Skip to content

[AMDGPU][True16][MC] true16 for v_cos_f16 #120639

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 2 commits into from
Jan 3, 2025
Merged

[AMDGPU][True16][MC] true16 for v_cos_f16 #120639

merged 2 commits into from
Jan 3, 2025

Conversation

broxigarchen
Copy link
Contributor

@broxigarchen broxigarchen commented Dec 19, 2024 •
edited
Loading

Support true16 format for v_cos_f16 in MC

@broxigarchen broxigarchen changed the title True16 for v_cos_f16 true16 for v_cos_f16 Dec 19, 2024
@broxigarchen broxigarchen changed the title true16 for v_cos_f16 [AMDGPU][True16][MC] true16 for v_cos_f16 Dec 19, 2024
@broxigarchen broxigarchen force-pushed the main-merge-true16-mc-vop1-more-instruction-2 branch from eace722 to 9aab516 Compare December 20, 2024 05:16
@broxigarchen broxigarchen marked this pull request as ready for review December 20, 2024 14:46
@llvmbot llvmbot added backend:AMDGPU mc Machine (object) code labels Dec 20, 2024
@llvmbot
Copy link
Member

llvmbot commented Dec 20, 2024

@llvm/pr-subscribers-backend-amdgpu

Author: Brox Chen (broxigarchen)

Changes

Support true16 format for v_cos_f16 in MC

Patch is 145.37 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/120639.diff

28 Files Affected:

  • (modified) llvm/lib/Target/AMDGPU/VOP1Instructions.td (+1-1)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1.s (+45-30)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s (+15-6)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s (+42)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s (+110-44)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16_from_vop1.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8_from_vop1.s (+17-8)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_from_vop1.s (+39-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1.s (+42-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_dpp16.s (+34-28)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_dpp8.s (+12-6)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_t16_err.s (+42)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_t16_promote.s (+110-44)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1.s (+39-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1_dpp16.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1_dpp8.s (+17-8)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1.txt (+48-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1_dpp16.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1_dpp8.txt (+15-2)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16_from_vop1.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8_from_vop1.txt (+20-4)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_from_vop1.txt (+42-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop1_dpp16.txt (+36-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop1_dpp8.txt (+11-2)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1.txt (+42-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1_dpp16.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1_dpp8.txt (+20-4)
diff --git a/llvm/lib/Target/AMDGPU/VOP1Instructions.td b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
index 1dd39be9e8d9c7..9620c9320f5f67 100644
--- a/llvm/lib/Target/AMDGPU/VOP1Instructions.td
+++ b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
@@ -1046,7 +1046,7 @@ defm V_TRUNC_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05d, "v_trunc_f1
 defm V_RNDNE_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05e, "v_rndne_f16">;
 defm V_FRACT_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05f, "v_fract_f16">;
 defm V_SIN_F16_fake16        : VOP1_Real_FULL_t16_gfx11_gfx12<0x060, "v_sin_f16">;
-defm V_COS_F16_fake16        : VOP1_Real_FULL_t16_gfx11_gfx12<0x061, "v_cos_f16">;
+defm V_COS_F16               : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x061, "v_cos_f16">;
 defm V_SAT_PK_U8_I16_fake16  : VOP1_Real_FULL_t16_gfx11_gfx12<0x062, "v_sat_pk_u8_i16">;
 defm V_CVT_NORM_I16_F16      : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x063, "v_cvt_norm_i16_f16">;
 defm V_CVT_NORM_U16_F16      : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x064, "v_cvt_norm_u16_f16">;
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
index 4e4dc6647daebd..fcc30eca6ae5a7 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
@@ -269,50 +269,65 @@ v_clz_i32_u32 v5, src_scc
 v_clz_i32_u32 v255, 0xaf123456
 // GFX11: v_clz_i32_u32_e32 v255, 0xaf123456      ; encoding: [0xff,0x72,0xfe,0x7f,0x56,0x34,0x12,0xaf]
 
-v_cos_f16 v5, v1
-// GFX11: v_cos_f16_e32 v5, v1                    ; encoding: [0x01,0xc3,0x0a,0x7e]
+v_cos_f16 v5.l, v1.l
+// GFX11: v_cos_f16_e32 v5.l, v1.l                ; encoding: [0x01,0xc3,0x0a,0x7e]
 
-v_cos_f16 v5, v127
-// GFX11: v_cos_f16_e32 v5, v127                  ; encoding: [0x7f,0xc3,0x0a,0x7e]
+v_cos_f16 v5.l, v127.l
+// GFX11: v_cos_f16_e32 v5.l, v127.l              ; encoding: [0x7f,0xc3,0x0a,0x7e]
 
-v_cos_f16 v5, s1
-// GFX11: v_cos_f16_e32 v5, s1                    ; encoding: [0x01,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, s1
+// GFX11: v_cos_f16_e32 v5.l, s1                  ; encoding: [0x01,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, s105
-// GFX11: v_cos_f16_e32 v5, s105                  ; encoding: [0x69,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, s105
+// GFX11: v_cos_f16_e32 v5.l, s105                ; encoding: [0x69,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, vcc_lo
-// GFX11: v_cos_f16_e32 v5, vcc_lo                ; encoding: [0x6a,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, vcc_lo
+// GFX11: v_cos_f16_e32 v5.l, vcc_lo              ; encoding: [0x6a,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, vcc_hi
-// GFX11: v_cos_f16_e32 v5, vcc_hi                ; encoding: [0x6b,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, vcc_hi
+// GFX11: v_cos_f16_e32 v5.l, vcc_hi              ; encoding: [0x6b,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, ttmp15
-// GFX11: v_cos_f16_e32 v5, ttmp15                ; encoding: [0x7b,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, ttmp15
+// GFX11: v_cos_f16_e32 v5.l, ttmp15              ; encoding: [0x7b,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, m0
-// GFX11: v_cos_f16_e32 v5, m0                    ; encoding: [0x7d,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, m0
+// GFX11: v_cos_f16_e32 v5.l, m0                  ; encoding: [0x7d,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, exec_lo
-// GFX11: v_cos_f16_e32 v5, exec_lo               ; encoding: [0x7e,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, exec_lo
+// GFX11: v_cos_f16_e32 v5.l, exec_lo             ; encoding: [0x7e,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, exec_hi
-// GFX11: v_cos_f16_e32 v5, exec_hi               ; encoding: [0x7f,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, exec_hi
+// GFX11: v_cos_f16_e32 v5.l, exec_hi             ; encoding: [0x7f,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, null
-// GFX11: v_cos_f16_e32 v5, null                  ; encoding: [0x7c,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, null
+// GFX11: v_cos_f16_e32 v5.l, null                ; encoding: [0x7c,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, -1
-// GFX11: v_cos_f16_e32 v5, -1                    ; encoding: [0xc1,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, -1
+// GFX11: v_cos_f16_e32 v5.l, -1                  ; encoding: [0xc1,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, 0.5
-// GFX11: v_cos_f16_e32 v5, 0.5                   ; encoding: [0xf0,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, 0.5
+// GFX11: v_cos_f16_e32 v5.l, 0.5                 ; encoding: [0xf0,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, src_scc
-// GFX11: v_cos_f16_e32 v5, src_scc               ; encoding: [0xfd,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, src_scc
+// GFX11: v_cos_f16_e32 v5.l, src_scc             ; encoding: [0xfd,0xc2,0x0a,0x7e]
 
-v_cos_f16 v127, 0xfe0b
-// GFX11: v_cos_f16_e32 v127, 0xfe0b              ; encoding: [0xff,0xc2,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+v_cos_f16 v127.l, 0xfe0b
+// GFX11: v_cos_f16_e32 v127.l, 0xfe0b            ; encoding: [0xff,0xc2,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+
+v_cos_f16 v5.l, v1.h
+// GFX11: v_cos_f16_e32 v5.l, v1.h                ; encoding: [0x81,0xc3,0x0a,0x7e]
+
+v_cos_f16 v5.l, v127.h
+// GFX11: v_cos_f16_e32 v5.l, v127.h              ; encoding: [0xff,0xc3,0x0a,0x7e]
+
+v_cos_f16 v127.l, 0.5
+// GFX11: v_cos_f16_e32 v127.l, 0.5               ; encoding: [0xf0,0xc2,0xfe,0x7e]
+
+v_cos_f16 v5.h, src_scc
+// GFX11: v_cos_f16_e32 v5.h, src_scc             ; encoding: [0xfd,0xc2,0x0a,0x7f]
+
+v_cos_f16 v127.h, 0xfe0b
+// GFX11: v_cos_f16_e32 v127.h, 0xfe0b            ; encoding: [0xff,0xc2,0xfe,0x7f,0x0b,0xfe,0x00,0x00]
 
 v_cos_f32 v5, v1
 // GFX11: v_cos_f32_e32 v5, v1                    ; encoding: [0x01,0x6d,0x0a,0x7e]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
index 98e4b29b25666e..c878cc05e3a3b9 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
@@ -212,47 +212,56 @@ v_clz_i32_u32 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
 v_clz_i32_u32 v255, v255 row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
 // GFX11: v_clz_i32_u32_dpp v255, v255 row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0x72,0xfe,0x7f,0xff,0x6f,0x05,0x30]
 
-v_cos_f16 v5, v1 quad_perm:[3,2,1,0]
-// GFX11: v_cos_f16_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1b,0x00,0xff]
+v_cos_f16 v5.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: v_cos_f16_dpp v5.l, v1.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1b,0x00,0xff]
 
-v_cos_f16 v5, v1 quad_perm:[0,1,2,3]
-// GFX11: v_cos_f16_dpp v5, v1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0xe4,0x00,0xff]
+v_cos_f16 v5.l, v1.l quad_perm:[0,1,2,3]
+// GFX11: v_cos_f16_dpp v5.l, v1.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0xe4,0x00,0xff]
 
-v_cos_f16 v5, v1 row_mirror
-// GFX11: v_cos_f16_dpp v5, v1 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x40,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_mirror
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x40,0x01,0xff]
 
-v_cos_f16 v5, v1 row_half_mirror
-// GFX11: v_cos_f16_dpp v5, v1 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x41,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_half_mirror
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x41,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shl:1
-// GFX11: v_cos_f16_dpp v5, v1 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x01,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shl:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x01,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shl:15
-// GFX11: v_cos_f16_dpp v5, v1 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x0f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shl:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x0f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shr:1
-// GFX11: v_cos_f16_dpp v5, v1 row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x11,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shr:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x11,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shr:15
-// GFX11: v_cos_f16_dpp v5, v1 row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shr:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_ror:1
-// GFX11: v_cos_f16_dpp v5, v1 row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x21,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_ror:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x21,0x01,0xff]
 
-v_cos_f16 v5, v1 row_ror:15
-// GFX11: v_cos_f16_dpp v5, v1 row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x2f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_ror:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x2f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: v_cos_f16_dpp v5, v1 row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x50,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x50,0x01,0xff]
 
-v_cos_f16 v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: v_cos_f16_dpp v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x5f,0x01,0x01]
+v_cos_f16 v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x5f,0x01,0x01]
 
-v_cos_f16 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: v_cos_f16_dpp v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x60,0x09,0x13]
+v_cos_f16 v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x60,0x09,0x13]
 
-v_cos_f16 v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: v_cos_f16_dpp v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+v_cos_f16 v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1
+// GFX11: v_cos_f16_dpp v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+
+v_cos_f16 v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_cos_f16_dpp v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x5f,0x01,0x01]
+
+v_cos_f16 v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
+// GFX11: v_cos_f16_dpp v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7f,0x81,0x60,0x09,0x13]
+
+v_cos_f16 v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
+// GFX11: v_cos_f16_dpp v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7f,0xff,0x6f,0x35,0x30]
 
 v_cos_f32 v5, v1 quad_perm:[3,2,1,0]
 // GFX11: v_cos_f32_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0x6c,0x0a,0x7e,0x01,0x1b,0x00,0xff]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
index ab4606af2bb35c..06232292c4e329 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
@@ -50,14 +50,23 @@ v_clz_i32_u32 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
 v_clz_i32_u32 v255, v255 dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_clz_i32_u32_dpp v255, v255 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0x72,0xfe,0x7f,0xff,0x00,0x00,0x00]
 
-v_cos_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_cos_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_cos_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_cos_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
 
-v_cos_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: v_cos_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_cos_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
 
-v_cos_f16 v127, v127 dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: v_cos_f16_dpp v127, v127 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+v_cos_f16 v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0]
+// GFX11: v_cos_f16_dpp v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+
+v_cos_f16 v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_cos_f16_dpp v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x77,0x39,0x05]
+
+v_cos_f16 v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_cos_f16_dpp v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7f,0x81,0x77,0x39,0x05]
+
+v_cos_f16 v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] fi:0
+// GFX11: v_cos_f16_dpp v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7f,0xff,0x00,0x00,0x00]
 
 v_cos_f32 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_cos_f32_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0x6c,0x0a,0x7e,0x01,0x77,0x39,0x05]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
index 4ae91340386b65..6ff3c26d89725e 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
@@ -47,6 +47,12 @@ v_ceil_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
 v_cos_f16_e32 v128, 0xfe0b
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
+v_cos_f16_e32 v128.h, 0xfe0b
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v128.l, 0xfe0b
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
 v_cos_f16_e32 v255, v1
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
@@ -56,6 +62,24 @@ v_cos_f16_e32 v255, v1 dpp8:[7,6,5,4,3,2,1,0]
 v_cos_f16_e32 v255, v1 quad_perm:[3,2,1,0]
 // GFX11: :[[@LINE-1]]:24: error: invalid operand for instruction
 
+v_cos_f16_e32 v255.h, v1.h
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.h, v1.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.h, v1.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
 v_cos_f16_e32 v5, v199
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
@@ -65,6 +89,24 @@ v_cos_f16_e32 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
 v_cos_f16_e32 v5, v199 quad_perm:[3,2,1,0]
 // GFX11: :[[@LINE-1]]:24: error: invalid operand for instruction
 
+v_cos_f16_e32 v5.h, v199.h
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.h, v199.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.h, v199.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
 v_cvt_f16_f32_e32 v128, 0xaf123456
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
index 1d441720280ca2..2bafba8ee00445 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
@@ -68,71 +68,137 @@ v_ceil_f16 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
 v_ceil_f16 v5, v199 quad_perm:[3,2,1,0]
 // GFX11: v_ceil_f16_e64_dpp v5, v199 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0xdc,0xd5,0xfa,0x00,0x00,0x00,0xc7,0x1b,0x00,0xff]
 
-v_cos_f16 v128, 0xfe0b
-// GFX11: v_cos_f16_e64 v128, 0xfe0b              ; encoding: [0x80,0x00,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
+v_cos_f16 v128.h, 0xfe0b
+// GFX11: v_cos_f16_e64 v128.h, 0xfe0b op_sel:[0,1] ; encoding: [0x80,0x40,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
-v_cos_f16 v255, -1
-// GFX11: v_cos_f16_e64 v255, -1                  ; encoding: [0xff,0x00,0xe1,0xd5,0xc1,0x00,0x00,0x00]
+v_cos_f16 v128.l, 0xfe0b
+// GFX11: v_cos_f16_e64 v128.l, 0xfe0b            ; encoding: [0x80,0x00,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
-v_cos_f16 v255, 0.5
-// GFX11: v_cos_f16_e64 v255, 0.5                 ; encoding: [0xff,0x00,0xe1,0xd5,0xf0,0x00,0x00,0x00]
+v_cos_f16 v255.h, -1
+// GFX11: v_cos_f16_e64 v255.h, -1 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0xc1,0x00,0x00,0x00]
 
-v_cos_f16 v255, exec_hi
-// GFX11: v_cos_f16_e64 v255, exec_hi             ; encoding: [0xff,0x00,0xe1,0xd5,0x7f,0x00,0x00,0x00]
+v_cos_f16 v255.h, 0.5
+// GFX11: v_cos_f16_e64 v255.h, 0.5 op_sel:[0,1]  ; encoding: [0xff,0x40,0xe1,0xd5,0xf0,0x00,0x00,0x00]
 
-v_cos_f16 v255, exec_lo
-// GFX11: v_cos_f16_e64 v255, exec_lo             ; encoding: [0xff,0x00,0xe1,0xd5,0x7e,0x00,0x00,0x00]
+v_cos_f16 v255.h, exec_hi
+// GFX11: v_cos_f16_e64 v255.h, exec_hi op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7f,0x00,0x00,0x00]
 
-v_cos_f16 v255, m0
-// GFX11: v_cos_f16_e64 v255, m0                  ; encoding: [0xff,0x00,0xe1,0xd5,0x7d,0x00,0x00,0x00]
+v_cos_f16 v255.h, exec_lo
+// GFX11: v_cos_f16_e64 v255.h, exec_lo op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7e,0x00,0x00,0x00]
 
-v_cos_f16 v255, null
-// GFX11: v_cos_f16_e64 v255, null                ; encoding: [0xff,0x00,0xe1,0xd5,0x7c,0x00,0x00,0x00]
+v_cos_f16 v255.h, m0
+// GFX11: v_cos_f16_e64 v255.h, m0 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0x7d,0x00,0x00,0x00]
 
-v_cos_f16 v255, s1
-// GFX11: v_cos_f16_e64 v255, s1                  ; encoding: [0xff,0x00,0xe1,0xd5,0x01,0x00,0x00,0x00]
+v_cos_f16 v255.h, null
+// GFX11: v_cos_f16_e64 v255.h, null op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7c,0x00,0x00,0x00]
 
-v_cos_f16 v255, s105
-// GFX11: v_cos_f16_e64 v255, s105                ; encoding: [0xff,0x00,0xe1,0xd5,0x69,0x00,0x00,0x00]
+v_cos_f16 v255.h, s1
+// GFX11: v_cos_f16_e64 v255.h, s1 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0x01,0x00,0x00,0x00]
 
-v_cos_f16 v255, src_scc
-// GFX11: v_cos_f16_e64 v255, src_scc             ; encoding: [0xff,0x00,0xe1,0xd5,0xfd,0x00,0x00,0x00]
+v_cos_f16 v255.h, s105
+// GFX11: v_cos_f16_e64 v255.h, s105 op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x69,0x00,0x00,0x00]
 
-v_cos_f16 v255, ttmp15
-// GFX11: v_cos_f16_e64 v255, ttmp15              ; encoding: [0xff,0x00,0xe1,0xd5,0x7b,0x00,0x00,0x00]
+v_cos_f16 v255.h, src_scc
+// GFX11: v_cos_f16_e64 v255.h, src_scc op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0xfd,0x00,0x00,0x00]
 
-v_cos_f16 v255, v1
-// GFX11: v_cos_f16_e64 v255, v1                  ; encoding: [0xff,0x00,0xe1,0xd5,0x01,0x01,0x00,0x00]
+v_cos_f16 v255.h, ttmp15
+// GFX11: v_cos_f16_e64 v255.h, ttmp15 op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7b,0x00,0x00,0x00]
 
-v_cos_f16 v255, v1 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_cos_f16_e64_dpp v255, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xff,0x00,0xe1,0xd5,0xe9,0x00,0x00,0x00,0x01,0x77,0x39,0x05]
+v_cos_f16 v255.h, v1.h
+// GFX11: v_cos_f16_e64 v255.h, v1.h op_sel:[1,1] ; encoding: [0xff,0x48,0xe1,0xd5,0x01,0x01,0x00,0x00]
 
-v_cos_f1...
[truncated]

@llvmbot
Copy link
Member

llvmbot commented Dec 20, 2024

@llvm/pr-subscribers-mc

Author: Brox Chen (broxigarchen)

Changes

Support true16 format for v_cos_f16 in MC

Patch is 145.37 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/120639.diff

28 Files Affected:

  • (modified) llvm/lib/Target/AMDGPU/VOP1Instructions.td (+1-1)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1.s (+45-30)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s (+15-6)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s (+42)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s (+110-44)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp16_from_vop1.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_dpp8_from_vop1.s (+17-8)
  • (modified) llvm/test/MC/AMDGPU/gfx11_asm_vop3_from_vop1.s (+39-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1.s (+42-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_dpp16.s (+34-28)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_dpp8.s (+12-6)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_t16_err.s (+42)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop1_t16_promote.s (+110-44)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1.s (+39-30)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1_dpp16.s (+37-28)
  • (modified) llvm/test/MC/AMDGPU/gfx12_asm_vop3_from_vop1_dpp8.s (+17-8)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1.txt (+48-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1_dpp16.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop1_dpp8.txt (+15-2)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp16_from_vop1.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_dpp8_from_vop1.txt (+20-4)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx11_dasm_vop3_from_vop1.txt (+42-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop1_dpp16.txt (+36-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop1_dpp8.txt (+11-2)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1.txt (+42-15)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1_dpp16.txt (+40-14)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx12_dasm_vop3_from_vop1_dpp8.txt (+20-4)
diff --git a/llvm/lib/Target/AMDGPU/VOP1Instructions.td b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
index 1dd39be9e8d9c7..9620c9320f5f67 100644
--- a/llvm/lib/Target/AMDGPU/VOP1Instructions.td
+++ b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
@@ -1046,7 +1046,7 @@ defm V_TRUNC_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05d, "v_trunc_f1
 defm V_RNDNE_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05e, "v_rndne_f16">;
 defm V_FRACT_F16_fake16      : VOP1_Real_FULL_t16_gfx11_gfx12<0x05f, "v_fract_f16">;
 defm V_SIN_F16_fake16        : VOP1_Real_FULL_t16_gfx11_gfx12<0x060, "v_sin_f16">;
-defm V_COS_F16_fake16        : VOP1_Real_FULL_t16_gfx11_gfx12<0x061, "v_cos_f16">;
+defm V_COS_F16               : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x061, "v_cos_f16">;
 defm V_SAT_PK_U8_I16_fake16  : VOP1_Real_FULL_t16_gfx11_gfx12<0x062, "v_sat_pk_u8_i16">;
 defm V_CVT_NORM_I16_F16      : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x063, "v_cvt_norm_i16_f16">;
 defm V_CVT_NORM_U16_F16      : VOP1_Real_FULL_t16_and_fake16_gfx11_gfx12<0x064, "v_cvt_norm_u16_f16">;
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
index 4e4dc6647daebd..fcc30eca6ae5a7 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
@@ -269,50 +269,65 @@ v_clz_i32_u32 v5, src_scc
 v_clz_i32_u32 v255, 0xaf123456
 // GFX11: v_clz_i32_u32_e32 v255, 0xaf123456      ; encoding: [0xff,0x72,0xfe,0x7f,0x56,0x34,0x12,0xaf]
 
-v_cos_f16 v5, v1
-// GFX11: v_cos_f16_e32 v5, v1                    ; encoding: [0x01,0xc3,0x0a,0x7e]
+v_cos_f16 v5.l, v1.l
+// GFX11: v_cos_f16_e32 v5.l, v1.l                ; encoding: [0x01,0xc3,0x0a,0x7e]
 
-v_cos_f16 v5, v127
-// GFX11: v_cos_f16_e32 v5, v127                  ; encoding: [0x7f,0xc3,0x0a,0x7e]
+v_cos_f16 v5.l, v127.l
+// GFX11: v_cos_f16_e32 v5.l, v127.l              ; encoding: [0x7f,0xc3,0x0a,0x7e]
 
-v_cos_f16 v5, s1
-// GFX11: v_cos_f16_e32 v5, s1                    ; encoding: [0x01,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, s1
+// GFX11: v_cos_f16_e32 v5.l, s1                  ; encoding: [0x01,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, s105
-// GFX11: v_cos_f16_e32 v5, s105                  ; encoding: [0x69,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, s105
+// GFX11: v_cos_f16_e32 v5.l, s105                ; encoding: [0x69,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, vcc_lo
-// GFX11: v_cos_f16_e32 v5, vcc_lo                ; encoding: [0x6a,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, vcc_lo
+// GFX11: v_cos_f16_e32 v5.l, vcc_lo              ; encoding: [0x6a,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, vcc_hi
-// GFX11: v_cos_f16_e32 v5, vcc_hi                ; encoding: [0x6b,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, vcc_hi
+// GFX11: v_cos_f16_e32 v5.l, vcc_hi              ; encoding: [0x6b,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, ttmp15
-// GFX11: v_cos_f16_e32 v5, ttmp15                ; encoding: [0x7b,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, ttmp15
+// GFX11: v_cos_f16_e32 v5.l, ttmp15              ; encoding: [0x7b,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, m0
-// GFX11: v_cos_f16_e32 v5, m0                    ; encoding: [0x7d,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, m0
+// GFX11: v_cos_f16_e32 v5.l, m0                  ; encoding: [0x7d,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, exec_lo
-// GFX11: v_cos_f16_e32 v5, exec_lo               ; encoding: [0x7e,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, exec_lo
+// GFX11: v_cos_f16_e32 v5.l, exec_lo             ; encoding: [0x7e,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, exec_hi
-// GFX11: v_cos_f16_e32 v5, exec_hi               ; encoding: [0x7f,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, exec_hi
+// GFX11: v_cos_f16_e32 v5.l, exec_hi             ; encoding: [0x7f,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, null
-// GFX11: v_cos_f16_e32 v5, null                  ; encoding: [0x7c,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, null
+// GFX11: v_cos_f16_e32 v5.l, null                ; encoding: [0x7c,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, -1
-// GFX11: v_cos_f16_e32 v5, -1                    ; encoding: [0xc1,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, -1
+// GFX11: v_cos_f16_e32 v5.l, -1                  ; encoding: [0xc1,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, 0.5
-// GFX11: v_cos_f16_e32 v5, 0.5                   ; encoding: [0xf0,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, 0.5
+// GFX11: v_cos_f16_e32 v5.l, 0.5                 ; encoding: [0xf0,0xc2,0x0a,0x7e]
 
-v_cos_f16 v5, src_scc
-// GFX11: v_cos_f16_e32 v5, src_scc               ; encoding: [0xfd,0xc2,0x0a,0x7e]
+v_cos_f16 v5.l, src_scc
+// GFX11: v_cos_f16_e32 v5.l, src_scc             ; encoding: [0xfd,0xc2,0x0a,0x7e]
 
-v_cos_f16 v127, 0xfe0b
-// GFX11: v_cos_f16_e32 v127, 0xfe0b              ; encoding: [0xff,0xc2,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+v_cos_f16 v127.l, 0xfe0b
+// GFX11: v_cos_f16_e32 v127.l, 0xfe0b            ; encoding: [0xff,0xc2,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+
+v_cos_f16 v5.l, v1.h
+// GFX11: v_cos_f16_e32 v5.l, v1.h                ; encoding: [0x81,0xc3,0x0a,0x7e]
+
+v_cos_f16 v5.l, v127.h
+// GFX11: v_cos_f16_e32 v5.l, v127.h              ; encoding: [0xff,0xc3,0x0a,0x7e]
+
+v_cos_f16 v127.l, 0.5
+// GFX11: v_cos_f16_e32 v127.l, 0.5               ; encoding: [0xf0,0xc2,0xfe,0x7e]
+
+v_cos_f16 v5.h, src_scc
+// GFX11: v_cos_f16_e32 v5.h, src_scc             ; encoding: [0xfd,0xc2,0x0a,0x7f]
+
+v_cos_f16 v127.h, 0xfe0b
+// GFX11: v_cos_f16_e32 v127.h, 0xfe0b            ; encoding: [0xff,0xc2,0xfe,0x7f,0x0b,0xfe,0x00,0x00]
 
 v_cos_f32 v5, v1
 // GFX11: v_cos_f32_e32 v5, v1                    ; encoding: [0x01,0x6d,0x0a,0x7e]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
index 98e4b29b25666e..c878cc05e3a3b9 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
@@ -212,47 +212,56 @@ v_clz_i32_u32 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
 v_clz_i32_u32 v255, v255 row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
 // GFX11: v_clz_i32_u32_dpp v255, v255 row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0x72,0xfe,0x7f,0xff,0x6f,0x05,0x30]
 
-v_cos_f16 v5, v1 quad_perm:[3,2,1,0]
-// GFX11: v_cos_f16_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1b,0x00,0xff]
+v_cos_f16 v5.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: v_cos_f16_dpp v5.l, v1.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1b,0x00,0xff]
 
-v_cos_f16 v5, v1 quad_perm:[0,1,2,3]
-// GFX11: v_cos_f16_dpp v5, v1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0xe4,0x00,0xff]
+v_cos_f16 v5.l, v1.l quad_perm:[0,1,2,3]
+// GFX11: v_cos_f16_dpp v5.l, v1.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0xe4,0x00,0xff]
 
-v_cos_f16 v5, v1 row_mirror
-// GFX11: v_cos_f16_dpp v5, v1 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x40,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_mirror
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x40,0x01,0xff]
 
-v_cos_f16 v5, v1 row_half_mirror
-// GFX11: v_cos_f16_dpp v5, v1 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x41,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_half_mirror
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x41,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shl:1
-// GFX11: v_cos_f16_dpp v5, v1 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x01,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shl:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x01,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shl:15
-// GFX11: v_cos_f16_dpp v5, v1 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x0f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shl:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x0f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shr:1
-// GFX11: v_cos_f16_dpp v5, v1 row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x11,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shr:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x11,0x01,0xff]
 
-v_cos_f16 v5, v1 row_shr:15
-// GFX11: v_cos_f16_dpp v5, v1 row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_shr:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x1f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_ror:1
-// GFX11: v_cos_f16_dpp v5, v1 row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x21,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_ror:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x21,0x01,0xff]
 
-v_cos_f16 v5, v1 row_ror:15
-// GFX11: v_cos_f16_dpp v5, v1 row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x2f,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_ror:15
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x2f,0x01,0xff]
 
-v_cos_f16 v5, v1 row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: v_cos_f16_dpp v5, v1 row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x50,0x01,0xff]
+v_cos_f16 v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x50,0x01,0xff]
 
-v_cos_f16 v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: v_cos_f16_dpp v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x5f,0x01,0x01]
+v_cos_f16 v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x5f,0x01,0x01]
 
-v_cos_f16 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: v_cos_f16_dpp v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x60,0x09,0x13]
+v_cos_f16 v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7e,0x01,0x60,0x09,0x13]
 
-v_cos_f16 v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: v_cos_f16_dpp v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+v_cos_f16 v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1
+// GFX11: v_cos_f16_dpp v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+
+v_cos_f16 v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_cos_f16_dpp v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xc2,0xfe,0x7e,0x7f,0x5f,0x01,0x01]
+
+v_cos_f16 v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
+// GFX11: v_cos_f16_dpp v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xc2,0x0a,0x7f,0x81,0x60,0x09,0x13]
+
+v_cos_f16 v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
+// GFX11: v_cos_f16_dpp v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xc2,0xfe,0x7f,0xff,0x6f,0x35,0x30]
 
 v_cos_f32 v5, v1 quad_perm:[3,2,1,0]
 // GFX11: v_cos_f32_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0x6c,0x0a,0x7e,0x01,0x1b,0x00,0xff]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
index ab4606af2bb35c..06232292c4e329 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
@@ -50,14 +50,23 @@ v_clz_i32_u32 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
 v_clz_i32_u32 v255, v255 dpp8:[0,0,0,0,0,0,0,0] fi:0
 // GFX11: v_clz_i32_u32_dpp v255, v255 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0x72,0xfe,0x7f,0xff,0x00,0x00,0x00]
 
-v_cos_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_cos_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_cos_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_cos_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
 
-v_cos_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: v_cos_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_cos_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_cos_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7e,0x01,0x77,0x39,0x05]
 
-v_cos_f16 v127, v127 dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: v_cos_f16_dpp v127, v127 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+v_cos_f16 v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0]
+// GFX11: v_cos_f16_dpp v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+
+v_cos_f16 v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_cos_f16_dpp v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xc2,0xfe,0x7e,0x7f,0x77,0x39,0x05]
+
+v_cos_f16 v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_cos_f16_dpp v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xc2,0x0a,0x7f,0x81,0x77,0x39,0x05]
+
+v_cos_f16 v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] fi:0
+// GFX11: v_cos_f16_dpp v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xc2,0xfe,0x7f,0xff,0x00,0x00,0x00]
 
 v_cos_f32 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
 // GFX11: v_cos_f32_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0x6c,0x0a,0x7e,0x01,0x77,0x39,0x05]
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
index 4ae91340386b65..6ff3c26d89725e 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
@@ -47,6 +47,12 @@ v_ceil_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
 v_cos_f16_e32 v128, 0xfe0b
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
+v_cos_f16_e32 v128.h, 0xfe0b
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v128.l, 0xfe0b
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
 v_cos_f16_e32 v255, v1
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
@@ -56,6 +62,24 @@ v_cos_f16_e32 v255, v1 dpp8:[7,6,5,4,3,2,1,0]
 v_cos_f16_e32 v255, v1 quad_perm:[3,2,1,0]
 // GFX11: :[[@LINE-1]]:24: error: invalid operand for instruction
 
+v_cos_f16_e32 v255.h, v1.h
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.h, v1.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.h, v1.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
+v_cos_f16_e32 v255.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
+
 v_cos_f16_e32 v5, v199
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
@@ -65,6 +89,24 @@ v_cos_f16_e32 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
 v_cos_f16_e32 v5, v199 quad_perm:[3,2,1,0]
 // GFX11: :[[@LINE-1]]:24: error: invalid operand for instruction
 
+v_cos_f16_e32 v5.h, v199.h
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.h, v199.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.h, v199.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
+v_cos_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:21: error: invalid operand for instruction
+
 v_cvt_f16_f32_e32 v128, 0xaf123456
 // GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
 
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
index 1d441720280ca2..2bafba8ee00445 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
@@ -68,71 +68,137 @@ v_ceil_f16 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
 v_ceil_f16 v5, v199 quad_perm:[3,2,1,0]
 // GFX11: v_ceil_f16_e64_dpp v5, v199 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0xdc,0xd5,0xfa,0x00,0x00,0x00,0xc7,0x1b,0x00,0xff]
 
-v_cos_f16 v128, 0xfe0b
-// GFX11: v_cos_f16_e64 v128, 0xfe0b              ; encoding: [0x80,0x00,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
+v_cos_f16 v128.h, 0xfe0b
+// GFX11: v_cos_f16_e64 v128.h, 0xfe0b op_sel:[0,1] ; encoding: [0x80,0x40,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
-v_cos_f16 v255, -1
-// GFX11: v_cos_f16_e64 v255, -1                  ; encoding: [0xff,0x00,0xe1,0xd5,0xc1,0x00,0x00,0x00]
+v_cos_f16 v128.l, 0xfe0b
+// GFX11: v_cos_f16_e64 v128.l, 0xfe0b            ; encoding: [0x80,0x00,0xe1,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
 
-v_cos_f16 v255, 0.5
-// GFX11: v_cos_f16_e64 v255, 0.5                 ; encoding: [0xff,0x00,0xe1,0xd5,0xf0,0x00,0x00,0x00]
+v_cos_f16 v255.h, -1
+// GFX11: v_cos_f16_e64 v255.h, -1 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0xc1,0x00,0x00,0x00]
 
-v_cos_f16 v255, exec_hi
-// GFX11: v_cos_f16_e64 v255, exec_hi             ; encoding: [0xff,0x00,0xe1,0xd5,0x7f,0x00,0x00,0x00]
+v_cos_f16 v255.h, 0.5
+// GFX11: v_cos_f16_e64 v255.h, 0.5 op_sel:[0,1]  ; encoding: [0xff,0x40,0xe1,0xd5,0xf0,0x00,0x00,0x00]
 
-v_cos_f16 v255, exec_lo
-// GFX11: v_cos_f16_e64 v255, exec_lo             ; encoding: [0xff,0x00,0xe1,0xd5,0x7e,0x00,0x00,0x00]
+v_cos_f16 v255.h, exec_hi
+// GFX11: v_cos_f16_e64 v255.h, exec_hi op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7f,0x00,0x00,0x00]
 
-v_cos_f16 v255, m0
-// GFX11: v_cos_f16_e64 v255, m0                  ; encoding: [0xff,0x00,0xe1,0xd5,0x7d,0x00,0x00,0x00]
+v_cos_f16 v255.h, exec_lo
+// GFX11: v_cos_f16_e64 v255.h, exec_lo op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7e,0x00,0x00,0x00]
 
-v_cos_f16 v255, null
-// GFX11: v_cos_f16_e64 v255, null                ; encoding: [0xff,0x00,0xe1,0xd5,0x7c,0x00,0x00,0x00]
+v_cos_f16 v255.h, m0
+// GFX11: v_cos_f16_e64 v255.h, m0 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0x7d,0x00,0x00,0x00]
 
-v_cos_f16 v255, s1
-// GFX11: v_cos_f16_e64 v255, s1                  ; encoding: [0xff,0x00,0xe1,0xd5,0x01,0x00,0x00,0x00]
+v_cos_f16 v255.h, null
+// GFX11: v_cos_f16_e64 v255.h, null op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7c,0x00,0x00,0x00]
 
-v_cos_f16 v255, s105
-// GFX11: v_cos_f16_e64 v255, s105                ; encoding: [0xff,0x00,0xe1,0xd5,0x69,0x00,0x00,0x00]
+v_cos_f16 v255.h, s1
+// GFX11: v_cos_f16_e64 v255.h, s1 op_sel:[0,1]   ; encoding: [0xff,0x40,0xe1,0xd5,0x01,0x00,0x00,0x00]
 
-v_cos_f16 v255, src_scc
-// GFX11: v_cos_f16_e64 v255, src_scc             ; encoding: [0xff,0x00,0xe1,0xd5,0xfd,0x00,0x00,0x00]
+v_cos_f16 v255.h, s105
+// GFX11: v_cos_f16_e64 v255.h, s105 op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x69,0x00,0x00,0x00]
 
-v_cos_f16 v255, ttmp15
-// GFX11: v_cos_f16_e64 v255, ttmp15              ; encoding: [0xff,0x00,0xe1,0xd5,0x7b,0x00,0x00,0x00]
+v_cos_f16 v255.h, src_scc
+// GFX11: v_cos_f16_e64 v255.h, src_scc op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0xfd,0x00,0x00,0x00]
 
-v_cos_f16 v255, v1
-// GFX11: v_cos_f16_e64 v255, v1                  ; encoding: [0xff,0x00,0xe1,0xd5,0x01,0x01,0x00,0x00]
+v_cos_f16 v255.h, ttmp15
+// GFX11: v_cos_f16_e64 v255.h, ttmp15 op_sel:[0,1] ; encoding: [0xff,0x40,0xe1,0xd5,0x7b,0x00,0x00,0x00]
 
-v_cos_f16 v255, v1 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_cos_f16_e64_dpp v255, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xff,0x00,0xe1,0xd5,0xe9,0x00,0x00,0x00,0x01,0x77,0x39,0x05]
+v_cos_f16 v255.h, v1.h
+// GFX11: v_cos_f16_e64 v255.h, v1.h op_sel:[1,1] ; encoding: [0xff,0x48,0xe1,0xd5,0x01,0x01,0x00,0x00]
 
-v_cos_f1...
[truncated]

@broxigarchen broxigarchen force-pushed the main-merge-true16-mc-vop1-more-instruction-2 branch from 9aab516 to d696c9c Compare January 3, 2025 16:46
@broxigarchen broxigarchen force-pushed the main-merge-true16-mc-vop1-more-instruction-2 branch from d696c9c to 367906a Compare January 3, 2025 16:48
@broxigarchen
Copy link
Contributor Author

Added a GFX12 runline in codegen test. Rebased

@broxigarchen broxigarchen merged commit bf274b3 into llvm:main Jan 3, 2025
8 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@arsenm arsenm arsenm approved these changes

@Sisyph Sisyph Awaiting requested review from Sisyph

@kosarev kosarev Awaiting requested review from kosarev

Assignees
No one assigned
Labels
backend:AMDGPU mc Machine (object) code
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@broxigarchen @llvmbot @arsenm