Skip to content

[AMDGPU] MC support for load monitor instructions on gfx1250 #150496

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
changpeng merged 3 commits into from
Jul 24, 2025
Merged

[AMDGPU] MC support for load monitor instructions on gfx1250 #150496

changpeng merged 3 commits into from
Jul 24, 2025

Conversation

@changpeng
Copy link
Contributor

@changpeng changpeng commented Jul 24, 2025

No description provided.

@llvmbot llvmbot added backend:AMDGPU llvm:mc Machine (object) code labels Jul 24, 2025
@llvmbot
Copy link
Member

llvmbot commented Jul 24, 2025 •
edited
Loading

@llvm/pr-subscribers-backend-amdgpu

@llvm/pr-subscribers-mc

Author: Changpeng Fang (changpeng)

Changes

Full diff: https://github.com/llvm/llvm-project/pull/150496.diff

3 Files Affected:

  • (modified) llvm/lib/Target/AMDGPU/FLATInstructions.td (+19)
  • (modified) llvm/test/MC/AMDGPU/gfx1250_asm_vflat.s (+88)
  • (modified) llvm/test/MC/Disassembler/AMDGPU/gfx1250_dasm_vflat.txt (+93)
diff --git a/llvm/lib/Target/AMDGPU/FLATInstructions.td b/llvm/lib/Target/AMDGPU/FLATInstructions.td
index db827f4fd7c46..8b1fc9f7383ef 100644
--- a/llvm/lib/Target/AMDGPU/FLATInstructions.td
+++ b/llvm/lib/Target/AMDGPU/FLATInstructions.td
@@ -1193,6 +1193,16 @@ defm SCRATCH_LOAD_LDS_USHORT : FLAT_Scratch_Load_LDS_Pseudo <"scratch_load_lds_u
 defm SCRATCH_LOAD_LDS_SSHORT : FLAT_Scratch_Load_LDS_Pseudo <"scratch_load_lds_sshort">;
 defm SCRATCH_LOAD_LDS_DWORD  : FLAT_Scratch_Load_LDS_Pseudo <"scratch_load_lds_dword">;
 
+let SubtargetPredicate = isGFX125xOnly in {
+defm FLAT_LOAD_MONITOR_B32    : FLAT_Flat_Load_Pseudo <"flat_load_monitor_b32",  VGPR_32>;
+defm FLAT_LOAD_MONITOR_B64    : FLAT_Flat_Load_Pseudo <"flat_load_monitor_b64",  VReg_64>;
+defm FLAT_LOAD_MONITOR_B128   : FLAT_Flat_Load_Pseudo <"flat_load_monitor_b128", VReg_128>;
+
+defm GLOBAL_LOAD_MONITOR_B32  : FLAT_Global_Load_Pseudo <"global_load_monitor_b32",  VGPR_32>;
+defm GLOBAL_LOAD_MONITOR_B64  : FLAT_Global_Load_Pseudo <"global_load_monitor_b64",  VReg_64>;
+defm GLOBAL_LOAD_MONITOR_B128 : FLAT_Global_Load_Pseudo <"global_load_monitor_b128", VReg_128>;
+} // End SubtargetPredicate = isGFX125xOnly
+
 let SubtargetPredicate = isGFX12Plus in {
   let Uses = [EXEC, M0] in {
     defm GLOBAL_LOAD_BLOCK  : FLAT_Global_Load_Pseudo <"global_load_block", VReg_1024>;
@@ -2045,6 +2055,7 @@ let WaveSizePredicate = isWave32,  OtherPredicates = [HasTransposeLoadF4F6Insts]
   defm : GlobalFLATLoadPats <GLOBAL_LOAD_TR6_B96, int_amdgcn_global_load_tr6_b96, v3i32>;
 }
 
+
 let SubtargetPredicate = HasAtomicFMinFMaxF32GlobalInsts, OtherPredicates = [HasFlatGlobalInsts] in {
 defm : GlobalFLATAtomicPats <"GLOBAL_ATOMIC_FMIN", "atomic_load_fmin_global", f32>;
 defm : GlobalFLATAtomicPats <"GLOBAL_ATOMIC_FMAX", "atomic_load_fmax_global", f32>;
@@ -3249,6 +3260,14 @@ defm TENSOR_STOP                      : VFLAT_Real_gfx1250<0x06f>;
 defm FLAT_PREFETCH_B8                 : VFLAT_Real_AllAddr_gfx1250<0x05d>;
 defm GLOBAL_PREFETCH_B8               : VFLAT_Real_AllAddr_gfx1250<0x05d>;
 
+defm FLAT_LOAD_MONITOR_B32            : VFLAT_Real_AllAddr_gfx1250<0x070>;
+defm FLAT_LOAD_MONITOR_B64            : VFLAT_Real_AllAddr_gfx1250<0x071>;
+defm FLAT_LOAD_MONITOR_B128           : VFLAT_Real_AllAddr_gfx1250<0x072>;
+
+defm GLOBAL_LOAD_MONITOR_B32          : VFLAT_Real_AllAddr_gfx1250<0x070>;
+defm GLOBAL_LOAD_MONITOR_B64          : VFLAT_Real_AllAddr_gfx1250<0x071>;
+defm GLOBAL_LOAD_MONITOR_B128         : VFLAT_Real_AllAddr_gfx1250<0x072>;
+
 defm GLOBAL_LOAD_TR_B128_w32          : VFLAT_Real_AllAddr_gfx1250<0x057, "global_load_tr16_b128">;
 defm GLOBAL_LOAD_TR_B64_w32           : VFLAT_Real_AllAddr_gfx1250<0x058, "global_load_tr8_b64">;
 
diff --git a/llvm/test/MC/AMDGPU/gfx1250_asm_vflat.s b/llvm/test/MC/AMDGPU/gfx1250_asm_vflat.s
index f073e9034208c..b9eb2d2dc8c70 100644
--- a/llvm/test/MC/AMDGPU/gfx1250_asm_vflat.s
+++ b/llvm/test/MC/AMDGPU/gfx1250_asm_vflat.s
@@ -161,6 +161,94 @@ global_prefetch_b8 v[2:3], off offset:64 th:TH_LOAD_NT_RT scope:SCOPE_DEV
 // GFX1250: global_prefetch_b8 v[2:3], off offset:64 th:TH_LOAD_NT_RT scope:SCOPE_DEV ; encoding: [0x7c,0x40,0x17,0xee,0x00,0x00,0x48,0x00,0x02,0x40,0x00,0x00]
 // GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
 
+global_load_monitor_b32 v1, v[2:3], off
+// GFX1250: global_load_monitor_b32 v1, v[2:3], off ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v[2:3], off offset:64
+// GFX1250: global_load_monitor_b32 v1, v[2:3], off offset:64 ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v[2:3], off offset:-64 th:TH_LOAD_NT_HT scope:SCOPE_DEV
+// GFX1250: global_load_monitor_b32 v1, v[2:3], off offset:-64 th:TH_LOAD_NT_HT scope:SCOPE_DEV ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x68,0x00,0x02,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v2, s[0:1]
+// GFX1250: global_load_monitor_b32 v1, v2, s[0:1] ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v2, s[0:1] offset:64
+// GFX1250: global_load_monitor_b32 v1, v2, s[0:1] offset:64 ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS
+// GFX1250: global_load_monitor_b32 v1, v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v[2:3], off
+// GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v[2:3], off offset:64
+// GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off offset:64 ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v[2:3], off offset:-64 th:TH_LOAD_HT scope:SCOPE_SE
+// GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off offset:-64 th:TH_LOAD_HT scope:SCOPE_SE ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x24,0x00,0x02,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v2, s[0:1]
+// GFX1250: encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v2, s[0:1] offset:64
+// GFX1250: global_load_monitor_b64 v[0:1], v2, s[0:1] offset:64 ; encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[0:1], v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS
+// GFX1250: global_load_monitor_b64 v[0:1], v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v[4:5], off
+// GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v[4:5], off offset:64
+// GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off offset:64 ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v[4:5], off offset:-64 th:TH_LOAD_NT
+// GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off offset:-64 th:TH_LOAD_NT ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x10,0x00,0x04,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v4, s[0:1]
+// GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] ; encoding:  [0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v4, s[0:1] offset:64
+// GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] offset:64 ; encoding: [0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b128 v[0:3], v4, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS
+// GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x80,0x1c,0xee,0x00,0x00,0x3c,0x00,0x04,0xc0,0xff,0xff]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b32 v1, v2, s[4:5] offset:64 scale_offset
+// GFX1250: global_load_monitor_b32 v1, v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x00,0x1c,0xee,0x01,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+global_load_monitor_b64 v[2:3], v2, s[4:5] offset:64 scale_offset
+// GFX1250: global_load_monitor_b64 v[2:3], v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x40,0x1c,0xee,0x02,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+flat_load_monitor_b32 v1, v[2:3]
+// GFX1250: flat_load_monitor_b32 v1, v[2:3] ; encoding: [0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
+flat_load_monitor_b32 v1, v[2:3] offset:64
+// GFX1250: flat_load_monitor_b32 v1, v[2:3] offset:64 ; encoding: [0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+// GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
+
 tensor_save s[0:1]
 // GFX1250: tensor_save s[0:1] ; encoding: [0x00,0x80,0x1b,0xee,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00]
 // GFX12-ERR: :[[@LINE-2]]:{{[0-9]+}}: error: instruction not supported on this GPU
diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx1250_dasm_vflat.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx1250_dasm_vflat.txt
index 2f74c69a7ed03..de7895fbef0cb 100644
--- a/llvm/test/MC/Disassembler/AMDGPU/gfx1250_dasm_vflat.txt
+++ b/llvm/test/MC/Disassembler/AMDGPU/gfx1250_dasm_vflat.txt
@@ -3084,6 +3084,99 @@
 # GFX1250: global_prefetch_b8 v[2:3], off th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x7c,0x40,0x17,0xee,0x00,0x00,0x3c,0x00,0x02,0x00,0x00,0x00]
 0x7c,0x40,0x17,0xee,0x00,0x00,0x3c,0x00,0x02,0x00,0x00,0x00
 
+# GFX1250: flat_load_monitor_b128 v[0:3], v[4:5]   ; encoding: [0x7c,0x80,0x1c,0xec,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00]
+0x7c,0x80,0x1c,0xec,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00
+
+# GFX1250: flat_load_monitor_b128 v[0:3], v[4:5] offset:64 ; encoding: [0x7c,0x80,0x1c,0xec,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00]
+0x7c,0x80,0x1c,0xec,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00
+
+# GFX1250: flat_load_monitor_b128 v[0:3], v[4:5] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x7c,0x80,0x1c,0xec,0x00,0x00,0x3c,0x00,0x04,0xc0,0xff,0xff]
+0x7c,0x80,0x1c,0xec,0x00,0x00,0x3c,0x00,0x04,0xc0,0xff,0xff
+
+# GFX1250: flat_load_monitor_b32 v1, v[2:3]        ; encoding: [0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: flat_load_monitor_b32 v1, v[2:3] offset:64 ; encoding: [0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x7c,0x00,0x1c,0xec,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: flat_load_monitor_b32 v1, v[2:3] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x7c,0x00,0x1c,0xec,0x01,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+0x7c,0x00,0x1c,0xec,0x01,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: flat_load_monitor_b64 v[0:1], v[2:3]    ; encoding: [0x7c,0x40,0x1c,0xec,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x7c,0x40,0x1c,0xec,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: flat_load_monitor_b64 v[0:1], v[2:3] offset:64 ; encoding: [0x7c,0x40,0x1c,0xec,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x7c,0x40,0x1c,0xec,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: flat_load_monitor_b64 v[0:1], v[2:3] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x7c,0x40,0x1c,0xec,0x00,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+0x7c,0x40,0x1c,0xec,0x00,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: flat_load_monitor_b32 v1, v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x00,0x1c,0xec,0x01,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+0x04,0x00,0x1c,0xec,0x01,0x00,0x01,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: flat_load_monitor_b64 v[2:3], v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x40,0x1c,0xec,0x02,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+0x04,0x40,0x1c,0xec,0x02,0x00,0x01,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00]
+0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off offset:64 ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00]
+0x7c,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b128 v[0:3], v[4:5], off offset:-64 th:TH_LOAD_NT ; encoding: [0x7c,0x80,0x1c,0xee,0x00,0x00,0x10,0x00,0x04,0xc0,0xff,0xff]
+0x7c,0x80,0x1c,0xee,0x00,0x00,0x10,0x00,0x04,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] ; encoding: [0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00]
+0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] offset:64 ; encoding: [0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00]
+0x00,0x80,0x1c,0xee,0x00,0x00,0x00,0x00,0x04,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b128 v[0:3], v4, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x80,0x1c,0xee,0x00,0x00,0x3c,0x00,0x04,0xc0,0xff,0xff]
+0x00,0x80,0x1c,0xee,0x00,0x00,0x3c,0x00,0x04,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b32 v1, v[2:3], off ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b32 v1, v[2:3], off offset:64 ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x7c,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b32 v1, v[2:3], off offset:-64 th:TH_LOAD_NT_HT scope:SCOPE_DEV ; encoding: [0x7c,0x00,0x1c,0xee,0x01,0x00,0x68,0x00,0x02,0xc0,0xff,0xff]
+0x7c,0x00,0x1c,0xee,0x01,0x00,0x68,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b32 v1, v2, s[0:1]  ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b32 v1, v2, s[0:1] offset:64 ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x00,0x00,0x1c,0xee,0x01,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b32 v1, v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x00,0x1c,0xee,0x01,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+0x00,0x00,0x1c,0xee,0x01,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off offset:64 ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x7c,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b64 v[0:1], v[2:3], off offset:-64 th:TH_LOAD_HT scope:SCOPE_SE ; encoding: [0x7c,0x40,0x1c,0xee,0x00,0x00,0x24,0x00,0x02,0xc0,0xff,0xff]
+0x7c,0x40,0x1c,0xee,0x00,0x00,0x24,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b64 v[0:1], v2, s[0:1] ; encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00]
+0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x00,0x00,0x00
+
+# GFX1250: global_load_monitor_b64 v[0:1], v2, s[0:1] offset:64 ; encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00]
+0x00,0x40,0x1c,0xee,0x00,0x00,0x00,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b64 v[0:1], v2, s[0:1] offset:-64 th:TH_LOAD_BYPASS scope:SCOPE_SYS ; encoding: [0x00,0x40,0x1c,0xee,0x00,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff]
+0x00,0x40,0x1c,0xee,0x00,0x00,0x3c,0x00,0x02,0xc0,0xff,0xff
+
+# GFX1250: global_load_monitor_b32 v1, v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x00,0x1c,0xee,0x01,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+0x04,0x00,0x1c,0xee,0x01,0x00,0x01,0x00,0x02,0x40,0x00,0x00
+
+# GFX1250: global_load_monitor_b64 v[2:3], v2, s[4:5] offset:64 scale_offset ; encoding: [0x04,0x40,0x1c,0xee,0x02,0x00,0x01,0x00,0x02,0x40,0x00,0x00]
+0x04,0x40,0x1c,0xee,0x02,0x00,0x01,0x00,0x02,0x40,0x00,0x00
+
 # GFX1250: tensor_save s[0:1] ; encoding: [0x00,0x80,0x1b,0xee,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00]
 0x00,0x80,0x1b,0xee,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00

@changpeng changpeng requested review from rampitec and shiltian July 24, 2025 18:36
rampitec
rampitec approved these changes Jul 24, 2025
View reviewed changes
Copy link
Collaborator

@rampitec rampitec left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM with a nit.

llvm/lib/Target/AMDGPU/FLATInstructions.td Outdated
defm : GlobalFLATLoadPats <GLOBAL_LOAD_TR6_B96, int_amdgcn_global_load_tr6_b96, v3i32>;
}


Copy link
Collaborator

@rampitec rampitec Jul 24, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Unrelated whitespace change.

@changpeng changpeng merged commit 34b6587 into llvm:main Jul 24, 2025
9 checks passed
@changpeng changpeng deleted the monitor branch July 24, 2025 19:27
@llvm-ci
Copy link
Collaborator

llvm-ci commented Jul 25, 2025

LLVM Buildbot has detected a new failure on builder llvm-clang-x86_64-expensive-checks-debian running on gribozavr4 while building llvm at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/16/builds/23200

Here is the relevant piece of the build log for the reference 
Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LLVM :: TableGen/RuntimeLibcallEmitter.td' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
/b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llvm-tblgen -gen-runtime-libcalls -I /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/../../include /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td | /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td # RUN: at line 1
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llvm-tblgen -gen-runtime-libcalls -I /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/../../include /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td
/b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td:98:16: error: CHECK-NEXT: expected string not found in input
// CHECK-NEXT: sqrtl_f80 = 7, // sqrtl
               ^
<stdin>:32:23: note: scanning from here
 calloc = 6, // calloc
                      ^
<stdin>:34:2: note: possible intended match here
 sqrtl_f80 = 8, // sqrtl
 ^

Input file: <stdin>
Check file: /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td

-dump-input=help explains the following input dump.

Input was:
<<<<<<
           .
           .
           .
          27:  ___memcpy = 1, // ___memcpy 
          28:  ___memset = 2, // ___memset 
          29:  __ashlsi3 = 3, // __ashlsi3 
          30:  __lshrdi3 = 4, // __lshrdi3 
          31:  bzero = 5, // bzero 
          32:  calloc = 6, // calloc 
next:98'0                           X error: no match found
          33:  sqrtl_f128 = 7, // sqrtl 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~
          34:  sqrtl_f80 = 8, // sqrtl 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
next:98'1      ?                        possible intended match
          35:  NumLibcallImpls = 9 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~
          36: }; 
next:98'0     ~~~
          37: } // End namespace RTLIB 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
          38: } // End namespace llvm 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~
          39: #endif 
next:98'0     ~~~~~~~
...

mahesh-attarde pushed a commit to mahesh-attarde/llvm-project that referenced this pull request Jul 28, 2025
@changpeng @mahesh-attarde
[AMDGPU] MC support for load monitor instructions on gfx1250 (llvm#15…
2cb6eb2 
…0496)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@rampitec rampitec rampitec approved these changes

@shiltian shiltian Awaiting requested review from shiltian

Assignees

No one assigned

Labels

backend:AMDGPU llvm:mc Machine (object) code

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

4 participants

@changpeng @llvmbot @llvm-ci @rampitec