forked from Imagelibrary/binutils-gdb
d8b79a1f24
Cygwin debugging does not support follow fork. There is currently no
interface between the debugger and the Cygwin runtime to be able to
intercept forks and execs. Consequently, testcases that try to
exercise fork/exec all FAIL, and several hit long cascading timeouts.
Add a new allow_fork_tests procedure, meant be be used with require,
and sprinkle it throughout testcases that exercise fork.
Note that some tests currently are skipped on targets other than
Linux, with something like:
# Until "set follow-fork-mode" and "catch vfork" are implemented on
# other targets...
#
if {![istarget "*-linux*"]} {
continue
}
However, some BSD ports also support fork debugging nowadays, and the
testcases were never adjusted... That is why the new allow_fork_tests
procedure doesn't look for linux.
With this patch, on Cygwin, I get this:
$ make check TESTS="*/*fork*.exp"
...
=== gdb Summary ===
# of expected passes 6
# of untested testcases 1
# of unsupported tests 31
Change-Id: I0c5e8c574d1f61b28d370c22a0b0b6bc3efaf978
115 lines
4.3 KiB
Plaintext
115 lines
4.3 KiB
Plaintext
# Copyright 2020-2025 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 handling a vfork while another inferior is running. The bug that
|
|
# prompted writing this test case was in the Linux native target. The target
|
|
# assumed that the vfork-done event it received was for the current inferior
|
|
# (an invalid assumption, the current inferior is the one randomly selected by
|
|
# do_target_wait (at the time of writing). This caused the target to drop the
|
|
# vfork-done event, because it was seen as unneeded and to restart the thread
|
|
# as if nothing happened. This however resulted in the thread running with
|
|
# breakpoints not inserted.
|
|
#
|
|
# To catch the bug, this test verifies that we can hit a breakpoint after a
|
|
# vfork call, while a second inferior runs in the background.
|
|
|
|
require allow_fork_tests
|
|
|
|
require !use_gdb_stub
|
|
|
|
standard_testfile .c -sleep.c
|
|
|
|
set srcfile_sleep $srcfile2
|
|
set binfile_sleep ${binfile}-sleep
|
|
|
|
# The reproducibility of the bug depends on which inferior randomly selects in
|
|
# do_target_wait when consuming the vfork-done event. Since GDB doesn't call
|
|
# srand(), we are likely to always see the same sequence of inferior selected by
|
|
# do_target_wait, which can hide the bug if you are not "lucky". To work
|
|
# around that, call vfork and hit the breakpoint in a loop, it makes it
|
|
# somewhat likely that the wrong inferior will be selected eventually.
|
|
set nr_loops 20
|
|
|
|
# Compile the main program that calls vfork and hits a breakpoint.
|
|
set opts [list debug additional_flags=-DNR_LOOPS=$nr_loops]
|
|
if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable \
|
|
$opts] != "" } {
|
|
untested "failed to compile"
|
|
return -1
|
|
}
|
|
|
|
# Compile the secondary program, which just sleeps.
|
|
if { [gdb_compile "${srcdir}/${subdir}/${srcfile_sleep}" "${binfile_sleep}" executable \
|
|
{debug}] != "" } {
|
|
untested "failed to compile"
|
|
return -1
|
|
}
|
|
|
|
# We exercise two methods of getting a second inferior to execute while the
|
|
# first one vforks. METHOD can be:
|
|
#
|
|
# - non-stop: start GDB with non-stop on and run the second inferior in
|
|
# background.
|
|
# - schedule-multiple: set "schedule-multiple on", this will make "continue"
|
|
# resume both inferiors.
|
|
proc do_test {method} {
|
|
save_vars { ::GDBFLAGS } {
|
|
if { $method == "non-stop" } {
|
|
append ::GDBFLAGS " -ex \"set non-stop on\""
|
|
}
|
|
clean_restart
|
|
}
|
|
|
|
# Start the second inferior in background.
|
|
gdb_test "add-inferior" "Added inferior 2.*"
|
|
gdb_test "inferior 2" "Switching to inferior 2 .*"
|
|
gdb_file_cmd ${::binfile_sleep}
|
|
if { $method == "non-stop" } {
|
|
gdb_test "run &" "Starting program: .*" "run inferior 2"
|
|
} else {
|
|
gdb_test "start" "Temporary breakpoint $::decimal, main .*" \
|
|
"start inferior 2"
|
|
}
|
|
|
|
# Start the first inferior.
|
|
gdb_test "inferior 1" "Switching to inferior 1 .*"
|
|
gdb_file_cmd ${::binfile}
|
|
gdb_test "break should_break_here" "Breakpoint $::decimal at .*"
|
|
gdb_test "start" "Thread 1.1 .* hit Temporary breakpoint.*" \
|
|
"start inferior 1"
|
|
|
|
# Only enable schedule-multiple this late, because of:
|
|
# https://sourceware.org/bugzilla/show_bug.cgi?id=28777
|
|
if { $method == "schedule-multiple" } {
|
|
gdb_test_no_output "set schedule-multiple on"
|
|
}
|
|
|
|
|
|
# Continue over vfork and until the breakpoint. The number of loops here
|
|
# matches the number of loops in the program. So if a breakpoint is missed
|
|
# at some point, a "continue" will wrongfully continue until the end of the
|
|
# program, which will fail the test.
|
|
for {set i 0} {$i < $::nr_loops} {incr i} {
|
|
with_test_prefix "i=$i" {
|
|
gdb_test "continue" \
|
|
"Thread 1.1 .* hit Breakpoint $::decimal, should_break_here.*"
|
|
}
|
|
}
|
|
}
|
|
|
|
foreach_with_prefix method {schedule-multiple non-stop} {
|
|
do_test $method
|
|
}
|