2015-01-01 09:32:14 +00:00
|
|
|
# Copyright 2009-2015 Free Software Foundation, Inc.
|
2009-09-01 12:06:55 +00:00
|
|
|
# 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
|
|
|
|
}
|
|
|
|
|
2015-01-09 11:04:19 +00:00
|
|
|
if {![can_spawn_for_attach]} {
|
2009-09-01 12:06:55 +00:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2012-06-21 20:46:25 +00:00
|
|
|
if [get_compiler_info] {
|
gdb/testsuite/
* config/monitor.exp (gdb_target_cmd): Remove semicolon after
'return'.
(gdb_target_monitor, gdb_load): Likewise.
* config/sid.exp (gdb_load): Likewise.
* config/slite.exp (gdb_load): Likewise.
* config/vx.exp (gdb_start, spawn_vxgdb): Likewise.
* gdb.ada/arrayidx.exp, gdb.ada/null_array.exp: Likewise.
* gdb.arch/mips-octeon-bbit.exp (single_step): Likewise.
(single_step_until): Likewise.
* gdb.arch/powerpc-d128-regs.exp: Likewise.
* gdb.arch/system-gcore.exp: Likewise.
* gdb.base/bigcore.exp (extract_heap): Likewise.
* gdb.base/break-on-linker-gcd-function.exp: Likewise.
* gdb.base/call-ar-st.exp: Likewise.
* gdb.base/call-rt-st.exp: Likewise.
* gdb.base/call-sc.exp, gdb.base/call-strs.exp: Likewise.
* gdb.base/callfuncs.exp, gdb.base/completion.exp: Likewise.
* gdb.base/condbreak.exp, gdb.base/constvars.exp: Likewise.
* gdb.base/corefile.exp: Likewise.
* gdb.base/dbx.exp (gdb_file_cmd): Likewise.
* gdb.base/exprs.exp, gdb.base/fileio.exp: Likewise.
* gdb.base/fixsection.exp: Likewise.
* gdb.base/funcargs.exp: Likewise.
* gdb.base/gcore-buffer-overflow.exp: Likewise.
* gdb.base/gcore-relro.exp: Likewise.
* gdb.base/gcore.exp, gdb.base/gdb11530.exp: Likewise.
* gdb.base/gdb11531.exp, gdb.base/gnu-ifunc.exp: Likewise.
* gdb.base/info-os.exp, gdb.base/info-proc.exp: Likewise.
* gdb.base/interp.exp, gdb.base/langs.exp:: Likewise.
* gdb.base/list.exp: Likewise.
(set_listsize): Likewise.
* gdb.base/logical.exp, gdb.base/mips_pro.exp: Likewise.
* gdb.base/miscexprs.exp, gdb.base/nodebug.exp: Likewise.
* gdb.base/opaque.exp, gdb.base/pointers.exp: Likewise.
* gdb.base/psymtab.exp, gdb.base/ptype.exp: Likewise.
* gdb.base/relational.exp, gdb.base/scope.exp: Likewise.
* gdb.base/setvar.exp: Likewise.
(test_set): Likewise.
* gdb.base/signals.exp, gdb.base/sizeof.exp: Likewise.
* gdb.base/solib-overlap.exp: Likewise.
* gdb.base/store.exp, gdb.base/structs.exp: Likewise.
* gdb.base/structs2.exp, gdb.base/volatile.exp: Likewise.
* gdb.base/watchpoint.exp (initialize): Likewise.
(test_simple_watchpoint): Likewise.
(test_disabling_watchpoints): Likewise.
(test_watchpoint_triggered_in_syscall): Likewise.
* gdb.base/whatis.exp, gdb.cp/ambiguous.exp: Likewise.
* gdb.cp/casts.exp, gdb.cp/ctti.exp: Likewise.
* gdb.cp/namespace.exp, gdb.cp/nsdecl.exp: Likewise.
* gdb.cp/psmang.exp, gdb.dwarf2/dw2-ranges.exp: Likewise.
* gdb.hp/gdb.aCC/optimize.exp: Likewise.
* gdb.hp/gdb.aCC/watch-cmd.exp: Likewise.
* gdb.hp/gdb.base-hp/callfwmall.exp: Likewise.
* gdb.hp/gdb.base-hp/pxdb.exp: Likewise.
* gdb.hp/gdb.compat/xdb1.exp: Likewise.
* gdb.hp/gdb.compat/xdb2.exp: Likewise.
* gdb.hp/gdb.compat/xdb3.exp: Likewise.
* gdb.hp/gdb.defects/bs14602.exp: Likewise.
* gdb.hp/gdb.defects/solib-d.exp: Likewise.
* gdb.mi/gdb792.exp: Likewise.
* gdb.mi/mi-inheritance-syntax-error.exp: Likewise.
* gdb.mi/mi-logging.exp, gdb.mi/mi-var-cp.exp : Likewise.
* gdb.mi/mi-var-rtti.exp, gdb.python/py-type.exp: Likewise.
* gdb.threads/gcore-thread.exp: Likewise.
(load_core): Likewise.
* gdb.threads/pthreads.exp (all_threads_running): Likewise.
(test_startup, check_control_c): Likewise.
* gdb.threads/sigstep-threads.exp: Likewise.
* gdb.threads/thread_check.exp: Likewise.
* gdb.trace/backtrace.exp, gdb.trace/change-loc.exp: Likewise.
* gdb.trace/circ.exp (run_trace_experiment): Likewise.
(set_a_tracepoint, trace_buffer_normal): Likewise.
(gdb_trace_circular_tests): Likewise.
* gdb.trace/collection.exp: Likewise.
* gdb.trace/disconnected-tracing.exp: Likewise.
* gdb.trace/infotrace.exp: Likewise.
* gdb.trace/mi-traceframe-changed.exp: Likewise.
* gdb.trace/mi-tracepoint-changed.exp: Likewise.
* gdb.trace/mi-tsv-changed.exp (test_create_delete_modify_tsv): Likewise.
* gdb.trace/packetlen.exp, gdb.trace/passc-dyn.exp: Likewise.
* gdb.trace/pending.exp, gdb.trace/report.exp: Likewise.
* gdb.trace/stap-trace.exp: Likewise.
* gdb.trace/status-stop.exp,gdb.trace/strace.exp: Likewise.
* gdb.trace/tfind.exp, gdb.trace/trace-break.exp: Likewise.
* gdb.trace/trace-buffer-size.exp: Likewise.
* gdb.trace/tspeed.exp, gdb.trace/tsv.exp: Likewise.
* gdb.trace/unavailable.exp: Likewise.
* gdb.trace/while-dyn.exp: Likewise.
* lib/fortran.exp (set_lang_fortran): Likewise.
* lib/gdb.exp (default_gdb_version, gdb_start_cmd): Likewise.
(gdb_breakpoint, gdb_reinitialize_dir): Likewise.
(default_gdb_start, get_compiler_info): Likewise.
(gdb_compile, gdb_compile_objc, gdb_reload, gdb_init): Likewise.
(get_debug_format, setup_xfail_format): Likewise.
(rerun_to_main, gdb_skip_float_test): Likewise.
(build_id_debug_filename_get, get_remotetimeout): Likewise.
* lib/java.exp (set_lang_java): Likewise.
* lib/mi-support.exp (default_mi_gdb_start): Likewise.
(mi_gdb_reinitialize_dir, mi_gdb_target_cmd): Likewise.
(mi_gdb_file_cmd, mi_gdb_test): Likewise.
(mi_run_cmd_full, mi_expect_interrupt): Likewise.
* lib/objc.exp (set_lang_objc): Likewise.
* lib/pascal.exp (set_lang_pascal): Likewise.
* lib/prompt.exp (default_prompt_gdb_start): Likewise.
* lib/trace-support.exp (gdb_trace_setactions, gdb_tfind_test): Likewise.
(gdb_readexpr, gdb_gettpnum, gdb_find_recursion_test_baseline): Likewise.
2013-03-14 13:34:06 +00:00
|
|
|
return -1
|
2009-09-01 12:06:55 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
# 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}/${testfile}.c
|
|
|
|
|
|
|
|
# Base addresses for `prelink -r' which should be compatible with both -m32 and
|
|
|
|
# -m64 targets. If it clashes with system prelinked libraries it would give
|
|
|
|
# false PASS.
|
|
|
|
# Prelink first lib1 at 0x40000000 and lib2 at 0x41000000.
|
|
|
|
# During second pass try lib1 at 0x50000000 and lib2 at 0x51000000.
|
2012-02-21 22:01:10 +00:00
|
|
|
foreach prelink_lib1 {0x40000000 0x50000000} { with_test_prefix "$prelink_lib1" {
|
2009-09-01 12:06:55 +00:00
|
|
|
set prelink_lib2 [format "0x%x" [expr $prelink_lib1 + 0x01000000]]
|
|
|
|
|
|
|
|
# Library file.
|
test suite update - gdb.base/s*
Convert files gdb.base/s*.exp to use standard_output_file et al.
* gdb.base/save-bp.exp, gdb.base/savedregs.exp,
gdb.base/scope.exp, gdb.base/sep.exp, gdb.base/sepsymtab.exp,
gdb.base/set-lang-auto.exp, gdb.base/setshow.exp,
gdb.base/setvar.exp, gdb.base/shlib-call.exp,
gdb.base/shreloc.exp, gdb.base/sigall.exp,
gdb.base/sigaltstack.exp, gdb.base/sigbpt.exp,
gdb.base/sigchld.exp, gdb.base/siginfo-addr.exp,
gdb.base/siginfo-infcall.exp, gdb.base/siginfo-obj.exp,
gdb.base/siginfo.exp, gdb.base/signals.exp, gdb.base/signest.exp,
gdb.base/signull.exp, gdb.base/sigrepeat.exp,
gdb.base/sigstep.exp, gdb.base/sizeof.exp,
gdb.base/skip-solib.exp, gdb.base/so-impl-ld.exp,
gdb.base/solib-display.exp, gdb.base/solib-nodir.exp,
gdb.base/solib-overlap.exp, gdb.base/solib-symbol.exp,
gdb.base/solib-weak.exp, gdb.base/source.exp,
gdb.base/stack-checking.exp, gdb.base/stale-infcall.exp,
gdb.base/stap-probe.exp, gdb.base/start.exp,
gdb.base/step-break.exp, gdb.base/step-bt.exp,
gdb.base/step-line.exp, gdb.base/step-resume-infcall.exp,
gdb.base/step-test.exp, gdb.base/structs.exp,
gdb.base/structs2.exp, gdb.base/structs3.exp,
gdb.base/symbol-without-target_section.exp: Use standard_testfile,
standard_output_file, prepare_for_testing, clean_restart.
2013-06-27 18:58:28 +00:00
|
|
|
set binfile_lib1 [standard_output_file ${libname}1-${prelink_lib1}.so]
|
2011-02-07 13:35:42 +00:00
|
|
|
set binfile_lib1_test_msg OBJDIR/${subdir}/${libname}1-${prelink_lib1}.so
|
test suite update - gdb.base/s*
Convert files gdb.base/s*.exp to use standard_output_file et al.
* gdb.base/save-bp.exp, gdb.base/savedregs.exp,
gdb.base/scope.exp, gdb.base/sep.exp, gdb.base/sepsymtab.exp,
gdb.base/set-lang-auto.exp, gdb.base/setshow.exp,
gdb.base/setvar.exp, gdb.base/shlib-call.exp,
gdb.base/shreloc.exp, gdb.base/sigall.exp,
gdb.base/sigaltstack.exp, gdb.base/sigbpt.exp,
gdb.base/sigchld.exp, gdb.base/siginfo-addr.exp,
gdb.base/siginfo-infcall.exp, gdb.base/siginfo-obj.exp,
gdb.base/siginfo.exp, gdb.base/signals.exp, gdb.base/signest.exp,
gdb.base/signull.exp, gdb.base/sigrepeat.exp,
gdb.base/sigstep.exp, gdb.base/sizeof.exp,
gdb.base/skip-solib.exp, gdb.base/so-impl-ld.exp,
gdb.base/solib-display.exp, gdb.base/solib-nodir.exp,
gdb.base/solib-overlap.exp, gdb.base/solib-symbol.exp,
gdb.base/solib-weak.exp, gdb.base/source.exp,
gdb.base/stack-checking.exp, gdb.base/stale-infcall.exp,
gdb.base/stap-probe.exp, gdb.base/start.exp,
gdb.base/step-break.exp, gdb.base/step-bt.exp,
gdb.base/step-line.exp, gdb.base/step-resume-infcall.exp,
gdb.base/step-test.exp, gdb.base/structs.exp,
gdb.base/structs2.exp, gdb.base/structs3.exp,
gdb.base/symbol-without-target_section.exp: Use standard_testfile,
standard_output_file, prepare_for_testing, clean_restart.
2013-06-27 18:58:28 +00:00
|
|
|
set binfile_lib2 [standard_output_file ${libname}2-${prelink_lib1}.so]
|
2011-02-07 13:35:42 +00:00
|
|
|
set binfile_lib2_test_msg OBJDIR/${subdir}/${libname}2-${prelink_lib1}.so
|
2009-09-01 12:06:55 +00:00
|
|
|
set lib_flags {debug}
|
|
|
|
# Binary file.
|
|
|
|
set binfile_base ${testfile}-${prelink_lib1}
|
test suite update - gdb.base/s*
Convert files gdb.base/s*.exp to use standard_output_file et al.
* gdb.base/save-bp.exp, gdb.base/savedregs.exp,
gdb.base/scope.exp, gdb.base/sep.exp, gdb.base/sepsymtab.exp,
gdb.base/set-lang-auto.exp, gdb.base/setshow.exp,
gdb.base/setvar.exp, gdb.base/shlib-call.exp,
gdb.base/shreloc.exp, gdb.base/sigall.exp,
gdb.base/sigaltstack.exp, gdb.base/sigbpt.exp,
gdb.base/sigchld.exp, gdb.base/siginfo-addr.exp,
gdb.base/siginfo-infcall.exp, gdb.base/siginfo-obj.exp,
gdb.base/siginfo.exp, gdb.base/signals.exp, gdb.base/signest.exp,
gdb.base/signull.exp, gdb.base/sigrepeat.exp,
gdb.base/sigstep.exp, gdb.base/sizeof.exp,
gdb.base/skip-solib.exp, gdb.base/so-impl-ld.exp,
gdb.base/solib-display.exp, gdb.base/solib-nodir.exp,
gdb.base/solib-overlap.exp, gdb.base/solib-symbol.exp,
gdb.base/solib-weak.exp, gdb.base/source.exp,
gdb.base/stack-checking.exp, gdb.base/stale-infcall.exp,
gdb.base/stap-probe.exp, gdb.base/start.exp,
gdb.base/step-break.exp, gdb.base/step-bt.exp,
gdb.base/step-line.exp, gdb.base/step-resume-infcall.exp,
gdb.base/step-test.exp, gdb.base/structs.exp,
gdb.base/structs2.exp, gdb.base/structs3.exp,
gdb.base/symbol-without-target_section.exp: Use standard_testfile,
standard_output_file, prepare_for_testing, clean_restart.
2013-06-27 18:58:28 +00:00
|
|
|
set binfile [standard_output_file ${binfile_base}]
|
2011-02-07 13:35:42 +00:00
|
|
|
set binfile_test_msg OBJDIR/${subdir}/${binfile_base}
|
2009-09-01 12:06:55 +00:00
|
|
|
set bin_flags [list debug shlib=${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] != "" } {
|
2011-02-07 13:35:42 +00:00
|
|
|
untested "Could not compile ${binfile_lib1_test_msg}, ${binfile_lib2_test_msg} or ${binfile_test_msg}."
|
2009-09-01 12:06:55 +00:00
|
|
|
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.
|
2011-02-07 13:35:42 +00:00
|
|
|
untested "Could not prelink ${binfile_lib1_test_msg} or ${binfile_lib2_test_msg}."
|
2009-09-01 12:06:55 +00:00
|
|
|
return -1
|
|
|
|
}
|
|
|
|
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
set test_spawn_id [spawn_wait_for_attach $binfile]
|
|
|
|
set testpid [spawn_id_get_pid $test_spawn_id]
|
2009-09-01 12:06:55 +00:00
|
|
|
|
|
|
|
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] != ""} {
|
2011-02-07 13:35:42 +00:00
|
|
|
untested "Could not recompile ${binfile_lib1_test_msg} or ${binfile_lib2_test_msg}."
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
kill_wait_spawned_process $test_spawn_id
|
2009-09-01 12:06:55 +00:00
|
|
|
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
|
|
|
|
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
kill_wait_spawned_process $test_spawn_id
|
2012-02-21 21:55:39 +00:00
|
|
|
}}
|