Add a compatible_vector_types_p target hook

One problem with adding an N-bit vector extension to an existing
architecture is to decide how N-bit vectors should be passed to
functions and returned from functions.  Allowing all N-bit vector
types to be passed in registers breaks backwards compatibility,
since N-bit vectors could be used (and emulated) before the vector
extension was added.  But always passing N-bit vectors on the
stack would be inefficient for things like vector libm functions.

For SVE we took the compromise position of predefining new SVE vector
types that are distinct from all existing vector types, including
GNU-style vectors.  The new types are passed and returned in an
efficient way while existing vector types are passed and returned
in the traditional way.  In the right circumstances, the two types
are inter-convertible.

The SVE types are created using:

      vectype = build_distinct_type_copy (vectype);
      SET_TYPE_STRUCTURAL_EQUALITY (vectype);
      TYPE_ARTIFICIAL (vectype) = 1;

The C frontend maintains this distinction, using VIEW_CONVERT_EXPR
to convert from one type to the other.  However, the distinction can
be lost during gimple, which treats two vector types with the same
mode, number of elements, and element type as equivalent.  And for
most targets that's the right thing to do.

This patch therefore adds a hook that lets the target choose
whether such vector types are indeed equivalent.

Note that the new tests fail for -mabi=ilp32 in the same way as other
ACLE-based tests.  I'm still planning to fix that as a follow-on.

2020-01-09  Richard Sandiford  <richard.sandiford@arm.com>

gcc/
	* target.def (compatible_vector_types_p): New target hook.
	* hooks.h (hook_bool_const_tree_const_tree_true): Declare.
	* hooks.c (hook_bool_const_tree_const_tree_true): New function.
	* doc/tm.texi.in (TARGET_COMPATIBLE_VECTOR_TYPES_P): New hook.
	* doc/tm.texi: Regenerate.
	* gimple-expr.c: Include target.h.
	(useless_type_conversion_p): Use targetm.compatible_vector_types_p.
	* config/aarch64/aarch64.c (aarch64_compatible_vector_types_p): New
	function.
	(TARGET_COMPATIBLE_VECTOR_TYPES_P): Define.
	* config/aarch64/aarch64-sve-builtins.cc (gimple_folder::convert_pred):
	Use the original predicate if it already has a suitable type.

gcc/testsuite/
	* gcc.target/aarch64/sve/pcs/gnu_vectors_1.c: New test.
	* gcc.target/aarch64/sve/pcs/gnu_vectors_2.c: Likewise.

From-SVN: r280047

1 files changed, 21 insertions, 0 deletions

diff --git a/gcc/doc/tm.texi b/gcc/doc/tm.texi
index 0e90b31..4aec468 100644
--- a/gcc/doc/tm.texi
+++ b/gcc/doc/tm.texi
@@ -4324,6 +4324,27 @@ insns involving vector mode @var{mode}.  At the very least, it
 must have move patterns for this mode.
 @end deftypefn
 
+@deftypefn {Target Hook} bool TARGET_COMPATIBLE_VECTOR_TYPES_P (const_tree @var{type1}, const_tree @var{type2})
+Return true if there is no target-specific reason for treating
+vector types @var{type1} and @var{type2} as distinct types.  The caller
+has already checked for target-independent reasons, meaning that the
+types are known to have the same mode, to have the same number of elements,
+and to have what the caller considers to be compatible element types.
+
+The main reason for defining this hook is to reject pairs of types
+that are handled differently by the target's calling convention.
+For example, when a new @var{N}-bit vector architecture is added
+to a target, the target may want to handle normal @var{N}-bit
+@code{VECTOR_TYPE} arguments and return values in the same way as
+before, to maintain backwards compatibility.  However, it may also
+provide new, architecture-specific @code{VECTOR_TYPE}s that are passed
+and returned in a more efficient way.  It is then important to maintain
+a distinction between the ``normal'' @code{VECTOR_TYPE}s and the new
+architecture-specific ones.
+
+The default implementation returns true, which is correct for most targets.
+@end deftypefn
+
 @deftypefn {Target Hook} opt_machine_mode TARGET_ARRAY_MODE (machine_mode @var{mode}, unsigned HOST_WIDE_INT @var{nelems})
 Return the mode that GCC should use for an array that has
 @var{nelems} elements, with each element having mode @var{mode}.