From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cagney To: Mark Kettenis Cc: gdb-patches@sourceware.cygnus.com, cagney@cygnus.com Subject: Re: [RFA] Testsuite addition for x86 linux GDB and SIGALRM fix Date: Wed, 27 Jun 2001 22:22:00 -0000 Message-id: <3B3ABD6E.1040304@cygnus.com> References: <200005192321.e4JNLEv13368@delius.kettenis.local> X-SW-Source: 2001-06/msg00450.html Anyone? Andrew > Here's the test I promised Andrew a while ago for the fix for the > problem reported by Jonathan Larmour: > > http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html > > The fix has already been checked in, the problem is still mentioned in > the TODO file (let's keep it there until this test has been added). > > I verified that some of these tests (the "stepi" and "nexti" tests) > do fail without my fix to infrun.c. > > I'm not sure to what extent the use of setitimer() is portable. > However, it is hard to come up with a test that doesn't use it. > > > 2000-05-20 Mark Kettenis > > Add tests for stepping with pending signals. > * gdb.base/step-alarm.exp: New file. > * gdb.base/step-alarm.c: New file. > > > --- /dev/null Thu Feb 19 16:30:24 1998 > +++ testsuite/gdb.base/step-alarm.exp Sat May 20 01:09:56 2000 > @@ -0,0 +1,226 @@ > +# Copyright (C) 1997, 1998, 1999, 2000 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 2 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, write to the Free Software > +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ > + > +# Please email any bugs, comments, and/or additions to this file to: > +# bug-gdb@gnu.org > + > +# use this to debug: > +# > +#log_user 1 > + > +# step-alarm.exp -- Expect script to test stepping in gdb > +# with a pending signals > + > +# Most of this script is copied over from step-test.exp. We run (almost) > +# the same tests, except that we set a timer and install a SIGALRM signal > +# handler. The addition of these tests was prompted by the following fix: > +# > +# 2000-05-01 Mark Kettenis > +# > +# * infrun.c (handle_inferior_event): Add missing call to keep_going > +# and missing return when handling an ordinary signal from the > +# inferior. > +# > +# for a problem where "stepi" didn't make any progress if a signal > +# was pending. > + > +if $tracelevel then { > + strace $tracelevel > +} > + > +set testfile step-alarm > +set srcfile ${srcdir}/${subdir}/${testfile}.c > +set binfile ${objdir}/${subdir}/${testfile} > + > +remote_exec build "rm -f ${binfile}" > +if { [gdb_compile "${srcfile}" "${binfile}" executable {debug}] != "" } { > + gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail." > +} > + > +gdb_exit > +gdb_start > +gdb_reinitialize_dir $srcdir/$subdir > +gdb_load ${binfile} > + > +if ![runto_main] then { > + fail "Can't run to main" > + return 0 > +} > + > +# Make sure that the signal handler is installed, and the timer is set. > +# > +gdb_test "break [gdb_get_line_number "w = 0"]" \ > + ".*Breakpoint.* at .*" \ > + "set breakpoint after timer activation" > +gdb_test "continue" \ > + ".*Breakpoint ${decimal},.*w = 0.*" \ > + "run until timer is activated" > + > + > +# Set a breakpoint at line 57, if stepi then finish fails, we would > +# run to the end of the program, which would mess up the rest of the tests. > + > +# Vanilla step/next > +# > +gdb_test "next" ".*${decimal}.*x = 1;.*" "next 1" > +gdb_test "step" ".*${decimal}.*y = 2;.*" "step 1" > + > +# With count > +# > +gdb_test "next 2" ".*${decimal}.*w = w.*2;.*" "next 2" > +gdb_test "step 3" ".*${decimal}.*z = z.*5;.*" "step 3" > +gdb_test "next" ".*${decimal}.*callee.*OVER.*" "next 3" > + > +# Step over call > +# > +gdb_test "next" ".*${decimal}.*callee.*INTO.*" "next over" > + > +# Step into call > +# > +gdb_test "step" ".*${decimal}.*myglob.*" "step into" > + > +# Step out of call > +# > +# I wonder if this is really portable. Are there any caller-saves > +# platforms, on which `finish' will return you to some kind of pop > +# instruction, which is attributed to the line containing the function > +# call? > + > +# On PA64, we end up at a different instruction than PA32. > +# On IA-64, we also end up on callee instead of on the next line due > +# to the restoration of the global pointer (which is a caller-save). > +if { [istarget "hppa2.0w-hp-hpux*"] || [istarget "ia64-*-*"]} { > + send_gdb "finish\n" > + gdb_expect { > + -re ".*${decimal}.*a.*5.*= a.*3.*$gdb_prompt $" { pass "step out 1" } > + -re ".*${decimal}.*callee.*INTO.*$gdb_prompt $" { pass "step out 2" } > + timeout { fail "step out" } > + } > +} else { > + gdb_test "finish" ".*${decimal}.*a.*5.*= a.*3.*" "step out" > +} > + > +### Testing nexti and stepi. > +### > +### test_i NAME COMMAND HERE THERE > +### > +### Send COMMAND to gdb over and over, while the output matches the > +### regexp HERE, followed by the gdb prompt. Pass if the output > +### eventually matches the regexp THERE, followed by the gdb prompt; > +### fail if we have to iterate more than a hundred times, we time out > +### talking to gdb, or we get output which is neither HERE nor THERE. :) > +### > +### Use NAME as the name of the test. > +### > +### The exact regexps used are "$HERE.*$gdb_prompt $" > +### and "$THERE.*$gdb_prompt $" > +### > +proc test_i {name command here there} { > + global gdb_prompt > + > + set i 0 > + while 1 { > + send_gdb "${command}\n" > + gdb_expect { > + -re "$here.*$gdb_prompt $" { > + # Okay, we're still on the same line. Just step again. > + } > + -re "$there.*$gdb_prompt $" { > + # We've reached the next line. Rah. > + pass "$name" > + return > + } > + -re "$gdb_prompt $" { > + # We got something else. Fail. > + fail "$name" > + return > + } > + timeout { > + fail "$name (timeout)" > + return > + } > + } > + > + # Have we gone for too many steps without seeing any progress? > + if {[incr i] >= 100} { > + fail "$name (no progress after 100 steps)" > + return > + } > + } > +} > + > +test_i "stepi to next line" "stepi" \ > + ".*${decimal}.*a.*5.* = a.*3" \ > + ".*${decimal}.*callee.*STEPI" > + > +test_i "stepi into function" "stepi" \ > + ".*${decimal}.*callee.*STEPI" \ > + ".*callee \\(\\) at .*step-alarm\\.c" > + > +# Continue to step until we reach the function's body. This makes it > +# more likely that we've actually completed the prologue, so "finish" > +# will work. > +test_i "stepi into function's first source line" "stepi" \ > + ".*${decimal}.*\\{" \ > + ".*${decimal}.*myglob" > + > +# Have to be careful here, if the finish does not work, > +# then we may run to the end of the program, which > +# will cause erroneous failures in the rest of the tests > +send_gdb "finish\n" > +gdb_expect { > + -re ".*(Program received|Program exited).*$gdb_prompt $" { > + # Oops... We ran to the end of the program... Better reset > + if {![runto_main]} then { > + fail "Can't run to main" > + return 0 > + } > + if {![runto step-alarm.c:57]} { > + fail "Can't run to line 57" > + return 0 > + } > + fail "stepi: finish call" > + } > + -re ".*${decimal}.*callee.*NEXTI.*$gdb_prompt $" { > + pass "stepi: finish call" > + } > + -re ".*${decimal}.*callee.*STEPI.*$gdb_prompt $" { > + # On PA64, we end up at a different instruction than PA32. > + # On IA-64, we end up on callee instead of on the following line due > + # to the restoration of the global pointer. > + if { [istarget "hppa2.0w-hp-hpux*"] || [istarget "ia64-*-*"] } { > + pass "stepi: finish call 2" > + } else { > + fail "stepi: finish call 2" > + return > + } > + } > + -re "$gdb_prompt $" { > + # We got something else. Fail. > + fail "stepi: finish call" > + return > + } > + timeout { > + fail "stepi: finish call (timeout)" > + return > + } > +} > + > +test_i "nexti over function" "nexti" \ > + ".*${decimal}.*callee.*NEXTI" \ > + ".*${decimal}.*y = w \\+ z;" > + > +return 0 > --- /dev/null Thu Feb 19 16:30:24 1998 > +++ testsuite/gdb.base/step-alarm.c Sat May 20 01:01:21 2000 > @@ -0,0 +1,65 @@ > +#include > +#include > +#include > + > +/* Test stepping with a pending signal. */ > + > +int myglob = 0; > + > +int callee (void) > +{ > + myglob++; > + return 0; > +} > + > +static void > +handler (int signum) > +{ > +} > + > +int > +main (void) > +{ > + struct sigaction sa; > + struct itimerval it; > + int w, x, y, z; > + int a[10]; > + > + sa.sa_handler = handler; > + sigemptyset (&sa.sa_mask); > + sa.sa_flags = 0; > + sigaction (SIGALRM, &sa, NULL); > + > + it.it_interval.tv_usec = 5000; > + it.it_interval.tv_sec = 0; > + it.it_value.tv_usec = 5000; > + it.it_value.tv_sec = 0; > + setitimer (ITIMER_REAL, &it, NULL); > + > + /* Test "next" and "step" */ > + w = 0; > + x = 1; > + y = 2; > + z = 3; > + w = w + 2; > + x = x + 3; > + y = y + 4; > + z = z + 5; > + > + /* Test that "next" goes over a call */ > + callee(); /* OVER */ > + > + /* Test that "step" doesn't */ > + callee(); /* INTO */ > + > + /* Test "stepi" */ > + a[5] = a[3] - a[4]; > + callee(); /* STEPI */ > + > + /* Test "nexti" */ > + callee(); /* NEXTI */ > + > + y = w + z; > + > + exit (0); > +} > >