr = 0xa2170, value = 0x5439980}
...
and indeed we do:
...
(gdb) p ps
$1 = (partial_symtab *) 0x231d1b0
...
but not in the symbols address map:
...
(gdb) set $symtab = ps->get_compunit_symtab (objfile)
(gdb) set $map = (addrmap_fixed *)$symtab.blockvector.map
(gdb) p /x $map.transitions[20]
$42 = {addr = 0x7ff8feb2993a, value = 0x0}
(gdb) p /x $map.transitions[21]
$43 = {addr = 0x7ff8feb322b0, value = 0x6155ef0}
...
The reason for the difference between the two address maps (who are based on
the same data), is this code in dwarf2_rnglists_process:
...
/* A not-uncommon case of bad debug info.
Don't pollute the addrmap with bad data. */
if (range_beginning + baseaddr == 0
&& !per_objfile->per_bfd->has_section_at_zero)
{
complaint (_(".debug_rnglists entry has start address of zero"
" [in module %s]"), objfile_name (objfile));
continue;
}
...
which triggers for the partial symbol table case (with unrelocated addresses),
but not for the symbol table case (with relocated addresses).
The difference is that in the latter case, baseaddr (initialized from
objfile->text_section_offset ()) isn't 0.
Fix this by removing the baseaddr part from the condition. Same for
dwarf2_ranges_process.
The effect is that:
- the complaint is triggered, and
- the warning / Internal error is no longer triggered.
Tested on x86_64-linux.
Co-Authored-By: Simon Marchi <simon.marchi@polymtl.ca>
gdb/ChangeLog:
2021-07-29 Simon Marchi <simon.marchi@polymtl.ca>
Tom de Vries <tdevries@suse.de>
PR symtab/28004
* gdb/dwarf2/read.c (dwarf2_rnglists_process, dwarf2_ranges_process):
Fix zero address complaint.
* gdb/testsuite/gdb.dwarf2/dw2-zero-range-shlib.c: New test.
* gdb/testsuite/gdb.dwarf2/dw2-zero-range.c: New test.
* gdb/testsuite/gdb.dwarf2/dw2-zero-range.exp: New file.
---
gdb/dwarf2/read.c | 10 +-
gdb/testsuite/gdb.dwarf2/dw2-zero-range-shlib.c | 21 +++
gdb/testsuite/gdb.dwarf2/dw2-zero-range.c | 24 +++
gdb/testsuite/gdb.dwarf2/dw2-zero-range.exp | 194 ++++++++++++++++++++++++
4 files changed, 241 insertions(+), 8 deletions(-)
diff --git a/gdb/dwarf2/read.c b/gdb/dwarf2/read.c
index 6f1b453ef45..de84c47b626 100644
--- a/gdb/dwarf2/read.c
+++ b/gdb/dwarf2/read.c
@@ -13631,7 +13631,6 @@ dwarf2_rnglists_process (unsigned offset, struct dwarf2_cu *cu,
/* Base address selection entry. */
gdb::optional<CORE_ADDR> base;
const gdb_byte *buffer;
- CORE_ADDR baseaddr;
bool overflow = false;
ULONGEST addr_index;
struct dwarf2_section_info *rnglists_section;
@@ -13648,8 +13647,6 @@ dwarf2_rnglists_process (unsigned offset, struct dwarf2_cu *cu,
}
buffer = rnglists_section->buffer + offset;
- baseaddr = objfile->text_section_offset ();
-
while (1)
{
/* Initialize it due to a false compiler warning. */
@@ -13791,7 +13788,7 @@ dwarf2_rnglists_process (unsigned offset, struct dwarf2_cu *cu,
/* A not-uncommon case of bad debug info.
Don't pollute the addrmap with bad data. */
- if (range_beginning + baseaddr == 0
+ if (range_beginning == 0
&& !per_objfile->per_bfd->has_section_at_zero)
{
complaint (_(".debug_rnglists entry has start address of zero"
@@ -13833,7 +13830,6 @@ dwarf2_ranges_process (unsigned offset, struct dwarf2_cu *cu, dwarf_tag tag,
gdb::optional<CORE_ADDR> base;
unsigned int dummy;
const gdb_byte *buffer;
- CORE_ADDR baseaddr;
if (cu_header->version >= 5)
return dwarf2_rnglists_process (offset, cu, tag, callback);
@@ -13849,8 +13845,6 @@ dwarf2_ranges_process (unsigned offset, struct dwarf2_cu *cu, dwarf_tag tag,
}
buffer = per_objfile->per_bfd->ranges.buffer + offset;
- baseaddr = objfile->text_section_offset ();
-
while (1)
{
CORE_ADDR range_beginning, range_end;
@@ -13901,7 +13895,7 @@ dwarf2_ranges_process (unsigned offset, struct dwarf2_cu *cu, dwarf_tag tag,
/* A not-uncommon case of bad debug info.
Don't pollute the addrmap with bad data. */
- if (range_beginning + baseaddr == 0
+ if (range_beginning == 0
&& !per_objfile->per_bfd->has_section_at_zero)
{
complaint (_(".debug_ranges entry has start address of zero"
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-zero-range-shlib.c b/gdb/testsuite/gdb.dwarf2/dw2-zero-range-shlib.c
new file mode 100644
index 00000000000..1ddc8d6677b
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-zero-range-shlib.c
@@ -0,0 +1,21 @@
+/*
+ Copyright 2021 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/>. */
+
+void
+foo (void)
+{
+ asm ("foo_label: .globl foo_label");
+}
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-zero-range.c b/gdb/testsuite/gdb.dwarf2/dw2-zero-range.c
new file mode 100644
index 00000000000..a32c9242c8d
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-zero-range.c
@@ -0,0 +1,24 @@
+/*
+ Copyright 2021 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/>. */
+
+extern void foo (void);
+
+void
+main (void)
+{
+ asm ("main_label: .globl main_label");
+ foo ();
+}
diff --git a/gdb/testsuite/gdb.dwarf2/dw2-zero-range.exp b/gdb/testsuite/gdb.dwarf2/dw2-zero-range.exp
new file mode 100644
index 00000000000..7b46c93d9b5
--- /dev/null
+++ b/gdb/testsuite/gdb.dwarf2/dw2-zero-range.exp
@@ -0,0 +1,194 @@
+# Copyright 2021 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/>.
+
+# Check that the ".debug_ranges entry has start address of zero" complaint
+# is triggered for a loaded shared lib.
+
+load_lib dwarf.exp
+
+# This test can only be run on targets which support DWARF-2 and use gas.
+if {![dwarf2_support]} {
+ verbose "Skipping $gdb_test_file_name."
+ return 0
+}
+
+if {[skip_shlib_tests]} {
+ return 0
+}
+
+# The .c files use __attribute__.
+if [get_compiler_info] {
+ return -1
+}
+if !$gcc_compiled {
+ verbose "Skipping $gdb_test_file_name."
+ return 0
+}
+
+standard_testfile .c -shlib.c -dw.S
+
+# Test with both a .debug_ranges section (DWARF 4) and a .debug_rnglists
+# section (DWARF 5).
+
+foreach_with_prefix ranges_sect {ranges rnglists} {
+ set asm_file [standard_output_file ${ranges_sect}-$srcfile3]
+
+ if { $ranges_sect == "ranges" } {
+ Dwarf::assemble $asm_file {
+ global srcdir subdir srcfile2
+ declare_labels ranges_label
+
+ cu {} {
+ compile_unit {
+ {language @DW_LANG_C}
+ {name $srcfile2}
+ {ranges ${ranges_label} DW_FORM_sec_offset}
+ } {
+ subprogram {
+ {external 1 flag}
+ {name foo}
+ }
+ }
+ }
+
+ ranges {is_64 [is_64_target]} {
+ ranges_label: sequence {
+ base 0
+ range 0 1
+ }
+ }
+ }
+ } elseif { $ranges_sect == "rnglists" } {
+ Dwarf::assemble $asm_file {
+ global srcdir subdir srcfile2
+ declare_labels rnglists_label
+
+ cu {
+ version 5
+ } {
+ compile_unit {
+ {language @DW_LANG_C}
+ {name $srcfile2}
+ {ranges ${rnglists_label} DW_FORM_sec_offset}
+ } {
+ subprogram {
+ {external 1 flag}
+ {name foo}
+ }
+ }
+ }
+
+ rnglists {
+ table {
+ rnglists_label: list_ {
+ start_end 0 1
+ }
+ }
+ }
+ }
+ } else {
+ error "invalid ranges section kind"
+ }
+
+ set lib1 [standard_output_file shr1.sl]
+ set lib_opts "nodebug"
+
+ set sources [list ${srcdir}/${subdir}/$srcfile2 $asm_file]
+ if { [gdb_compile_shlib $sources ${lib1} $lib_opts] != "" } {
+ untested "failed to compile"
+ return -1
+ }
+
+ set exec_opts [list debug shlib=${lib1}]
+ set sources ${srcdir}/${subdir}/${srcfile}
+ if { [gdb_compile $sources ${binfile} executable \
+ $exec_opts] != ""} {
+ untested "failed to compile"
+ return -1
+ }
+
+ clean_restart $binfile
+
+ set readnow_p [readnow]
+
+ # Don't load the symbols for $lib1 during runto_main.
+ # Instead, we do this afterwards using "sharedlibrary $lib1".
+ gdb_test_no_output "set auto-solib-add off"
+
+ if { ![runto_main] } {
+ fail "cannot run to main."
+ return -1
+ }
+
+ set complaint_re ".debug_${ranges_sect} entry has start address of zero"
+ set re \
+ "During symbol reading: $complaint_re \\\[in module \[^\r\n\]*\\\]"
+
+ # Test for presence of complaint, with lib1 relocated.
+ with_complaints 1 {
+ set test "Zero address complaint - relocated - psymtab"
+ set have_complaint 0
+ gdb_test_multiple "sharedlibrary $lib1" $test {
+ -re -wrap $re {
+ set have_complaint 1
+ }
+ -re -wrap "" {
+ }
+ }
+
+ if { $have_complaint } {
+ pass $test
+ } else {
+ set index [have_index $lib1]
+ if { $index == "gdb_index" } {
+ kfail symtab/28159 $test
+ } else {
+ fail $test
+ }
+ }
+ }
+
+ if { ! $readnow_p } {
+ with_complaints 1 {
+ gdb_test "maint expand-symtabs $srcfile2" $re \
+ "Zero address complaint - relocated - symtab"
+ }
+ }
+
+ clean_restart
+ # Test for presence of complaint, with lib1 unrelocated.
+ with_complaints 1 {
+ gdb_load $lib1
+ set test "Zero address complaint - unrelocated - psymtab"
+ set have_complaint [regexp $re.* $gdb_file_cmd_msg]
+ if { $have_complaint } {
+ pass $test
+ } else {
+ set index [have_index $lib1]
+ if { $index == "gdb_index" } {
+ kfail symtab/28159 $test
+ } else {
+ fail $test
+ }
+ }
+ }
+
+ if { ! $readnow_p } {
+ with_complaints 1 {
+ gdb_test "maint expand-symtabs $srcfile2" $re \
+ "Zero address complaint - unrelocated - symtab"
+ }
+ }
+}
next prev parent reply other threads:[~2021-08-01 18:00 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-29 21:23 Tom de Vries
2021-07-30 1:25 ` Simon Marchi via Gdb-patches
2021-07-30 1:48 ` Simon Marchi via Gdb-patches
2021-07-30 15:25 ` Tom de Vries
2021-07-30 17:50 ` Simon Marchi via Gdb-patches
2021-08-01 17:59 ` Tom de Vries [this message]
2021-08-02 13:09 ` Simon Marchi via Gdb-patches
2021-08-06 14:43 ` Tom de Vries via Gdb-patches
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=31b389e1-fbfc-cdfa-14dd-5a4a99167430@suse.de \
--to=tdevries@suse.de \
--cc=gdb-patches@sourceware.org \
--cc=simon.marchi@polymtl.ca \
/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