diff options
| author | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2013-02-20 12:42:54 +0000 |
|---|---|---|
| committer | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2013-02-20 12:42:54 +0000 |
| commit | 52981c4b6016d9f0e295e0771ec0a50dd073b4b3 (patch) | |
| tree | ccb38cb2e26c468e41f6c039be8acb00bd60b19b /test/Analysis/CostModel | |
| parent | 1479c9bb392325688b72e5829bbb7939c4a079a4 (diff) | |
| download | external_llvm-52981c4b6016d9f0e295e0771ec0a50dd073b4b3.zip external_llvm-52981c4b6016d9f0e295e0771ec0a50dd073b4b3.tar.gz external_llvm-52981c4b6016d9f0e295e0771ec0a50dd073b4b3.tar.bz2 | |
I optimized the following patterns:
sext <4 x i1> to <4 x i64>
sext <4 x i8> to <4 x i64>
sext <4 x i16> to <4 x i64>
I'm running Combine on SIGN_EXTEND_IN_REG and revert SEXT patterns:
(sext_in_reg (v4i64 anyext (v4i32 x )), ExtraVT) -> (v4i64 sext (v4i32 sext_in_reg (v4i32 x , ExtraVT)))
The sext_in_reg (v4i32 x) may be lowered to shl+sar operations.
The "sar" does not exist on 64-bit operation, so lowering sext_in_reg (v4i64 x) has no vector solution.
I also added a cost of this operations to the AVX costs table.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@175619 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test/Analysis/CostModel')
| -rw-r--r-- | test/Analysis/CostModel/X86/cast.ll | 12 |
1 files changed, 11 insertions, 1 deletions
diff --git a/test/Analysis/CostModel/X86/cast.ll b/test/Analysis/CostModel/X86/cast.ll index cedc682..bacc778 100644 --- a/test/Analysis/CostModel/X86/cast.ll +++ b/test/Analysis/CostModel/X86/cast.ll @@ -44,6 +44,10 @@ define i32 @zext_sext(<8 x i1> %in) { %B = zext <8 x i16> undef to <8 x i32> ;CHECK: cost of 1 {{.*}} sext %C = sext <4 x i32> undef to <4 x i64> + ;CHECK: cost of 8 {{.*}} sext + %C1 = sext <4 x i8> undef to <4 x i64> + ;CHECK: cost of 8 {{.*}} sext + %C2 = sext <4 x i16> undef to <4 x i64> ;CHECK: cost of 1 {{.*}} zext %D = zext <4 x i32> undef to <4 x i64> @@ -59,7 +63,7 @@ define i32 @zext_sext(<8 x i1> %in) { ret i32 undef } -define i32 @masks(<8 x i1> %in) { +define i32 @masks8(<8 x i1> %in) { ;CHECK: cost of 6 {{.*}} zext %Z = zext <8 x i1> %in to <8 x i32> ;CHECK: cost of 9 {{.*}} sext @@ -67,3 +71,9 @@ define i32 @masks(<8 x i1> %in) { ret i32 undef } +define i32 @masks4(<4 x i1> %in) { + ;CHECK: cost of 8 {{.*}} sext + %S = sext <4 x i1> %in to <4 x i64> + ret i32 undef +} + |