[XRay][x86_64] Define a tail exit trampoline.

Summary:
We define a new trampoline that's a hybrid between the exit and entry
trampolines with the following properties:

- Saves all of the callee-saved registers according to the x86_64
  calling conventions.
- Indicate to the log handler function being called that this is a
  function exit event.

This fixes a bug that is a result of not saving enough of the register
states, and that the log handler is clobbering registers that would be
used by the function being tail-exited into manifesting as runtime
errors.

Reviewers: rSerge, echristo, majnemer

Subscribers: mehdi_amini, llvm-commits

Differential Revision: https://reviews.llvm.org/D26020

llvm-svn: 285787

compiler-rt/lib/xray/xray_interface_internal.h

@@ -61,6 +61,7 @@ extern "C" {
 // basis. See xray_trampoline_*.S files for implementations.
 extern void __xray_FunctionEntry();
 extern void __xray_FunctionExit();
+extern void __xray_FunctionTailExit();
 }
 
 #endif

compiler-rt/lib/xray/xray_trampoline_x86_64.S

@@ -110,3 +110,63 @@ __xray_FunctionExit:
 .Ltmp3:
 	.size __xray_FunctionExit, .Ltmp3-__xray_FunctionExit
 	.cfi_endproc
+
+	.global __xray_FunctionTailExit
+	.align 16, 0x90
+	.type __xray_FunctionTailExit,@function
+__xray_FunctionTailExit:
+	.cfi_startproc
+	// Save the important registers as in the entry trampoline, but indicate that
+	// this is an exit. In the future, we will introduce a new entry type that
+	// differentiates between a normal exit and a tail exit, but we'd have to do
+	// this and increment the version number for the header.
+	pushq %rbp
+	.cfi_def_cfa_offset 16
+	subq $200, %rsp
+	movupd	%xmm0, 184(%rsp)
+	movupd	%xmm1, 168(%rsp)
+	movupd	%xmm2, 152(%rsp)
+	movupd	%xmm3, 136(%rsp)
+	movupd	%xmm4, 120(%rsp)
+	movupd	%xmm5, 104(%rsp)
+	movupd	%xmm6, 88(%rsp)
+	movupd	%xmm7, 72(%rsp)
+	movq	%rdi, 64(%rsp)
+	movq  %rax, 56(%rsp)
+	movq  %rdx, 48(%rsp)
+	movq	%rsi, 40(%rsp)
+	movq	%rcx, 32(%rsp)
+	movq	%r8, 24(%rsp)
+	movq	%r9, 16(%rsp)
+
+	movq	_ZN6__xray19XRayPatchedFunctionE(%rip), %rax
+	testq %rax,%rax
+	je	.Ltmp4
+
+	movl	%r10d, %edi
+	movl	$1, %esi
+	callq	*%rax
+
+.Ltmp4:
+  // Restore the registers.
+	movupd	184(%rsp), %xmm0
+	movupd	168(%rsp), %xmm1
+	movupd	152(%rsp), %xmm2
+	movupd	136(%rsp), %xmm3
+	movupd	120(%rsp), %xmm4
+	movupd	104(%rsp), %xmm5
+	movupd	88(%rsp) , %xmm6
+	movupd	72(%rsp) , %xmm7
+	movq	64(%rsp), %rdi
+	movq  56(%rsp), %rax
+	movq  48(%rsp), %rdx
+	movq	40(%rsp), %rsi
+	movq	32(%rsp), %rcx
+	movq	24(%rsp), %r8
+	movq	16(%rsp), %r9
+	addq	$200, %rsp
+	popq	%rbp
+	retq
+.Ltmp5:
+  .size __xray_FunctionTailExit, .Ltmp5-__xray_FunctionTailExit
+	.cfi_endproc

compiler-rt/lib/xray/xray_x86_64.cc

@@ -114,12 +114,9 @@ bool patchFunctionExit(const bool Enable, const uint32_t FuncId,
 bool patchFunctionTailExit(const bool Enable, const uint32_t FuncId,
                            const XRaySledEntry &Sled) {
   // Here we do the dance of replacing the tail call sled with a similar
-  // sequence as the entry sled, but calls the exit sled instead, so we can
-  // treat tail call exits as if they were normal exits.
-  //
-  // FIXME: In the future we'd need to distinguish between non-tail exits and
-  // tail exits for better information preservation.
-  int64_t TrampolineOffset = reinterpret_cast<int64_t>(__xray_FunctionExit) -
+  // sequence as the entry sled, but calls the tail exit sled instead.
+  int64_t TrampolineOffset =
+      reinterpret_cast<int64_t>(__xray_FunctionTailExit) -
       (static_cast<int64_t>(Sled.Address) + 11);
   if (TrampolineOffset < MinOffset || TrampolineOffset > MaxOffset) {
     Report("XRay Exit trampoline (%p) too far from sled (%p); distance = "