# This testcase is part of GDB, the GNU debugger. # Copyright 2004-2017 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 . # Check that GDB can trigger and backtrace SIGSEGV signal stacks # caused by both accessing (data) and executing (code) at address # zero. # On function descriptor architectures, a zero descriptor, instead of # a NULL pointer, is used. That way the NULL code test always # contains a zero code reference. # For recovery, sigjmp/longjmp are used. # This also tests backtrace/gdb1476. if [target_info exists gdb,nosignals] { verbose "Skipping signull.exp because of nosignals." continue } standard_testfile .c if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } { untested signull.exp return -1 } clean_restart ${binfile} # # Run to `main' where we begin our tests. # if ![runto_main] then { gdb_suppress_tests } # If we can examine what's at memory address 0, it is possible that we # could also execute it. This could probably make us run away, # executing random code, which could have all sorts of ill effects, # especially on targets without an MMU. Don't run the tests in that # case. if { [is_address_zero_readable] } { untested "Memory at address 0 is possibly executable" return } # If an attempt to call a NULL pointer leaves the inferior in main, # then function pointers are descriptors, probe this and remember the # result. gdb_test_no_output "set test = code_entry_point" \ "set for function pointer probe" set test "probe function pointer" set function_pointer code_entry_point set signame "SIGSEGV" gdb_test_multiple "continue" "$test" { -re "Program received signal SIGSEGV.*bowler .*$gdb_prompt $" { set function_pointer code_descriptor pass "$test (function descriptor)" } -re "Program received signal SIGSEGV.*0.*$gdb_prompt $" { pass "$test (function entry-point)" } -re "Program received signal SIGBUS.*0.*$gdb_prompt $" { set signame SIGBUS pass "$test (function entry-point)" } } # Re-start from scratch, breakpoint the bowler so that control is # regained after each test, and run up to that. rerun_to_main gdb_test "break bowler" gdb_test "break keeper" # By default Stop:Yes Print:Yes Pass:Yes gdb_test "handle SIGSEGV" "SIGSEGV.*Yes.*Yes.*Yes.*Segmentation fault" gdb_test "handle SIGBUS" "SIGBUS.*Yes.*Yes.*Yes.*Bus error" # For the given signal type, check that: the SIGSEGV occures; a # backtrace from the SEGV works; the sigsegv is delivered; a backtrace # through the SEGV works. proc test_segv { name tag bt_from_segv bt_from_keeper } { global signame gdb_test continue "Breakpoint.* bowler.*" "${name}; start with the bowler" # NB: Don't use $tag in the testname - changes across systems. gdb_test_no_output "set test = $tag" "${name}; select the pointer type" gdb_test continue "Program received signal ${signame}.*" \ "${name}; take the ${signame}" gdb_test backtrace $bt_from_segv "${name}; backtrace from ${signame}" gdb_test continue "Breakpoint.* keeper.*" "${name}; continue to the keeper" gdb_test backtrace $bt_from_keeper "${name}; backtrace from keeper through ${signame}" } test_segv "data read" data_read \ {#0 .* bowler .*#1 .* main .*} \ {#0 .* keeper .*#1 .* handler .*#2 .* bowler .*#3 .* main .*} test_segv "data write" data_write \ {#0 .* bowler .*#1 .* main .*} \ {#0 .* keeper .*#1 .* handler .*#2 .* bowler .*#3 .* main .*} test_segv code $function_pointer \ {#0 .* 0x0+ .*#1 .* bowler .*#2 .* main .*} \ {#0 .* keeper .*#1 .* handler .*#2 .* 0x0+ .*#3 .* bowler .*#4 .* main .*}