https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90496
--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed with GCC 8:
Program received signal SIGSEGV, Segmentation fault.
0x0000000001ce24bb in unix_lbasename (name=0x0)
at /space/rguenther/src/svn/gcc-8-branch/libiberty/lbasename.c:53
53 for (base = name; *name; name++)
(gdb) bt
#0 0x0000000001ce24bb in unix_lbasename (name=0x0)
at /space/rguenther/src/svn/gcc-8-branch/libiberty/lbasename.c:53
#1 0x0000000001ce2561 in lbasename (name=0x0)
at /space/rguenther/src/svn/gcc-8-branch/libiberty/lbasename.c:82
#2 0x0000000000ef0468 in output_stack_usage ()
at /space/rguenther/src/svn/gcc-8-branch/gcc/toplev.c:1024
#3 0x0000000000aef817 in rest_of_handle_thread_prologue_and_epilogue ()
at /space/rguenther/src/svn/gcc-8-branch/gcc/function.c:6546
#4 0x0000000000aef865 in (anonymous
namespace)::pass_thread_prologue_and_epilogue::execute (this=0x2a7a6d0)
at /space/rguenther/src/svn/gcc-8-branch/gcc/function.c:6576
where we seem to not consider that a location might have a NULL loc.file:
(gdb) p debug_tree (current_function_decl)
<function_decl 0x7ffff669cc00 _GLOBAL__sub_I_00099_0_main
type <function_type 0x7ffff689ac78
type <void_type 0x7ffff688ef18 void VOID
align:8 warn_if_not_align:0 symtab:0 alias-set -1
structural-equality
pointer_to_this <pointer_type 0x7ffff68950a8>>
QI
size <integer_cst 0x7ffff687acd8 constant 8>
unit-size <integer_cst 0x7ffff687acf0 constant 1>
align:8 warn_if_not_align:0 symtab:0 alias-set -1 structural-equality
arg-types <tree_list 0x7ffff688daa0 value <void_type 0x7ffff688ef18
void>>
pointer_to_this <pointer_type 0x7ffff6698348>>
asm_written used static ignored uninlinable QI (null):0:0 align:8
warn_if_not_align:0 initial <block 0x7ffff66a2060>
result <result_decl 0x7ffff6884258 D.4688 type <void_type 0x7ffff688ef18
void>
VOID (null):0:0
align:8 warn_if_not_align:0 context <function_decl 0x7ffff669cc00
_GLOBAL__sub_I_00099_0_main>>
struct-function 0x7ffff66a3160>
(gdb) p current_function_decl->decl_minimal.locus
$5 = 0
(gdb) p current_function_decl->decl_common.artificial_flag
$6 = 1
it probably doesn't make sense to print stack usage for compiler-generated
functions? Or at least UNKNOWN_LOCATION should be handled better.
Eric?
Probably not really LTO specific.
