ARM: fix WoA PEI instruction selection

The ARM::BLX instruction is an ARM mode instruction.  The Windows on ARM target
is limited to Thumb instructions.  Correctly use the thumb mode tBLXr
instruction.  This would manifest as an errant write into the object file as the
instruction is 4-bytes in length rather than 2.  The result would be a corrupted
object file that would eventually result in an executable that would crash at
runtime.

llvm-svn: 208152

llvm/lib/Target/ARM/ARMFrameLowering.cpp

@@ -319,7 +319,8 @@ void ARMFrameLowering::emitPrologue(MachineFunction &MF) const {
       BuildMI(MBB, MBBI, dl, TII.get(ARM::t2MOVi32imm), ARM::R12)
         .addExternalSymbol("__chkstk");
 
-      BuildMI(MBB, MBBI, dl, TII.get(ARM::BLX))
+      BuildMI(MBB, MBBI, dl, TII.get(ARM::tBLXr))
+        .addImm((unsigned)ARMCC::AL).addReg(0)
         .addReg(ARM::R12, RegState::Kill)
         .addReg(ARM::R4, RegState::Implicit);
       break;

llvm/test/CodeGen/ARM/Windows/chkstk-movw-movt-isel.ll

@@ -0,0 +1,27 @@
+; RUN: llc -mtriple thumbv7--windows-itanium -code-model large -filetype obj -o - %s \
+; RUN:    | llvm-objdump -no-show-raw-insn -d - | FileCheck %s
+
+; ModuleID = 'reduced.c'
+target datalayout = "e-m:e-p:32:32-i1:8:32-i8:8:32-i16:16:32-i64:64-v128:64:128-a:0:32-n32-S64"
+target triple = "thumbv7--windows-itanium"
+
+define arm_aapcs_vfpcc i8 @isel(i32 %i) {
+entry:
+  %i.addr = alloca i32, align 4
+  %buffer = alloca [4096 x i8], align 1
+  store i32 %i, i32* %i.addr, align 4
+  %0 = load i32* %i.addr, align 4
+  %rem = urem i32 %0, 4096
+  %arrayidx = getelementptr inbounds [4096 x i8]* %buffer, i32 0, i32 %rem
+  %1 = load volatile i8* %arrayidx, align 1
+  ret i8 %1
+}
+
+; CHECK-LABEL: isel
+; CHECK: push {r4, r5}
+; CHECK: movw r4, #{{\d*}}
+; CHECK: movw r12, #0
+; CHECK: movt r12, #0
+; CHECK: blx r12
+; CHECK: sub.w sp, sp, r4
+