# Copyright (C) 1988, 1990, 1991, 1992, 1994 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@prep.ai.mit.edu # This file was written by Rob Savoye. (rob@cygnus.com) if $tracelevel then { strace $tracelevel } set prms_id 0 set bug_id 0 # are we on a target board if ![isnative] then { return } if [istarget "m68k*-*-hpux*"] then { # The top-level makefile passes CFLAGS= (no -g) for hp300. This probably # should be fixed (it is only needed for gcc bootstrapping, not gdb), # but until then..... setup_xfail "*-*-*" fail "cannot test self if compiled without debug info" return -1 } # Not all of the lines of code near the start of main are executed for # every machine. Also, optimization may reorder some of the lines. # So all we do is try to step or next over everything until we get # to a line that we know is always executed. proc do_steps_and_nexts {} { global prompt for {set count 0} {$count < 20} {incr count} { send "list\n" expect { -re ".*symarg = NULL.*$prompt $" { set description "step over symarg initialization" set command "step" } -re ".*execarg = NULL.*$prompt $" { set description "step over execarg initialization" set command "step" } -re ".*corearg = NULL.*$prompt $" { set description "step over corearg initialization" set command "step" } -re ".*cdarg = NULL.*$prompt $" { set description "step over cdarg initialization" set command "step" } -re ".*ttyarg = NULL.*$prompt $" { set description "step over ttyarg initialization" set command "step" } -re ".*time_at_startup = get_run_time.*$prompt $" { set description "next over get_run_time and everything it calls" set command "next" } -re ".*START_PROGRESS.*$prompt $" { set description "next over START_PROGRESS and everything it calls" set command "next" } -re ".*mac_init.*$prompt $" { set description "next over mac_init and everything it calls" set command "next" } -re ".*init_malloc.*$prompt $" { set description "next over init_malloc and everything it calls" set command "next" } -re ".*SET_TOP_LEVEL.*$prompt $" { set description "next over SET_TOP_LEVEL call" set command "next" } -re ".*cmdsize = 1.*$prompt $" { set description "step over cmdsize initialization" set command "next" } -re ".*cmdarg = .* xmalloc.*$prompt $" { set description "next over cmdarg initialization via xmalloc" set command "next" } -re ".*ncmd = 0.*$prompt $" { set description "next over ncmd initialization" set command "next" } -re ".*dirsize = 1.*$prompt $" { set description "next over dirsize initialization" set command "next" } -re ".*dirarg = .* xmalloc.*$prompt $" { return } -re "\[ \t\]+\{\r\n$prompt $" { setup_xfail "mips-*-irix5*" fail "$description ended up at odd location" } -re ".*main.c.*No such file or directory" { setup_xfail "rs6000-*-aix3*" fail "must be able to list source lines" return } default { fail "unknown source line near main" return } } send "$command\n" expect { -re ".*No such file or directory.\r\n$prompt $" { fail "$description (no source available)" } -re ".*A file or directory .* does not exist..\r\n$prompt $" { fail "$description (no source available)" } -re ".*$prompt $" { pass "$description" } timeout { fail "$description (timeout)" } } } } proc test_with_self {} { global prompt global tool global det_file global decimal global timeout # load yourself into the debugger # This can take a relatively long time, particularly for testing where # the executable is being accessed over a network, or where gdb does not # support partial symbols for a particular target and has to load the # entire symbol table. Set the timeout to 10 minutes, which should be # adequate for most environments (it *has* timed out with 5 min on a # SPARCstation SLC under moderate load, so this isn't unreasonable). # After gdb is started, set the timeout to 30 seconds for the duration # of this test, and then back to the original value. set oldtimeout $timeout set timeout 600 verbose "Timeout is now $timeout seconds" 2 if {[gdb_load "./x$tool"] <0} then { set timeout $oldtimeout verbose "Timeout is now $timeout seconds" 2 return -1 } set timeout 30 verbose "Timeout is now $timeout seconds" 2 # disassemble yourself send "x/10i main\n" expect { -re "x/10i.*main.*main.$decimal.*main.$decimal.*$prompt $"\ { pass "Disassemble main" } -re ".*$prompt $" { fail "Disassemble main" } timeout { fail "(timeout) Disassemble main" } } set description "set breakpoint at main" send "break main\n" expect { -re "Breakpoint.*at.* file.*, line.*$prompt $" { pass "$description" } -re ".*$prompt $" { fail "$description" } timeout { fail "$description (timeout)" } } # We'll need this when we send a ^C to GDB. Need to do it before we # run the program and gdb starts saving and restoring tty states. # On Ultrix, we don't need it and it is really slow (because shell_escape # doesn't use vfork). if ![istarget "*-*-ultrix*"] then { gdb_test "shell stty intr '^C'" "" \ "set interrupt character in test_with_self" } # FIXME: If we put this after the run to main, the first list # command doesn't print the same line as the current line where # gdb is stopped. send "set listsize 1\n" ; expect -re "$prompt $" # run yourself # It may take a very long time for the inferior gdb to start (lynx), # so we bump it back up for the duration of this command. set timeout 600 set description "run until breakpoint at main" send "run -nw\n" expect { -re "Starting program.*Breakpoint \[0-9\]+,.*main .argc.*argv.* at .*main.c:.*$prompt $" { pass "$description" } -re "Starting program.*Breakpoint \[0-9\]+,.*main .argc.*argv.*$prompt $" { xfail "$description (line numbers scrambled?)" } -re "vfork: No more processes.*$prompt $" { fail "$description (out of virtual memory)" set timeout $oldtimeout verbose "Timeout is now $timeout seconds" 2 return -1 } -re ".*$prompt $" { fail "$description" set timeout $oldtimeout verbose "Timeout is now $timeout seconds" 2 return -1 } timeout { fail "$description (timeout)" } } set timeout 30 verbose "Timeout is now $timeout seconds" 2 # do we have a version number ? send "print version\n" expect { -re ".\[0-9\]+ = +0x.*\[0-9.\]+.*$prompt $" { pass "printed version" } -re ".\[0-9\]+ = +.+ +0x.*\[0-9.\]+.*$prompt $" { pass "printed version with cast" } -re ".*$prompt $" { fail "printed version" } timeout { fail "(timeout) printed version" } } do_steps_and_nexts set description "print a string" send "print \"foo\"\n" expect { -re ".\[0-9\]+ = \"foo\".*$prompt $" { pass "$description" } -re ".*$prompt $" { fail "$description" } timeout { fail "$description (timeout)" } } # do_steps_and_nexts left us ready to execute an xmalloc call, # so give that a try. # If we don't actually enter the xmalloc call when we give a # step command that seems like a genuine bug. It seems to happen # on most RISC processors. setup_xfail "alpha-*-*" "hppa*-*-*" "mips-*-*" set description "step into xmalloc call" send "step\n" expect { -re "xmalloc.*size=.*at.*utils.c.*$prompt $" { pass "$description" } -re ".*No such file or directory.\r\n$prompt $" { pass "$description (no source available)" } -re "A file or directory .* does not exist..\r\n$prompt $" { pass "$description (no source available)" } -re ".*$prompt $" { fail "$description" } timeout { fail "$description (timeout)" } } # start the "xgdb" process send "continue\n" expect { -re "GNU gdb \[0-9\.\]*.* Copyright \[0-9\]* Free Software Foundation, Inc.* GDB is free software, covered by the GNU General Public License, and you are.* welcome to change it and/or distribute copies of it under certain conditions.* Type \"show copying\" to see the conditions.* There is absolutely no warranty for GDB. Type \"show warranty\" for details.* This GDB was configured as .*$prompt $"\ { pass "xgdb is at prompt" } -re "GDB is free software and you are welcome to distribute copies of it.* under certain conditions; type \"show copying\" to see the conditions..* There is absolutely no warranty for GDB; type \"show warranty\" for details..* GDB.*Copyright \[0-9\]+ Free Software Foundation, Inc..*$prompt $"\ { pass "xgdb is at prompt (obsolescent gdb)" } -re ".*$prompt $" { fail "xgdb is at prompt" } timeout { fail "(timeout) xgdb is at prompt" } } # set xgdb prompt so we can tell which is which send "set prompt (xgdb) \n" expect { -re "\[(\]xgdb\[)\].*\[(\]xgdb\[)\] $" { pass "Set xgdb prompt" } -re ".*$prompt $" { fail "Set xgdb prompt" } default { fail "(timeout) Set xgdb prompt" } } # kill the xgdb process set description "send ^C to child process" send "\003" expect { -re "Program received signal SIGINT.*$prompt $" { pass "$description" } -re ".*$prompt $" { fail "$description" } timeout { fail "$description (timeout)" } } set description "send SIGINT signal to child process" send "signal SIGINT\n" expect { -re "Continuing with signal SIGINT.*$prompt $" { pass "$description" } -re ".*$prompt $" { fail "$description" } timeout { fail "$description (timeout)" } } # get a stack trace # # This fails on some linux systems for unknown reasons. On the # systems where it fails, sometimes it works fine when run manually. # The testsuite failures may not be limited to just aout systems. setup_xfail "i*86-*-linuxaout" "hppa*-*-hpux*" set description "backtrace through signal handler" send "backtrace\n" expect { -re "#0.*read.*in main \\(.*\\) at .*main\\.c.*$prompt $" { pass "$description" } -re ".*$prompt $" { # On the alpha, we hit the infamous problem about gdb # being unable to get the frame pointer (mentioned in # gdb/README). As it is intermittent, there is no way to # XFAIL it which will give us an XPASS if the problem goes # away. setup_xfail "alpha*-*-osf*" fail "$description" } timeout { fail "$description (timeout)" } } gdb_exit # Set the timeout back to the value it had when we were called. set timeout $oldtimeout verbose "Timeout is now $timeout seconds" 2 # Restart gdb in case next test expects it to be started already. gdb_start return 0 } # Find a pathname to a file that we would execute if the shell was asked # to run $arg using the current PATH. proc find_gdb { arg } { # If the arg directly specifies an existing executable file, then # simply use it. if [file executable $arg] then { return $arg } set result [which $arg] if [string match "/" [ string range $result 0 0 ]] then { return $result } # If everything fails, just return the unqualified pathname as default # and hope for best. return $arg } # Run the test with self. # Copy the file executable file in case this OS doesn't like to edit it's own # text space. set GDB_FULLPATH [find_gdb $GDB] # Remove any old copy lying around. catch "exec rm -f ./x$tool" if ![file executable $GDB_FULLPATH] then { fail "couldn't convert $GDB to absolute pathname to make local copy" } else { if [catch "exec cp $GDB_FULLPATH ./x$tool"] then { fail "couldn't copy $GDB_FULLPATH to current directory" } else { verbose "\t\tCopied $GDB_FULLPATH to ./x$tool" if {[test_with_self] <0} then { warning "Couldn't test self" catch "exec rm -f ./x$tool" return -1 } catch "exec rm -f ./x$tool" } }