Intel Software Network articles Feed

Intel® Math Kernel Library (Intel® MKL) : What's deprecated?

Thu, 01 Dec 2011 00:00:00 -0800

Please see below for details on the deprecated functionalities in Intel® Math Kernel Library (Intel® MKL).

Forthcoming Intel® MKL 11.0 (Expected date of release: Autumn 2012)

• Intel® MKL 11.0 will have backward incompatibility with Intel® MKL 10.2 update 3

• Open MP static library on Microsoft* Windows

• GMP* Arithmetic functions

• Intel® MKL Reference Manual will be removed from product package

• Intel® Pentium® III Processors will no longer be supported

• Red Hat* EL4 and PGI* Fortran 10.x support will be dropped

• PGI* Fortran 77 support will be removed

Please send us your comments at Intel® Premier Support or the Intel® MKL user forum.

Intel(R) C++ Compiler Compatibility with Microsoft* Visual Studio Express*

Thu, 16 Jun 2011 21:00:00 -0700

The Microsoft* Visual Studio Express Edition* does not provide the necessary interface for the Intel(R) C++ Compiler plug-in for Visual Studio IDE. To install and use the Intel C++ Compiler in Visual Studio IDE you will need Visual Studio Standard Edition* or higher. When using the Intel C++ Compiler with Visual Studio Exrpess you can only compile and build applications from an Intel Compiler command prompt.

There are no constraints in using the Intel(R) libraries such as Intel(R) IPP or TBB with Visual Studio Express. You can download stand-alone versions of the above libraries, add the library "include" and "library" directories in the IDE and compile using the Visual Studio Express Compiler.

mathimf.h is incompatible with system math.h

Mon, 30 May 2011 08:30:00 -0700

Reference number: DPD200166165

Problem : mathimf.h problem in VS2010

Environment : VS2010 , Intel C++ compiler

Affected compilers: 12.0 u/d (1 to 4)

Sample code:

#pragma once

#include 
using std::string;

#include 

class bccb
{
bccb(void);
~bccb(void);
};


bccb::bccb(void)
{
}

bccb::~bccb(void)
{
}

Problem Description:--
When above is compiled with Intel compiler in VS2010 mode, following error:--

# error " is incompatible with system !"

With VS2008, the code compiles fine.

Solutions/workarounds:--

1. Use math.h instead of mathimf.h, and in case of a needed Intel-specific functions, their definitions can be taken from mathimf.h and inserted into application code.

2. Standard recommendation is to avoid usage of math.h and mathimf.h in the same compilation unit
by separating code that depends on math.h from the code that depends on mathimf.h into dirrerent files.
The above code though has the problem that the use is not avoided. This is an issue needed to be fixed.

3. First, try to include before any other header. This will automatically prevent futher inclusions of .
So, include . But still icl in VS2010 mode does not compile, and this will be fixed in later release. should never be mistakenly used before , as this would definitely cause problems due to below reason:--

In C language,
MS math.h deprecates usage of some c99-functions.
The issue here is that MS does not support ISO C99.
Also a number of float and long double functions are defined
through calls to corresponding double-precision functions,
giving performance or accuracy loss compared to float/long-double functions
existing in Intel math library.

In C++ language,
MS redefines a number of c99-functions so that they are inlined.

Resolution status:
This issue will be fixed in the later 12.0 release and next major release 12.1.
So, including before explicitly or inadvertently as in sample code will build fine with VS2010 in later 12.x .

[DISCLAIMER: The information on this web site is intended for hardware system manufacturers and software developers. Intel does not warrant the accuracy, completeness or utility of any information on this site. Intel may make changes to the information or the site at any time without notice. Intel makes no commitment to update the information at this site. ALL INFORMATION PROVIDED ON THIS WEBSITE IS PROVIDED "as is" without any express, implied, or statutory warranty of any kind including but not limited to warranties of merchantability, non-infringement of intellectual property, or fitness for any particular purpose. Independent companies manufacture the third-party products that are mentioned on this site. Intel is not responsible for the quality or performance of third-party products and makes no representation or warranty regarding such products. The third-party supplier remains solely responsible for the design, manufacture, sale and functionality of its products. Intel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. *Other names and brands may be claimed as the property of others.]

C++0x Features Supported by Intel(R) C++ Compiler

Tue, 07 Dec 2010 00:00:00 -0800

Starting in 11.0 the Intel(R) C++ Compiler has supported some of the C++0x features. With the latest release of Intel C++ Composer XE for Windows, Linux* and Mac OS* X, and the Intel Parallel Composer 2011, more C++0x features are supported.

C++0x Core Language Features	Version 11.1	Version 12.0	Version 12.0 Update 6
Rvalue references v2	No	Yes	Yes
Rvalue references for *this	No	No	No
Initialization of class objects by rvalues	Yes	Yes	Yes
static_assert	Yes	Yes	Yes
auto	No	Yes	Yes
Multi-declarator auto	No	Yes	Yes
Removing old auto	No	Yes	Yes
Trailing return types	No	Yes	Yes
Lambdas V1.0	No	Yes	Yes
decltype	No	Yes	Yes
Right angle brackets	Yes	Yes	Yes
Extern templates	Yes	Yes	Yes
nullptr	No	No	yes
Strongly typed enums	Partial	Partial	Partial
Forward declared enums	Partial	Partial	Partial
Extended friend declarations	Partial	Yes	Yes
Local and unnamed types as template arguments	No	Yes	Yes
deleted and defaulted functions	No	Yes	Yes
Allow typename outside of templates	No	Yes	Yes
variadic templates	No	No	Yes
New-style SFINAE (N2634)	No	No	Yes
alias templates	No	No	Yes
Late-specified return types	No	No	Yes
Default template arguments for function templates	No	No	Yes
Standard attributes (N2761)	No	No	Yes
char16_t/char32_t types	No	No	Partial on Windows; Yes on Linux & Mac OS X
C++0x Core Language Features: Concurrency
exception_ptr	No	Yes	Yes
Thread-local storage	Partial	Partial	Partial
C++0x Core Language Features: C99
__func__	Yes	Yes	Yes
C99 preprocessor	Yes	Yes	Yes
long long	Yes	Yes	Yes

On Windows environment when using Intel C++ compiler with Visual Studio 2010*, the C++0x features supported by Visual C++ 2010 are enabled by default. Use "/Qstd=c++0x" to turn on the support for all other cases.
On Linux or Mac OS X environment use "-std=c++0x".

Please read the documentation for detail information under /Qstd (-std on Linux* or Mac OS* X) option.

See C99 Support in Intel(R) C++ Compiler article for supported C99 features.

C99 Support in Intel(R) C++ Compiler

Wed, 01 Dec 2010 00:00:00 -0800

Most of the C99 features are supported by Intel(R) C++ Composer XE for Windows*, Linux* and Mac OS* X, and Intel Parallel Composer 2011.

The option to turn on C99 support:

/Qstd=c99 on Windows
-std=c99 on Linux and Mac OS X
The default is C89 instead

The following C99 features are supported:

restricted pointers (restrict keyword)
variable-length Arrays
flexible array members
complex number support (_Complex keyword)
hexadecimal floating-point constants
compound literals
designated initializers
mixed declarations and code
macros with a variable number of arguments
inline functions (inline keyword)
boolean type (_Bool keyword)

The following features are not supported:

long double (128-bit representations)

See C++0x Features Supported by Intel(R) C++ Compiler for supported C++0x features.

C interface Support for LAPACK

Sat, 06 Nov 2010 11:30:00 -0700

Intel® MKL 10.3, we have extended C support and added C language interface to LAPACK rotuines.

Please refer the C interface to LAPACK technical paper for more details.

Please follow the link, you can find the online documentations and C LAPACK examples.

OpenMP static library has been deprecated since Intel® IPP 7.0

Thu, 02 Sep 2010 09:00:00 -0700

OpenMP static library (libomp5*.lib on Windows* and libomp5*.a on Linux) has been deprecated since Intel® IPP 7.0.
In the figure shown below the different versions of the static OpenMP libraries are shown for a Windows* installation of Intel® Parallel Studio.

We recommend you to use the dynamic version of the OpenMP library instead of the static version.

The dynamic OpenMP library is located in Parallel Studio 2011\Composer\compiler\lib\ Folder. For e.g: the ia32 dynamic OpenMP library is located in Parallel Studio 2011\Composer\compiler\lib\ia32 folder in a Windows* installation as shown in Figure below.

Using C/C++ Atomics

Tue, 31 Aug 2010 21:00:00 -0700

Introduction

The Intel® C++ Compiler 12.0 available in Intel® Parallel Composer 2011 and Intel® C++ Composer XE now supports the atomic data types. These standard data types provide a standard way of implementing fine grained atomic locking in C and C++ which can be useful for many different parallel methodologies, from native threads to Intel® Threading Building Blocks to OpenMP* to Intel® Cilk™ Plus. This article covers the basic syntax and how it works to ensure threadsafe accesses of shared data.

Atomics Requirements

Intel® Parallel Composer 2011
#include the or header files

atomic_flag

The atomic_flag data type is the basis for all atomic data types. It provides two operations that can be done atomically - a test and set operation, and a clear operation. The flag is a bool that is set to 0 when clear() is used. For test_and_set, the current value of the bool is returned and then the flag is set to 1.

The syntax for this is:

std::atomic_flag x;
bool y = x.test_and_set();
x.clear();

Other atomic data types

In short, void* (as atomic_address), bool (atomic_bool), and various char/integer types are supported. These can also be alternatively declared in C++ using atomic where T is void*, bool, or the char/int type.

atomic_bool supports the following operations:

bool load()
-- Atomically return the value of atomic_bool. Can also happen when using the "=" assignment operator or during conversions.

void store(bool)
-- Atomically set the value of atomic_bool. Can also happen when using the "=" assignment operator or during conversions.

bool exchange(bool)
-- Atomically swap the values of the current atomic_bool with the boolean input. Return the old value of the atomic bool.

bool compare_exchange_weak(bool & expected, bool desired)
-- Atomically: compare the current value of atomic_bool with expected. If equal, set atomic_bool value to value of desired and return true. If not equal, set expected to desired value and return false. Intended for use in loops.

bool compare_exchange_strong(bool & expected, bool desired)
-- Atomically: compare the current value of atomic_bool with expected. If equal, set atomic_bool value to value of desired and return true. If not equal, set expected to desired value and return false. Intended for use outside of loops.

atomic_address supports the above atomic_bool operations for void* types, and also the following:

Operator +=(void*)
-- Atomically increment the atomic_address by the input number of bytes and store

Operator -=(void*)
-- Atomically decrement the atomic_address by the input number of bytes and store

Other atomic types:

The other types that are supported:

atomic_char16_t
atomic_char32_t
atomic_wchar_t
atomic_char
atomic_schar
atomic_uchar
atomic_short
atomic_ushort
atomic_int
atomic_uint
atomic_long
atomic_ulong
atomic_llong
atomic_ullong

These data types support the atomic_bool and atomic_address functions above for their data type, plus the following:

Operator ++:
-- Atomically increment by 1

Operator --:
-- Atomically decrement by 1

Operator &= (T)
-- Atomically bitwise-and with input and store

Operator |= (T)
-- Atomically bitwise-or with input and store

Operator ^= (T)
-- Atomically bitwise-xor with input and store

Technical Support

Visit our online Community Support User Forums and Knowledge Base to get all of the help you need from our own tools and parallelism experts, and your fellow developers. Go tohttp://software.intel.com/en-us/articles/intel-software-developer-support/to start your search.

For product and purchase information visit:

http://intel.com/software/products

Intel and the Intel logo, are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries.

*Other names and brands may be claimed as the property of others.

INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL’S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT. Intel products are not intended for use in medical, life saving, life sustaining applications. Intel may make changes to specifications and product descriptions at any time, without notice.

Copyright © 2010, Intel Corporation. All Rights Reserved.

Valgrind* memcheck Gives Uninitialized Value False Positives on String Routines

Tue, 20 Apr 2010 00:00:00 -0700

Problem : When using valgrind* to do memory checking of an executable built with the Intel® C++ Compiler 11.0 or 11.1, you may get false positive messages on string routines that look like the following:

==14292== Conditional jump or move depends on uninitialised value(s)
==14292==    at 0x3D0F41442D: strlen (in /lib64/ld-2.5.so)

==14292== Conditional jump or move depends on uninitialised value(s)
==14292==    at 0x3D0F40A2E5: _dl_relocate_object (in /lib64/ld-2.5.so)

An example of a code that would get false positives is as follows:

#include 
#include 

int main() {
   std::cout << "The result of strcmp of two identical strings is " << std::strcmp("String A", "String A") << std::endl;
   return(0);
}

Environment : Windows*, Linux* or Mac OS* on IA-32 or Intel(R) 64-based systems

Root Cause : This is a known issue with valgrind which they have documented on their documentation page under Known Limitations†. Valgrind gives false positives on the highly optimized string routines the Intel® C++ Compiler provides for systems that support Intel(R) Streaming SIMD Extensions 2 (this environment is targeted by default with the 11.x compilers).

Resolution : Please contact valgrind if you need a resolution. Alternative solutions include:

Building your application with /arch:IA32 (Windows*) or -mia32 (Linux*/Mac OS*). This will likely affect performance.
If on Windows*, you can use Intel® Parallel Inspector which does not have this issue.

† This link will take you off the Intel website. Intel is not responsible for its content.

Intel® IPP Library 6.1 Fixes List

Thu, 04 Feb 2010 08:00:00 -0800

The tables below summarize specific customer issues or feature requests that have been addressed by the indicated product releases. Some items span multiple architectures and/or operating systems and some relate only to a single architecture or operating system.

NOTE: The issues, defects, bug reports, and feature requests summarized below represent specific issues with specific test cases. An item listed here does not imply that it necessarily applies to your application(s). If your situation does not match the specific test case you may not have experienced the error or problem associated with that update. It is not possible to describe the details of every issue and its specific test case in these tables.

Providing a complete description of each item in the list below is impractical. For that reason we ask that you post a message on the IPP user forum and reference this page and the "DPD" number of interest if you have further questions regarding the information in these tables. A DPD number with more information has a link.

These tables are updated regularly to correct errors and omissions.

NOTE: The bugs listed below are for both the IPP library and the IPP sample code. Always upgrade the IPP library and sample code at the same time.

IPP v6.1 update 6 (30 Jul 2010)

DPD200155165        UMC simple_player app does not work with Windows Aero interface.
DPD200154198        Data Compression error in ZLIB (deflate.c) when source buffer size >=48K.
DPD200154900        Data compression error with deflate() under certain variable block size conditions.
DPD200153276        Change alignment in H.264 UMC decoder to DECLALIGN macros insted of __ICL alignment.
DPD200087211        Documentation: Example 9-4 ippiFilterColumn() needs to include border information.
DPD200154874        Documentation: ippsFIR() descriptions are misleading or incorrect in description of "numIter" parameter.
DPD200186071        Documentation: Correct example 6-1 using ippiRGBToYUV() -- wrong ROI size: should be {3,3} not {9,3).

IPP v6.1 update 5 (9 Apr 2010)

DPD200150333        Output image has small corruption on top edge when applying WarpAffine with dst roi size (1920 x 240).
DPD200150883        Delphi sample code error on loading file name (sample has been updated).
DPD200089782        JPEG grayscale lossless compresion distortion.
DPD200150199        Error in DIB_PAD_BYTES in UIC picnic sample.
DPD200151794        Corrected results for the ippGetNumCoresOnDie() function on multi-core processors.
DPD200151952        IPP-zlib compress generates spurious compressed data that cannot be decompressed.
DPD200089070        UMC H264EncoderFrameList_InsertFrame() fails allocating 15MB frame buffer, returns null pointer.
DPD200092946        Picnic sample application produces invalid JPEG 2000 files when saving in lossless format.
DPD200187089        ippiRGBToGray_8u_AC4C1R and ippiRGBToGray_8u_C3C1R translate RGB(255,255,255) to 254.

IPP v6.1 update 4 (10 Feb 2010)

DPD200149498        Memory leak in H.264 UMC sample code with resolution change on same instance of decoder.
DPD200150122        IPP_BZIP2 data corruption -- data sensitive and isolated to a 3mb fragment of a specific binary file.
DPD200149570        IPP samples redistribution file and EULA updated for clarification.
DPD200087613        ippsFilterNoiseDetect* "ascending noise detection" problem fixed. PESQ was increased by ~0.01.
DPD200148808        Cryptography function ippsECCPComparePoint* contains a misprint: "equial" should be "equal."

IPP v6.1 update 3 (26 Nov 2009)

DPD200140668        ippiResizeSqrPixel_8u_C1R has differing results depending on CPU (rounding error problem)
DPD200134986        DFT performance optimization for lengths 8*n, 16*n, 13*n, p*n (p>50)
DPD200084751        Correct ippiColorToGray_8u_C3C1R rounding errors
DPD200086319        ippiSegmentWatershed_8u16u should use the value IPP_MAX_16U for border pixels but uses 8u instead
DPD200084408        Problem with IPP linking in kernel mode for Intel64 (em64t red zone in ring0 and interrupts)
DPD200087170        ippiCopy_8u_C1R performance on v8 architecture (v8 code slower in Core 2 Duo than t7 and w7)
DPD200141560        Feature request: BZIP2 with ipp_ prefixes

IPP v6.1 update 2 (18 Oct 2009)

DPD200084538        UIC sample picnic.exe (64 bits) cannot be started
DPD200085298        Linking problem on 64-bit Mac OS X with IPP fuctions
DPD200084177        Image artifacts in jpeg2000 image
DPD200140266        DMIP crashes when reusing graphs
DPD200140194        BWT error
DPD200138989        Running application in debug causes heap corruption at the end of functions that use UIC classes
DPD200133386        mp3 encoder produces fuzzy results with 8k sampleing rate PCM
DPD200136808        IPP C# wrapper bugs
DPD200134685        ippiAbs_32f_AC4IR failed with 300x300
DPD200134689        C++ sample ippi.hpp missing ippiCopy_16u_C1R
DPD200134691        Remove description of ippiGrayDilateBorder/ippiGrayErodeBorder
DPD200132032        ROI is reversed in ippiResizeSqrPixel_x_P
DPD200132261        Fix the jpegview makefile for OpenMP from MS VC.NET 2005
DPD200128941        Small destination block size optimization
DPD200132721        Crypto sample error: add -lguide -lpthread in build_ssl_ipp.sh
DPD200133648        ippiSwapChannels_8u_C41R crashes with small image width
DPD200133653        OverFlow in 10-bit DCT ippiDCT8x8Inv_16s_C11()
DPD200134288        Bug in WarpBilinearBack function with NN interpolation
DPD200134697        Problem with ippiFilterGaussBorder_32f_C1R
DPD200133374        Correction to custom DLL section of the User Guide
DPD200133665        Document: Output for example 12-27 in ippsman.pdf is incorrect
DPD200134755        Incorrect decoding for IPP MJPEG sample in UMC
DPD200130124        JPEG color conversion functions
DPD200133693        Line 741 of umc_frame_constuctor.cpp
DPD200134726        Document error: ippiMirror returns ippStsSizeErr
DPD200134712        Distored JPEG output
DPD200134959        Unexpected result with ippiColorToGray with integer inputs
DPD200134967        UMC::ColorSparceConversion YUY2->YV12 error

IPP v6.1 update 1 (02 Jul 2009)

DPD200133747        ippiForegroundGaussian does not produce expected results
DPD200134979        ResizeSqrPixel function crash when size<=6
DPD200133708        ippiResizeSqrPixel_8u_C1R performance issue
DPD200132114        ippiQualityIndex_16u32f_C3R() error
DPD200131846        Inconsistent use of const specifier for some IPP functions
DPD200133631        Improve scripts to display clear error message
DPD200131102        No explanation of nonpic libs (linux) in documentation
DPD200134358        NHM (Intel Nehalem microarchitecture) function optimizations
DPD200133920        ippsFIRSparse is slower than using SSE compiler intrinsics
DPD200134752        Incorrect conversion of RGB to Gray
DPD200134693        redist.txt missing ipp*merged_t.lib
DPD200134748        Redistribute and document libmmds.lib
DPD200134969        Gather interface for accumlulating AAD for AES-GCM
DPD200136077        Sample build problem: file not found: libiomp5.dylib
DPD200082620        G711 bug in PLC
DPD200134973        UMC::ColorSpaceConversion YUY2 -> YV12 error
DPD200134862        Bug in umc_video_processing.cpp
DPD200134779        ippsCIS_32fc_All does not work
DPD200134356        IPP ZLIB fails to compress image after gzrewind
DPD200134270        ippiNormDiff_L1_8u_C1R error
DPD200133638        ippsNorm_L1_16s32f bug
DPD200132645        MPEG4 decoding problem
DPD200132341        Makefile error for UMC
DPD200131704        Resize function error
DPD200131142        ippiUndistortRadial function gives wrong results
DPD200130292        Performance issue for ippsAccCovarianceMatrix_32f64f_D2
DPD200130035        Add notes for ippiDCTQuantInv8x8LS_JPEG_16s16u functions
DPD200133943        Improve MPEG2Mux
DPD200133939        IppiCrossCorrValid_NormLevel_8u32f
DPD200130294        Performance issue for IPP String processing function on EM64T systems
DPD200134815        ippsResamplePolyphase_32f performance issue
DPD200134750        Internal function cpMul_BNU_FullSize does not obey the IPF software conventions
DPD200134716        ippiCrossCorrFull_NormLevel gives unexpected results
DPD200134707        Crash inside ippiInterpolateAverage16x16_8u_C1IR
DPD200133569        Intermediate computational errors
DPD200129130        Wiener filter MaskSize error
DPD200129122        2D media filter performance
DPD200134399        Incorrect Redist.txt file
DPD200134254        Modify IPP ZLIB sample to get better performance
DPD200128923        Watershed segmentation for 32 bit float images added
DPD200134378        Crypto sample does not work with latest OpenSSL package
DPD200133670        Problems encoding 16u_C1 jpeg2000
DPD200130345        Add destructive versions of ippsNthMaxElement_32f_I

IPP v6.1 (25 Apr 2009)

DPD200134199        Feature request: Improve the UMC sample code buiding script
DPD200134679        V-plane Corrupted when input 640x480 YUV422 into encoder
DPD200134370        Bug in 'ippiRGBToYCbCr_8u_C3R'
DPD200134362        ippiTrueDistanceTransform_8u32f_C1R caps the maximum distance value at 1000.0
DPD200134360        ippiDecodeHuffman8x8_Direct_JPEG_1u163_C1 reads beyond input data
DPD200134387        Cannot use libipp_z.a and libipp_bzip2.a at the same time
DPD200134597        Error in color conversion function 'ippiRGBToYCbCr420_8u_C3P3R' for 'u8' variant
DPD200134372        Read error in 'ippiDCTQuantInv8x8To2x2LS_JPEG_16s8u_C1R' for JPEG decoding
DPD200134341        'ippiCompColorKey_8u_C4R' does not work correctly for some ROI widths
DPD200134447        Remove ippStaticInitBest() from Getting_started.htm
DPD200134092        IPP-Crypto- support to RSA_SSA1.5 nor RSA_PKCSv1.5
DPD200133906        ippiCrossCorrValid_NormLevel_32f_C1R on EM64T system
DPD200134376        Bug in colors for planar destinations at lower resolution option in 'JPEGView'
DPD200134386        Issue with ippiTrueDistanceTransform_8u32f_C1R function
DPD200133691        A number of UMC MPEG-4 decoder issues (IPP v. 5.2)
DPD200134366        Missing build64.sh file from data compression sample
DPD200133646        ippsRandomRunform_8u does not produce the upper border value
DPD200133366        Resolution not saved in jpeg
DPD200134179        ippGetStatusString crash in C# sample code
DPD200133821        USC_EC delay issues
DPD200134368        Bug in the state of CABAC in that mode
DPD200133819        Add pixel aspect ratio as a configuration parameter
DPD200134284        Add info about RTI removal to the release notes
DPD200133689        A number of issues for the H.264 UMC decoder
DPD200134397        H264 encoder problem after running many hours
DPD200134487        UMC H264 parser memory leak