diff options
| author | Manman Ren <mren@apple.com> | 2012-08-10 23:43:32 +0000 |
|---|---|---|
| committer | Manman Ren <mren@apple.com> | 2012-08-10 23:43:32 +0000 |
| commit | 743a2cff0422031da007bcdc70ced28c09057aea (patch) | |
| tree | f0c095097969d0d24518d4a56bd8a6d0e03d42cf | |
| parent | eb74c08192d9c9425b2d8cf08852e9e787a87881 (diff) | |
X86: when we are auto-detecting the subtarget features, make sure we turn on
FeatureFastUAMem for Nehalem, Westmere and Sandy Bridge.
FeatureFastUAMem is already on if we pass in nehalem or westmere as a command
argument.
rdar: 7252306
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@161717 91177308-0d34-0410-b5e6-96231b3b80d8
| -rw-r--r-- | lib/Target/X86/X86Subtarget.cpp | 7 | ||||
| -rw-r--r-- | test/CodeGen/X86/alignment-2.ll | 4 |
2 files changed, 8 insertions, 3 deletions
diff --git a/lib/Target/X86/X86Subtarget.cpp b/lib/Target/X86/X86Subtarget.cpp index ae2229827b..d7e3efc4f9 100644 --- a/lib/Target/X86/X86Subtarget.cpp +++ b/lib/Target/X86/X86Subtarget.cpp @@ -246,8 +246,11 @@ void X86Subtarget::AutoDetectSubtargetFeatures() { } // If it's Nehalem, unaligned memory access is fast. - // FIXME: Nehalem is family 6. Also include Westmere and later processors? - if (Family == 15 && Model == 26) { + // Include Westmere and Sandy Bridge as well. + // FIXME: add later processors. + if ((Family == 6 && Model == 26) || + (Family == 6 && Model == 44) || + (Family == 6 && Model == 42)) { IsUAMemFast = true; ToggleFeature(X86::FeatureFastUAMem); } diff --git a/test/CodeGen/X86/alignment-2.ll b/test/CodeGen/X86/alignment-2.ll index cc709b52d9..1f9e85cbb7 100644 --- a/test/CodeGen/X86/alignment-2.ll +++ b/test/CodeGen/X86/alignment-2.ll @@ -18,7 +18,9 @@ define signext i8 @do_lo_list() nounwind optsize ssp { bb: ; CHECK: do_lo_list -; CHECK-NOT: movaps +; Make sure we do not use movaps for the global variable. +; It is okay to use movaps for writing the local variable on stack. +; CHECK-NOT: movaps {{[0-9]*}}(%{{[a-z]*}}), {{%xmm[0-9]}} %myopt = alloca %struct.printQueryOpt, align 4 %tmp = bitcast %struct.printQueryOpt* %myopt to i8* call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* bitcast (%struct.printQueryOpt* getelementptr inbounds (%struct._psqlSettings* @pset, i32 0, i32 4) to i8*), i32 76, i32 4, i1 false) |