SecBSD/ports
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
ports/multimedia/dav1d/patches/patch-src_x86_itx16_avx2_asm

644 lines
24 KiB
Text
Raw Normal View History

SecBSD's official ports repository
2023-08-16 22:26:55 +00:00
Index: src/x86/itx16_avx2.asm
--- src/x86/itx16_avx2.asm.orig
+++ src/x86/itx16_avx2.asm
@@ -360,6 +360,7 @@ cglobal idct_4x4_internal_10bpc, 0, 7, 6, dst, stride,
pshufb m0, m2
jmp tx2q
.pass2:
+ _CET_ENDBR
vextracti128 xm1, m0, 1
WRAP_XMM IDCT4_1D_PACKED_WORD 0, 1, 2, 3, 4, 5
packssdw xm5, xm5 ; pw_2048
@@ -445,6 +446,7 @@ cglobal iadst_4x4_internal_10bpc, 0, 7, 6, dst, stride
%endif
jmp tx2q
.pass2:
+ _CET_ENDBR
lea r6, [deint_shuf+128]
vextracti128 xm1, m0, 1
call m(iadst_4x4_internal_8bpc).main
@@ -497,6 +499,7 @@ cglobal iflipadst_4x4_internal_10bpc, 0, 7, 6, dst, st
vinserti128 m1, m6, xm4, 1
jmp m(iadst_4x4_internal_10bpc).pass1_end
.pass2:
+ _CET_ENDBR
lea r6, [deint_shuf+128]
vextracti128 xm1, m0, 1
call m(iadst_4x4_internal_8bpc).main
@@ -545,6 +548,7 @@ cglobal iidentity_4x4_internal_10bpc, 0, 7, 6, dst, st
pshufb m0, m3
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m1, [pw_1697x8]
movq xm2, [dstq+strideq*0]
movhps xm2, [dstq+strideq*1]
@@ -585,6 +589,7 @@ cglobal idct_4x4_internal_12bpc, 0, 7, 8, dst, stride,
vpermd m1, m3, m0
jmp m(iadst_4x4_internal_12bpc).pass1_end2
.pass2:
+ _CET_ENDBR
vpbroadcastd m5, [pd_2048]
vpermq m0, m0, q3120
vpermq m1, m1, q3120
@@ -624,6 +629,7 @@ cglobal iadst_4x4_internal_12bpc, 0, 7, 8, dst, stride
pminsd m1, m4
jmp tx2q
.pass2:
+ _CET_ENDBR
call .main_pass2
vinserti128 m0, m4, xm6, 1
vinserti128 m1, m2, xm3, 1
@@ -680,6 +686,7 @@ cglobal iflipadst_4x4_internal_12bpc, 0, 7, 8, dst, st
vinserti128 m2, m6, xm4, 1
jmp m(iadst_4x4_internal_12bpc).pass1_end
.pass2:
+ _CET_ENDBR
call m(iadst_4x4_internal_12bpc).main_pass2
vinserti128 m0, m3, xm2, 1
vinserti128 m1, m6, xm4, 1
@@ -706,6 +713,7 @@ cglobal iidentity_4x4_internal_12bpc, 0, 7, 8, dst, st
paddd m2, m3
jmp m(iadst_4x4_internal_12bpc).pass1_end2
.pass2:
+ _CET_ENDBR
; m0 = in0 in1
; m1 = in2 in3
vpbroadcastd m3, [pd_5793]
@@ -771,6 +779,7 @@ cglobal idct_4x8_internal_10bpc, 0, 7, 8, dst, stride,
IDCT4_1D 0, 1, 2, 3, 4, 5, 6, 7
jmp tx2q
.pass2:
+ _CET_ENDBR
packssdw m0, m2
packssdw m1, m3
lea r6, [deint_shuf+128]
@@ -828,6 +837,7 @@ cglobal iadst_4x8_internal_10bpc, 0, 7, 8, dst, stride
REPX {psrad x, 12}, m0, m1, m2, m3
jmp tx2q
.pass2:
+ _CET_ENDBR
call .pass2_main
mova xm4, [pw_2048_m2048]
REPX {pmulhrsw x, xm4}, xm0, xm1, xm2, xm3
@@ -949,6 +959,7 @@ cglobal iflipadst_4x8_internal_10bpc, 0, 7, 8, dst, st
paddd m3, m5, m4
jmp m(iadst_4x8_internal_10bpc).pass1_end
.pass2:
+ _CET_ENDBR
call m(iadst_4x8_internal_10bpc).pass2_main
mova xm4, [pw_2048_m2048]
REPX {pmulhrsw x, xm4}, xm3, xm2, xm1, xm0
@@ -1002,6 +1013,7 @@ cglobal iidentity_4x8_internal_10bpc, 0, 7, 8, dst, st
REPX {psrad x, 12}, m0, m1, m2, m3
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m6, [pixel_10bpc_max]
call .pass2_end
RET
@@ -1058,6 +1070,7 @@ INV_TXFM_4X8_FN dct, flipadst, 12
cglobal idct_4x8_internal_12bpc, 0, 7, 10, dst, stride, c, eob, tx2
jmp m(idct_4x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -1104,6 +1117,7 @@ cglobal iadst_4x8_internal_12bpc, 0, 7, 10, dst, strid
REPX {psrad x, 11}, m0, m1, m2, m3
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -1185,6 +1199,7 @@ cglobal iflipadst_4x8_internal_12bpc, 0, 7, 10, dst, s
psrad m3, m4, 1
jmp m(iadst_4x8_internal_12bpc).pass1_end
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -1206,6 +1221,7 @@ INV_TXFM_4X8_FN identity, identity, 12
cglobal iidentity_4x8_internal_12bpc, 0, 7, 10, dst, stride, c, eob, tx2
jmp m(iidentity_4x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
; m0 = in0 in1
; m1 = in2 in3
; m2 = in4 in5
@@ -1248,6 +1264,7 @@ cglobal idct_4x16_internal_10bpc, 0, 7, 11, dst, strid
REPX {psrad x, 1}, m0, m1, m2, m3, m4, m5, m6, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
packssdw m0, m4
packssdw m1, m5
packssdw m2, m6
@@ -1355,6 +1372,7 @@ cglobal iadst_4x16_internal_10bpc, 0, 7, 11, dst, stri
psrad m7, 13
jmp tx2q
.pass2:
+ _CET_ENDBR
call .pass2_main
vpbroadcastd m5, [pw_2048]
vpbroadcastd m8, [pixel_10bpc_max]
@@ -1535,6 +1553,7 @@ cglobal iflipadst_4x16_internal_10bpc, 0, 7, 11, dst,
psrad m7, m8, 13
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(iadst_4x16_internal_10bpc).pass2_main
vpbroadcastd m5, [pw_2048]
vpbroadcastd m8, [pixel_10bpc_max]
@@ -1596,6 +1615,7 @@ cglobal iidentity_4x16_internal_10bpc, 0, 7, 11, dst,
REPX {psrad x, 13}, m0, m4, m1, m5, m2, m6, m3, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
packssdw m0, m4
packssdw m1, m5
packssdw m2, m6
@@ -1666,6 +1686,7 @@ INV_TXFM_4X16_FN dct, flipadst, 12
cglobal idct_4x16_internal_12bpc, 0, 7, 14, dst, stride, c, eob, tx2
jmp m(idct_4x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
punpckldq m8, m0, m1
punpckhdq m0, m1
punpckldq m9, m2, m3
@@ -1725,6 +1746,7 @@ cglobal iadst_4x16_internal_12bpc, 0, 7, 14, dst, stri
psrad m7, 12
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -1819,6 +1841,7 @@ cglobal iflipadst_4x16_internal_12bpc, 0, 7, 14, dst,
psrad m7, m8, 12
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -1876,6 +1899,7 @@ cglobal iidentity_4x16_internal_12bpc, 0, 7, 14, dst,
REPX {psrad x, 1 }, m2, m6, m3, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -1950,6 +1974,7 @@ cglobal idct_8x4_internal_10bpc, 0, 7, 10, dst, stride
pshufd m3, m3, q1032
jmp tx2q
.pass2:
+ _CET_ENDBR
vbroadcasti128 m4, [deint_shuf]
packssdw m0, m1
packssdw m2, m3
@@ -1997,6 +2022,7 @@ cglobal iadst_8x4_internal_10bpc, 0, 7, 10, dst, strid
psignd m1, m6 ; out2 out3
jmp tx2q
.pass2:
+ _CET_ENDBR
call .pass2_main
vpermq m0, m0, q3120 ; out0 out1
vpermq m2, m1, q3120 ; out2 out3
@@ -2064,6 +2090,7 @@ cglobal iflipadst_8x4_internal_10bpc, 0, 5, 10, dst, s
psignd m2, m5, m6
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(iadst_8x4_internal_10bpc).pass2_main
vpermq m2, m0, q2031
vpermq m0, m1, q2031
@@ -2088,6 +2115,7 @@ cglobal iidentity_8x4_internal_10bpc, 0, 7, 10, dst, s
REPX {paddd x, x }, m0, m1, m2, m3
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m5, [pixel_10bpc_max]
vpbroadcastd m4, [pw_1697x8]
packssdw m0, m1
@@ -2135,6 +2163,7 @@ cglobal idct_8x4_internal_12bpc, 0, 7, 10, dst, stride
vpbroadcastd m9, [clip_20b_max]
jmp m(idct_8x4_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -2159,6 +2188,7 @@ cglobal iadst_8x4_internal_12bpc, 0, 7, 10, dst, strid
psignd m1, m6 ; out2 out3
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -2221,6 +2251,7 @@ cglobal iflipadst_8x4_internal_12bpc, 0, 5, 10, dst, s
psignd m2, m5, m6
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m8, [clip_18b_min]
vpbroadcastd m9, [clip_18b_max]
REPX {pmaxsd x, m8}, m0, m1, m2, m3
@@ -2241,6 +2272,7 @@ INV_TXFM_8X4_FN identity, identity, 12
cglobal iidentity_8x4_internal_12bpc, 0, 7, 10, dst, stride, c, eob, tx2
jmp m(iidentity_8x4_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
; m0 = in0 in1 (interleaved)
; m1 = in2 in3 (interleaved)
; m2 = in4 in5 (interleaved)
@@ -2352,6 +2384,7 @@ cglobal idct_8x8_internal_10bpc, 0, 7, 14, dst, stride
call .round_shift1
jmp tx2q
.pass2:
+ _CET_ENDBR
call .transpose_8x8_packed
call m(idct_8x8_internal_8bpc).main
vpbroadcastd m12, [pw_2048]
@@ -2471,6 +2504,7 @@ cglobal iadst_8x8_internal_10bpc, 0, 7, 14, dst, strid
call .main_end
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_8x8_internal_10bpc).transpose_8x8_packed
pshufd m4, m0, q1032
pshufd m5, m1, q1032
@@ -2532,6 +2566,7 @@ cglobal iflipadst_8x8_internal_10bpc, 0, 7, 14, dst, s
call .main_end
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_8x8_internal_10bpc).transpose_8x8_packed
pshufd m4, m0, q1032
pshufd m5, m1, q1032
@@ -2588,6 +2623,7 @@ cglobal iidentity_8x8_internal_10bpc, 0, 7, 14, dst, s
mova m7, [cq+32*7]
jmp tx2q
.pass2:
+ _CET_ENDBR
packssdw m3, m7
vpbroadcastd m7, [pixel_10bpc_max]
.pass2_main:
@@ -2678,6 +2714,7 @@ cglobal idct_8x8_internal_12bpc, 0, 7, 14, dst, stride
vpbroadcastd m13, [clip_20b_max]
jmp m(idct_8x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -2722,6 +2759,7 @@ cglobal iadst_8x8_internal_12bpc, 0, 7, 14, dst, strid
vpbroadcastd m13, [clip_20b_max]
jmp m(iadst_8x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_main
.pass2_end:
packssdw m0, m1
@@ -2769,6 +2807,7 @@ cglobal iflipadst_8x8_internal_12bpc, 0, 7, 14, dst, s
vpbroadcastd m13, [clip_20b_max]
jmp m(iflipadst_8x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call m(iadst_8x8_internal_12bpc).pass2_main
packssdw m7, m7, m6
packssdw m6, m1, m0
@@ -2791,6 +2830,7 @@ INV_TXFM_8X8_FN identity, identity, 12
cglobal iidentity_8x8_internal_12bpc, 0, 7, 14, dst, stride, c, eob, tx2
jmp m(iidentity_8x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
packssdw m3, m7
vpbroadcastd m7, [pixel_12bpc_max]
jmp m(iidentity_8x8_internal_10bpc).pass2_main
@@ -2849,6 +2889,7 @@ cglobal idct_8x16_internal_10bpc, 0, 7, 16, dst, strid
REPX {mova x, m8}, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call .transpose
call m(idct_8x16_internal_8bpc).main
vpbroadcastd m12, [pw_2048]
@@ -3046,6 +3087,7 @@ cglobal iadst_8x16_internal_10bpc, 0, 7, 16, dst, stri
REPX {mova x, m8}, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_8x16_internal_10bpc).transpose
call m(iadst_8x16_internal_8bpc).main
call m(iadst_8x16_internal_8bpc).main_pass2_end
@@ -3112,6 +3154,7 @@ cglobal iflipadst_8x16_internal_10bpc, 0, 7, 16, dst,
REPX {mova x, m8}, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_8x16_internal_10bpc).transpose
call m(iadst_8x16_internal_8bpc).main
call m(iadst_8x16_internal_8bpc).main_pass2_end
@@ -3179,6 +3222,7 @@ cglobal iidentity_8x16_internal_10bpc, 0, 7, 16, dst,
m8, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
packssdw m0, m8
packssdw m1, m9
packssdw m2, m10
@@ -3244,6 +3288,7 @@ cglobal idct_8x16_internal_12bpc, 0, 7, 16, 32*8, dst,
vpbroadcastd m13, [clip_20b_max]
jmp m(idct_8x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
lea r6, [rsp+32*4]
call .transpose
vpbroadcastd m12, [clip_18b_min]
@@ -3338,6 +3383,7 @@ cglobal iadst_8x16_internal_12bpc, 0, 7, 16, 32*8, dst
vpbroadcastd m13, [clip_20b_max]
jmp m(iadst_8x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
lea r6, [rsp+32*4]
call .pass2_main
call m(iadst_16x8_internal_10bpc).pass1_rotations
@@ -3392,6 +3438,7 @@ cglobal iflipadst_8x16_internal_12bpc, 0, 7, 16, 32*8,
vpbroadcastd m13, [clip_20b_max]
jmp m(iflipadst_8x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
lea r6, [rsp+32*4]
call m(iadst_8x16_internal_12bpc).pass2_main
call m(iflipadst_16x8_internal_10bpc).pass1_rotations
@@ -3405,6 +3452,7 @@ INV_TXFM_8X16_FN identity, identity, 0, 12
cglobal iidentity_8x16_internal_12bpc, 0, 7, 16, 32*8, dst, stride, c, eob, tx2
jmp m(iidentity_8x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_main
packssdw m0, m8
packssdw m1, m9
@@ -3521,6 +3569,7 @@ cglobal idct_16x4_internal_10bpc, 0, 7, 14, dst, strid
REPX {psrad x, 1 }, m0, m1, m2, m3, m4, m5, m6, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
call .transpose_4x16_packed
lea r6, [deint_shuf+128]
call m(idct_16x4_internal_8bpc).main
@@ -3653,6 +3702,7 @@ cglobal iadst_16x4_internal_10bpc, 0, 7, 14, dst, stri
REPX {psrad x, 1}, m0, m1, m2, m3, m4, m5, m6, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x4_internal_10bpc).transpose_4x16_packed
lea r6, [deint_shuf+128]
call m(iadst_16x4_internal_8bpc).main
@@ -3738,6 +3788,7 @@ cglobal iflipadst_16x4_internal_10bpc, 0, 7, 14, dst,
paddd m0, m8, m11
jmp m(iadst_16x4_internal_10bpc).pass1_end
.pass2:
+ _CET_ENDBR
call m(idct_16x4_internal_10bpc).transpose_4x16_packed
lea r6, [deint_shuf+128]
call m(iadst_16x4_internal_8bpc).main
@@ -3772,6 +3823,7 @@ cglobal iidentity_16x4_internal_10bpc, 0, 7, 14, dst,
REPX {psrad x, 12}, m0, m1, m2, m3, m4, m5, m6, m7
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x4_internal_10bpc).transpose_4x16_packed
vpbroadcastd m7, [pw_1697x8]
pmulhrsw m4, m7, m0
@@ -3794,6 +3846,7 @@ cglobal idct_16x4_internal_12bpc, 0, 7, 14, dst, strid
vpbroadcastd m9, [clip_20b_max]
jmp m(idct_16x4_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -3843,6 +3896,7 @@ cglobal iadst_16x4_internal_12bpc, 0, 7, 14, dst, stri
vpbroadcastd m13, [clip_20b_max]
jmp m(iadst_16x4_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_main
REPX {vpermq x, x, q3120}, m0, m1, m2, m3
REPX {pmulhrsw x, m4}, m0, m1, m2, m3
@@ -3898,6 +3952,7 @@ cglobal iflipadst_16x4_internal_12bpc, 0, 7, 14, dst,
vpbroadcastd m13, [clip_20b_max]
jmp m(iflipadst_16x4_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call m(iadst_16x4_internal_12bpc).pass2_main
vpermq m7, m0, q3120
vpermq m6, m1, q3120
@@ -3947,6 +4002,7 @@ cglobal iidentity_16x4_internal_12bpc, 0, 7, 14, dst,
paddd m7, m13
jmp tx2q
.pass2:
+ _CET_ENDBR
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
REPX {pmaxsd x, m12}, m0, m1, m2, m3, m4, m5, m6, m7
@@ -4014,6 +4070,7 @@ cglobal idct_16x8_internal_10bpc, 0, 7, 16, 32*8, dst,
m8, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call .transpose
call m(idct_16x8_internal_8bpc).main
vpbroadcastd m10, [pw_2048]
@@ -4144,6 +4201,7 @@ cglobal iadst_16x8_internal_10bpc, 0, 7, 16, 32*8, dst
REPX {psrad x, 12}, m4, m5, m6, m7, m8, m9, m10, m11
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x8_internal_10bpc).transpose
call m(iadst_16x8_internal_8bpc).main
call m(iadst_16x8_internal_8bpc).main_pass2_end
@@ -4356,6 +4414,7 @@ cglobal iflipadst_16x8_internal_10bpc, 0, 7, 16, 32*8,
call .pass1_rotations
jmp m(iadst_16x8_internal_10bpc).pass1_end
.pass2:
+ _CET_ENDBR
call m(idct_16x8_internal_10bpc).transpose
call m(iadst_16x8_internal_8bpc).main
call m(iadst_16x8_internal_8bpc).main_pass2_end
@@ -4443,6 +4502,7 @@ cglobal iidentity_16x8_internal_10bpc, 0, 7, 16, 32*8,
m8, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x8_internal_10bpc).transpose
vpbroadcastd m10, [pw_4096]
jmp m(idct_16x8_internal_10bpc).end
@@ -4457,6 +4517,7 @@ cglobal idct_16x8_internal_12bpc, 0, 7, 16, 32*8, dst,
vpbroadcastd m13, [clip_20b_max]
jmp m(idct_16x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_main
RET
ALIGN function_align
@@ -4522,6 +4583,7 @@ cglobal iadst_16x8_internal_12bpc, 0, 7, 16, 32*8, dst
vpbroadcastd m14, [clip_20b_max]
jmp m(iadst_16x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_main
call m(idct_16x8_internal_12bpc).end
RET
@@ -4564,6 +4626,7 @@ cglobal iflipadst_16x8_internal_12bpc, 0, 7, 16, 32*8,
vpbroadcastd m14, [clip_20b_max]
jmp m(iflipadst_16x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call m(iadst_16x8_internal_12bpc).pass2_main
packssdw m13, m0, [cq+32* 8]
packssdw m12, m1, [cq+32* 9]
@@ -4591,6 +4654,7 @@ INV_TXFM_16X8_FN identity, identity, 12
cglobal iidentity_16x8_internal_12bpc, 0, 7, 16, 32*8, dst, stride, c, eob, tx2
jmp m(iidentity_16x8_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call m(idct_16x8_internal_10bpc).transpose2
vpbroadcastd m10, [pw_4096]
pmulhrsw m0, m10
@@ -4708,6 +4772,7 @@ cglobal idct_16x16_internal_10bpc, 0, 7, 16, 32*24, ds
m8, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call .transpose
lea r6, [pw_5+128]
mova [rsp], m15
@@ -4944,6 +5009,7 @@ cglobal iadst_16x16_internal_10bpc, 0, 7, 16, 32*24, d
sub r6, 32*8
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x16_internal_10bpc).transpose
lea r6, [pw_5+128]
mova [rsp], m15
@@ -5077,6 +5143,7 @@ cglobal iflipadst_16x16_internal_10bpc, 0, 7, 16, 32*2
paddd m1, [r6-32*3]
jmp m(iadst_16x16_internal_10bpc).pass1_end
.pass2:
+ _CET_ENDBR
call m(idct_16x16_internal_10bpc).transpose
lea r6, [pw_5+128]
mova [rsp], m15
@@ -5163,6 +5230,7 @@ cglobal iidentity_16x16_internal_10bpc, 0, 7, 16, 32*2
m8, m9, m10, m11, m12, m13, m14, m15
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(idct_16x16_internal_10bpc).transpose
mova [cq+32*0], m15
@@ -5191,6 +5259,7 @@ cglobal idct_16x16_internal_12bpc, 0, 7, 16, 32*24, ds
vpbroadcastd m13, [clip_20b_max]
jmp m(idct_16x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
mova [cq+32* 8], m8
mova [cq+32* 9], m9
mova [cq+32*10], m10
@@ -5318,6 +5387,7 @@ cglobal iadst_16x16_internal_12bpc, 0, 7, 16, 32*24, d
vpbroadcastd m14, [clip_20b_max]
jmp m(iadst_16x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call .pass2_part1
call m(iadst_16x8_internal_10bpc).pass1_rotations
call .pass2_part2
@@ -5487,6 +5557,7 @@ cglobal iflipadst_16x16_internal_12bpc, 0, 7, 16, 32*2
vpbroadcastd m14, [clip_20b_max]
jmp m(iflipadst_16x16_internal_10bpc).pass1
.pass2:
+ _CET_ENDBR
call m(iadst_16x16_internal_12bpc).pass2_part1
call m(iflipadst_16x8_internal_10bpc).pass1_rotations
call m(iadst_16x16_internal_12bpc).pass2_part2
@@ -5564,6 +5635,7 @@ cglobal iidentity_16x16_internal_12bpc, 0, 7, 16, 32*2
mova m7, [cq+64*1]
jmp tx2q
.pass2:
+ _CET_ENDBR
call m(iidentity_8x16_internal_12bpc).pass2_main
call m(idct_16x16_internal_10bpc).transpose_fast
test eobd, eobd
@@ -6200,6 +6272,7 @@ cglobal inv_txfm_add_dct_dct_8x32_12bpc, 4, 7, 0, dst,
.eob171:
call .pass1_main
.pass2:
+ _CET_ENDBR
mov cq, r4
vpbroadcastd m12, [clip_18b_min]
vpbroadcastd m13, [clip_18b_max]
@@ -6428,6 +6501,7 @@ cglobal inv_txfm_add_dct_dct_32x8_10bpc, 4, 7, 0, dst,
RET
ALIGN function_align
.pass2:
+ _CET_ENDBR
call m(idct_16x8_internal_8bpc).main
REPX {pmulhrsw x, m11}, m0, m1, m2, m3
call m(idct_16x8_internal_10bpc).write_16x4_start
@@ -7240,6 +7314,7 @@ cglobal inv_txfm_add_dct_dct_32x32_10bpc, 4, 7, 0, dst
cmp r6, r4
jl .fast_loop
.pass2:
+ _CET_ENDBR
lea r3, [rsp+32*3]
mov r4, r6
lea r5, [r6+32*8]
@@ -7525,6 +7600,7 @@ cglobal inv_txfm_add_dct_dct_16x64_10bpc, 4, 7, 0, dst
cmp r6, r4
jl .fast_loop
.pass2:
+ _CET_ENDBR
lea r6, [pw_5+128]
mova m0, [rsp+32* 2] ; in0
mova m1, [rsp+32* 6] ; in4
@@ -7868,6 +7944,7 @@ cglobal inv_txfm_add_dct_dct_32x64_10bpc, 4, 7, 0, dst
cmp r6, r4
jl .fast_loop
.pass2:
+ _CET_ENDBR
lea r6, [pw_5 + 128]
mov r10, rsp
lea r8, [strideq*4]
@@ -8084,6 +8161,7 @@ cglobal inv_txfm_add_dct_dct_64x16_10bpc, 4, 7, 0, dst
dec r3d
jg .fast_loop
.pass2:
+ _CET_ENDBR
lea r7, [r6-32*64]
lea r4, [r6-32*32]
lea r6, [pw_5+128]
@@ -8317,6 +8395,7 @@ cglobal inv_txfm_add_dct_dct_64x32_10bpc, 4, 7, 0, dst
cmp r6, r4
jl .fast_loop
.pass2:
+ _CET_ENDBR
lea r7, [r6-32*32]
lea r5, [r6+32*8]
lea r6, [pw_5+128]
@@ -8461,6 +8540,7 @@ cglobal inv_txfm_add_dct_dct_64x64_10bpc, 4, 7, 0, dst
cmp r6, r4
jl .fast_loop
.pass2:
+ _CET_ENDBR
lea r10, [r6-32*32]
lea r6, [pw_5+128]
lea r8, [strideq*4]
Reference in a new issue Copy permalink