Commit 0b45fd7a by Richard Guenther Committed by Richard Biener

re PR tree-optimization/37005 (GNAT Bug Box for cd2a24e.adb:37 at tree-vrp.c:392)

2008-08-28  Richard Guenther  <rguenther@suse.de>

	PR middle-end/37005
	* fold-const.c (maybe_canonicalize_comparison_1): Require
	undefined overflow only for canonicalizing A +- CST cmp CST.
	Make sure to not generate new constants that are not inside
	their TYPE_MIN/MAX_VALUE range.
	(maybe_canonicalize_comparison): Remove undefined overflow
	checking from here.
	(fold_binary): Remove now duplicate folding.

From-SVN: r139702
parent 4717882f
2008-08-28 Richard Guenther <rguenther@suse.de>
PR middle-end/37005
* fold-const.c (maybe_canonicalize_comparison_1): Require
undefined overflow only for canonicalizing A +- CST cmp CST.
Make sure to not generate new constants that are not inside
their TYPE_MIN/MAX_VALUE range.
(maybe_canonicalize_comparison): Remove undefined overflow
checking from here.
(fold_binary): Remove now duplicate folding.
2008-08-27 Paolo Carlini <paolo.carlini@oracle.com>
* c-typeck.c (convert_for_assignment): Adjust WARN_FOR_ASSIGNMENT
......
......@@ -8289,9 +8289,14 @@ maybe_canonicalize_comparison_1 (enum tree_code code, tree type,
int sgn0;
bool swap = false;
/* Match A +- CST code arg1 and CST code arg1. */
if (!(((code0 == MINUS_EXPR
|| code0 == PLUS_EXPR)
/* Match A +- CST code arg1 and CST code arg1. We can change the
first form only if overflow is undefined. */
if (!((TYPE_OVERFLOW_UNDEFINED (TREE_TYPE (arg0))
/* In principle pointers also have undefined overflow behavior,
but that causes problems elsewhere. */
&& !POINTER_TYPE_P (TREE_TYPE (arg0))
&& (code0 == MINUS_EXPR
|| code0 == PLUS_EXPR)
&& TREE_CODE (TREE_OPERAND (arg0, 1)) == INTEGER_CST)
|| code0 == INTEGER_CST))
return NULL_TREE;
......@@ -8352,9 +8357,21 @@ maybe_canonicalize_comparison_1 (enum tree_code code, tree type,
*strict_overflow_p = true;
}
/* Now build the constant reduced in magnitude. */
/* Now build the constant reduced in magnitude. But not if that
would produce one outside of its types range. */
if (INTEGRAL_TYPE_P (TREE_TYPE (cst0))
&& ((sgn0 == 1
&& TYPE_MIN_VALUE (TREE_TYPE (cst0))
&& tree_int_cst_equal (cst0, TYPE_MIN_VALUE (TREE_TYPE (cst0))))
|| (sgn0 == -1
&& TYPE_MAX_VALUE (TREE_TYPE (cst0))
&& tree_int_cst_equal (cst0, TYPE_MAX_VALUE (TREE_TYPE (cst0))))))
/* We cannot swap the comparison here as that would cause us to
endlessly recurse. */
return NULL_TREE;
t = int_const_binop (sgn0 == -1 ? PLUS_EXPR : MINUS_EXPR,
cst0, build_int_cst (TREE_TYPE (cst0), 1), 0);
cst0, build_int_cst (TREE_TYPE (cst0), 1), 0);
if (code0 != INTEGER_CST)
t = fold_build2 (code0, TREE_TYPE (arg0), TREE_OPERAND (arg0, 0), t);
......@@ -8380,12 +8397,6 @@ maybe_canonicalize_comparison (enum tree_code code, tree type,
const char * const warnmsg = G_("assuming signed overflow does not occur "
"when reducing constant in comparison");
/* In principle pointers also have undefined overflow behavior,
but that causes problems elsewhere. */
if (!TYPE_OVERFLOW_UNDEFINED (TREE_TYPE (arg0))
|| POINTER_TYPE_P (TREE_TYPE (arg0)))
return NULL_TREE;
/* Try canonicalization by simplifying arg0. */
strict_overflow_p = false;
t = maybe_canonicalize_comparison_1 (code, type, arg0, arg1,
......@@ -12424,29 +12435,6 @@ fold_binary (enum tree_code code, tree type, tree op0, tree op1)
}
}
/* Change X >= C to X > (C - 1) and X < C to X <= (C - 1) if C > 0.
This transformation affects the cases which are handled in later
optimizations involving comparisons with non-negative constants. */
if (TREE_CODE (arg1) == INTEGER_CST
&& TREE_CODE (arg0) != INTEGER_CST
&& tree_int_cst_sgn (arg1) > 0)
{
if (code == GE_EXPR)
{
arg1 = const_binop (MINUS_EXPR, arg1,
build_int_cst (TREE_TYPE (arg1), 1), 0);
return fold_build2 (GT_EXPR, type, arg0,
fold_convert (TREE_TYPE (arg0), arg1));
}
if (code == LT_EXPR)
{
arg1 = const_binop (MINUS_EXPR, arg1,
build_int_cst (TREE_TYPE (arg1), 1), 0);
return fold_build2 (LE_EXPR, type, arg0,
fold_convert (TREE_TYPE (arg0), arg1));
}
}
/* Comparisons with the highest or lowest possible integer of
the specified precision will have known values. */
{
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment