[Qemu-devel] [PATCH v2 00/10] A64 SIMD patchset one: ld/st, C3.6.1..C3.6.7

[Qemu-devel] [PATCH v2 00/10] A64 SIMD patchset one: ld/st, C3.6.1..C3.6.7

[Peter Maydell]

This is an initial set of patches which make a start on SIMD (Neon)
emulation in the A64 decoder. The patches implement all the SIMD
load/store operations, provide a decoder skeleton for the SIMD
dp instructions, and implement all the instructions in the ARM ARM's
groupings C3.6.1 through C3.6.7.

Changes v1->v2:
 * minor tweaks as suggested by rth (noted more specifically
   below the fold in the affected patches)
 * reworked EXT to do vector accesses outside do_ext64 and
   thus avoid accessing the middle element twice
 * fixed incorrect handling of rn==sp, cleaned up tcg_rn code in
   the ld/st single and multiple patches

thanks
-- PMM

Alex Bennée (4):
  target-arm: A64: Add SIMD ld/st multiple
  target-arm: A64: Add decode skeleton for SIMD data processing insns
  target-arm: A64: Add SIMD copy operations
  target-arm: A64: Add SIMD modified immediate group

Michael Matz (3):
  target-arm: A64: Add SIMD TBL/TBLX
  target-arm: A64: Add SIMD ZIP/UZP/TRN
  target-arm: A64: Add SIMD across-lanes instructions

Peter Maydell (3):
  target-arm: A64: Add SIMD ld/st single
  target-arm: A64: Add SIMD EXT
  target-arm: A64: Add SIMD scalar copy instructions

 target-arm/helper-a64.c    |   31 +
 target-arm/helper-a64.h    |    1 +
 target-arm/translate-a64.c | 1454 +++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 1477 insertions(+), 9 deletions(-)

-- 
1.8.5

[Qemu-devel] [PATCH v2 01/10] target-arm: A64: Add SIMD ld/st multiple

[Peter Maydell]

From: Alex Bennée <alex.bennee@linaro.org>

This adds support support for the SIMD load/store
multiple category of instructions.

This also brings in a couple of helper functions for manipulating
sections of the SIMD registers:

  * do_vec_get - fetch value from a slice of a vector register
  * do_vec_set - set a slice of a vector register

which use vec_reg_offset for consistent processing of offsets in an
endian aware manner. There are also additional helpers:

  * do_vec_ld - load value into SIMD
  * do_vec_st - store value from SIMD

which load or store a slice of a vector register to memory.
These don't zero extend like the fp variants.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
public v1->v2 changes:
 removed stray double-spacing
 fixed incorrect handling of rn==sp, cleaned up tcg_rn code
---
 target-arm/translate-a64.c | 250 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 248 insertions(+), 2 deletions(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index cf80c46..e4fdf00 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -308,6 +308,28 @@ static TCGv_i64 read_cpu_reg_sp(DisasContext *s, int reg, int sf)
     return v;
 }
 
+/* Return the offset into CPUARMState of an element of specified
+ * size, 'element' places in from the least significant end of
+ * the FP/vector register Qn.
+ */
+static inline int vec_reg_offset(int regno, int element, TCGMemOp size)
+{
+    int offs = offsetof(CPUARMState, vfp.regs[regno * 2]);
+#ifdef HOST_WORDS_BIGENDIAN
+    /* This is complicated slightly because vfp.regs[2n] is
+     * still the low half and  vfp.regs[2n+1] the high half
+     * of the 128 bit vector, even on big endian systems.
+     * Calculate the offset assuming a fully bigendian 128 bits,
+     * then XOR to account for the order of the two 64 bit halves.
+     */
+    offs += (16 - ((element + 1) * (1 << size)));
+    offs ^= 8;
+#else
+    offs += element * (1 << size);
+#endif
+    return offs;
+}
+
 /* Return the offset into CPUARMState of a slice (from
  * the least significant end) of FP register Qn (ie
  * Dn, Sn, Hn or Bn).
@@ -661,6 +683,108 @@ static void do_fp_ld(DisasContext *s, int destidx, TCGv_i64 tcg_addr, int size)
 }
 
 /*
+ * Vector load/store helpers.
+ *
+ * The principal difference between this and a FP load is that we don't
+ * zero extend as we are filling a partial chunk of the vector register.
+ * These functions don't support 128 bit loads/stores, which would be
+ * normal load/store operations.
+ */
+
+/* Get value of an element within a vector register */
+static void read_vec_element(DisasContext *s, TCGv_i64 tcg_dest, int srcidx,
+                             int element, TCGMemOp memop)
+{
+    int vect_off = vec_reg_offset(srcidx, element, memop & MO_SIZE);
+    switch (memop) {
+    case MO_8:
+        tcg_gen_ld8u_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_16:
+        tcg_gen_ld16u_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_32:
+        tcg_gen_ld32u_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_8|MO_SIGN:
+        tcg_gen_ld8s_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_16|MO_SIGN:
+        tcg_gen_ld16s_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_32|MO_SIGN:
+        tcg_gen_ld32s_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    case MO_64:
+    case MO_64|MO_SIGN:
+        tcg_gen_ld_i64(tcg_dest, cpu_env, vect_off);
+        break;
+    default:
+        g_assert_not_reached();
+    }
+}
+
+/* Set value of an element within a vector register */
+static void write_vec_element(DisasContext *s, TCGv_i64 tcg_src, int destidx,
+                              int element, TCGMemOp memop)
+{
+    int vect_off = vec_reg_offset(destidx, element, memop & MO_SIZE);
+    switch (memop) {
+    case MO_8:
+        tcg_gen_st8_i64(tcg_src, cpu_env, vect_off);
+        break;
+    case MO_16:
+        tcg_gen_st16_i64(tcg_src, cpu_env, vect_off);
+        break;
+    case MO_32:
+        tcg_gen_st32_i64(tcg_src, cpu_env, vect_off);
+        break;
+    case MO_64:
+        tcg_gen_st_i64(tcg_src, cpu_env, vect_off);
+        break;
+    default:
+        g_assert_not_reached();
+    }
+}
+
+/* Clear the high 64 bits of a 128 bit vector (in general non-quad
+ * vector ops all need to do this).
+ */
+static void clear_vec_high(DisasContext *s, int rd)
+{
+    TCGv_i64 tcg_zero = tcg_const_i64(0);
+
+    write_vec_element(s, tcg_zero, rd, 1, MO_64);
+    tcg_temp_free_i64(tcg_zero);
+}
+
+/* Store from vector register to memory */
+static void do_vec_st(DisasContext *s, int srcidx, int element,
+                      TCGv_i64 tcg_addr, int size)
+{
+    TCGMemOp memop = MO_TE + size;
+    TCGv_i64 tcg_tmp = tcg_temp_new_i64();
+
+    read_vec_element(s, tcg_tmp, srcidx, element, size);
+    tcg_gen_qemu_st_i64(tcg_tmp, tcg_addr, get_mem_index(s), memop);
+
+    tcg_temp_free_i64(tcg_tmp);
+}
+
+/* Load from memory to vector register */
+static void do_vec_ld(DisasContext *s, int destidx, int element,
+                      TCGv_i64 tcg_addr, int size)
+{
+    TCGMemOp memop = MO_TE + size;
+    TCGv_i64 tcg_tmp = tcg_temp_new_i64();
+
+    tcg_gen_qemu_ld_i64(tcg_tmp, tcg_addr, get_mem_index(s), memop);
+    write_vec_element(s, tcg_tmp, destidx, element, size);
+
+    tcg_temp_free_i64(tcg_tmp);
+}
+
+/*
  * This utility function is for doing register extension with an
  * optional shift. You will likely want to pass a temporary for the
  * destination register. See DecodeRegExtend() in the ARM ARM.
@@ -1835,10 +1959,132 @@ static void disas_ldst_reg(DisasContext *s, uint32_t insn)
     }
 }
 
-/* AdvSIMD load/store multiple structures */
+/* C3.3.1 AdvSIMD load/store multiple structures
+ *
+ *  31  30  29           23 22  21         16 15    12 11  10 9    5 4    0
+ * +---+---+---------------+---+-------------+--------+------+------+------+
+ * | 0 | Q | 0 0 1 1 0 0 0 | L | 0 0 0 0 0 0 | opcode | size |  Rn  |  Rt  |
+ * +---+---+---------------+---+-------------+--------+------+------+------+
+ *
+ * C3.3.2 AdvSIMD load/store multiple structures (post-indexed)
+ *
+ *  31  30  29           23 22  21  20     16 15    12 11  10 9    5 4    0
+ * +---+---+---------------+---+---+---------+--------+------+------+------+
+ * | 0 | Q | 0 0 1 1 0 0 1 | L | 0 |   Rm    | opcode | size |  Rn  |  Rt  |
+ * +---+---+---------------+---+---+---------+--------+------+------+------+
+ *
+ * Rt: first (or only) SIMD&FP register to be transferred
+ * Rn: base address or SP
+ * Rm (post-index only): post-index register (when !31) or size dependent #imm
+ */
 static void disas_ldst_multiple_struct(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rt = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int size = extract32(insn, 10, 2);
+    int opcode = extract32(insn, 12, 4);
+    bool is_store = !extract32(insn, 22, 1);
+    bool is_postidx = extract32(insn, 23, 1);
+    bool is_q = extract32(insn, 30, 1);
+    TCGv_i64 tcg_addr, tcg_rn;
+
+    int ebytes = 1 << size;
+    int elements = (is_q ? 128 : 64) / (8 << size);
+    int rpt;    /* num iterations */
+    int selem;  /* structure elements */
+    int r;
+
+    if (extract32(insn, 31, 1) || extract32(insn, 21, 1)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    /* From the shared decode logic */
+    switch (opcode) {
+    case 0x0:
+        rpt = 1;
+        selem = 4;
+        break;
+    case 0x2:
+        rpt = 4;
+        selem = 1;
+        break;
+    case 0x4:
+        rpt = 1;
+        selem = 3;
+        break;
+    case 0x6:
+        rpt = 3;
+        selem = 1;
+        break;
+    case 0x7:
+        rpt = 1;
+        selem = 1;
+        break;
+    case 0x8:
+        rpt = 1;
+        selem = 2;
+        break;
+    case 0xa:
+        rpt = 2;
+        selem = 1;
+        break;
+    default:
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (size == 3 && !is_q && selem != 1) {
+        /* reserved */
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (rn == 31) {
+        gen_check_sp_alignment(s);
+    }
+
+    tcg_rn = cpu_reg_sp(s, rn);
+    tcg_addr = tcg_temp_new_i64();
+    tcg_gen_mov_i64(tcg_addr, tcg_rn);
+
+    for (r = 0; r < rpt; r++) {
+        int e;
+        for (e = 0; e < elements; e++) {
+            int tt = (rt + r) % 32;
+            int xs;
+            for (xs = 0; xs < selem; xs++) {
+                if (is_store) {
+                    do_vec_st(s, tt, e, tcg_addr, size);
+                } else {
+                    do_vec_ld(s, tt, e, tcg_addr, size);
+
+                    /* For non-quad operations, setting a slice of the low
+                     * 64 bits of the register clears the high 64 bits (in
+                     * the ARM ARM pseudocode this is implicit in the fact
+                     * that 'rval' is a 64 bit wide variable). We optimize
+                     * by noticing that we only need to do this the first
+                     * time we touch a register.
+                     */
+                    if (!is_q && e == 0 && (r == 0 || xs == selem - 1)) {
+                        clear_vec_high(s, tt);
+                    }
+                }
+                tcg_gen_addi_i64(tcg_addr, tcg_addr, ebytes);
+                tt = (tt + 1) % 32;
+            }
+        }
+    }
+
+    if (is_postidx) {
+        int rm = extract32(insn, 16, 5);
+        if (rm == 31) {
+            tcg_gen_mov_i64(tcg_rn, tcg_addr);
+        } else {
+            tcg_gen_add_i64(tcg_rn, tcg_rn, cpu_reg(s, rm));
+        }
+    }
+    tcg_temp_free_i64(tcg_addr);
 }
 
 /* AdvSIMD load/store single structure */
-- 
1.8.5

[Qemu-devel] [PATCH v2 02/10] target-arm: A64: Add SIMD ld/st single

[Peter Maydell]

Implement the SIMD ld/st single structure instructions.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
public v1->v2 changes:
 * same tcg_rn cleanup/fix as ld/st multiple
---
 target-arm/translate-a64.c | 144 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 142 insertions(+), 2 deletions(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index e4fdf00..6e31740 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -2087,10 +2087,150 @@ static void disas_ldst_multiple_struct(DisasContext *s, uint32_t insn)
     tcg_temp_free_i64(tcg_addr);
 }
 
-/* AdvSIMD load/store single structure */
+/* C3.3.3 AdvSIMD load/store single structure
+ *
+ *  31  30  29           23 22 21 20       16 15 13 12  11  10 9    5 4    0
+ * +---+---+---------------+-----+-----------+-----+---+------+------+------+
+ * | 0 | Q | 0 0 1 1 0 1 0 | L R | 0 0 0 0 0 | opc | S | size |  Rn  |  Rt  |
+ * +---+---+---------------+-----+-----------+-----+---+------+------+------+
+ *
+ * C3.3.4 AdvSIMD load/store single structure (post-indexed)
+ *
+ *  31  30  29           23 22 21 20       16 15 13 12  11  10 9    5 4    0
+ * +---+---+---------------+-----+-----------+-----+---+------+------+------+
+ * | 0 | Q | 0 0 1 1 0 1 1 | L R |     Rm    | opc | S | size |  Rn  |  Rt  |
+ * +---+---+---------------+-----+-----------+-----+---+------+------+------+
+ *
+ * Rt: first (or only) SIMD&FP register to be transferred
+ * Rn: base address or SP
+ * Rm (post-index only): post-index register (when !31) or size dependent #imm
+ * index = encoded in Q:S:size dependent on size
+ *
+ * lane_size = encoded in R, opc
+ * transfer width = encoded in opc, S, size
+ */
 static void disas_ldst_single_struct(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rt = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int size = extract32(insn, 10, 2);
+    int S = extract32(insn, 12, 1);
+    int opc = extract32(insn, 13, 3);
+    int R = extract32(insn, 21, 1);
+    int is_load = extract32(insn, 22, 1);
+    int is_postidx = extract32(insn, 23, 1);
+    int is_q = extract32(insn, 30, 1);
+
+    int scale = extract32(opc, 1, 2);
+    int selem = (extract32(opc, 0, 1) << 1 | R) + 1;
+    bool replicate = false;
+    int index = is_q << 3 | S << 2 | size;
+    int ebytes, xs;
+    TCGv_i64 tcg_addr, tcg_rn;
+
+    switch (scale) {
+    case 3:
+        if (!is_load || S) {
+            unallocated_encoding(s);
+            return;
+        }
+        scale = size;
+        replicate = true;
+        break;
+    case 0:
+        break;
+    case 1:
+        if (extract32(size, 0, 1)) {
+            unallocated_encoding(s);
+            return;
+        }
+        index >>= 1;
+        break;
+    case 2:
+        if (extract32(size, 1, 1)) {
+            unallocated_encoding(s);
+            return;
+        }
+        if (!extract32(size, 0, 1)) {
+            index >>= 2;
+        } else {
+            if (S) {
+                unallocated_encoding(s);
+                return;
+            }
+            index >>= 3;
+            scale = 3;
+        }
+        break;
+    default:
+        g_assert_not_reached();
+    }
+
+    ebytes = 1 << scale;
+
+    if (rn == 31) {
+        gen_check_sp_alignment(s);
+    }
+
+    tcg_rn = cpu_reg_sp(s, rn);
+    tcg_addr = tcg_temp_new_i64();
+    tcg_gen_mov_i64(tcg_addr, tcg_rn);
+
+    for (xs = 0; xs < selem; xs++) {
+        if (replicate) {
+            /* Load and replicate to all elements */
+            uint64_t mulconst;
+            TCGv_i64 tcg_tmp = tcg_temp_new_i64();
+
+            tcg_gen_qemu_ld_i64(tcg_tmp, tcg_addr,
+                                get_mem_index(s), MO_TE + scale);
+            switch (scale) {
+            case 0:
+                mulconst = 0x0101010101010101ULL;
+                break;
+            case 1:
+                mulconst = 0x0001000100010001ULL;
+                break;
+            case 2:
+                mulconst = 0x0000000100000001ULL;
+                break;
+            case 3:
+                mulconst = 0;
+                break;
+            default:
+                g_assert_not_reached();
+            }
+            if (mulconst) {
+                tcg_gen_muli_i64(tcg_tmp, tcg_tmp, mulconst);
+            }
+            write_vec_element(s, tcg_tmp, rt, 0, MO_64);
+            if (is_q) {
+                write_vec_element(s, tcg_tmp, rt, 1, MO_64);
+            } else {
+                clear_vec_high(s, rt);
+            }
+            tcg_temp_free_i64(tcg_tmp);
+        } else {
+            /* Load/store one element per register */
+            if (is_load) {
+                do_vec_ld(s, rt, index, tcg_addr, MO_TE + scale);
+            } else {
+                do_vec_st(s, rt, index, tcg_addr, MO_TE + scale);
+            }
+        }
+        tcg_gen_addi_i64(tcg_addr, tcg_addr, ebytes);
+        rt = (rt + 1) % 32;
+    }
+
+    if (is_postidx) {
+        int rm = extract32(insn, 16, 5);
+        if (rm == 31) {
+            tcg_gen_mov_i64(tcg_rn, tcg_addr);
+        } else {
+            tcg_gen_add_i64(tcg_rn, tcg_rn, cpu_reg(s, rm));
+        }
+    }
+    tcg_temp_free_i64(tcg_addr);
 }
 
 /* C3.3 Loads and stores */
-- 
1.8.5

[Qemu-devel] [PATCH v2 03/10] target-arm: A64: Add decode skeleton for SIMD data processing insns

[Peter Maydell]

From: Alex Bennée <alex.bennee@linaro.org>

Add decode skeleton and function placeholders for all the SIMD data
processing instructions. Due to the complexity of this part of the
table the normal extract and switch approach gets very messy very
quickly, so we use a simple data-driven pattern-and-mask approach.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
v1->v2:
 * made data_proc_simd table const
 * fixed errors in insn diagram comments
 * fixed coding style for fn pointer call
---
 target-arm/translate-a64.c | 306 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 305 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 6e31740..8060c9f 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -61,6 +61,17 @@ enum a64_shift_type {
     A64_SHIFT_TYPE_ROR = 3
 };
 
+/* Table based decoder typedefs - used when the relevant bits for decode
+ * are too awkwardly scattered across the instruction (eg SIMD).
+ */
+typedef void AArch64DecodeFn(DisasContext *s, uint32_t insn);
+
+typedef struct AArch64DecodeTable {
+    uint32_t pattern;
+    uint32_t mask;
+    AArch64DecodeFn *disas_fn;
+} AArch64DecodeTable;
+
 /* initialize TCG globals.  */
 void a64_translate_init(void)
 {
@@ -846,6 +857,31 @@ static inline void gen_check_sp_alignment(DisasContext *s)
 }
 
 /*
+ * This provides a simple table based table lookup decoder. It is
+ * intended to be used when the relevant bits for decode are too
+ * awkwardly placed and switch/if based logic would be confusing and
+ * deeply nested. Since it's a linear search through the table, tables
+ * should be kept small.
+ *
+ * It returns the first handler where insn & mask == pattern, or
+ * NULL if there is no match.
+ * The table is terminated by an empty mask (i.e. 0)
+ */
+static inline AArch64DecodeFn *lookup_disas_fn(const AArch64DecodeTable *table,
+                                               uint32_t insn)
+{
+    const AArch64DecodeTable *tptr = table;
+
+    while (tptr->mask) {
+        if ((insn & tptr->mask) == tptr->pattern) {
+            return tptr->disas_fn;
+        }
+        tptr++;
+    }
+    return NULL;
+}
+
+/*
  * the instruction disassembly implemented here matches
  * the instruction encoding classifications in chapter 3 (C3)
  * of the ARM Architecture Reference Manual (DDI0487A_a)
@@ -4610,13 +4646,281 @@ static void disas_data_proc_fp(DisasContext *s, uint32_t insn)
     }
 }
 
+/* C3.6.1 EXT
+ *   31  30 29         24 23 22  21 20  16 15  14  11 10  9    5 4    0
+ * +---+---+-------------+-----+---+------+---+------+---+------+------+
+ * | 0 | Q | 1 0 1 1 1 0 | op2 | 0 |  Rm  | 0 | imm4 | 0 |  Rn  |  Rd  |
+ * +---+---+-------------+-----+---+------+---+------+---+------+------+
+ */
+static void disas_simd_ext(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.2 TBL/TBX
+ *   31  30 29         24 23 22  21 20  16 15  14 13  12  11 10 9    5 4    0
+ * +---+---+-------------+-----+---+------+---+-----+----+-----+------+------+
+ * | 0 | Q | 0 0 1 1 1 0 | op2 | 0 |  Rm  | 0 | len | op | 0 0 |  Rn  |  Rd  |
+ * +---+---+-------------+-----+---+------+---+-----+----+-----+------+------+
+ */
+static void disas_simd_tb(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.3 ZIP/UZP/TRN
+ *   31  30 29         24 23  22  21 20   16 15 14 12 11 10 9    5 4    0
+ * +---+---+-------------+------+---+------+---+------------------+------+
+ * | 0 | Q | 0 0 1 1 1 0 | size | 0 |  Rm  | 0 | opc | 1 0 |  Rn  |  Rd  |
+ * +---+---+-------------+------+---+------+---+------------------+------+
+ */
+static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.4 AdvSIMD across lanes
+ *   31  30  29 28       24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +---+---+---+-----------+------+-----------+--------+-----+------+------+
+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 1 0 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +---+---+---+-----------+------+-----------+--------+-----+------+------+
+ */
+static void disas_simd_across_lanes(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.5 AdvSIMD copy
+ *   31  30  29  28             21 20  16 15  14  11 10  9    5 4    0
+ * +---+---+----+-----------------+------+---+------+---+------+------+
+ * | 0 | Q | op | 0 1 1 1 0 0 0 0 | imm5 | 0 | imm4 | 1 |  Rn  |  Rd  |
+ * +---+---+----+-----------------+------+---+------+---+------+------+
+ */
+static void disas_simd_copy(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.6 AdvSIMD modified immediate
+ *  31  30   29  28                 19 18 16 15   12  11  10  9     5 4    0
+ * +---+---+----+---------------------+-----+-------+----+---+-------+------+
+ * | 0 | Q | op | 0 1 1 1 1 0 0 0 0 0 | abc | cmode | o2 | 1 | defgh |  Rd  |
+ * +---+---+----+---------------------+-----+-------+----+---+-------+------+
+ */
+static void disas_simd_mod_imm(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.7 AdvSIMD scalar copy
+ *  31 30  29  28             21 20  16 15  14  11 10  9    5 4    0
+ * +-----+----+-----------------+------+---+------+---+------+------+
+ * | 0 1 | op | 1 1 1 1 0 0 0 0 | imm5 | 0 | imm4 | 1 |  Rn  |  Rd  |
+ * +-----+----+-----------------+------+---+------+---+------+------+
+ */
+static void disas_simd_scalar_copy(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.8 AdvSIMD scalar pairwise
+ *  31 30  29 28       24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +-----+---+-----------+------+-----------+--------+-----+------+------+
+ * | 0 1 | U | 1 1 1 1 0 | size | 1 1 0 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +-----+---+-----------+------+-----------+--------+-----+------+------+
+ */
+static void disas_simd_scalar_pairwise(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.9 AdvSIMD scalar shift by immediate
+ *  31 30  29 28         23 22  19 18  16 15    11  10 9    5 4    0
+ * +-----+---+-------------+------+------+--------+---+------+------+
+ * | 0 1 | U | 1 1 1 1 1 0 | immh | immb | opcode | 1 |  Rn  |  Rd  |
+ * +-----+---+-------------+------+------+--------+---+------+------+
+ */
+static void disas_simd_scalar_shift_imm(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.10 AdvSIMD scalar three different
+ *  31 30  29 28       24 23  22  21 20  16 15    12 11 10 9    5 4    0
+ * +-----+---+-----------+------+---+------+--------+-----+------+------+
+ * | 0 1 | U | 1 1 1 1 0 | size | 1 |  Rm  | opcode | 0 0 |  Rn  |  Rd  |
+ * +-----+---+-----------+------+---+------+--------+-----+------+------+
+ */
+static void disas_simd_scalar_three_reg_diff(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.11 AdvSIMD scalar three same
+ *  31 30  29 28       24 23  22  21 20  16 15    11  10 9    5 4    0
+ * +-----+---+-----------+------+---+------+--------+---+------+------+
+ * | 0 1 | U | 1 1 1 1 0 | size | 1 |  Rm  | opcode | 1 |  Rn  |  Rd  |
+ * +-----+---+-----------+------+---+------+--------+---+------+------+
+ */
+static void disas_simd_scalar_three_reg_same(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.12 AdvSIMD scalar two reg misc
+ *  31 30  29 28       24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +-----+---+-----------+------+-----------+--------+-----+------+------+
+ * | 0 1 | U | 1 1 1 1 0 | size | 1 0 0 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +-----+---+-----------+------+-----------+--------+-----+------+------+
+ */
+static void disas_simd_scalar_two_reg_misc(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.13 AdvSIMD scalar x indexed element
+ *  31 30  29 28       24 23  22 21  20  19  16 15 12  11  10 9    5 4    0
+ * +-----+---+-----------+------+---+---+------+-----+---+---+------+------+
+ * | 0 1 | U | 1 1 1 1 1 | size | L | M |  Rm  | opc | H | 0 |  Rn  |  Rd  |
+ * +-----+---+-----------+------+---+---+------+-----+---+---+------+------+
+ */
+static void disas_simd_scalar_indexed(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.14 AdvSIMD shift by immediate
+ *  31  30   29 28         23 22  19 18  16 15    11  10 9    5 4    0
+ * +---+---+---+-------------+------+------+--------+---+------+------+
+ * | 0 | Q | U | 0 1 1 1 1 0 | immh | immb | opcode | 1 |  Rn  |  Rd  |
+ * +---+---+---+-------------+------+------+--------+---+------+------+
+ */
+static void disas_simd_shift_imm(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.15 AdvSIMD three different
+ *   31  30  29 28       24 23  22  21 20  16 15    12 11 10 9    5 4    0
+ * +---+---+---+-----------+------+---+------+--------+-----+------+------+
+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 |  Rm  | opcode | 0 0 |  Rn  |  Rd  |
+ * +---+---+---+-----------+------+---+------+--------+-----+------+------+
+ */
+static void disas_simd_three_reg_diff(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.16 AdvSIMD three same
+ *  31  30  29  28       24 23  22  21 20  16 15    11  10 9    5 4    0
+ * +---+---+---+-----------+------+---+------+--------+---+------+------+
+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 |  Rm  | opcode | 1 |  Rn  |  Rd  |
+ * +---+---+---+-----------+------+---+------+--------+---+------+------+
+ */
+static void disas_simd_three_reg_same(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.17 AdvSIMD two reg misc
+ *   31  30  29 28       24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +---+---+---+-----------+------+-----------+--------+-----+------+------+
+ * | 0 | Q | U | 0 1 1 1 0 | size | 1 0 0 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +---+---+---+-----------+------+-----------+--------+-----+------+------+
+ */
+static void disas_simd_two_reg_misc(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.18 AdvSIMD vector x indexed element
+ *   31  30  29 28       24 23  22 21  20  19  16 15 12  11  10 9    5 4    0
+ * +---+---+---+-----------+------+---+---+------+-----+---+---+------+------+
+ * | 0 | Q | U | 0 1 1 1 1 | size | L | M |  Rm  | opc | H | 0 |  Rn  |  Rd  |
+ * +---+---+---+-----------+------+---+---+------+-----+---+---+------+------+
+ */
+static void disas_simd_indexed_vector(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.19 Crypto AES
+ *  31             24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +-----------------+------+-----------+--------+-----+------+------+
+ * | 0 1 0 0 1 1 1 0 | size | 1 0 1 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +-----------------+------+-----------+--------+-----+------+------+
+ */
+static void disas_crypto_aes(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.20 Crypto three-reg SHA
+ *  31             24 23  22  21 20  16  15 14    12 11 10 9    5 4    0
+ * +-----------------+------+---+------+---+--------+-----+------+------+
+ * | 0 1 0 1 1 1 1 0 | size | 0 |  Rm  | 0 | opcode | 0 0 |  Rn  |  Rd  |
+ * +-----------------+------+---+------+---+--------+-----+------+------+
+ */
+static void disas_crypto_three_reg_sha(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6.21 Crypto two-reg SHA
+ *  31             24 23  22 21       17 16    12 11 10 9    5 4    0
+ * +-----------------+------+-----------+--------+-----+------+------+
+ * | 0 1 0 1 1 1 1 0 | size | 1 0 1 0 0 | opcode | 1 0 |  Rn  |  Rd  |
+ * +-----------------+------+-----------+--------+-----+------+------+
+ */
+static void disas_crypto_two_reg_sha(DisasContext *s, uint32_t insn)
+{
+    unsupported_encoding(s, insn);
+}
+
+/* C3.6 Data processing - SIMD, inc Crypto
+ *
+ * As the decode gets a little complex we are using a table based
+ * approach for this part of the decode.
+ */
+static const AArch64DecodeTable data_proc_simd[] = {
+    /* pattern  ,  mask     ,  fn                        */
+    { 0x0e200400, 0x9f200400, disas_simd_three_reg_same },
+    { 0x0e200000, 0x9f200c00, disas_simd_three_reg_diff },
+    { 0x0e200800, 0x9f3e0c00, disas_simd_two_reg_misc },
+    { 0x0e300800, 0x9f3e0c00, disas_simd_across_lanes },
+    { 0x0e000400, 0x9fe08400, disas_simd_copy },
+    { 0x0f000000, 0x9f000400, disas_simd_indexed_vector },
+    /* simd_mod_imm decode is a subset of simd_shift_imm, so must precede it */
+    { 0x0f000400, 0x9ff80400, disas_simd_mod_imm },
+    { 0x0f000400, 0x9f800400, disas_simd_shift_imm },
+    { 0x0e000000, 0xbf208c00, disas_simd_tb },
+    { 0x0e000800, 0xbf208c00, disas_simd_zip_trn },
+    { 0x2e000000, 0xbf208400, disas_simd_ext },
+    { 0x5e200400, 0xdf200400, disas_simd_scalar_three_reg_same },
+    { 0x5e200000, 0xdf200c00, disas_simd_scalar_three_reg_diff },
+    { 0x5e200800, 0xdf3e0c00, disas_simd_scalar_two_reg_misc },
+    { 0x5e300800, 0xdf3e0c00, disas_simd_scalar_pairwise },
+    { 0x5e000400, 0xdfe08400, disas_simd_scalar_copy },
+    { 0x5f000000, 0xdf000400, disas_simd_scalar_indexed },
+    { 0x5f000400, 0xdf800400, disas_simd_scalar_shift_imm },
+    { 0x4e280800, 0xff3e0c00, disas_crypto_aes },
+    { 0x5e000000, 0xff208c00, disas_crypto_three_reg_sha },
+    { 0x5e280800, 0xff3e0c00, disas_crypto_two_reg_sha },
+    { 0x00000000, 0x00000000, NULL }
+};
+
 static void disas_data_proc_simd(DisasContext *s, uint32_t insn)
 {
     /* Note that this is called with all non-FP cases from
      * table C3-6 so it must UNDEF for entries not specifically
      * allocated to instructions in that table.
      */
-    unsupported_encoding(s, insn);
+    AArch64DecodeFn *fn = lookup_disas_fn(&data_proc_simd[0], insn);
+    if (fn) {
+        fn(s, insn);
+    } else {
+        unallocated_encoding(s);
+    }
 }
 
 /* C3.6 Data processing - SIMD and floating point */
-- 
1.8.5

[Qemu-devel] [PATCH v2 04/10] target-arm: A64: Add SIMD EXT

[Peter Maydell]

Add support for the SIMD EXT instruction (the only one in its
group, C3.6.1).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
v1->v2:
 * reworked to do the vector element reads outside do_ext64;
   this turned out to be a substantial enough change that I haven't
   carried rth's reviewed-by tag across
---
 target-arm/translate-a64.c | 80 +++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 79 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 8060c9f..219af78 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -4646,6 +4646,25 @@ static void disas_data_proc_fp(DisasContext *s, uint32_t insn)
     }
 }
 
+static void do_ext64(DisasContext *s, TCGv_i64 tcg_left, TCGv_i64 tcg_right,
+                     int pos)
+{
+    /* Extract 64 bits from the middle of two concatenated 64 bit
+     * vector register slices left:right. The extracted bits start
+     * at 'pos' bits into the right (least significant) side.
+     * We return the result in tcg_right, and guarantee not to
+     * trash tcg_left.
+     */
+    TCGv_i64 tcg_tmp = tcg_temp_new_i64();
+    assert(pos > 0 && pos < 64);
+
+    tcg_gen_shri_i64(tcg_right, tcg_right, pos);
+    tcg_gen_shli_i64(tcg_tmp, tcg_left, 64 - pos);
+    tcg_gen_or_i64(tcg_right, tcg_right, tcg_tmp);
+
+    tcg_temp_free_i64(tcg_tmp);
+}
+
 /* C3.6.1 EXT
  *   31  30 29         24 23 22  21 20  16 15  14  11 10  9    5 4    0
  * +---+---+-------------+-----+---+------+---+------+---+------+------+
@@ -4654,7 +4673,66 @@ static void disas_data_proc_fp(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_ext(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int is_q = extract32(insn, 30, 1);
+    int op2 = extract32(insn, 22, 2);
+    int imm4 = extract32(insn, 11, 4);
+    int rm = extract32(insn, 16, 5);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    int pos = imm4 << 3;
+    TCGv_i64 tcg_resl, tcg_resh;
+
+    if (op2 != 0 || (!is_q && extract32(imm4, 3, 1))) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    tcg_resh = tcg_temp_new_i64();
+    tcg_resl = tcg_temp_new_i64();
+
+    /* Vd gets bits starting at pos bits into Vm:Vn. This is
+     * either extracting 128 bits from a 128:128 concatenation, or
+     * extracting 64 bits from a 64:64 concatenation.
+     */
+    if (!is_q) {
+        read_vec_element(s, tcg_resl, rn, 0, MO_64);
+        if (pos != 0) {
+            read_vec_element(s, tcg_resh, rm, 0, MO_64);
+            read_vec_element(s, tcg_resl, rn, 0, MO_64);
+            do_ext64(s, tcg_resh, tcg_resl, pos);
+        }
+        tcg_gen_movi_i64(tcg_resh, 0);
+    } else {
+        TCGv_i64 tcg_hh;
+        typedef struct {
+            int reg;
+            int elt;
+        } EltPosns;
+        EltPosns eltposns[] = { {rn, 0}, {rn, 1}, {rm, 0}, {rm, 1} };
+        EltPosns *elt = eltposns;
+
+        if (pos >= 64) {
+            elt++;
+            pos -= 64;
+        }
+
+        read_vec_element(s, tcg_resl, elt->reg, elt->elt, MO_64);
+        elt++;
+        read_vec_element(s, tcg_resh, elt->reg, elt->elt, MO_64);
+        elt++;
+        if (pos != 0) {
+            do_ext64(s, tcg_resh, tcg_resl, pos);
+            tcg_hh = tcg_temp_new_i64();
+            read_vec_element(s, tcg_hh, elt->reg, elt->elt, MO_64);
+            do_ext64(s, tcg_hh, tcg_resh, pos);
+            tcg_temp_free_i64(tcg_hh);
+        }
+    }
+
+    write_vec_element(s, tcg_resl, rd, 0, MO_64);
+    tcg_temp_free_i64(tcg_resl);
+    write_vec_element(s, tcg_resh, rd, 1, MO_64);
+    tcg_temp_free_i64(tcg_resh);
 }
 
 /* C3.6.2 TBL/TBX
-- 
1.8.5

[Qemu-devel] [PATCH v2 05/10] target-arm: A64: Add SIMD TBL/TBLX

[Peter Maydell]

From: Michael Matz <matz@suse.de>

Add support for the SIMD TBL/TBLX instructions (group C3.6.2).

Signed-off-by: Michael Matz <matz@suse.de>
[PMM: rewritten to do more of the decode in translate-a64.c,
 and to do only one 64 bit pass at a time in the helper]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
v1->v2: change rn and numregs args to helper to i32
---
 target-arm/helper-a64.c    | 31 ++++++++++++++++++++++++++
 target-arm/helper-a64.h    |  1 +
 target-arm/translate-a64.c | 55 +++++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 86 insertions(+), 1 deletion(-)

diff --git a/target-arm/helper-a64.c b/target-arm/helper-a64.c
index 4ce0d01..6ca958a 100644
--- a/target-arm/helper-a64.c
+++ b/target-arm/helper-a64.c
@@ -122,3 +122,34 @@ uint64_t HELPER(vfp_cmped_a64)(float64 x, float64 y, void *fp_status)
 {
     return float_rel_to_flags(float64_compare(x, y, fp_status));
 }
+
+uint64_t HELPER(simd_tbl)(CPUARMState *env, uint64_t result, uint64_t indices,
+                          uint32_t rn, uint32_t numregs)
+{
+    /* Helper function for SIMD TBL and TBX. We have to do the table
+     * lookup part for the 64 bits worth of indices we're passed in.
+     * result is the initial results vector (either zeroes for TBL
+     * or some guest values for TBX), rn the register number where
+     * the table starts, and numregs the number of registers in the table.
+     * We return the results of the lookups.
+     */
+    int shift;
+
+    for (shift = 0; shift < 64; shift += 8) {
+        int index = extract64(indices, shift, 8);
+        if (index < 16 * numregs) {
+            /* Convert index (a byte offset into the virtual table
+             * which is a series of 128-bit vectors concatenated)
+             * into the correct vfp.regs[] element plus a bit offset
+             * into that element, bearing in mind that the table
+             * can wrap around from V31 to V0.
+             */
+            int elt = (rn * 2 + (index >> 3)) % 64;
+            int bitidx = (index & 7) * 8;
+            uint64_t val = extract64(env->vfp.regs[elt], bitidx, 8);
+
+            result = deposit64(result, shift, 8, val);
+        }
+    }
+    return result;
+}
diff --git a/target-arm/helper-a64.h b/target-arm/helper-a64.h
index bca19f3..99832ee 100644
--- a/target-arm/helper-a64.h
+++ b/target-arm/helper-a64.h
@@ -26,3 +26,4 @@ DEF_HELPER_3(vfp_cmps_a64, i64, f32, f32, ptr)
 DEF_HELPER_3(vfp_cmpes_a64, i64, f32, f32, ptr)
 DEF_HELPER_3(vfp_cmpd_a64, i64, f64, f64, ptr)
 DEF_HELPER_3(vfp_cmped_a64, i64, f64, f64, ptr)
+DEF_HELPER_FLAGS_5(simd_tbl, TCG_CALL_NO_RWG_SE, i64, env, i64, i64, i32, i32)
diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 219af78..3eae175 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -4743,7 +4743,60 @@ static void disas_simd_ext(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_tb(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int op2 = extract32(insn, 22, 2);
+    int is_q = extract32(insn, 30, 1);
+    int rm = extract32(insn, 16, 5);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    int is_tblx = extract32(insn, 12, 1);
+    int len = extract32(insn, 13, 2);
+    TCGv_i64 tcg_resl, tcg_resh, tcg_idx;
+    TCGv_i32 tcg_regno, tcg_numregs;
+
+    if (op2 != 0) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    /* This does a table lookup: for every byte element in the input
+     * we index into a table formed from up to four vector registers,
+     * and then the output is the result of the lookups. Our helper
+     * function does the lookup operation for a single 64 bit part of
+     * the input.
+     */
+    tcg_resl = tcg_temp_new_i64();
+    tcg_resh = tcg_temp_new_i64();
+
+    if (is_tblx) {
+        read_vec_element(s, tcg_resl, rd, 0, MO_64);
+    } else {
+        tcg_gen_movi_i64(tcg_resl, 0);
+    }
+    if (is_tblx && is_q) {
+        read_vec_element(s, tcg_resh, rd, 1, MO_64);
+    } else {
+        tcg_gen_movi_i64(tcg_resh, 0);
+    }
+
+    tcg_idx = tcg_temp_new_i64();
+    tcg_regno = tcg_const_i32(rn);
+    tcg_numregs = tcg_const_i32(len + 1);
+    read_vec_element(s, tcg_idx, rm, 0, MO_64);
+    gen_helper_simd_tbl(tcg_resl, cpu_env, tcg_resl, tcg_idx,
+                        tcg_regno, tcg_numregs);
+    if (is_q) {
+        read_vec_element(s, tcg_idx, rm, 1, MO_64);
+        gen_helper_simd_tbl(tcg_resh, cpu_env, tcg_resh, tcg_idx,
+                            tcg_regno, tcg_numregs);
+    }
+    tcg_temp_free_i64(tcg_idx);
+    tcg_temp_free_i32(tcg_regno);
+    tcg_temp_free_i32(tcg_numregs);
+
+    write_vec_element(s, tcg_resl, rd, 0, MO_64);
+    tcg_temp_free_i64(tcg_resl);
+    write_vec_element(s, tcg_resh, rd, 1, MO_64);
+    tcg_temp_free_i64(tcg_resh);
 }
 
 /* C3.6.3 ZIP/UZP/TRN
-- 
1.8.5

[Qemu-devel] [PATCH v2 06/10] target-arm: A64: Add SIMD ZIP/UZP/TRN

[Peter Maydell]

From: Michael Matz <matz@suse.de>

Add support for the SIMD ZIP/UZIP/TRN instruction group
(C3.6.3).

Signed-off-by: Michael Matz <matz@suse.de>
[PMM: use new do_vec_get/set etc functions and generally update to new
 codebase standards; refactor to pull per-element loop outside switch]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
 target-arm/translate-a64.c | 76 +++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 75 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 3eae175..1c28c0f 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -4807,7 +4807,81 @@ static void disas_simd_tb(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rd = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int rm = extract32(insn, 16, 5);
+    int size = extract32(insn, 22, 2);
+    /* opc field bits [1:0] indicate ZIP/UZP/TRN;
+     * bit 2 indicates 1 vs 2 variant of the insn.
+     */
+    int opcode = extract32(insn, 12, 2);
+    bool part = extract32(insn, 14, 1);
+    bool is_q = extract32(insn, 30, 1);
+    int esize = 8 << size;
+    int i, ofs;
+    int datasize = is_q ? 128 : 64;
+    int elements = datasize / esize;
+    TCGv_i64 tcg_res, tcg_resl, tcg_resh;
+
+    if (opcode == 0 || (size == 3 && !is_q)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    tcg_resl = tcg_const_i64(0);
+    tcg_resh = tcg_const_i64(0);
+    tcg_res = tcg_temp_new_i64();
+
+    for (i = 0; i < elements; i++) {
+        switch (opcode) {
+        case 1: /* UZP1/2 */
+        {
+            int midpoint = elements / 2;
+            if (i < midpoint) {
+                read_vec_element(s, tcg_res, rn, 2 * i + part, size);
+            } else {
+                read_vec_element(s, tcg_res, rm,
+                                 2 * (i - midpoint) + part, size);
+            }
+            break;
+        }
+        case 2: /* TRN1/2 */
+            if (i & 1) {
+                read_vec_element(s, tcg_res, rm, (i & ~1) + part, size);
+            } else {
+                read_vec_element(s, tcg_res, rn, (i & ~1) + part, size);
+            }
+            break;
+        case 3: /* ZIP1/2 */
+        {
+            int base = part * elements / 2;
+            if (i & 1) {
+                read_vec_element(s, tcg_res, rm, base + (i >> 1), size);
+            } else {
+                read_vec_element(s, tcg_res, rn, base + (i >> 1), size);
+            }
+            break;
+        }
+        default:
+            g_assert_not_reached();
+        }
+
+        ofs = i * esize;
+        if (ofs < 64) {
+            tcg_gen_shli_i64(tcg_res, tcg_res, ofs);
+            tcg_gen_or_i64(tcg_resl, tcg_resl, tcg_res);
+        } else {
+            tcg_gen_shli_i64(tcg_res, tcg_res, ofs - 64);
+            tcg_gen_or_i64(tcg_resh, tcg_resh, tcg_res);
+        }
+    }
+
+    tcg_temp_free_i64(tcg_res);
+
+    write_vec_element(s, tcg_resl, rd, 0, MO_64);
+    tcg_temp_free_i64(tcg_resl);
+    write_vec_element(s, tcg_resh, rd, 1, MO_64);
+    tcg_temp_free_i64(tcg_resh);
 }
 
 /* C3.6.4 AdvSIMD across lanes
-- 
1.8.5

[Qemu-devel] [PATCH v2 07/10] target-arm: A64: Add SIMD across-lanes instructions

[Peter Maydell]

From: Michael Matz <matz@suse.de>

Add support for the SIMD "across lanes" instruction group (C3.6.4).

Signed-off-by: Michael Matz <matz@suse.de>
[PMM: Updated to current codebase, added fp min/max ops,
 added unallocated encoding checks]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
 target-arm/translate-a64.c | 177 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 176 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 1c28c0f..94f00c0 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -4884,6 +4884,29 @@ static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
     tcg_temp_free_i64(tcg_resh);
 }
 
+static void do_minmaxop(DisasContext *s, TCGv_i32 tcg_elt1, TCGv_i32 tcg_elt2,
+                        int opc, bool is_min, TCGv_ptr fpst)
+{
+    /* Helper function for disas_simd_across_lanes: do a single precision
+     * min/max operation on the specified two inputs,
+     * and return the result in tcg_elt1.
+     */
+    if (opc == 0xc) {
+        if (is_min) {
+            gen_helper_vfp_minnums(tcg_elt1, tcg_elt1, tcg_elt2, fpst);
+        } else {
+            gen_helper_vfp_maxnums(tcg_elt1, tcg_elt1, tcg_elt2, fpst);
+        }
+    } else {
+        assert(opc == 0xf);
+        if (is_min) {
+            gen_helper_vfp_mins(tcg_elt1, tcg_elt1, tcg_elt2, fpst);
+        } else {
+            gen_helper_vfp_maxs(tcg_elt1, tcg_elt1, tcg_elt2, fpst);
+        }
+    }
+}
+
 /* C3.6.4 AdvSIMD across lanes
  *   31  30  29 28       24 23  22 21       17 16    12 11 10 9    5 4    0
  * +---+---+---+-----------+------+-----------+--------+-----+------+------+
@@ -4892,7 +4915,159 @@ static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_across_lanes(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rd = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int size = extract32(insn, 22, 2);
+    int opcode = extract32(insn, 12, 5);
+    bool is_q = extract32(insn, 30, 1);
+    bool is_u = extract32(insn, 29, 1);
+    bool is_fp = false;
+    bool is_min = false;
+    int esize;
+    int elements;
+    int i;
+    TCGv_i64 tcg_res, tcg_elt;
+
+    switch (opcode) {
+    case 0x1b: /* ADDV */
+        if (is_u) {
+            unallocated_encoding(s);
+            return;
+        }
+        /* fall through */
+    case 0x3: /* SADDLV, UADDLV */
+    case 0xa: /* SMAXV, UMAXV */
+    case 0x1a: /* SMINV, UMINV */
+        if (size == 3 || (size == 2 && !is_q)) {
+            unallocated_encoding(s);
+            return;
+        }
+        break;
+    case 0xc: /* FMAXNMV, FMINNMV */
+    case 0xf: /* FMAXV, FMINV */
+        if (!is_u || !is_q || extract32(size, 0, 1)) {
+            unallocated_encoding(s);
+            return;
+        }
+        /* Bit 1 of size field encodes min vs max, and actual size is always
+         * 32 bits: adjust the size variable so following code can rely on it
+         */
+        is_min = extract32(size, 1, 1);
+        is_fp = true;
+        size = 2;
+        break;
+    default:
+        unallocated_encoding(s);
+        return;
+    }
+
+    esize = 8 << size;
+    elements = (is_q ? 128 : 64) / esize;
+
+    tcg_res = tcg_temp_new_i64();
+    tcg_elt = tcg_temp_new_i64();
+
+    /* These instructions operate across all lanes of a vector
+     * to produce a single result. We can guarantee that a 64
+     * bit intermediate is sufficient:
+     *  + for [US]ADDLV the maximum element size is 32 bits, and
+     *    the result type is 64 bits
+     *  + for FMAX*V, FMIN*V, ADDV the intermediate type is the
+     *    same as the element size, which is 32 bits at most
+     * For the integer operations we can choose to work at 64
+     * or 32 bits and truncate at the end; for simplicity
+     * we use 64 bits always. The floating point
+     * ops do require 32 bit intermediates, though.
+     */
+    if (!is_fp) {
+        read_vec_element(s, tcg_res, rn, 0, size | (is_u ? 0 : MO_SIGN));
+
+        for (i = 1; i < elements; i++) {
+            read_vec_element(s, tcg_elt, rn, i, size | (is_u ? 0 : MO_SIGN));
+
+            switch (opcode) {
+            case 0x03: /* SADDLV / UADDLV */
+            case 0x1b: /* ADDV */
+                tcg_gen_add_i64(tcg_res, tcg_res, tcg_elt);
+                break;
+            case 0x0a: /* SMAXV / UMAXV */
+                tcg_gen_movcond_i64(is_u ? TCG_COND_GEU : TCG_COND_GE,
+                                    tcg_res,
+                                    tcg_res, tcg_elt, tcg_res, tcg_elt);
+                break;
+            case 0x1a: /* SMINV / UMINV */
+                tcg_gen_movcond_i64(is_u ? TCG_COND_LEU : TCG_COND_LE,
+                                    tcg_res,
+                                    tcg_res, tcg_elt, tcg_res, tcg_elt);
+                break;
+                break;
+            default:
+                g_assert_not_reached();
+            }
+
+        }
+    } else {
+        /* Floating point ops which work on 32 bit (single) intermediates.
+         * Note that correct NaN propagation requires that we do these
+         * operations in exactly the order specified by the pseudocode.
+         */
+        TCGv_i32 tcg_elt1 = tcg_temp_new_i32();
+        TCGv_i32 tcg_elt2 = tcg_temp_new_i32();
+        TCGv_i32 tcg_elt3 = tcg_temp_new_i32();
+        TCGv_ptr fpst = get_fpstatus_ptr();
+
+        assert(esize == 32);
+        assert(elements == 4);
+
+        read_vec_element(s, tcg_elt, rn, 0, MO_32);
+        tcg_gen_trunc_i64_i32(tcg_elt1, tcg_elt);
+        read_vec_element(s, tcg_elt, rn, 1, MO_32);
+        tcg_gen_trunc_i64_i32(tcg_elt2, tcg_elt);
+
+        do_minmaxop(s, tcg_elt1, tcg_elt2, opcode, is_min, fpst);
+
+        read_vec_element(s, tcg_elt, rn, 2, MO_32);
+        tcg_gen_trunc_i64_i32(tcg_elt2, tcg_elt);
+        read_vec_element(s, tcg_elt, rn, 3, MO_32);
+        tcg_gen_trunc_i64_i32(tcg_elt3, tcg_elt);
+
+        do_minmaxop(s, tcg_elt2, tcg_elt3, opcode, is_min, fpst);
+
+        do_minmaxop(s, tcg_elt1, tcg_elt2, opcode, is_min, fpst);
+
+        tcg_gen_extu_i32_i64(tcg_res, tcg_elt1);
+        tcg_temp_free_i32(tcg_elt1);
+        tcg_temp_free_i32(tcg_elt2);
+        tcg_temp_free_i32(tcg_elt3);
+        tcg_temp_free_ptr(fpst);
+    }
+
+    tcg_temp_free_i64(tcg_elt);
+
+    /* Now truncate the result to the width required for the final output */
+    if (opcode == 0x03) {
+        /* SADDLV, UADDLV: result is 2*esize */
+        size++;
+    }
+
+    switch (size) {
+    case 0:
+        tcg_gen_ext8u_i64(tcg_res, tcg_res);
+        break;
+    case 1:
+        tcg_gen_ext16u_i64(tcg_res, tcg_res);
+        break;
+    case 2:
+        tcg_gen_ext32u_i64(tcg_res, tcg_res);
+        break;
+    case 3:
+        break;
+    default:
+        g_assert_not_reached();
+    }
+
+    write_fp_dreg(s, rd, tcg_res);
+    tcg_temp_free_i64(tcg_res);
 }
 
 /* C3.6.5 AdvSIMD copy
-- 
1.8.5

[Qemu-devel] [PATCH v2 08/10] target-arm: A64: Add SIMD copy operations

[Peter Maydell]

From: Alex Bennée <alex.bennee@linaro.org>

This adds support for the all the AdvSIMD vector copy operations
(ARM ARM 3.6.5).

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
 target-arm/translate-a64.c | 210 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 209 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 94f00c0..85daca9 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -5070,6 +5070,173 @@ static void disas_simd_across_lanes(DisasContext *s, uint32_t insn)
     tcg_temp_free_i64(tcg_res);
 }
 
+/* C6.3.31 DUP (Element, Vector)
+ *
+ *  31  30   29              21 20    16 15        10  9    5 4    0
+ * +---+---+-------------------+--------+-------------+------+------+
+ * | 0 | Q | 0 0 1 1 1 0 0 0 0 |  imm5  | 0 0 0 0 0 1 |  Rn  |  Rd  |
+ * +---+---+-------------------+--------+-------------+------+------+
+ *
+ * size: encoded in imm5 (see ARM ARM LowestSetBit())
+ */
+static void handle_simd_dupe(DisasContext *s, int is_q, int rd, int rn,
+                             int imm5)
+{
+    int size = ctz32(imm5);
+    int esize = 8 << size;
+    int elements = (is_q ? 128 : 64) / esize;
+    int index, i;
+    TCGv_i64 tmp;
+
+    if (size > 3 || (size == 3 && !is_q)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    index = imm5 >> (size + 1);
+
+    tmp = tcg_temp_new_i64();
+    read_vec_element(s, tmp, rn, index, size);
+
+    for (i = 0; i < elements; i++) {
+        write_vec_element(s, tmp, rd, i, size);
+    }
+
+    if (!is_q) {
+        clear_vec_high(s, rd);
+    }
+
+    tcg_temp_free_i64(tmp);
+}
+
+/* C6.3.32 DUP (General)
+ *
+ *  31  30   29              21 20    16 15        10  9    5 4    0
+ * +---+---+-------------------+--------+-------------+------+------+
+ * | 0 | Q | 0 0 1 1 1 0 0 0 0 |  imm5  | 0 0 0 0 1 1 |  Rn  |  Rd  |
+ * +---+---+-------------------+--------+-------------+------+------+
+ *
+ * size: encoded in imm5 (see ARM ARM LowestSetBit())
+ */
+static void handle_simd_dupg(DisasContext *s, int is_q, int rd, int rn,
+                             int imm5)
+{
+    int size = ctz32(imm5);
+    int esize = 8 << size;
+    int elements = (is_q ? 128 : 64)/esize;
+    int i = 0;
+
+    if (size > 3 || ((size == 3) && !is_q)) {
+        unallocated_encoding(s);
+        return;
+    }
+    for (i = 0; i < elements; i++) {
+        write_vec_element(s, cpu_reg(s, rn), rd, i, size);
+    }
+    if (!is_q) {
+        clear_vec_high(s, rd);
+    }
+}
+
+/* C6.3.150 INS (Element)
+ *
+ *  31                   21 20    16 15  14    11  10 9    5 4    0
+ * +-----------------------+--------+------------+---+------+------+
+ * | 0 1 1 0 1 1 1 0 0 0 0 |  imm5  | 0 |  imm4  | 1 |  Rn  |  Rd  |
+ * +-----------------------+--------+------------+---+------+------+
+ *
+ * size: encoded in imm5 (see ARM ARM LowestSetBit())
+ * index: encoded in imm5<4:size+1>
+ */
+static void handle_simd_inse(DisasContext *s, int rd, int rn,
+                             int imm4, int imm5)
+{
+    int size = ctz32(imm5);
+    int src_index, dst_index;
+    TCGv_i64 tmp;
+
+    if (size > 3) {
+        unallocated_encoding(s);
+        return;
+    }
+    dst_index = extract32(imm5, 1+size, 5);
+    src_index = extract32(imm4, size, 4);
+
+    tmp = tcg_temp_new_i64();
+
+    read_vec_element(s, tmp, rn, src_index, size);
+    write_vec_element(s, tmp, rd, dst_index, size);
+
+    tcg_temp_free_i64(tmp);
+}
+
+
+/* C6.3.151 INS (General)
+ *
+ *  31                   21 20    16 15        10  9    5 4    0
+ * +-----------------------+--------+-------------+------+------+
+ * | 0 1 0 0 1 1 1 0 0 0 0 |  imm5  | 0 0 0 1 1 1 |  Rn  |  Rd  |
+ * +-----------------------+--------+-------------+------+------+
+ *
+ * size: encoded in imm5 (see ARM ARM LowestSetBit())
+ * index: encoded in imm5<4:size+1>
+ */
+static void handle_simd_insg(DisasContext *s, int rd, int rn, int imm5)
+{
+    int size = ctz32(imm5);
+    int idx;
+
+    if (size > 3) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    idx = extract32(imm5, 1 + size, 4 - size);
+    write_vec_element(s, cpu_reg(s, rn), rd, idx, size);
+}
+
+/*
+ * C6.3.321 UMOV (General)
+ * C6.3.237 SMOV (General)
+ *
+ *  31  30   29              21 20    16 15    12   10 9    5 4    0
+ * +---+---+-------------------+--------+-------------+------+------+
+ * | 0 | Q | 0 0 1 1 1 0 0 0 0 |  imm5  | 0 0 1 U 1 1 |  Rn  |  Rd  |
+ * +---+---+-------------------+--------+-------------+------+------+
+ *
+ * U: unsigned when set
+ * size: encoded in imm5 (see ARM ARM LowestSetBit())
+ */
+static void handle_simd_umov_smov(DisasContext *s, int is_q, int is_signed,
+                                  int rn, int rd, int imm5)
+{
+    int size = ctz32(imm5);
+    int element;
+    TCGv_i64 tcg_rd;
+
+    /* Check for UnallocatedEncodings */
+    if (is_signed) {
+        if (size > 2 || (size == 2 && !is_q)) {
+            unallocated_encoding(s);
+            return;
+        }
+    } else {
+        if (size > 3
+            || (size < 3 && is_q)
+            || (size == 3 && !is_q)) {
+            unallocated_encoding(s);
+            return;
+        }
+    }
+    element = extract32(imm5, 1+size, 4);
+
+    tcg_rd = cpu_reg(s, rd);
+    read_vec_element(s, tcg_rd, rn, element, size | (is_signed ? MO_SIGN : 0));
+    if (is_signed && !is_q) {
+        tcg_gen_ext32u_i64(tcg_rd, tcg_rd);
+    }
+}
+
 /* C3.6.5 AdvSIMD copy
  *   31  30  29  28             21 20  16 15  14  11 10  9    5 4    0
  * +---+---+----+-----------------+------+---+------+---+------+------+
@@ -5078,7 +5245,48 @@ static void disas_simd_across_lanes(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_copy(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rd = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int imm4 = extract32(insn, 11, 4);
+    int op = extract32(insn, 29, 1);
+    int is_q = extract32(insn, 30, 1);
+    int imm5 = extract32(insn, 16, 5);
+
+    if (op) {
+        if (is_q) {
+            /* INS (element) */
+            handle_simd_inse(s, rd, rn, imm4, imm5);
+        } else {
+            unallocated_encoding(s);
+        }
+    } else {
+        switch (imm4) {
+        case 0:
+            /* DUP (element - vector) */
+            handle_simd_dupe(s, is_q, rd, rn, imm5);
+            break;
+        case 1:
+            /* DUP (general) */
+            handle_simd_dupg(s, is_q, rd, rn, imm5);
+            break;
+        case 3:
+            if (is_q) {
+                /* INS (general) */
+                handle_simd_insg(s, rd, rn, imm5);
+            } else {
+                unallocated_encoding(s);
+            }
+            break;
+        case 5:
+        case 7:
+            /* UMOV/SMOV (is_q indicates 32/64; imm4 indicates signedness) */
+            handle_simd_umov_smov(s, is_q, (imm4 == 5), rn, rd, imm5);
+            break;
+        default:
+            unallocated_encoding(s);
+            break;
+        }
+    }
 }
 
 /* C3.6.6 AdvSIMD modified immediate
-- 
1.8.5

[Qemu-devel] [PATCH v2 09/10] target-arm: A64: Add SIMD modified immediate group

[Peter Maydell]

From: Alex Bennée <alex.bennee@linaro.org>

This patch adds support for the AdvSIMD modified immediate group
(C3.6.6) with all its suboperations (movi, orr, fmov, mvni, bic).

Signed-off-by: Alexander Graf <agraf@suse.de>
[AJB: new decode struct, minor bug fixes, optimisation]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
v1->v2: use bitfield_replicate() rather than opencoded shift-and-or
---
 target-arm/translate-a64.c | 120 ++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 119 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 85daca9..5f56db9 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -5294,10 +5294,128 @@ static void disas_simd_copy(DisasContext *s, uint32_t insn)
  * +---+---+----+---------------------+-----+-------+----+---+-------+------+
  * | 0 | Q | op | 0 1 1 1 1 0 0 0 0 0 | abc | cmode | o2 | 1 | defgh |  Rd  |
  * +---+---+----+---------------------+-----+-------+----+---+-------+------+
+ *
+ * There are a number of operations that can be carried out here:
+ *   MOVI - move (shifted) imm into register
+ *   MVNI - move inverted (shifted) imm into register
+ *   ORR  - bitwise OR of (shifted) imm with register
+ *   BIC  - bitwise clear of (shifted) imm with register
  */
 static void disas_simd_mod_imm(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rd = extract32(insn, 0, 5);
+    int cmode = extract32(insn, 12, 4);
+    int cmode_3_1 = extract32(cmode, 1, 3);
+    int cmode_0 = extract32(cmode, 0, 1);
+    int o2 = extract32(insn, 11, 1);
+    uint64_t abcdefgh = extract32(insn, 5, 5) | (extract32(insn, 16, 3) << 5);
+    bool is_neg = extract32(insn, 29, 1);
+    bool is_q = extract32(insn, 30, 1);
+    uint64_t imm = 0;
+    TCGv_i64 tcg_rd, tcg_imm;
+    int i;
+
+    if (o2 != 0 || ((cmode == 0xf) && is_neg && !is_q)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    /* See AdvSIMDExpandImm() in ARM ARM */
+    switch (cmode_3_1) {
+    case 0: /* Replicate(Zeros(24):imm8, 2) */
+    case 1: /* Replicate(Zeros(16):imm8:Zeros(8), 2) */
+    case 2: /* Replicate(Zeros(8):imm8:Zeros(16), 2) */
+    case 3: /* Replicate(imm8:Zeros(24), 2) */
+    {
+        int shift = cmode_3_1 * 8;
+        imm = bitfield_replicate(abcdefgh << shift, 32);
+        break;
+    }
+    case 4: /* Replicate(Zeros(8):imm8, 4) */
+    case 5: /* Replicate(imm8:Zeros(8), 4) */
+    {
+        int shift = (cmode_3_1 & 0x1) * 8;
+        imm = bitfield_replicate(abcdefgh << shift, 16);
+        break;
+    }
+    case 6:
+        if (cmode_0) {
+            /* Replicate(Zeros(8):imm8:Ones(16), 2) */
+            imm = (abcdefgh << 16) | 0xffff;
+        } else {
+            /* Replicate(Zeros(16):imm8:Ones(8), 2) */
+            imm = (abcdefgh << 8) | 0xff;
+        }
+        imm = bitfield_replicate(imm, 32);
+        break;
+    case 7:
+        if (!cmode_0 && !is_neg) {
+            imm = bitfield_replicate(abcdefgh, 8);
+        } else if (!cmode_0 && is_neg) {
+            int i;
+            imm = 0;
+            for (i = 0; i < 8; i++) {
+                if ((abcdefgh) & (1 << i)) {
+                    imm |= 0xffULL << (i * 8);
+                }
+            }
+        } else if (cmode_0) {
+            if (is_neg) {
+                imm = (abcdefgh & 0x3f) << 48;
+                if (abcdefgh & 0x80) {
+                    imm |= 0x8000000000000000ULL;
+                }
+                if (abcdefgh & 0x40) {
+                    imm |= 0x3fc0000000000000ULL;
+                } else {
+                    imm |= 0x4000000000000000ULL;
+                }
+            } else {
+                imm = (abcdefgh & 0x3f) << 19;
+                if (abcdefgh & 0x80) {
+                    imm |= 0x80000000;
+                }
+                if (abcdefgh & 0x40) {
+                    imm |= 0x3e000000;
+                } else {
+                    imm |= 0x40000000;
+                }
+                imm |= (imm << 32);
+            }
+        }
+        break;
+    }
+
+    if (cmode_3_1 != 7 && is_neg) {
+        imm = ~imm;
+    }
+
+    tcg_imm = tcg_const_i64(imm);
+    tcg_rd = new_tmp_a64(s);
+
+    for (i = 0; i < 2; i++) {
+        int foffs = i ? fp_reg_hi_offset(rd) : fp_reg_offset(rd, MO_64);
+
+        if (i == 1 && !is_q) {
+            /* non-quad ops clear high half of vector */
+            tcg_gen_movi_i64(tcg_rd, 0);
+        } else if ((cmode & 0x9) == 0x1 || (cmode & 0xd) == 0x9) {
+            tcg_gen_ld_i64(tcg_rd, cpu_env, foffs);
+            if (is_neg) {
+                /* AND (BIC) */
+                tcg_gen_and_i64(tcg_rd, tcg_rd, tcg_imm);
+            } else {
+                /* ORR */
+                tcg_gen_or_i64(tcg_rd, tcg_rd, tcg_imm);
+            }
+        } else {
+            /* MOVI */
+            tcg_gen_mov_i64(tcg_rd, tcg_imm);
+        }
+        tcg_gen_st_i64(tcg_rd, cpu_env, foffs);
+    }
+
+    tcg_temp_free_i64(tcg_imm);
 }
 
 /* C3.6.7 AdvSIMD scalar copy
-- 
1.8.5

[Qemu-devel] [PATCH v2 10/10] target-arm: A64: Add SIMD scalar copy instructions

[Peter Maydell]

Add support for the SIMD scalar copy instruction group (C3.6.7),
which consists of the single instruction DUP (element, scalar).

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <rth@twiddle.net>
---
v1->v2: fixed error in insn diagram comment
---
 target-arm/translate-a64.c | 42 +++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 41 insertions(+), 1 deletion(-)

diff --git a/target-arm/translate-a64.c b/target-arm/translate-a64.c
index 5f56db9..2870675 100644
--- a/target-arm/translate-a64.c
+++ b/target-arm/translate-a64.c
@@ -5109,6 +5109,35 @@ static void handle_simd_dupe(DisasContext *s, int is_q, int rd, int rn,
     tcg_temp_free_i64(tmp);
 }
 
+/* C6.3.31 DUP (element, scalar)
+ *  31                   21 20    16 15        10  9    5 4    0
+ * +-----------------------+--------+-------------+------+------+
+ * | 0 1 0 1 1 1 1 0 0 0 0 |  imm5  | 0 0 0 0 0 1 |  Rn  |  Rd  |
+ * +-----------------------+--------+-------------+------+------+
+ */
+static void handle_simd_dupes(DisasContext *s, int rd, int rn,
+                              int imm5)
+{
+    int size = ctz32(imm5);
+    int index;
+    TCGv_i64 tmp;
+
+    if (size > 3) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    index = imm5 >> (size + 1);
+
+    /* This instruction just extracts the specified element and
+     * zero-extends it into the bottom of the destination register.
+     */
+    tmp = tcg_temp_new_i64();
+    read_vec_element(s, tmp, rn, index, size);
+    write_fp_dreg(s, rd, tmp);
+    tcg_temp_free_i64(tmp);
+}
+
 /* C6.3.32 DUP (General)
  *
  *  31  30   29              21 20    16 15        10  9    5 4    0
@@ -5426,7 +5455,18 @@ static void disas_simd_mod_imm(DisasContext *s, uint32_t insn)
  */
 static void disas_simd_scalar_copy(DisasContext *s, uint32_t insn)
 {
-    unsupported_encoding(s, insn);
+    int rd = extract32(insn, 0, 5);
+    int rn = extract32(insn, 5, 5);
+    int imm4 = extract32(insn, 11, 4);
+    int imm5 = extract32(insn, 16, 5);
+    int op = extract32(insn, 29, 1);
+
+    if (op != 0 || imm4 != 0) {
+        unallocated_encoding(s);
+    }
+
+    /* DUP (element, scalar) */
+    handle_simd_dupes(s, rd, rn, imm5);
 }
 
 /* C3.6.8 AdvSIMD scalar pairwise
-- 
1.8.5

Re: [Qemu-devel] [PATCH v2 01/10] target-arm: A64: Add SIMD ld/st multiple

[Richard Henderson]

On 01/13/2014 06:13 AM, Peter Maydell wrote:
> From: Alex Bennée <alex.bennee@linaro.org>
> 
> This adds support support for the SIMD load/store
> multiple category of instructions.
> 
> This also brings in a couple of helper functions for manipulating
> sections of the SIMD registers:
> 
>   * do_vec_get - fetch value from a slice of a vector register
>   * do_vec_set - set a slice of a vector register
> 
> which use vec_reg_offset for consistent processing of offsets in an
> endian aware manner. There are also additional helpers:
> 
>   * do_vec_ld - load value into SIMD
>   * do_vec_st - store value from SIMD
> 
> which load or store a slice of a vector register to memory.
> These don't zero extend like the fp variants.
> 
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
> public v1->v2 changes:
>  removed stray double-spacing
>  fixed incorrect handling of rn==sp, cleaned up tcg_rn code
> ---
>  target-arm/translate-a64.c | 250 ++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 248 insertions(+), 2 deletions(-)

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v2 02/10] target-arm: A64: Add SIMD ld/st single

[Richard Henderson]

On 01/13/2014 06:13 AM, Peter Maydell wrote:
> Implement the SIMD ld/st single structure instructions.
> 
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
> public v1->v2 changes:
>  * same tcg_rn cleanup/fix as ld/st multiple
> ---
>  target-arm/translate-a64.c | 144 ++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 142 insertions(+), 2 deletions(-)


Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v2 03/10] target-arm: A64: Add decode skeleton for SIMD data processing insns

[Richard Henderson]

On 01/13/2014 06:13 AM, Peter Maydell wrote:
> From: Alex Bennée <alex.bennee@linaro.org>
> 
> Add decode skeleton and function placeholders for all the SIMD data
> processing instructions. Due to the complexity of this part of the
> table the normal extract and switch approach gets very messy very
> quickly, so we use a simple data-driven pattern-and-mask approach.
> 
> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> Reviewed-by: Richard Henderson <rth@twiddle.net>
> ---
> v1->v2:
>  * made data_proc_simd table const
>  * fixed errors in insn diagram comments
>  * fixed coding style for fn pointer call
> ---
>  target-arm/translate-a64.c | 306 ++++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 305 insertions(+), 1 deletion(-)


Reviewed-by: Richard Henderson <rth@twiddle.net>


r~

Re: [Qemu-devel] [PATCH v2 04/10] target-arm: A64: Add SIMD EXT

[Richard Henderson]

On 01/13/2014 06:13 AM, Peter Maydell wrote:
> +    if (!is_q) {
> +        read_vec_element(s, tcg_resl, rn, 0, MO_64);
> +        if (pos != 0) {
> +            read_vec_element(s, tcg_resh, rm, 0, MO_64);
> +            read_vec_element(s, tcg_resl, rn, 0, MO_64);

Duplicate resl read.

Otherwise,

Reviewed-by: Richard Henderson <rth@twiddle.net>


r~