On 03/10/13 17:17, Tom de Vries wrote:
> we need to emit it before the FUNCTION_BEG insn-note
> (rough proof-of-concept patch attached), such that no .loc will be generated
> for
> it.
I investigated further, and now I think it's a regression caused by the fix for
PR47028.
Attached patch works f
Richard,
( see also related discussion
http://gcc.gnu.org/ml/gcc-patches/2010-12/msg01570.html )
Consider break.c (minimized from gdb/testsuite/gdb.base/break.c):
...
void *v;
void a (void *x) { }
void b (void) { }
int
main (int argc)
{
if (argc == 12345)
{
a (v);
return 1;
