cmd/asm, cmd/internal/obj/ppc64: avoid unnecessary load zeros

When instructions add, and, or, xor, and movd have
constant operands in some cases more instructions are
generated than necessary by the assembler.

This adds more opcode/operand combinations to the optab
and improves the code generation for the cases where the
size and sign of the constant allows the use of 1
instructions instead of 2.

Example of previous code:
	oris r3, r0, 0
	ori  r3, r3, 65533

now:
	ori r3, r0, 65533

This does not significantly reduce the overall binary size
because the improvement depends on the constant value.
Some procedures show a 1-2% reduction in size. This improvement
could also be significant in cases where the extra instructions
occur in a critical loop.

Testcase ppc64enc.s was added to cmd/asm/internal/asm/testdata
with the variations affected by this change.

Updates #23845

Change-Id: I7fdf2320c95815d99f2755ba77d0c6921cd7fad7
Reviewed-on: https://go-review.googlesource.com/95135
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>

2 files changed, 91 insertions, 0 deletions

diff --git a/src/cmd/asm/internal/asm/endtoend_test.go b/src/cmd/asm/internal/asm/endtoend_test.go
index d8a447d958..093d5c40ed 100644
--- a/src/cmd/asm/internal/asm/endtoend_test.go
+++ b/src/cmd/asm/internal/asm/endtoend_test.go
@@ -411,6 +411,10 @@ func TestPPC64EndToEnd(t *testing.T) {
 	testEndToEnd(t, "ppc64", "ppc64")
 }
 
+func TestPPC64Encoder(t *testing.T) {
+	testEndToEnd(t, "ppc64", "ppc64enc")
+}
+
 func TestS390XEndToEnd(t *testing.T) {
 	testEndToEnd(t, "s390x", "s390x")
 }
diff --git a/src/cmd/asm/internal/asm/testdata/ppc64enc.s b/src/cmd/asm/internal/asm/testdata/ppc64enc.s
new file mode 100644
index 0000000000..f39c0728eb
--- /dev/null
+++ b/src/cmd/asm/internal/asm/testdata/ppc64enc.s
@@ -0,0 +1,87 @@
+// Copyright 2018 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Initial set of opcode combinations based on
+// improvements to processing of constant
+// operands.
+
+// Full set will be added at a later date.
+
+#include "../../../../../runtime/textflag.h"
+
+TEXT asmtest(SB),DUPOK|NOSPLIT,$0
+	// move constants
+	MOVD $1, R3                     // 38600001
+	MOVD $-1, R4                    // 3880ffff
+	MOVD $65535, R5                 // 6005ffff
+	MOVD $65536, R6                 // 64060001
+	MOVD $-32767, R5                // 38a08001
+	MOVD $-32768, R6                // 38c08000
+	MOVD $1234567, R5               // 6405001260a5d687
+
+	// add constants
+	ADD $1, R3                      // 38630001
+	ADD $1, R3, R4                  // 38830001
+	ADD $-1, R4                     // 3884ffff
+	ADD $-1, R4, R5                 // 38a4ffff
+	ADD $65535, R5                  // 601fffff7cbf2a14
+	ADD $65535, R5, R6              // 601fffff7cdf2a14
+	ADD $65536, R6                  // 3cc60001
+	ADD $65536, R6, R7              // 3ce60001
+	ADD $-32767, R5                 // 38a58001
+	ADD $-32767, R5, R4             // 38858001
+	ADD $-32768, R6                 // 38c68000
+	ADD $-32768, R6, R5             // 38a68000
+	ADD $1234567, R5                // 641f001263ffd6877cbf2a14
+	ADD $1234567, R5, R6            // 641f001263ffd6877cdf2a14
+
+	// and constants
+	ANDCC $1, R3                    // 70630001
+	ANDCC $1, R3, R4                // 70640001
+	ANDCC $-1, R4                   // 3be0ffff7fe42039
+	ANDCC $-1, R4, R5               // 3be0ffff7fe52039
+	ANDCC $65535, R5                // 70a5ffff
+	ANDCC $65535, R5, R6            // 70a6ffff
+	ANDCC $65536, R6                // 74c60001
+	ANDCC $65536, R6, R7            // 74c70001
+	ANDCC $-32767, R5               // 3be080017fe52839
+	ANDCC $-32767, R5, R4           // 3be080017fe42839
+	ANDCC $-32768, R6               // 3be080007fe63039
+	ANDCC $-32768, R5, R6           // 3be080007fe62839
+	ANDCC $1234567, R5              // 641f001263ffd6877fe52839
+	ANDCC $1234567, R5, R6          // 641f001263ffd6877fe62839
+
+	// or constants
+	OR $1, R3                       // 60630001
+	OR $1, R3, R4                   // 60640001
+	OR $-1, R4                      // 3be0ffff7fe42378
+	OR $-1, R4, R5                  // 3be0ffff7fe52378
+	OR $65535, R5                   // 60a5ffff
+	OR $65535, R5, R6               // 60a6ffff
+	OR $65536, R6                   // 64c60001
+	OR $65536, R6, R7               // 64c70001
+	OR $-32767, R5                  // 3be080017fe52b78
+	OR $-32767, R5, R6              // 3be080017fe62b78
+	OR $-32768, R6                  // 3be080007fe63378
+	OR $-32768, R6, R7              // 3be080007fe73378
+	OR $1234567, R5                 // 641f001263ffd6877fe52b78
+	OR $1234567, R5, R3             // 641f001263ffd6877fe32b78
+
+	// or constants
+	XOR $1, R3                      // 68630001
+	XOR $1, R3, R4                  // 68640001
+	XOR $-1, R4                     // 3be0ffff7fe42278
+	XOR $-1, R4, R5                 // 3be0ffff7fe52278
+	XOR $65535, R5                  // 68a5ffff
+	XOR $65535, R5, R6              // 68a6ffff
+	XOR $65536, R6                  // 6cc60001
+	XOR $65536, R6, R7              // 6cc70001
+	XOR $-32767, R5                 // 3be080017fe52a78
+	XOR $-32767, R5, R6             // 3be080017fe62a78
+	XOR $-32768, R6                 // 3be080007fe63278
+	XOR $-32768, R6, R7             // 3be080007fe73278
+	XOR $1234567, R5                // 641f001263ffd6877fe52a78
+	XOR $1234567, R5, R3            // 641f001263ffd6877fe32a78
+
+	RET