#define DEBUG_TYPE "jit"
#include "X86JITInfo.h"
#include "X86Relocations.h"
+#include "X86Subtarget.h"
#include "llvm/CodeGen/MachineCodeEmitter.h"
#include "llvm/Config/alloca.h"
#include <cstdlib>
-#include <iostream>
using namespace llvm;
#ifdef _MSC_VER
ASMPREFIX "X86CompilationCallback:\n"
"pushl %ebp\n"
"movl %esp, %ebp\n" // Standard prologue
-#if FASTCC_NUM_INT_ARGS_INREGS > 0
"pushl %eax\n"
- "pushl %edx\n" // Save EAX/EDX
-#endif
+ "pushl %edx\n" // Save EAX/EDX/ECX
+ "pushl %ecx\n"
#if defined(__APPLE__)
"andl $-16, %esp\n" // Align ESP on 16-byte boundary
#endif
"movl %ebp, (%esp)\n"
"call " ASMPREFIX "X86CompilationCallback2\n"
"movl %ebp, %esp\n" // Restore ESP
-#if FASTCC_NUM_INT_ARGS_INREGS > 0
- "subl $8, %esp\n"
+ "subl $12, %esp\n"
+ "popl %ecx\n"
+ "popl %edx\n"
+ "popl %eax\n"
+ "popl %ebp\n"
+ "ret\n");
+
+ // Same as X86CompilationCallback but also saves XMM argument registers.
+ void X86CompilationCallback_SSE(void);
+ asm(
+ ".text\n"
+ ".align 8\n"
+ ".globl " ASMPREFIX "X86CompilationCallback_SSE\n"
+ ASMPREFIX "X86CompilationCallback_SSE:\n"
+ "pushl %ebp\n"
+ "movl %esp, %ebp\n" // Standard prologue
+ "pushl %eax\n"
+ "pushl %edx\n" // Save EAX/EDX/ECX
+ "pushl %ecx\n"
+ "andl $-16, %esp\n" // Align ESP on 16-byte boundary
+ // Save all XMM arg registers
+ "subl $64, %esp\n"
+ "movaps %xmm0, (%esp)\n"
+ "movaps %xmm1, 16(%esp)\n"
+ "movaps %xmm2, 32(%esp)\n"
+ "movaps %xmm3, 48(%esp)\n"
+ "subl $16, %esp\n"
+ "movl 4(%ebp), %eax\n" // Pass prev frame and return address
+ "movl %eax, 4(%esp)\n"
+ "movl %ebp, (%esp)\n"
+ "call " ASMPREFIX "X86CompilationCallback2\n"
+ "addl $16, %esp\n"
+ "movaps 48(%esp), %xmm3\n"
+ "movaps 32(%esp), %xmm2\n"
+ "movaps 16(%esp), %xmm1\n"
+ "movaps (%esp), %xmm0\n"
+ "movl %ebp, %esp\n" // Restore ESP
+ "subl $12, %esp\n"
+ "popl %ecx\n"
"popl %edx\n"
"popl %eax\n"
-#endif
"popl %ebp\n"
"ret\n");
#else
__asm {
push eax
push edx
+ push ecx
call X86CompilationCallback2
+ pop ecx
pop edx
pop eax
ret
#else // Not an i386 host
void X86CompilationCallback() {
- std::cerr << "Cannot call X86CompilationCallback() on a non-x86 arch!\n";
+ assert(0 && "Cannot call X86CompilationCallback() on a non-x86 arch!\n");
abort();
}
#endif
#ifdef _MSC_VER
extern "C" void X86CompilationCallback2() {
assert(sizeof(size_t) == 4); // FIXME: handle Win64
- unsigned *RetAddrLoc = (unsigned *)_AddressOfReturnAddress();
- RetAddrLoc += 3; // skip over ret addr, edx, eax
- unsigned RetAddr = *RetAddrLoc;
+ intptr_t *RetAddrLoc = (intptr_t *)_AddressOfReturnAddress();
+ RetAddrLoc += 4; // skip over ret addr, edx, eax, ecx
+ intptr_t RetAddr = *RetAddrLoc;
#else
extern "C" void X86CompilationCallback2(intptr_t *StackPtr, intptr_t RetAddr) {
intptr_t *RetAddrLoc = &StackPtr[1];
bool isStub = ((unsigned char*)RetAddr)[0] == 0xCD;
// The call instruction should have pushed the return value onto the stack...
+#ifdef __x86_64__
+ RetAddr--; // Backtrack to the reference itself...
+#else
RetAddr -= 4; // Backtrack to the reference itself...
+#endif
#if 0
- DEBUG(std::cerr << "In callback! Addr=" << (void*)RetAddr
- << " ESP=" << (void*)StackPtr
- << ": Resolving call to function: "
- << TheVM->getFunctionReferencedName((void*)RetAddr) << "\n");
+ DOUT << "In callback! Addr=" << (void*)RetAddr
+ << " ESP=" << (void*)StackPtr
+ << ": Resolving call to function: "
+ << TheVM->getFunctionReferencedName((void*)RetAddr) << "\n";
#endif
// Sanity check to make sure this really is a call instruction.
+#ifdef __x86_64__
+ assert(((unsigned char*)RetAddr)[-2] == 0x41 &&"Not a call instr!");
+ assert(((unsigned char*)RetAddr)[-1] == 0xFF &&"Not a call instr!");
+#else
assert(((unsigned char*)RetAddr)[-1] == 0xE8 &&"Not a call instr!");
+#endif
intptr_t NewVal = (intptr_t)JITCompilerFunction((void*)RetAddr);
// Rewrite the call target... so that we don't end up here every time we
// execute the call.
- *(unsigned *)RetAddr = (unsigned)(NewVal-RetAddr-4);
+#ifdef __x86_64__
+ *(intptr_t *)(RetAddr - 0xa) = NewVal;
+#else
+ *(intptr_t *)RetAddr = (intptr_t)(NewVal-RetAddr-4);
+#endif
if (isStub) {
// If this is a stub, rewrite the call into an unconditional branch
// instruction so that two return addresses are not pushed onto the stack
// when the requested function finally gets called. This also makes the
// 0xCD byte (interrupt) dead, so the marker doesn't effect anything.
+#ifdef __x86_64__
+ ((unsigned char*)RetAddr)[0] = (2 | (4 << 3) | (3 << 6));
+#else
((unsigned char*)RetAddr)[-1] = 0xE9;
+#endif
}
// Change the return address to reexecute the call instruction...
+#ifdef __x86_64__
+ *RetAddrLoc -= 0xd;
+#else
*RetAddrLoc -= 5;
+#endif
}
TargetJITInfo::LazyResolverFn
X86JITInfo::getLazyResolverFunction(JITCompilerFn F) {
JITCompilerFunction = F;
+
+#if (defined(__i386__) || defined(i386) || defined(_M_IX86)) && \
+ !defined(_MSC_VER) && !defined(__x86_64__)
+ unsigned EAX = 0, EBX = 0, ECX = 0, EDX = 0;
+ union {
+ unsigned u[3];
+ char c[12];
+ } text;
+
+ if (!X86::GetCpuIDAndInfo(0, &EAX, text.u+0, text.u+2, text.u+1)) {
+ // FIXME: support for AMD family of processors.
+ if (memcmp(text.c, "GenuineIntel", 12) == 0) {
+ X86::GetCpuIDAndInfo(0x1, &EAX, &EBX, &ECX, &EDX);
+ if ((EDX >> 25) & 0x1)
+ return X86CompilationCallback_SSE;
+ }
+ }
+#endif
+
return X86CompilationCallback;
}
void *X86JITInfo::emitFunctionStub(void *Fn, MachineCodeEmitter &MCE) {
// Note, we cast to intptr_t here to silence a -pedantic warning that
// complains about casting a function pointer to a normal pointer.
- if (Fn != (void*)(intptr_t)X86CompilationCallback) {
- MCE.startFunctionStub(5);
+#if (defined(__i386__) || defined(i386) || defined(_M_IX86)) && \
+ !defined(_MSC_VER) && !defined(__x86_64__)
+ bool NotCC = (Fn != (void*)(intptr_t)X86CompilationCallback &&
+ Fn != (void*)(intptr_t)X86CompilationCallback_SSE);
+#else
+ bool NotCC = Fn != (void*)(intptr_t)X86CompilationCallback;
+#endif
+ if (NotCC) {
+#ifdef __x86_64__
+ MCE.startFunctionStub(13, 4);
+ MCE.emitByte(0x49); // REX prefix
+ MCE.emitByte(0xB8+2); // movabsq r10
+ MCE.emitWordLE(((unsigned *)&Fn)[0]);
+ MCE.emitWordLE(((unsigned *)&Fn)[1]);
+ MCE.emitByte(0x41); // REX prefix
+ MCE.emitByte(0xFF); // jmpq *r10
+ MCE.emitByte(2 | (4 << 3) | (3 << 6));
+#else
+ MCE.startFunctionStub(5, 4);
MCE.emitByte(0xE9);
MCE.emitWordLE((intptr_t)Fn-MCE.getCurrentPCValue()-4);
+#endif
return MCE.finishFunctionStub(0);
}
- MCE.startFunctionStub(6);
+#ifdef __x86_64__
+ MCE.startFunctionStub(14, 4);
+ MCE.emitByte(0x49); // REX prefix
+ MCE.emitByte(0xB8+2); // movabsq r10
+ MCE.emitWordLE(((unsigned *)&Fn)[0]);
+ MCE.emitWordLE(((unsigned *)&Fn)[1]);
+ MCE.emitByte(0x41); // REX prefix
+ MCE.emitByte(0xFF); // callq *r10
+ MCE.emitByte(2 | (2 << 3) | (3 << 6));
+#else
+ MCE.startFunctionStub(6, 4);
MCE.emitByte(0xE8); // Call with 32 bit pc-rel destination...
MCE.emitWordLE((intptr_t)Fn-MCE.getCurrentPCValue()-4);
+#endif
MCE.emitByte(0xCD); // Interrupt - Just a marker identifying the stub!
return MCE.finishFunctionStub(0);
// in memory.
*((unsigned*)RelocPos) += (unsigned)ResultPtr;
break;
+ case X86::reloc_absolute_dword:
+ *((intptr_t*)RelocPos) += ResultPtr;
+ break;
}
}
}
projects / oota-llvm.git / blobdiff