(* Auto-generate ARM Neon intrinsics header file.
|
(* Auto-generate ARM Neon intrinsics header file.
|
Copyright (C) 2006, 2007, 2009 Free Software Foundation, Inc.
|
Copyright (C) 2006, 2007, 2009 Free Software Foundation, Inc.
|
Contributed by CodeSourcery.
|
Contributed by CodeSourcery.
|
|
|
This file is part of GCC.
|
This file is part of GCC.
|
|
|
GCC is free software; you can redistribute it and/or modify it under
|
GCC is free software; you can redistribute it and/or modify it under
|
the terms of the GNU General Public License as published by the Free
|
the terms of the GNU General Public License as published by the Free
|
Software Foundation; either version 3, or (at your option) any later
|
Software Foundation; either version 3, or (at your option) any later
|
version.
|
version.
|
|
|
GCC is distributed in the hope that it will be useful, but WITHOUT ANY
|
GCC is distributed in the hope that it will be useful, but WITHOUT ANY
|
WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
|
FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
|
for more details.
|
for more details.
|
|
|
You should have received a copy of the GNU General Public License
|
You should have received a copy of the GNU General Public License
|
along with GCC; see the file COPYING3. If not see
|
along with GCC; see the file COPYING3. If not see
|
.
|
.
|
|
|
This is an O'Caml program. The O'Caml compiler is available from:
|
This is an O'Caml program. The O'Caml compiler is available from:
|
|
|
http://caml.inria.fr/
|
http://caml.inria.fr/
|
|
|
Or from your favourite OS's friendly packaging system. Tested with version
|
Or from your favourite OS's friendly packaging system. Tested with version
|
3.09.2, though other versions will probably work too.
|
3.09.2, though other versions will probably work too.
|
|
|
Compile with:
|
Compile with:
|
ocamlc -c neon.ml
|
ocamlc -c neon.ml
|
ocamlc -o neon-gen neon.cmo neon-gen.ml
|
ocamlc -o neon-gen neon.cmo neon-gen.ml
|
|
|
Run with:
|
Run with:
|
./neon-gen > arm_neon.h
|
./neon-gen > arm_neon.h
|
*)
|
*)
|
|
|
open Neon
|
open Neon
|
|
|
(* The format codes used in the following functions are documented at:
|
(* The format codes used in the following functions are documented at:
|
http://caml.inria.fr/pub/docs/manual-ocaml/libref/Format.html\
|
http://caml.inria.fr/pub/docs/manual-ocaml/libref/Format.html\
|
#6_printflikefunctionsforprettyprinting
|
#6_printflikefunctionsforprettyprinting
|
(one line, remove the backslash.)
|
(one line, remove the backslash.)
|
*)
|
*)
|
|
|
(* Following functions can be used to approximate GNU indentation style. *)
|
(* Following functions can be used to approximate GNU indentation style. *)
|
let start_function () =
|
let start_function () =
|
Format.printf "@[";
|
Format.printf "@[";
|
ref 0
|
ref 0
|
|
|
let end_function nesting =
|
let end_function nesting =
|
match !nesting with
|
match !nesting with
|
0 -> Format.printf "@;@;@]"
|
0 -> Format.printf "@;@;@]"
|
| _ -> failwith ("Bad nesting (ending function at level "
|
| _ -> failwith ("Bad nesting (ending function at level "
|
^ (string_of_int !nesting) ^ ")")
|
^ (string_of_int !nesting) ^ ")")
|
|
|
let open_braceblock nesting =
|
let open_braceblock nesting =
|
begin match !nesting with
|
begin match !nesting with
|
0 -> Format.printf "@,@<0>{@[@,"
|
0 -> Format.printf "@,@<0>{@[@,"
|
| _ -> Format.printf "@,@[ @<0>{@[@,"
|
| _ -> Format.printf "@,@[ @<0>{@[@,"
|
end;
|
end;
|
incr nesting
|
incr nesting
|
|
|
let close_braceblock nesting =
|
let close_braceblock nesting =
|
decr nesting;
|
decr nesting;
|
match !nesting with
|
match !nesting with
|
0 -> Format.printf "@]@,@<0>}"
|
0 -> Format.printf "@]@,@<0>}"
|
| _ -> Format.printf "@]@,@<0>}@]"
|
| _ -> Format.printf "@]@,@<0>}@]"
|
|
|
let print_function arity fnname body =
|
let print_function arity fnname body =
|
let ffmt = start_function () in
|
let ffmt = start_function () in
|
Format.printf "__extension__ static __inline ";
|
Format.printf "__extension__ static __inline ";
|
let inl = "__attribute__ ((__always_inline__))" in
|
let inl = "__attribute__ ((__always_inline__))" in
|
begin match arity with
|
begin match arity with
|
Arity0 ret ->
|
Arity0 ret ->
|
Format.printf "%s %s@,%s (void)" (string_of_vectype ret) inl fnname
|
Format.printf "%s %s@,%s (void)" (string_of_vectype ret) inl fnname
|
| Arity1 (ret, arg0) ->
|
| Arity1 (ret, arg0) ->
|
Format.printf "%s %s@,%s (%s __a)" (string_of_vectype ret) inl fnname
|
Format.printf "%s %s@,%s (%s __a)" (string_of_vectype ret) inl fnname
|
(string_of_vectype arg0)
|
(string_of_vectype arg0)
|
| Arity2 (ret, arg0, arg1) ->
|
| Arity2 (ret, arg0, arg1) ->
|
Format.printf "%s %s@,%s (%s __a, %s __b)"
|
Format.printf "%s %s@,%s (%s __a, %s __b)"
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype arg1)
|
(string_of_vectype arg1)
|
| Arity3 (ret, arg0, arg1, arg2) ->
|
| Arity3 (ret, arg0, arg1, arg2) ->
|
Format.printf "%s %s@,%s (%s __a, %s __b, %s __c)"
|
Format.printf "%s %s@,%s (%s __a, %s __b, %s __c)"
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype arg1) (string_of_vectype arg2)
|
(string_of_vectype arg1) (string_of_vectype arg2)
|
| Arity4 (ret, arg0, arg1, arg2, arg3) ->
|
| Arity4 (ret, arg0, arg1, arg2, arg3) ->
|
Format.printf "%s %s@,%s (%s __a, %s __b, %s __c, %s __d)"
|
Format.printf "%s %s@,%s (%s __a, %s __b, %s __c, %s __d)"
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype ret) inl fnname (string_of_vectype arg0)
|
(string_of_vectype arg1) (string_of_vectype arg2)
|
(string_of_vectype arg1) (string_of_vectype arg2)
|
(string_of_vectype arg3)
|
(string_of_vectype arg3)
|
end;
|
end;
|
open_braceblock ffmt;
|
open_braceblock ffmt;
|
let rec print_lines = function
|
let rec print_lines = function
|
[] -> ()
|
[] -> ()
|
| [line] -> Format.printf "%s" line
|
| [line] -> Format.printf "%s" line
|
| line::lines -> Format.printf "%s@," line; print_lines lines in
|
| line::lines -> Format.printf "%s@," line; print_lines lines in
|
print_lines body;
|
print_lines body;
|
close_braceblock ffmt;
|
close_braceblock ffmt;
|
end_function ffmt
|
end_function ffmt
|
|
|
let return_by_ptr features = List.mem ReturnPtr features
|
let return_by_ptr features = List.mem ReturnPtr features
|
|
|
let union_string num elts base =
|
let union_string num elts base =
|
let itype = inttype_for_array num elts in
|
let itype = inttype_for_array num elts in
|
let iname = string_of_inttype itype
|
let iname = string_of_inttype itype
|
and sname = string_of_vectype (T_arrayof (num, elts)) in
|
and sname = string_of_vectype (T_arrayof (num, elts)) in
|
Printf.sprintf "union { %s __i; %s __o; } %s" sname iname base
|
Printf.sprintf "union { %s __i; %s __o; } %s" sname iname base
|
|
|
let rec signed_ctype = function
|
let rec signed_ctype = function
|
T_uint8x8 | T_poly8x8 -> T_int8x8
|
T_uint8x8 | T_poly8x8 -> T_int8x8
|
| T_uint8x16 | T_poly8x16 -> T_int8x16
|
| T_uint8x16 | T_poly8x16 -> T_int8x16
|
| T_uint16x4 | T_poly16x4 -> T_int16x4
|
| T_uint16x4 | T_poly16x4 -> T_int16x4
|
| T_uint16x8 | T_poly16x8 -> T_int16x8
|
| T_uint16x8 | T_poly16x8 -> T_int16x8
|
| T_uint32x2 -> T_int32x2
|
| T_uint32x2 -> T_int32x2
|
| T_uint32x4 -> T_int32x4
|
| T_uint32x4 -> T_int32x4
|
| T_uint64x1 -> T_int64x1
|
| T_uint64x1 -> T_int64x1
|
| T_uint64x2 -> T_int64x2
|
| T_uint64x2 -> T_int64x2
|
(* Cast to types defined by mode in arm.c, not random types pulled in from
|
(* Cast to types defined by mode in arm.c, not random types pulled in from
|
the header in use. This fixes incompatible pointer errors when
|
the header in use. This fixes incompatible pointer errors when
|
compiling with C++. *)
|
compiling with C++. *)
|
| T_uint8 | T_int8 -> T_intQI
|
| T_uint8 | T_int8 -> T_intQI
|
| T_uint16 | T_int16 -> T_intHI
|
| T_uint16 | T_int16 -> T_intHI
|
| T_uint32 | T_int32 -> T_intSI
|
| T_uint32 | T_int32 -> T_intSI
|
| T_uint64 | T_int64 -> T_intDI
|
| T_uint64 | T_int64 -> T_intDI
|
| T_float32 -> T_floatSF
|
| T_float32 -> T_floatSF
|
| T_poly8 -> T_intQI
|
| T_poly8 -> T_intQI
|
| T_poly16 -> T_intHI
|
| T_poly16 -> T_intHI
|
| T_arrayof (n, elt) -> T_arrayof (n, signed_ctype elt)
|
| T_arrayof (n, elt) -> T_arrayof (n, signed_ctype elt)
|
| T_ptrto elt -> T_ptrto (signed_ctype elt)
|
| T_ptrto elt -> T_ptrto (signed_ctype elt)
|
| T_const elt -> T_const (signed_ctype elt)
|
| T_const elt -> T_const (signed_ctype elt)
|
| x -> x
|
| x -> x
|
|
|
let add_cast ctype cval =
|
let add_cast ctype cval =
|
let stype = signed_ctype ctype in
|
let stype = signed_ctype ctype in
|
if ctype <> stype then
|
if ctype <> stype then
|
Printf.sprintf "(%s) %s" (string_of_vectype stype) cval
|
Printf.sprintf "(%s) %s" (string_of_vectype stype) cval
|
else
|
else
|
cval
|
cval
|
|
|
let cast_for_return to_ty = "(" ^ (string_of_vectype to_ty) ^ ")"
|
let cast_for_return to_ty = "(" ^ (string_of_vectype to_ty) ^ ")"
|
|
|
(* Return a tuple of a list of declarations to go at the start of the function,
|
(* Return a tuple of a list of declarations to go at the start of the function,
|
and a list of statements needed to return THING. *)
|
and a list of statements needed to return THING. *)
|
let return arity return_by_ptr thing =
|
let return arity return_by_ptr thing =
|
match arity with
|
match arity with
|
Arity0 (ret) | Arity1 (ret, _) | Arity2 (ret, _, _) | Arity3 (ret, _, _, _)
|
Arity0 (ret) | Arity1 (ret, _) | Arity2 (ret, _, _) | Arity3 (ret, _, _, _)
|
| Arity4 (ret, _, _, _, _) ->
|
| Arity4 (ret, _, _, _, _) ->
|
match ret with
|
match ret with
|
T_arrayof (num, vec) ->
|
T_arrayof (num, vec) ->
|
if return_by_ptr then
|
if return_by_ptr then
|
let sname = string_of_vectype ret in
|
let sname = string_of_vectype ret in
|
[Printf.sprintf "%s __rv;" sname],
|
[Printf.sprintf "%s __rv;" sname],
|
[thing ^ ";"; "return __rv;"]
|
[thing ^ ";"; "return __rv;"]
|
else
|
else
|
let uname = union_string num vec "__rv" in
|
let uname = union_string num vec "__rv" in
|
[uname ^ ";"], ["__rv.__o = " ^ thing ^ ";"; "return __rv.__i;"]
|
[uname ^ ";"], ["__rv.__o = " ^ thing ^ ";"; "return __rv.__i;"]
|
| T_void -> [], [thing ^ ";"]
|
| T_void -> [], [thing ^ ";"]
|
| _ ->
|
| _ ->
|
[], ["return " ^ (cast_for_return ret) ^ thing ^ ";"]
|
[], ["return " ^ (cast_for_return ret) ^ thing ^ ";"]
|
|
|
let rec element_type ctype =
|
let rec element_type ctype =
|
match ctype with
|
match ctype with
|
T_arrayof (_, v) -> element_type v
|
T_arrayof (_, v) -> element_type v
|
| _ -> ctype
|
| _ -> ctype
|
|
|
let params return_by_ptr ps =
|
let params return_by_ptr ps =
|
let pdecls = ref [] in
|
let pdecls = ref [] in
|
let ptype t p =
|
let ptype t p =
|
match t with
|
match t with
|
T_arrayof (num, elts) ->
|
T_arrayof (num, elts) ->
|
let uname = union_string num elts (p ^ "u") in
|
let uname = union_string num elts (p ^ "u") in
|
let decl = Printf.sprintf "%s = { %s };" uname p in
|
let decl = Printf.sprintf "%s = { %s };" uname p in
|
pdecls := decl :: !pdecls;
|
pdecls := decl :: !pdecls;
|
p ^ "u.__o"
|
p ^ "u.__o"
|
| _ -> add_cast t p in
|
| _ -> add_cast t p in
|
let plist = match ps with
|
let plist = match ps with
|
Arity0 _ -> []
|
Arity0 _ -> []
|
| Arity1 (_, t1) -> [ptype t1 "__a"]
|
| Arity1 (_, t1) -> [ptype t1 "__a"]
|
| Arity2 (_, t1, t2) -> [ptype t1 "__a"; ptype t2 "__b"]
|
| Arity2 (_, t1, t2) -> [ptype t1 "__a"; ptype t2 "__b"]
|
| Arity3 (_, t1, t2, t3) -> [ptype t1 "__a"; ptype t2 "__b"; ptype t3 "__c"]
|
| Arity3 (_, t1, t2, t3) -> [ptype t1 "__a"; ptype t2 "__b"; ptype t3 "__c"]
|
| Arity4 (_, t1, t2, t3, t4) ->
|
| Arity4 (_, t1, t2, t3, t4) ->
|
[ptype t1 "__a"; ptype t2 "__b"; ptype t3 "__c"; ptype t4 "__d"] in
|
[ptype t1 "__a"; ptype t2 "__b"; ptype t3 "__c"; ptype t4 "__d"] in
|
match ps with
|
match ps with
|
Arity0 ret | Arity1 (ret, _) | Arity2 (ret, _, _) | Arity3 (ret, _, _, _)
|
Arity0 ret | Arity1 (ret, _) | Arity2 (ret, _, _) | Arity3 (ret, _, _, _)
|
| Arity4 (ret, _, _, _, _) ->
|
| Arity4 (ret, _, _, _, _) ->
|
if return_by_ptr then
|
if return_by_ptr then
|
!pdecls, add_cast (T_ptrto (element_type ret)) "&__rv.val[0]" :: plist
|
!pdecls, add_cast (T_ptrto (element_type ret)) "&__rv.val[0]" :: plist
|
else
|
else
|
!pdecls, plist
|
!pdecls, plist
|
|
|
let modify_params features plist =
|
let modify_params features plist =
|
let is_flipped =
|
let is_flipped =
|
List.exists (function Flipped _ -> true | _ -> false) features in
|
List.exists (function Flipped _ -> true | _ -> false) features in
|
if is_flipped then
|
if is_flipped then
|
match plist with
|
match plist with
|
[ a; b ] -> [ b; a ]
|
[ a; b ] -> [ b; a ]
|
| _ ->
|
| _ ->
|
failwith ("Don't know how to flip args " ^ (String.concat ", " plist))
|
failwith ("Don't know how to flip args " ^ (String.concat ", " plist))
|
else
|
else
|
plist
|
plist
|
|
|
(* !!! Decide whether to add an extra information word based on the shape
|
(* !!! Decide whether to add an extra information word based on the shape
|
form. *)
|
form. *)
|
let extra_word shape features paramlist bits =
|
let extra_word shape features paramlist bits =
|
let use_word =
|
let use_word =
|
match shape with
|
match shape with
|
All _ | Long | Long_noreg _ | Wide | Wide_noreg _ | Narrow
|
All _ | Long | Long_noreg _ | Wide | Wide_noreg _ | Narrow
|
| By_scalar _ | Wide_scalar | Wide_lane | Binary_imm _ | Long_imm
|
| By_scalar _ | Wide_scalar | Wide_lane | Binary_imm _ | Long_imm
|
| Narrow_imm -> true
|
| Narrow_imm -> true
|
| _ -> List.mem InfoWord features
|
| _ -> List.mem InfoWord features
|
in
|
in
|
if use_word then
|
if use_word then
|
paramlist @ [string_of_int bits]
|
paramlist @ [string_of_int bits]
|
else
|
else
|
paramlist
|
paramlist
|
|
|
(* Bit 0 represents signed (1) vs unsigned (0), or float (1) vs poly (0).
|
(* Bit 0 represents signed (1) vs unsigned (0), or float (1) vs poly (0).
|
Bit 1 represents floats & polynomials (1), or ordinary integers (0).
|
Bit 1 represents floats & polynomials (1), or ordinary integers (0).
|
Bit 2 represents rounding (1) vs none (0). *)
|
Bit 2 represents rounding (1) vs none (0). *)
|
let infoword_value elttype features =
|
let infoword_value elttype features =
|
let bits01 =
|
let bits01 =
|
match elt_class elttype with
|
match elt_class elttype with
|
Signed | ConvClass (Signed, _) | ConvClass (_, Signed) -> 0b001
|
Signed | ConvClass (Signed, _) | ConvClass (_, Signed) -> 0b001
|
| Poly -> 0b010
|
| Poly -> 0b010
|
| Float -> 0b011
|
| Float -> 0b011
|
| _ -> 0b000
|
| _ -> 0b000
|
and rounding_bit = if List.mem Rounding features then 0b100 else 0b000 in
|
and rounding_bit = if List.mem Rounding features then 0b100 else 0b000 in
|
bits01 lor rounding_bit
|
bits01 lor rounding_bit
|
|
|
(* "Cast" type operations will throw an exception in mode_of_elt (actually in
|
(* "Cast" type operations will throw an exception in mode_of_elt (actually in
|
elt_width, called from there). Deal with that here, and generate a suffix
|
elt_width, called from there). Deal with that here, and generate a suffix
|
with multiple modes (). *)
|
with multiple modes (). *)
|
let rec mode_suffix elttype shape =
|
let rec mode_suffix elttype shape =
|
try
|
try
|
let mode = mode_of_elt elttype shape in
|
let mode = mode_of_elt elttype shape in
|
string_of_mode mode
|
string_of_mode mode
|
with MixedMode (dst, src) ->
|
with MixedMode (dst, src) ->
|
let dstmode = mode_of_elt dst shape
|
let dstmode = mode_of_elt dst shape
|
and srcmode = mode_of_elt src shape in
|
and srcmode = mode_of_elt src shape in
|
string_of_mode dstmode ^ string_of_mode srcmode
|
string_of_mode dstmode ^ string_of_mode srcmode
|
|
|
let print_variant opcode features shape name (ctype, asmtype, elttype) =
|
let print_variant opcode features shape name (ctype, asmtype, elttype) =
|
let bits = infoword_value elttype features in
|
let bits = infoword_value elttype features in
|
let modesuf = mode_suffix elttype shape in
|
let modesuf = mode_suffix elttype shape in
|
let return_by_ptr = return_by_ptr features in
|
let return_by_ptr = return_by_ptr features in
|
let pdecls, paramlist = params return_by_ptr ctype in
|
let pdecls, paramlist = params return_by_ptr ctype in
|
let paramlist' = modify_params features paramlist in
|
let paramlist' = modify_params features paramlist in
|
let paramlist'' = extra_word shape features paramlist' bits in
|
let paramlist'' = extra_word shape features paramlist' bits in
|
let parstr = String.concat ", " paramlist'' in
|
let parstr = String.concat ", " paramlist'' in
|
let builtin = Printf.sprintf "__builtin_neon_%s%s (%s)"
|
let builtin = Printf.sprintf "__builtin_neon_%s%s (%s)"
|
(builtin_name features name) modesuf parstr in
|
(builtin_name features name) modesuf parstr in
|
let rdecls, stmts = return ctype return_by_ptr builtin in
|
let rdecls, stmts = return ctype return_by_ptr builtin in
|
let body = pdecls @ rdecls @ stmts
|
let body = pdecls @ rdecls @ stmts
|
and fnname = (intrinsic_name name) ^ "_" ^ (string_of_elt elttype) in
|
and fnname = (intrinsic_name name) ^ "_" ^ (string_of_elt elttype) in
|
print_function ctype fnname body
|
print_function ctype fnname body
|
|
|
(* When this function processes the element types in the ops table, it rewrites
|
(* When this function processes the element types in the ops table, it rewrites
|
them in a list of tuples (a,b,c):
|
them in a list of tuples (a,b,c):
|
a : C type as an "arity", e.g. Arity1 (T_poly8x8, T_poly8x8)
|
a : C type as an "arity", e.g. Arity1 (T_poly8x8, T_poly8x8)
|
b : Asm type : a single, processed element type, e.g. P16. This is the
|
b : Asm type : a single, processed element type, e.g. P16. This is the
|
type which should be attached to the asm opcode.
|
type which should be attached to the asm opcode.
|
c : Variant type : the unprocessed type for this variant (e.g. in add
|
c : Variant type : the unprocessed type for this variant (e.g. in add
|
instructions which don't care about the sign, b might be i16 and c
|
instructions which don't care about the sign, b might be i16 and c
|
might be s16.)
|
might be s16.)
|
*)
|
*)
|
|
|
let print_op (opcode, features, shape, name, munge, types) =
|
let print_op (opcode, features, shape, name, munge, types) =
|
let sorted_types = List.sort compare types in
|
let sorted_types = List.sort compare types in
|
let munged_types = List.map
|
let munged_types = List.map
|
(fun elt -> let c, asm = munge shape elt in c, asm, elt) sorted_types in
|
(fun elt -> let c, asm = munge shape elt in c, asm, elt) sorted_types in
|
List.iter
|
List.iter
|
(fun variant -> print_variant opcode features shape name variant)
|
(fun variant -> print_variant opcode features shape name variant)
|
munged_types
|
munged_types
|
|
|
let print_ops ops =
|
let print_ops ops =
|
List.iter print_op ops
|
List.iter print_op ops
|
|
|
(* Output type definitions. Table entries are:
|
(* Output type definitions. Table entries are:
|
cbase : "C" name for the type.
|
cbase : "C" name for the type.
|
abase : "ARM" base name for the type (i.e. int in int8x8_t).
|
abase : "ARM" base name for the type (i.e. int in int8x8_t).
|
esize : element size.
|
esize : element size.
|
enum : element count.
|
enum : element count.
|
*)
|
*)
|
|
|
let deftypes () =
|
let deftypes () =
|
let typeinfo = [
|
let typeinfo = [
|
(* Doubleword vector types. *)
|
(* Doubleword vector types. *)
|
"__builtin_neon_qi", "int", 8, 8;
|
"__builtin_neon_qi", "int", 8, 8;
|
"__builtin_neon_hi", "int", 16, 4;
|
"__builtin_neon_hi", "int", 16, 4;
|
"__builtin_neon_si", "int", 32, 2;
|
"__builtin_neon_si", "int", 32, 2;
|
"__builtin_neon_di", "int", 64, 1;
|
"__builtin_neon_di", "int", 64, 1;
|
"__builtin_neon_sf", "float", 32, 2;
|
"__builtin_neon_sf", "float", 32, 2;
|
"__builtin_neon_poly8", "poly", 8, 8;
|
"__builtin_neon_poly8", "poly", 8, 8;
|
"__builtin_neon_poly16", "poly", 16, 4;
|
"__builtin_neon_poly16", "poly", 16, 4;
|
"__builtin_neon_uqi", "uint", 8, 8;
|
"__builtin_neon_uqi", "uint", 8, 8;
|
"__builtin_neon_uhi", "uint", 16, 4;
|
"__builtin_neon_uhi", "uint", 16, 4;
|
"__builtin_neon_usi", "uint", 32, 2;
|
"__builtin_neon_usi", "uint", 32, 2;
|
"__builtin_neon_udi", "uint", 64, 1;
|
"__builtin_neon_udi", "uint", 64, 1;
|
|
|
(* Quadword vector types. *)
|
(* Quadword vector types. *)
|
"__builtin_neon_qi", "int", 8, 16;
|
"__builtin_neon_qi", "int", 8, 16;
|
"__builtin_neon_hi", "int", 16, 8;
|
"__builtin_neon_hi", "int", 16, 8;
|
"__builtin_neon_si", "int", 32, 4;
|
"__builtin_neon_si", "int", 32, 4;
|
"__builtin_neon_di", "int", 64, 2;
|
"__builtin_neon_di", "int", 64, 2;
|
"__builtin_neon_sf", "float", 32, 4;
|
"__builtin_neon_sf", "float", 32, 4;
|
"__builtin_neon_poly8", "poly", 8, 16;
|
"__builtin_neon_poly8", "poly", 8, 16;
|
"__builtin_neon_poly16", "poly", 16, 8;
|
"__builtin_neon_poly16", "poly", 16, 8;
|
"__builtin_neon_uqi", "uint", 8, 16;
|
"__builtin_neon_uqi", "uint", 8, 16;
|
"__builtin_neon_uhi", "uint", 16, 8;
|
"__builtin_neon_uhi", "uint", 16, 8;
|
"__builtin_neon_usi", "uint", 32, 4;
|
"__builtin_neon_usi", "uint", 32, 4;
|
"__builtin_neon_udi", "uint", 64, 2
|
"__builtin_neon_udi", "uint", 64, 2
|
] in
|
] in
|
List.iter
|
List.iter
|
(fun (cbase, abase, esize, enum) ->
|
(fun (cbase, abase, esize, enum) ->
|
let attr =
|
let attr =
|
match enum with
|
match enum with
|
1 -> ""
|
1 -> ""
|
| _ -> Printf.sprintf "\t__attribute__ ((__vector_size__ (%d)))"
|
| _ -> Printf.sprintf "\t__attribute__ ((__vector_size__ (%d)))"
|
(esize * enum / 8) in
|
(esize * enum / 8) in
|
Format.printf "typedef %s %s%dx%d_t%s;@\n" cbase abase esize enum attr)
|
Format.printf "typedef %s %s%dx%d_t%s;@\n" cbase abase esize enum attr)
|
typeinfo;
|
typeinfo;
|
Format.print_newline ();
|
Format.print_newline ();
|
(* Extra types not in . *)
|
(* Extra types not in . *)
|
Format.printf "typedef float float32_t;\n";
|
Format.printf "typedef float float32_t;\n";
|
Format.printf "typedef __builtin_neon_poly8 poly8_t;\n";
|
Format.printf "typedef __builtin_neon_poly8 poly8_t;\n";
|
Format.printf "typedef __builtin_neon_poly16 poly16_t;\n"
|
Format.printf "typedef __builtin_neon_poly16 poly16_t;\n"
|
|
|
(* Output structs containing arrays, for load & store instructions etc. *)
|
(* Output structs containing arrays, for load & store instructions etc. *)
|
|
|
let arrtypes () =
|
let arrtypes () =
|
let typeinfo = [
|
let typeinfo = [
|
"int", 8; "int", 16;
|
"int", 8; "int", 16;
|
"int", 32; "int", 64;
|
"int", 32; "int", 64;
|
"uint", 8; "uint", 16;
|
"uint", 8; "uint", 16;
|
"uint", 32; "uint", 64;
|
"uint", 32; "uint", 64;
|
"float", 32; "poly", 8;
|
"float", 32; "poly", 8;
|
"poly", 16
|
"poly", 16
|
] in
|
] in
|
let writestruct elname elsize regsize arrsize =
|
let writestruct elname elsize regsize arrsize =
|
let elnum = regsize / elsize in
|
let elnum = regsize / elsize in
|
let structname =
|
let structname =
|
Printf.sprintf "%s%dx%dx%d_t" elname elsize elnum arrsize in
|
Printf.sprintf "%s%dx%dx%d_t" elname elsize elnum arrsize in
|
let sfmt = start_function () in
|
let sfmt = start_function () in
|
Format.printf "typedef struct %s" structname;
|
Format.printf "typedef struct %s" structname;
|
open_braceblock sfmt;
|
open_braceblock sfmt;
|
Format.printf "%s%dx%d_t val[%d];" elname elsize elnum arrsize;
|
Format.printf "%s%dx%d_t val[%d];" elname elsize elnum arrsize;
|
close_braceblock sfmt;
|
close_braceblock sfmt;
|
Format.printf " %s;" structname;
|
Format.printf " %s;" structname;
|
end_function sfmt;
|
end_function sfmt;
|
in
|
in
|
for n = 2 to 4 do
|
for n = 2 to 4 do
|
List.iter
|
List.iter
|
(fun (elname, elsize) ->
|
(fun (elname, elsize) ->
|
writestruct elname elsize 64 n;
|
writestruct elname elsize 64 n;
|
writestruct elname elsize 128 n)
|
writestruct elname elsize 128 n)
|
typeinfo
|
typeinfo
|
done
|
done
|
|
|
let print_lines = List.iter (fun s -> Format.printf "%s@\n" s)
|
let print_lines = List.iter (fun s -> Format.printf "%s@\n" s)
|
|
|
(* Do it. *)
|
(* Do it. *)
|
|
|
let _ =
|
let _ =
|
print_lines [
|
print_lines [
|
"/* ARM NEON intrinsics include file. This file is generated automatically";
|
"/* ARM NEON intrinsics include file. This file is generated automatically";
|
" using neon-gen.ml. Please do not edit manually.";
|
" using neon-gen.ml. Please do not edit manually.";
|
"";
|
"";
|
" Copyright (C) 2006, 2007, 2009 Free Software Foundation, Inc.";
|
" Copyright (C) 2006, 2007, 2009 Free Software Foundation, Inc.";
|
" Contributed by CodeSourcery.";
|
" Contributed by CodeSourcery.";
|
"";
|
"";
|
" This file is part of GCC.";
|
" This file is part of GCC.";
|
"";
|
"";
|
" GCC is free software; you can redistribute it and/or modify it";
|
" GCC is free software; you can redistribute it and/or modify it";
|
" under the terms of the GNU General Public License as published";
|
" under the terms of the GNU General Public License as published";
|
" by the Free Software Foundation; either version 3, or (at your";
|
" by the Free Software Foundation; either version 3, or (at your";
|
" option) any later version.";
|
" option) any later version.";
|
"";
|
"";
|
" GCC is distributed in the hope that it will be useful, but WITHOUT";
|
" GCC is distributed in the hope that it will be useful, but WITHOUT";
|
" ANY WARRANTY; without even the implied warranty of MERCHANTABILITY";
|
" ANY WARRANTY; without even the implied warranty of MERCHANTABILITY";
|
" or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public";
|
" or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public";
|
" License for more details.";
|
" License for more details.";
|
"";
|
"";
|
" Under Section 7 of GPL version 3, you are granted additional";
|
" Under Section 7 of GPL version 3, you are granted additional";
|
" permissions described in the GCC Runtime Library Exception, version";
|
" permissions described in the GCC Runtime Library Exception, version";
|
" 3.1, as published by the Free Software Foundation.";
|
" 3.1, as published by the Free Software Foundation.";
|
"";
|
"";
|
" You should have received a copy of the GNU General Public License and";
|
" You should have received a copy of the GNU General Public License and";
|
" a copy of the GCC Runtime Library Exception along with this program;";
|
" a copy of the GCC Runtime Library Exception along with this program;";
|
" see the files COPYING3 and COPYING.RUNTIME respectively. If not, see";
|
" see the files COPYING3 and COPYING.RUNTIME respectively. If not, see";
|
" . */";
|
" . */";
|
"";
|
"";
|
"#ifndef _GCC_ARM_NEON_H";
|
"#ifndef _GCC_ARM_NEON_H";
|
"#define _GCC_ARM_NEON_H 1";
|
"#define _GCC_ARM_NEON_H 1";
|
"";
|
"";
|
"#ifndef __ARM_NEON__";
|
"#ifndef __ARM_NEON__";
|
"#error You must enable NEON instructions (e.g. -mfloat-abi=softfp -mfpu=neon) to use arm_neon.h";
|
"#error You must enable NEON instructions (e.g. -mfloat-abi=softfp -mfpu=neon) to use arm_neon.h";
|
"#else";
|
"#else";
|
"";
|
"";
|
"#ifdef __cplusplus";
|
"#ifdef __cplusplus";
|
"extern \"C\" {";
|
"extern \"C\" {";
|
"#endif";
|
"#endif";
|
"";
|
"";
|
"#include ";
|
"#include ";
|
""];
|
""];
|
deftypes ();
|
deftypes ();
|
arrtypes ();
|
arrtypes ();
|
Format.print_newline ();
|
Format.print_newline ();
|
print_ops ops;
|
print_ops ops;
|
Format.print_newline ();
|
Format.print_newline ();
|
print_ops reinterp;
|
print_ops reinterp;
|
print_lines [
|
print_lines [
|
"#ifdef __cplusplus";
|
"#ifdef __cplusplus";
|
"}";
|
"}";
|
"#endif";
|
"#endif";
|
"#endif";
|
"#endif";
|
"#endif"]
|
"#endif"]
|
|
|
