6502/llvm-6502
1
0
Fork 0
mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-11-08 19:06:39 +00:00
Code Issues Projects Releases Wiki Activity
d288259ccd
llvm-6502/test/CodeGen/X86/sandybridge-loads.ll

40 lines
1.2 KiB
LLVM
Raw Normal View History

On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
; RUN: llc -march=x86-64 -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -o - < %s | FileCheck %s
Update to more CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change. All changes were made by the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" grep -q "^; *RUN: *llc.*debug" $NAME && continue grep -q "^; *RUN:.*llvm-objdump" $NAME && continue grep -q "^; *RUN: *opt.*" $NAME && continue TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\([A-Za-z0-9_-]*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC[:]* *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME done This script catches a superset of the cases caught by the script associated with commit r186280. It initially found some false positives due to unusual constructs in a minority of tests; all such cases were disambiguated first in commit r186621. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186624 91177308-0d34-0410-b5e6-96231b3b80d8
2013-07-18 22:47:09 +00:00
;CHECK-LABEL: wideloads:
On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
;CHECK: vmovaps
;CHECK: vinsertf128
On Sandybridge split unaligned 256bit stores into two xmm-sized stores. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172894 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-19 08:38:41 +00:00
;CHECK: vmovaps
On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
;CHECK-NOT: vinsertf128
;CHECK: ret
define void @wideloads(<8 x float>* %a, <8 x float>* %b, <8 x float>* %c) nounwind uwtable noinline ssp {
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230794 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-27 21:17:42 +00:00
%v0 = load <8 x float>, <8 x float>* %a, align 16 ; <---- unaligned!
%v1 = load <8 x float>, <8 x float>* %b, align 32 ; <---- aligned!
On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
%m0 = fcmp olt <8 x float> %v1, %v0
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230794 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-27 21:17:42 +00:00
%v2 = load <8 x float>, <8 x float>* %c, align 32 ; <---- aligned!
On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
%m1 = fcmp olt <8 x float> %v2, %v0
%mand = and <8 x i1> %m1, %m0
%r = zext <8 x i1> %mand to <8 x i32>
On Sandybridge split unaligned 256bit stores into two xmm-sized stores. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172894 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-19 08:38:41 +00:00
store <8 x i32> %r, <8 x i32>* undef, align 32
ret void
}
; CHECK: widestores
; loads:
; CHECK: vmovaps
; CHECK: vmovaps
; stores:
; CHECK: vmovaps
; CHECK: vextractf128
; CHECK: vmovaps
;CHECK: ret
define void @widestores(<8 x float>* %a, <8 x float>* %b, <8 x float>* %c) nounwind uwtable noinline ssp {
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230794 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-27 21:17:42 +00:00
%v0 = load <8 x float>, <8 x float>* %a, align 32
%v1 = load <8 x float>, <8 x float>* %b, align 32
On Sandybridge split unaligned 256bit stores into two xmm-sized stores. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172894 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-19 08:38:41 +00:00
store <8 x float> %v0, <8 x float>* %b, align 32 ; <--- aligned
store <8 x float> %v1, <8 x float>* %a, align 16 ; <--- unaligned
On Sandybridge loading unaligned 256bits using two XMM loads (vmovups and vinsertf128) is faster than using a single vmovups instruction. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@172868 91177308-0d34-0410-b5e6-96231b3b80d8
2013-01-18 23:10:30 +00:00
ret void
}
Reference in New Issue Copy Permalink