On Thu, Nov 7, 2013 at 1:01 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1383807695 -19800
# Node ID 34ba8955747b66dcf3471fa216d15b97a3b07e0c
# Parent 93cccbe49a93dd4c054ef06aca76974948793613
added pixelsub_ps C primitive and function pointer
-- Forwarded message --
From: Steve Borho st...@borho.org
Date: Thu, Nov 7, 2013 at 1:51 PM
Subject: Re: [x265] [PATCH] added pixelsub_ps C primitive and function
pointer creation
To: Development for x265 x265-devel@videolan.org
On Thu, Nov 7, 2013 at 1:01 AM,
Hi,
The patch has implemented VBV methods for x265.
1.add parameter vbv-maxrate vbv-bufsize vbv-init into cmd line;
2.implement vbv methods.
The patch doesn't include vbv algorithms based on lookahead, it will be added
in future. And perhaps CU level ratecontrol will also be added later.
I
Since the default preset is medium, shouldnt x265_param_default apply
medium settings only? So that the if (!strcmp(preset, medium)) block here
is empty??
In that case, the preset (if specified) will change only the ones which are
different from medium?
Also, there are a few params here that
1.
# HG changeset patch
# User Min Chen chenm...@163.com
# Date 1383819472 -28800
# Node ID 31ec0bc30ea383dce0768c162e81db2aef05c8f5
# Parent 2cf0e87eaf6e3d7679d2f6c278a7bb353a6d7826
asm: the pixel value in blockcopy_ps is saturation by calcRecon, so asm can use
packuswb
diff -r 2cf0e87eaf6e -r
# HG changeset patch
# User Min Chen chenm...@163.com
# Date 1383819496 -28800
# Node ID ed6c59ee1ee351e83debaf254406ae4a33422396
# Parent 31ec0bc30ea383dce0768c162e81db2aef05c8f5
cleanup: remove unused blockcpy_sc
diff -r 31ec0bc30ea3 -r ed6c59ee1ee3 source/common/pixel.cpp
---
right, except loop is large
At 2013-11-07 15:38:25,yuva...@multicorewareinc.com wrote:
# HG changeset patch
# User Yuvaraj Venkatesh yuva...@multicorewareinc.com
# Date 1383809838 -19800
# Thu Nov 07 13:07:18 2013 +0530
# Node ID a8b449fb406d60fa674cfc71d7207bf0d2162175
# Parent
# HG changeset patch
# User Yuvaraj Venkatesh yuva...@multicorewareinc.com
# Date 1383822620 -19800
# Thu Nov 07 16:40:20 2013 +0530
# Node ID 519267d0fce9bf4b2f5ec52fe6ddc08a274f16bd
# Parent a8b449fb406d60fa674cfc71d7207bf0d2162175
asm: assembly code for pixel_sad_x3_64xN
diff -r
# HG changeset patch
# User Aarthi Thirumalai
# Date 1383823017 -19800
# Thu Nov 07 16:46:57 2013 +0530
# Node ID 0823d243b4e2f9eee6323a9750eeb811bccbad77
# Parent 85002898f5b4308547af6ce464bbdff5f360fa13
aq: bug fix.extend border of TComPic::m_origPicYuv to a multiple of 16
diff -r
# HG changeset patch
# User Sumalatha Polureddy
# Date 1383823751 -19800
# Node ID a54b30b16e83048a7a2ef5584e1e1c9682216075
# Parent 0a1b379be359cbcf76140ac392104c856a037c78
no-rdo: giving weightage to the cost of all CU's and neighbour CU's for early
exit
Early exit is done when CU cost at
# HG changeset patch
# User Aarthi Thirumalai
# Date 1383825146 -19800
# Thu Nov 07 17:22:26 2013 +0530
# Node ID e87bef56a1b6ed9a7a20be7d7021c178c8ddc42a
# Parent 0823d243b4e2f9eee6323a9750eeb811bccbad77
tcompicyuv: add right boundary padding while applying bottom row padding.
diff -r
right
At 2013-11-07 19:12:33,yuva...@multicorewareinc.com wrote:
# HG changeset patch
# User Yuvaraj Venkatesh yuva...@multicorewareinc.com
# Date 1383822620 -19800
# Thu Nov 07 16:40:20 2013 +0530
# Node ID 519267d0fce9bf4b2f5ec52fe6ddc08a274f16bd
# Parent
# HG changeset patch
# User Praveen Tiwari
# Date 1383828382 -19800
# Node ID d71f86b1c58b4fc9f8a3ffeaaef45c60f8bcc468
# Parent 6de480fbbfb94a709adb18dcd3fbef9169de7a97
unit test code for blockfil_s_c function
diff -r 6de480fbbfb9 -r d71f86b1c58b source/test/pixelharness.cpp
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383828996 -19800
# Node ID f2af7af43dfcb08135a08e755f654314a89efae7
# Parent d71f86b1c58b4fc9f8a3ffeaaef45c60f8bcc468
asm code for blockfil_s, 4x4
diff -r d71f86b1c58b -r f2af7af43dfc source/common/x86/asm-primitives.cpp
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383830968 -19800
# Node ID b6f08fc7aef3141e2049701515d23169ef6e232d
# Parent f2af7af43dfcb08135a08e755f654314a89efae7
asm code for blockfil_s, 8x8
diff -r f2af7af43dfc -r b6f08fc7aef3 source/common/x86/asm-primitives.cpp
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383833451 -19800
# Node ID fed3fbe5e9f1942da657957821a5d1bb396f3d37
# Parent b6f08fc7aef3141e2049701515d23169ef6e232d
asm code for blockfil_s, 16x16
diff -r b6f08fc7aef3 -r fed3fbe5e9f1 source/common/x86/asm-primitives.cpp
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383835016 -19800
# Node ID 2848c581fbc3483c951610311e8ffb8ab66c1064
# Parent fed3fbe5e9f1942da657957821a5d1bb396f3d37
asm code for blockfil_s, 32x32
diff -r fed3fbe5e9f1 -r 2848c581fbc3 source/common/x86/asm-primitives.cpp
---
+movr3d, %2
%2/8
+
+ subr3d,8
+ jnz.loop
dec r3d___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel
# HG changeset patch
# User Praveen Tiwari
# Date 1383837973 -19800
# Node ID 7fd763aee9dfcbe2305582c4389c617855e2ee90
# Parent 2848c581fbc3483c951610311e8ffb8ab66c1064
modified unit test code for blockfil_s function
diff -r 2848c581fbc3 -r 7fd763aee9df source/test/pixelharness.cpp
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383838295 -19800
# Node ID 3db284158e681c643d92df531ea5a6311125bb5a
# Parent 7fd763aee9dfcbe2305582c4389c617855e2ee90
blockfil_s, 16x16 sub replaced with dec
diff -r 7fd763aee9df -r 3db284158e68 source/common/x86/blockcopy8.asm
---
# HG changeset patch
# User Praveen Tiwari
# Date 1383838424 -19800
# Node ID b3409a47634c5f29a9978e357e676637ab2ffd25
# Parent 3db284158e681c643d92df531ea5a6311125bb5a
blockfil_s, 32x32 sub replaced with dec
diff -r 3db284158e68 -r b3409a47634c source/common/x86/blockcopy8.asm
---
memset(ref_dest, 0, sizeof(ref_dest));
memset(opt_dest, 0, sizeof(opt_dest));
don't initial to zero
+opt(opt_dest + j, 64, value);
+ref(ref_dest + j, 64, value);
+
+if (memcmp(ref_dest + j, opt_dest + j, (64 * 64 * sizeof(int16_t)) -
j * sizeof(int16_t)))
# HG changeset patch
# User Nabajit Deka
# Date 1383838838 -19800
# Thu Nov 07 21:10:38 2013 +0530
# Node ID a56c53581344df95e54f9cda919419f1d1ad0850
# Parent 85002898f5b4308547af6ce464bbdff5f360fa13
Bug fix for luma vpp asm routines.Also incorporated review comment changes.
diff -r
# HG changeset patch
# User Murugan Vairavel muru...@multicorewareinc.com
# Date 1383839350 -19800
# Thu Nov 07 21:19:10 2013 +0530
# Node ID d3517392c9845395895f6dbd62fb1a009df4df2f
# Parent 93cccbe49a93dd4c054ef06aca76974948793613
asm: pixel_sub_ps routine for 4x2 block size
diff -r
# HG changeset patch
# User Praveen Tiwari
# Date 1383840171 -19800
# Node ID 4410965d4b9db9285a9047435ebff660a031e474
# Parent b3409a47634c5f29a9978e357e676637ab2ffd25
blockfil_s, changed initial vale to 0xCD
diff -r b3409a47634c -r 4410965d4b9d source/test/pixelharness.cpp
---
+;-
+; void pixel_sub_ps_c_4x2(int16_t *dest, intptr_t destride, pixel *src0,
pixel *src1, intptr_t srcstride0, intptr_t srcstride1);
+;-
This
Applied to code.
Regards,
Praveen Tiwari
On Thu, Nov 7, 2013 at 8:09 PM, chen chenm...@163.com wrote:
+movr3d, %2
%2/8
+
+ subr3d,8
+ jnz.loop
dec r3d
___
x265-devel mailing list
# HG changeset patch
# User Murugan Vairavel muru...@multicorewareinc.com
# Date 1383840463 -19800
# Thu Nov 07 21:37:43 2013 +0530
# Node ID 318720b33850878ff5bb2bbbceb53e53944725ad
# Parent d3517392c9845395895f6dbd62fb1a009df4df2f
asm: pixel_sub_ps routine for 4x4 block size
diff -r
+movdm0, [r2]
+movdm1, [r3]
+pmovzxbwm2, m0
+pmovzxbwm3, m1
+
+movdm0, [r2 + r4]
+movdm1, [r3 + r5]
+pmovzxbwm4, m0
+pmovzxbwm5, m1
+
+movdm0, [r2 + 2 * r4]
+movdm1, [r3 + 2 * r5]
+pmovzxbw
On Thu, Nov 7, 2013 at 4:07 AM, Deepthi Nandakumar
deep...@multicorewareinc.com wrote:
Since the default preset is medium, shouldnt x265_param_default apply
medium settings only? So that the if (!strcmp(preset, medium)) block here
is empty??
In that case, the preset (if specified) will
On Thu, Nov 7, 2013 at 5:17 AM, Aarthi Thirumalai
aar...@multicorewareinc.com wrote:
# HG changeset patch
# User Aarthi Thirumalai
# Date 1383823017 -19800
# Thu Nov 07 16:46:57 2013 +0530
# Node ID 0823d243b4e2f9eee6323a9750eeb811bccbad77
# Parent
On Thu, Nov 7, 2013 at 6:01 AM, yuva...@multicorewareinc.com wrote:
# HG changeset patch
# User Yuvaraj Venkatesh yuva...@multicorewareinc.com
# Date 1383825592 -19800
# Thu Nov 07 17:29:52 2013 +0530
# Node ID d956fd7741f2314a12bbbf529796589d4d6388bf
# Parent
On Thu, Nov 7, 2013 at 2:54 AM, Praveen Tiwari prav...@multicorewareinc.com
wrote:
-- Forwarded message --
From: Steve Borho st...@borho.org
Date: Thu, Nov 7, 2013 at 1:51 PM
Subject: Re: [x265] [PATCH] added pixelsub_ps C primitive and function
pointer creation
To:
On Thu, Nov 7, 2013 at 6:56 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1383828996 -19800
# Node ID f2af7af43dfcb08135a08e755f654314a89efae7
# Parent d71f86b1c58b4fc9f8a3ffeaaef45c60f8bcc468
asm code for blockfil_s, 4x4
blockfill has two l
On Thu, Nov 7, 2013 at 5:29 AM, sumala...@multicorewareinc.com wrote:
# HG changeset patch
# User Sumalatha Polureddy
# Date 1383823751 -19800
# Node ID a54b30b16e83048a7a2ef5584e1e1c9682216075
# Parent 0a1b379be359cbcf76140ac392104c856a037c78
no-rdo: giving weightage to the cost of all
On Thu, Nov 7, 2013 at 9:41 AM, naba...@multicorewareinc.com wrote:
# HG changeset patch
# User Nabajit Deka
# Date 1383838838 -19800
# Thu Nov 07 21:10:38 2013 +0530
# Node ID a56c53581344df95e54f9cda919419f1d1ad0850
# Parent 85002898f5b4308547af6ce464bbdff5f360fa13
Bug fix for luma
I have a few questions.
1. Do we need so many local variables?
2. Why are we adding outTempCU-cost to totalCost and then comparing
against outBestCU-cost? That doesnt make much sense to me. AFAIk,
outTempCU does not contain any valid data - we should remove this.
3. Should we be adding
-- Forwarded message --
From: Steve Borho st...@borho.org
Date: 2013/11/8
Subject: Re: [x265] [PATCH] asm code for blockfil_s, 4x4
To: Development for x265 x265-devel@videolan.org
On Thu, Nov 7, 2013 at 6:56 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
#
On Fri, Nov 8, 2013 at 8:03 AM, Deepthi Nandakumar
deep...@multicorewareinc.com wrote:
I have a few questions.
1. Do we need so many local variables?
As Steve suggested, totalCostNeigh and totalCountNeigh can be added
inside each if () and will try to remove costCU, costCUAbove ,
On Nov 7, 2013 11:24 PM, Steve Borho st...@borho.org wrote:
On Thu, Nov 7, 2013 at 5:17 AM, Aarthi Thirumalai
aar...@multicorewareinc.com wrote:
# HG changeset patch
# User Aarthi Thirumalai
# Date 1383823017 -19800
# Thu Nov 07 16:46:57 2013 +0530
# Node ID
# HG changeset patch
# User Min Chen chenm...@163.com
# Date 1383894022 -28800
# Node ID 2bcf52ff55e649eabaff8a54ff940cdbdda60a5d
# Parent fd721a5ba06302858fa39739b4adc67ae86f7be3
testbench: update for blockcopy, these function use dynamic range [0,255]
diff -r fd721a5ba063 -r 2bcf52ff55e6
On Thu, Nov 7, 2013 at 10:50 PM, Aarthi Priya Thirumalai
aar...@multicorewareinc.com wrote:
On Nov 7, 2013 11:24 PM, Steve Borho st...@borho.org wrote:
On Thu, Nov 7, 2013 at 5:17 AM, Aarthi Thirumalai
aar...@multicorewareinc.com wrote:
# HG changeset patch
# User Aarthi
I noticed that enabling Adaptive Quantization (--aq-mode 1) may or may not
lead to visible artefacts. The dependencies are not yet certain. But a
user in the doom9 forum (vood007) mentioned a relation to presets, as I
reported artefacts with a command line not using any preset, but a clean
# HG changeset patch
# User Min Chen chenm...@163.com
# Date 1383897233 -28800
# Node ID 6b3b0ca2dd0ee3e3cfe43ec3142c54c3e5f9788a
# Parent 2ef0f2fb1fe58ae021004c8de0062f602d03b0e9
testbench: update for blockcopy, these function use dynamic range [0,255]
diff -r 2ef0f2fb1fe5 -r 6b3b0ca2dd0e
44 matches
Mail list logo
