[Bug java/18190] New: [4.0 regression] primitive array optimization is gone

[Bug java/18190] New: [4.0 regression] primitive array optimization is gone

[tromey at gcc dot gnu dot org]

Sample source:

public class t
{
  public static final int[] x = { 5, 7, 9, 11 };
}

In gcc 3.3 (and 3.4 and earlier than 3.3), this code compiled
to something like:

.LJv0.0:
	.long	_Jv_intVTable
	.long	4
	.long	5
	.long	7
	.long	9
	.long	11

Now we actually emit code for this.

The culprits are a combination of this patch to gcj:

2004-07-08  Richard Henderson  <rth@redhat.com>

	* expr.c (case_identity, get_primitive_array_vtable,
	java_expand_expr, emit_init_test_initialization): Remove.
	* java-tree.h (java_expand_expr): Remove.
	* lang.c (LANG_HOOKS_EXPAND_EXPR): Remove.

and this patch to libgcj (which removed the primitive vtables):

2004-07-23  Bryce McKinlay  <mckinlay@redhat.com>

	* prims.cc (_Jv_InitPrimClass): Don't create an array class.
	(_Jv_CreateJavaVM): Don't pass array vtable parameter to
	_Jv_InitPrimClass.
	(DECLARE_PRIM_TYPE): Don't declare array vtables.
	* include/jvm.h (struct _Jv_ArrayVTable): Removed.
	* java/lang/Class.h (_Jv_InitPrimClass): Update friend declaration.

-- 
           Summary: [4.0 regression] primitive array optimization is gone
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: java
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: tromey at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org,java-prs at gcc dot gnu
                    dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-10-28 01:39 -------
Confirmed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |missed-optimization
   Last reconfirmed|0000-00-00 00:00:00         |2004-10-28 01:39:42
               date|                            |
   Target Milestone|---                         |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-11-04 04:53 -------
Some of the code from java_expand_expr should be moved into java_gimplify_new_array_init.
And maybe fix up libjava code or maybe just storing the array's vtable by using a store.  I might look 
into doing this tomorrow.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2004-10-28 01:39:42         |2004-11-04 04:53:33
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[mckinlay at redhat dot com]

------- Additional Comments From mckinlay at redhat dot com  2004-11-04 18:46 -------
Here's my thoughts about this:

- This optimization only ever worked for source compilation. Bytecode compilers
always emit array initializers as code, so for byte compilation it makes no
difference.

- I don't see a strong reason not to reference the vtable symbols, but if we
decide to remove my patch then we need to be careful that the original bug
remains fixed - see: http://gcc.gnu.org/ml/java-patches/2004-q3/msg00343.html

- Is this optimization really worth worrying about? I'm pretty sure that
performance-wise, the difference is insignificant - binary size is what we
should be concerned about here. Is a binary that initializes arrays in code
significantly larger?



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-11-04 21:19 -------
Note I attached the front-end part so the only thing left is to fix up libjava to emit the vtables again.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-11-04 23:06 -------
I finished up porting the other part of the expand part.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |minor
           Priority|P2                          |P3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0 regression] primitive array optimization is gone

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-01-19 18:52 -------
Ada and Java bugs are not release-critical; therefore, I've removed the target
milsetone.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.0.0                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0/4.1 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0/4.1 regression] primitive array optimization is gone

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-08-22 01:47 -------
Removing target milestone per last comment.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|4.1.0                       |---


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190

[Bug java/18190] [4.0/4.1 regression] primitive array optimization is gone

[pinskia at gcc dot gnu dot org]

-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P5
   Target Milestone|---                         |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18190