path: root/0128-Backport-SME-aarch64-Fix-nosimd-handling-of-FPR-move.patch

From 81a4b464d01cf00f8b355115588e67bf2c021acd Mon Sep 17 00:00:00 2001
From: Richard Sandiford <richard.sandiford@arm.com>
Date: Wed, 7 Sep 2022 10:52:04 +0100
Subject: [PATCH 029/157] [Backport][SME] aarch64: Fix +nosimd handling of FPR
 moves

Reference: https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=d6106132907f6bd01109f2616d20a87edecc6fc6

8-bit and 16-bit FPR moves would ICE for +nosimd+fp, and some other
moves would handle FPR<-zero inefficiently.  This is very much a
niche case at the moment, but something like it becomes more
important with SME streaming mode.

The si, di and vector tests already passed, they're just included for
completeness.

We're a bit inconsistent about whether alternatives involving FPRs
are marked with arch==fp or arch=* (i.e. default).  E.g. FPR loads
and stores are sometimes * and sometimes fp.

IMO * makes more sense.  FPRs should not be used at all without
TARGET_FLOAT, so TARGET_FLOAT represents the base architecture
when FPRs are enabled.  I think it's more useful if non-default
arches represent a genuine restriction.

gcc/
	* config/aarch64/aarch64.md (*mov<SHORT:mode>_aarch64): Extend
	w<-w, r<-w and w<-r alternatives to !simd, using 32-bit moves
	in that case.  Extend w<-r to w<-Z.
	(*mov<HFBF:mode>_aarch64): Likewise, but with Y instead of Z.
	(*movti_aarch64): Use an FMOV from XZR for w<-Z if MOVI is not
	available.
	(define_split): Do not apply the floating-point immediate-to-register
	split to zeros, even if MOVI is not available.

gcc/testsuite/
	* gcc.target/aarch64/movqi_1.c: New test.
	* gcc.target/aarch64/movhi_1.c: Likewise.
	* gcc.target/aarch64/movsi_1.c: Likewise.
	* gcc.target/aarch64/movdi_2.c: Likewise.
	* gcc.target/aarch64/movti_2.c: Likewise.
	* gcc.target/aarch64/movhf_1.c: Likewise.
	* gcc.target/aarch64/movsf_1.c: Likewise.
	* gcc.target/aarch64/movdf_1.c: Likewise.
	* gcc.target/aarch64/movtf_2.c: Likewise.
	* gcc.target/aarch64/movv8qi_1.c: Likewise.
	* gcc.target/aarch64/movv16qi_1.c: Likewise.
---
 gcc/config/aarch64/aarch64.md                 | 38 ++++----
 gcc/testsuite/gcc.target/aarch64/movdf_1.c    | 53 ++++++++++++
 gcc/testsuite/gcc.target/aarch64/movdi_2.c    | 61 +++++++++++++
 gcc/testsuite/gcc.target/aarch64/movhf_1.c    | 53 ++++++++++++
 gcc/testsuite/gcc.target/aarch64/movhi_1.c    | 61 +++++++++++++
 gcc/testsuite/gcc.target/aarch64/movqi_1.c    | 61 +++++++++++++
 gcc/testsuite/gcc.target/aarch64/movsf_1.c    | 53 ++++++++++++
 gcc/testsuite/gcc.target/aarch64/movsi_1.c    | 61 +++++++++++++
 gcc/testsuite/gcc.target/aarch64/movtf_2.c    | 81 +++++++++++++++++
 gcc/testsuite/gcc.target/aarch64/movti_2.c    | 86 +++++++++++++++++++
 gcc/testsuite/gcc.target/aarch64/movv16qi_1.c | 82 ++++++++++++++++++
 gcc/testsuite/gcc.target/aarch64/movv8qi_1.c  | 55 ++++++++++++
 12 files changed, 729 insertions(+), 16 deletions(-)
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movdf_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movdi_2.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movhf_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movhi_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movqi_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movsf_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movsi_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movtf_2.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movti_2.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movv16qi_1.c
 create mode 100644 gcc/testsuite/gcc.target/aarch64/movv8qi_1.c

diff --git a/gcc/config/aarch64/aarch64.md b/gcc/config/aarch64/aarch64.md
index 7ee26284d..7267a74d6 100644
--- a/gcc/config/aarch64/aarch64.md
+++ b/gcc/config/aarch64/aarch64.md
@@ -1201,7 +1201,7 @@
 
 (define_insn "*mov<mode>_aarch64"
   [(set (match_operand:SHORT 0 "nonimmediate_operand" "=r,r,    w,r  ,r,w, m,m,r,w,w")
-	(match_operand:SHORT 1 "aarch64_mov_operand"  " r,M,D<hq>,Usv,m,m,rZ,w,w,r,w"))]
+	(match_operand:SHORT 1 "aarch64_mov_operand"  " r,M,D<hq>,Usv,m,m,rZ,w,w,rZ,w"))]
   "(register_operand (operands[0], <MODE>mode)
     || aarch64_reg_or_zero (operands[1], <MODE>mode))"
 {
@@ -1225,11 +1225,11 @@
      case 7:
        return "str\t%<size>1, %0";
      case 8:
-       return "umov\t%w0, %1.<v>[0]";
+       return TARGET_SIMD ? "umov\t%w0, %1.<v>[0]" : "fmov\t%w0, %s1";
      case 9:
-       return "dup\t%0.<Vallxd>, %w1";
+       return TARGET_SIMD ? "dup\t%0.<Vallxd>, %w1" : "fmov\t%s0, %w1";
      case 10:
-       return "dup\t%<Vetype>0, %1.<v>[0]";
+       return TARGET_SIMD ? "dup\t%<Vetype>0, %1.<v>[0]" : "fmov\t%s0, %s1";
      default:
        gcc_unreachable ();
      }
@@ -1237,7 +1237,7 @@
   ;; The "mov_imm" type for CNT is just a placeholder.
   [(set_attr "type" "mov_reg,mov_imm,neon_move,mov_imm,load_4,load_4,store_4,
 		     store_4,neon_to_gp<q>,neon_from_gp<q>,neon_dup")
-   (set_attr "arch" "*,*,simd,sve,*,*,*,*,simd,simd,simd")]
+   (set_attr "arch" "*,*,simd,sve,*,*,*,*,*,*,*")]
 )
 
 (define_expand "mov<mode>"
@@ -1399,14 +1399,15 @@
 
 (define_insn "*movti_aarch64"
   [(set (match_operand:TI 0
-	 "nonimmediate_operand"  "=   r,w,w, r,w,r,m,m,w,m")
+	 "nonimmediate_operand"  "=   r,w,w,w, r,w,r,m,m,w,m")
 	(match_operand:TI 1
-	 "aarch64_movti_operand" " rUti,Z,r, w,w,m,r,Z,m,w"))]
+	 "aarch64_movti_operand" " rUti,Z,Z,r, w,w,m,r,Z,m,w"))]
   "(register_operand (operands[0], TImode)
     || aarch64_reg_or_zero (operands[1], TImode))"
   "@
    #
    movi\\t%0.2d, #0
+   fmov\t%d0, xzr
    #
    #
    mov\\t%0.16b, %1.16b
@@ -1415,11 +1416,11 @@
    stp\\txzr, xzr, %0
    ldr\\t%q0, %1
    str\\t%q1, %0"
-  [(set_attr "type" "multiple,neon_move,f_mcr,f_mrc,neon_logic_q, \
+  [(set_attr "type" "multiple,neon_move,f_mcr,f_mcr,f_mrc,neon_logic_q, \
 		             load_16,store_16,store_16,\
                              load_16,store_16")
-   (set_attr "length" "8,4,8,8,4,4,4,4,4,4")
-   (set_attr "arch" "*,simd,*,*,simd,*,*,*,fp,fp")]
+   (set_attr "length" "8,4,4,8,8,4,4,4,4,4,4")
+   (set_attr "arch" "*,simd,*,*,*,simd,*,*,*,fp,fp")]
 )
 
 ;; Split a TImode register-register or register-immediate move into
@@ -1458,16 +1459,19 @@
 )
 
 (define_insn "*mov<mode>_aarch64"
-  [(set (match_operand:HFBF 0 "nonimmediate_operand" "=w,w  , w,?r,w,w  ,w  ,w,m,r,m ,r")
-	(match_operand:HFBF 1 "general_operand"      "Y ,?rY,?r, w,w,Ufc,Uvi,m,w,m,rY,r"))]
+  [(set (match_operand:HFBF 0 "nonimmediate_operand" "=w,w  ,w ,w  ,?r,?r,w,w,w  ,w  ,w,m,r,m ,r")
+	(match_operand:HFBF 1 "general_operand"      "Y ,?rY,?r,?rY, w, w,w,w,Ufc,Uvi,m,w,m,rY,r"))]
   "TARGET_FLOAT && (register_operand (operands[0], <MODE>mode)
     || aarch64_reg_or_fp_zero (operands[1], <MODE>mode))"
   "@
    movi\\t%0.4h, #0
    fmov\\t%h0, %w1
    dup\\t%w0.4h, %w1
+   fmov\\t%s0, %w1
    umov\\t%w0, %1.h[0]
+   fmov\\t%w0, %s1
    mov\\t%0.h[0], %1.h[0]
+   fmov\\t%s0, %s1
    fmov\\t%h0, %1
    * return aarch64_output_scalar_simd_mov_immediate (operands[1], HImode);
    ldr\\t%h0, %1
@@ -1475,9 +1479,10 @@
    ldrh\\t%w0, %1
    strh\\t%w1, %0
    mov\\t%w0, %w1"
-  [(set_attr "type" "neon_move,f_mcr,neon_move,neon_to_gp, neon_move,fconsts, \
-		     neon_move,f_loads,f_stores,load_4,store_4,mov_reg")
-   (set_attr "arch" "simd,fp16,simd,simd,simd,fp16,simd,*,*,*,*,*")]
+  [(set_attr "type" "neon_move,f_mcr,neon_move,f_mcr,neon_to_gp,f_mrc,
+		     neon_move,fmov,fconsts,neon_move,f_loads,f_stores,
+		     load_4,store_4,mov_reg")
+   (set_attr "arch" "simd,fp16,simd,*,simd,*,simd,*,fp16,simd,*,*,*,*,*")]
 )
 
 (define_insn "*movsf_aarch64"
@@ -1530,10 +1535,11 @@
 
 (define_split
   [(set (match_operand:GPF_HF 0 "nonimmediate_operand")
-	(match_operand:GPF_HF 1 "general_operand"))]
+	(match_operand:GPF_HF 1 "const_double_operand"))]
   "can_create_pseudo_p ()
    && !aarch64_can_const_movi_rtx_p (operands[1], <MODE>mode)
    && !aarch64_float_const_representable_p (operands[1])
+   && !aarch64_float_const_zero_rtx_p (operands[1])
    &&  aarch64_float_const_rtx_p (operands[1])"
   [(const_int 0)]
   {
diff --git a/gcc/testsuite/gcc.target/aarch64/movdf_1.c b/gcc/testsuite/gcc.target/aarch64/movdf_1.c
new file mode 100644
index 000000000..a51ded1d6
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movdf_1.c
@@ -0,0 +1,53 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+/*
+** fpr_to_fpr:
+**	fmov	d0, d1
+**	ret
+*/
+double
+fpr_to_fpr (double q0, double q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	d0, x0
+**	ret
+*/
+double
+gpr_to_fpr ()
+{
+  register double x0 asm ("x0");
+  asm volatile ("" : "=r" (x0));
+  return x0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	d0, xzr
+**	ret
+*/
+double
+zero_to_fpr ()
+{
+  return 0;
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	x0, d0
+**	ret
+*/
+void
+fpr_to_gpr (double q0)
+{
+  register double x0 asm ("x0");
+  x0 = q0;
+  asm volatile ("" :: "r" (x0));
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movdi_2.c b/gcc/testsuite/gcc.target/aarch64/movdi_2.c
new file mode 100644
index 000000000..dd3fc3e8a
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movdi_2.c
@@ -0,0 +1,61 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+#include <stdint.h>
+
+/*
+** fpr_to_fpr:
+**	fmov	d0, d1
+**	ret
+*/
+void
+fpr_to_fpr (void)
+{
+  register uint64_t q0 asm ("q0");
+  register uint64_t q1 asm ("q1");
+  asm volatile ("" : "=w" (q1));
+  q0 = q1;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	d0, x0
+**	ret
+*/
+void
+gpr_to_fpr (uint64_t x0)
+{
+  register uint64_t q0 asm ("q0");
+  q0 = x0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** zero_to_fpr:
+**	fmov	d0, xzr
+**	ret
+*/
+void
+zero_to_fpr ()
+{
+  register uint64_t q0 asm ("q0");
+  q0 = 0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	x0, d0
+**	ret
+*/
+uint64_t
+fpr_to_gpr ()
+{
+  register uint64_t q0 asm ("q0");
+  asm volatile ("" : "=w" (q0));
+  return q0;
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movhf_1.c b/gcc/testsuite/gcc.target/aarch64/movhf_1.c
new file mode 100644
index 000000000..cae25d4e5
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movhf_1.c
@@ -0,0 +1,53 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+/*
+** fpr_to_fpr:
+**	fmov	s0, s1
+**	ret
+*/
+_Float16
+fpr_to_fpr (_Float16 q0, _Float16 q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	s0, w0
+**	ret
+*/
+_Float16
+gpr_to_fpr ()
+{
+  register _Float16 w0 asm ("w0");
+  asm volatile ("" : "=r" (w0));
+  return w0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+_Float16
+zero_to_fpr ()
+{
+  return 0;
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	w0, s0
+**	ret
+*/
+void
+fpr_to_gpr (_Float16 q0)
+{
+  register _Float16 w0 asm ("w0");
+  w0 = q0;
+  asm volatile ("" :: "r" (w0));
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movhi_1.c b/gcc/testsuite/gcc.target/aarch64/movhi_1.c
new file mode 100644
index 000000000..8017abc5f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movhi_1.c
@@ -0,0 +1,61 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+#include <stdint.h>
+
+/*
+** fpr_to_fpr:
+**	fmov	s0, s1
+**	ret
+*/
+void
+fpr_to_fpr (void)
+{
+  register uint16_t q0 asm ("q0");
+  register uint16_t q1 asm ("q1");
+  asm volatile ("" : "=w" (q1));
+  q0 = q1;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	s0, w0
+**	ret
+*/
+void
+gpr_to_fpr (uint16_t w0)
+{
+  register uint16_t q0 asm ("q0");
+  q0 = w0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+void
+zero_to_fpr ()
+{
+  register uint16_t q0 asm ("q0");
+  q0 = 0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	w0, s0
+**	ret
+*/
+uint16_t
+fpr_to_gpr ()
+{
+  register uint16_t q0 asm ("q0");
+  asm volatile ("" : "=w" (q0));
+  return q0;
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movqi_1.c b/gcc/testsuite/gcc.target/aarch64/movqi_1.c
new file mode 100644
index 000000000..401a79630
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movqi_1.c
@@ -0,0 +1,61 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+#include <stdint.h>
+
+/*
+** fpr_to_fpr:
+**	fmov	s0, s1
+**	ret
+*/
+void
+fpr_to_fpr (void)
+{
+  register uint8_t q0 asm ("q0");
+  register uint8_t q1 asm ("q1");
+  asm volatile ("" : "=w" (q1));
+  q0 = q1;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	s0, w0
+**	ret
+*/
+void
+gpr_to_fpr (uint8_t w0)
+{
+  register uint8_t q0 asm ("q0");
+  q0 = w0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+void
+zero_to_fpr ()
+{
+  register uint8_t q0 asm ("q0");
+  q0 = 0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	w0, s0
+**	ret
+*/
+uint8_t
+fpr_to_gpr ()
+{
+  register uint8_t q0 asm ("q0");
+  asm volatile ("" : "=w" (q0));
+  return q0;
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movsf_1.c b/gcc/testsuite/gcc.target/aarch64/movsf_1.c
new file mode 100644
index 000000000..09715aa4f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movsf_1.c
@@ -0,0 +1,53 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+/*
+** fpr_to_fpr:
+**	fmov	s0, s1
+**	ret
+*/
+float
+fpr_to_fpr (float q0, float q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	s0, w0
+**	ret
+*/
+float
+gpr_to_fpr ()
+{
+  register float w0 asm ("w0");
+  asm volatile ("" : "=r" (w0));
+  return w0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+float
+zero_to_fpr ()
+{
+  return 0;
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	w0, s0
+**	ret
+*/
+void
+fpr_to_gpr (float q0)
+{
+  register float w0 asm ("w0");
+  w0 = q0;
+  asm volatile ("" :: "r" (w0));
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movsi_1.c b/gcc/testsuite/gcc.target/aarch64/movsi_1.c
new file mode 100644
index 000000000..5314139aa
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movsi_1.c
@@ -0,0 +1,61 @@
+/* { dg-do assemble } */
+/* { dg-options "-O --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+#include <stdint.h>
+
+/*
+** fpr_to_fpr:
+**	fmov	s0, s1
+**	ret
+*/
+void
+fpr_to_fpr (void)
+{
+  register uint32_t q0 asm ("q0");
+  register uint32_t q1 asm ("q1");
+  asm volatile ("" : "=w" (q1));
+  q0 = q1;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	s0, w0
+**	ret
+*/
+void
+gpr_to_fpr (uint32_t w0)
+{
+  register uint32_t q0 asm ("q0");
+  q0 = w0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+void
+zero_to_fpr ()
+{
+  register uint32_t q0 asm ("q0");
+  q0 = 0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	w0, s0
+**	ret
+*/
+uint32_t
+fpr_to_gpr ()
+{
+  register uint32_t q0 asm ("q0");
+  asm volatile ("" : "=w" (q0));
+  return q0;
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movtf_2.c b/gcc/testsuite/gcc.target/aarch64/movtf_2.c
new file mode 100644
index 000000000..38b16358d
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movtf_2.c
@@ -0,0 +1,81 @@
+/* { dg-do assemble } */
+/* { dg-require-effective-target large_long_double } */
+/* { dg-options "-O -mtune=neoverse-v1 --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+/*
+** fpr_to_fpr:
+**	sub	sp, sp, #16
+**	str	q1, \[sp\]
+**	ldr	q0, \[sp\]
+**	add	sp, sp, #?16
+**	ret
+*/
+long double
+fpr_to_fpr (long double q0, long double q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:	{ target aarch64_little_endian }
+**	fmov	d0, x0
+**	fmov	v0.d\[1\], x1
+**	ret
+*/
+/*
+** gpr_to_fpr:	{ target aarch64_big_endian }
+**	fmov	d0, x1
+**	fmov	v0.d\[1\], x0
+**	ret
+*/
+long double
+gpr_to_fpr ()
+{
+  register long double x0 asm ("x0");
+  asm volatile ("" : "=r" (x0));
+  return x0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	s0, wzr
+**	ret
+*/
+long double
+zero_to_fpr ()
+{
+  return 0;
+}
+
+/*
+** fpr_to_gpr:	{ target aarch64_little_endian }
+** (
+**	fmov	x0, d0
+**	fmov	x1, v0.d\[1\]
+** |
+**	fmov	x1, v0.d\[1\]
+**	fmov	x0, d0
+** )
+**	ret
+*/
+/*
+** fpr_to_gpr:	{ target aarch64_big_endian }
+** (
+**	fmov	x1, d0
+**	fmov	x0, v0.d\[1\]
+** |
+**	fmov	x0, v0.d\[1\]
+**	fmov	x1, d0
+** )
+**	ret
+*/
+void
+fpr_to_gpr (long double q0)
+{
+  register long double x0 asm ("x0");
+  x0 = q0;
+  asm volatile ("" :: "r" (x0));
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movti_2.c b/gcc/testsuite/gcc.target/aarch64/movti_2.c
new file mode 100644
index 000000000..c393b1220
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movti_2.c
@@ -0,0 +1,86 @@
+/* { dg-do assemble } */
+/* { dg-options "-O -mtune=neoverse-v1 --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+/*
+** fpr_to_fpr:
+**	sub	sp, sp, #16
+**	str	q1, \[sp\]
+**	ldr	q0, \[sp\]
+**	add	sp, sp, #?16
+**	ret
+*/
+void
+fpr_to_fpr (void)
+{
+  register __int128_t q0 asm ("q0");
+  register __int128_t q1 asm ("q1");
+  asm volatile ("" : "=w" (q1));
+  q0 = q1;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** gpr_to_fpr:	{ target aarch64_little_endian }
+**	fmov	d0, x0
+**	fmov	v0.d\[1\], x1
+**	ret
+*/
+/*
+** gpr_to_fpr:	{ target aarch64_big_endian }
+**	fmov	d0, x1
+**	fmov	v0.d\[1\], x0
+**	ret
+*/
+void
+gpr_to_fpr (__int128_t x0)
+{
+  register __int128_t q0 asm ("q0");
+  q0 = x0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** zero_to_fpr:
+**	fmov	d0, xzr
+**	ret
+*/
+void
+zero_to_fpr ()
+{
+  register __int128_t q0 asm ("q0");
+  q0 = 0;
+  asm volatile ("" :: "w" (q0));
+}
+
+/*
+** fpr_to_gpr:	{ target aarch64_little_endian }
+** (
+**	fmov	x0, d0
+**	fmov	x1, v0.d\[1\]
+** |
+**	fmov	x1, v0.d\[1\]
+**	fmov	x0, d0
+** )
+**	ret
+*/
+/*
+** fpr_to_gpr:	{ target aarch64_big_endian }
+** (
+**	fmov	x1, d0
+**	fmov	x0, v0.d\[1\]
+** |
+**	fmov	x0, v0.d\[1\]
+**	fmov	x1, d0
+** )
+**	ret
+*/
+__int128_t
+fpr_to_gpr ()
+{
+  register __int128_t q0 asm ("q0");
+  asm volatile ("" : "=w" (q0));
+  return q0;
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movv16qi_1.c b/gcc/testsuite/gcc.target/aarch64/movv16qi_1.c
new file mode 100644
index 000000000..8a6afb13b
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movv16qi_1.c
@@ -0,0 +1,82 @@
+/* { dg-do assemble } */
+/* { dg-options "-O -mtune=neoverse-v1 --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+typedef unsigned char v16qi __attribute__((vector_size(16)));
+
+/*
+** fpr_to_fpr:
+**	sub	sp, sp, #16
+**	str	q1, \[sp\]
+**	ldr	q0, \[sp\]
+**	add	sp, sp, #?16
+**	ret
+*/
+v16qi
+fpr_to_fpr (v16qi q0, v16qi q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:	{ target aarch64_little_endian }
+**	fmov	d0, x0
+**	fmov	v0.d\[1\], x1
+**	ret
+*/
+/*
+** gpr_to_fpr:	{ target aarch64_big_endian }
+**	fmov	d0, x1
+**	fmov	v0.d\[1\], x0
+**	ret
+*/
+v16qi
+gpr_to_fpr ()
+{
+  register v16qi x0 asm ("x0");
+  asm volatile ("" : "=r" (x0));
+  return x0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	d0, xzr
+**	ret
+*/
+v16qi
+zero_to_fpr ()
+{
+  return (v16qi) {};
+}
+
+/*
+** fpr_to_gpr:	{ target aarch64_little_endian }
+** (
+**	fmov	x0, d0
+**	fmov	x1, v0.d\[1\]
+** |
+**	fmov	x1, v0.d\[1\]
+**	fmov	x0, d0
+** )
+**	ret
+*/
+/*
+** fpr_to_gpr:	{ target aarch64_big_endian }
+** (
+**	fmov	x1, d0
+**	fmov	x0, v0.d\[1\]
+** |
+**	fmov	x0, v0.d\[1\]
+**	fmov	x1, d0
+** )
+**	ret
+*/
+void
+fpr_to_gpr (v16qi q0)
+{
+  register v16qi x0 asm ("x0");
+  x0 = q0;
+  asm volatile ("" :: "r" (x0));
+}
diff --git a/gcc/testsuite/gcc.target/aarch64/movv8qi_1.c b/gcc/testsuite/gcc.target/aarch64/movv8qi_1.c
new file mode 100644
index 000000000..4c97e6fbc
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/movv8qi_1.c
@@ -0,0 +1,55 @@
+/* { dg-do assemble } */
+/* { dg-options "-O -mtune=neoverse-v1 --save-temps" } */
+/* { dg-final { check-function-bodies "**" "" "" } } */
+
+#pragma GCC target "+nothing+nosimd+fp"
+
+typedef unsigned char v8qi __attribute__((vector_size(8)));
+
+/*
+** fpr_to_fpr:
+**	fmov	d0, d1
+**	ret
+*/
+v8qi
+fpr_to_fpr (v8qi q0, v8qi q1)
+{
+  return q1;
+}
+
+/*
+** gpr_to_fpr:
+**	fmov	d0, x0
+**	ret
+*/
+v8qi
+gpr_to_fpr ()
+{
+  register v8qi x0 asm ("x0");
+  asm volatile ("" : "=r" (x0));
+  return x0;
+}
+
+/*
+** zero_to_fpr:
+**	fmov	d0, xzr
+**	ret
+*/
+v8qi
+zero_to_fpr ()
+{
+  return (v8qi) {};
+}
+
+/*
+** fpr_to_gpr:
+**	fmov	x0, d0
+**	ret
+*/
+void
+fpr_to_gpr (v8qi q0)
+{
+  register v8qi x0 asm ("x0");
+  x0 = q0;
+  asm volatile ("" :: "r" (x0));
+}
-- 
2.33.0