source/libvpx/vp9/common/arm/neon/vp9_mb_lpf_neon.asm - Issue 168343002: libvpx: Pull from upstream

Side by Side Diff: source/libvpx/vp9/common/arm/neon/vp9_mb_lpf_neon.asm

Issue 168343002: libvpx: Pull from upstream (Closed) Base URL: svn://svn.chromium.org/chrome/trunk/deps/third_party/libvpx/

Patch Set: libvpx: Pull from upstream Created 6 years, 10 months ago

Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.

Jump to:

View unified diff | Download patch | Annotate | Revision Log

OLD	NEW
1 ;	1 ;

2 ; Copyright (c) 2013 The WebM project authors. All Rights Reserved.	2 ; Copyright (c) 2013 The WebM project authors. All Rights Reserved.

3 ;	3 ;

4 ; Use of this source code is governed by a BSD-style license	4 ; Use of this source code is governed by a BSD-style license

5 ; that can be found in the LICENSE file in the root of the source	5 ; that can be found in the LICENSE file in the root of the source

6 ; tree. An additional intellectual property rights grant can be found	6 ; tree. An additional intellectual property rights grant can be found

7 ; in the file PATENTS. All contributing project authors may	7 ; in the file PATENTS. All contributing project authors may

8 ; be found in the AUTHORS file in the root of the source tree.	8 ; be found in the AUTHORS file in the root of the source tree.

9 ;	9 ;

10	10

(...skipping 421 matching lines...) Expand 10 before \| Expand all \| Expand 10 after Loading...
432 vqsub.s8 d26, d26, d29 ; oq1 = clamp(qs1 - filter)	432 vqsub.s8 d26, d26, d29 ; oq1 = clamp(qs1 - filter)

433	433

434 veor d24, d24, d22 ; *f_op0 = u^0x80	434 veor d24, d24, d22 ; *f_op0 = u^0x80

435 veor d23, d23, d22 ; *f_oq0 = u^0x80	435 veor d23, d23, d22 ; *f_oq0 = u^0x80

436 veor d25, d25, d22 ; *f_op1 = u^0x80	436 veor d25, d25, d22 ; *f_op1 = u^0x80

437 veor d26, d26, d22 ; *f_oq1 = u^0x80	437 veor d26, d26, d22 ; *f_oq1 = u^0x80

438	438

439 tst r7, #1	439 tst r7, #1

440 bxne lr	440 bxne lr

441	441

	442 orrs r5, r5, r6 ; Check for 0

	443 orreq r7, r7, #2 ; Only do mbfilter branch

	444

442 ; mbfilter flat && mask branch	445 ; mbfilter flat && mask branch

443 ; TODO(fgalligan): Can I decrease the cycles shifting to consective d's	446 ; TODO(fgalligan): Can I decrease the cycles shifting to consective d's

444 ; and using vibt on the q's?	447 ; and using vibt on the q's?

445 vmov.u8 d29, #2	448 vmov.u8 d29, #2

446 vaddl.u8 q15, d7, d8 ; op2 = p0 + q0	449 vaddl.u8 q15, d7, d8 ; op2 = p0 + q0

447 vmlal.u8 q15, d4, d27 ; op2 = p0 + q0 + p3 * 3	450 vmlal.u8 q15, d4, d27 ; op2 = p0 + q0 + p3 * 3

448 vmlal.u8 q15, d5, d29 ; op2 = p0 + q0 + p3 * 3 + p2 * 2	451 vmlal.u8 q15, d5, d29 ; op2 = p0 + q0 + p3 * 3 + p2 * 2

449 vaddl.u8 q10, d4, d5	452 vaddl.u8 q10, d4, d5

450 vaddw.u8 q15, d6 ; op2=p1 + p0 + q0 + p3 * 3 + p2 *2	453 vaddw.u8 q15, d6 ; op2=p1 + p0 + q0 + p3 * 3 + p2 *2

451 vaddl.u8 q14, d6, d9	454 vaddl.u8 q14, d6, d9

(...skipping 142 matching lines...) Expand 10 before \| Expand all \| Expand 10 after Loading...
594 vbif d23, d11, d17 ; oq3 \|= q3 & ~(f2 & f & m)	597 vbif d23, d11, d17 ; oq3 \|= q3 & ~(f2 & f & m)

595 vqrshrn.u16 d3, q15, #4 ; w_oq6	598 vqrshrn.u16 d3, q15, #4 ; w_oq6

596 vbif d1, d12, d17 ; oq4 \|= q4 & ~(f2 & f & m)	599 vbif d1, d12, d17 ; oq4 \|= q4 & ~(f2 & f & m)

597 vbif d2, d13, d17 ; oq5 \|= q5 & ~(f2 & f & m)	600 vbif d2, d13, d17 ; oq5 \|= q5 & ~(f2 & f & m)

598 vbif d3, d14, d17 ; oq6 \|= q6 & ~(f2 & f & m)	601 vbif d3, d14, d17 ; oq6 \|= q6 & ~(f2 & f & m)

599	602

600 bx lr	603 bx lr

601 ENDP ; \|vp9_wide_mbfilter_neon\|	604 ENDP ; \|vp9_wide_mbfilter_neon\|

602	605

603 END	606 END

OLD	NEW