AMDGPU/GlobalISel: Fix clamping shifts with 16-bit insts

llvm-svn: 352599

diff --git a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
index 434fcc5..0da4607 100644 (file)
--- a/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
@@ -320,9 +320,10 @@ AMDGPULegalizerInfo::AMDGPULegalizerInfo(const GCNSubtarget &ST,
   // be more flexible with the shift amount type.
   auto &Shifts = getActionDefinitionsBuilder({G_SHL, G_LSHR, G_ASHR})
     .legalFor({{S32, S32}, {S64, S32}});
-  if (ST.has16BitInsts())
+  if (ST.has16BitInsts()) {
     Shifts.legalFor({{S16, S32}, {S16, S16}});
-  else
+    Shifts.clampScalar(0, S16, S64);
+  } else
     Shifts.clampScalar(0, S32, S64);
   Shifts.clampScalar(1, S32, S32);
 

diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-ashr.mir b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-ashr.mir
index 814891d..eb1cc74 100644 (file)
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-ashr.mir
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-ashr.mir
@@ -250,3 +250,59 @@ body: |
     %5:_(s32) = G_ANYEXT %4
     $vgpr0 = COPY %5
 ...
+
+---
+name: test_ashr_i8_i8
+body: |
+  bb.0:
+    liveins: $vgpr0, $vgpr1
+
+    ; SI-LABEL: name: test_ashr_i8_i8
+    ; SI: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; SI: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; SI: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 24
+    ; SI: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY]](s32)
+    ; SI: [[SHL:%[0-9]+]]:_(s32) = G_SHL [[COPY2]], [[C]](s32)
+    ; SI: [[ASHR:%[0-9]+]]:_(s32) = G_ASHR [[SHL]], [[C]](s32)
+    ; SI: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; SI: [[COPY3:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; SI: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY3]], [[C1]]
+    ; SI: [[ASHR1:%[0-9]+]]:_(s32) = G_ASHR [[ASHR]], [[AND]](s32)
+    ; SI: [[COPY4:%[0-9]+]]:_(s32) = COPY [[ASHR1]](s32)
+    ; SI: $vgpr0 = COPY [[COPY4]](s32)
+    ; VI-LABEL: name: test_ashr_i8_i8
+    ; VI: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; VI: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; VI: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+    ; VI: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[C]](s32)
+    ; VI: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[COPY]](s32)
+    ; VI: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[TRUNC1]], [[TRUNC]](s16)
+    ; VI: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[SHL]], [[TRUNC]](s16)
+    ; VI: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; VI: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; VI: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[C1]]
+    ; VI: [[ASHR1:%[0-9]+]]:_(s16) = G_ASHR [[ASHR]], [[AND]](s32)
+    ; VI: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR1]](s16)
+    ; VI: $vgpr0 = COPY [[ANYEXT]](s32)
+    ; GFX9-LABEL: name: test_ashr_i8_i8
+    ; GFX9: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; GFX9: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; GFX9: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 8
+    ; GFX9: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[C]](s32)
+    ; GFX9: [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[COPY]](s32)
+    ; GFX9: [[SHL:%[0-9]+]]:_(s16) = G_SHL [[TRUNC1]], [[TRUNC]](s16)
+    ; GFX9: [[ASHR:%[0-9]+]]:_(s16) = G_ASHR [[SHL]], [[TRUNC]](s16)
+    ; GFX9: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; GFX9: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; GFX9: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[C1]]
+    ; GFX9: [[ASHR1:%[0-9]+]]:_(s16) = G_ASHR [[ASHR]], [[AND]](s32)
+    ; GFX9: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[ASHR1]](s16)
+    ; GFX9: $vgpr0 = COPY [[ANYEXT]](s32)
+    %0:_(s32) = COPY $vgpr0
+    %1:_(s32) = COPY $vgpr1
+    %2:_(s8) = G_TRUNC %0
+    %3:_(s8) = G_TRUNC %1
+    %4:_(s8) = G_ASHR %2, %3
+    %5:_(s32) = G_ANYEXT %4
+    $vgpr0 = COPY %5
+...

diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-lshr.mir b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-lshr.mir
index d0999fc..15d5084 100644 (file)
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-lshr.mir
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-lshr.mir
@@ -247,3 +247,58 @@ body: |
     %5:_(s32) = G_ANYEXT %4
     $vgpr0 = COPY %5
 ...
+
+---
+name: test_lshr_i8_i8
+body: |
+  bb.0:
+    liveins: $vgpr0, $vgpr1
+
+    ; SI-LABEL: name: test_lshr_i8_i8
+    ; SI: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; SI: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; SI: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; SI: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY]](s32)
+    ; SI: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[C]]
+    ; SI: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; SI: [[COPY3:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; SI: [[AND1:%[0-9]+]]:_(s32) = G_AND [[COPY3]], [[C1]]
+    ; SI: [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[AND]], [[AND1]](s32)
+    ; SI: [[COPY4:%[0-9]+]]:_(s32) = COPY [[LSHR]](s32)
+    ; SI: $vgpr0 = COPY [[COPY4]](s32)
+    ; VI-LABEL: name: test_lshr_i8_i8
+    ; VI: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; VI: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; VI: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; VI: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY]](s32)
+    ; VI: [[COPY3:%[0-9]+]]:_(s32) = COPY [[C]](s32)
+    ; VI: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[COPY3]]
+    ; VI: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[AND]](s32)
+    ; VI: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; VI: [[COPY4:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; VI: [[AND1:%[0-9]+]]:_(s32) = G_AND [[COPY4]], [[C1]]
+    ; VI: [[LSHR:%[0-9]+]]:_(s16) = G_LSHR [[TRUNC]], [[AND1]](s32)
+    ; VI: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[LSHR]](s16)
+    ; VI: $vgpr0 = COPY [[ANYEXT]](s32)
+    ; GFX9-LABEL: name: test_lshr_i8_i8
+    ; GFX9: [[COPY:%[0-9]+]]:_(s32) = COPY $vgpr0
+    ; GFX9: [[COPY1:%[0-9]+]]:_(s32) = COPY $vgpr1
+    ; GFX9: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; GFX9: [[COPY2:%[0-9]+]]:_(s32) = COPY [[COPY]](s32)
+    ; GFX9: [[COPY3:%[0-9]+]]:_(s32) = COPY [[C]](s32)
+    ; GFX9: [[AND:%[0-9]+]]:_(s32) = G_AND [[COPY2]], [[COPY3]]
+    ; GFX9: [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[AND]](s32)
+    ; GFX9: [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 255
+    ; GFX9: [[COPY4:%[0-9]+]]:_(s32) = COPY [[COPY1]](s32)
+    ; GFX9: [[AND1:%[0-9]+]]:_(s32) = G_AND [[COPY4]], [[C1]]
+    ; GFX9: [[LSHR:%[0-9]+]]:_(s16) = G_LSHR [[TRUNC]], [[AND1]](s32)
+    ; GFX9: [[ANYEXT:%[0-9]+]]:_(s32) = G_ANYEXT [[LSHR]](s16)
+    ; GFX9: $vgpr0 = COPY [[ANYEXT]](s32)
+    %0:_(s32) = COPY $vgpr0
+    %1:_(s32) = COPY $vgpr1
+    %2:_(s8) = G_TRUNC %0
+    %3:_(s8) = G_TRUNC %1
+    %4:_(s8) = G_LSHR %2, %3
+    %5:_(s32) = G_ANYEXT %4
+    $vgpr0 = COPY %5
+...

diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-shl.mir b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-shl.mir
index 63b0476..fb94ab8 100644 (file)
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-shl.mir
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-shl.mir
@@ -265,3 +265,18 @@ body: |
     %5:_(s32) = G_ANYEXT %4
     $vgpr0 = COPY %5
 ...
+
+---
+name: test_shl_i8_i8
+body: |
+  bb.0:
+    liveins: $vgpr0, $vgpr1
+
+    %0:_(s32) = COPY $vgpr0
+    %1:_(s32) = COPY $vgpr1
+    %2:_(s8) = G_TRUNC %0
+    %3:_(s8) = G_TRUNC %1
+    %4:_(s8) = G_SHL %2, %3
+    %5:_(s32) = G_ANYEXT %4
+    $vgpr0 = COPY %5
+...