diff options
| author | Evan Cheng <evan.cheng@apple.com> | 2009-12-18 07:40:29 +0000 |
|---|---|---|
| committer | Evan Cheng <evan.cheng@apple.com> | 2009-12-18 07:40:29 +0000 |
| commit | 400073d5467b79534d8c63b0d996a55e4252ff4b (patch) | |
| tree | f7204e84da8877e7b062f05bcb1878a05108b44e /lib/Target/X86/X86Subtarget.cpp | |
| parent | 3a5d409f3c2eccf1d1f0a4616023760829a4db67 (diff) | |
On recent Intel u-arch's, folding loads into some unary SSE instructions can
be non-optimal. To be precise, we should avoid folding loads if the instructions
only update part of the destination register, and the non-updated part is not
needed. e.g. cvtss2sd, sqrtss. Unfolding the load from these instructions breaks
the partial register dependency and it can improve performance. e.g.
movss (%rdi), %xmm0
cvtss2sd %xmm0, %xmm0
instead of
cvtss2sd (%rdi), %xmm0
An alternative method to break dependency is to clear the register first. e.g.
xorps %xmm0, %xmm0
cvtss2sd (%rdi), %xmm0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@91672 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'lib/Target/X86/X86Subtarget.cpp')
| -rw-r--r-- | lib/Target/X86/X86Subtarget.cpp | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/lib/Target/X86/X86Subtarget.cpp b/lib/Target/X86/X86Subtarget.cpp index 75cdbada1b..4db3fdb82e 100644 --- a/lib/Target/X86/X86Subtarget.cpp +++ b/lib/Target/X86/X86Subtarget.cpp @@ -266,6 +266,7 @@ void X86Subtarget::AutoDetectSubtargetFeatures() { unsigned Model = 0; DetectFamilyModel(EAX, Family, Model); IsBTMemSlow = IsAMD || (Family == 6 && Model >= 13); + BreakSSEDep = IsIntel; GetCpuIDAndInfo(0x80000001, &EAX, &EBX, &ECX, &EDX); HasX86_64 = (EDX >> 29) & 0x1; @@ -286,6 +287,7 @@ X86Subtarget::X86Subtarget(const std::string &TT, const std::string &FS, , HasFMA3(false) , HasFMA4(false) , IsBTMemSlow(false) + , BreakSSEDep(false) , DarwinVers(0) , stackAlignment(8) // FIXME: this is a known good value for Yonah. How about others? |