6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-11-19 01:13:25 +00:00
Code Issues Projects Releases Wiki Activity
10ac96b276
llvm-6502/test/CodeGen/X86/2004-02-12-Memcpy.ll

26 lines
832 B
LLVM
Raw Normal View History

Fix a number of byval / memcpy / memset related codegen issues. 1. x86-64 byval alignment should be max of 8 and alignment of type. Previously the code was not doing what the commit message was saying. 2. Do not use byte repeat move and store operations. These are slow. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55139 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-21 21:00:15 +00:00
; RUN: llvm-as < %s | llc -march=x86 -mtriple=i686-pc-linux-gnu | grep movs | count 1
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
call libc memcpy/memset if array size is bigger then threshold. Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41479 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27 10:18:20 +00:00
@A = global [32 x i32] zeroinitializer
@B = global [32 x i32] zeroinitializer
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
call libc memcpy/memset if array size is bigger then threshold. Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41479 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27 10:18:20 +00:00
declare void @llvm.memcpy.i32(i8*, i8*, i32, i32)
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
Fix a number of byval / memcpy / memset related codegen issues. 1. x86-64 byval alignment should be max of 8 and alignment of type. Previously the code was not doing what the commit message was saying. 2. Do not use byte repeat move and store operations. These are slow. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55139 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-21 21:00:15 +00:00
define void @main() nounwind {
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
; dword copy
call libc memcpy/memset if array size is bigger then threshold. Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41479 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27 10:18:20 +00:00
call void @llvm.memcpy.i32(i8* bitcast ([32 x i32]* @A to i8*),
i8* bitcast ([32 x i32]* @B to i8*),
i32 128, i32 4 )
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
; word copy
call libc memcpy/memset if array size is bigger then threshold. Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41479 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27 10:18:20 +00:00
call void @llvm.memcpy.i32( i8* bitcast ([32 x i32]* @A to i8*),
i8* bitcast ([32 x i32]* @B to i8*),
i32 128, i32 2 )
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
; byte copy
call libc memcpy/memset if array size is bigger then threshold. Coping 100MB array (after a warmup) shows that glibc 2.6.1 implementation on x86-64 (core 2) is 30% faster (from 0.270917s to 0.188079s) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@41479 91177308-0d34-0410-b5e6-96231b3b80d8
2007-08-27 10:18:20 +00:00
call void @llvm.memcpy.i32( i8* bitcast ([32 x i32]* @A to i8*),
i8* bitcast ([32 x i32]* @B to i8*),
i32 128, i32 1 )
Test that we get rep movs when calling memcpy git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@11352 91177308-0d34-0410-b5e6-96231b3b80d8
2004-02-12 17:53:43 +00:00
ret void
}
Reference in New Issue Copy Permalink