Johann 93166c5e51 neon: vpx_quantize_b_32x32 ...

With skip block the neon is about twice as fast as C.

The neon has no shortcut for coeff < zbin so it always takes the
same amount of time. Even if the C can take the shortcut, it is over
twice as fast in neon. If it can't, that gap increases to over 10x.

BUG=webm:1426

Change-Id: I400722146c1b5a5f6289f67d85fd642463d2bfc6

2017-08-08 14:05:18 -07:00

..

avg_neon.c

avg_neon: fix 4x4, update 8x8

2017-07-10 15:29:34 -07:00

avg_pred_neon.c

comp_avg_pred neon: used by sub pixel avg variance

2017-05-30 22:47:34 +00:00

deblock_neon.c

postproc: vpx_mbpost_proc_down_neon

2017-01-09 10:21:56 -08:00

fdct16x16_neon.c

fdct16x16 neon optimization

2017-06-07 14:59:55 -07:00

fdct32x32_neon.c

fdct neon: 32x32_rd

2017-07-06 06:30:50 -07:00

fdct_neon.c

neon fdct: 4x4 implementation

2017-05-17 07:38:18 -07:00

fdct_partial_neon.c

neon: consolidate horizontal adds

2017-07-10 15:29:13 -07:00

fwd_txfm_neon.c

partial fdct neon: move 8x8_1 and enable hbd tests

2017-06-28 15:37:43 -07:00

hadamard_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

highbd_idct4x4_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct8x8_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct16x16_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct32x32_34_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct32x32_135_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct32x32_1024_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_idct32x32_add_neon.c

Update highbd idct functions arguments to use uint16_t dst

2017-05-03 13:59:16 -07:00

highbd_intrapred_neon.c

Add high bitdepth intra prediction NEON optimization (mode tm)

2016-11-15 14:19:46 -08:00

highbd_loopfilter_neon.c

cosmetics,*loopfilter_neon.c: s/tranpose/transpose/

2016-10-12 16:12:56 -07:00

highbd_vpx_convolve8_neon.c

Update highbd convolve functions arguments to use uint16_t src/dst

2017-04-25 14:22:19 -07:00

highbd_vpx_convolve_avg_neon.c

Update highbd convolve functions arguments to use uint16_t src/dst

2017-04-25 14:22:19 -07:00

highbd_vpx_convolve_copy_neon.c

Update highbd convolve functions arguments to use uint16_t src/dst

2017-04-25 14:22:19 -07:00

highbd_vpx_convolve_neon.c

remove DECLARE_ALIGNED from neon code

2017-05-26 10:41:57 -07:00

idct4x4_1_add_neon.asm

Cosmetics by unifying dest_stride to stride in idct

2016-12-12 15:13:22 -08:00

idct4x4_1_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct4x4_add_neon.asm

Cosmetics by unifying dest_stride to stride in idct

2016-12-12 15:13:22 -08:00

idct4x4_add_neon.c

neon 4 byte helper functions

2017-05-15 13:42:31 -07:00

idct8x8_1_add_neon.c

Clean DC only idct NEON intrinsics

2016-12-28 13:51:44 -08:00

idct8x8_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct16x16_1_add_neon.c

Clean DC only idct NEON intrinsics

2016-12-28 13:51:44 -08:00

idct16x16_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct32x32_1_add_neon.c

Clean DC only idct NEON intrinsics

2016-12-28 13:51:44 -08:00

idct32x32_34_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct32x32_135_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct32x32_add_neon.c

move neon load/stores to a new file

2017-05-15 08:29:43 -07:00

idct_neon.asm

enable vpx_idct16x16_10_add_neon in hbd builds

2016-12-06 16:09:19 -08:00

idct_neon.h

remove DECLARE_ALIGNED from neon code

2017-05-26 10:41:57 -07:00

intrapred_neon_asm.asm

Replace prefix vp9_ with vpx_ for intra prediction functions

2015-07-27 13:42:06 -07:00

intrapred_neon.c

Merge "Add 32x32 d45 and 8x8, 16x16, 32x32 d135 NEON intra prediction"

2016-11-22 23:20:53 +00:00

loopfilter_4_neon.asm

Check in vpx_lpf_vertical_4_dual_neon() assembly

2016-12-02 15:54:30 -08:00

loopfilter_8_neon.asm

NEON asm of vpx_lpf_{horizontal,vertical}_8_dual_neon()

2016-08-16 08:50:57 -07:00

loopfilter_16_neon.asm

Refactor vpx lpf NEON files (step 2/2)

2016-09-30 09:56:28 -07:00

loopfilter_neon.c

cosmetics,*loopfilter_neon.c: s/tranpose/transpose/

2016-10-12 16:12:56 -07:00

mem_neon.h

sub pel variance neon: 4x block sizes

2017-05-22 14:40:01 -07:00

quantize_neon.c

neon: vpx_quantize_b_32x32

2017-08-08 14:05:18 -07:00

sad4d_neon.c

sad4d neon: 64x[32,64]

2017-07-12 13:26:39 +00:00

sad_neon.c

neon: consolidate horizontal adds

2017-07-10 15:29:13 -07:00

save_reg_neon.asm

replace by VSTM/VLDM to reduce one of VST1/VLD1

2016-07-28 23:01:38 +00:00

subpel_variance_neon.c

subpel variance neon: reduce stack usage

2017-05-24 13:28:13 -07:00

subtract_neon.c

vpx_dsp: apply clang-format

2016-07-25 14:14:19 -07:00

sum_neon.h

sad4d neon: 64x[32,64]

2017-07-12 13:26:39 +00:00

transpose_neon.h

Add vpx_highbd_idct32x32_135_add_neon()

2017-03-16 22:37:55 -07:00

variance_neon.c

neon: consolidate horizontal adds

2017-07-10 15:29:13 -07:00

vpx_convolve8_avg_neon_asm.asm

VPX: removed step checks from neon convolve code

2015-08-12 16:46:53 -07:00

vpx_convolve8_neon_asm.asm

VPX: removed step checks from neon convolve code

2015-08-12 16:46:53 -07:00

vpx_convolve8_neon.c

add vpx high bitdepth convolve8 NEON intrinsics optimization

2016-10-17 15:23:54 -07:00

vpx_convolve_avg_neon_asm.asm

Code refactor on InterpKernel

2015-07-31 10:27:33 -07:00

vpx_convolve_avg_neon.c

Refine vpx_convolve_copy_neon() and vpx_convolve_avg_neon()

2016-09-29 16:19:39 -07:00

vpx_convolve_copy_neon_asm.asm

Code refactor on InterpKernel

2015-07-31 10:27:33 -07:00

vpx_convolve_copy_neon.c

Refine vpx_convolve_copy_neon() and vpx_convolve_avg_neon()

2016-09-29 16:19:39 -07:00

vpx_convolve_neon.c

remove DECLARE_ALIGNED from neon code

2017-05-26 10:41:57 -07:00