This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: [tree-ssa] Bootstrap broken on PPC
> >Just a note, the changes below don't actually make it more aggressive
> >(those were committed a few days ago, and they were bootstrapped on
> >ppc64), they just speed it up.
>
> I also just re-bootstrapped a tree without the changes in Diego's
> changelog, except for the tree-ssa-pre.c changes.
> It bootstraps fine on ppc64.
By inspecting the patches I notticed the following pasto. It may affect
64bit targets where we do have more conversions and we theoretically can
manage to create nested conversion... I will commit obvious patch.
I will try to build crosscompiler and try to reproduce it there now.
x86_64 bootstrap seems to pass this file with no problems.
Honza
Index: fold-const.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/fold-const.c,v
retrieving revision 1.213.2.58
diff -c -3 -p -r1.213.2.58 fold-const.c
*** fold-const.c 7 Nov 2003 13:02:06 -0000 1.213.2.58
--- fold-const.c 8 Nov 2003 18:19:32 -0000
*************** nondestructive_fold_binary_to_constant (
*** 9145,9151 ****
&& (TREE_CODE (TREE_OPERAND (op0, 0)) == ADDR_EXPR
|| (TREE_CODE (TREE_OPERAND (op0, 0)) == NOP_EXPR
&& (TREE_CODE (TREE_OPERAND (TREE_OPERAND (op0, 0), 0))
! == NOP_EXPR)))
&& TREE_CODE (TREE_OPERAND (op0, 1)) == INTEGER_CST)
{
return build (PLUS_EXPR, type, TREE_OPERAND (op0, 0),
--- 9145,9151 ----
&& (TREE_CODE (TREE_OPERAND (op0, 0)) == ADDR_EXPR
|| (TREE_CODE (TREE_OPERAND (op0, 0)) == NOP_EXPR
&& (TREE_CODE (TREE_OPERAND (TREE_OPERAND (op0, 0), 0))
! == ADDR_EXPR)))
&& TREE_CODE (TREE_OPERAND (op0, 1)) == INTEGER_CST)
{
return build (PLUS_EXPR, type, TREE_OPERAND (op0, 0),