/*
 * Copyright © 2008 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 *
 * Author:
 *    Zou Nan hai <nanhai.zou@intel.com>
 *    Zhang Hua jun <huajun.zhang@intel.com>
 *    Xing Dong sheng <dongsheng.xing@intel.com>
 *
 */
mov (8) g115.0<1>UD g1.0<8,8,1>UD {align1};

and.nz (1) null g2.0<1,1,1>UD 0x1UD{align1};
(f0) jmpi direct_idct;
add (1) g2.8<1>UD g115.8<1,1,1>UD 0UD{align1};
send (16) 0 g3.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g4.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g5.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g6.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g7.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g8.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g9.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g10.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g11.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g12.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g13.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g14.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g15.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g16.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g17.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g18.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g19.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g20.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g21.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g22.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g23<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g24.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g25.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
add (1) g2.8<1>UD g2.8<1,1,1>UD 32UD {align1};
send (16) 0 g26.0<1>UD g2<16,16,1>UD read(3, 0, 2, 0) mlen 1 rlen 1 { align1 };
direct_idct:
mov (1) g1.8<1>UD 0x00F000FUD {align1};
mov.sat (16) g3.0<2>UB g3.0<16,16,1>W {align1};
mov.sat (16) g4.0<2>UB g4.0<16,16,1>W {align1};
mov.sat (16) g5.0<2>UB g5.0<16,16,1>W {align1};
mov.sat (16) g6.0<2>UB g6.0<16,16,1>W {align1};
mov.sat (16) g7.0<2>UB g7.0<16,16,1>W {align1};
mov.sat (16) g8.0<2>UB g8.0<16,16,1>W {align1};
mov.sat (16) g9.0<2>UB g9.0<16,16,1>W {align1};
mov.sat (16) g10.0<2>UB g10.0<16,16,1>W {align1};
mov.sat (16) g11.0<2>UB g11.0<16,16,1>W {align1};
mov.sat (16) g12.0<2>UB g12.0<16,16,1>W {align1};
mov.sat (16) g13.0<2>UB g13.0<16,16,1>W {align1};
mov.sat (16) g14.0<2>UB g14.0<16,16,1>W {align1};
mov.sat (16) g15.0<2>UB g15.0<16,16,1>W {align1};
mov.sat (16) g16.0<2>UB g16.0<16,16,1>W {align1};
mov.sat (16) g17.0<2>UB g17.0<16,16,1>W {align1};
mov.sat (16) g18.0<2>UB g18.0<16,16,1>W {align1};

and.nz (1) null g1.30<1,1,1>UB 0x1UW{align1};
(f0) jmpi field_dct;
//Frame IDCT
mov (16) m1.0<1>UB g3.0<16,16,2>UB {align1};
mov (16) m1.16<1>UB g4.0<16,16,2>UB {align1};
mov (16) m2.0<1>UB g5.0<16,16,2>UB {align1};
mov (16) m2.16<1>UB g6.0<16,16,2>UB {align1};
mov (16) m3.0<1>UB g7.0<16,16,2>UB {align1};
mov (16) m3.16<1>UB g8.0<16,16,2>UB {align1};
mov (16) m4.0<1>UB g9.0<16,16,2>UB {align1};
mov (16) m4.16<1>UB g10.0<16,16,2>UB {align1};
mov (16) m5.0<1>UB g11.0<16,16,2>UB {align1};
mov (16) m5.16<1>UB g12.0<16,16,2>UB {align1};
mov (16) m6.0<1>UB g13.0<16,16,2>UB {align1};
mov (16) m6.16<1>UB g14.0<16,16,2>UB {align1};
mov (16) m7.0<1>UB g15.0<16,16,2>UB {align1};
mov (16) m7.16<1>UB g16.0<16,16,2>UB {align1};
mov (16) m8.0<1>UB g17.0<16,16,2>UB {align1};
mov (16) m8.16<1>UB g18.0<16,16,2>UB {align1};
jmpi write_back;

field_dct:
//Field IDCT
mov (16) m1.0<1>UB g3.0<16,16,2>UB {align1};
mov (16) m1.16<1>UB g11.0<16,16,2>UB {align1};
mov (16) m2.0<1>UB g4.0<16,16,2>UB {align1};
mov (16) m2.16<1>UB g12.0<16,16,2>UB {align1};
mov (16) m3.0<1>UB g5.0<16,16,2>UB {align1};
mov (16) m3.16<1>UB g13.0<16,16,2>UB {align1};
mov (16) m4.0<1>UB g6.0<16,16,2>UB {align1};
mov (16) m4.16<1>UB g14.0<16,16,2>UB {align1};
mov (16) m5.0<1>UB g7.0<16,16,2>UB {align1};
mov (16) m5.16<1>UB g15.0<16,16,2>UB {align1};
mov (16) m6.0<1>UB g8.0<16,16,2>UB {align1};
mov (16) m6.16<1>UB g16.0<16,16,2>UB {align1};
mov (16) m7.0<1>UB g9.0<16,16,2>UB {align1};
mov (16) m7.16<1>UB g17.0<16,16,2>UB {align1};
mov (16) m8.0<1>UB g10.0<16,16,2>UB {align1};
mov (16) m8.16<1>UB g18.0<16,16,2>UB {align1};

write_back:
send (16) 0 acc0<1>UW g1<8,8,1>UW write(0,0,2,0) mlen 9 rlen 0 {align1};
//U
mov (1) g1.8<1>UD 0x0070007UD  { align1 };
shr (2) g1.0<1>UD g115.0<2,2,1>UD 1D {align1};
mov.sat (16) g19.0<2>UB g19.0<16,16,1>W {align1};
mov.sat (16) g20.0<2>UB g20.0<16,16,1>W {align1};
mov.sat (16) g21.0<2>UB g21.0<16,16,1>W {align1};
mov.sat (16) g22.0<2>UB g22.0<16,16,1>W {align1};
mov (16) m1.0<1>UB g19.0<16,16,2>UB {align1};
mov (16) m1.16<1>UB g20.0<16,16,2>UB {align1};
mov (16) m2.0<1>UB g21.0<16,16,2>UB {align1};
mov (16) m2.16<1>UB g22.0<16,16,2>UB {align1};
send (16) 0 acc0<1>UW g1<8,8,1>UW write(1, 0, 2, 0) mlen 3 rlen 0 { align1 };
//V
mov.sat (16) g23.0<2>UB g23.0<16,16,1>W {align1};
mov.sat (16) g24.0<2>UB g24.0<16,16,1>W {align1};
mov.sat (16) g25.0<2>UB g25.0<16,16,1>W {align1};
mov.sat (16) g26.0<2>UB g26.0<16,16,1>W {align1};
mov (16) m1.0<1>UB g23.0<16,16,2>UB {align1};
mov (16) m1.16<1>UB g24.0<16,16,2>UB {align1};
mov (16) m2.0<1>UB g25.0<16,16,2>UB {align1};
mov (16) m2.16<1>UB g26.0<16,16,2>UB {align1};
send (16) 0 acc0<1>UW g1<8,8,1>UW write(2, 0, 2, 0) mlen 3 rlen 0 { align1 };

send (16) 0 acc0<1>UW g0<8,8,1>UW
        thread_spawner(0, 0, 0) mlen 1 rlen 0 { align1 EOT};