public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c++/14167] New: Unneeded C++ types are output in debug info due to use of static constants
@ 2004-02-17 4:50 drow at gcc dot gnu dot org
2004-02-17 5:07 ` [Bug c++/14167] " pinskia at gcc dot gnu dot org
` (2 more replies)
0 siblings, 3 replies; 9+ messages in thread
From: drow at gcc dot gnu dot org @ 2004-02-17 4:50 UTC (permalink / raw)
To: gcc-bugs
Take this testcase:
class Class1 {
static const int var1 = 1;
static const int var2 = var1;
};
class Class2 {
static const int var1 = 1;
static const int var2 = 1;
};
Compile it with -gdwarf-2. Debug information for Class2 will not be emitted,
which is OK, because Class2 is unused. Debug information for Class1 will be
emitted, because Class1::var1 is marked as used in the initializer for Class1::var2,
which causes Class1::var1 to be emitted, which causes all of Class1 to be emitted.
It gets marked used in finish_id_expression as I expected:
2581 if (TREE_CODE (decl) == VAR_DECL
2582 || TREE_CODE (decl) == PARM_DECL
2583 || TREE_CODE (decl) == RESULT_DECL)
2584 mark_used (decl);
Is the used marker necessary in integral constant-expressions? Where else?
This is the source of about 60% of the (60k!) debug information emitted for the
"trivial" C++ program:
#include <iostream>
int main() { return 0; }
--
Summary: Unneeded C++ types are output in debug info due to use
of static constants
Product: gcc
Version: 3.4.0
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: drow at gcc dot gnu dot org
CC: gcc-bugs at gcc dot gnu dot org
GCC build triplet: i686-pc-linux-gnu
GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread[parent not found: <bug-14167-3264@http.gcc.gnu.org/bugzilla/>]
* [Bug c++/14167] Unneeded C++ types are output in debug info due to use of static constants
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
@ 2006-01-21 16:27 ` drow at gcc dot gnu dot org
2006-01-21 16:37 ` drow at gcc dot gnu dot org
` (3 subsequent siblings)
4 siblings, 0 replies; 9+ messages in thread
From: drow at gcc dot gnu dot org @ 2006-01-21 16:27 UTC (permalink / raw)
To: gcc-bugs
------- Comment #3 from drow at gcc dot gnu dot org 2006-01-21 16:27 -------
I don't have G++ 3.4 handy at the moment. 3.3 emits about three and a half
times as large of a .debug_info section as 4.0, however, so I believe this is
fixed.
There's still lots of unneeded debug information output, but I believe I've got
another open PR describing that.
--
drow at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread* [Bug c++/14167] Unneeded C++ types are output in debug info due to use of static constants
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
2006-01-21 16:27 ` drow at gcc dot gnu dot org
@ 2006-01-21 16:37 ` drow at gcc dot gnu dot org
2006-01-29 18:12 ` pinskia at gcc dot gnu dot org
` (2 subsequent siblings)
4 siblings, 0 replies; 9+ messages in thread
From: drow at gcc dot gnu dot org @ 2006-01-21 16:37 UTC (permalink / raw)
To: gcc-bugs
------- Comment #4 from drow at gcc dot gnu dot org 2006-01-21 16:37 -------
Oops, not so clear after all.
Debug information for Class1 is no longer emitted, but debug information for
Class1::var1 is still emitted. It just doesn't trigger the output of the
containing class any more.
If I add a real use of var1:
class Class1 {
public:
static const int var1 = 1;
static const int var2 = var1;
};
class Class2 {
static const int var1 = 1;
static const int var2 = 1;
};
int main(void)
{
return Class1::var1;
}
Still no change. We get a declaration for var1, but no definition for it, and
not even a declaration of the containing class. It looks like this:
<1><244>: Abbrev Number: 5 (DW_TAG_variable)
DW_AT_name : var1
DW_AT_decl_file : 1
DW_AT_decl_line : 3
DW_AT_MIPS_linkage_name: _ZN6Class14var1E
DW_AT_type : <219>
DW_AT_external : 1
DW_AT_declaration : 1
DW_AT_const_value : 1
That's pretty useless. It doesn't let a debugger print Class1::var1.
--
drow at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |REOPENED
Resolution|FIXED |
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread* [Bug c++/14167] Unneeded C++ types are output in debug info due to use of static constants
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
2006-01-21 16:27 ` drow at gcc dot gnu dot org
2006-01-21 16:37 ` drow at gcc dot gnu dot org
@ 2006-01-29 18:12 ` pinskia at gcc dot gnu dot org
2006-08-22 19:35 ` jason at gcc dot gnu dot org
2006-08-22 20:03 ` drow at gcc dot gnu dot org
4 siblings, 0 replies; 9+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2006-01-29 18:12 UTC (permalink / raw)
To: gcc-bugs
--
pinskia at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|REOPENED |NEW
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread* [Bug c++/14167] Unneeded C++ types are output in debug info due to use of static constants
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
` (2 preceding siblings ...)
2006-01-29 18:12 ` pinskia at gcc dot gnu dot org
@ 2006-08-22 19:35 ` jason at gcc dot gnu dot org
2006-08-22 20:03 ` drow at gcc dot gnu dot org
4 siblings, 0 replies; 9+ messages in thread
From: jason at gcc dot gnu dot org @ 2006-08-22 19:35 UTC (permalink / raw)
To: gcc-bugs
------- Comment #5 from jason at gcc dot gnu dot org 2006-08-22 19:34 -------
[basic.def.odr] says that an object that appears in an integral constant
expression is not thereby "used", and it seems reasonable to extend that to not
setting TREE_USED.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread* [Bug c++/14167] Unneeded C++ types are output in debug info due to use of static constants
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
` (3 preceding siblings ...)
2006-08-22 19:35 ` jason at gcc dot gnu dot org
@ 2006-08-22 20:03 ` drow at gcc dot gnu dot org
4 siblings, 0 replies; 9+ messages in thread
From: drow at gcc dot gnu dot org @ 2006-08-22 20:03 UTC (permalink / raw)
To: gcc-bugs
------- Comment #6 from drow at gcc dot gnu dot org 2006-08-22 20:03 -------
The use in main should still set TREE_USED, though; should TREE_USED on a
member ought to cause the class to be emitted?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14167
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2006-08-22 20:03 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2004-02-17 4:50 [Bug c++/14167] New: Unneeded C++ types are output in debug info due to use of static constants drow at gcc dot gnu dot org
2004-02-17 5:07 ` [Bug c++/14167] " pinskia at gcc dot gnu dot org
2004-08-11 22:10 ` pinskia at gcc dot gnu dot org
2005-05-12 2:06 ` pinskia at gcc dot gnu dot org
[not found] <bug-14167-3264@http.gcc.gnu.org/bugzilla/>
2006-01-21 16:27 ` drow at gcc dot gnu dot org
2006-01-21 16:37 ` drow at gcc dot gnu dot org
2006-01-29 18:12 ` pinskia at gcc dot gnu dot org
2006-08-22 19:35 ` jason at gcc dot gnu dot org
2006-08-22 20:03 ` drow at gcc dot gnu dot org
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).