From: Jan Kratochvil <jan.kratochvil@redhat.com>
To: Paul Pluzhnikov <ppluzhnikov@google.com>
Cc: gdb-patches@sourceware.org
Subject: Re: [patch] [new testcase] Regression on qsort_cmp
Date: Tue, 25 Aug 2009 19:15:00 -0000 [thread overview]
Message-ID: <20090825185510.GA29860@host0.dyn.jankratochvil.net> (raw)
In-Reply-To: <8ac60eac0908251020w721af576tc18830a09449ea3c@mail.gmail.com>
On Tue, 25 Aug 2009 19:20:30 +0200, Paul Pluzhnikov wrote:
> fixed by the "remove assertions and deal with possible unexpected overlaps"
> patch I am working on :)
Missed the thread, OK, great:
http://sourceware.org/ml/gdb-patches/2009-08/msg00253.html
> In order to run this test on x86_64, I had to do this:
>
> - set lib_flags {debug}
> + set lib_flags {debug additional_flags=-fpic}
>
> Otherwise, I get
> gdb.base/solib-overlap-lib.c.o: relocation R_X86_64_32 against `a
> local symbol' can not be used when making a shared object; recompile
> with -fPIC
> and the test goes UNTESTED.
There was a bug in the .exp file:
proc gdb_compile_shlib:
# Build a shared library from SOURCES. You must use get_compiler_info
# first.
I did not use it. Fixed.
(It worked for me in the trivial case of an empty function.)
Thanks,
Jan
2009-08-25 Jan Kratochvil <jan.kratochvil@redhat.com>
* gdb.base/solib-overlap.exp, gdb.base/solib-overlap-lib.c,
gdb.base/solib-overlap-main.c: New.
--- /dev/null
+++ b/gdb/testsuite/gdb.base/solib-overlap-lib.c
@@ -0,0 +1,27 @@
+/* Copyright 2009 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/>.
+
+ Contributed by Jan Kratochvil <jan.kratochvil@redhat.com>. */
+
+void
+libsym (void)
+{
+}
+
+#ifdef SYMB
+void
+libsymb (void)
+{
+}
+#endif
--- /dev/null
+++ b/gdb/testsuite/gdb.base/solib-overlap-main.c
@@ -0,0 +1,25 @@
+/* Copyright 2009 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/>.
+
+ Contributed by Jan Kratochvil <jan.kratochvil@redhat.com>. */
+
+#include <unistd.h>
+
+int
+main (void)
+{
+ sleep (60);
+
+ return 1;
+}
--- /dev/null
+++ b/gdb/testsuite/gdb.base/solib-overlap.exp
@@ -0,0 +1,139 @@
+# Copyright 2009 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/>.
+#
+# Contributed by Jan Kratochvil <jan.kratochvil@redhat.com>.
+
+# Test GDB can cope with two libraries loaded with overlapping VMA ranges.
+# Prelink libraries first so they can be loaded and their native address.
+# In such case `struct linkmap'.l_addr will be zero. Provide different
+# unprelinked library files on the disk which have zero-based VMAs. These
+# different files should have their .dynamic section at a different offset in
+# page size so that we get for
+# warning: .dynamic section for "..." is not at the expected address
+# the reason
+# (wrong library or version mismatch?)
+# and not:
+# difference appears to be caused by prelink, adjusting expectations
+# In such case both disk libraries will be loaded at VMAs starting at zero.
+
+if [skip_shlib_tests] {
+ return 0
+}
+
+# Are we on a target board? It is required for attaching to a process.
+if [is_remote target] {
+ return 0
+}
+
+if [get_compiler_info binfile-unused] {
+ return -1;
+}
+
+# Library file.
+set libname "solib-overlap-lib"
+set srcfile_lib ${srcdir}/${subdir}/${libname}.c
+# Binary file.
+set testfile "solib-overlap-main"
+set srcfile ${srcdir}/${subdir}/${testfile}.c
+
+# Base addresses for `prelink -r' which should be compatible with both -m32 and
+# -m64 targets. If it clashes with system prelinked libraries it would give
+# false PASS.
+# Prelink first lib1 at 0x40000000 and lib2 at 0x41000000.
+# During second pass try lib1 at 0x50000000 and lib2 at 0x51000000.
+foreach prelink_lib1 {0x40000000 0x50000000} {
+ set prelink_lib2 [format "0x%x" [expr $prelink_lib1 + 0x01000000]]
+
+ set old_prefix $pf_prefix
+ lappend pf_prefix "$prelink_lib1:"
+
+ # Library file.
+ set binfile_lib1 ${objdir}/${subdir}/${libname}1-${prelink_lib1}.so
+ set binfile_lib2 ${objdir}/${subdir}/${libname}2-${prelink_lib1}.so
+ set lib_flags {debug}
+ # Binary file.
+ set binfile_base ${testfile}-${prelink_lib1}
+ set binfile ${objdir}/${subdir}/${binfile_base}
+ set bin_flags [list debug shlib=${binfile_lib1} shlib=${binfile_lib2}]
+ set escapedbinfile [string_to_regexp ${binfile}]
+
+ if { [gdb_compile_shlib ${srcfile_lib} ${binfile_lib1} $lib_flags] != ""
+ || [gdb_compile_shlib ${srcfile_lib} ${binfile_lib2} $lib_flags] != ""
+ || [gdb_compile ${srcfile} ${binfile} executable $bin_flags] != "" } {
+ untested "Could not compile ${binfile_lib1}, ${binfile_lib2} or ${binfile}."
+ return -1
+ }
+
+ if {[catch "system \"prelink -N -r ${prelink_lib1} ${binfile_lib1}\""] != 0
+ || [catch "system \"prelink -N -r ${prelink_lib2} ${binfile_lib2}\""] != 0} {
+ # Maybe we don't have prelink.
+ untested "Could not prelink ${binfile_lib1} or ${binfile_lib2}."
+ return -1
+ }
+
+ # Start the program running and then wait for a bit, to be sure
+ # that it can be attached to.
+
+ set testpid [eval exec $binfile &]
+ sleep 2
+ if { [istarget "*-*-cygwin*"] } {
+ # testpid is the Cygwin PID, GDB uses the Windows PID, which might be
+ # different due to the way fork/exec works.
+ set testpid [ exec ps -e | gawk "{ if (\$1 == $testpid) print \$4; }" ]
+ }
+
+ remote_exec build "mv -f ${binfile_lib1} ${binfile_lib1}-running"
+ remote_exec build "mv -f ${binfile_lib2} ${binfile_lib2}-running"
+
+ # Provide another exported function name to cause different sizes of sections.
+ lappend lib_flags additional_flags=-DSYMB
+
+ if { [gdb_compile_shlib ${srcfile_lib} ${binfile_lib1} $lib_flags] != ""
+ || [gdb_compile_shlib ${srcfile_lib} ${binfile_lib2} $lib_flags] != ""} {
+ untested "Could not recompile ${binfile_lib1} or ${binfile_lib2}."
+ remote_exec build "kill -9 ${testpid}"
+ return -1
+ }
+
+ clean_restart ${binfile_base}
+ # This testcase currently does not support remote targets.
+ # gdb_load_shlibs ${binfile_lib1} ${binfile_lib2}
+
+ # Here we should get:
+ # warning: .dynamic section for ".../solib-overlap-lib1.so" is not at the expected address (wrong library or version mismatch?)
+ # warning: .dynamic section for ".../solib-overlap-lib2.so" is not at the expected address (wrong library or version mismatch?)
+
+ set test attach
+ gdb_test_multiple "attach $testpid" $test {
+ -re "Attaching to program.*`?$escapedbinfile'?, process $testpid.*$gdb_prompt $" {
+ pass $test
+ }
+ -re "Attaching to program.*`?$escapedbinfile\.exe'?, process $testpid.*\[Switching to thread $testpid\..*\].*$gdb_prompt $" {
+ # Response expected on Cygwin
+ pass $test
+ }
+ }
+
+ # Detach the process.
+
+ gdb_test "detach" "Detaching from program: .*$escapedbinfile, process $testpid"
+
+ # Wait a bit for gdb to finish detaching
+
+ sleep 5
+
+ remote_exec build "kill -9 ${testpid}"
+
+ set pf_prefix $old_prefix
+}
next prev parent reply other threads:[~2009-08-25 18:55 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <m3bpmvfro1.fsf@fleche.redhat.com>
[not found] ` <200908041804.30240.thiago.bauermann@gmail.com>
[not found] ` <20090805044504.GJ4370@adacore.com>
[not found] ` <4A7ABAF0.7010004@Sun.COM>
[not found] ` <C11EE4DE-A0A0-4A9E-A86E-9153A7AC81C8@adacore.com>
[not found] ` <4A7FFAB3.7080903@Sun.COM>
[not found] ` <m33a7z8qq5.fsf@fleche.redhat.com>
[not found] ` <8ac60eac0908101320k6cc20e9age92656cba254a5d4@mail.gmail.com>
[not found] ` <4A812B8B.8040202@Sun.COM>
[not found] ` <20090825084613.GA22242@host0.dyn.jankratochvil.net>
2009-08-25 9:40 ` [patch] [new testcase] Regression on qsort_cmp [Re: status of Darwin support] Jan Kratochvil
2009-08-25 17:51 ` Paul Pluzhnikov
2009-08-25 19:15 ` Jan Kratochvil [this message]
2009-08-31 22:56 ` [patch] [new testcase] Regression on qsort_cmp Tom Tromey
2009-09-01 12:08 ` Jan Kratochvil
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=20090825185510.GA29860@host0.dyn.jankratochvil.net \
--to=jan.kratochvil@redhat.com \
--cc=gdb-patches@sourceware.org \
--cc=ppluzhnikov@google.com \
/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