x86: movhps support (#1444)

* x86: MOVHPS support

Closes #1432.

* Update manticore/native/cpu/x86.py

Co-Authored-By: Eric Hennenfent <ecapstone@gmail.com>

* tests: Add movhps tests

* tests: Fix tests

* tests: Remove relative import

Works locally, fails on CI.

* tests: Fix movhps shellcode

Was using qword and not qword ptr.

* tests: Swap assertions

Author: woodruffw

manticore/native/cpu/x86.py

@@ -5507,6 +5507,27 @@ def MOVHPD(cpu, dest, src):
             value = Operators.EXTRACT(dest.read(), 0, 64)  # low part
             dest.write(Operators.CONCAT(128, src.read(), value))
 
+    @instruction
+    def MOVHPS(cpu, dest, src):
+        """
+        Moves high packed single-precision floating-point value.
+
+        Moves two packed single-precision floating-point values from the source operand
+        (second operand) to the destination operand (first operand). The source and destination
+        operands can be an XMM register or a 64-bit memory location. The instruction allows
+        single-precision floating-point values to be moved to and from the high quadword of
+        an XMM register and memory. It cannot be used for register to register or memory to
+        memory moves. When the destination operand is an XMM register, the low quadword
+        of the register remains unchanged.
+        """
+        if src.size == 128:
+            assert dest.size == 64
+            dest.write(Operators.EXTRACT(src.read(), 64, 64))
+        else:
+            assert src.size == 64 and dest.size == 128
+            value = Operators.EXTRACT(dest.read(), 0, 64)  # low part
+            dest.write(Operators.CONCAT(128, src.read(), value))
+
     @instruction
     def PSUBB(cpu, dest, src):
         """

tests/native/test_cpu_manual.py

@@ -1128,6 +1128,53 @@ def test_SAR_3_symbolic(self):
             temp_cs.add(condition == False)
             self.assertFalse(solver.check(temp_cs))
 
+    def test_MOVHPS_1(self):
+        mem = Memory32()
+        cpu = I386Cpu(mem)
+        mem.mmap(0x0041e000, 0x1000, 'rwx')
+
+        # 3.14
+        mem[0x0041e000] = '\x40'
+        mem[0x0041e001] = '\x48'
+        mem[0x0041e002] = '\xf5'
+        mem[0x0041e003] = '\xc3'
+
+        # 6.28
+        mem[0x0041e004] = '\x40'
+        mem[0x0041e005] = '\xc8'
+        mem[0x0041e006] = '\xf5'
+        mem[0x0041e007] = '\xc3'
+
+        # movhps xmm0, qword ptr [eax]
+        mem[0x0041e10a] = '\x0f'
+        mem[0x0041e10b] = '\x16'
+        mem[0x0041e10c] = '\x00'
+
+        cpu.RIP = 0x41e10a
+        cpu.EAX = 0x41e000
+        cpu.XMM0 = 0x0000000000000000ffffffffffffffff
+        cpu.execute()
+
+        self.assertEqual(cpu.XMM0, 0xc3f5c840c3f54840ffffffffffffffff)
+
+    def test_MOVHPS_2(self):
+        mem = Memory32()
+        cpu = I386Cpu(mem)
+        mem.mmap(0x0041e000, 0x1000, 'rwx')
+
+        # movhps qword ptr [eax], xmm1
+        mem[0x0041e10a] = '\x0f'
+        mem[0x0041e10b] = '\x17'
+        mem[0x0041e10c] = '\x08'
+
+        cpu.RIP = 0x41e10a
+        cpu.EAX = 0x41e000
+        cpu.XMM1 = 0x4048f5c340c8f5c3ffffffffffffffff
+        cpu.execute()
+
+        self.assertItemsEqual(mem[0x41e000:0x41e004], to_bytelist(b'\x40\xc8\xf5\xc3'))
+        self.assertItemsEqual(mem[0x41e004:0x41e008], to_bytelist(b'\x40\x48\xf5\xc3'))
+
     def test_symbolic_instruction(self):
         cs = ConstraintSet()
         mem = SMemory32(cs)