[llvm] 4e3edef - Use pragmas to work around MSVC x86_32 debug miscompile bug

[Reid Kleckner via llvm-commits]

Author: Reid Kleckner
Date: 2020-09-10T14:50:01-07:00
New Revision: 4e3edef4b8b637c0c76897497eb7c66f00157210

URL: https://github.com/llvm/llvm-project/commit/4e3edef4b8b637c0c76897497eb7c66f00157210
DIFF: https://github.com/llvm/llvm-project/commit/4e3edef4b8b637c0c76897497eb7c66f00157210.diff

LOG: Use pragmas to work around MSVC x86_32 debug miscompile bug

Halide users reported this here: https://llvm.org/pr46176
I reported the issue to MSVC here:
https://developercommunity.visualstudio.com/content/problem/1179643/msvc-copies-overaligned-non-trivially-copyable-par.html

This codepath is apparently not covered by LLVM's unit tests, so I added
coverage in a unit test.

If we want to support this configuration going forward, it means that is
in general not safe to pass a SmallVector<T, N> by value if alignof(T)
is greater than 4. This doesn't appear to come up often because passing
a SmallVector by value is inefficient and not idiomatic: it copies the
inline storage. In this case, the SmallVector<LLT,4> is captured by
value by a lambda, and the lambda is passed by value into std::function,
and that's how we hit the bug.

Differential Revision: https://reviews.llvm.org/D87475

Added: 
    

Modified: 
    llvm/lib/CodeGen/GlobalISel/LegalityPredicates.cpp
    llvm/unittests/CodeGen/GlobalISel/LegalizerInfoTest.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/CodeGen/GlobalISel/LegalityPredicates.cpp b/llvm/lib/CodeGen/GlobalISel/LegalityPredicates.cpp
index 17bce517814d..e25705e0e101 100644
--- a/llvm/lib/CodeGen/GlobalISel/LegalityPredicates.cpp
+++ b/llvm/lib/CodeGen/GlobalISel/LegalityPredicates.cpp
@@ -10,6 +10,17 @@
 //
 //===----------------------------------------------------------------------===//
 
+// Disable optimizations to work around MSVC debug mode bug in 32-bit:
+// https://developercommunity.visualstudio.com/content/problem/1179643/msvc-copies-overaligned-non-trivially-copyable-par.html
+// FIXME: Remove this when the issue is closed.
+#if defined(_MSC_VER) && !defined(__clang__) && defined(_M_IX86)
+// We have to disable runtime checks in order to enable optimizations. This is
+// done for the entire file because the problem is actually observed in STL
+// template functions.
+#pragma runtime_checks("", off)
+#pragma optimize("gs", on)
+#endif
+
 #include "llvm/CodeGen/GlobalISel/LegalizerInfo.h"
 
 using namespace llvm;

diff  --git a/llvm/unittests/CodeGen/GlobalISel/LegalizerInfoTest.cpp b/llvm/unittests/CodeGen/GlobalISel/LegalizerInfoTest.cpp
index 7fd2ea453a2a..ac9112fe5aa4 100644
--- a/llvm/unittests/CodeGen/GlobalISel/LegalizerInfoTest.cpp
+++ b/llvm/unittests/CodeGen/GlobalISel/LegalizerInfoTest.cpp
@@ -406,3 +406,13 @@ TEST(LegalizerInfoTest, MMOAlignment) {
                                   32, 8, AtomicOrdering::NotAtomic }));
   }
 }
+
+// This code sequence doesn't do anything, but it covers a previously uncovered
+// codepath that used to crash in MSVC x86_32 debug mode.
+TEST(LegalizerInfoTest, MSVCDebugMiscompile) {
+  const LLT S1 = LLT::scalar(1);
+  const LLT P0 = LLT::pointer(0, 32);
+  LegalizerInfo LI;
+  auto Builder = LI.getActionDefinitionsBuilder(TargetOpcode::G_PTRTOINT);
+  (void)Builder.legalForCartesianProduct({S1}, {P0});
+}