Sort returned variables to make sure `overflow` is last
[soc.git] / src / soc / decoder / isa / caller.py
index 57e7825254b611650705487d7592f8428156f0f8..bf84e05a4f6e5e4f383e89f51c3c7a2852f0a43e 100644 (file)
@@ -1,14 +1,30 @@
+"""core of the python-based POWER9 simulator
+
+this is part of a cycle-accurate POWER9 simulator.  its primary purpose is
+not speed, it is for both learning and educational purposes, as well as
+a method of verifying the HDL.
+
+related bugs:
+
+* https://bugs.libre-soc.org/show_bug.cgi?id=424
+"""
+
 from functools import wraps
+from copy import copy
 from soc.decoder.orderedset import OrderedSet
 from soc.decoder.selectable_int import (FieldSelectableInt, SelectableInt,
                                         selectconcat)
-from soc.decoder.power_enums import spr_dict, XER_bits
-from soc.decoder.helpers import exts
+from soc.decoder.power_enums import (spr_dict, spr_byname, XER_bits,
+                                     insns, MicrOp)
+from soc.decoder.helpers import exts, gtu, ltu
+from soc.consts import PIb, MSRb  # big-endian (PowerISA versions)
+
 from collections import namedtuple
 import math
+import sys
 
 instruction_info = namedtuple('instruction_info',
-                              'func read_regs uninit_regs write_regs ' + \
+                              'func read_regs uninit_regs write_regs ' +
                               'special_regs op_fields form asmregs')
 
 special_sprs = {
@@ -19,71 +35,129 @@ special_sprs = {
     'VRSAVE': 256}
 
 
+def swap_order(x, nbytes):
+    x = x.to_bytes(nbytes, byteorder='little')
+    x = int.from_bytes(x, byteorder='big', signed=False)
+    return x
+
+
+REG_SORT_ORDER = {
+    # TODO (lkcl): adjust other registers that should be in a particular order
+    # probably CA, CA32, and CR
+    "RT": 0,
+    "RA": 0,
+    "RB": 0,
+    "RS": 0,
+    "CR": 0,
+    "LR": 0,
+    "CTR": 0,
+    "TAR": 0,
+    "CA": 0,
+    "CA32": 0,
+    "MSR": 0,
+
+    "overflow": 1,
+}
+
+
 def create_args(reglist, extra=None):
-    args = OrderedSet()
-    for reg in reglist:
-        args.add(reg)
-    args = list(args)
-    if extra:
-        args = [extra] + args
-    return args
+    regset = OrderedSet(reglist)
+    retval = []
+    for reg in regset:
+        retval.append(reg)
+    retval.sort(key=lambda reg: REG_SORT_ORDER[reg])
+    if extra is not None:
+        return [extra] + retval
+    return retval
 
 
 class Mem:
 
-    def __init__(self, bytes_per_word=8):
+    def __init__(self, row_bytes=8, initial_mem=None):
         self.mem = {}
-        self.bytes_per_word = bytes_per_word
-        self.word_log2 = math.ceil(math.log2(bytes_per_word))
-
-    def _get_shifter_mask(self, width, remainder):
-        shifter = ((self.bytes_per_word - width) - remainder) * \
+        self.bytes_per_word = row_bytes
+        self.word_log2 = math.ceil(math.log2(row_bytes))
+        print("Sim-Mem", initial_mem, self.bytes_per_word, self.word_log2)
+        if not initial_mem:
+            return
+
+        # different types of memory data structures recognised (for convenience)
+        if isinstance(initial_mem, list):
+            initial_mem = (0, initial_mem)
+        if isinstance(initial_mem, tuple):
+            startaddr, mem = initial_mem
+            initial_mem = {}
+            for i, val in enumerate(mem):
+                initial_mem[startaddr + row_bytes*i] = (val, row_bytes)
+
+        for addr, (val, width) in initial_mem.items():
+            #val = swap_order(val, width)
+            self.st(addr, val, width, swap=False)
+
+    def _get_shifter_mask(self, wid, remainder):
+        shifter = ((self.bytes_per_word - wid) - remainder) * \
             8  # bits per byte
-        mask = (1 << (width * 8)) - 1
+        # XXX https://bugs.libre-soc.org/show_bug.cgi?id=377
+        # BE/LE mode?
+        shifter = remainder * 8
+        mask = (1 << (wid * 8)) - 1
+        print("width,rem,shift,mask", wid, remainder, hex(shifter), hex(mask))
         return shifter, mask
 
     # TODO: Implement ld/st of lesser width
-    def ld(self, address, width=8):
+    def ld(self, address, width=8, swap=True, check_in_mem=False):
+        print("ld from addr 0x{:x} width {:d}".format(address, width))
         remainder = address & (self.bytes_per_word - 1)
         address = address >> self.word_log2
         assert remainder & (width - 1) == 0, "Unaligned access unsupported!"
         if address in self.mem:
             val = self.mem[address]
+        elif check_in_mem:
+            return None
         else:
             val = 0
+        print("mem @ 0x{:x} rem {:d} : 0x{:x}".format(address, remainder, val))
 
         if width != self.bytes_per_word:
             shifter, mask = self._get_shifter_mask(width, remainder)
+            print("masking", hex(val), hex(mask << shifter), shifter)
             val = val & (mask << shifter)
             val >>= shifter
-        print("Read {:x} from addr {:x}".format(val, address))
+        if swap:
+            val = swap_order(val, width)
+        print("Read 0x{:x} from addr 0x{:x}".format(val, address))
         return val
 
-    def st(self, address, value, width=8):
-        remainder = address & (self.bytes_per_word - 1)
-        address = address >> self.word_log2
+    def st(self, addr, v, width=8, swap=True):
+        staddr = addr
+        remainder = addr & (self.bytes_per_word - 1)
+        addr = addr >> self.word_log2
+        print("Writing 0x{:x} to ST 0x{:x} "
+              "memaddr 0x{:x}/{:x}".format(v, staddr, addr, remainder, swap))
         assert remainder & (width - 1) == 0, "Unaligned access unsupported!"
-        print("Writing {:x} to addr {:x}".format(value, address))
+        if swap:
+            v = swap_order(v, width)
         if width != self.bytes_per_word:
-            if address in self.mem:
-                val = self.mem[address]
+            if addr in self.mem:
+                val = self.mem[addr]
             else:
                 val = 0
             shifter, mask = self._get_shifter_mask(width, remainder)
             val &= ~(mask << shifter)
-            val |= value << shifter
-            self.mem[address] = val
+            val |= v << shifter
+            self.mem[addr] = val
         else:
-            self.mem[address] = value
+            self.mem[addr] = v
+        print("mem @ 0x{:x}: 0x{:x}".format(addr, self.mem[addr]))
 
     def __call__(self, addr, sz):
-        val = self.ld(addr.value, sz)
-        print ("memread", addr, sz, val)
+        val = self.ld(addr.value, sz, swap=False)
+        print("memread", addr, sz, val)
         return SelectableInt(val, sz*8)
 
     def memassign(self, addr, sz, val):
-        print ("memassign", addr, sz, val)
-        self.st(addr.value, val.value, sz)
+        print("memassign", addr, sz, val)
+        self.st(addr.value, val.value, sz, swap=False)
 
 
 class GPR(dict):
@@ -100,7 +174,7 @@ class GPR(dict):
         self.form = form
 
     def getz(self, rnum):
-        #rnum = rnum.value # only SelectableInt allowed
+        # rnum = rnum.value # only SelectableInt allowed
         print("GPR getzero", rnum)
         if rnum == 0:
             return SelectableInt(0, 64)
@@ -124,6 +198,7 @@ class GPR(dict):
             s = ' '.join(s)
             print("reg", "%2d" % i, s)
 
+
 class PC:
     def __init__(self, pc_init=0):
         self.CIA = SelectableInt(pc_init, 64)
@@ -140,38 +215,111 @@ class SPR(dict):
     def __init__(self, dec2, initial_sprs={}):
         self.sd = dec2
         dict.__init__(self)
-        self.update(initial_sprs)
+        for key, v in initial_sprs.items():
+            if isinstance(key, SelectableInt):
+                key = key.value
+            key = special_sprs.get(key, key)
+            if isinstance(key, int):
+                info = spr_dict[key]
+            else:
+                info = spr_byname[key]
+            if not isinstance(v, SelectableInt):
+                v = SelectableInt(v, info.length)
+            self[key] = v
 
     def __getitem__(self, key):
+        print("get spr", key)
+        print("dict", self.items())
         # if key in special_sprs get the special spr, otherwise return key
         if isinstance(key, SelectableInt):
             key = key.value
+        if isinstance(key, int):
+            key = spr_dict[key].SPR
         key = special_sprs.get(key, key)
+        if key == 'HSRR0':  # HACK!
+            key = 'SRR0'
+        if key == 'HSRR1':  # HACK!
+            key = 'SRR1'
         if key in self:
-            return dict.__getitem__(self, key)
+            res = dict.__getitem__(self, key)
         else:
-            info = spr_dict[key]
-            return SelectableInt(0, info.length)
+            if isinstance(key, int):
+                info = spr_dict[key]
+            else:
+                info = spr_byname[key]
+            dict.__setitem__(self, key, SelectableInt(0, info.length))
+            res = dict.__getitem__(self, key)
+        print("spr returning", key, res)
+        return res
 
     def __setitem__(self, key, value):
         if isinstance(key, SelectableInt):
             key = key.value
+        if isinstance(key, int):
+            key = spr_dict[key].SPR
+            print("spr key", key)
         key = special_sprs.get(key, key)
+        if key == 'HSRR0':  # HACK!
+            self.__setitem__('SRR0', value)
+        if key == 'HSRR1':  # HACK!
+            self.__setitem__('SRR1', value)
+        print("setting spr", key, value)
         dict.__setitem__(self, key, value)
 
     def __call__(self, ridx):
         return self[ridx]
-        
-        
+
 
 class ISACaller:
     # decoder2 - an instance of power_decoder2
     # regfile - a list of initial values for the registers
-    def __init__(self, decoder2, regfile, initial_sprs={}, initial_cr=0):
+    # initial_{etc} - initial values for SPRs, Condition Register, Mem, MSR
+    # respect_pc - tracks the program counter.  requires initial_insns
+    def __init__(self, decoder2, regfile, initial_sprs=None, initial_cr=0,
+                 initial_mem=None, initial_msr=0,
+                 initial_insns=None, respect_pc=False,
+                 disassembly=None,
+                 initial_pc=0,
+                 bigendian=False):
+
+        self.bigendian = bigendian
+        self.halted = False
+        self.respect_pc = respect_pc
+        if initial_sprs is None:
+            initial_sprs = {}
+        if initial_mem is None:
+            initial_mem = {}
+        if initial_insns is None:
+            initial_insns = {}
+            assert self.respect_pc == False, "instructions required to honor pc"
+
+        print("ISACaller insns", respect_pc, initial_insns, disassembly)
+        print("ISACaller initial_msr", initial_msr)
+
+        # "fake program counter" mode (for unit testing)
+        self.fake_pc = 0
+        disasm_start = 0
+        if not respect_pc:
+            if isinstance(initial_mem, tuple):
+                self.fake_pc = initial_mem[0]
+                disasm_start = self.fake_pc
+        else:
+            disasm_start = initial_pc
+
+        # disassembly: we need this for now (not given from the decoder)
+        self.disassembly = {}
+        if disassembly:
+            for i, code in enumerate(disassembly):
+                self.disassembly[i*4 + disasm_start] = code
+
+        # set up registers, instruction memory, data memory, PC, SPRs, MSR
         self.gpr = GPR(decoder2, regfile)
-        self.mem = Mem()
+        self.mem = Mem(row_bytes=8, initial_mem=initial_mem)
+        self.imem = Mem(row_bytes=4, initial_mem=initial_insns)
         self.pc = PC()
         self.spr = SPR(decoder2, initial_sprs)
+        self.msr = SelectableInt(initial_msr, 64)  # underlying reg
+
         # TODO, needed here:
         # FPR (same as GPR except for FP nums)
         # 4.2.2 p124 FPSCR (definitely "separate" - not in SPR)
@@ -186,28 +334,35 @@ class ISACaller:
         # 3.2.3 p46 p232 VRSAVE (actually SPR #256)
 
         # create CR then allow portions of it to be "selectable" (below)
-        self._cr = SelectableInt(initial_cr, 64) # underlying reg
-        self.cr = FieldSelectableInt(self._cr, list(range(32,64)))
+        #rev_cr = int('{:016b}'.format(initial_cr)[::-1], 2)
+        self.cr = SelectableInt(initial_cr, 64)  # underlying reg
+        #self.cr = FieldSelectableInt(self._cr, list(range(32, 64)))
 
         # "undefined", just set to variable-bit-width int (use exts "max")
-        self.undefined = SelectableInt(0, 256) # TODO, not hard-code 256!
-
-        self.namespace = {'GPR': self.gpr,
-                          'MEM': self.mem,
-                          'SPR': self.spr,
-                          'memassign': self.memassign,
-                          'NIA': self.pc.NIA,
-                          'CIA': self.pc.CIA,
-                          'CR': self.cr,
-                          'undefined': self.undefined,
-                          'mode_is_64bit': True,
-                          'SO': XER_bits['SO']
-                          }
+        self.undefined = SelectableInt(0, 256)  # TODO, not hard-code 256!
+
+        self.namespace = {}
+        self.namespace.update(self.spr)
+        self.namespace.update({'GPR': self.gpr,
+                               'MEM': self.mem,
+                               'SPR': self.spr,
+                               'memassign': self.memassign,
+                               'NIA': self.pc.NIA,
+                               'CIA': self.pc.CIA,
+                               'CR': self.cr,
+                               'MSR': self.msr,
+                               'undefined': self.undefined,
+                               'mode_is_64bit': True,
+                               'SO': XER_bits['SO']
+                               })
+
+        # update pc to requested start point
+        self.set_pc(initial_pc)
 
         # field-selectable versions of Condition Register TODO check bitranges?
         self.crl = []
         for i in range(8):
-            bits = tuple(range(i*4, (i+1)*4))# errr... maybe?
+            bits = tuple(range(i*4+32, (i+1)*4+32))  # errr... maybe?
             _cr = FieldSelectableInt(self.cr, bits)
             self.crl.append(_cr)
             self.namespace["CR%d" % i] = _cr
@@ -215,6 +370,37 @@ class ISACaller:
         self.decoder = decoder2.dec
         self.dec2 = decoder2
 
+    def TRAP(self, trap_addr=0x700, trap_bit=PIb.TRAP):
+        print("TRAP:", hex(trap_addr), hex(self.namespace['MSR'].value))
+        # store CIA(+4?) in SRR0, set NIA to 0x700
+        # store MSR in SRR1, set MSR to um errr something, have to check spec
+        self.spr['SRR0'].value = self.pc.CIA.value
+        self.spr['SRR1'].value = self.namespace['MSR'].value
+        self.trap_nia = SelectableInt(trap_addr, 64)
+        self.spr['SRR1'][trap_bit] = 1  # change *copy* of MSR in SRR1
+
+        # set exception bits.  TODO: this should, based on the address
+        # in figure 66 p1065 V3.0B and the table figure 65 p1063 set these
+        # bits appropriately.  however it turns out that *for now* in all
+        # cases (all trap_addrs) the exact same thing is needed.
+        self.msr[MSRb.IR] = 0
+        self.msr[MSRb.DR] = 0
+        self.msr[MSRb.FE0] = 0
+        self.msr[MSRb.FE1] = 0
+        self.msr[MSRb.EE] = 0
+        self.msr[MSRb.RI] = 0
+        self.msr[MSRb.SF] = 1
+        self.msr[MSRb.TM] = 0
+        self.msr[MSRb.VEC] = 0
+        self.msr[MSRb.VSX] = 0
+        self.msr[MSRb.PR] = 0
+        self.msr[MSRb.FP] = 0
+        self.msr[MSRb.PMM] = 0
+        self.msr[MSRb.TEs] = 0
+        self.msr[MSRb.TEe] = 0
+        self.msr[MSRb.UND] = 0
+        self.msr[MSRb.LE] = 1
+
     def memassign(self, ea, sz, val):
         self.mem.memassign(ea, sz, val)
 
@@ -231,59 +417,290 @@ class ISACaller:
             else:
                 sig = getattr(fields, name)
             val = yield sig
-            if name in ['BF', 'BFA']:
+            # these are all opcode fields involved in index-selection of CR,
+            # and need to do "standard" arithmetic.  CR[BA+32] for example
+            # would, if using SelectableInt, only be 5-bit.
+            if name in ['BF', 'BFA', 'BC', 'BA', 'BB', 'BT', 'BI']:
                 self.namespace[name] = val
             else:
                 self.namespace[name] = SelectableInt(val, sig.width)
 
         self.namespace['XER'] = self.spr['XER']
         self.namespace['CA'] = self.spr['XER'][XER_bits['CA']].value
+        self.namespace['CA32'] = self.spr['XER'][XER_bits['CA32']].value
 
-    def handle_carry_(self, inputs, outputs):
-        inv_a = yield self.dec2.e.invert_a
+    def handle_carry_(self, inputs, outputs, already_done):
+        inv_a = yield self.dec2.e.do.invert_in
         if inv_a:
             inputs[0] = ~inputs[0]
+
+        imm_ok = yield self.dec2.e.do.imm_data.ok
+        if imm_ok:
+            imm = yield self.dec2.e.do.imm_data.data
+            inputs.append(SelectableInt(imm, 64))
         assert len(outputs) >= 1
-        output = outputs[0]
-        gts = [(x > output) == SelectableInt(1, 1) for x in inputs]
-        print(gts)
-        if any(gts):
-            cy = True
+        print("outputs", repr(outputs))
+        if isinstance(outputs, list) or isinstance(outputs, tuple):
+            output = outputs[0]
         else:
-            cy = False
-        self.spr['XER'][XER_bits['CA']] = cy
-
+            output = outputs
+        gts = []
+        for x in inputs:
+            print("gt input", x, output)
+            gt = (gtu(x, output))
+            gts.append(gt)
+        print(gts)
+        cy = 1 if any(gts) else 0
+        print("CA", cy, gts)
+        if not (1 & already_done):
+            self.spr['XER'][XER_bits['CA']] = cy
 
+        print("inputs", already_done, inputs)
         # 32 bit carry
-        gts = [(x[32:64] > output[32:64]) == SelectableInt(1, 1)
-               for x in inputs]
-        cy32 = 1 if any(gts) else 0
-        self.spr['XER'][XER_bits['CA32']] = cy32
-
+        # ARGH... different for OP_ADD... *sigh*...
+        op = yield self.dec2.e.do.insn_type
+        if op == MicrOp.OP_ADD.value:
+            res32 = (output.value & (1 << 32)) != 0
+            a32 = (inputs[0].value & (1 << 32)) != 0
+            if len(inputs) >= 2:
+                b32 = (inputs[1].value & (1 << 32)) != 0
+            else:
+                b32 = False
+            cy32 = res32 ^ a32 ^ b32
+            print("CA32 ADD", cy32)
+        else:
+            gts = []
+            for x in inputs:
+                print("input", x, output)
+                print("     x[32:64]", x, x[32:64])
+                print("     o[32:64]", output, output[32:64])
+                gt = (gtu(x[32:64], output[32:64])) == SelectableInt(1, 1)
+                gts.append(gt)
+            cy32 = 1 if any(gts) else 0
+            print("CA32", cy32, gts)
+        if not (2 & already_done):
+            self.spr['XER'][XER_bits['CA32']] = cy32
+
+    def handle_overflow(self, inputs, outputs, div_overflow):
+        if hasattr(self.dec2.e.do, "invert_in"):
+            inv_a = yield self.dec2.e.do.invert_in
+            if inv_a:
+                inputs[0] = ~inputs[0]
+
+        imm_ok = yield self.dec2.e.do.imm_data.ok
+        if imm_ok:
+            imm = yield self.dec2.e.do.imm_data.data
+            inputs.append(SelectableInt(imm, 64))
+        assert len(outputs) >= 1
+        print("handle_overflow", inputs, outputs, div_overflow)
+        if len(inputs) < 2 and div_overflow is None:
+            return
+
+        # div overflow is different: it's returned by the pseudo-code
+        # because it's more complex than can be done by analysing the output
+        if div_overflow is not None:
+            ov, ov32 = div_overflow, div_overflow
+        # arithmetic overflow can be done by analysing the input and output
+        elif len(inputs) >= 2:
+            output = outputs[0]
+
+            # OV (64-bit)
+            input_sgn = [exts(x.value, x.bits) < 0 for x in inputs]
+            output_sgn = exts(output.value, output.bits) < 0
+            ov = 1 if input_sgn[0] == input_sgn[1] and \
+                output_sgn != input_sgn[0] else 0
+
+            # OV (32-bit)
+            input32_sgn = [exts(x.value, 32) < 0 for x in inputs]
+            output32_sgn = exts(output.value, 32) < 0
+            ov32 = 1 if input32_sgn[0] == input32_sgn[1] and \
+                output32_sgn != input32_sgn[0] else 0
+
+        self.spr['XER'][XER_bits['OV']] = ov
+        self.spr['XER'][XER_bits['OV32']] = ov32
+        so = self.spr['XER'][XER_bits['SO']]
+        so = so | ov
+        self.spr['XER'][XER_bits['SO']] = so
 
     def handle_comparison(self, outputs):
         out = outputs[0]
+        assert isinstance(out, SelectableInt), \
+            "out zero not a SelectableInt %s" % repr(outputs)
+        print("handle_comparison", out.bits, hex(out.value))
+        # TODO - XXX *processor* in 32-bit mode
+        # https://bugs.libre-soc.org/show_bug.cgi?id=424
+        # if is_32bit:
+        #    o32 = exts(out.value, 32)
+        #    print ("handle_comparison exts 32 bit", hex(o32))
         out = exts(out.value, out.bits)
+        print("handle_comparison exts", hex(out))
         zero = SelectableInt(out == 0, 1)
         positive = SelectableInt(out > 0, 1)
         negative = SelectableInt(out < 0, 1)
-        SO = SelectableInt(0, 1)
+        SO = self.spr['XER'][XER_bits['SO']]
+        print("handle_comparison SO", SO)
         cr_field = selectconcat(negative, positive, zero, SO)
         self.crl[0].eq(cr_field)
 
     def set_pc(self, pc_val):
         self.namespace['NIA'] = SelectableInt(pc_val, 64)
         self.pc.update(self.namespace)
-        
+
+    def setup_one(self):
+        """set up one instruction
+        """
+        if self.respect_pc:
+            pc = self.pc.CIA.value
+        else:
+            pc = self.fake_pc
+        self._pc = pc
+        ins = self.imem.ld(pc, 4, False, True)
+        if ins is None:
+            raise KeyError("no instruction at 0x%x" % pc)
+        print("setup: 0x%x 0x%x %s" % (pc, ins & 0xffffffff, bin(ins)))
+        print("CIA NIA", self.respect_pc, self.pc.CIA.value, self.pc.NIA.value)
+
+        yield self.dec2.dec.raw_opcode_in.eq(ins & 0xffffffff)
+        yield self.dec2.dec.bigendian.eq(self.bigendian)
+        yield self.dec2.state.msr.eq(self.msr.value)
+        yield self.dec2.state.pc.eq(pc)
+
+    def execute_one(self):
+        """execute one instruction
+        """
+        # get the disassembly code for this instruction
+        code = self.disassembly[self._pc]
+        print("sim-execute", hex(self._pc), code)
+        opname = code.split(' ')[0]
+        yield from self.call(opname)
+
+        if not self.respect_pc:
+            self.fake_pc += 4
+        print("execute one, CIA NIA", self.pc.CIA.value, self.pc.NIA.value)
+
+    def get_assembly_name(self):
+        # TODO, asmregs is from the spec, e.g. add RT,RA,RB
+        # see http://bugs.libre-riscv.org/show_bug.cgi?id=282
+        dec_insn = yield self.dec2.e.do.insn
+        asmcode = yield self.dec2.dec.op.asmcode
+        print("get assembly name asmcode", asmcode, hex(dec_insn))
+        asmop = insns.get(asmcode, None)
+        int_op = yield self.dec2.dec.op.internal_op
+
+        # sigh reconstruct the assembly instruction name
+        if hasattr(self.dec2.e.do, "oe"):
+            ov_en = yield self.dec2.e.do.oe.oe
+            ov_ok = yield self.dec2.e.do.oe.ok
+        else:
+            ov_en = False
+            ov_ok = False
+        if hasattr(self.dec2.e.do, "rc"):
+            rc_en = yield self.dec2.e.do.rc.rc
+            rc_ok = yield self.dec2.e.do.rc.ok
+        else:
+            rc_en = False
+            rc_ok = False
+        # grrrr have to special-case MUL op (see DecodeOE)
+        print("ov %d en %d rc %d en %d op %d" %
+              (ov_ok, ov_en, rc_ok, rc_en, int_op))
+        if int_op in [MicrOp.OP_MUL_H64.value, MicrOp.OP_MUL_H32.value]:
+            print("mul op")
+            if rc_en & rc_ok:
+                asmop += "."
+        else:
+            if not asmop.endswith("."):  # don't add "." to "andis."
+                if rc_en & rc_ok:
+                    asmop += "."
+        if hasattr(self.dec2.e.do, "lk"):
+            lk = yield self.dec2.e.do.lk
+            if lk:
+                asmop += "l"
+        print("int_op", int_op)
+        if int_op in [MicrOp.OP_B.value, MicrOp.OP_BC.value]:
+            AA = yield self.dec2.dec.fields.FormI.AA[0:-1]
+            print("AA", AA)
+            if AA:
+                asmop += "a"
+        spr_msb = yield from self.get_spr_msb()
+        if int_op == MicrOp.OP_MFCR.value:
+            if spr_msb:
+                asmop = 'mfocrf'
+            else:
+                asmop = 'mfcr'
+        # XXX TODO: for whatever weird reason this doesn't work
+        # https://bugs.libre-soc.org/show_bug.cgi?id=390
+        if int_op == MicrOp.OP_MTCRF.value:
+            if spr_msb:
+                asmop = 'mtocrf'
+            else:
+                asmop = 'mtcrf'
+        return asmop
+
+    def get_spr_msb(self):
+        dec_insn = yield self.dec2.e.do.insn
+        return dec_insn & (1 << 20) != 0  # sigh - XFF.spr[-1]?
 
     def call(self, name):
+        name = name.strip()  # remove spaces if not already done so
+        if self.halted:
+            print("halted - not executing", name)
+            return
+
         # TODO, asmregs is from the spec, e.g. add RT,RA,RB
         # see http://bugs.libre-riscv.org/show_bug.cgi?id=282
+        asmop = yield from self.get_assembly_name()
+        print("call", name, asmop)
+
+        # check privileged
+        int_op = yield self.dec2.dec.op.internal_op
+        spr_msb = yield from self.get_spr_msb()
+
+        instr_is_privileged = False
+        if int_op in [MicrOp.OP_ATTN.value,
+                      MicrOp.OP_MFMSR.value,
+                      MicrOp.OP_MTMSR.value,
+                      MicrOp.OP_MTMSRD.value,
+                      # TODO: OP_TLBIE
+                      MicrOp.OP_RFID.value]:
+            instr_is_privileged = True
+        if int_op in [MicrOp.OP_MFSPR.value,
+                      MicrOp.OP_MTSPR.value] and spr_msb:
+            instr_is_privileged = True
+
+        print("is priv", instr_is_privileged, hex(self.msr.value),
+              self.msr[MSRb.PR])
+        # check MSR priv bit and whether op is privileged: if so, throw trap
+        if instr_is_privileged and self.msr[MSRb.PR] == 1:
+            self.TRAP(0x700, PIb.PRIV)
+            self.namespace['NIA'] = self.trap_nia
+            self.pc.update(self.namespace)
+            return
+
+        # check halted condition
+        if name == 'attn':
+            self.halted = True
+            return
+
+        # check illegal instruction
+        illegal = False
+        if name not in ['mtcrf', 'mtocrf']:
+            illegal = name != asmop
+
+        if illegal:
+            print("illegal", name, asmop)
+            self.TRAP(0x700, PIb.ILLEG)
+            self.namespace['NIA'] = self.trap_nia
+            self.pc.update(self.namespace)
+            print("name %s != %s - calling ILLEGAL trap, PC: %x" %
+                  (name, asmop, self.pc.CIA.value))
+            return
+
         info = self.instrs[name]
         yield from self.prep_namespace(info.form, info.op_fields)
 
         # preserve order of register names
-        input_names = create_args(list(info.read_regs) + list(info.uninit_regs))
+        input_names = create_args(list(info.read_regs) +
+                                  list(info.uninit_regs))
         print(input_names)
 
         # main registers (RT, RA ...)
@@ -302,28 +719,83 @@ class ISACaller:
             else:
                 inputs.append(self.namespace[special])
 
+        # clear trap (trap) NIA
+        self.trap_nia = None
+
         print(inputs)
         results = info.func(self, *inputs)
         print(results)
 
-        rc_en = yield self.dec2.e.rc.data
+        # "inject" decorator takes namespace from function locals: we need to
+        # overwrite NIA being overwritten (sigh)
+        if self.trap_nia is not None:
+            self.namespace['NIA'] = self.trap_nia
+
+        print("after func", self.namespace['CIA'], self.namespace['NIA'])
+
+        # detect if CA/CA32 already in outputs (sra*, basically)
+        already_done = 0
+        if info.write_regs:
+            output_names = create_args(info.write_regs)
+            for name in output_names:
+                if name == 'CA':
+                    already_done |= 1
+                if name == 'CA32':
+                    already_done |= 2
+
+        print("carry already done?", bin(already_done))
+        if hasattr(self.dec2.e.do, "output_carry"):
+            carry_en = yield self.dec2.e.do.output_carry
+        else:
+            carry_en = False
+        if carry_en:
+            yield from self.handle_carry_(inputs, results, already_done)
+
+        # detect if overflow was in return result
+        overflow = None
+        if info.write_regs:
+            for name, output in zip(output_names, results):
+                if name == 'overflow':
+                    overflow = output
+
+        if hasattr(self.dec2.e.do, "oe"):
+            ov_en = yield self.dec2.e.do.oe.oe
+            ov_ok = yield self.dec2.e.do.oe.ok
+        else:
+            ov_en = False
+            ov_ok = False
+        print("internal overflow", overflow, ov_en, ov_ok)
+        if ov_en & ov_ok:
+            yield from self.handle_overflow(inputs, results, overflow)
+
+        if hasattr(self.dec2.e.do, "rc"):
+            rc_en = yield self.dec2.e.do.rc.rc
+        else:
+            rc_en = False
         if rc_en:
             self.handle_comparison(results)
-        carry_en = yield self.dec2.e.output_carry
-        yield from self.handle_carry_(inputs, results)
 
         # any modified return results?
         if info.write_regs:
-            output_names = create_args(info.write_regs)
             for name, output in zip(output_names, results):
+                if name == 'overflow':  # ignore, done already (above)
+                    continue
                 if isinstance(output, int):
                     output = SelectableInt(output, 256)
-                if name in info.special_regs:
-                    print('writing special %s' % name, output)
+                if name in ['CA', 'CA32']:
+                    if carry_en:
+                        print("writing %s to XER" % name, output)
+                        self.spr['XER'][XER_bits[name]] = output.value
+                    else:
+                        print("NOT writing %s to XER" % name, output)
+                elif name in info.special_regs:
+                    print('writing special %s' % name, output, special_sprs)
                     if name in special_sprs:
                         self.spr[name] = output
                     else:
                         self.namespace[name].eq(output)
+                    if name == 'MSR':
+                        print('msr written', hex(self.msr.value))
                 else:
                     regnum = yield getattr(self.decoder, name)
                     print('writing reg %d %s' % (regnum, str(output)))
@@ -331,12 +803,23 @@ class ISACaller:
                         output = SelectableInt(output.value, 64)
                     self.gpr[regnum] = output
 
-        # update program counter
+        print("end of call", self.namespace['CIA'], self.namespace['NIA'])
+        # UPDATE program counter
         self.pc.update(self.namespace)
 
 
 def inject():
-    """ Decorator factory. """
+    """Decorator factory.
+
+    this decorator will "inject" variables into the function's namespace,
+    from the *dictionary* in self.namespace.  it therefore becomes possible
+    to make it look like a whole stack of variables which would otherwise
+    need "self." inserted in front of them (*and* for those variables to be
+    added to the instance) "appear" in the function.
+
+    "self.namespace['SI']" for example becomes accessible as just "SI" but
+    *only* inside the function, when decorated.
+    """
     def variable_injector(func):
         @wraps(func)
         def decorator(*args, **kwargs):
@@ -345,14 +828,17 @@ def inject():
             except AttributeError:
                 func_globals = func.func_globals  # Earlier versions.
 
-            context = args[0].namespace
+            context = args[0].namespace  # variables to be injected
             saved_values = func_globals.copy()  # Shallow copy of dict.
             func_globals.update(context)
             result = func(*args, **kwargs)
+            print("globals after", func_globals['CIA'], func_globals['NIA'])
+            print("args[0]", args[0].namespace['CIA'],
+                  args[0].namespace['NIA'])
             args[0].namespace = func_globals
             #exec (func.__code__, func_globals)
 
-            #finally:
+            # finally:
             #    func_globals = saved_values  # Undo changes.
 
             return result
@@ -360,4 +846,3 @@ def inject():
         return decorator
 
     return variable_injector
-