x86emul: support AVX512BW pack insns

No further test harness additions - what is there is good enough for
these rather "regular" insns.

Signed-off-by: Jan Beulich <jbeulich@suse.com>
Acked-by: Andrew Cooper <andrew.cooper3@citrix.com>

tools/tests/x86_emulator/evex-disp8.c

@@ -306,6 +306,10 @@ static const struct test avx512bw_all[] = {
     INSN(movdqu8,     f2,   0f, 7f,    vl,    b, vl),
     INSN(movdqu16,    f2,   0f, 6f,    vl,    w, vl),
     INSN(movdqu16,    f2,   0f, 7f,    vl,    w, vl),
+    INSN(packssdw,    66,   0f, 6b,    vl, d_nb, vl),
+    INSN(packsswb,    66,   0f, 63,    vl,    w, vl),
+    INSN(packusdw,    66, 0f38, 2b,    vl, d_nb, vl),
+    INSN(packuswb,    66,   0f, 67,    vl,    w, vl),
     INSN(paddb,       66,   0f, fc,    vl,    b, vl),
     INSN(paddsb,      66,   0f, ec,    vl,    b, vl),
     INSN(paddsw,      66,   0f, ed,    vl,    w, vl),

tools/tests/x86_emulator/simd.h

@@ -361,6 +361,10 @@ OVR(pextrw);
 OVR(pinsrb);
 OVR(pinsrw);
 #  ifdef __AVX512VL__
+OVR(packssdw);
+OVR(packsswb);
+OVR(packusdw);
+OVR(packuswb);
 OVR(pmaddwd);
 OVR(pmovsxbw);
 OVR(pmovzxbw);

xen/arch/x86/x86_emulate/x86_emulate.c

@@ -453,7 +453,7 @@ static const struct ext0f38_table {
     [0x25] = { .simd_size = simd_other, .two_op = 1, .d8s = d8s_vl_by_2 },
     [0x26 ... 0x29] = { .simd_size = simd_packed_int, .d8s = d8s_vl },
     [0x2a] = { .simd_size = simd_packed_int, .two_op = 1, .d8s = d8s_vl },
-    [0x2b] = { .simd_size = simd_packed_int },
+    [0x2b] = { .simd_size = simd_packed_int, .d8s = d8s_vl },
     [0x2c ... 0x2d] = { .simd_size = simd_packed_fp },
     [0x2e ... 0x2f] = { .simd_size = simd_packed_fp, .to_mem = 1 },
     [0x30] = { .simd_size = simd_other, .two_op = 1, .d8s = d8s_vl_by_2 },
@@ -6720,6 +6720,8 @@ x86_emulate(
     case X86EMUL_OPC_EVEX_66(0x0f, 0x69): /* vpunpckhwd [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
         op_bytes = 16 << evex.lr;
         /* fall through */
+    case X86EMUL_OPC_EVEX_66(0x0f, 0x63): /* vpacksswb [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+    case X86EMUL_OPC_EVEX_66(0x0f, 0x67): /* vpackuswb [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
     case X86EMUL_OPC_EVEX_66(0x0f, 0xd1): /* vpsrlw xmm/m128,[xyz]mm,[xyz]mm{k} */
     case X86EMUL_OPC_EVEX_66(0x0f, 0xe1): /* vpsraw xmm/m128,[xyz]mm,[xyz]mm{k} */
     case X86EMUL_OPC_EVEX_66(0x0f, 0xf1): /* vpsllw xmm/m128,[xyz]mm,[xyz]mm{k} */
@@ -6781,6 +6783,12 @@ x86_emulate(
         avx512_vlen_check(false);
         goto simd_zmm;
 
+    case X86EMUL_OPC_EVEX_66(0x0f, 0x6b): /* vpackssdw [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+    case X86EMUL_OPC_EVEX_66(0x0f38, 0x2b): /* vpackusdw [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
+        generate_exception_if(evex.w || evex.brs, EXC_UD);
+        fault_suppression = false;
+        goto avx512f_no_sae;
+
     case X86EMUL_OPC_EVEX_66(0x0f, 0x6c): /* vpunpcklqdq [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
     case X86EMUL_OPC_EVEX_66(0x0f, 0x6d): /* vpunpckhqdq [xyz]mm/mem,[xyz]mm,[xyz]mm{k} */
         fault_suppression = false;