[APInt] Fixed bug where APInt(UINT32_MAX, 0) would blow up when being constructed.

This was due to arithmetic overflow in the getNumBits() computation. Now we
cast BitWidth to a uint64_t so that does not occur during the computation. After
the computation is complete, the uint64_t is truncated when the function
returns.

I know that this is not something that is likely to happen, but it *IS* a valid
input and we should not blow up.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@199609 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Michael Gottesman 2014-01-19 20:33:38 +00:00
parent b45edea9b3
commit e7413972a4
2 changed files with 9 additions and 1 deletions

View File

@ -1265,7 +1265,7 @@ public:
/// \returns the number of words to hold the integer value with a given bit
/// width.
static unsigned getNumWords(unsigned BitWidth) {
return (BitWidth + APINT_BITS_PER_WORD - 1) / APINT_BITS_PER_WORD;
return ((uint64_t)BitWidth + APINT_BITS_PER_WORD - 1) / APINT_BITS_PER_WORD;
}
/// \brief Compute the number of active bits in the value

View File

@ -623,6 +623,14 @@ TEST(APIntTest, arrayAccess) {
}
}
TEST(APIntTest, LargeAPIntConstruction) {
// Check that we can properly construct very large APInt. It is very
// unlikely that people will ever do this, but it is a legal input,
// so we should not crash on it.
APInt A9(UINT32_MAX, 0);
EXPECT_FALSE(A9.getBoolValue());
}
TEST(APIntTest, nearestLogBase2) {
// Single word check.