SIMD Types: ABI Considerations [N4395]
نویسنده
چکیده
This document discusses the ABI implications from the SIMD types described in [N4184] and [N4185]. I investigate strategies to automatically adapt between different translation units compiled for different microarchitectures of the same architecture. None of the strategies lead to a solution without surprises. The solutions section therefore looks at how the default vector type may need to be declared to make ABI incompatibilities a conscious choice of the user.
منابع مشابه
SIMD Types Example: Matrix Multiplication [N4454]
This document describes one possible implementation of a matrix class and matrix multiplication using the data-parallel SIMD types introduced in [N4184]. The example shows the basic use of SIMD types for manual transformation of a loop over scalars to a loop with increased stride using SIMD vector loads and stores and SIMD operations in the loop body.
متن کاملPorting Linux to x86-64
x86-64 is a 64-bit extension for the IA32 architecture, which is supported by the next generation of AMD CPUs. New features include 64-bit pointers, a 48-bit address space, 16 general purpose 64-bit integer registers, 16 SSE (Streaming SIMD Extensions) registers, and a compatibility mode to support old binaries. The Linux kernel port to x86-64 is based on the existing IA32 port with some extens...
متن کاملExploiting SIMD for complex numerical predicates
We study the use of SIMD instructions to support complex conjunctive numerical predicates. Compared to previous studies, we aim to model more realistic use scenarios, where different data types, different comparison operations, and different predicate types can be mixed in a single filtering clause. Moreover, the evaluation of the predicates on a set of columns can take advantage of multiple pr...
متن کاملDissertation Performance Portable Short Vector Transforms
ing from Special Machine Features In the context of this thesis all short vector SIMD extensions feature the functionality required in intermediate level building blocks. However, the implementation of such building blocks depends on special features of the target architecture. For instance, a complex reordering operation like a permutation has to be implemented using register-register permutat...
متن کاملAdapting Tree Structures for Processing with SIMD Instructions
In this paper, we accelerate the processing of tree-based index structures by using SIMD instructions. We adapt the B+-Tree and prefix B-Tree (trie) by changing the search algorithm on inner nodes from binary search to k-ary search. The k-ary search enables the use of SIMD instructions, which are commonly available on most modern processors today. The main challenge for using SIMD instructions ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2015