Hello Vittorio,
Sorry for the late reply, all of us were on leave due to the Diwali
festival in India.
Thanks for the patch, will run some basic test and push the patch.
Regards,
Praveen
On Wed, Nov 7, 2018 at 12:35 AM Vittorio Giovara
wrote:
>
>
> On Thu, Nov 1, 2018 at 5:34 PM Vittorio
Thanks! I messed up the syntax.
On Wed, Oct 31, 2018 at 5:45 PM Andrey Semashev
wrote:
> On 10/31/18 2:33 PM, prav...@multicorewareinc.com wrote:
> > # HG changeset patch
> > # User Praveen Tiwari
> > # Date 1540983948 -19800
> > # Wed Oct 31 16:35:
Hello Jeffrey,
You can find all C primitives in source/common folder.
SAD C primitives ares in source/common/pixel.cpp.
Thanks,
Praveen
On Wed, Sep 5, 2018 at 12:23 PM, Mario *LigH* Rohkrämer
wrote:
> Jeffrey Chen schrieb am 04.09.2018 um 23:57:
>
>> Hi, I would like to configure the sad
Hello Min,
Thanks for the suggestion, we will run some tests and let you know if any
change is required here. Thanks.
Regards,
Praveen Tiwari
On Sat, Jun 2, 2018 at 9:18 AM, chen wrote:
> There have series performance issues, such as,
>
> uint32_t sum = (uint32_t)pow((outOfBound
, Andrey Semashev <andrey.semas...@gmail.com>
wrote:
> On Thu, May 3, 2018 at 7:37 PM, Pradeep Ramachandran
> <prad...@multicorewareinc.com> wrote:
> >
> > On Thu, May 3, 2018 at 2:23 PM, <prav...@multicorewareinc.com> wrote:
> >>
> >>
Thanks for reporting, we are looking at the issue, will send a fix soon.
Regards,
Praveen Tiwari
On Thu, Apr 12, 2018 at 2:31 AM, Mario Rohkrämer <cont...@ligh.de> wrote:
> Am 07.04.2018, 04:29 Uhr, schrieb <mythr...@multicorewareinc.com>:
>
> This series of patches enables
Your request is on the way, soon we will share the performance related
details. Thanks.
Regards,
Praveen Tiwari
On Fri, Apr 6, 2018 at 9:36 PM, Vittorio Giovara <vittorio.giov...@gmail.com
> wrote:
> just curious, what kind of general speed improvement does this give?
> I could have
Please ignore this patch I messed an update. I will resend this soon. Thanks
On Mon, Nov 27, 2017 at 5:11 PM, <prav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1511167656 -19800
> # Mon No
,
Praveen Tiwari
___
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel
-- Forwarded message --
From: chen
Date: Tue, Nov 21, 2017 at 10:07 AM
Subject: Re: [x265] [PATCH] intra: sse4 version of strong intra smoothing
To: Development for x265
>diff -r a7c2f80c18af -r 973560d58dfb
-- Forwarded message --
From:
Date: Tue, May 2, 2017 at 3:16 PM
Subject: [x265] [PATCH 3 of 3] SEA motion search:integralv functions avx2
implementation
To: x265-devel@videolan.org
# HG changeset patch
# User Vignesh Vijayakumar
# Date 1493121121
-- Forwarded message --
From:
Date: 2017-05-02 15:16 GMT+05:30
Subject: [x265] [PATCH 2 of 3] SEA motion search:Add testbench for
integralv functions
To: x265-devel@videolan.org
# HG changeset patch
# User Vignesh Vijayakumar
# Date 1493358749
-- Forwarded message --
From:
Date: Tue, May 2, 2017 at 3:16 PM
Subject: [x265] [PATCH 1 of 3] SEA motion search:Setup asm primitives for
integral calculation
To: x265-devel@videolan.org
# HG changeset patch
# User Vignesh Vijayakumar
# Date
Hi Mario,
Sorry for late reply, you have shared an interesting and useful
information. Currently we are doing some experimental refactoring over the
ASM code base, so it might take some time. Hoping to receive more post like
this.
Regards,
Praveen Tiwari
On Wed, Mar 1, 2017 at 8:21 PM, Mario
Please, ignore this patch. Thanks.
On Thu, Nov 17, 2016 at 8:51 PM, <prav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1479128885 -19800
> # Mon Nov 14 18:38:05 2016 +0530
>
bit-depth version and compare with one bit-depth version,
> but the output are still matched in both 10 and 12 bit.
>
> Regards,
> Min
>
> At 2016-09-22 14:39:50,"Praveen Tiwari" <prav...@multicorewareinc.com>
> wrote:
>
> Hi Min,
>
> After this pat
Hi Min,
After this patch outputs are changing, tested for following command line
for 10-bit and 12-bit outputs.
--input=NebutaFestival_2560x1600_60_10bit_crop.yuv --input-res=2560x1600
--fps=60 --numa-pools="NULL" --output-depth=12 --hash=1 -o NFOut12.hevc
Regards,
Praveen
On Thu, Sep
Please ignore this this behaviour is not required for linux systems.
Thanks.
Regards,
Praveen
On Wed, Sep 7, 2016 at 5:19 PM, <prav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1473246754 -19800
>
h https://patches.videolan.org/patch/13495/ (it fixes
> also this warning)?
>
>
> W dniu 2016-05-30 o 14:45, prav...@multicorewareinc.com pisze:
> > # HG changeset patch
> > # User Praveen Tiwari <prav...@multicorewareinc.com>
> > # Date 1
rav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1463655478 -19800
> # Thu May 19 16:27:58 2016 +0530
> # Node ID 9a6ab28b736e1167ac26977d7da8ab2d23cc296f
> # Parent aca781339b4c8dae94ff7da73f18cd44
Please ignore this sending updated patch. thanks.
Regards,
Praveen
On Tue, May 17, 2016 at 7:17 PM, <prav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1463492830 -19800
> # Tue May 17 19:17:1
Please ignore this sending updated patch. Thanks
Regards,
Praveen
On Tue, May 17, 2016 at 8:01 PM, Pradeep Ramachandran <
prad...@multicorewareinc.com> wrote:
>
> On Tue, May 17, 2016 at 7:07 PM, <prav...@multicorewareinc.com> wrote:
>
>> # HG changeset patch
>
reinc.com wrote:
># HG changeset patch
># User Praveen Tiwari <prav...@multicorewareinc.com>
># Date 1457448163 -19800
># Tue Mar 08 20:12:43 2016 +0530
># Node ID 519441d72cf723dc3b279a91a6080f329729cb49
># Parent 0e1b6472c05e3a53538d8e064e502d8a7508eb6e
>motion.c
Please ignore the patch need to update. Thanks.
Regards,
Praveen
On Tue, Mar 8, 2016 at 10:57 AM, <prav...@multicorewareinc.com> wrote:
> # HG changeset patch
> # User Praveen Tiwari <prav...@multicorewareinc.com>
> # Date 1457356750 -19800
> # Mon Mar 07 18:49:1
-- Forwarded message --
From: aasaipr...@multicorewareinc.com
Date: Mon, Jun 29, 2015 at 4:51 PM
Subject: [x265] [PATCH] asm: avx2 code for weight_sp() 16bpp
To: x265-devel@videolan.org
# HG changeset patch
# User Aasaipriya Chandran aasaipr...@multicorewareinc.com
# Date
on this.
On Fri, Jun 26, 2015 at 5:31 PM, Praveen Tiwari
prav...@multicorewareinc.com wrote:
-- Forwarded message --
From: raj...@multicorewareinc.com
Date: Fri, Jun 26, 2015 at 3:14 PM
Subject: [x265] [PATCH] asm: pixelavg_pp[8xN] avx2 code for 10bpp
To: x265-devel
-- Forwarded message --
From: raj...@multicorewareinc.com
Date: Fri, Jun 26, 2015 at 3:14 PM
Subject: [x265] [PATCH] asm: pixelavg_pp[8xN] avx2 code for 10bpp
To: x265-devel@videolan.org
# HG changeset patch
# User Rajesh Paulrajraj...@multicorewareinc.com
# Date 1435311076
wrote:
I tried using vinserti128. But that reduces the performance than this one.
So i kept this version.
On Fri, Jun 26, 2015 at 3:37 PM, Praveen Tiwari
prav...@multicorewareinc.com wrote:
-- Forwarded message --
From: raj...@multicorewareinc.com
Date: Fri, Jun 26
-- Forwarded message --
From: raj...@multicorewareinc.com
Date: Fri, Jun 26, 2015 at 3:14 PM
Subject: [x265] [PATCH] asm: pixelavg_pp[8xN] avx2 code for 10bpp
To: x265-devel@videolan.org
# HG changeset patch
# User Rajesh Paulrajraj...@multicorewareinc.com
# Date 1435311076
Please ignore duplicate patch (second), send my mistake.
Regards,
Praveen
On Fri, Mar 27, 2015 at 10:41 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1427356204 -19800
# Thu Mar 26 13:20:04 2015 +0530
# Branch
Please ignore duplicate patch (second), send my mistake.
Regards,
Praveen
On Fri, Mar 27, 2015 at 10:41 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 142736 -19800
# Thu Mar 26 14:23:20 2015 +0530
# Branch
Please ignore, need to add performance data in commit message.
Regards,
Praveen
On Thu, Mar 12, 2015 at 6:50 PM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1426165765 -19800
# Node ID
Updated this patch on tip.
Thanks,
Praveen
On Tue, Mar 10, 2015 at 10:53 AM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1425964751 -19800
# Node ID f97dfb483647d573cbcab9a4f007ac2aa89c9066
# Parent
-- Forwarded message --
From: sumala...@multicorewareinc.com
Date: Wed, Mar 11, 2015 at 2:24 PM
Subject: [x265] [PATCH] asm: avx2 code for sad[32x32] for 8bpp
To: x265-devel@videolan.org
# HG changeset patch
# User Sumalatha Polureddysumala...@multicorewareinc.com
# Date
the compiler will not use two 'mova' instruction internally rather
than just using once? Can be depend on the compiler here for this
optimization? Even syntax of 'vpermd' does not allows this.
At 2015-03-10 13:58:50,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav
-- Forwarded message --
From: chen chenm...@163.com
Date: Wed, Mar 11, 2015 at 6:32 AM
Subject: Re: [x265] [PATCH] asm: intra_pred_ang16_34
To: Development for x265 x265-devel@videolan.org
same speed to old version
This avx2 version of asm code eliminates following instruction
-- Forwarded message --
From: chen chenm...@163.com
Date: Wed, Mar 11, 2015 at 6:32 AM
Subject: Re: [x265] [PATCH] asm: intra_pred_ang16_2
To: Development for x265 x265-devel@videolan.org
same speed to old version
This avx2 version of asm code eliminates following instruction on
-- Forwarded message --
From: chen chenm...@163.com
Date: Wed, Mar 11, 2015 at 6:09 AM
Subject: Re: [x265] [PATCH] asm: intra_pred_ang8_24 8bpp, improved 206.33c
- 177.70c over SSE version
To: Development for x265 x265-devel@videolan.org
+c_ang8_mode_24: db 5, 27, 5, 27, 5,
Updated the code with more optimization.
Regards,
Praveen
On Sat, Mar 7, 2015 at 3:31 AM, chen chenm...@163.com wrote:
right
At 2015-03-06 14:16:23,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1425622433 -19800
Update the patch with more optimization.
Regards,
Praveen
On Sat, Mar 7, 2015 at 3:40 AM, chen chenm...@163.com wrote:
right
At 2015-03-06 15:50:38,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1425628229 -19800
Updated the patch as per suggestions.
Regards,
Praveen
On Sat, Mar 7, 2015 at 3:57 AM, chen chenm...@163.com wrote:
At 2015-03-06 17:24:05,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1425633836 -19800
# Node ID
,Praveen Tiwari prav...@multicorewareinc.com
wrote:
-- Forwarded message --
From: chen chenm...@163.com
Date: Wed, Feb 25, 2015 at 7:38 PM
Subject: Re: [x265] [PATCH Review Only] asm-avx2: intra_pred_ang8_33,
improved 265.79c - 185.43c over sse4 asm code
To: Development for x265
,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari prav...@multicorewareinc.com
# Date 1424854196 -19800
# Node ID 177fe9372668b4824c291e967349664766688179
# Parent 02bac78bde961d60d180e59b5260fad93b98d9b4
asm-avx2: intra_pred_ang8_33, improved 265.79c - 185.43c over sse4
-- Forwarded message --
From: chen chenm...@163.com
Date: Thu, Feb 5, 2015 at 5:55 PM
Subject: Re: [x265] [PATCH] blockcopy_pp_12x32: SSE2 asm code optimization
To: Development for x265 x265-devel@videolan.org
this code is right
but could you try use general register move (rN,
Sent updated patch. Thanks.
Regards,
Praveen
On Mon, Feb 2, 2015 at 4:39 PM, chen chenm...@163.com wrote:
At 2015-02-02 16:55:16,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1422867249 -19800
# Branch stable
# Node ID
If it is only 64x64, then definitely it is range issue when we are finally
accumulating sum of all sad calculations. It make more obvious with 64x64
because more number of accumulation is here. Algorithm issue must have
reflected in other partition also.
Regards,
Praveen
On Fri, Jan 9, 2015 at
tab_LumaCoeffVer_32 table of this name is already in file, redefining here
will cause build error. Please, verify and update patch.
On Thu, Nov 20, 2014 at 2:49 PM, Divya Manivannan
di...@multicorewareinc.com wrote:
# HG changeset patch
# User Divya Manivannan di...@multicorewareinc.com
#
patch
# User Praveen Tiwari
# Date 1416299427 -19800
# Node ID 706fa4af912bc1610478de8f09a651ae3e58624c
# Parent 2f0062f0791b822fa932712a56e6b0a14e976d91
refactorizaton of the transform/quant path.
This patch involves scaling down the DCT/IDCT coefficients from int32_t
to int16_t
as they can
patch
# User Praveen Tiwari
# Date 1416299427 -19800
# Node ID 706fa4af912bc1610478de8f09a651ae3e58624c
# Parent 2f0062f0791b822fa932712a56e6b0a14e976d91
refactorizaton of the transform/quant path.
This patch involves scaling down the DCT/IDCT coefficients from int32_t
to int16_t
as they can
changeset patch
# User Praveen Tiwari
# Date 1416402744 -19800
# Node ID 0ef14321fb144362b609d51f2d7c58f7db757ceb
# Parent 706fa4af912bc1610478de8f09a651ae3e58624c
disable denoiseDct asm code until fixed for Mac OS
with denoise disabled, it finds the next failing primitive:
$ ./test
Crashing on vc11-x86-8bpp, Release mode. Min, can you check your code ?
Regards,
Praveen
On Fri, Oct 31, 2014 at 4:16 AM, Min Chen chenm...@163.com wrote:
# HG changeset patch
# User Min Chen chenm...@163.com
# Date 1414709200 25200
# Node ID 5d0b20f6e4de0b59b8c3306793c7267e01b9a41b
#
-16 17:20:13,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1413451199 -19800
# Node ID 858be8d7d7176ab6c6d01cf92d00c8478fe99b34
# Parent 79702581ec824a2a375aebe228d69c3930aeea96
weight_pp avx2 asm code, improved from 8608.65 cycles to 5138.09 cycles over
sse
Seems we missed out something here, I tested this patch at my end outputs
are deterministic with --pmode but still non-deterministic without --pmode
option. Steve/Deepthi please verify at your end before pushing it. I used
the following cli:
y4mInputs\park_joy_1280x720p50.y4m --tune=ssim --psnr
-- Forwarded message --
From: Steve Borho st...@borho.org
Date: Mon, Sep 15, 2014 at 4:28 PM
Subject: Re: [x265] [PATCH] denoiseDct: unit test code
To: Development for x265 x265-devel@videolan.org
On 09/15, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen
You can push 16x16 and 32x32 also they are good in performance but they
need a bit more improvement, I will be sending improvement patch soon.
Regards,
Praveen Tiwari
On Thu, Sep 11, 2014 at 11:29 AM, Deepthi Nandakumar
deep...@multicorewareinc.com wrote:
Would be better to combine this asm
Ignore It, need to correct commit message.
Regards,
Praveen Tiwari
On Thu, Sep 11, 2014 at 4:41 PM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1410433904 -19800
# Node ID 5740ec22db67267bfca97fbba07ef9239802d2b0
# Parent
-- Forwarded message --
From: chen chenm...@163.com
Date: Wed, Sep 10, 2014 at 12:14 PM
Subject: Re: [x265] Fwd: [PATCH] copy_cnt_4: faster AVX2 code
To: Development for x265 x265-devel@videolan.org
At 2014-09-10 09:34:31,Praveen Tiwari prav...@multicorewareinc.com
wrote
vinserti128 ?
At 2014-09-09 16:37:23,prav...@multicorewareinc.com wrote:
# HG changeset patch # User Praveen Tiwari # Date 1410251834 -19800
# Node ID d011073f35258cb2f0ad95db6038c2d9fb840b27 # Parent
ebb84e9dbb0fa0e8c4c9304b2efd57f8ac3d0c05 copy_cnt_4: faster AVX2 code
diff -r ebb84e9dbb0f -r
Thanks, just sent a fix for it.
Regards,
Praveen
On Tue, Aug 12, 2014 at 7:18 PM, chen chenm...@163.com wrote:
-X265_CHECK((int)numSig == primitives.count_nonzero(coeff, 1
log2TrSize * 2), numSig differ\n);
+/* This section of code is to safely convert int32_t
I think you are testing with asm code enabled. Assembly code has it's own
table, it nothing to do with constant 'g_t8' at
source/Lib/TLibCommon/TComRom.cpp (only for C code). Check dct8.asm file
for asm tables.
Regards,
Praveen Tiwari
On Wed, May 28, 2014 at 5:15 AM, Paulo André Oliveira
(1), W(5), W(1), W(3), W(1), W(5), W(1),
W(4), W(5), W(2), W(5), W(4), W(5), W(2), W(5),
W(3), W(1), W(5), W(1), W(3), W(1), W(5), W(1)
what is logic behind such arrangement ?
Regards,
Praveen Tiwari
On Sat, May 10, 2014 at 8:12 AM, Jason Garrett-Glaser ja...@x264.comwrote
-- Forwarded message --
From: Jason Garrett-Glaser ja...@x264.com
Date: Thu, May 8, 2014 at 5:08 PM
Subject: Re: [x265] [PATCH] noise reduction feature, ported from x264
To: Development for x265 x265-devel@videolan.org
This only seems to have 4x4 and 8x8 transform sizes; how does
This is new patch same changes in other modes, but I have given same commit
message perhaps that's why it seems confusing. Do I need to send as an
attachment ?
On Thu, Feb 27, 2014 at 4:28 PM, Deepthi Nandakumar
deep...@multicorewareinc.com wrote:
The earlier patch was pushed, Praveen. Can
Oh, just left by mistake. I commented old code to test correctness of new
code, I will update the patch.
On Thu, Feb 27, 2014 at 3:33 AM, chen chenm...@163.com wrote:
At 2014-02-26 20:28:52,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1393417704
Min, I have sent the updated full patch.
Regards,
Praveen Tiwari
On Wed, Dec 4, 2013 at 8:58 PM, chen chenm...@163.com wrote:
can you send a full patch, not patch to patch
At 2013-12-04 22:50:05,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date
sorry, I removed wrong pointer initialization, I will fix it in next patch,
don't merge it.
On Fri, Nov 22, 2013 at 4:34 PM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1385118266 -19800
# Node ID f2b8bcaf435c00d835cd4389063ed09d22e7be28
# Parent
Merged, sent implementation.
Regards,
Praveen Tiwari
On Wed, Nov 20, 2013 at 6:08 PM, chen chenm...@163.com wrote:
At 2013-11-20 19:45:24,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1384947915 -19800
# Node ID
Please, ignore this patch old code is also fine. Some other bug.
Regards,
Praveen Tiwari
On Tue, Nov 12, 2013 at 3:09 PM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1384249182 -19800
# Node ID 40695de368b6c890fa27a08c8e5a277c9682149c
# Parent
I mistyped one partition size, instead of 8x6 it will be 8x8, rest are
correct.
Regards,
Praveen Tiwari
On Mon, Nov 11, 2013 at 2:58 PM, prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1384162089 -19800
# Node ID
Fixed.
Regards,
Praveen Tiwari
On Mon, Nov 11, 2013 at 4:06 PM, chen chenm...@163.com wrote:
+movu m1, [r2]
+punpcklbw m2, m1,m0
Here have a hide register copy, try to avoid it by SSE4.1 pmovzxbw m2, m1
+movu [r0], m2
Replaced.
Regards,
Praveen Tiwari
On Mon, Nov 11, 2013 at 7:02 PM, chen chenm...@163.com wrote:
+movd m0,[r2]
+pmovzxbw m0,m0
+pextrd [r0], m0, 0
same as movd
___
x265-devel mailing list
Sent Patch.
Regards,
Praveen Tiwari
On Mon, Nov 11, 2013 at 6:54 PM, chen chenm...@163.com wrote:
+;-
+; void blockcopy_ps_%1x%2(int16_t *dest, intptr_t destStride, pixel *src,
intptr_t srcStride
# User Praveen Tiwari
# Date 1383903250 -19800
# Node ID 1e6bf52b6e3471b81e636569daa667f6dec9838a
# Parent 44ac213169c906eab5cba6b4aba876391b81da99
blockcopy_sp_4x8, optimized asm code
diff -r 44ac213169c9 -r 1e6bf52b6e34 source/common/x86/blockcopy8.asm
--- a/source/common/x86/blockcopy8.asm Fri
-- Forwarded message --
From: chen chenm...@163.com
Date: Fri, Nov 8, 2013 at 4:30 PM
Subject: Re: [x265] [PATCH] blockcopy_sp_8x2, optimized asm code
To: Development for x265 x265-devel@videolan.org
+movh [r0], m0
+movhps [r0 + r1], m0
change movh to movlps is
for .asm files?
t 2013-11-08 21:32:05,prav...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1383917516 -19800
# Node ID 662664f0863b38b838a15867745c5564f574fb09
# Parent 227a5666e08869d36e07a75f3db95dd94c774715
blockcopy_sp_16xN, optimized asm code
diff -r 227a5666e088
...@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 creation
diff -r 93cccbe49a93 -r 34ba8955747b
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
# User Praveen Tiwari
# Date 1383828996 -19800
# Node ID f2af7af43dfcb08135a08e755f654314a89efae7
# Parent d71f86b1c58b4fc9f8a3ffeaaef45c60f8bcc468
asm code for blockfil_s, 4x4
blockfill has two l
Actually I named all pointers with blockfill (two I) and function with
blockfil (one I), perhaps
Fixed.
Regards,
Praveen Tiwari
On Wed, Nov 6, 2013 at 8:09 PM, chen chenm...@163.com wrote:
+ movd [r0 + 2 * r1], m3
+ pextrwr6,m3,2
+ mov [r0 + 2 * r1 + 4], r6w
SSE4.1 support below:
pextrw[r0 + 2 * r1 + 4], m3,2
-- Forwarded message --
From: dnyanesh...@multicorewareinc.com
Date: Wed, Oct 30, 2013 at 7:47 PM
Subject: [x265] [PATCH] asm: assembly code for pixel_sad_12x16
To: x265-devel@videolan.org
# HG changeset patch
# User Dnyaneshwar Gorade dnyanesh...@multicorewareinc.com
# Date
-- Forwarded message --
From: yuva...@multicorewareinc.com
Date: Wed, Oct 30, 2013 at 2:38 PM
Subject: [x265] [PATCH] assembly code for pixel_sad_x3_24x32
To: x265-devel@videolan.org
# HG changeset patch
# User Yuvaraj Venkatesh yuva...@multicorewareinc.com
# Date 1383124045
-- Forwarded message --
From: Steve Borho st...@borho.org
Date: Mon, Oct 28, 2013 at 11:55 PM
Subject: Re: [x265] [PATCH 4 of 4] asm: interp_8tap_v_sp for
ipfilter_sp[FILTER_V_S_P_8]
To: Development for x265 x265-devel@videolan.org
On Mon, Oct 28, 2013 at 9:24 AM, Min Chen
I tried using stride 64 for both the source and dest buffers, which is
perfectly reasonable, and the 2xN primitives failed their unit test which
tells me they need to be fixed prior to using them in the encoder.
Sent patch for fix.
___
x265-devel
+templateint N, int width
+void interp_horiz_pp(pixel *src, intptr_t srcStride, pixel *dst, intptr_t
dstStride, int height, int coeffIdx)
+{
+int cStride = 1;
+short const * coeff= g_chromaFilter[coeffIdx];
+src -= (N / 2 - 1) * cStride;
+coeffIdx;
+int offset;
+
...@multicorewareinc.com wrote:
# HG changeset patch
# User Praveen Tiwari
# Date 1381510220 -19800
# Node ID 5a9160e8b0bdc3117c2417bc29453077488efd8e
# Parent c6d89dc62e191f56f63dbcb1781a6494da50a70d
chroma 4XN block, coeffIdex insted of coeff pointer
diff -r c6d89dc62e19 -r 5a9160e8b0bd source/common/x86
ohh... It will be movacoef2, [tab_coeff + coeffIdx * 16].
On Fri, Oct 11, 2013 at 11:21 PM, Praveen Tiwari
prav...@multicorewareinc.com wrote:
I have just missed to change the line movacoef2,
[tab_coeff + 16] (I was just testing for coeffIdex 1 ) I will make
for (int x = 0; x bx; x += 16)
{
-Vec16uc word0, word1;
-Vec8s word3, word4;
-word0.load_a(src0 + x);
-word1.load_a(src1 + x);
-word3 = extend_low(word0) - extend_low(word1);
-
suppose, during execution width comes less than 8 like 5, then we would
like to run our code section which handles the reaming width (_end_col:)
not the whole code (handle multiple of 8 and renaming width part, it will
computed twice in this case and corrupting some (8 - widthleft) dst[] old
88 matches
Mail list logo