replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
324,998 Commits 2 Branches 0 Tags 1.8 GiB
C++ 43.7%
LLVM 32.9%
C 12.3%
Assembly 7.4%
Python 1.1%
Other 2.2%
Go to file
Andrea Di Biagio c9649eb9da [X86][BtVer2] Fix latency/throughput of scalar integer MUL instructions. 
Single operand MUL instructions that implicitly set EAX have the following
latency/throughput profile (see below):

imul %cl              # latency: 3cy - uOPs: 1 - 1 JMul
imul %cx              # latency: 3cy - uOPs: 3 - 3 JMul
imul %ecx             # latency: 3cy - uOPs: 2 - 2 JMul
imul %rcx             # latency: 6cy - uOPs: 2 - 4 JMul

mul %cl               # latency: 3cy - uOPs: 1 - 1 JMul
mul %cx               # latency: 3cy - uOPs: 3 - 3 JMul
mul %ecx              # latency: 3cy - uOPs: 2 - 2 JMul
mul %rcx              # latency: 6cy - uOPs: 2 - 4 JMul

Excluding the 64bit variant, which has a latency of 6cy, every other instruction
has a latency of 3cy. However, the number of decoded macro-opcodes (as well as
the resource cyles) depend on the MUL size.

The two operand MULs have a more predictable profile (see below):

imul %dx, %dx         # latency: 3cy - uOPs: 1 - 1 JMul
imul %edx, %edx       # latency: 3cy - uOPs: 1 - 1 JMul
imul %rdx, %rdx       # latency: 6cy - uOPs: 1 - 4 JMul

imul $3, %dx, %dx     # latency: 4cy - uOPs: 2 - 2 JMul
imul $3, %ecx, %ecx   # latency: 3cy - uOPs: 1 - 1 JMul
imul $3, %rdx, %rdx   # latency: 6cy - uOPs: 1 - 4 JMul

This patch updates the values in the Jaguar scheduling model and regenerates
llvm-mca tests.

Differential Revision: https://reviews.llvm.org/D66547

llvm-svn: 369661
 		2019-08-22 15:20:16 +00:00
clang Revert r369402 "win: Enable /Zc:twoPhase by default if targeting MSVC 2017 update 3 or newer" 2019-08-22 13:15:36 +00:00
clang-tools-extra [clangd] Send suppported codeActionKinds to the client. 2019-08-22 14:53:45 +00:00
compiler-rt Revert "[GWP-ASan] Remove c++ standard lib dependency." 2019-08-22 07:03:38 +00:00
debuginfo-tests Make nrvo-string test more robust. 2019-06-27 20:38:37 +00:00
libclc Revert "Rename F_{None,Text,Append} to OF_{None,Text,Append}. NFC" 2019-08-05 17:23:23 +00:00
libcxx libcxx: Rename last two .hpp files in libcxx to .h 2019-08-21 22:38:38 +00:00
libcxxabi Provide a meaningful diagnostic when LLVM_PATH doesn't point to a directory 2019-08-19 23:51:26 +00:00
libunwind [runtimes] Don't depend on libpthread on Android 2019-07-22 20:41:03 +00:00
lld [lld-link] implement -lto-obj-path 2019-08-21 18:24:59 +00:00
lldb [lldb] Remove ')' to fix the build 2019-08-22 15:18:40 +00:00
llgo IR: Support parsing numeric block ids, and emit them in textual output. 2019-03-22 18:27:13 +00:00
llvm [X86][BtVer2] Fix latency/throughput of scalar integer MUL instructions. 2019-08-22 15:20:16 +00:00
openmp Force honoring nthreads-var and thread-limit-var inside teams construct on host 2019-08-20 19:39:17 +00:00
parallel-libs
…
polly [Polly] Migrate llvm::make_unique to std::make_unique 2019-08-14 22:28:12 +00:00
pstl [pstl] Allow customizing whether per-TU insulation is provided 2019-08-13 12:49:00 +00:00
.arcconfig
…
.clang-format
…
.clang-tidy Disable tidy checks with too many hits 2019-02-01 11:20:13 +00:00
.gitignore Change /build to /build* in top-level .gitignore. 2019-08-02 02:14:08 +00:00
README.md
…

README.md

The LLVM Compiler Infrastructure

This directory and its subdirectories contain source code for LLVM, a toolkit for the construction of highly optimized compilers, optimizers, and runtime environments.