Arguments spilled on the stack before a function call may have

alignment requirements, for example in the case of vectors.
These requirements are exploited by the code generator by using
move instructions that have similar alignment requirements, e.g.,
movaps on x86.

Although the code generator properly aligns the arguments with
respect to the displacement of the stack pointer it computes,
the displacement itself may cause misalignment. For example if
we have

%3 = load <16 x float>, <16 x float>* %1, align 64
call void @bar(<16 x float> %3, i32 0)

the x86 back-end emits:

movaps  32(%ecx), %xmm2
movaps  (%ecx), %xmm0
movaps  16(%ecx), %xmm1
movaps  48(%ecx), %xmm3
subl    $20, %esp       <-- if %esp was 16-byte aligned before this instruction, it no longer will be afterwards 
movaps  %xmm3, (%esp)   <-- movaps requires 16-byte alignment, while %esp is not aligned as such.
movl    $0, 16(%esp)
calll   __bar

To solve this, we need to make sure that the computed value with which
the stack pointer is changed is a multiple af the maximal alignment seen
during its computation. With this change we get proper alignment:

subl    $32, %esp
movaps  %xmm3, (%esp)

Differential Revision: http://reviews.llvm.org/D12337

llvm-svn: 248786

llvm/include/llvm/CodeGen/CallingConvLower.h

@@ -201,6 +201,7 @@ private:
   LLVMContext &Context;
 
   unsigned StackOffset;
+  unsigned MaxStackArgAlign;
   SmallVector<uint32_t, 16> UsedRegs;
   SmallVector<CCValAssign, 4> PendingLocs;
 
@@ -270,7 +271,18 @@ public:
   CallingConv::ID getCallingConv() const { return CallingConv; }
   bool isVarArg() const { return IsVarArg; }
 
-  unsigned getNextStackOffset() const { return StackOffset; }
+  /// getNextStackOffset - Return the next stack offset such that all stack
+  /// slots satisfy their alignment requirements.
+  unsigned getNextStackOffset() const {
+    return StackOffset;
+  }
+
+  /// getAlignedCallFrameSize - Return the size of the call frame needed to
+  /// be able to store all arguments and such that the alignment requirement
+  /// of each of the arguments is satisfied.
+  unsigned getAlignedCallFrameSize() const {
+    return RoundUpToAlignment(StackOffset, MaxStackArgAlign);
+  }
 
   /// isAllocated - Return true if the specified register (or an alias) is
   /// allocated.
@@ -400,9 +412,10 @@ public:
   /// and alignment.
   unsigned AllocateStack(unsigned Size, unsigned Align) {
     assert(Align && ((Align - 1) & Align) == 0); // Align is power of 2.
-    StackOffset = ((StackOffset + Align - 1) & ~(Align - 1));
+    StackOffset = RoundUpToAlignment(StackOffset, Align);
     unsigned Result = StackOffset;
     StackOffset += Size;
+    MaxStackArgAlign = std::max(Align, MaxStackArgAlign);
     MF.getFrameInfo()->ensureMaxAlignment(Align);
     return Result;
   }

llvm/lib/CodeGen/CallingConvLower.cpp

@@ -32,6 +32,7 @@ CCState::CCState(CallingConv::ID CC, bool isVarArg, MachineFunction &mf,
       CallOrPrologue(Unknown) {
   // No stack is used.
   StackOffset = 0;
+  MaxStackArgAlign = 1;
 
   clearByValRegsInfo();
   UsedRegs.resize((TRI.getNumRegs()+31)/32);
@@ -192,6 +193,7 @@ static bool isValueTypeInRegForCC(CallingConv::ID CC, MVT VT) {
 void CCState::getRemainingRegParmsForType(SmallVectorImpl<MCPhysReg> &Regs,
                                           MVT VT, CCAssignFn Fn) {
   unsigned SavedStackOffset = StackOffset;
+  unsigned SavedMaxStackArgAlign = MaxStackArgAlign;
   unsigned NumLocs = Locs.size();
 
   // Set the 'inreg' flag if it is used for this calling convention.
@@ -223,6 +225,7 @@ void CCState::getRemainingRegParmsForType(SmallVectorImpl<MCPhysReg> &Regs,
   // as allocated so that future queries don't return the same registers, i.e.
   // when i64 and f64 are both passed in GPRs.
   StackOffset = SavedStackOffset;
+  MaxStackArgAlign = SavedMaxStackArgAlign;
   Locs.resize(NumLocs);
 }
 

llvm/lib/Target/X86/X86FastISel.cpp

@@ -2906,7 +2906,7 @@ bool X86FastISel::fastLowerCall(CallLoweringInfo &CLI) {
   CCInfo.AnalyzeCallOperands(OutVTs, OutFlags, CC_X86);
 
   // Get a count of how many bytes are to be pushed on the stack.
-  unsigned NumBytes = CCInfo.getNextStackOffset();
+  unsigned NumBytes = CCInfo.getAlignedCallFrameSize();
 
   // Issue CALLSEQ_START
   unsigned AdjStackDown = TII.getCallFrameSetupOpcode();

llvm/lib/Target/X86/X86ISelLowering.cpp

@@ -3019,7 +3019,7 @@ X86TargetLowering::LowerCall(TargetLowering::CallLoweringInfo &CLI,
   CCInfo.AnalyzeCallOperands(Outs, CC_X86);
 
   // Get a count of how many bytes are to be pushed on the stack.
-  unsigned NumBytes = CCInfo.getNextStackOffset();
+  unsigned NumBytes = CCInfo.getAlignedCallFrameSize();
   if (IsSibcall)
     // This is a sibcall. The memory operands are available in caller's
     // own caller's stack.

llvm/test/CodeGen/X86/win32-spill-xmm.ll

@@ -0,0 +1,40 @@
+; RUN: llc -mcpu=generic -mtriple=i686-pc-windows-msvc -mattr=+sse < %s | FileCheck %s
+
+; Check proper alignment of spilled vector
+
+; CHECK-LABEL: spill_ok
+; CHECK: subl    $32, %esp
+; CHECK: movaps  %xmm3, (%esp)
+; CHECK: movl    $0, 16(%esp)
+; CHECK: calll   _bar
+define void @spill_ok(i32, <16 x float> *) {
+entry:
+  %2 = alloca i32, i32 %0
+  %3 = load <16 x float>, <16 x float> * %1, align 64
+  tail call void @bar(<16 x float> %3, i32 0) nounwind
+  ret void
+}
+
+declare void @bar(<16 x float> %a, i32 %b)
+
+; Check that proper alignment of spilled vector does not affect vargs
+
+; CHECK-LABEL: vargs_not_affected
+; CHECK: leal    28(%ebp), %eax
+define i32 @vargs_not_affected(<4 x float> %v, i8* %f, ...) {
+entry:
+  %ap = alloca i8*, align 4
+  %0 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %0)
+  %argp.cur = load i8*, i8** %ap, align 4
+  %argp.next = getelementptr inbounds i8, i8* %argp.cur, i32 4
+  store i8* %argp.next, i8** %ap, align 4
+  %1 = bitcast i8* %argp.cur to i32*
+  %2 = load i32, i32* %1, align 4
+  call void @llvm.va_end(i8* %0)
+  ret i32 %2
+}
+
+declare void @llvm.va_start(i8*)
+
+declare void @llvm.va_end(i8*)