and positive: positive, because it could be directly computed to be positive;
negative, because the nsw flags means it is either negative or undefined (the
multiplication always overflowed).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145104
91177308-0d34-0410-b5e6-
96231b3b80d8
APInt::getHighBitsSet(BitWidth, LeadZ);
KnownZero &= Mask;
APInt::getHighBitsSet(BitWidth, LeadZ);
KnownZero &= Mask;
- if (isKnownNonNegative)
+ // Only make use of no-wrap flags if we failed to compute the sign bit
+ // directly. This matters if the multiplication always overflows, in
+ // which case we prefer to follow the result of the direct computation,
+ // though as the program is invoking undefined behaviour we can choose
+ // whatever we like here.
+ if (isKnownNonNegative && !KnownOne.isNegative())
KnownZero.setBit(BitWidth - 1);
KnownZero.setBit(BitWidth - 1);
- else if (isKnownNegative)
+ else if (isKnownNegative && !KnownZero.isNegative())
KnownOne.setBit(BitWidth - 1);
return;
KnownOne.setBit(BitWidth - 1);
return;
--- /dev/null
+; RUN: opt < %s -instsimplify
+
+; The mul can be proved to always overflow (turning a negative value
+; into a positive one) and thus results in undefined behaviour. At
+; the same time we were deducing from the nsw flag that that mul could
+; be assumed to have a negative value (since if not it has an undefined
+; value, which can be taken to be negative). We were reporting the mul
+; as being both positive and negative, firing an assertion!
+define i1 @test1(i32 %a) {
+entry:
+ %0 = or i32 %a, 1
+ %1 = shl i32 %0, 31
+ %2 = mul nsw i32 %1, 4
+ %3 = and i32 %2, -4
+ %4 = icmp ne i32 %3, 0
+ ret i1 %4
+}