From: Simon Pilgrim Date: Sat, 25 Oct 2014 08:11:20 +0000 (+0000) Subject: [X86][SSE] Vector integer/float conversion memory folding X-Git-Url: http://plrg.eecs.uci.edu/git/?p=oota-llvm.git;a=commitdiff_plain;h=c31aaa5a3fcef2851898fb30c61c16a70564079a;hp=ae7aa60b2dfa7bba9bb6852b0f8fad0134b18018;ds=sidebyside [X86][SSE] Vector integer/float conversion memory folding Tidied up some entries in the folding tables so that they are under the correct comment section (they were categorised as AVX2 instructions when they're AVX1). Minor patch agreed with qcolombet. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220613 91177308-0d34-0410-b5e6-96231b3b80d8 --- diff --git a/lib/Target/X86/X86InstrInfo.cpp b/lib/Target/X86/X86InstrInfo.cpp index ff41e48c8f2..b5cbee566f2 100644 --- a/lib/Target/X86/X86InstrInfo.cpp +++ b/lib/Target/X86/X86InstrInfo.cpp @@ -566,6 +566,13 @@ X86InstrInfo::X86InstrInfo(X86Subtarget &STI) { X86::VMOVUPSYrr, X86::VMOVUPSYrm, 0 }, { X86::VPERMILPDYri, X86::VPERMILPDYmi, 0 }, { X86::VPERMILPSYri, X86::VPERMILPSYmi, 0 }, + { X86::VRCPPSYr, X86::VRCPPSYm, 0 }, + { X86::VRCPPSYr_Int, X86::VRCPPSYm_Int, 0 }, + { X86::VRSQRTPSYr, X86::VRSQRTPSYm, 0 }, + { X86::VSQRTPDYr, X86::VSQRTPDYm, 0 }, + { X86::VSQRTPSYr, X86::VSQRTPSYm, 0 }, + { X86::VBROADCASTSSYrr, X86::VBROADCASTSSYrm, TB_NO_REVERSE }, + { X86::VBROADCASTSDYrr, X86::VBROADCASTSDYrm, TB_NO_REVERSE }, // AVX2 foldable instructions { X86::VPABSBrr256, X86::VPABSBrm256, 0 }, @@ -574,13 +581,6 @@ X86InstrInfo::X86InstrInfo(X86Subtarget &STI) { X86::VPSHUFDYri, X86::VPSHUFDYmi, 0 }, { X86::VPSHUFHWYri, X86::VPSHUFHWYmi, 0 }, { X86::VPSHUFLWYri, X86::VPSHUFLWYmi, 0 }, - { X86::VRCPPSYr, X86::VRCPPSYm, 0 }, - { X86::VRCPPSYr_Int, X86::VRCPPSYm_Int, 0 }, - { X86::VRSQRTPSYr, X86::VRSQRTPSYm, 0 }, - { X86::VSQRTPDYr, X86::VSQRTPDYm, 0 }, - { X86::VSQRTPSYr, X86::VSQRTPSYm, 0 }, - { X86::VBROADCASTSSYrr, X86::VBROADCASTSSYrm, TB_NO_REVERSE }, - { X86::VBROADCASTSDYrr, X86::VBROADCASTSDYrm, TB_NO_REVERSE }, // BMI/BMI2/LZCNT/POPCNT/TBM foldable instructions { X86::BEXTR32rr, X86::BEXTR32rm, 0 },