From: Evan Cheng Date: Fri, 17 Feb 2006 00:00:58 +0000 (+0000) Subject: Remove the entry about using movapd for SSE reg-reg moves. X-Git-Url: http://plrg.eecs.uci.edu/git/?a=commitdiff_plain;h=dc8acb6420051b09a41742a0fcee2648fe688671;p=oota-llvm.git Remove the entry about using movapd for SSE reg-reg moves. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@26245 91177308-0d34-0410-b5e6-96231b3b80d8 --- diff --git a/lib/Target/X86/README.txt b/lib/Target/X86/README.txt index 50d87bc173a..b794176fa53 100644 --- a/lib/Target/X86/README.txt +++ b/lib/Target/X86/README.txt @@ -356,12 +356,6 @@ This may just be a matter of using 'test' to write bigger patterns for X86cmp. //===---------------------------------------------------------------------===// -Evaluate whether using movapd for SSE reg-reg moves is faster than using -movsd/movss for them. It may eliminate false partial register dependences by -writing the whole result register. - -//===---------------------------------------------------------------------===// - SSE should implement 'select_cc' using 'emulated conditional moves' that use pcmp/pand/pandn/por to do a selection instead of a conditional branch: