///***************************************************************************** //* //* Copyright (C) 2012 Ittiam Systems Pvt Ltd, Bangalore //* //* Licensed under the Apache License, Version 2.0 (the "License"); //* you may not use this file except in compliance with the License. //* You may obtain a copy of the License at: //* //* http://www.apache.org/licenses/LICENSE-2.0 //* //* Unless required by applicable law or agreed to in writing, software //* distributed under the License is distributed on an "AS IS" BASIS, //* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. //* See the License for the specific language governing permissions and //* limitations under the License. //* //*****************************************************************************/ ///** //******************************************************************************* //* @file //* ihevc_intra_pred_luma_mode_27_to_33.s //* //* @brief //* contains function definition for intra prediction interpolation filters //* //* //* @author //* parthiban v //* //* @par list of functions: //* - ihevc_intra_pred_luma_mode_27_to_33() //* //* @remarks //* none //* //******************************************************************************* //*/ // ///** //******************************************************************************* //* //* @brief //* intra prediction interpolation filter for luma mode 27 to mode 33 //* //* @par description: //* intraprediction for mode 27 to 33 (positive angle, vertical mode ) with //* .extern neighboring samples location pointed by 'pu1_ref' to the tu //* block location pointed by 'pu1_dst' //* //* @param[in] pu1_src //* uword8 pointer to the source //* //* @param[out] pu1_dst //* uword8 pointer to the destination //* //* @param[in] src_strd //* integer source stride //* //* @param[in] dst_strd //* integer destination stride //* //* @param[in] nt //* integer transform block size //* //* @param[in] mode //* integer intraprediction mode //* //* @returns //* //* @remarks //* none //* //******************************************************************************* //*/ //void ihevc_intra_pred_luma_mode_27_to_33(uword8 *pu1_ref, // word32 src_strd, // uword8 *pu1_dst, // word32 dst_strd, // word32 nt, // word32 mode) //**************variables vs registers***************************************** //x0 => *pu1_ref //x1 => src_strd //x2 => *pu1_dst //x3 => dst_strd .text .align 4 .include "ihevc_neon_macros.s" .globl ihevc_intra_pred_luma_mode_27_to_33_av8 .extern gai4_ihevc_ang_table .extern gau1_ihevc_planar_factor .type ihevc_intra_pred_luma_mode_27_to_33_av8, %function ihevc_intra_pred_luma_mode_27_to_33_av8: // stmfd sp!, {x4-x12, x14} //stack stores the values of the arguments stp d9,d10,[sp,#-16]! stp d12,d13,[sp,#-16]! stp d14,d15,[sp,#-16]! stp x19, x20,[sp,#-16]! adrp x6, :got:gai4_ihevc_ang_table //loads word32 gai4_ihevc_ang_table[35] ldr x6, [x6, #:got_lo12:gai4_ihevc_ang_table] lsl x7,x4,#1 //two_nt add x8,x6,x5,lsl #2 //*gai4_ihevc_ang_table[mode] ldr w9, [x8] //intra_pred_ang = gai4_ihevc_ang_table[mode] sxtw x9,w9 adrp x1, :got:gau1_ihevc_planar_factor //used for ((row + 1) * intra_pred_ang) row values ldr x1, [x1, #:got_lo12:gau1_ihevc_planar_factor] add x6,x1,#1 tst x4,#7 add x8,x0,x7 //pu1_ref + two_nt mov x14,#0 //row mov x12,x4 bne core_loop_4 core_loop_8: add x8,x8,#1 //pu1_ref_main_idx += (two_nt + 1) dup v0.8b,w9 //intra_pred_ang lsr x12, x4, #3 //divide by 8 movi v1.8b, #32 mul x7, x4, x12 movi v6.8h, #31 //lsl x12,x3,#3 mov x1,x8 //sub x12,x12,x4 mov x5,x4 mov x11,#1 prologue: ld1 {v3.8b},[x6] //loads the row value umull v2.8h, v3.8b, v0.8b //pos = ((row + 1) * intra_pred_ang) and v4.16b, v2.16b , v6.16b //dup_const_fract(fract = pos & (31)) xtn v4.8b, v4.8h shrn v5.8b, v2.8h,#5 //idx = pos >> 5 dup v31.8b, v4.b[0] add x0,x2,x3 umov w14, v5.s[0] //(i row)extract idx to the r register sxtw x14,w14 dup v29.8b, v4.b[1] //(ii) and x9,x14,#0xff //(i row) get the last byte add x10,x8,x9 //(i row)*pu1_ref[ref_main_idx] asr x14,x14,#8 //(ii)shift by 8 ld1 {v23.8b},[x10],x11 //(i row)ref_main_idx and x9,x14,#0xff //(ii)get the last byte asr x14,x14,#8 //(iii) ld1 {v9.8b},[x10] //(i row)ref_main_idx_1 add x12,x8,x9 //(ii)*pu1_ref[ref_main_idx] and x9,x14,#0xff //(iii) sub v30.8b, v1.8b , v31.8b //32-fract(dup_const_32_fract) add x10,x8,x9 //(iii)*pu1_ref[ref_main_idx] ld1 {v12.8b},[x12],x11 //(ii)ref_main_idx umull v10.8h, v23.8b, v30.8b //(i row)vmull_u8(ref_main_idx, dup_const_32_fract) ld1 {v13.8b},[x12] //(ii)ref_main_idx_1 umlal v10.8h, v9.8b, v31.8b //(i row)vmull_u8(ref_main_idx_1, dup_const_fract) asr x14,x14,#8 //(iv) dup v27.8b, v4.b[2] //(iii) sub v28.8b, v1.8b , v29.8b //(ii)32-fract(dup_const_32_fract) and x9,x14,#0xff //(iv) dup v25.8b, v4.b[3] //(iv) umull v14.8h, v12.8b, v28.8b //(ii)vmull_u8(ref_main_idx, dup_const_32_fract) add x12,x8,x9 //(iv)*pu1_ref[ref_main_idx] ld1 {v16.8b},[x10],x11 //(iii)ref_main_idx umlal v14.8h, v13.8b, v29.8b //(ii)vmull_u8(ref_main_idx_1, dup_const_fract) ld1 {v17.8b},[x10] //(iii)ref_main_idx_1 rshrn v10.8b, v10.8h,#5 //(i row)shift_res = vrshrn_n_u16(add_res, 5) ld1 {v20.8b},[x12],x11 //(iv)ref_main_idx sub v26.8b, v1.8b , v27.8b //(iii)32-fract(dup_const_32_fract) ld1 {v21.8b},[x12] //(iv)ref_main_idx_1 dup v31.8b, v4.b[4] //(v) umull v18.8h, v16.8b, v26.8b //(iii)vmull_u8(ref_main_idx, dup_const_32_fract) umov w14, v5.s[1] //extract idx to the r register sxtw x14,w14 umlal v18.8h, v17.8b, v27.8b //(iii)vmull_u8(ref_main_idx_1, dup_const_fract) st1 {v10.8b},[x2],#8 //(i row) rshrn v14.8b, v14.8h,#5 //(ii)shift_res = vrshrn_n_u16(add_res, 5) and x9,x14,#0xff //(v) dup v29.8b, v4.b[5] //(vi) add x10,x8,x9 //(v)*pu1_ref[ref_main_idx] ld1 {v23.8b},[x10],x11 //(v)ref_main_idx sub v24.8b, v1.8b , v25.8b //(iv)32-fract(dup_const_32_fract) asr x14,x14,#8 //(vi) umull v22.8h, v20.8b, v24.8b //(iv)vmull_u8(ref_main_idx, dup_const_32_fract) and x9,x14,#0xff //(vi) ld1 {v9.8b},[x10] //(v)ref_main_idx_1 umlal v22.8h, v21.8b, v25.8b //(iv)vmull_u8(ref_main_idx_1, dup_const_fract) st1 {v14.8b},[x0],x3 //(ii) rshrn v18.8b, v18.8h,#5 //(iii)shift_res = vrshrn_n_u16(add_res, 5) add x12,x8,x9 //(vi)*pu1_ref[ref_main_idx] dup v27.8b, v4.b[6] //(vii) asr x14,x14,#8 //(vii) and x9,x14,#0xff //(vii) sub v30.8b, v1.8b , v31.8b //(v)32-fract(dup_const_32_fract) add x10,x8,x9 //(vii)*pu1_ref[ref_main_idx] ld1 {v12.8b},[x12],x11 //(vi)ref_main_idx umull v10.8h, v23.8b, v30.8b //(v)vmull_u8(ref_main_idx, dup_const_32_fract) ld1 {v13.8b},[x12] //(vi)ref_main_idx_1 umlal v10.8h, v9.8b, v31.8b //(v)vmull_u8(ref_main_idx_1, dup_const_fract) st1 {v18.8b},[x0],x3 //(iii) rshrn v22.8b, v22.8h,#5 //(iv)shift_res = vrshrn_n_u16(add_res, 5) asr x14,x14,#8 //(viii) dup v25.8b, v4.b[7] //(viii) and x9,x14,#0xff //(viii) ld1 {v16.8b},[x10],x11 //(vii)ref_main_idx sub v28.8b, v1.8b , v29.8b //(vi)32-fract(dup_const_32_fract) ld1 {v17.8b},[x10] //(vii)ref_main_idx_1 umull v14.8h, v12.8b, v28.8b //(vi)vmull_u8(ref_main_idx, dup_const_32_fract) add x12,x8,x9 //(viii)*pu1_ref[ref_main_idx] umlal v14.8h, v13.8b, v29.8b //(vi)vmull_u8(ref_main_idx_1, dup_const_fract) subs x4,x4,#8 st1 {v22.8b},[x0],x3 //(iv) rshrn v10.8b, v10.8h,#5 //(v)shift_res = vrshrn_n_u16(add_res, 5) ld1 {v20.8b},[x12],x11 //(viii)ref_main_idx sub v26.8b, v1.8b , v27.8b //(vii)32-fract(dup_const_32_fract) ld1 {v21.8b},[x12] //(viii)ref_main_idx_1 umull v18.8h, v16.8b, v26.8b //(vii)vmull_u8(ref_main_idx, dup_const_32_fract) add x20,x8,#8 csel x8, x20, x8,gt umlal v18.8h, v17.8b, v27.8b //(vii)vmull_u8(ref_main_idx_1, dup_const_fract) sub x20,x7,#8 csel x7, x20, x7,gt st1 {v10.8b},[x0],x3 //(v) rshrn v14.8b, v14.8h,#5 //(vi)shift_res = vrshrn_n_u16(add_res, 5) beq epilogue ld1 {v5.8b},[x6] //loads the row value umull v2.8h, v5.8b, v0.8b //pos = ((row + 1) * intra_pred_ang) and v4.16b, v2.16b , v6.16b //dup_const_fract(fract = pos & (31)) xtn v4.8b, v4.8h shrn v3.8b, v2.8h,#5 //idx = pos >> 5 umov w14, v3.s[0] //(i)extract idx to the r register sxtw x14,w14 and x9,x14,#0xff //(i) add x10,x8,x9 //(i)*pu1_ref[ref_main_idx] kernel_8_rows: asr x14,x14,#8 //(ii) dup v31.8b, v4.b[0] subs x4,x4,#8 ld1 {v23.8b},[x10],x11 //(i)ref_main_idx sub v24.8b, v1.8b , v25.8b //(viii)32-fract(dup_const_32_fract) and x9,x14,#0xff //(ii) add x20,x6,#8 //increment the row value csel x6, x20, x6,le ld1 {v9.8b},[x10] //(i)ref_main_idx_1 umull v22.8h, v20.8b, v24.8b //(viii)vmull_u8(ref_main_idx, dup_const_32_fract) add x12,x8,x9 //(ii)*pu1_ref[ref_main_idx] ld1 {v5.8b},[x6] //loads the row value umlal v22.8h, v21.8b, v25.8b //(viii)vmull_u8(ref_main_idx_1, dup_const_fract) asr x14,x14,#8 //(iii) dup v29.8b, v4.b[1] //(ii) rshrn v18.8b, v18.8h,#5 //(vii)shift_res = vrshrn_n_u16(add_res, 5) and x9,x14,#0xff //(iii) st1 {v14.8b},[x0],x3 //(vi) sub v30.8b, v1.8b , v31.8b //(i)32-fract(dup_const_32_fract) add x10,x8,x9 //(iii)*pu1_ref[ref_main_idx] ld1 {v12.8b},[x12],x11 //(ii)ref_main_idx umull v10.8h, v23.8b, v30.8b //(i)vmull_u8(ref_main_idx, dup_const_32_fract) asr x14,x14,#8 //(iv) ld1 {v13.8b},[x12] //(ii)ref_main_idx_1 umlal v10.8h, v9.8b, v31.8b //(i)vmull_u8(ref_main_idx_1, dup_const_fract) and x9,x14,#0xff //(iv) umov w14, v3.s[1] //extract idx to the r register sxtw x14,w14 rshrn v22.8b, v22.8h,#5 //(viii)shift_res = vrshrn_n_u16(add_res, 5) dup v27.8b, v4.b[2] //(iii) sub v28.8b, v1.8b , v29.8b //(ii)32-fract(dup_const_32_fract) csel x4, x5, x4,le //reload nt ld1 {v16.8b},[x10],x11 //(iii)ref_main_idx umull v14.8h, v12.8b, v28.8b //(ii)vmull_u8(ref_main_idx, dup_const_32_fract) add x12,x8,x9 //(iv)*pu1_ref[ref_main_idx] st1 {v18.8b},[x0],x3 //(vii) umlal v14.8h, v13.8b, v29.8b //(ii)vmull_u8(ref_main_idx_1, dup_const_fract) ld1 {v17.8b},[x10] //(iii)ref_main_idx_1 rshrn v10.8b, v10.8h,#5 //(i)shift_res = vrshrn_n_u16(add_res, 5) dup v25.8b, v4.b[3] //(iv) umull v2.8h, v5.8b, v0.8b //pos = ((row + 1) * intra_pred_ang) st1 {v22.8b},[x0] //(viii) sub v26.8b, v1.8b , v27.8b //(iii)32-fract(dup_const_32_fract) ld1 {v20.8b},[x12],x11 //(iv)ref_main_idx umull v18.8h, v16.8b, v26.8b //(iii)vmull_u8(ref_main_idx, dup_const_32_fract) add x0,x2,x3 ld1 {v21.8b},[x12] //(iv)ref_main_idx_1 umlal v18.8h, v17.8b, v27.8b //(iii)vmull_u8(ref_main_idx_1, dup_const_fract) and x9,x14,#0xff //(v) dup v31.8b, v4.b[4] //(v) rshrn v14.8b, v14.8h,#5 //(ii)shift_res = vrshrn_n_u16(add_res, 5) add x10,x8,x9 //(v)*pu1_ref[ref_main_idx] st1 {v10.8b},[x2],#8 //(i) sub v24.8b, v1.8b , v25.8b //(iv)32-fract(dup_const_32_fract) asr x14,x14,#8 //(vi) dup v29.8b, v4.b[5] //(vi) umull v22.8h, v20.8b, v24.8b //(iv)vmull_u8(ref_main_idx, dup_const_32_fract) and x9,x14,#0xff //(vi) dup v27.8b, v4.b[6] //(vii) umlal v22.8h, v21.8b, v25.8b //(iv)vmull_u8(ref_main_idx_1, dup_const_fract) add x12,x8,x9 //(vi)*pu1_ref[ref_main_idx] dup v25.8b, v4.b[7] //(viii) rshrn v18.8b, v18.8h,#5 //(iii)shift_res = vrshrn_n_u16(add_res, 5) asr x14,x14,#8 //(vii) ld1 {v23.8b},[x10],x11 //(v)ref_main_idx and v4.16b, v2.16b , v6.16b //dup_const_fract(fract = pos & (31)) and x9,x14,#0xff //(vii) ld1 {v9.8b},[x10] //(v)ref_main_idx_1 shrn v3.8b, v2.8h,#5 //idx = pos >> 5 asr x14,x14,#8 //(viii) st1 {v14.8b},[x0],x3 //(ii) rshrn v22.8b, v22.8h,#5 //(iv)shift_res = vrshrn_n_u16(add_res, 5) add x10,x8,x9 //(vii)*pu1_ref[ref_main_idx] ld1 {v12.8b},[x12],x11 //(vi)ref_main_idx sub v30.8b, v1.8b , v31.8b //(v)32-fract(dup_const_32_fract) and x9,x14,#0xff //(viii) ld1 {v13.8b},[x12] //(vi)ref_main_idx_1 umull v10.8h, v23.8b, v30.8b //(v)vmull_u8(ref_main_idx, dup_const_32_fract) umov w14, v3.s[0] //(i)extract idx to the r register sxtw x14,w14 umlal v10.8h, v9.8b, v31.8b //(v)vmull_u8(ref_main_idx_1, dup_const_fract) add x12,x8,x9 //(viii)*pu1_ref[ref_main_idx] ld1 {v16.8b},[x10],x11 //(vii)ref_main_idx sub v28.8b, v1.8b , v29.8b //(vi)32-fract(dup_const_32_fract) st1 {v18.8b},[x0],x3 //(iii) umull v14.8h, v12.8b, v28.8b //(vi)vmull_u8(ref_main_idx, dup_const_32_fract) csel x8, x1, x8,le //reload the source to pu1_src+2nt ld1 {v17.8b},[x10] //(vii)ref_main_idx_1 umlal v14.8h, v13.8b, v29.8b //(vi)vmull_u8(ref_main_idx_1, dup_const_fract) add x20,x8,#8 //increment the source next set 8 columns in same row csel x8, x20, x8,gt ld1 {v20.8b},[x12],x11 //(viii)ref_main_idx rshrn v10.8b, v10.8h,#5 //(v)shift_res = vrshrn_n_u16(add_res, 5) ld1 {v21.8b},[x12] //(viii)ref_main_idx_1 sub v26.8b, v1.8b , v27.8b //(vii)32-fract(dup_const_32_fract) lsl x20, x3,#3 csel x12,x20,x12,le st1 {v22.8b},[x0],x3 //(iv) umull v18.8h, v16.8b, v26.8b //(vii)vmull_u8(ref_main_idx, dup_const_32_fract) sub x20,x12,x5 csel x12, x20, x12,le st1 {v10.8b},[x0],x3 //(v) umlal v18.8h, v17.8b, v27.8b //(vii)vmull_u8(ref_main_idx_1, dup_const_fract) add x20,x2,x12 //increment the dst pointer to 8*dst_strd - nt csel x2, x20, x2,le xtn v4.8b, v4.8h rshrn v14.8b, v14.8h,#5 //(vi)shift_res = vrshrn_n_u16(add_res, 5) and x9,x14,#0xff //(i) subs x7,x7,#8 add x10,x8,x9 //(i)*pu1_ref[ref_main_idx] bne kernel_8_rows epilogue: st1 {v14.8b},[x0],x3 //(vi) rshrn v18.8b, v18.8h,#5 //(vii)shift_res = vrshrn_n_u16(add_res, 5) sub v24.8b, v1.8b , v25.8b //(viii)32-fract(dup_const_32_fract) umull v22.8h, v20.8b, v24.8b //(viii)vmull_u8(ref_main_idx, dup_const_32_fract) umlal v22.8h, v21.8b, v25.8b //(viii)vmull_u8(ref_main_idx_1, dup_const_fract) st1 {v18.8b},[x0],x3 //(vii) rshrn v22.8b, v22.8h,#5 //(viii)shift_res = vrshrn_n_u16(add_res, 5) st1 {v22.8b},[x0],x3 //(viii) b end_loops core_loop_4: add x10,x8,#1 //pu1_ref_main_idx += (two_nt + 1) add x11,x8,#2 //pu1_ref_main_idx_1 += (two_nt + 2) mov x8,#0 add x5,x8,#1 //row + 1 mul x5, x5, x9 //pos = ((row + 1) * intra_pred_ang) and x5,x5,#31 //fract = pos & (31) cmp x14,x5 //if(fract_prev > fract) add x20,x10,#1 //pu1_ref_main_idx += 1 csel x10, x20, x10,gt add x11,x10,#1 //pu1_ref_main_idx_1 += 1 dup v0.8b,w5 //dup_const_fract sub x20,x5,#32 neg x4, x20 dup v1.8b,w4 //dup_const_32_fract //inner_loop_4 ld1 {v2.s}[0],[x10] //ref_main_idx add x8,x8,#1 mov x14,x5 //fract_prev = fract ld1 {v3.s}[0],[x11] //ref_main_idx_1 add x5,x8,#1 //row + 1 mul x5, x5, x9 //pos = ((row + 1) * intra_pred_ang) and x5,x5,#31 //fract = pos & (31) cmp x14,x5 //if(fract_prev > fract) add x20,x10,#1 //pu1_ref_main_idx += 1 csel x10, x20, x10,gt add x11,x10,#1 //pu1_ref_main_idx_1 += 1 dup v6.8b,w5 //dup_const_fract umull v4.8h, v2.8b, v1.8b //vmull_u8(ref_main_idx, dup_const_32_fract) sub x20,x5,#32 neg x4, x20 dup v7.8b,w4 //dup_const_32_fract umlal v4.8h, v3.8b, v0.8b //vmull_u8(ref_main_idx_1, dup_const_fract) ld1 {v23.s}[0],[x10] //ref_main_idx add x8,x8,#1 ld1 {v9.s}[0],[x11] //ref_main_idx_1 rshrn v4.8b, v4.8h,#5 //shift_res = vrshrn_n_u16(add_res, 5) mov x14,x5 //fract_prev = fract add x5,x8,#1 //row + 1 mul x5, x5, x9 //pos = ((row + 1) * intra_pred_ang) and x5,x5,#31 //fract = pos & (31) cmp x14,x5 //if(fract_prev > fract) add x20,x10,#1 //pu1_ref_main_idx += 1 csel x10, x20, x10,gt add x11,x10,#1 //pu1_ref_main_idx_1 += 1 dup v12.8b,w5 //dup_const_fract umull v10.8h, v23.8b, v7.8b //vmull_u8(ref_main_idx, dup_const_32_fract) sub x20,x5,#32 neg x4, x20 dup v13.8b,w4 //dup_const_32_fract umlal v10.8h, v9.8b, v6.8b //vmull_u8(ref_main_idx_1, dup_const_fract) ld1 {v14.s}[0],[x10] //ref_main_idx add x8,x8,#1 st1 {v4.s}[0],[x2],x3 rshrn v10.8b, v10.8h,#5 //shift_res = vrshrn_n_u16(add_res, 5) ld1 {v15.s}[0],[x11] //ref_main_idx_1 mov x14,x5 //fract_prev = fract add x5,x8,#1 //row + 1 mul x5, x5, x9 //pos = ((row + 1) * intra_pred_ang) and x5,x5,#31 //fract = pos & (31) cmp x14,x5 //if(fract_prev > fract) add x20,x10,#1 //pu1_ref_main_idx += 1 csel x10, x20, x10,gt add x11,x10,#1 //pu1_ref_main_idx_1 += 1 dup v18.8b,w5 //dup_const_fract umull v16.8h, v14.8b, v13.8b //vmull_u8(ref_main_idx, dup_const_32_fract) sub x20,x5,#32 neg x4, x20 dup v19.8b,w4 //dup_const_32_fract umlal v16.8h, v15.8b, v12.8b //vmull_u8(ref_main_idx_1, dup_const_fract) ld1 {v20.s}[0],[x10] //ref_main_idx st1 {v10.s}[0],[x2],x3 rshrn v16.8b, v16.8h,#5 //shift_res = vrshrn_n_u16(add_res, 5) ld1 {v21.s}[0],[x11] //ref_main_idx_1 umull v22.8h, v20.8b, v19.8b //vmull_u8(ref_main_idx, dup_const_32_fract) umlal v22.8h, v21.8b, v18.8b //vmull_u8(ref_main_idx_1, dup_const_fract) st1 {v16.s}[0],[x2],x3 rshrn v22.8b, v22.8h,#5 //shift_res = vrshrn_n_u16(add_res, 5) st1 {v22.s}[0],[x2],x3 end_loops: // ldmfd sp!,{x4-x12,x15} //reload the registers from sp ldp x19, x20,[sp],#16 ldp d14,d15,[sp],#16 ldp d12,d13,[sp],#16 ldp d9,d10,[sp],#16 ret