AVX-512: Add a test case for EVEX encoded instructions

This was converted from a gas testsuite.
(gas/testsuite/gas/i386/x86-64-avx512f-intel.d)
A python script that is used for converting is also included.

Signed-off-by: Jin Kyu Song <jin.kyu.song@intel.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>

test/gas2nasm.py

@@ -0,0 +1,96 @@
+#!/usr/bin/env python -tt
+# -*- python -*-
+# Convert gas testsuite file to NASM test asm file
+# usage >
+# python gas2nasm.py -i input_gas_file -o output_nasm_file -b bits
+# e.g. python gas2nasm.py -i x86-64-avx512f-intel.d -o avx512f.asm -b 64
+
+import sys
+import os
+import optparse
+import re
+
+def setup():
+    parser = optparse.OptionParser()
+    parser.add_option('-i', dest='input', action='store',
+            default="",
+            help='Name for input gas testsuite file.')
+    parser.add_option('-o', dest='output', action='store',
+            default="",
+            help='Name for output NASM test asm file.')
+    parser.add_option('-b', dest='bits', action='store',
+            default="",
+            help='Bits for output ASM file.')
+    (options, args) =  parser.parse_args()
+    return options
+
+def read(options):
+    with open(options.input, 'rb') as f:
+        recs = []
+        for line in f:
+            if line[0] == '[':
+                d = []
+                strr = line[16:].partition('   ')
+                if strr[1] == '':
+                    strr = line[16:].partition('\t')
+                l = strr[0].strip()
+                r = strr[2].strip()
+                # Filter out Pseudo-op / vex instructions until those are added in insns.dat
+                if not (re.match('vcmp.+[ps][ds]', r) or re.match('vpcmp[^u]+u?[dq]', r) or r[0] == 'k'):
+                    d.append(l)
+                    d.append(r)
+                    recs.append(d)
+    return recs
+
+def commas(recs):
+    replace_tbl = {' PTR':'', '\\':'', 'MM':'', '{':'\{', '}':'\}', 'XWORD':'OWORD'}
+    reccommas = []
+    for insn in recs:
+        new = []
+        byte = '0x' + insn[0].replace(' ', ', 0x')
+        for rep in replace_tbl.keys():
+            insn[1] = insn[1].replace(rep, replace_tbl[rep])
+        mnemonic = insn[1]
+
+        # gas size specifier for gather and scatter insturctions seems wrong. just remove them.
+        if 'gather' in insn[1] or 'scatter' in insn[1]:
+            mnemonic = mnemonic.replace('ZWORD', '')
+
+        new.append(byte)
+        new.append(mnemonic)
+        reccommas.append(new)
+    return reccommas
+
+# The spaces reserved here can be adjusted according to the output string length.
+# maxlen printed out at the end of the process will give a hint for it.
+outstrfmt = "testcase\t{ %-70s }, { %-60s }\n"
+
+macro = "%macro testcase 2\n %ifdef BIN\n  db %1\n %endif\n %ifdef SRC\n  %2\n %endif\n%endmacro\n\n\n"
+
+def write(data, options):
+    if options.output:
+        with open(options.output, 'wb') as out:
+            out.write(macro)
+            if options.bits:
+                out.write('bits ' + options.bits + '\n\n')
+            for insn in data:
+                outstr = outstrfmt % tuple(insn)
+                out.write(outstr)
+
+if __name__ == "__main__":
+    options = setup()
+    recs = read(options)
+    print "AVX3.1 instructions"
+
+    recs = commas(recs)
+
+    write(recs, options)
+
+    maxlen = [0,0,0,0,0,0,0,0]
+    for insn in recs:
+#print insn[0] + '\t<-\t' + insn[1]
+        print outstrfmt[:-1] % tuple(insn)
+        for i, strstr in enumerate(insn):
+            if maxlen[i] < len(strstr): maxlen[i] = len(strstr)
+
+    print maxlen