[clang] [compiler-rt] [llvm] [XRay][RISCV] RISCV support for XRay (PR #117368)

[Craig Topper via cfe-commits]

================
@@ -453,11 +475,71 @@ bool RISCVAsmPrinter::runOnMachineFunction(MachineFunction &MF) {
   SetupMachineFunction(MF);
   emitFunctionBody();
 
+  // Emit the XRay table
+  emitXRayTable();
+
   if (EmittedOptionArch)
     RTS.emitDirectiveOptionPop();
   return false;
 }
 
+void RISCVAsmPrinter::LowerPATCHABLE_FUNCTION_ENTER(const MachineInstr *MI) {
+  emitSled(MI, SledKind::FUNCTION_ENTER);
+}
+
+void RISCVAsmPrinter::LowerPATCHABLE_FUNCTION_EXIT(const MachineInstr *MI) {
+  emitSled(MI, SledKind::FUNCTION_EXIT);
+}
+
+void RISCVAsmPrinter::LowerPATCHABLE_TAIL_CALL(const MachineInstr *MI) {
+  emitSled(MI, SledKind::TAIL_CALL);
+}
+
+void RISCVAsmPrinter::emitSled(const MachineInstr *MI, SledKind Kind) {
+  // We want to emit the jump instruction and the nops constituting the sled.
+  // The format is as follows:
+  // .Lxray_sled_N
+  //   ALIGN
+  //   J .tmpN
+  //   25 or 33 C.NOP instructions
+  // .tmpN
+
+  // The following variable holds the count of the number of NOPs to be patched
+  // in for XRay instrumentation during compilation.
+  // Note that RV64 and RV32 each has a sled of 68 and 52 bytes, respectively.
+  // Assuming we're using JAL to jump to .tmpN, then we only need
+  // (68 - 4)/2 = 32 NOPs for RV64 and (52 - 4)/2 = 24 for RV32. However, there
+  // is a chance that we'll use C.JAL instead, so an additional NOP is needed.
+  const uint8_t NoopsInSledCount =
+      MI->getParent()->getParent()->getSubtarget<RISCVSubtarget>().is64Bit()
+          ? 33
+          : 25;
+
+  OutStreamer->emitCodeAlignment(Align(4), &getSubtargetInfo());
+  auto CurSled = OutContext.createTempSymbol("xray_sled_", true);
+  OutStreamer->emitLabel(CurSled);
+  auto Target = OutContext.createTempSymbol();
+
+  const MCExpr *TargetExpr = MCSymbolRefExpr::create(
+      Target, MCSymbolRefExpr::VariantKind::VK_None, OutContext);
+
+  // Emit "J bytes" instruction, which jumps over the nop sled to the actual
+  // start of function.
+  EmitToStreamer(
+      *OutStreamer,
+      MCInstBuilder(RISCV::JAL).addReg(RISCV::X0).addExpr(TargetExpr));
+
+  // Emit NOP instructions
+  for (int8_t I = 0; I < NoopsInSledCount; ++I)
----------------
topperc wrote:

If compressed instructions aren't enabled this emits too many bytes doesn't it?

https://github.com/llvm/llvm-project/pull/117368