[llvm] [AMDGPU][True16][MC] true16 for v_alignbit_b32 (PR #119409)

Brox Chen via llvm-commits llvm-commits at lists.llvm.org
Wed Feb 26 13:51:09 PST 2025


https://github.com/broxigarchen updated https://github.com/llvm/llvm-project/pull/119409

>From ff222e2bb4c94bdf718c4ca5f68bb6915637215e Mon Sep 17 00:00:00 2001
From: guochen2 <guochen2 at amd.com>
Date: Tue, 10 Dec 2024 01:20:40 -0500
Subject: [PATCH] VOP3 v_alignbit_b32 mc support true16

---
 llvm/lib/Target/AMDGPU/SIInstructions.td      | 115 +++++++++++++++---
 llvm/lib/Target/AMDGPU/VOP3Instructions.td    |   9 +-
 .../AMDGPU/GlobalISel/inst-select-fshr.mir    |  11 +-
 llvm/test/CodeGen/AMDGPU/bswap.ll             |  28 +++--
 llvm/test/MC/AMDGPU/gfx11_asm_vop3.s          |  29 +++--
 llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16.s    |  67 ++++++----
 llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8.s     |  41 ++++---
 llvm/test/MC/AMDGPU/gfx12_asm_vop3.s          |   3 +
 llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp16.s    |   3 +
 llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp8.s     |   3 +
 .../Disassembler/AMDGPU/gfx11_dasm_vop3.txt   |  16 ++-
 .../AMDGPU/gfx11_dasm_vop3_dpp16.txt          |  31 ++++-
 .../AMDGPU/gfx11_dasm_vop3_dpp8.txt           |  16 ++-
 .../Disassembler/AMDGPU/gfx12_dasm_vop3.txt   |  16 ++-
 .../AMDGPU/gfx12_dasm_vop3_dpp16.txt          |  36 +++++-
 .../AMDGPU/gfx12_dasm_vop3_dpp8.txt           |  21 +++-
 16 files changed, 343 insertions(+), 102 deletions(-)

diff --git a/llvm/lib/Target/AMDGPU/SIInstructions.td b/llvm/lib/Target/AMDGPU/SIInstructions.td
index dc5158352e77a..e8daa4b542e02 100644
--- a/llvm/lib/Target/AMDGPU/SIInstructions.td
+++ b/llvm/lib/Target/AMDGPU/SIInstructions.td
@@ -2499,6 +2499,7 @@ def : AMDGPUPat <
                $src1), sub1)
 >;
 
+let True16Predicate = NotHasTrue16BitInsts in {
 def : ROTRPattern <V_ALIGNBIT_B32_e64>;
 
 def : GCNPat<(i32 (trunc (srl i64:$src0, (and i32:$src1, (i32 31))))),
@@ -2508,6 +2509,42 @@ def : GCNPat<(i32 (trunc (srl i64:$src0, (and i32:$src1, (i32 31))))),
 def : GCNPat<(i32 (trunc (srl i64:$src0, (i32 ShiftAmt32Imm:$src1)))),
           (V_ALIGNBIT_B32_e64 (i32 (EXTRACT_SUBREG (i64 $src0), sub1)),
                           (i32 (EXTRACT_SUBREG (i64 $src0), sub0)), $src1)>;
+} // end True16Predicate = NotHasTrue16BitInsts
+
+let True16Predicate = UseFakeTrue16Insts in {
+def : GCNPat <
+  (rotr i32:$src0, i32:$src1),
+  (V_ALIGNBIT_B32_fake16_e64 /* src0_modifiers */ 0, $src0,
+                             /* src1_modifiers */ 0, $src0,
+                             /* src2_modifiers */ 0,
+                             $src1, /* clamp */ 0, /* op_sel */ 0)
+>;
+
+def : GCNPat<(i32 (trunc (srl i64:$src0, (and i32:$src1, (i32 31))))),
+     (V_ALIGNBIT_B32_fake16_e64 0, /* src0_modifiers */
+                               (i32 (EXTRACT_SUBREG (i64 $src0), sub1)),
+                                0, /* src1_modifiers */
+                               (i32 (EXTRACT_SUBREG (i64 $src0), sub0)),
+                                0, /* src2_modifiers */
+                                $src1, /* clamp */ 0, /* op_sel */ 0)
+>;
+
+def : GCNPat<(i32 (trunc (srl i64:$src0, (i32 ShiftAmt32Imm:$src1)))),
+     (V_ALIGNBIT_B32_fake16_e64 0, /* src0_modifiers */
+                               (i32 (EXTRACT_SUBREG (i64 $src0), sub1)),
+                                0, /* src1_modifiers */
+                               (i32 (EXTRACT_SUBREG (i64 $src0), sub0)),
+                                0, /* src2_modifiers */
+                                $src1, /* clamp */ 0, /* op_sel */ 0)
+>;
+
+def : GCNPat<(fshr i32:$src0, i32:$src1, i32:$src2),
+     (V_ALIGNBIT_B32_fake16_e64 /* src0_modifiers */ 0, $src0,
+                                /* src1_modifiers */ 0, $src1,
+                                /* src2_modifiers */ 0,
+                                $src2, /* clamp */ 0, /* op_sel */ 0)
+>;
+} // end True16Predicate = UseFakeTrue16Insts
 
 /********** ====================== **********/
 /**********   Indirect addressing  **********/
@@ -3015,6 +3052,7 @@ def : GCNPat <
                     (i32 (EXTRACT_SUBREG $a, sub0))), (i32 1))
 >;
 
+let True16Predicate = NotHasTrue16BitInsts in
 def : GCNPat <
   (i32 (bswap i32:$a)),
   (V_BFI_B32_e64 (S_MOV_B32 (i32 0x00ff00ff)),
@@ -3022,28 +3060,61 @@ def : GCNPat <
              (V_ALIGNBIT_B32_e64 VSrc_b32:$a, VSrc_b32:$a, (i32 8)))
 >;
 
-// FIXME: This should have been narrowed to i32 during legalization.
-// This pattern should also be skipped for GlobalISel
+let True16Predicate = UseFakeTrue16Insts in
+def : GCNPat <
+  (i32 (bswap i32:$a)),
+  (V_BFI_B32_e64 (S_MOV_B32 (i32 0x00ff00ff)),
+             (V_ALIGNBIT_B32_fake16_e64 0, /* src0_modifiers */
+                                        VSrc_b32:$a,
+                                        0, /* src1_modifiers */
+                                        VSrc_b32:$a,
+                                        0, /* src2_modifiers */
+                                        (i32 24), /* clamp */ 0, /* op_sel */ 0),
+             (V_ALIGNBIT_B32_fake16_e64 0, /* src0_modifiers */
+                                        VSrc_b32:$a,
+                                        0, /* src1_modifiers */
+                                        VSrc_b32:$a,
+                                        0, /* src2_modifiers */
+                                        (i32 8), /* clamp */ 0, /* op_sel */ 0))
+>;
+
+class AlignBit32Inst<dag op1, dag op2, dag op3, bit isTrue16> {
+  defvar inst = !if(isTrue16, V_ALIGNBIT_B32_fake16_e64, V_ALIGNBIT_B32_e64);
+  defvar NoMods = !if(isTrue16, (inst 0), (inst));
+  dag ret = !con(NoMods, (inst op1), NoMods, (inst op2),
+                     NoMods, (inst op3), NoMods, NoMods);
+}
+
+multiclass bswapi64ExtPat<bit hasTrue16> {
 def : GCNPat <
   (i64 (bswap i64:$a)),
   (REG_SEQUENCE VReg_64,
   (V_BFI_B32_e64 (S_MOV_B32 (i32 0x00ff00ff)),
-             (V_ALIGNBIT_B32_e64 (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
-                             (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
-                             (i32 24)),
-             (V_ALIGNBIT_B32_e64 (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
-                             (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
-                             (i32 8))),
+     AlignBit32Inst<(i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
+                     (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
+                     (i32 24), hasTrue16>.ret,
+     AlignBit32Inst<(i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
+                     (i32 (EXTRACT_SUBREG VReg_64:$a, sub1)),
+                     (i32 8), hasTrue16>.ret),
   sub0,
   (V_BFI_B32_e64 (S_MOV_B32 (i32 0x00ff00ff)),
-             (V_ALIGNBIT_B32_e64 (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
-                             (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
-                             (i32 24)),
-             (V_ALIGNBIT_B32_e64 (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
-                             (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
-                             (i32 8))),
+     AlignBit32Inst<(i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
+                     (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
+                     (i32 24), hasTrue16>.ret,
+     AlignBit32Inst<(i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
+                     (i32 (EXTRACT_SUBREG VReg_64:$a, sub0)),
+                     (i32 8), hasTrue16>.ret),
   sub1)
 >;
+}
+
+// FIXME: This should have been narrowed to i32 during legalization.
+// This pattern should also be skipped for GlobalISel
+let True16Predicate = NotHasTrue16BitInsts in
+defm : bswapi64ExtPat</*hasTrue16*/0>;
+
+let True16Predicate = UseFakeTrue16Insts in
+defm : bswapi64ExtPat</*hasTrue16*/1>;
 
 // FIXME: The AddedComplexity should not be needed, but in GlobalISel
 // the BFI pattern ends up taking precedence without it.
@@ -3456,8 +3527,7 @@ def : GCNPat <
 
 // Take the upper 16 bits from V[0] and the lower 16 bits from V[1]
 // Special case, can use V_ALIGNBIT (always uses encoded literal)
-foreach p = [NotHasTrue16BitInsts, UseFakeTrue16Insts] in
-let True16Predicate = p in {
+let True16Predicate = NotHasTrue16BitInsts in
 def : GCNPat <
   (vecTy (DivergentBinFrag<build_vector>
     (Ty !if(!eq(Ty, i16),
@@ -3467,7 +3537,19 @@ def : GCNPat <
     (V_ALIGNBIT_B32_e64 VGPR_32:$b, VGPR_32:$a, (i32 16))
 >;
 
+let True16Predicate = UseFakeTrue16Insts in
+def : GCNPat <
+  (vecTy (DivergentBinFrag<build_vector>
+    (Ty !if(!eq(Ty, i16),
+      (Ty (trunc (srl VGPR_32:$a, (i32 16)))),
+      (Ty (bitconvert (i16 (trunc (srl VGPR_32:$a, (i32 16)))))))),
+    (Ty VGPR_32:$b))),
+    (V_ALIGNBIT_B32_fake16_e64 0, VGPR_32:$b, 0, VGPR_32:$a, 0, (i16 16), 0, 0)
+>;
+
 // Take the upper 16 bits from each VGPR_32 and concat them
+foreach p = [NotHasTrue16BitInsts, UseFakeTrue16Insts] in
+let True16Predicate = p in
 def : GCNPat <
   (vecTy (DivergentBinFrag<build_vector>
     (Ty !if(!eq(Ty, i16),
@@ -3478,7 +3560,6 @@ def : GCNPat <
       (Ty (bitconvert (i16 (trunc (srl VGPR_32:$b, (i32 16)))))))))),
   (V_PERM_B32_e64 VGPR_32:$b, VGPR_32:$a, (S_MOV_B32 (i32 0x07060302)))
 >;
-}
 
 } // end foreach Ty
 
diff --git a/llvm/lib/Target/AMDGPU/VOP3Instructions.td b/llvm/lib/Target/AMDGPU/VOP3Instructions.td
index 1447804871809..b80ad366b2796 100644
--- a/llvm/lib/Target/AMDGPU/VOP3Instructions.td
+++ b/llvm/lib/Target/AMDGPU/VOP3Instructions.td
@@ -211,9 +211,12 @@ defm V_CUBEMA_F32 : VOP3Inst <"v_cubema_f32", VOP3_Profile<VOP_F32_F32_F32_F32>,
 defm V_BFE_U32 : VOP3Inst <"v_bfe_u32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfe_u32>;
 defm V_BFE_I32 : VOP3Inst <"v_bfe_i32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfe_i32>;
 defm V_BFI_B32 : VOP3Inst <"v_bfi_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, AMDGPUbfi>;
-defm V_ALIGNBIT_B32 : VOP3Inst <"v_alignbit_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, fshr>;
+defm V_ALIGNBIT_B32 : VOP3Inst_t16_with_profiles <"v_alignbit_b32",
+                                                   VOP3_Profile<VOP_I32_I32_I32_I32>,
+                                                   VOP3_Profile_True16<VOP_I32_I32_I32_I16, VOP3_OPSEL>,
+                                                   VOP3_Profile_Fake16<VOP_I32_I32_I32_I16, VOP3_OPSEL>,
+                                                   fshr, null_frag>;
 
-let True16Predicate = NotHasTrue16BitInsts in
 defm V_ALIGNBYTE_B32 : VOP3Inst <"v_alignbyte_b32", VOP3_Profile<VOP_I32_I32_I32_I32>, int_amdgcn_alignbyte>;
 let True16Predicate = UseRealTrue16Insts in
 defm V_ALIGNBYTE_B32_t16 : VOP3Inst <"v_alignbyte_b32_t16", VOP3_Profile_True16<VOP_I32_I32_I32_I16, VOP3_OPSEL>>;
@@ -1726,7 +1729,7 @@ defm V_BFI_B32             : VOP3_Realtriple_gfx11_gfx12<0x212>;
 defm V_FMA_F32             : VOP3_Realtriple_gfx11_gfx12<0x213>;
 defm V_FMA_F64             : VOP3_Real_Base_gfx11_gfx12<0x214>;
 defm V_LERP_U8             : VOP3_Realtriple_gfx11_gfx12<0x215>;
-defm V_ALIGNBIT_B32        : VOP3_Realtriple_gfx11_gfx12<0x216>;
+defm V_ALIGNBIT_B32        : VOP3_Realtriple_t16_and_fake16_gfx11_gfx12<0x216, "v_alignbit_b32">;
 defm V_ALIGNBYTE_B32       : VOP3_Realtriple_t16_and_fake16_gfx11_gfx12<0x217, "v_alignbyte_b32">;
 defm V_MULLIT_F32          : VOP3_Realtriple_gfx11_gfx12<0x218>;
 defm V_MIN3_F32            : VOP3_Realtriple_gfx11<0x219>;
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-fshr.mir b/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-fshr.mir
index 4291744d85732..1fb67fe17cb0a 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-fshr.mir
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/inst-select-fshr.mir
@@ -3,7 +3,7 @@
 # RUN: llc -mtriple=amdgcn -mcpu=fiji -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=GCN %s
 # RUN: llc -mtriple=amdgcn -mcpu=gfx900 -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=GCN %s
 # RUN: llc -mtriple=amdgcn -mcpu=gfx1010 -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=GCN %s
-# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=GCN %s
+# RUN: llc -mtriple=amdgcn -mcpu=gfx1100 -run-pass=instruction-select -verify-machineinstrs -o - %s | FileCheck -check-prefix=GFX11 %s
 
 ---
 
@@ -23,6 +23,15 @@ body: |
     ; GCN-NEXT: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
     ; GCN-NEXT: [[V_ALIGNBIT_B32_e64_:%[0-9]+]]:vgpr_32 = V_ALIGNBIT_B32_e64 [[COPY]], [[COPY1]], [[COPY2]], implicit $exec
     ; GCN-NEXT: S_ENDPGM 0, implicit [[V_ALIGNBIT_B32_e64_]]
+    ;
+    ; GFX11-LABEL: name: fshr_s32
+    ; GFX11: liveins: $vgpr0, $vgpr1, $vgpr2
+    ; GFX11-NEXT: {{  $}}
+    ; GFX11-NEXT: [[COPY:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+    ; GFX11-NEXT: [[COPY1:%[0-9]+]]:vgpr_32 = COPY $vgpr1
+    ; GFX11-NEXT: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr2
+    ; GFX11-NEXT: [[V_ALIGNBIT_B32_fake16_e64_:%[0-9]+]]:vgpr_32 = V_ALIGNBIT_B32_fake16_e64 0, [[COPY]], 0, [[COPY1]], 0, [[COPY2]], 0, 0, implicit $exec
+    ; GFX11-NEXT: S_ENDPGM 0, implicit [[V_ALIGNBIT_B32_fake16_e64_]]
     %0:vgpr(s32) = COPY $vgpr0
     %1:vgpr(s32) = COPY $vgpr1
     %2:vgpr(s32) = COPY $vgpr2
diff --git a/llvm/test/CodeGen/AMDGPU/bswap.ll b/llvm/test/CodeGen/AMDGPU/bswap.ll
index a95a1aba0c914..b98ca1ff55f59 100644
--- a/llvm/test/CodeGen/AMDGPU/bswap.ll
+++ b/llvm/test/CodeGen/AMDGPU/bswap.ll
@@ -737,15 +737,25 @@ define i64 @v_bswap_i48(i64 %src) {
 ; VI-NEXT:    v_lshrrev_b32_e32 v1, 16, v2
 ; VI-NEXT:    s_setpc_b64 s[30:31]
 ;
-; GFX11-LABEL: v_bswap_i48:
-; GFX11:       ; %bb.0:
-; GFX11-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
-; GFX11-NEXT:    v_perm_b32 v2, 0, v0, 0x10203
-; GFX11-NEXT:    v_perm_b32 v0, 0, v1, 0x10203
-; GFX11-NEXT:    s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
-; GFX11-NEXT:    v_lshrrev_b32_e32 v1, 16, v2
-; GFX11-NEXT:    v_alignbit_b32 v0, v2, v0, 16
-; GFX11-NEXT:    s_setpc_b64 s[30:31]
+; GFX11-REAL16-LABEL: v_bswap_i48:
+; GFX11-REAL16:       ; %bb.0:
+; GFX11-REAL16-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-REAL16-NEXT:    v_perm_b32 v2, 0, v0, 0x10203
+; GFX11-REAL16-NEXT:    v_perm_b32 v1, 0, v1, 0x10203
+; GFX11-REAL16-NEXT:    s_delay_alu instid0(VALU_DEP_1)
+; GFX11-REAL16-NEXT:    v_lshrrev_b64 v[0:1], 16, v[1:2]
+; GFX11-REAL16-NEXT:    v_lshrrev_b32_e32 v1, 16, v2
+; GFX11-REAL16-NEXT:    s_setpc_b64 s[30:31]
+;
+; GFX11-FAKE16-LABEL: v_bswap_i48:
+; GFX11-FAKE16:       ; %bb.0:
+; GFX11-FAKE16-NEXT:    s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
+; GFX11-FAKE16-NEXT:    v_perm_b32 v2, 0, v0, 0x10203
+; GFX11-FAKE16-NEXT:    v_perm_b32 v0, 0, v1, 0x10203
+; GFX11-FAKE16-NEXT:    s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
+; GFX11-FAKE16-NEXT:    v_lshrrev_b32_e32 v1, 16, v2
+; GFX11-FAKE16-NEXT:    v_alignbit_b32 v0, v2, v0, 16
+; GFX11-FAKE16-NEXT:    s_setpc_b64 s[30:31]
   %trunc = trunc i64 %src to i48
   %bswap = call i48 @llvm.bswap.i48(i48 %trunc)
   %zext = zext i48 %bswap to i64
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop3.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop3.s
index 857a1359b00d9..ce452957b0198 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop3.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop3.s
@@ -416,11 +416,11 @@ v_alignbit_b32 v5, s1, v255, s3
 v_alignbit_b32 v5, s105, s105, s105
 // GFX11: v_alignbit_b32 v5, s105, s105, s105     ; encoding: [0x05,0x00,0x16,0xd6,0x69,0xd2,0xa4,0x01]
 
-v_alignbit_b32 v5, vcc_lo, ttmp15, v3
-// GFX11: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l
+// GFX11: v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
 
-v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255
-// GFX11: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l
+// GFX11: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
 
 v_alignbit_b32 v5, ttmp15, src_scc, ttmp15
 // GFX11: v_alignbit_b32 v5, ttmp15, src_scc, ttmp15 ; encoding: [0x05,0x00,0x16,0xd6,0x7b,0xfa,0xed,0x01]
@@ -449,6 +449,9 @@ v_alignbit_b32 v5, src_scc, vcc_lo, -1
 v_alignbit_b32 v255, 0xaf123456, vcc_hi, null
 // GFX11: v_alignbit_b32 v255, 0xaf123456, vcc_hi, null ; encoding: [0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf]
 
+v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h
+// GFX11: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+
 v_alignbyte_b32 v5, v1, v2, s3
 // GFX11: v_alignbyte_b32 v5, v1, v2, s3          ; encoding: [0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00]
 
@@ -543,13 +546,13 @@ v_and_b16 v255.l, 0xfe0b, vcc_hi
 // GFX11: v_and_b16 v255.l, 0xfe0b, vcc_hi        ; encoding: [0xff,0x00,0x62,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
 v_and_b16 v5.l, v1.h, v2.l
-// GFX11: [0x05,0x08,0x62,0xd7,0x01,0x05,0x02,0x00]
+// GFX11: v_and_b16 v5.l, v1.h, v2.l op_sel:[1,0,0] ; encoding: [0x05,0x08,0x62,0xd7,0x01,0x05,0x02,0x00]
 
 v_and_b16 v5.l, v255.l, v255.h
-// GFX11: [0x05,0x10,0x62,0xd7,0xff,0xff,0x03,0x00]
+// GFX11: v_and_b16 v5.l, v255.l, v255.h op_sel:[0,1,0] ; encoding: [0x05,0x10,0x62,0xd7,0xff,0xff,0x03,0x00]
 
 v_and_b16 v255.h, 0xfe0b, vcc_hi
-// GFX11: [0xff,0x40,0x62,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
+// GFX11: v_and_b16 v255.h, 0xfe0b, vcc_hi op_sel:[0,0,1] ; encoding: [0xff,0x40,0x62,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
 v_and_or_b32 v5, v1, v2, s3
 // GFX11: v_and_or_b32 v5, v1, v2, s3             ; encoding: [0x05,0x00,0x57,0xd6,0x01,0x05,0x0e,0x00]
@@ -5577,13 +5580,13 @@ v_or_b16 v255.l, 0xfe0b, vcc_hi
 // GFX11: v_or_b16 v255.l, 0xfe0b, vcc_hi         ; encoding: [0xff,0x00,0x63,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
 v_or_b16 v5.l, v1.h, v2.l
-// GFX11: [0x05,0x08,0x63,0xd7,0x01,0x05,0x02,0x00]
+// GFX11: v_or_b16 v5.l, v1.h, v2.l op_sel:[1,0,0] ; encoding: [0x05,0x08,0x63,0xd7,0x01,0x05,0x02,0x00]
 
 v_or_b16 v5.l, v255.l, v255.h
-// GFX11: [0x05,0x10,0x63,0xd7,0xff,0xff,0x03,0x00]
+// GFX11: v_or_b16 v5.l, v255.l, v255.h op_sel:[0,1,0] ; encoding: [0x05,0x10,0x63,0xd7,0xff,0xff,0x03,0x00]
 
 v_or_b16 v255.h, 0xfe0b, vcc_hi
-// GFX11: [0xff,0x40,0x63,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
+// GFX11: v_or_b16 v255.h, 0xfe0b, vcc_hi op_sel:[0,0,1] ; encoding: [0xff,0x40,0x63,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
 v_pack_b32_f16 v5, v1.l, v2.l
 // GFX11: v_pack_b32_f16 v5, v1.l, v2.l           ; encoding: [0x05,0x00,0x11,0xd7,0x01,0x05,0x02,0x00]
@@ -6632,10 +6635,10 @@ v_xor_b16 v255.l, 0xfe0b, vcc_hi
 // GFX11: v_xor_b16 v255.l, 0xfe0b, vcc_hi        ; encoding: [0xff,0x00,0x64,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
 v_xor_b16 v5.l, v1.h, v2.l
-// GFX11: [0x05,0x08,0x64,0xd7,0x01,0x05,0x02,0x00]
+// GFX11: v_xor_b16 v5.l, v1.h, v2.l op_sel:[1,0,0] ; encoding: [0x05,0x08,0x64,0xd7,0x01,0x05,0x02,0x00]
 
 v_xor_b16 v5.l, v255.l, v255.h
-// GFX11: [0x05,0x10,0x64,0xd7,0xff,0xff,0x03,0x00]
+// GFX11: v_xor_b16 v5.l, v255.l, v255.h op_sel:[0,1,0] ; encoding: [0x05,0x10,0x64,0xd7,0xff,0xff,0x03,0x00]
 
 v_xor_b16 v255.h, 0xfe0b, vcc_hi
-// GFX11: [0xff,0x40,0x64,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
+// GFX11: v_xor_b16 v255.h, 0xfe0b, vcc_hi op_sel:[0,0,1] ; encoding: [0xff,0x40,0x64,0xd7,0xff,0xd6,0x00,0x00,0x0b,0xfe,0x00,0x00]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16.s
index 1864996b26028..6b514a02d843f 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16.s
@@ -321,20 +321,23 @@ v_add_nc_u16_e64_dpp v5.l, v1.l, v2.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bou
 v_add_nc_u16_e64_dpp v255.h, v255.l, v255.l clamp row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
 // GFX11: v_add_nc_u16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] clamp row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0xc0,0x03,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0]
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3]
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.h row_shl:1
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
 
 v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:15
 // GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x0f,0x01,0xff]
@@ -345,7 +348,7 @@ v_alignbit_b32_e64_dpp v5, v1, v2, vcc_hi row_shr:1
 v_alignbit_b32_e64_dpp v5, v1, v2, vcc_lo row_shr:15
 // GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, vcc_lo row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xaa,0x01,0x01,0x1f,0x01,0xff]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, ttmp15 row_ror:1
+v_alignbit_b32_e64_dpp v5, v1, v2, ttmp15 row_ror:1 row_mask:0xf bank_mask:0xf
 // GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, ttmp15 row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xee,0x01,0x01,0x21,0x01,0xff]
 
 v_alignbit_b32_e64_dpp v5, v1, v2, exec_hi row_ror:15
@@ -363,6 +366,24 @@ v_alignbit_b32_e64_dpp v5, v1, v2, -1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bou
 v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
 // GFX11: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x05,0x30]
 
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_mirror
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, s3 row_half_mirror
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x00,0x01,0x41,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:1
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x01,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, ttmp15 row_shl:15
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, ttmp15 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xee,0x01,0x01,0x0f,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, m0 row_ror:1
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, m0 row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xf6,0x01,0x01,0x21,0x01,0xff]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.h row_mirror
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+
 v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0]
 // GFX11: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 
@@ -466,16 +487,16 @@ v_and_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0
 // GFX11: v_and_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x00,0x62,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_and_b16_e64_dpp v5.h, v1.h, v2.h row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: [0x05,0x58,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
+// GFX11: v_and_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x58,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
 
 v_and_b16_e64_dpp v5.l, v1.h, v2.l row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: [0x05,0x08,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
+// GFX11: v_and_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0x05,0x08,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
 
 v_and_b16_e64_dpp v5.l, v1.l, v2.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: [0x05,0x10,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
+// GFX11: v_and_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0x05,0x10,0x62,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
 
 v_and_b16_e64_dpp v255.h, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: [0xff,0x40,0x62,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
+// GFX11: v_and_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x40,0x62,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_and_or_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0]
 // GFX11: v_and_or_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x57,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
@@ -4272,16 +4293,16 @@ v_or_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0
 // GFX11: v_or_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x00,0x63,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_or_b16_e64_dpp v5.h, v1.h, v2.h row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: [0x05,0x58,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
+// GFX11: v_or_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x58,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
 
 v_or_b16_e64_dpp v5.l, v1.h, v2.l row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: [0x05,0x08,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
+// GFX11: v_or_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0x05,0x08,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
 
 v_or_b16_e64_dpp v5.l, v1.l, v2.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: [0x05,0x10,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
+// GFX11: v_or_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0x05,0x10,0x63,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
 
 v_or_b16_e64_dpp v255.h, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: [0xff,0x40,0x63,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
+// GFX11: v_or_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x40,0x63,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_pack_b32_f16_e64_dpp v5, v1.l, v2.l quad_perm:[3,2,1,0]
 // GFX11: v_pack_b32_f16_e64_dpp v5, v1.l, v2.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x11,0xd7,0xfa,0x04,0x02,0x00,0x01,0x1b,0x00,0xff]
@@ -5008,16 +5029,16 @@ v_xor_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0
 // GFX11: v_xor_b16_e64_dpp v255.l, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x00,0x64,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_xor_b16_e64_dpp v5.h, v1.h, v2.h row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: [0x05,0x58,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
+// GFX11: v_xor_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x58,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x50,0x01,0xff]
 
 v_xor_b16_e64_dpp v5.l, v1.h, v2.l row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: [0x05,0x08,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
+// GFX11: v_xor_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0x05,0x08,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x5f,0x01,0x01]
 
 v_xor_b16_e64_dpp v5.l, v1.l, v2.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: [0x05,0x10,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
+// GFX11: v_xor_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0x05,0x10,0x64,0xd7,0xfa,0x04,0x02,0x00,0x01,0x60,0x09,0x13]
 
 v_xor_b16_e64_dpp v255.h, v255.l, v255.l row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: [0xff,0x40,0x64,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
+// GFX11: v_xor_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x40,0x64,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x05,0x30]
 
 v_cvt_pk_norm_i16_f16_e64_dpp v5, -v1.h, |v2.l| op_sel:[1,0,0] row_xmask:0 row_mask:0x1 bank_mask:0x3
 // GFX11: v_cvt_pk_norm_i16_f16_e64_dpp v5, -v1.h, |v2.l| op_sel:[1,0,0] row_xmask:0 row_mask:0x1 bank_mask:0x3 ; encoding: [0x05,0x0a,0x12,0xd7,0xfa,0x04,0x02,0x20,0x01,0x60,0x01,0x13]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8.s
index 300e5ef22f5ae..80ee725db319f 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8.s
@@ -154,11 +154,11 @@ v_add_nc_u16_e64_dpp v5.l, v1.l, v2.h dpp8:[7,6,5,4,3,2,1,0] fi:1
 v_add_nc_u16_e64_dpp v255.h, v255.l, v255.l clamp dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_add_nc_u16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] clamp dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0xc0,0x03,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 
-v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
 
 v_alignbit_b32_e64_dpp v5, v1, v2, s105 dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s105 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xa6,0x01,0x01,0x77,0x39,0x05]
@@ -187,6 +187,15 @@ v_alignbit_b32_e64_dpp v5, v1, v2, -1 dpp8:[7,6,5,4,3,2,1,0] fi:1
 v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x00,0x16,0xd6,0xe9,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00]
 
+v_alignbit_b32_e64_dpp v5, v1, v2, s3 dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x00,0x01,0x77,0x39,0x05]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, m0 dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, m0 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xf6,0x01,0x01,0x77,0x39,0x05]
+
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+
 v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 
@@ -239,16 +248,16 @@ v_and_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_and_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x00,0x62,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_and_b16_e64_dpp v5.h, v1.h, v2.h dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x58,0x62,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_and_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x58,0x62,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_and_b16_e64_dpp v5.l, v1.h, v2.l dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x08,0x62,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_and_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x08,0x62,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_and_b16_e64_dpp v5.l, v1.l, v2.h dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: [0x05,0x10,0x62,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_and_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0x05,0x10,0x62,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_and_b16_e64_dpp v255.h, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: [0xff,0x40,0x62,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
+// GFX11: v_and_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x40,0x62,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_and_or_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_and_or_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x57,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
@@ -2855,16 +2864,16 @@ v_or_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_or_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x00,0x63,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_or_b16_e64_dpp v5.h, v1.h, v2.h dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x58,0x63,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_or_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x58,0x63,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_or_b16_e64_dpp v5.l, v1.h, v2.l dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x08,0x63,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_or_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x08,0x63,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_or_b16_e64_dpp v5.l, v1.l, v2.h dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: [0x05,0x10,0x63,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_or_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0x05,0x10,0x63,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_or_b16_e64_dpp v255.h, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: [0xff,0x40,0x63,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
+// GFX11: v_or_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x40,0x63,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_pack_b32_f16_e64_dpp v5, v1.l, v2.l dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_pack_b32_f16_e64_dpp v5, v1.l, v2.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x11,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
@@ -3263,16 +3272,16 @@ v_xor_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_xor_b16_e64_dpp v255.l, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x00,0x64,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_xor_b16_e64_dpp v5.h, v1.h, v2.h dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x58,0x64,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_xor_b16_e64_dpp v5.h, v1.h, v2.h op_sel:[1,1,1] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x58,0x64,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_xor_b16_e64_dpp v5.l, v1.h, v2.l dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: [0x05,0x08,0x64,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_xor_b16_e64_dpp v5.l, v1.h, v2.l op_sel:[1,0,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x08,0x64,0xd7,0xe9,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_xor_b16_e64_dpp v5.l, v1.l, v2.h dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: [0x05,0x10,0x64,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
+// GFX11: v_xor_b16_e64_dpp v5.l, v1.l, v2.h op_sel:[0,1,0] dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0x05,0x10,0x64,0xd7,0xea,0x04,0x02,0x00,0x01,0x77,0x39,0x05]
 
 v_xor_b16_e64_dpp v255.h, v255.l, v255.l dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: [0xff,0x40,0x64,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
+// GFX11: v_xor_b16_e64_dpp v255.h, v255.l, v255.l op_sel:[0,0,1] dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x40,0x64,0xd7,0xe9,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 v_cvt_pk_norm_i16_f16_e64_dpp v5, -v1.h, |v2.l| op_sel:[1,0,0] dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_cvt_pk_norm_i16_f16_e64_dpp v5, -v1.h, |v2.l| op_sel:[1,0,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x0a,0x12,0xd7,0xe9,0x04,0x02,0x20,0x01,0x77,0x39,0x05]
diff --git a/llvm/test/MC/AMDGPU/gfx12_asm_vop3.s b/llvm/test/MC/AMDGPU/gfx12_asm_vop3.s
index 72d201e060df7..afd4af92b4f27 100644
--- a/llvm/test/MC/AMDGPU/gfx12_asm_vop3.s
+++ b/llvm/test/MC/AMDGPU/gfx12_asm_vop3.s
@@ -407,6 +407,9 @@ v_alignbit_b32 v5, src_scc, vcc_lo, -1
 v_alignbit_b32 v255, 0xaf123456, vcc_hi, null
 // GFX12: v_alignbit_b32 v255, 0xaf123456, vcc_hi, null ; encoding: [0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf]
 
+v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h
+// GFX12: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+
 v_alignbyte_b32 v5, v1, v2, s3
 // GFX12: v_alignbyte_b32 v5, v1, v2, s3          ; encoding: [0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00]
 
diff --git a/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp16.s b/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp16.s
index 08be9225890c2..cfd01ee02aa7e 100644
--- a/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp16.s
@@ -437,6 +437,9 @@ v_alignbit_b32_e64_dpp v5, v1, v2, -1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bou
 v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
 // GFX12: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x05,0x30]
 
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.h row_mirror
+// GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+
 v_alignbyte_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0]
 // GFX12: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 
diff --git a/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp8.s b/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp8.s
index 0523b0fe8c9a4..b6d395c8ede89 100644
--- a/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx12_asm_vop3_dpp8.s
@@ -249,6 +249,9 @@ v_alignbit_b32_e64_dpp v5, v1, v2, -1 dpp8:[7,6,5,4,3,2,1,0] fi:1
 v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX12: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xff,0x00,0x16,0xd6,0xe9,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00]
 
+v_alignbit_b32_e64_dpp v5, v1, v2, v255.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+
 v_alignbyte_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0]
 // GFX12: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3.txt
index b74128b21f563..1b29ddcb20a91 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3.txt
@@ -463,10 +463,16 @@
 # GFX11: v_alignbit_b32 v5, s105, s105, s105     ; encoding: [0x05,0x00,0x16,0xd6,0x69,0xd2,0xa4,0x01]
 
 0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04
-# GFX11: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W32-REAL16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W64-REAL16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
 
 0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf
-# GFX11: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
 
 0x05,0x00,0x16,0xd6,0x7b,0xfa,0xed,0x01
 # GFX11: v_alignbit_b32 v5, ttmp15, src_scc, ttmp15 ; encoding: [0x05,0x00,0x16,0xd6,0x7b,0xfa,0xed,0x01]
@@ -495,6 +501,12 @@
 0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf
 # GFX11: v_alignbit_b32 v255, 0xaf123456, vcc_hi, null ; encoding: [0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf]
 
+0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf
+# W32-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+
 0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00
 # GFX11: v_alignbyte_b32 v5, v1, v2, s3          ; encoding: [0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00]
 
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16.txt
index 8e7122b902326..0f47d33761475 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16.txt
@@ -186,19 +186,34 @@
 # GFX11: v_add_nc_i32_e64_dpp v255, v255, v255 clamp row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:1 fi:1 ; encoding: [0xff,0x80,0x26,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x0d,0x30]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x0f,0x01,0xff
 # GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x0f,0x01,0xff]
@@ -227,6 +242,12 @@
 0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x0d,0x30
 # GFX11: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:1 fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x0d,0x30]
 
+0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+
 0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff
 # W32-REAL16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 # W32-FAKE16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8.txt
index f67eb32385407..67bafd71d01e5 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8.txt
@@ -102,10 +102,16 @@
 # GFX11: v_add_nc_i32_e64_dpp v255, v255, v255 clamp dpp8:[0,0,0,0,0,0,0,0] fi:1 ; encoding: [0xff,0x80,0x26,0xd7,0xea,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05
-# GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0xa6,0x01,0x01,0x77,0x39,0x05
 # GFX11: v_alignbit_b32_e64_dpp v5, v1, v2, s105 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xa6,0x01,0x01,0x77,0x39,0x05]
@@ -134,6 +140,12 @@
 0xff,0x00,0x16,0xd6,0xea,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00
 # GFX11: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xea,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00]
 
+0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+
 0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05
 # W32-REAL16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 # W32-FAKE16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3.txt
index 3d5e78c86bc22..e7d92af7511f5 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3.txt
@@ -427,10 +427,16 @@
 # GFX12: v_alignbit_b32 v5, s105, s105, s105     ; encoding: [0x05,0x00,0x16,0xd6,0x69,0xd2,0xa4,0x01]
 
 0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04
-# GFX12: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W32-REAL16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W64-REAL16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3.l ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_lo, ttmp15, v3   ; encoding: [0x05,0x00,0x16,0xd6,0x6a,0xf6,0x0c,0x04]
 
 0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf
-# GFX12: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.l ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 ; encoding: [0x05,0x00,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
 
 0x05,0x00,0x16,0xd6,0x7b,0xfa,0xed,0x01
 # GFX12: v_alignbit_b32 v5, ttmp15, src_scc, ttmp15 ; encoding: [0x05,0x00,0x16,0xd6,0x7b,0xfa,0xed,0x01]
@@ -459,6 +465,12 @@
 0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf
 # GFX12: v_alignbit_b32 v255, 0xaf123456, vcc_hi, null ; encoding: [0xff,0x00,0x16,0xd6,0xff,0xd6,0xf0,0x01,0x56,0x34,0x12,0xaf]
 
+0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf
+# W32-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W32-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-REAL16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255.h op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+# W64-FAKE16: v_alignbit_b32 v5, vcc_hi, 0xaf123456, v255 op_sel:[0,0,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0x6b,0xfe,0xfd,0x07,0x56,0x34,0x12,0xaf]
+
 0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00
 # GFX12: v_alignbyte_b32 v5, v1, v2, s3          ; encoding: [0x05,0x00,0x17,0xd6,0x01,0x05,0x0e,0x00]
 
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp16.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp16.txt
index c63c09b6b24c5..a84e0e1734ff5 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp16.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp16.txt
@@ -195,22 +195,40 @@
 # GFX12: v_add_nc_i32_e64_dpp v255, v255, v255 clamp row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:1 fi:1 ; encoding: [0xff,0x80,0x26,0xd7,0xfa,0xfe,0x03,0x00,0xff,0x6f,0x0d,0x30]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, s3, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, s3, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, s3, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, s3, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, s3, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x06,0x0c,0x04,0x01,0x1b,0x00,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0xe4,0x00,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x40,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x41,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x01,0x01,0xff]
 
 0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x0f,0x01,0xff
 # GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, s105 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x16,0xd6,0xfa,0x04,0xa6,0x01,0x01,0x0f,0x01,0xff]
@@ -239,6 +257,12 @@
 0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x0d,0x30
 # GFX12: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:1 fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xfa,0xfe,0xf7,0x03,0xff,0x6f,0x0d,0x30]
 
+0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x20,0x16,0xd6,0xfa,0x04,0xfe,0x07,0x01,0x40,0x01,0xff]
+
 0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff
 # W32-REAL16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
 # W32-FAKE16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0x17,0xd6,0xfa,0x04,0x0e,0x04,0x01,0x1b,0x00,0xff]
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp8.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp8.txt
index 4182cd93f813c..e4bef0a17385c 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp8.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_dpp8.txt
@@ -111,13 +111,22 @@
 # GFX12: v_add_nc_i32_e64_dpp v255, v255, v255 clamp dpp8:[0,0,0,0,0,0,0,0] fi:1 ; encoding: [0xff,0x80,0x26,0xd7,0xea,0xfe,0x03,0x00,0xff,0x00,0x00,0x00]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 
 0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, s3, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, s3, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, s3, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, s3, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, s3, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x06,0x0c,0x04,0x01,0x77,0x39,0x05]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05
-# GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
 
 0x05,0x00,0x16,0xd6,0xe9,0x04,0xa6,0x01,0x01,0x77,0x39,0x05
 # GFX12: v_alignbit_b32_e64_dpp v5, v1, v2, s105 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x16,0xd6,0xe9,0x04,0xa6,0x01,0x01,0x77,0x39,0x05]
@@ -146,6 +155,12 @@
 0xff,0x00,0x16,0xd6,0xea,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00
 # GFX12: v_alignbit_b32_e64_dpp v255, v255, v255, src_scc dpp8:[0,0,0,0,0,0,0,0] fi:1 ; encoding: [0xff,0x00,0x16,0xd6,0xea,0xfe,0xf7,0x03,0xff,0x00,0x00,0x00]
 
+0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05
+# W32-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W32-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-REAL16: v_alignbit_b32_e64_dpp v5, v1, v2, v255.h op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+# W64-FAKE16: v_alignbit_b32_e64_dpp v5, v1, v2, v255 op_sel:[0,0,1,0] dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x20,0x16,0xd6,0xe9,0x04,0xfe,0x07,0x01,0x77,0x39,0x05]
+
 0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05
 # W32-REAL16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]
 # W32-FAKE16: v_alignbyte_b32_e64_dpp v5, v1, v2, v3 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0x05,0x00,0x17,0xd6,0xe9,0x04,0x0e,0x04,0x01,0x77,0x39,0x05]



More information about the llvm-commits mailing list