OpenE2K/binutils-gdb
12
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

* blockframe.c (create_new_frame, get_prev_frame_info):

Browse Source 
Use the function name when calling IN_SIGTRAMP.
	* config/m68k/tm-m68k.h (SIG_PC_FP_OFFSET, SIG_SP_FP_OFFSET):
	Define for correct handling of bachtraces through _sigtramp.
	* m68k-tdep.c (m68k_find_saved_regs):  Adjust saved sp for fake
	sigtramp frames.
	* mipsread.c (parse_type):  Handle corrupt TIR info with complaint
	instead of core dump.
	* mipsread.c (parse_partial_symbols):  Put static symbols into the
	mimimal symbol table, use proper mst_types for all minimal symbols.
	* stack.c (frame_info, print_frame_info):  Use the starting source
	line number on a line boundary if backtracing through sigtramp.
This commit is contained in:
Peter Schauer 1993-08-16 10:03:57 +00:00
parent 4fd0b84ea3
commit dff84871e9
1 changed files with 31 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
gdb/stack.c
View File

@ -161,8 +161,29 @@ print_frame_info (fi, level, source, args)
register char *funname = 0;
enum language funlang = language_unknown;
int numargs;
char buf[MAX_REGISTER_RAW_SIZE];
CORE_ADDR sp;
if (PC_IN_CALL_DUMMY (fi->pc, read_sp (), fi->frame))
/* Get the value of SP_REGNUM relative to the frame. */
get_saved_register (buf, (int *)NULL, (CORE_ADDR *)NULL,
FRAME_INFO_ID (fi), SP_REGNUM, (enum lval_type *)NULL);
sp = extract_address (buf, REGISTER_RAW_SIZE (SP_REGNUM));
/* This is not a perfect test, because if a function alloca's some
memory, puts some code there, and then jumps into it, then the test
will succeed even though there is no call dummy. A better
solution would be to keep track of where the call dummies are.
Probably the best way to do that is by setting a breakpoint.c
breakpoint at the end of the call dummy (wanted anyway, to clean
up wait_for_inferior). Then we know that the sizeof (CALL_DUMMY)
(or some such) bytes before that breakpoint are a call dummy.
Only problem I see with this approach is figuring out to get rid
of the breakpoint whenever the call dummy vanishes (e.g.
return_command, or longjmp out of the called function), which we
probably can solve (it's very similar to figuring out when a
watchpoint on a local variable goes out of scope if it is being
watched via something like a 386 debug register). */
if (PC_IN_CALL_DUMMY (fi->pc, sp, fi->frame))
{
/* Do this regardless of SOURCE because we don't have any source
to list for this frame. */
@ -181,7 +202,8 @@ print_frame_info (fi, level, source, args)
return;
}
sal = find_pc_line (fi->pc, fi->next != NULL);
sal = find_pc_line (fi->pc,
fi->next != NULL && fi->next->signal_handler_caller == 0);
func = find_pc_function (fi->pc);
if (func)
{
@ -191,13 +213,15 @@ print_frame_info (fi, level, source, args)
is compiled with debugging symbols, and the "foo.o" symbol
that is supposed to tell us where the file with debugging symbols
ends has been truncated by ar because it is longer than 15
characters).
characters). This also occurs if the user uses asm() to create
a function but not stabs for it (in a file compiled -g).
So look in the minimal symbol tables as well, and if it comes
up with a larger address for the function use that instead.
I don't think this can ever cause any problems; there shouldn't
be any minimal symbols in the middle of a function.
FIXME: (Not necessarily true. What about text labels) */
be any minimal symbols in the middle of a function; if this is
ever changed many parts of GDB will need to be changed (and we'll
create a find_pc_minimal_function or some such). */
struct minimal_symbol *msymbol = lookup_minimal_symbol_by_pc (fi->pc);
if (msymbol != NULL
@ -417,7 +441,8 @@ frame_info (addr_exp, from_tty)
error ("Invalid frame specified.");
fi = get_frame_info (frame);
sal = find_pc_line (fi->pc, fi->next != NULL);
sal = find_pc_line (fi->pc,
fi->next != NULL && fi->next->signal_handler_caller == 0);
func = get_frame_function (frame);
s = find_pc_symtab(fi->pc);
if (func)