[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits #142355

usha1830 · 2025-06-02T10:17:23Z

Code ported from SelectionDAG::ComputeNumSignBits

llvmbot · 2025-06-02T10:18:01Z

@llvm/pr-subscribers-backend-aarch64

@llvm/pr-subscribers-llvm-globalisel

Author: Usha Gupta (usha1830)

Changes

Code ported from SelectionDAG::ComputeNumSignBits

Full diff: https://github.com/llvm/llvm-project/pull/142355.diff

7 Files Affected:

(modified) llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp (+21)
(modified) llvm/test/CodeGen/AArch64/GlobalISel/knownbits-concat.mir (+2-2)
(modified) llvm/test/CodeGen/AArch64/GlobalISel/legalize-min-max.mir (+16-24)
(modified) llvm/test/CodeGen/AArch64/aarch64-dup-ext.ll (-2)
(modified) llvm/test/CodeGen/AArch64/fcmp.ll (+7-11)
(modified) llvm/test/CodeGen/AArch64/scmp.ll (-4)
(modified) llvm/test/CodeGen/AArch64/ucmp.ll (-4)

diff --git a/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp b/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp
index 692befde71cb1..50c2f452ee18e 100644
--- a/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp
@@ -1958,6 +1958,27 @@ unsigned GISelValueTracking::computeNumSignBits(Register R,
     }
     break;
   }
+  case TargetOpcode::G_CONCAT_VECTORS: {
+    if (MRI.getType(MI.getOperand(0).getReg()).isScalableVector())
+      break;
+    FirstAnswer = std::numeric_limits<unsigned>::max();
+    // Determine the minimum number of sign bits across all demanded
+    // elts of the input vectors. Early out if the result is already 1.
+    unsigned NumSubVectorElts =
+        MRI.getType(MI.getOperand(1).getReg()).getNumElements();
+    unsigned NumSubVectors = MI.getNumOperands() - 1;
+    for (unsigned i = 0; (i < NumSubVectors); ++i) {
+      APInt DemandedSub =
+          DemandedElts.extractBits(NumSubVectorElts, i * NumSubVectorElts);
+      if (!DemandedSub)
+        continue;
+      unsigned Tmp2 = computeNumSignBits(MI.getOperand(i + 1).getReg(),
+                                         DemandedSub, Depth + 1);
+
+      FirstAnswer = std::min(FirstAnswer, Tmp2);
+    }
+    break;
+  }
   case TargetOpcode::G_SHUFFLE_VECTOR: {
     // Collect the minimum number of sign bits that are shared by every vector
     // element referenced by the shuffle.
diff --git a/llvm/test/CodeGen/AArch64/GlobalISel/knownbits-concat.mir b/llvm/test/CodeGen/AArch64/GlobalISel/knownbits-concat.mir
index 85129e9639a6a..03f7533c7d604 100644
--- a/llvm/test/CodeGen/AArch64/GlobalISel/knownbits-concat.mir
+++ b/llvm/test/CodeGen/AArch64/GlobalISel/knownbits-concat.mir
@@ -42,7 +42,7 @@ body: |
   ; CHECK-NEXT: %1:_ KnownBits:???????? SignBits:1
   ; CHECK-NEXT: %sext0:_ KnownBits:???????????????? SignBits:9
   ; CHECK-NEXT: %sext1:_ KnownBits:???????????????? SignBits:9
-  ; CHECK-NEXT: %res:_ KnownBits:???????????????? SignBits:1
+  ; CHECK-NEXT: %res:_ KnownBits:???????????????? SignBits:9
     %0:_(<2 x s8>) = COPY $h0
     %1:_(<2 x s8>) = COPY $h1
     %sext0:_(<2 x s16>) = G_SEXT %0
@@ -60,7 +60,7 @@ body: |
   ; CHECK-NEXT: %1:_ KnownBits:???????? SignBits:1
   ; CHECK-NEXT: %zext0:_ KnownBits:00000000???????? SignBits:8
   ; CHECK-NEXT: %sext1:_ KnownBits:???????????????? SignBits:9
-  ; CHECK-NEXT: %res:_ KnownBits:???????????????? SignBits:1
+  ; CHECK-NEXT: %res:_ KnownBits:???????????????? SignBits:8
     %0:_(<2 x s8>) = COPY $h0
     %1:_(<2 x s8>) = COPY $h1
     %zext0:_(<2 x s16>) = G_ZEXT %0
diff --git a/llvm/test/CodeGen/AArch64/GlobalISel/legalize-min-max.mir b/llvm/test/CodeGen/AArch64/GlobalISel/legalize-min-max.mir
index 4a3457a35256d..fae979d38c6ea 100644
--- a/llvm/test/CodeGen/AArch64/GlobalISel/legalize-min-max.mir
+++ b/llvm/test/CodeGen/AArch64/GlobalISel/legalize-min-max.mir
@@ -249,14 +249,12 @@ body: |
     ; CHECK-NEXT: [[DEF:%[0-9]+]]:_(<2 x s64>) = G_IMPLICIT_DEF
     ; CHECK-NEXT: [[ICMP:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(slt), [[DEF]](<2 x s64>), [[DEF]]
     ; CHECK-NEXT: [[ICMP1:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(slt), [[DEF]](<2 x s64>), [[DEF]]
-    ; CHECK-NEXT: [[SEXT_INREG:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP]], 1
-    ; CHECK-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP1]], 1
     ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 -1
     ; CHECK-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s64>) = G_BUILD_VECTOR [[C]](s64), [[C]](s64)
-    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG1]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG]]
-    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG1]]
+    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP1]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP]]
+    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP1]]
     ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR]]
     ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR1]]
     ; CHECK-NEXT: [[OR:%[0-9]+]]:_(<2 x s64>) = G_OR [[AND]], [[AND2]]
@@ -521,14 +519,12 @@ body: |
     ; CHECK-NEXT: [[DEF:%[0-9]+]]:_(<2 x s64>) = G_IMPLICIT_DEF
     ; CHECK-NEXT: [[ICMP:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(ult), [[DEF]](<2 x s64>), [[DEF]]
     ; CHECK-NEXT: [[ICMP1:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(ult), [[DEF]](<2 x s64>), [[DEF]]
-    ; CHECK-NEXT: [[SEXT_INREG:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP]], 1
-    ; CHECK-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP1]], 1
     ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 -1
     ; CHECK-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s64>) = G_BUILD_VECTOR [[C]](s64), [[C]](s64)
-    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG1]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG]]
-    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG1]]
+    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP1]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP]]
+    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP1]]
     ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR]]
     ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR1]]
     ; CHECK-NEXT: [[OR:%[0-9]+]]:_(<2 x s64>) = G_OR [[AND]], [[AND2]]
@@ -793,14 +789,12 @@ body: |
     ; CHECK-NEXT: [[DEF:%[0-9]+]]:_(<2 x s64>) = G_IMPLICIT_DEF
     ; CHECK-NEXT: [[ICMP:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(sgt), [[DEF]](<2 x s64>), [[DEF]]
     ; CHECK-NEXT: [[ICMP1:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(sgt), [[DEF]](<2 x s64>), [[DEF]]
-    ; CHECK-NEXT: [[SEXT_INREG:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP]], 1
-    ; CHECK-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP1]], 1
     ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 -1
     ; CHECK-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s64>) = G_BUILD_VECTOR [[C]](s64), [[C]](s64)
-    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG1]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG]]
-    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG1]]
+    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP1]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP]]
+    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP1]]
     ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR]]
     ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR1]]
     ; CHECK-NEXT: [[OR:%[0-9]+]]:_(<2 x s64>) = G_OR [[AND]], [[AND2]]
@@ -1065,14 +1059,12 @@ body: |
     ; CHECK-NEXT: [[DEF:%[0-9]+]]:_(<2 x s64>) = G_IMPLICIT_DEF
     ; CHECK-NEXT: [[ICMP:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(ugt), [[DEF]](<2 x s64>), [[DEF]]
     ; CHECK-NEXT: [[ICMP1:%[0-9]+]]:_(<2 x s64>) = G_ICMP intpred(ugt), [[DEF]](<2 x s64>), [[DEF]]
-    ; CHECK-NEXT: [[SEXT_INREG:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP]], 1
-    ; CHECK-NEXT: [[SEXT_INREG1:%[0-9]+]]:_(<2 x s64>) = G_SEXT_INREG [[ICMP1]], 1
     ; CHECK-NEXT: [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 -1
     ; CHECK-NEXT: [[BUILD_VECTOR:%[0-9]+]]:_(<2 x s64>) = G_BUILD_VECTOR [[C]](s64), [[C]](s64)
-    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[SEXT_INREG1]], [[BUILD_VECTOR]]
-    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG]]
-    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[SEXT_INREG1]]
+    ; CHECK-NEXT: [[XOR:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[XOR1:%[0-9]+]]:_(<2 x s64>) = G_XOR [[ICMP1]], [[BUILD_VECTOR]]
+    ; CHECK-NEXT: [[AND:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP]]
+    ; CHECK-NEXT: [[AND1:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[ICMP1]]
     ; CHECK-NEXT: [[AND2:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR]]
     ; CHECK-NEXT: [[AND3:%[0-9]+]]:_(<2 x s64>) = G_AND [[DEF]], [[XOR1]]
     ; CHECK-NEXT: [[OR:%[0-9]+]]:_(<2 x s64>) = G_OR [[AND]], [[AND2]]
diff --git a/llvm/test/CodeGen/AArch64/aarch64-dup-ext.ll b/llvm/test/CodeGen/AArch64/aarch64-dup-ext.ll
index 295863f18fd41..be79135c8b831 100644
--- a/llvm/test/CodeGen/AArch64/aarch64-dup-ext.ll
+++ b/llvm/test/CodeGen/AArch64/aarch64-dup-ext.ll
@@ -335,8 +335,6 @@ define void @typei1_orig(i64 %a, ptr %p, ptr %q) {
 ; CHECK-GI-NEXT:    cmtst v0.8h, v0.8h, v0.8h
 ; CHECK-GI-NEXT:    mvn v1.16b, v1.16b
 ; CHECK-GI-NEXT:    uzp1 v0.16b, v0.16b, v1.16b
-; CHECK-GI-NEXT:    shl v0.16b, v0.16b, #7
-; CHECK-GI-NEXT:    sshr v0.16b, v0.16b, #7
 ; CHECK-GI-NEXT:    str q0, [x1]
 ; CHECK-GI-NEXT:    ret
   %tmp = xor <16 x i1> zeroinitializer, <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>
diff --git a/llvm/test/CodeGen/AArch64/fcmp.ll b/llvm/test/CodeGen/AArch64/fcmp.ll
index e16ea9883a0c4..b9d65fa0314c4 100644
--- a/llvm/test/CodeGen/AArch64/fcmp.ll
+++ b/llvm/test/CodeGen/AArch64/fcmp.ll
@@ -969,8 +969,6 @@ define <4 x i32> @v4f64_i32(<4 x double> %a, <4 x double> %b, <4 x i32> %d, <4 x
 ; CHECK-GI-NEXT:    fcmgt v0.2d, v2.2d, v0.2d
 ; CHECK-GI-NEXT:    fcmgt v1.2d, v3.2d, v1.2d
 ; CHECK-GI-NEXT:    uzp1 v0.4s, v0.4s, v1.4s
-; CHECK-GI-NEXT:    shl v0.4s, v0.4s, #31
-; CHECK-GI-NEXT:    sshr v0.4s, v0.4s, #31
 ; CHECK-GI-NEXT:    bsl v0.16b, v4.16b, v5.16b
 ; CHECK-GI-NEXT:    ret
 entry:
@@ -1291,8 +1289,6 @@ define <8 x half> @v8f16_half(<8 x half> %a, <8 x half> %b, <8 x half> %d, <8 x
 ; CHECK-GI-NOFP16-NEXT:    fcmgt v4.4s, v5.4s, v4.4s
 ; CHECK-GI-NOFP16-NEXT:    fcmgt v0.4s, v1.4s, v0.4s
 ; CHECK-GI-NOFP16-NEXT:    uzp1 v0.8h, v4.8h, v0.8h
-; CHECK-GI-NOFP16-NEXT:    shl v0.8h, v0.8h, #15
-; CHECK-GI-NOFP16-NEXT:    sshr v0.8h, v0.8h, #15
 ; CHECK-GI-NOFP16-NEXT:    bsl v0.16b, v2.16b, v3.16b
 ; CHECK-GI-NOFP16-NEXT:    ret
 ;
@@ -1341,9 +1337,9 @@ define <16 x half> @v16f16_half(<16 x half> %a, <16 x half> %b, <16 x half> %d,
 ; CHECK-GI-NOFP16-NEXT:    fcvtl v16.4s, v0.4h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl2 v0.4s, v0.8h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl v17.4s, v1.4h
-; CHECK-GI-NOFP16-NEXT:    fcvtl2 v1.4s, v1.8h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl v18.4s, v2.4h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl2 v2.4s, v2.8h
+; CHECK-GI-NOFP16-NEXT:    fcvtl2 v1.4s, v1.8h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl v19.4s, v3.4h
 ; CHECK-GI-NOFP16-NEXT:    fcvtl2 v3.4s, v3.8h
 ; CHECK-GI-NOFP16-NEXT:    fcmgt v16.4s, v18.4s, v16.4s
@@ -1352,12 +1348,12 @@ define <16 x half> @v16f16_half(<16 x half> %a, <16 x half> %b, <16 x half> %d,
 ; CHECK-GI-NOFP16-NEXT:    fcmgt v1.4s, v3.4s, v1.4s
 ; CHECK-GI-NOFP16-NEXT:    uzp1 v0.8h, v16.8h, v0.8h
 ; CHECK-GI-NOFP16-NEXT:    uzp1 v1.8h, v2.8h, v1.8h
-; CHECK-GI-NOFP16-NEXT:    shl v0.8h, v0.8h, #15
-; CHECK-GI-NOFP16-NEXT:    shl v1.8h, v1.8h, #15
-; CHECK-GI-NOFP16-NEXT:    sshr v0.8h, v0.8h, #15
-; CHECK-GI-NOFP16-NEXT:    sshr v1.8h, v1.8h, #15
-; CHECK-GI-NOFP16-NEXT:    bsl v0.16b, v4.16b, v6.16b
-; CHECK-GI-NOFP16-NEXT:    bsl v1.16b, v5.16b, v7.16b
+; CHECK-GI-NOFP16-NEXT:    and v2.16b, v4.16b, v0.16b
+; CHECK-GI-NOFP16-NEXT:    bic v0.16b, v6.16b, v0.16b
+; CHECK-GI-NOFP16-NEXT:    and v3.16b, v5.16b, v1.16b
+; CHECK-GI-NOFP16-NEXT:    bic v1.16b, v7.16b, v1.16b
+; CHECK-GI-NOFP16-NEXT:    orr v0.16b, v2.16b, v0.16b
+; CHECK-GI-NOFP16-NEXT:    orr v1.16b, v3.16b, v1.16b
 ; CHECK-GI-NOFP16-NEXT:    ret
 ;
 ; CHECK-GI-FP16-LABEL: v16f16_half:
diff --git a/llvm/test/CodeGen/AArch64/scmp.ll b/llvm/test/CodeGen/AArch64/scmp.ll
index 7a73578f43e80..73cbd2cbeee48 100644
--- a/llvm/test/CodeGen/AArch64/scmp.ll
+++ b/llvm/test/CodeGen/AArch64/scmp.ll
@@ -310,10 +310,6 @@ define <16 x i8> @signOf_neon_scmp(<8 x i16> %s0_lo, <8 x i16> %s0_hi, <8 x i16>
 ; CHECK-GI-NEXT:    cmgt v1.8h, v3.8h, v1.8h
 ; CHECK-GI-NEXT:    uzp1 v0.16b, v0.16b, v1.16b
 ; CHECK-GI-NEXT:    uzp1 v1.16b, v4.16b, v5.16b
-; CHECK-GI-NEXT:    shl v0.16b, v0.16b, #7
-; CHECK-GI-NEXT:    shl v1.16b, v1.16b, #7
-; CHECK-GI-NEXT:    sshr v0.16b, v0.16b, #7
-; CHECK-GI-NEXT:    sshr v1.16b, v1.16b, #7
 ; CHECK-GI-NEXT:    sub v0.16b, v0.16b, v1.16b
 ; CHECK-GI-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/AArch64/ucmp.ll b/llvm/test/CodeGen/AArch64/ucmp.ll
index ad46e4abc477c..af8225307fedd 100644
--- a/llvm/test/CodeGen/AArch64/ucmp.ll
+++ b/llvm/test/CodeGen/AArch64/ucmp.ll
@@ -348,10 +348,6 @@ define <16 x i8> @signOf_neon(<8 x i16> %s0_lo, <8 x i16> %s0_hi, <8 x i16> %s1_
 ; CHECK-GI-NEXT:    cmhi v1.8h, v3.8h, v1.8h
 ; CHECK-GI-NEXT:    uzp1 v0.16b, v0.16b, v1.16b
 ; CHECK-GI-NEXT:    uzp1 v1.16b, v4.16b, v5.16b
-; CHECK-GI-NEXT:    shl v0.16b, v0.16b, #7
-; CHECK-GI-NEXT:    shl v1.16b, v1.16b, #7
-; CHECK-GI-NEXT:    sshr v0.16b, v0.16b, #7
-; CHECK-GI-NEXT:    sshr v1.16b, v1.16b, #7
 ; CHECK-GI-NEXT:    sub v0.16b, v0.16b, v1.16b
 ; CHECK-GI-NEXT:    ret
 entry:

llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp

davemgreen

Thanks - LGTM

llvm-ci · 2025-06-04T10:26:07Z

LLVM Buildbot has detected a new failure on builder lldb-x86_64-debian running on lldb-x86_64-debian while building llvm at step 6 "test".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/162/builds/23827

Here is the relevant piece of the build log for the reference

Step 6 (test) failure: build (failure)
...
UNSUPPORTED: lldb-shell :: ScriptInterpreter/Python/Crashlog/json.test (2983 of 2992)
UNSUPPORTED: lldb-shell :: Heap/heap-cstr.test (2984 of 2992)
UNSUPPORTED: lldb-shell :: ScriptInterpreter/Lua/breakpoint_oneline_callback.test (2985 of 2992)
UNSUPPORTED: lldb-shell :: ScriptInterpreter/Python/Crashlog/parser_text.test (2986 of 2992)
UNSUPPORTED: lldb-shell :: ScriptInterpreter/Lua/breakpoint_function_callback.test (2987 of 2992)
PASS: lldb-api :: api/multithreaded/TestMultithreaded.py (2988 of 2992)
PASS: lldb-api :: terminal/TestEditlineCompletions.py (2989 of 2992)
PASS: lldb-api :: commands/process/attach/TestProcessAttach.py (2990 of 2992)
PASS: lldb-api :: repl/clang/TestClangREPL.py (2991 of 2992)
TIMEOUT: lldb-shell :: Settings/TestCxxFrameFormatRecursive.test (2992 of 2992)
******************** TEST 'lldb-shell :: Settings/TestCxxFrameFormatRecursive.test' FAILED ********************
Exit Code: -9
Timeout: Reached timeout of 600 seconds

Command Output (stderr):
--
split-file /home/worker/2.0.1/lldb-x86_64-debian/llvm-project/lldb/test/Shell/Settings/TestCxxFrameFormatRecursive.test /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp # RUN: at line 6
+ split-file /home/worker/2.0.1/lldb-x86_64-debian/llvm-project/lldb/test/Shell/Settings/TestCxxFrameFormatRecursive.test /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp
/home/worker/2.0.1/lldb-x86_64-debian/build/bin/clang --target=specify-a-target-or-use-a-_host-substitution --target=x86_64-unknown-linux-gnu -pthread -fmodules-cache-path=/home/worker/2.0.1/lldb-x86_64-debian/build/lldb-test-build.noindex/module-cache-clang/lldb-shell -g -gdwarf /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp/main.cpp -o /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp.out # RUN: at line 7
+ /home/worker/2.0.1/lldb-x86_64-debian/build/bin/clang --target=specify-a-target-or-use-a-_host-substitution --target=x86_64-unknown-linux-gnu -pthread -fmodules-cache-path=/home/worker/2.0.1/lldb-x86_64-debian/build/lldb-test-build.noindex/module-cache-clang/lldb-shell -g -gdwarf /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp/main.cpp -o /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp.out
clang: warning: argument unused during compilation: '-fmodules-cache-path=/home/worker/2.0.1/lldb-x86_64-debian/build/lldb-test-build.noindex/module-cache-clang/lldb-shell' [-Wunused-command-line-argument]
/home/worker/2.0.1/lldb-x86_64-debian/build/bin/lldb --no-lldbinit -S /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/lit-lldb-init-quiet -o "settings set interpreter.stop-command-source-on-error false"        -x -b -s /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp/commands.input /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp.out -o exit 2>&1        | /home/worker/2.0.1/lldb-x86_64-debian/build/bin/FileCheck /home/worker/2.0.1/lldb-x86_64-debian/llvm-project/lldb/test/Shell/Settings/TestCxxFrameFormatRecursive.test # RUN: at line 8
+ /home/worker/2.0.1/lldb-x86_64-debian/build/bin/lldb --no-lldbinit -S /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/lit-lldb-init-quiet -o 'settings set interpreter.stop-command-source-on-error false' -x -b -s /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp/commands.input /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/Shell/Settings/Output/TestCxxFrameFormatRecursive.test.tmp.out -o exit
+ /home/worker/2.0.1/lldb-x86_64-debian/build/bin/FileCheck /home/worker/2.0.1/lldb-x86_64-debian/llvm-project/lldb/test/Shell/Settings/TestCxxFrameFormatRecursive.test

--

********************
********************
Timed Out Tests (1):
  lldb-shell :: Settings/TestCxxFrameFormatRecursive.test


Testing Time: 637.92s

Total Discovered Tests: 33276
  Skipped          :     2 (0.01%)
  Unsupported      :   458 (1.38%)
  Passed           : 32790 (98.54%)
  Expectedly Failed:    25 (0.08%)
  Timed Out        :     1 (0.00%)
FAILED: tools/lldb/test/CMakeFiles/check-lldb /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test/CMakeFiles/check-lldb 
cd /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test && /usr/bin/python3 /home/worker/2.0.1/lldb-x86_64-debian/build/./bin/llvm-lit -v /home/worker/2.0.1/lldb-x86_64-debian/build/tools/lldb/test
ninja: build stopped: subcommand failed.

usha1830 · 2025-06-04T11:01:50Z

1 test failed due to timeout. Looks like an intermittent and unrelated issue

Timed Out Tests (1):
  lldb-shell :: Settings/TestCxxFrameFormatRecursive.test

jayfoad · 2025-06-04T11:00:34Z

llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp

+    unsigned NumSubVectorElts =
+        MRI.getType(MI.getOperand(1).getReg()).getNumElements();
+    unsigned NumSubVectors = MI.getNumOperands() - 1;
+    for (unsigned i = 0; i < NumSubVectors; ++i) {


Suggested change

for (unsigned i = 0; i < NumSubVectors; ++i) {

for (unsigned I = 0; I < NumSubVectors; ++I) {

@jayfoad

…tyle consistency (#143113) Following up on a comment on #142355. Updated other instances in the file as well. @jayfoad

@jayfoad

…g.cpp for style consistency (#143113) Following up on a comment on llvm/llvm-project#142355. Updated other instances in the file as well. @jayfoad

…#142355) Code ported from SelectionDAG::ComputeNumSignBits

@jayfoad

…tyle consistency (llvm#143113) Following up on a comment on llvm#142355. Updated other instances in the file as well. @jayfoad

…#142355) Code ported from SelectionDAG::ComputeNumSignBits

@jayfoad

…tyle consistency (llvm#143113) Following up on a comment on llvm#142355. Updated other instances in the file as well. @jayfoad

@jayfoad

…tyle consistency (llvm#143113) Following up on a comment on llvm#142355. Updated other instances in the file as well. @jayfoad

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits

88cbf67

llvmbot added backend:AArch64 llvm:globalisel labels Jun 2, 2025

davemgreen requested review from jayfoad, arsenm and davemgreen June 2, 2025 10:20

davemgreen reviewed Jun 2, 2025

View reviewed changes

llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp Outdated Show resolved Hide resolved

llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp Outdated Show resolved Hide resolved

llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp Show resolved Hide resolved

Address review comments

078976c

davemgreen approved these changes Jun 4, 2025

View reviewed changes

davemgreen merged commit cf348e8 into llvm:main Jun 4, 2025
11 checks passed

jayfoad reviewed Jun 4, 2025

View reviewed changes

usha1830 mentioned this pull request Jun 6, 2025

[GISel] [NFC] Capitalize loop indices in GISelValueTracking.cpp for style consistency #143113

Merged

arsenm pushed a commit that referenced this pull request Jun 6, 2025

[GISel] [NFC] Capitalize loop indices in GISelValueTracking.cpp for s…

d5a1f49

…tyle consistency (#143113) Following up on a comment on #142355. Updated other instances in the file as well. @jayfoad

rorth pushed a commit to rorth/llvm-project that referenced this pull request Jun 11, 2025

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits (llvm…

6ae7905

…#142355) Code ported from SelectionDAG::ComputeNumSignBits

DhruvSrivastavaX pushed a commit to DhruvSrivastavaX/lldb-for-aix that referenced this pull request Jun 12, 2025

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits (llvm…

1b08956

…#142355) Code ported from SelectionDAG::ComputeNumSignBits

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits #142355

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits #142355

Uh oh!

usha1830 commented Jun 2, 2025

Uh oh!

llvmbot commented Jun 2, 2025 •

edited

Loading

Uh oh!

Uh oh!

Uh oh!

Uh oh!

davemgreen left a comment

Uh oh!

Uh oh!

llvm-ci commented Jun 4, 2025

Uh oh!

usha1830 commented Jun 4, 2025

Uh oh!

jayfoad Jun 4, 2025

Uh oh!

Uh oh!

	for (unsigned i = 0; i < NumSubVectors; ++i) {
	for (unsigned I = 0; I < NumSubVectors; ++I) {

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits #142355

[GlobalISel] Add G_CONCAT_VECTOR handling in computeNumSignBits #142355

Uh oh!

Conversation

usha1830 commented Jun 2, 2025

Uh oh!

llvmbot commented Jun 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

davemgreen left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

llvm-ci commented Jun 4, 2025

Uh oh!

usha1830 commented Jun 4, 2025

Uh oh!

jayfoad Jun 4, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

llvmbot commented Jun 2, 2025 •

edited

Loading