Beyond LLVM 3.3 having performance optimizations, one of many other features coming to this next compiler infrastructure update is greater support for Intel's AVX2 instruction set extensions.
Our latest benchmarks of the new GCC 4.8 compiler involve running through the CPU optimizations for the latest-generation Intel "Ivy Bridge" processors.
C++14 is the next update for the C++ programming language. While slated as only a minor extension to C++11, there are several new features being proposed.
Recent compiler testing of the latest LLVM/Clang 3.3 SVN code-base has yielded some significant performance boosts for some common C/C++ benchmarks against LLVM/Clang 3.2.
Following from this weekend's Intel Core i7 990X benchmarks on GCC 4.8, here are benchmarks of an ASUS Ultrabook with an Intel Core i3 "Ivy Bridge" benchmark as we explore how the new GCC 4.8 compiler affects its performance.
With yesterday's release of GCC 4.8.0, here are some new benchmarks. For this first roundabout are comparative benchmarks between GCC 4.7.2 and GCC 4.8.0 for an Intel Core i7 990X "Gulftown" system running Ubuntu Linux.
GCC 4.8 has been officially released today as the annual major update to the GNU Compiler Collection.
With GCC 4.8 using C++ as its implementation language of the compiler, some have questioned whether the compiler is as fast as when written in C. Here's some benchmarks showing C vs. C++ performance with GCC.
Last year at the LLVM developers' meeting it was proposed by an Apple engineer the concept of "modules" for C code in LLVM/Clang to replace the common development approach for C/C++ languages of including header files and passing the library to the linker. LLVM modules seek to take a different approach.
Compiler benchmarks at Phoronix commonly look at the performance of resulting binaries while less of a focus is the compilation time and binary sizes. However, a developer has carried out GCC benchmarks of the compilation times and binary sizes in different scenarios for GCC releases going from GCC 4.2 to the upcoming GCC 4.8.
Jakub Jelinek of Red Hat announced the first release candidate of GCC 4.8.0 on Saturday morning.
Developers from the member companies involved with Linaro are planning continued improvements to the LLVM compiler infrastructure that will benefit ARM developers and customers. With this ARM LLVM work there should also be Gallium3D LLVMpipe enhancements.
Building the Linux kernel with LLVM/Clang rather than GCC continues to be a big focus within the embedded Linux community.
AESOP is a new auto-parallelizing C/C++ compiler for shared memory systems. This new open-source compiler was written at the University of Maryland and is now available to the public.
After already having integrated TILE-Gx support into GCC 4.7, Tilera is now calling for the mainlining of its TILE-Gx back-end into LLVM. The LLVM Tile-Gx back-end is needed for the company's forthcoming many-core processor.
How's the NVIDIA Tegra 3 performance with its four Cortex-A9 cores performing under the forthcoming GCC 4.8 release? Here's some new benchmarks, similar to the recent compiler testing with the ARM Cortex-A15.
While Intel Haswell processors won't even be released to the public until later this year, compiler developers have already been working on supporting the new instruction set extensions of Haswell for more than one year. GCC developers already have early Intel Broadwell support ready for GCC 4.8.
Are there any performance improvements in store for GCC 4.8 as it affects the ARMv7 Cortex-A15 processor on SoCs like the Samsung Exynos 5 Dual? Here's some benchmark results to find out.
LLVM's Clang C/C++ compiler front-end is nearing feature completion for supporting C++11, the latest C++ ISO standard.
While LLVM and Clang (and related LLVM projects) remain in heavy development for the 3.3 cycle, up today are some initial compiler benchmarks of LLVM/Clang 3.3 SVN compared to the current stable release.
Aside from improvements to Link-Time Optimizations, run-time library improvements, and a new optimization level, the coming release of GCC 4.8 also features support for yet-to-be-out AMD hardware. AMD's Steamroller "Bulldozer 3" processor is already supported with compiler optimizations and so is AMD Jaguar, their new low-power APU that's rumoured to be in the next-generation consoles.
The release of LLVM 3.3 is still months away, but one of its features already are notable improvements to its loop vectorizer.
Code Synthesis has released version 2.2 of their ODB C++ Object-Relational Mapping (ORM) code.
CLDOC is a document generator for C and C++ projects to create documentation automatically out of comments as well as creating XML-based descriptions of the API. CLDOC is a new project but already sounds promising and is being powered by LLVM's Clang C/C++ front-end for its parser.
It's going on two years since the release of PCC 1.0, but there hasn't been any follow-on Portable C Compiler release nor is there much public-facing development activity happening.
PathScale, the company that's focused on providing high-performance compiler solutions, is hoping to speed up traditional software packages by automatically leveraging the graphics processor when compiling software with the PathScale ENZO compiler.
GCC 4.8 will feature a few improvements when it comes to LTO, a.k.a. Link-Time Optimization, but will this reflect in any greater performance for the resulting binaries?
The AArch64 back-end to LLVM that provides support for the compiler infrastructure to target ARMv8 64-bit hardware, is now enabled within the default build.
Another interesting open-source project has tipped up that is powered by the LLVM compiler infrastructure.
GCC has had support for 64-bit ARM, a.k.a. AArch64, going back to last summer for using the open-source compiler with next-generation ARMv8 hardware. Being merged today is finally support for the LLVM compiler infrastructure with an experimental 64-bit ARM/AArch64 back-end target.
729 Compiler news articles published on Phoronix.