cmd/internal/obj/arm64: fix the bug of incorrect handling negative offset of LDP/STP/LDPW/STPW

The current assembler will report error when the negative offset is in
the range of [-256, 0) and is not the multiples of 4/8.

The fix introduces C_NSAUTO_8, C_NSAUTO_4 and C_NAUTO4K. C_NPAUTO
includes C_NSAUTO_8 instead of C_NSAUTO, C_NAUTO4K includes C_NSAUTO_8,
C_NSAUTO_4 and C_NSAUTO. So that assembler will encode the negative offset
that is greater than -4095 and is not the multiples of 4/8 as two instructions.

Add the test cases.

Fixed #24471

Change-Id: I42f34e3b8a9fc52c9e8b41504294271aafade639
Reviewed-on: https://go-review.googlesource.com/102635
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

src/cmd/asm/internal/asm/testdata/arm64.s

@@ -540,8 +540,12 @@ again:
 	LDP	1024(RSP), (R1, R2) // fb031091610b40a9
 	LDP.W	8(RSP), (R1, R2)    // e18bc0a9
 	LDP.P	8(RSP), (R1, R2)    // e18bc0a8
+	LDP	-31(R0), (R1, R2)   // 1b7c00d1610b40a9
+	LDP	-4(R0), (R1, R2)    // 1b1000d1610b40a9
+	LDP	-8(R0), (R1, R2)    // 01887fa9
 	LDP	x(SB), (R1, R2)
 	LDP	x+8(SB), (R1, R2)
+	LDPW	-5(R0), (R1, R2)    // 1b1400d1610b4029
 	LDPW	(R0), (R1, R2)      // 01084029
 	LDPW	4(R0), (R1, R2)     // 01884029
 	LDPW	-4(R0), (R1, R2)    // 01887f29
@@ -579,6 +583,7 @@ again:
 	STP.W	(R3, R4), 8(R5)     // a39080a9
 	STP.P	(R3, R4), 8(R5)     // a39080a8
 	STP	(R3, R4), -8(R5)    // a3903fa9
+	STP	(R3, R4), -4(R5)    // bb1000d1631300a9
 	STP	(R3, R4), 11(R0)    // 1b2c0091631300a9
 	STP	(R3, R4), 1024(R0)  // 1b001091631300a9
 	STP	(R3, R4), (RSP)     // e31300a9
@@ -595,6 +600,7 @@ again:
 	STPW.W	(R3, R4), 4(R5)     // a3908029
 	STPW.P	(R3, R4), 4(R5)     // a3908028
 	STPW	(R3, R4), -4(R5)    // a3903f29
+	STPW	(R3, R4), -5(R5)    // bb1400d163130029
 	STPW	(R3, R4), 11(R0)    // 1b2c009163130029
 	STPW	(R3, R4), 1024(R0)  // 1b00109163130029
 	STPW	(R3, R4), (RSP)     // e3130029

src/cmd/internal/obj/arm64/a.out.go

@@ -425,8 +425,11 @@ const (
 	C_LBRA
 
 	C_ZAUTO      // 0(RSP)
-	C_NPAUTO     // -512 <= x < 0, 0 mod 8
+	C_NSAUTO_8   // -256 <= x < 0, 0 mod 8
+	C_NSAUTO_4   // -256 <= x < 0, 0 mod 4
 	C_NSAUTO     // -256 <= x < 0
+	C_NPAUTO     // -512 <= x < 0, 0 mod 8
+	C_NAUTO4K    // -4095 <= x < 0
 	C_PSAUTO_8   // 0 to 255, 0 mod 8
 	C_PSAUTO_4   // 0 to 255, 0 mod 4
 	C_PSAUTO     // 0 to 255
@@ -451,8 +454,11 @@ const (
 	C_LEXT
 
 	C_ZOREG    // 0(R)
-	C_NPOREG // must mirror NPAUTO, etc
+	C_NSOREG_8 // must mirror C_NSAUTO_8, etc
+	C_NSOREG_4
 	C_NSOREG
+	C_NPOREG
+	C_NOREG4K
 	C_PSOREG_8
 	C_PSOREG_4
 	C_PSOREG

src/cmd/internal/obj/arm64/anames7.go

@@ -37,8 +37,11 @@ var cnames7 = []string{
 	"SBRA",
 	"LBRA",
 	"ZAUTO",
-	"NPAUTO",
+	"NSAUTO_8",
+	"NSAUTO_4",
 	"NSAUTO",
+	"NPAUTO",
+	"NAUTO4K",
 	"PSAUTO_8",
 	"PSAUTO_4",
 	"PSAUTO",
@@ -61,8 +64,11 @@ var cnames7 = []string{
 	"SEXT16",
 	"LEXT",
 	"ZOREG",
-	"NPOREG",
+	"NSOREG_8",
+	"NSOREG_4",
 	"NSOREG",
+	"NPOREG",
+	"NOREG4K",
 	"PSOREG_8",
 	"PSOREG_4",
 	"PSOREG",