[llvm] [Transforms] Resolve FIXME: Pick the smallest legal type that fits (PR #79158)

via llvm-commits llvm-commits at lists.llvm.org
Sat Jan 27 10:01:44 PST 2024


https://github.com/AtariDreams updated https://github.com/llvm/llvm-project/pull/79158

>From ea8d0a523a88134a104e2d14bc64effad310674b Mon Sep 17 00:00:00 2001
From: Rose <83477269+AtariDreams at users.noreply.github.com>
Date: Tue, 23 Jan 2024 10:15:46 -0500
Subject: [PATCH] [Transforms] Resolve FIXME: Pick the smallest legal type that
 fits

Pick the type based on the smallest bit-width possible, using DataLayout.
---
 .../llvm/Transforms/Scalar/Float2Int.h        |   2 +-
 llvm/lib/Transforms/Scalar/Float2Int.cpp      |  42 ++-
 llvm/test/Transforms/Float2Int/smaller-int.ll | 351 ++++++++++++++++++
 3 files changed, 376 insertions(+), 19 deletions(-)
 create mode 100644 llvm/test/Transforms/Float2Int/smaller-int.ll

diff --git a/llvm/include/llvm/Transforms/Scalar/Float2Int.h b/llvm/include/llvm/Transforms/Scalar/Float2Int.h
index 83be329bed60ba..337e229efcf379 100644
--- a/llvm/include/llvm/Transforms/Scalar/Float2Int.h
+++ b/llvm/include/llvm/Transforms/Scalar/Float2Int.h
@@ -44,7 +44,7 @@ class Float2IntPass : public PassInfoMixin<Float2IntPass> {
   std::optional<ConstantRange> calcRange(Instruction *I);
   void walkBackwards();
   void walkForwards();
-  bool validateAndTransform();
+  bool validateAndTransform(const DataLayout &DL);
   Value *convert(Instruction *I, Type *ToTy);
   void cleanup();
 
diff --git a/llvm/lib/Transforms/Scalar/Float2Int.cpp b/llvm/lib/Transforms/Scalar/Float2Int.cpp
index ccca8bcc1a56ac..01218f6d26e271 100644
--- a/llvm/lib/Transforms/Scalar/Float2Int.cpp
+++ b/llvm/lib/Transforms/Scalar/Float2Int.cpp
@@ -311,7 +311,7 @@ void Float2IntPass::walkForwards() {
 }
 
 // If there is a valid transform to be done, do it.
-bool Float2IntPass::validateAndTransform() {
+bool Float2IntPass::validateAndTransform(const DataLayout &DL) {
   bool MadeChange = false;
 
   // Iterate over every disjoint partition of the def-use graph.
@@ -321,8 +321,8 @@ bool Float2IntPass::validateAndTransform() {
     Type *ConvertedToTy = nullptr;
 
     // For every member of the partition, union all the ranges together.
-    for (auto MI = ECs.member_begin(It), ME = ECs.member_end();
-         MI != ME; ++MI) {
+    for (auto MI = ECs.member_begin(It), ME = ECs.member_end(); MI != ME;
+         ++MI) {
       Instruction *I = *MI;
       auto SeenI = SeenInsts.find(I);
       if (SeenI == SeenInsts.end())
@@ -352,8 +352,8 @@ bool Float2IntPass::validateAndTransform() {
 
     // If the set was empty, or we failed, or the range is poisonous,
     // bail out.
-    if (ECs.member_begin(It) == ECs.member_end() || Fail ||
-        R.isFullSet() || R.isSignWrappedSet())
+    if (ECs.member_begin(It) == ECs.member_end() || Fail || R.isFullSet() ||
+        R.isSignWrappedSet())
       continue;
     assert(ConvertedToTy && "Must have set the convertedtoty by this point!");
 
@@ -370,24 +370,29 @@ bool Float2IntPass::validateAndTransform() {
     // Do we need more bits than are in the mantissa of the type we converted
     // to? semanticsPrecision returns the number of mantissa bits plus one
     // for the sign bit.
-    unsigned MaxRepresentableBits
-      = APFloat::semanticsPrecision(ConvertedToTy->getFltSemantics()) - 1;
+    unsigned MaxRepresentableBits =
+        APFloat::semanticsPrecision(ConvertedToTy->getFltSemantics()) - 1;
     if (MinBW > MaxRepresentableBits) {
       LLVM_DEBUG(dbgs() << "F2I: Value not guaranteed to be representable!\n");
       continue;
     }
-    if (MinBW > 64) {
-      LLVM_DEBUG(
-          dbgs() << "F2I: Value requires more than 64 bits to represent!\n");
-      continue;
-    }
 
-    // OK, R is known to be representable. Now pick a type for it.
-    // FIXME: Pick the smallest legal type that will fit.
-    Type *Ty = (MinBW > 32) ? Type::getInt64Ty(*Ctx) : Type::getInt32Ty(*Ctx);
+    // OK, R is known to be representable.
+    // Pick the smallest legal type that will fit.
+    Type *Ty = DL.getSmallestLegalIntType(*Ctx, MinBW);
+    if (!Ty) {
+      if (MinBW > 64) {
+        LLVM_DEBUG(dbgs() << "F2I: Value requires more than bits to represent "
+                             "than the target supports!\n");
+        continue;
+      }
+
+      // Every supported target supports 64 and 32-bit
+      // integers.
+      Ty = (MinBW > 32) ? Type::getInt64Ty(*Ctx) : Type::getInt32Ty(*Ctx);
+    }
 
-    for (auto MI = ECs.member_begin(It), ME = ECs.member_end();
-         MI != ME; ++MI)
+    for (auto MI = ECs.member_begin(It), ME = ECs.member_end(); MI != ME; ++MI)
       convert(*MI, Ty);
     MadeChange = true;
   }
@@ -491,7 +496,8 @@ bool Float2IntPass::runImpl(Function &F, const DominatorTree &DT) {
   walkBackwards();
   walkForwards();
 
-  bool Modified = validateAndTransform();
+  const DataLayout &DL = F.getParent()->getDataLayout();
+  bool Modified = validateAndTransform(DL);
   if (Modified)
     cleanup();
   return Modified;
diff --git a/llvm/test/Transforms/Float2Int/smaller-int.ll b/llvm/test/Transforms/Float2Int/smaller-int.ll
new file mode 100644
index 00000000000000..9376943c6ea94f
--- /dev/null
+++ b/llvm/test/Transforms/Float2Int/smaller-int.ll
@@ -0,0 +1,351 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes='float2int' -S | FileCheck %s
+target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
+
+;
+; Positive tests
+;
+
+define i16 @simple1(i8 %a) {
+; CHECK-LABEL: @simple1(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[T21:%.*]] = add i16 [[TMP1]], 1
+; CHECK-NEXT:    ret i16 [[T21]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fadd float %t1, 1.0
+  %t3 = fptoui float %t2 to i16
+  ret i16 %t3
+}
+
+define i8 @simple2(i8 %a) {
+; CHECK-LABEL: @simple2(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[T21:%.*]] = sub i16 [[TMP1]], 1
+; CHECK-NEXT:    [[TMP2:%.*]] = trunc i16 [[T21]] to i8
+; CHECK-NEXT:    ret i8 [[TMP2]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fsub float %t1, 1.0
+  %t3 = fptoui float %t2 to i8
+  ret i8 %t3
+}
+
+define i32 @simple3(i8 %a) {
+; CHECK-LABEL: @simple3(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[T21:%.*]] = sub i16 [[TMP1]], 1
+; CHECK-NEXT:    [[TMP2:%.*]] = zext i16 [[T21]] to i32
+; CHECK-NEXT:    ret i32 [[TMP2]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fsub float %t1, 1.0
+  %t3 = fptoui float %t2 to i32
+  ret i32 %t3
+}
+
+define i1 @cmp(i8 %a, i8 %b) {
+; CHECK-LABEL: @cmp(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[TMP2:%.*]] = zext i8 [[B:%.*]] to i16
+; CHECK-NEXT:    [[T31:%.*]] = icmp slt i16 [[TMP1]], [[TMP2]]
+; CHECK-NEXT:    ret i1 [[T31]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = uitofp i8 %b to float
+  %t3 = fcmp ult float %t1, %t2
+  ret i1 %t3
+}
+
+define i32 @simple4(i32 %a) {
+; CHECK-LABEL: @simple4(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i32 [[A:%.*]] to i64
+; CHECK-NEXT:    [[T21:%.*]] = add i64 [[TMP1]], 1
+; CHECK-NEXT:    [[TMP2:%.*]] = trunc i64 [[T21]] to i32
+; CHECK-NEXT:    ret i32 [[TMP2]]
+;
+  %t1 = uitofp i32 %a to double
+  %t2 = fadd double %t1, 1.0
+  %t3 = fptoui double %t2 to i32
+  ret i32 %t3
+}
+
+define i32 @simple5(i8 %a, i8 %b) {
+; CHECK-LABEL: @simple5(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i32
+; CHECK-NEXT:    [[TMP2:%.*]] = zext i8 [[B:%.*]] to i32
+; CHECK-NEXT:    [[T31:%.*]] = add i32 [[TMP1]], 1
+; CHECK-NEXT:    [[T42:%.*]] = mul i32 [[T31]], [[TMP2]]
+; CHECK-NEXT:    ret i32 [[T42]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = uitofp i8 %b to float
+  %t3 = fadd float %t1, 1.0
+  %t4 = fmul float %t3, %t2
+  %t5 = fptoui float %t4 to i32
+  ret i32 %t5
+}
+
+define i32 @simple6(i8 %a, i8 %b) {
+; CHECK-LABEL: @simple6(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i32
+; CHECK-NEXT:    [[TMP2:%.*]] = zext i8 [[B:%.*]] to i32
+; CHECK-NEXT:    [[T31:%.*]] = sub i32 0, [[TMP1]]
+; CHECK-NEXT:    [[T42:%.*]] = mul i32 [[T31]], [[TMP2]]
+; CHECK-NEXT:    ret i32 [[T42]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = uitofp i8 %b to float
+  %t3 = fneg float %t1
+  %t4 = fmul float %t3, %t2
+  %t5 = fptoui float %t4 to i32
+  ret i32 %t5
+}
+
+; The two chains don't interact - failure of one shouldn't
+; cause failure of the other.
+
+define i32 @multi1(i8 %a, i8 %b, i8 %c, float %d) {
+; CHECK-LABEL: @multi1(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[TMP2:%.*]] = zext i8 [[B:%.*]] to i16
+; CHECK-NEXT:    [[FC:%.*]] = uitofp i8 [[C:%.*]] to float
+; CHECK-NEXT:    [[X1:%.*]] = add i16 [[TMP1]], [[TMP2]]
+; CHECK-NEXT:    [[TMP3:%.*]] = zext i16 [[X1]] to i32
+; CHECK-NEXT:    [[Z:%.*]] = fadd float [[FC]], [[D:%.*]]
+; CHECK-NEXT:    [[W:%.*]] = fptoui float [[Z]] to i32
+; CHECK-NEXT:    [[R:%.*]] = add i32 [[TMP3]], [[W]]
+; CHECK-NEXT:    ret i32 [[R]]
+;
+  %fa = uitofp i8 %a to float
+  %fb = uitofp i8 %b to float
+  %fc = uitofp i8 %c to float
+  %x = fadd float %fa, %fb
+  %y = fptoui float %x to i32
+  %z = fadd float %fc, %d
+  %w = fptoui float %z to i32
+  %r = add i32 %y, %w
+  ret i32 %r
+}
+
+define i16 @simple_negzero(i8 %a) {
+; CHECK-LABEL: @simple_negzero(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[T21:%.*]] = add i16 [[TMP1]], 0
+; CHECK-NEXT:    ret i16 [[T21]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fadd fast float %t1, -0.0
+  %t3 = fptoui float %t2 to i16
+  ret i16 %t3
+}
+
+define i32 @simple_negative(i8 %call) {
+; CHECK-LABEL: @simple_negative(
+; CHECK-NEXT:    [[TMP1:%.*]] = sext i8 [[CALL:%.*]] to i16
+; CHECK-NEXT:    [[MUL1:%.*]] = mul i16 [[TMP1]], -3
+; CHECK-NEXT:    [[TMP2:%.*]] = trunc i16 [[MUL1]] to i8
+; CHECK-NEXT:    [[CONV3:%.*]] = sext i8 [[TMP2]] to i32
+; CHECK-NEXT:    ret i32 [[CONV3]]
+;
+  %conv1 = sitofp i8 %call to float
+  %mul = fmul float %conv1, -3.000000e+00
+  %conv2 = fptosi float %mul to i8
+  %conv3 = sext i8 %conv2 to i32
+  ret i32 %conv3
+}
+
+define i16 @simple_fneg(i8 %a) {
+; CHECK-LABEL: @simple_fneg(
+; CHECK-NEXT:    [[TMP1:%.*]] = zext i8 [[A:%.*]] to i16
+; CHECK-NEXT:    [[T21:%.*]] = sub i16 0, [[TMP1]]
+; CHECK-NEXT:    ret i16 [[T21]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fneg fast float %t1
+  %t3 = fptoui float %t2 to i16
+  ret i16 %t3
+}
+
+;
+; Negative tests
+;
+
+; The two chains intersect, which means because one fails, no
+; transform can occur.
+
+define i32 @neg_multi1(i8 %a, i8 %b, i8 %c, float %d) {
+; CHECK-LABEL: @neg_multi1(
+; CHECK-NEXT:    [[FA:%.*]] = uitofp i8 [[A:%.*]] to float
+; CHECK-NEXT:    [[FC:%.*]] = uitofp i8 [[C:%.*]] to float
+; CHECK-NEXT:    [[X:%.*]] = fadd float [[FA]], [[FC]]
+; CHECK-NEXT:    [[Y:%.*]] = fptoui float [[X]] to i32
+; CHECK-NEXT:    [[Z:%.*]] = fadd float [[FC]], [[D:%.*]]
+; CHECK-NEXT:    [[W:%.*]] = fptoui float [[Z]] to i32
+; CHECK-NEXT:    [[R:%.*]] = add i32 [[Y]], [[W]]
+; CHECK-NEXT:    ret i32 [[R]]
+;
+  %fa = uitofp i8 %a to float
+  %fc = uitofp i8 %c to float
+  %x = fadd float %fa, %fc
+  %y = fptoui float %x to i32
+  %z = fadd float %fc, %d
+  %w = fptoui float %z to i32
+  %r = add i32 %y, %w
+  ret i32 %r
+}
+
+; The i32 * i32 = i64, which has 64 bits, which is greater than the 52 bits
+; that can be exactly represented in a double.
+
+define i64 @neg_muld(i32 %a, i32 %b) {
+; CHECK-LABEL: @neg_muld(
+; CHECK-NEXT:    [[FA:%.*]] = uitofp i32 [[A:%.*]] to double
+; CHECK-NEXT:    [[FB:%.*]] = uitofp i32 [[B:%.*]] to double
+; CHECK-NEXT:    [[MUL:%.*]] = fmul double [[FA]], [[FB]]
+; CHECK-NEXT:    [[R:%.*]] = fptoui double [[MUL]] to i64
+; CHECK-NEXT:    ret i64 [[R]]
+;
+  %fa = uitofp i32 %a to double
+  %fb = uitofp i32 %b to double
+  %mul = fmul double %fa, %fb
+  %r = fptoui double %mul to i64
+  ret i64 %r
+}
+
+; The i16 * i16 = i32, which can't be represented in a float, but can in a
+; double. This should fail, as the written code uses floats, not doubles so
+; the original result may be inaccurate.
+
+define i32 @neg_mulf(i16 %a, i16 %b) {
+; CHECK-LABEL: @neg_mulf(
+; CHECK-NEXT:    [[FA:%.*]] = uitofp i16 [[A:%.*]] to float
+; CHECK-NEXT:    [[FB:%.*]] = uitofp i16 [[B:%.*]] to float
+; CHECK-NEXT:    [[MUL:%.*]] = fmul float [[FA]], [[FB]]
+; CHECK-NEXT:    [[R:%.*]] = fptoui float [[MUL]] to i32
+; CHECK-NEXT:    ret i32 [[R]]
+;
+  %fa = uitofp i16 %a to float
+  %fb = uitofp i16 %b to float
+  %mul = fmul float %fa, %fb
+  %r = fptoui float %mul to i32
+  ret i32 %r
+}
+
+; "false" doesn't have an icmp equivalent.
+
+define i1 @neg_cmp(i8 %a, i8 %b) {
+; CHECK-LABEL: @neg_cmp(
+; CHECK-NEXT:    [[T1:%.*]] = uitofp i8 [[A:%.*]] to float
+; CHECK-NEXT:    [[T2:%.*]] = uitofp i8 [[B:%.*]] to float
+; CHECK-NEXT:    [[T3:%.*]] = fcmp false float [[T1]], [[T2]]
+; CHECK-NEXT:    ret i1 [[T3]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = uitofp i8 %b to float
+  %t3 = fcmp false float %t1, %t2
+  ret i1 %t3
+}
+
+; Division isn't a supported operator.
+
+define i16 @neg_div(i8 %a) {
+; CHECK-LABEL: @neg_div(
+; CHECK-NEXT:    [[T1:%.*]] = uitofp i8 [[A:%.*]] to float
+; CHECK-NEXT:    [[T2:%.*]] = fdiv float [[T1]], 1.000000e+00
+; CHECK-NEXT:    [[T3:%.*]] = fptoui float [[T2]] to i16
+; CHECK-NEXT:    ret i16 [[T3]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fdiv float %t1, 1.0
+  %t3 = fptoui float %t2 to i16
+  ret i16 %t3
+}
+
+; 1.2 is not an integer.
+
+define i16 @neg_remainder(i8 %a) {
+; CHECK-LABEL: @neg_remainder(
+; CHECK-NEXT:    [[T1:%.*]] = uitofp i8 [[A:%.*]] to float
+; CHECK-NEXT:    [[T2:%.*]] = fadd float [[T1]], 1.250000e+00
+; CHECK-NEXT:    [[T3:%.*]] = fptoui float [[T2]] to i16
+; CHECK-NEXT:    ret i16 [[T3]]
+;
+  %t1 = uitofp i8 %a to float
+  %t2 = fadd float %t1, 1.25
+  %t3 = fptoui float %t2 to i16
+  ret i16 %t3
+}
+
+; i80 > i64, which is the largest bitwidth handleable by default.
+
+define i80 @neg_toolarge(i80 %a) {
+; CHECK-LABEL: @neg_toolarge(
+; CHECK-NEXT:    [[T1:%.*]] = uitofp i80 [[A:%.*]] to fp128
+; CHECK-NEXT:    [[T2:%.*]] = fadd fp128 [[T1]], [[T1]]
+; CHECK-NEXT:    [[T3:%.*]] = fptoui fp128 [[T2]] to i80
+; CHECK-NEXT:    ret i80 [[T3]]
+;
+  %t1 = uitofp i80 %a to fp128
+  %t2 = fadd fp128 %t1, %t1
+  %t3 = fptoui fp128 %t2 to i80
+  ret i80 %t3
+}
+
+; The sequence %t1..%t3 cannot be converted because %t4 uses %t2.
+
+define i32 @neg_calluser(i32 %value) {
+; CHECK-LABEL: @neg_calluser(
+; CHECK-NEXT:    [[T1:%.*]] = sitofp i32 [[VALUE:%.*]] to double
+; CHECK-NEXT:    [[T2:%.*]] = fadd double [[T1]], 1.000000e+00
+; CHECK-NEXT:    [[T3:%.*]] = fcmp olt double [[T2]], 0.000000e+00
+; CHECK-NEXT:    [[T4:%.*]] = tail call double @g(double [[T2]])
+; CHECK-NEXT:    [[T5:%.*]] = fptosi double [[T4]] to i32
+; CHECK-NEXT:    [[T6:%.*]] = zext i1 [[T3]] to i32
+; CHECK-NEXT:    [[T7:%.*]] = add i32 [[T6]], [[T5]]
+; CHECK-NEXT:    ret i32 [[T7]]
+;
+  %t1 = sitofp i32 %value to double
+  %t2 = fadd double %t1, 1.0
+  %t3 = fcmp olt double %t2, 0.000000e+00
+  %t4 = tail call double @g(double %t2)
+  %t5 = fptosi double %t4 to i32
+  %t6 = zext i1 %t3 to i32
+  %t7 = add i32 %t6, %t5
+  ret i32 %t7
+}
+
+declare double @g(double)
+
+define <4 x i16> @neg_vector(<4 x i8> %a) {
+; CHECK-LABEL: @neg_vector(
+; CHECK-NEXT:    [[T1:%.*]] = uitofp <4 x i8> [[A:%.*]] to <4 x float>
+; CHECK-NEXT:    [[T2:%.*]] = fptoui <4 x float> [[T1]] to <4 x i16>
+; CHECK-NEXT:    ret <4 x i16> [[T2]]
+;
+  %t1 = uitofp <4 x i8> %a to <4 x float>
+  %t2 = fptoui <4 x float> %t1 to <4 x i16>
+  ret <4 x i16> %t2
+}
+
+; Don't crash while processing unreachable (non-standard) IR.
+
+define void @PR38502() {
+; CHECK-LABEL: @PR38502(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    ret void
+; CHECK:       bogusBB:
+; CHECK-NEXT:    [[INC1:%.*]] = fadd double [[INC:%.*]], 1.000000e+00
+; CHECK-NEXT:    [[INC]] = fadd double [[INC1]], 1.000000e+00
+; CHECK-NEXT:    [[TOBOOL:%.*]] = fcmp une double [[INC]], 0.000000e+00
+; CHECK-NEXT:    br label [[BOGUSBB:%.*]]
+;
+entry:
+  ret void
+
+bogusBB:                                          ; preds = %bogusBB
+  %inc1 = fadd double %inc, 1.000000e+00
+  %inc = fadd double %inc1, 1.000000e+00
+  %tobool = fcmp une double %inc, 0.000000e+00
+  br label %bogusBB
+}



More information about the llvm-commits mailing list