[PATCH] D28284: [Legalizer] Fix fp-to-uint to fp-tosint promotion assertion.

[Tim Shen via Phabricator via llvm-commits]

timshen created this revision.
timshen added reviewers: jmolloy, nadav, echristo, kbarton.
timshen added a subscriber: llvm-commits.
Herald added subscribers: nemanjai, mehdi_amini.

When promoting fp-to-uint16 to fp-to-sint32, the result is actually zero
extended. For example, given double 65534.0, without legalization:

  fp-to-uint16: 65534.0 -> 0xfffe

With the legalization:

  fp-to-sint32: 65534.0 -> 0x0000fffe

Without this patch, legalization wrongly emits a signed extend assertion,
which is consumed by later icmp instruction, and cause miscompile.

Note that the floating point value must be in [0, 65535), otherwise the
behavior is undefined.

This patch reverts r279223 behavior and adds more tests and
documentations.

In PR29041's context, James Molloy mentioned that:

  We don't need to mask because conversion from float->uint8_t is
  undefined if the integer part of the float value is not representable in
  uint8_t. Therefore we can assume this doesn't happen!

which is totally true and good, because fptoui is documented clearly to
have undefined behavior when overflow/underflow happens. We should take
the advantage of this behavior so that we can save unnecessary mask
instructions.


https://reviews.llvm.org/D28284

Files:
  llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
  llvm/test/CodeGen/AArch64/fptouint-i8-zext.ll
  llvm/test/CodeGen/PowerPC/fp64-to-int16.ll


Index: llvm/test/CodeGen/PowerPC/fp64-to-int16.ll
===================================================================
--- /dev/null
+++ llvm/test/CodeGen/PowerPC/fp64-to-int16.ll
@@ -0,0 +1,21 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -O0 < %s | FileCheck %s
+target triple = "powerpc64le--linux-gnu"
+
+define i1 @Test(double %a) {
+; CHECK-LABEL: Test:
+; CHECK:       # BB#0: # %entry
+; CHECK-NEXT:    xscvdpsxws 1, 1
+; CHECK-NEXT:    mfvsrwz 3, 1
+; CHECK-NEXT:    xori 3, 3, 65534
+; CHECK-NEXT:    cntlzw 3, 3
+; CHECK-NEXT:    srwi 3, 3, 5
+; CHECK-NEXT:    # implicit-def: %X4
+; CHECK-NEXT:    mr 4, 3
+; CHECK-NEXT:    mr 3, 4
+; CHECK-NEXT:    blr
+entry:
+  %conv = fptoui double %a to i16
+  %cmp = icmp eq i16 %conv, -2
+  ret i1 %cmp
+}
Index: llvm/test/CodeGen/AArch64/fptouint-i8-zext.ll
===================================================================
--- llvm/test/CodeGen/AArch64/fptouint-i8-zext.ll
+++ llvm/test/CodeGen/AArch64/fptouint-i8-zext.ll
@@ -3,9 +3,11 @@
 target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
 target triple = "aarch64"
 
+; If the float value is negative or too large, the result is undefined anyway;
+; otherwise, fcvtzs must returns a value in [0, 256), which guarantees zext.
+
 ; CHECK-LABEL: float_char_int_func:
 ; CHECK: fcvtzs [[A:w[0-9]+]], s0
-; CHECK-NEXT: and w0, [[A]], #0xff
 ; CHECK-NEXT: ret
 define i32 @float_char_int_func(float %infloatVal) {
 entry:
Index: llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
===================================================================
--- llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
+++ llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp
@@ -428,7 +428,11 @@
   // Assert that the converted value fits in the original type.  If it doesn't
   // (eg: because the value being converted is too big), then the result of the
   // original operation was undefined anyway, so the assert is still correct.
-  return DAG.getNode(NewOpc == ISD::FP_TO_UINT ?
+  //
+  // NOTE: fp-to-uint to fp-to-sint promotion guarantees zero extend. For example:
+  //   before legalization: fp-to-uint16, 65534. -> 0xfffe
+  //   after legalization: fp-to-sint32, 65534. -> 0x0000fffe
+  return DAG.getNode(N->getOpcode() == ISD::FP_TO_UINT ?
                      ISD::AssertZext : ISD::AssertSext, dl, NVT, Res,
                      DAG.getValueType(N->getValueType(0).getScalarType()));
 }


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D28284.83026.patch
Type: text/x-patch
Size: 2478 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170104/092c0a3f/attachment.bin>