target/arm: Decode aa64 armv8.1 scalar three same extra

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20180228193125.20577-5-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

1 files changed, 109 insertions, 0 deletions

diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c
new file mode 100644
index 0000000000..9febdff69c
--- /dev/null
+++ b/target/arm/vec_helper.c
@@ -0,0 +1,109 @@
+/*
+ * ARM AdvSIMD / SVE Vector Operations
+ *
+ * Copyright (c) 2018 Linaro
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "qemu/osdep.h"
+#include "cpu.h"
+#include "exec/exec-all.h"
+#include "exec/helper-proto.h"
+#include "tcg/tcg-gvec-desc.h"
+
+
+#define SET_QC() env->vfp.xregs[ARM_VFP_FPSCR] |= CPSR_Q
+
+/* Signed saturating rounding doubling multiply-accumulate high half, 16-bit */
+static uint16_t inl_qrdmlah_s16(CPUARMState *env, int16_t src1,
+                                int16_t src2, int16_t src3)
+{
+    /* Simplify:
+     * = ((a3 << 16) + ((e1 * e2) << 1) + (1 << 15)) >> 16
+     * = ((a3 << 15) + (e1 * e2) + (1 << 14)) >> 15
+     */
+    int32_t ret = (int32_t)src1 * src2;
+    ret = ((int32_t)src3 << 15) + ret + (1 << 14);
+    ret >>= 15;
+    if (ret != (int16_t)ret) {
+        SET_QC();
+        ret = (ret < 0 ? -0x8000 : 0x7fff);
+    }
+    return ret;
+}
+
+uint32_t HELPER(neon_qrdmlah_s16)(CPUARMState *env, uint32_t src1,
+                                  uint32_t src2, uint32_t src3)
+{
+    uint16_t e1 = inl_qrdmlah_s16(env, src1, src2, src3);
+    uint16_t e2 = inl_qrdmlah_s16(env, src1 >> 16, src2 >> 16, src3 >> 16);
+    return deposit32(e1, 16, 16, e2);
+}
+
+/* Signed saturating rounding doubling multiply-subtract high half, 16-bit */
+static uint16_t inl_qrdmlsh_s16(CPUARMState *env, int16_t src1,
+                                int16_t src2, int16_t src3)
+{
+    /* Similarly, using subtraction:
+     * = ((a3 << 16) - ((e1 * e2) << 1) + (1 << 15)) >> 16
+     * = ((a3 << 15) - (e1 * e2) + (1 << 14)) >> 15
+     */
+    int32_t ret = (int32_t)src1 * src2;
+    ret = ((int32_t)src3 << 15) - ret + (1 << 14);
+    ret >>= 15;
+    if (ret != (int16_t)ret) {
+        SET_QC();
+        ret = (ret < 0 ? -0x8000 : 0x7fff);
+    }
+    return ret;
+}
+
+uint32_t HELPER(neon_qrdmlsh_s16)(CPUARMState *env, uint32_t src1,
+                                  uint32_t src2, uint32_t src3)
+{
+    uint16_t e1 = inl_qrdmlsh_s16(env, src1, src2, src3);
+    uint16_t e2 = inl_qrdmlsh_s16(env, src1 >> 16, src2 >> 16, src3 >> 16);
+    return deposit32(e1, 16, 16, e2);
+}
+
+/* Signed saturating rounding doubling multiply-accumulate high half, 32-bit */
+uint32_t HELPER(neon_qrdmlah_s32)(CPUARMState *env, int32_t src1,
+                                  int32_t src2, int32_t src3)
+{
+    /* Simplify similarly to int_qrdmlah_s16 above.  */
+    int64_t ret = (int64_t)src1 * src2;
+    ret = ((int64_t)src3 << 31) + ret + (1 << 30);
+    ret >>= 31;
+    if (ret != (int32_t)ret) {
+        SET_QC();
+        ret = (ret < 0 ? INT32_MIN : INT32_MAX);
+    }
+    return ret;
+}
+
+/* Signed saturating rounding doubling multiply-subtract high half, 32-bit */
+uint32_t HELPER(neon_qrdmlsh_s32)(CPUARMState *env, int32_t src1,
+                                  int32_t src2, int32_t src3)
+{
+    /* Simplify similarly to int_qrdmlsh_s16 above.  */
+    int64_t ret = (int64_t)src1 * src2;
+    ret = ((int64_t)src3 << 31) - ret + (1 << 30);
+    ret >>= 31;
+    if (ret != (int32_t)ret) {
+        SET_QC();
+        ret = (ret < 0 ? INT32_MIN : INT32_MAX);
+    }
+    return ret;
+}