1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-21 20:12:56 +02:00
llvm-mirror/test
Cong Hou 0b6d5e284f Add a flag vectorizer-maximize-bandwidth in loop vectorizer to enable using larger vectorization factor.
To be able to maximize the bandwidth during vectorization, this patch provides a new flag vectorizer-maximize-bandwidth. When it is turned on, the vectorizer will determine the vectorization factor (VF) using the smallest instead of widest type in the loop. To avoid increasing register pressure too much, estimates of the register usage for different VFs are calculated so that we only choose a VF when its register usage doesn't exceed the number of available registers.

This is the second attempt to submit this patch. The first attempt got a test failure on ARM. This patch is updated to try to fix the failure (more specifically, by handling the case when VF=1).

Differential revision: http://reviews.llvm.org/D8943

llvm-svn: 251850
2015-11-02 22:53:48 +00:00
..
Analysis [SCEV] Fix PR25369 2015-11-02 02:06:01 +00:00
Assembler
Bindings
Bitcode Revert "Don't assert if materializing before seeing any function bodies" 2015-10-30 00:00:58 +00:00
BugPoint
CodeGen In MachineBlockPlacement, filter cold blocks off the loop chain when profile data is available. 2015-11-02 21:24:00 +00:00
DebugInfo [LLVMSymbolize] Use symbol table only if function linkage name was requested. 2015-10-26 20:12:29 +00:00
Examples
ExecutionEngine RuntimeDyld: add COFF i386 support 2015-11-01 01:26:15 +00:00
Feature [Inliner] Don't inline through callsites with operand bundles 2015-10-23 20:09:55 +00:00
FileCheck
Instrumentation [PGO] Do not emit runtime hook user function for Linux 2015-10-29 04:08:31 +00:00
Integer
JitListener
LibDriver
Linker Revert "Support for ThinLTO function importing and symbol linking." 2015-11-02 22:17:32 +00:00
LTO
MC MachO: support tvOS and watchOS version min commands in llvm-objdump 2015-11-02 21:26:58 +00:00
Object Tests: be slightly more specific to avoid conflict with path. 2015-10-26 13:40:03 +00:00
Other
SymbolRewriter
TableGen
tools Implemented the code to make llvm-nm’s -g option work. 2015-10-30 19:55:32 +00:00
Transforms Add a flag vectorizer-maximize-bandwidth in loop vectorizer to enable using larger vectorization factor. 2015-11-02 22:53:48 +00:00
Unit
Verifier Add verification for align, dereferenceable, dereferenceable_or_null load metadata 2015-10-09 17:41:29 +00:00
YAMLParser
.clang-format
CMakeLists.txt [CMake] check-llvm requires llvm-pdbdump. 2015-10-15 13:22:38 +00:00
lit.cfg Suppress LLVM::tools/llvm-symbolizer/coff-dwarf.test for mingw, for now. 2015-10-10 02:57:02 +00:00
lit.site.cfg.in
Makefile
Makefile.tests
TestRunner.sh