//
// The LLVM Compiler Infrastructure
//
-// This file was developed by the LLVM research group and is distributed under
-// the University of Illinois Open Source License. See LICENSE.TXT for details.
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
#define DEBUG_TYPE "jit"
#include "PPCJITInfo.h"
#include "PPCRelocations.h"
-#include "llvm/CodeGen/MachineCodeEmitter.h"
-#include "llvm/Config/alloca.h"
+#include "PPCTargetMachine.h"
+#include "llvm/Function.h"
+#include "llvm/System/Memory.h"
#include "llvm/Support/Debug.h"
-#include <set>
+#include "llvm/Support/ErrorHandling.h"
+#include "llvm/Support/raw_ostream.h"
using namespace llvm;
static TargetJITInfo::JITCompilerFn JITCompilerFunction;
extern "C" void PPC64CompilationCallback();
#if (defined(__POWERPC__) || defined (__ppc__) || defined(_POWER)) && \
- !defined(__ppc64__)
+ !(defined(__ppc64__) || defined(__FreeBSD__))
// CompilationCallback stub - We can't use a C function with inline assembly in
// it, because we the prolog/epilog inserted by GCC won't work for us. Instead,
// write our own wrapper, which does things our way, so we have complete control
// FIXME: need to save v[0-19] for altivec?
// FIXME: could shrink frame
// Set up a proper stack frame
+ // FIXME Layout
+ // PowerPC64 ABI linkage - 24 bytes
+ // parameters - 32 bytes
+ // 13 double registers - 104 bytes
+ // 8 int registers - 32 bytes
"mflr r0\n"
"stw r0, 8(r1)\n"
"stwu r1, -208(r1)\n"
"mtlr r2\n"
"bctr\n"
);
+
+#elif defined(__PPC__) && !defined(__ppc64__)
+// Linux & FreeBSD / PPC 32 support
+
+// CompilationCallback stub - We can't use a C function with inline assembly in
+// it, because we the prolog/epilog inserted by GCC won't work for us. Instead,
+// write our own wrapper, which does things our way, so we have complete control
+// over register saving and restoring.
+asm(
+ ".text\n"
+ ".align 2\n"
+ ".globl PPC32CompilationCallback\n"
+"PPC32CompilationCallback:\n"
+ // Make space for 8 ints r[3-10] and 8 doubles f[1-8] and the
+ // FIXME: need to save v[0-19] for altivec?
+ // FIXME: could shrink frame
+ // Set up a proper stack frame
+ // FIXME Layout
+ // 8 double registers - 64 bytes
+ // 8 int registers - 32 bytes
+ "mflr 0\n"
+ "stw 0, 4(1)\n"
+ "stwu 1, -104(1)\n"
+ // Save all int arg registers
+ "stw 10, 100(1)\n" "stw 9, 96(1)\n"
+ "stw 8, 92(1)\n" "stw 7, 88(1)\n"
+ "stw 6, 84(1)\n" "stw 5, 80(1)\n"
+ "stw 4, 76(1)\n" "stw 3, 72(1)\n"
+ // Save all call-clobbered FP regs.
+ "stfd 8, 64(1)\n"
+ "stfd 7, 56(1)\n" "stfd 6, 48(1)\n"
+ "stfd 5, 40(1)\n" "stfd 4, 32(1)\n"
+ "stfd 3, 24(1)\n" "stfd 2, 16(1)\n"
+ "stfd 1, 8(1)\n"
+ // Arguments to Compilation Callback:
+ // r3 - our lr (address of the call instruction in stub plus 4)
+ // r4 - stub's lr (address of instruction that called the stub plus 4)
+ // r5 - is64Bit - always 0.
+ "mr 3, 0\n"
+ "lwz 5, 104(1)\n" // stub's frame
+ "lwz 4, 4(5)\n" // stub's lr
+ "li 5, 0\n" // 0 == 32 bit
+ "bl PPCCompilationCallbackC\n"
+ "mtctr 3\n"
+ // Restore all int arg registers
+ "lwz 10, 100(1)\n" "lwz 9, 96(1)\n"
+ "lwz 8, 92(1)\n" "lwz 7, 88(1)\n"
+ "lwz 6, 84(1)\n" "lwz 5, 80(1)\n"
+ "lwz 4, 76(1)\n" "lwz 3, 72(1)\n"
+ // Restore all FP arg registers
+ "lfd 8, 64(1)\n"
+ "lfd 7, 56(1)\n" "lfd 6, 48(1)\n"
+ "lfd 5, 40(1)\n" "lfd 4, 32(1)\n"
+ "lfd 3, 24(1)\n" "lfd 2, 16(1)\n"
+ "lfd 1, 8(1)\n"
+ // Pop 3 frames off the stack and branch to target
+ "lwz 1, 104(1)\n"
+ "lwz 0, 4(1)\n"
+ "mtlr 0\n"
+ "bctr\n"
+ );
#else
void PPC32CompilationCallback() {
- assert(0 && "This is not a power pc, you can't execute this!");
- abort();
+ llvm_unreachable("This is not a power pc, you can't execute this!");
}
#endif
);
#else
void PPC64CompilationCallback() {
- assert(0 && "This is not a power pc, you can't execute this!");
- abort();
+ llvm_unreachable("This is not a power pc, you can't execute this!");
}
#endif
// Rewrite the stub with an unconditional branch to the target, for any users
// who took the address of the stub.
EmitBranchToAt((intptr_t)StubCallAddr, (intptr_t)Target, false, is64Bit);
+ sys::Memory::InvalidateInstructionCache(StubCallAddr, 7*4);
// Put the address of the target function to call and the address to return to
// after calling the target function in a place that is easy to get on the
return is64Bit ? PPC64CompilationCallback : PPC32CompilationCallback;
}
-void *PPCJITInfo::emitFunctionStub(void *Fn, MachineCodeEmitter &MCE) {
+TargetJITInfo::StubLayout PPCJITInfo::getStubLayout() {
+ // The stub contains up to 10 4-byte instructions, aligned at 4 bytes: 3
+ // instructions to save the caller's address if this is a lazy-compilation
+ // stub, plus a 1-, 4-, or 7-instruction sequence to load an arbitrary address
+ // into a register and jump through it.
+ StubLayout Result = {10*4, 4};
+ return Result;
+}
+
+#if (defined(__POWERPC__) || defined (__ppc__) || defined(_POWER)) && \
+defined(__APPLE__)
+extern "C" void sys_icache_invalidate(const void *Addr, size_t len);
+#endif
+
+void *PPCJITInfo::emitFunctionStub(const Function* F, void *Fn,
+ JITCodeEmitter &JCE) {
// If this is just a call to an external function, emit a branch instead of a
// call. The code is the same except for one bit of the last instruction.
if (Fn != (void*)(intptr_t)PPC32CompilationCallback &&
Fn != (void*)(intptr_t)PPC64CompilationCallback) {
- MCE.startFunctionStub(7*4);
- intptr_t Addr = (intptr_t)MCE.getCurrentPCValue();
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- EmitBranchToAt(Addr, (intptr_t)Fn, false, is64Bit);
- return MCE.finishFunctionStub(0);
+ void *Addr = (void*)JCE.getCurrentPCValue();
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ EmitBranchToAt((intptr_t)Addr, (intptr_t)Fn, false, is64Bit);
+ sys::Memory::InvalidateInstructionCache(Addr, 7*4);
+ return Addr;
}
- MCE.startFunctionStub(10*4);
+ void *Addr = (void*)JCE.getCurrentPCValue();
if (is64Bit) {
- MCE.emitWordBE(0xf821ffb1); // stdu r1,-80(r1)
- MCE.emitWordBE(0x7d6802a6); // mflr r11
- MCE.emitWordBE(0xf9610060); // std r11, 96(r1)
+ JCE.emitWordBE(0xf821ffb1); // stdu r1,-80(r1)
+ JCE.emitWordBE(0x7d6802a6); // mflr r11
+ JCE.emitWordBE(0xf9610060); // std r11, 96(r1)
+ } else if (TM.getSubtargetImpl()->isDarwinABI()){
+ JCE.emitWordBE(0x9421ffe0); // stwu r1,-32(r1)
+ JCE.emitWordBE(0x7d6802a6); // mflr r11
+ JCE.emitWordBE(0x91610028); // stw r11, 40(r1)
} else {
- MCE.emitWordBE(0x9421ffe0); // stwu r1,-32(r1)
- MCE.emitWordBE(0x7d6802a6); // mflr r11
- MCE.emitWordBE(0x91610028); // stw r11, 40(r1)
+ JCE.emitWordBE(0x9421ffe0); // stwu r1,-32(r1)
+ JCE.emitWordBE(0x7d6802a6); // mflr r11
+ JCE.emitWordBE(0x91610024); // stw r11, 36(r1)
}
- intptr_t Addr = (intptr_t)MCE.getCurrentPCValue();
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- MCE.emitWordBE(0);
- EmitBranchToAt(Addr, (intptr_t)Fn, true, is64Bit);
- return MCE.finishFunctionStub(0);
+ intptr_t BranchAddr = (intptr_t)JCE.getCurrentPCValue();
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ JCE.emitWordBE(0);
+ EmitBranchToAt(BranchAddr, (intptr_t)Fn, true, is64Bit);
+ sys::Memory::InvalidateInstructionCache(Addr, 10*4);
+ return Addr;
}
unsigned *RelocPos = (unsigned*)Function + MR->getMachineCodeOffset()/4;
intptr_t ResultPtr = (intptr_t)MR->getResultPointer();
switch ((PPC::RelocationType)MR->getRelocationType()) {
- default: assert(0 && "Unknown relocation type!");
+ default: llvm_unreachable("Unknown relocation type!");
case PPC::reloc_pcrel_bx:
// PC-relative relocation for b and bl instructions.
ResultPtr = (ResultPtr-(intptr_t)RelocPos) >> 2;
void PPCJITInfo::replaceMachineCodeForFunction(void *Old, void *New) {
EmitBranchToAt((intptr_t)Old, (intptr_t)New, false, is64Bit);
+ sys::Memory::InvalidateInstructionCache(Old, 7*4);
}