[AMDGPU][True16][MC] true16 for v_rndne_f16 #120691
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@llvm/pr-subscribers-mc @llvm/pr-subscribers-backend-amdgpu Author: Brox Chen (broxigarchen) ChangesSupport true16 format for v_rndne_b16 in MC Patch is 147.62 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/120691.diff 28 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/VOP1Instructions.td b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
index 1dd39be9e8d9c7..943b44c17d7443 100644
--- a/llvm/lib/Target/AMDGPU/VOP1Instructions.td
+++ b/llvm/lib/Target/AMDGPU/VOP1Instructions.td
@@ -1043,7 +1043,7 @@ defm V_FLOOR_F16_fake16 : VOP1_Real_FULL_t16_gfx11_gfx12<0x05b, "v_floor_f1
defm V_CEIL_F16_t16 : VOP1_Real_FULL_t16_gfx11_gfx12<0x05c, "v_ceil_f16">;
defm V_CEIL_F16_fake16 : VOP1_Real_FULL_t16_gfx11_gfx12<0x05c, "v_ceil_f16">;
defm V_TRUNC_F16_fake16 : VOP1_Real_FULL_t16_gfx11_gfx12<0x05d, "v_trunc_f16">;
-defm V_RNDNE_F16_fake16 : VOP1_Real_FULL_t16_gfx11_gfx12<0x05e, "v_rndne_f16">;
+defm V_RNDNE_F16 : VOP1_Real_FULL_t16_and_fake16_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">;
diff --git a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
index 4e4dc6647daebd..dcec10b04d1601 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1.s
@@ -2978,50 +2978,65 @@ v_readfirstlane_b32 ttmp15, v1
v_readfirstlane_b32 null, v255
// GFX11: v_readfirstlane_b32 null, v255 ; encoding: [0xff,0x05,0xf8,0x7e]
-v_rndne_f16 v5, v1
-// GFX11: v_rndne_f16_e32 v5, v1 ; encoding: [0x01,0xbd,0x0a,0x7e]
+v_rndne_f16 v5.l, v1.l
+// GFX11: v_rndne_f16_e32 v5.l, v1.l ; encoding: [0x01,0xbd,0x0a,0x7e]
-v_rndne_f16 v5, v127
-// GFX11: v_rndne_f16_e32 v5, v127 ; encoding: [0x7f,0xbd,0x0a,0x7e]
+v_rndne_f16 v5.l, v127.l
+// GFX11: v_rndne_f16_e32 v5.l, v127.l ; encoding: [0x7f,0xbd,0x0a,0x7e]
-v_rndne_f16 v5, s1
-// GFX11: v_rndne_f16_e32 v5, s1 ; encoding: [0x01,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, s1
+// GFX11: v_rndne_f16_e32 v5.l, s1 ; encoding: [0x01,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, s105
-// GFX11: v_rndne_f16_e32 v5, s105 ; encoding: [0x69,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, s105
+// GFX11: v_rndne_f16_e32 v5.l, s105 ; encoding: [0x69,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, vcc_lo
-// GFX11: v_rndne_f16_e32 v5, vcc_lo ; encoding: [0x6a,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, vcc_lo
+// GFX11: v_rndne_f16_e32 v5.l, vcc_lo ; encoding: [0x6a,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, vcc_hi
-// GFX11: v_rndne_f16_e32 v5, vcc_hi ; encoding: [0x6b,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, vcc_hi
+// GFX11: v_rndne_f16_e32 v5.l, vcc_hi ; encoding: [0x6b,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, ttmp15
-// GFX11: v_rndne_f16_e32 v5, ttmp15 ; encoding: [0x7b,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, ttmp15
+// GFX11: v_rndne_f16_e32 v5.l, ttmp15 ; encoding: [0x7b,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, m0
-// GFX11: v_rndne_f16_e32 v5, m0 ; encoding: [0x7d,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, m0
+// GFX11: v_rndne_f16_e32 v5.l, m0 ; encoding: [0x7d,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, exec_lo
-// GFX11: v_rndne_f16_e32 v5, exec_lo ; encoding: [0x7e,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, exec_lo
+// GFX11: v_rndne_f16_e32 v5.l, exec_lo ; encoding: [0x7e,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, exec_hi
-// GFX11: v_rndne_f16_e32 v5, exec_hi ; encoding: [0x7f,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, exec_hi
+// GFX11: v_rndne_f16_e32 v5.l, exec_hi ; encoding: [0x7f,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, null
-// GFX11: v_rndne_f16_e32 v5, null ; encoding: [0x7c,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, null
+// GFX11: v_rndne_f16_e32 v5.l, null ; encoding: [0x7c,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, -1
-// GFX11: v_rndne_f16_e32 v5, -1 ; encoding: [0xc1,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, -1
+// GFX11: v_rndne_f16_e32 v5.l, -1 ; encoding: [0xc1,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, 0.5
-// GFX11: v_rndne_f16_e32 v5, 0.5 ; encoding: [0xf0,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, 0.5
+// GFX11: v_rndne_f16_e32 v5.l, 0.5 ; encoding: [0xf0,0xbc,0x0a,0x7e]
-v_rndne_f16 v5, src_scc
-// GFX11: v_rndne_f16_e32 v5, src_scc ; encoding: [0xfd,0xbc,0x0a,0x7e]
+v_rndne_f16 v5.l, src_scc
+// GFX11: v_rndne_f16_e32 v5.l, src_scc ; encoding: [0xfd,0xbc,0x0a,0x7e]
-v_rndne_f16 v127, 0xfe0b
-// GFX11: v_rndne_f16_e32 v127, 0xfe0b ; encoding: [0xff,0xbc,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+v_rndne_f16 v127.l, 0xfe0b
+// GFX11: v_rndne_f16_e32 v127.l, 0xfe0b ; encoding: [0xff,0xbc,0xfe,0x7e,0x0b,0xfe,0x00,0x00]
+
+v_rndne_f16 v5.l, v1.h
+// GFX11: v_rndne_f16_e32 v5.l, v1.h ; encoding: [0x81,0xbd,0x0a,0x7e]
+
+v_rndne_f16 v5.l, v127.h
+// GFX11: v_rndne_f16_e32 v5.l, v127.h ; encoding: [0xff,0xbd,0x0a,0x7e]
+
+v_rndne_f16 v127.l, 0.5
+// GFX11: v_rndne_f16_e32 v127.l, 0.5 ; encoding: [0xf0,0xbc,0xfe,0x7e]
+
+v_rndne_f16 v5.h, src_scc
+// GFX11: v_rndne_f16_e32 v5.h, src_scc ; encoding: [0xfd,0xbc,0x0a,0x7f]
+
+v_rndne_f16 v127.h, 0xfe0b
+// GFX11: v_rndne_f16_e32 v127.h, 0xfe0b ; encoding: [0xff,0xbc,0xfe,0x7f,0x0b,0xfe,0x00,0x00]
v_rndne_f32 v5, v1
// GFX11: v_rndne_f32_e32 v5, v1 ; encoding: [0x01,0x47,0x0a,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..2879e1e79c6eca 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp16.s
@@ -2354,47 +2354,56 @@ v_rcp_iflag_f32 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
v_rcp_iflag_f32 v255, -|v255| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
// GFX11: v_rcp_iflag_f32_dpp v255, -|v255| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0x56,0xfe,0x7f,0xff,0x6f,0x35,0x30]
-v_rndne_f16 v5, v1 quad_perm:[3,2,1,0]
-// GFX11: v_rndne_f16_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x1b,0x00,0xff]
+v_rndne_f16 v5.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: v_rndne_f16_dpp v5.l, v1.l quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x1b,0x00,0xff]
-v_rndne_f16 v5, v1 quad_perm:[0,1,2,3]
-// GFX11: v_rndne_f16_dpp v5, v1 quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0xe4,0x00,0xff]
+v_rndne_f16 v5.l, v1.l quad_perm:[0,1,2,3]
+// GFX11: v_rndne_f16_dpp v5.l, v1.l quad_perm:[0,1,2,3] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0xe4,0x00,0xff]
-v_rndne_f16 v5, v1 row_mirror
-// GFX11: v_rndne_f16_dpp v5, v1 row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x40,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_mirror
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x40,0x01,0xff]
-v_rndne_f16 v5, v1 row_half_mirror
-// GFX11: v_rndne_f16_dpp v5, v1 row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x41,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_half_mirror
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_half_mirror row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x41,0x01,0xff]
-v_rndne_f16 v5, v1 row_shl:1
-// GFX11: v_rndne_f16_dpp v5, v1 row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x01,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_shl:1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_shl:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x01,0x01,0xff]
-v_rndne_f16 v5, v1 row_shl:15
-// GFX11: v_rndne_f16_dpp v5, v1 row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x0f,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_shl:15
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_shl:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x0f,0x01,0xff]
-v_rndne_f16 v5, v1 row_shr:1
-// GFX11: v_rndne_f16_dpp v5, v1 row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x11,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_shr:1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_shr:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x11,0x01,0xff]
-v_rndne_f16 v5, v1 row_shr:15
-// GFX11: v_rndne_f16_dpp v5, v1 row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x1f,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_shr:15
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_shr:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x1f,0x01,0xff]
-v_rndne_f16 v5, v1 row_ror:1
-// GFX11: v_rndne_f16_dpp v5, v1 row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x21,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_ror:1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_ror:1 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x21,0x01,0xff]
-v_rndne_f16 v5, v1 row_ror:15
-// GFX11: v_rndne_f16_dpp v5, v1 row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x2f,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_ror:15
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_ror:15 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x2f,0x01,0xff]
-v_rndne_f16 v5, v1 row_share:0 row_mask:0xf bank_mask:0xf
-// GFX11: v_rndne_f16_dpp v5, v1 row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x50,0x01,0xff]
+v_rndne_f16 v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_share:0 row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x50,0x01,0xff]
-v_rndne_f16 v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1
-// GFX11: v_rndne_f16_dpp v5, v1 row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x5f,0x01,0x01]
+v_rndne_f16 v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x5f,0x01,0x01]
-v_rndne_f16 v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
-// GFX11: v_rndne_f16_dpp v5, v1 row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x60,0x09,0x13]
+v_rndne_f16 v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xbc,0x0a,0x7e,0x01,0x60,0x09,0x13]
-v_rndne_f16 v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
-// GFX11: v_rndne_f16_dpp v127, -|v127| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xbc,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+v_rndne_f16 v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1
+// GFX11: v_rndne_f16_dpp v127.l, -|v127.l| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xbc,0xfe,0x7e,0x7f,0x6f,0x35,0x30]
+
+v_rndne_f16 v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1
+// GFX11: v_rndne_f16_dpp v127.l, v127.l row_share:15 row_mask:0x0 bank_mask:0x1 ; encoding: [0xfa,0xbc,0xfe,0x7e,0x7f,0x5f,0x01,0x01]
+
+v_rndne_f16 v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 fi:0
+// GFX11: v_rndne_f16_dpp v5.h, v1.h row_xmask:0 row_mask:0x1 bank_mask:0x3 bound_ctrl:1 ; encoding: [0xfa,0xbc,0x0a,0x7f,0x81,0x60,0x09,0x13]
+
+v_rndne_f16 v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 bound_ctrl:0 fi:1
+// GFX11: v_rndne_f16_dpp v127.h, -|v127.h| row_xmask:15 row_mask:0x3 bank_mask:0x0 fi:1 ; encoding: [0xfa,0xbc,0xfe,0x7f,0xff,0x6f,0x35,0x30]
v_rndne_f32 v5, v1 quad_perm:[3,2,1,0]
// GFX11: v_rndne_f32_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0xfa,0x46,0x0a,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..ae49ebb06a3209 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_dpp8.s
@@ -548,14 +548,23 @@ v_rcp_iflag_f32 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
v_rcp_iflag_f32 v255, v255 dpp8:[0,0,0,0,0,0,0,0] fi:0
// GFX11: v_rcp_iflag_f32_dpp v255, v255 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0x56,0xfe,0x7f,0xff,0x00,0x00,0x00]
-v_rndne_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
-// GFX11: v_rndne_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xbc,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_rndne_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_rndne_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xbc,0x0a,0x7e,0x01,0x77,0x39,0x05]
-v_rndne_f16 v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1
-// GFX11: v_rndne_f16_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xbc,0x0a,0x7e,0x01,0x77,0x39,0x05]
+v_rndne_f16 v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_rndne_f16_dpp v5.l, v1.l dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xbc,0x0a,0x7e,0x01,0x77,0x39,0x05]
-v_rndne_f16 v127, v127 dpp8:[0,0,0,0,0,0,0,0] fi:0
-// GFX11: v_rndne_f16_dpp v127, v127 dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xbc,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+v_rndne_f16 v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0]
+// GFX11: v_rndne_f16_dpp v127.l, v127.l dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xbc,0xfe,0x7e,0x7f,0x00,0x00,0x00]
+
+v_rndne_f16 v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: v_rndne_f16_dpp v127.l, v127.l dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0xbc,0xfe,0x7e,0x7f,0x77,0x39,0x05]
+
+v_rndne_f16 v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1
+// GFX11: v_rndne_f16_dpp v5.h, v1.h dpp8:[7,6,5,4,3,2,1,0] fi:1 ; encoding: [0xea,0xbc,0x0a,0x7f,0x81,0x77,0x39,0x05]
+
+v_rndne_f16 v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] fi:0
+// GFX11: v_rndne_f16_dpp v127.h, v127.h dpp8:[0,0,0,0,0,0,0,0] ; encoding: [0xe9,0xbc,0xfe,0x7f,0xff,0x00,0x00,0x00]
v_rndne_f32 v5, v1 dpp8:[7,6,5,4,3,2,1,0]
// GFX11: v_rndne_f32_dpp v5, v1 dpp8:[7,6,5,4,3,2,1,0] ; encoding: [0xe9,0x46,0x0a,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..8a36768f5c811d 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_err.s
@@ -647,6 +647,12 @@ v_rcp_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
v_rndne_f16_e32 v128, 0xfe0b
// GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
+v_rndne_f16_e32 v128.h, 0xfe0b
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v128.l, 0xfe0b
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
v_rndne_f16_e32 v255, v1
// GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
@@ -656,6 +662,24 @@ v_rndne_f16_e32 v255, v1 dpp8:[7,6,5,4,3,2,1,0]
v_rndne_f16_e32 v255, v1 quad_perm:[3,2,1,0]
// GFX11: :[[@LINE-1]]:26: error: invalid operand for instruction
+v_rndne_f16_e32 v255.h, v1.h
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v255.h, v1.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v255.h, v1.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v255.l, v1.l
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v255.l, v1.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
+v_rndne_f16_e32 v255.l, v1.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:17: error: invalid operand for instruction
+
v_rndne_f16_e32 v5, v199
// GFX11: :[[@LINE-1]]:1: error: operands are not valid for this GPU or mode
@@ -665,6 +689,24 @@ v_rndne_f16_e32 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
v_rndne_f16_e32 v5, v199 quad_perm:[3,2,1,0]
// GFX11: :[[@LINE-1]]:26: error: invalid operand for instruction
+v_rndne_f16_e32 v5.h, v199.h
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
+v_rndne_f16_e32 v5.h, v199.h dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
+v_rndne_f16_e32 v5.h, v199.h quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
+v_rndne_f16_e32 v5.l, v199.l
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
+v_rndne_f16_e32 v5.l, v199.l dpp8:[7,6,5,4,3,2,1,0]
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
+v_rndne_f16_e32 v5.l, v199.l quad_perm:[3,2,1,0]
+// GFX11: :[[@LINE-1]]:23: error: invalid operand for instruction
+
v_rsq_f16_e32 v128.h, 0xfe0b
// GFX11: :[[@LINE-1]]:15: error: invalid operand for instruction
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..9175dd07bf6211 100644
--- a/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
+++ b/llvm/test/MC/AMDGPU/gfx11_asm_vop1_t16_promote.s
@@ -1670,71 +1670,137 @@ v_rcp_f16 v5, v199 dpp8:[7,6,5,4,3,2,1,0]
v_rcp_f16 v5, v199 quad_perm:[3,2,1,0]
// GFX11: v_rcp_f16_e64_dpp v5, v199 quad_perm:[3,2,1,0] row_mask:0xf bank_mask:0xf ; encoding: [0x05,0x00,0xd4,0xd5,0xfa,0x00,0x00,0x00,0xc7,0x1b,0x00,0xff]
-v_rndne_f16 v128, 0xfe0b
-// GFX11: v_rndne_f16_e64 v128, 0xfe0b ; encoding: [0x80,0x00,0xde,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
+v_rndne_f16 v128.h, 0xfe0b
+// GFX11: v_rndne_f16_e64 v128.h, 0xfe0b op_sel:[0,1] ; encoding: [0x80,0x40,0xde,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
-v_rndne_f16 v255, -1
-// GFX11: v_rndne_f16_e64 v255, -1 ; encoding: [0xff,0x00,0xde,0xd5,0xc1,0x00,0x00,0x00]
+v_rndne_f16 v128.l, 0xfe0b
+// GFX11: v_rndne_f16_e64 v128.l, 0xfe0b ; encoding: [0x80,0x00,0xde,0xd5,0xff,0x00,0x00,0x00,0x0b,0xfe,0x00,0x00]
-v_rndne_f16 v255, 0.5
-// GFX11: v_rndne_f16_e64 v255, 0.5 ; encoding: [0xff,0x00,0xde,0xd5,0xf0,0x00,0x00,0x00]
+v_rndne_f16 v255.h, -1
+// GFX11: v_rndne_f16_e64 v255.h, -1 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0xc1,0x00,0x00,0x00]
-v_rndne_f16 v255, exec_hi
-// GFX11: v_rndne_f16_e64 v255, exec_hi ; encoding: [0xff,0x00,0xde,0xd5,0x7f,0x00,0x00,0x00]
+v_rndne_f16 v255.h, 0.5
+// GFX11: v_rndne_f16_e64 v255.h, 0.5 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0xf0,0x00,0x00,0x00]
-v_rndne_f16 v255, exec_lo
-// GFX11: v_rndne_f16_e64 v255, exec_lo ; encoding: [0xff,0x00,0xde,0xd5,0x7e,0x00,0x00,0x00]
+v_rndne_f16 v255.h, exec_hi
+// GFX11: v_rndne_f16_e64 v255.h, exec_hi op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x7f,0x00,0x00,0x00]
-v_rndne_f16 v255, m0
-// GFX11: v_rndne_f16_e64 v255, m0 ; encoding: [0xff,0x00,0xde,0xd5,0x7d,0x00,0x00,0x00]
+v_rndne_f16 v255.h, exec_lo
+// GFX11: v_rndne_f16_e64 v255.h, exec_lo op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x7e,0x00,0x00,0x00]
-v_rndne_f16 v255, null
-// GFX11: v_rndne_f16_e64 v255, null ; encoding: [0xff,0x00,0xde,0xd5,0x7c,0x00,0x00,0x00]
+v_rndne_f16 v255.h, m0
+// GFX11: v_rndne_f16_e64 v255.h, m0 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x7d,0x00,0x00,0x00]
-v_rndne_f16 v255, s1
-// GFX11: v_rndne_f16_e64 v255, s1 ; encoding: [0xff,0x00,0xde,0xd5,0x01,0x00,0x00,0x00]
+v_rndne_f16 v255.h, null
+// GFX11: v_rndne_f16_e64 v255.h, null op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x7c,0x00,0x00,0x00]
-v_rndne_f16 v255, s105
-// GFX11: v_rndne_f16_e64 v255, s105 ; encoding: [0xff,0x00,0xde,0xd5,0x69,0x00,0x00,0x00]
+v_rndne_f16 v255.h, s1
+// GFX11: v_rndne_f16_e64 v255.h, s1 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x01,0x00,0x00,0x00]
-v_rndne_f16 v255, src_scc
-// GFX11: v_rndne_f16_e64 v255, src_scc ; encoding: [0xff,0x00,0xde,0xd5,0xfd,0x00,0x00,0x00]
+v_rndne_f16 v255.h, s105
+// GFX11: v_rndne_f16_e64 v255.h, s105 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x69,0x00,0x00,0x00]
-v_rndne_f16 v255, ttmp15
-// GFX11: v_rndne_f16_e64 v255, ttmp15 ; encoding: [0xff,0x00,0xde,0xd5,0x7b,0x00,0x00,0x00]
+v_rndne_f16 v255.h, src_scc
+// GFX11: v_rndne_f16_e64 v255.h, src_scc op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0xfd,0x00,0x00,0x00]
-v_rndne_f16 v255, v1
-// GFX11: v_rndne_f16_e64 v255, v1 ; encoding: [0xff,0x00,0xde,0xd5,0x01,0x01,0x00,0x00]
+v_rndne_f16 v255.h, ttmp15
+// GFX11: v_rndne_f16_e64 v255.h, ttmp15 op_sel:[0,1] ; encoding: [0xff,0x40,0xde,0xd5,0x7b,0x00,0x00,0x00]
-v_rndne_f16 v2...
[truncated]
|
arsenm
approved these changes
Dec 23, 2024
broxigarchen
force-pushed
the
main-merge-true16-mc-vop1-more-instruction-8
branch
from
c1d1d18 to
480eb39
Compare
|
Added a GFX12 runline for codegen test. resolved conflict |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Support true16 format for v_rndne_b16 in MC