[llvm] [bolt][aarch64] Change indirect call instrumentation snippet (PR #180229)

[YongKang Zhu via llvm-commits]

================
@@ -2625,39 +2626,62 @@ class AArch64MCPlusBuilder : public MCPlusBuilder {
                                                      MCSymbol *HandlerFuncAddr,
                                                      int CallSiteID,
                                                      MCContext *Ctx) override {
-    InstructionListType Insts;
     // Code sequence used to enter indirect call instrumentation helper:
-    //   stp x0, x1, [sp, #-16]! createPushRegisters
-    //   mov target x0  convertIndirectCallToLoad -> orr x0 target xzr
+    //   stp x0, x1, [sp, #-16]! createPushRegisters  (1)
+    //   mov target, x0  convertIndirectCallToLoad -> orr x0 target xzr
     //   mov x1 CallSiteID createLoadImmediate ->
     //   movk    x1, #0x0, lsl #48
     //   movk    x1, #0x0, lsl #32
     //   movk    x1, #0x0, lsl #16
     //   movk    x1, #0x0
-    //   stp x0, x1, [sp, #-16]!
-    //   bl *HandlerFuncAddr createIndirectCall ->
+    //   stp x0, x1, [sp, #-16]!    (2)
+    //   str x30, [sp, #-16]!    (3)
     //   adr x0 *HandlerFuncAddr -> adrp + add
-    //   blr x0
+    //   blr x0   (__bolt_instr_ind_call_handler_func)
+    //   ldr x30, [sp], #16   (3)
+    //   ldr target ; restore target value (2)
+    //   ldp x0, x1, [sp], #16   (1)
+    //   blr target
+
+    InstructionListType Insts;
     Insts.emplace_back();
-    createPushRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+    createPushRegisters(Insts.back(), getIntArgRegister(0),
+                        getIntArgRegister(1));
     Insts.emplace_back(CallInst);
-    convertIndirectCallToLoad(Insts.back(), AArch64::X0);
+    convertIndirectCallToLoad(Insts.back(), getIntArgRegister(0));
     InstructionListType LoadImm =
         createLoadImmediate(getIntArgRegister(1), CallSiteID);
     Insts.insert(Insts.end(), LoadImm.begin(), LoadImm.end());
     Insts.emplace_back();
-    createPushRegisters(Insts.back(), AArch64::X0, AArch64::X1);
+    createPushRegisters(Insts.back(), getIntArgRegister(0),
+                        getIntArgRegister(1));
+    Insts.emplace_back();
+    storeReg(Insts.back(), AArch64::LR, getSpRegister(/*Size*/ 8));
     Insts.resize(Insts.size() + 2);
-    InstructionListType Addr =
-        materializeAddress(HandlerFuncAddr, Ctx, AArch64::X0);
+    InstructionListType Addr = materializeAddress(
+        HandlerFuncAddr, Ctx, CallInst.getOperand(0).getReg());
----------------
yozhu wrote:

The comment above says `x0` will be used to hold the handler function address, while here we use the register in the original indirect call instruction (`CallInst`).

https://github.com/llvm/llvm-project/pull/180229