projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 8d1cf22) | patch
[AArch64] Match float round and convert to int instructions.
authorGeoff Berry <gberry@codeaurora.org>
Tue, 28 Jul 2015 15:24:10 +0000 (15:24 +0000)
committerGeoff Berry <gberry@codeaurora.org>
Tue, 28 Jul 2015 15:24:10 +0000 (15:24 +0000)
commit1a0868d554dbd70b4809686939adae58b76e826a
tree95a226e7343903b0f38e4d13e995fab0ea700b59tree | snapshot
parent8d1cf2218d1d03171a4716816196975959aceb5fcommit | diff
[AArch64] Match float round and convert to int instructions.

Summary:
Add patterns for doing floating point round with various rounding modes
followed by conversion to int as a single FCVT* instruction.

Reviewers: t.p.northover, jmolloy

Subscribers: aemerson, rengolin, mcrosier, llvm-commits

Differential Revision: http://reviews.llvm.org/D11424

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243422 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/AArch64/AArch64ISelDAGToDAG.cpp diff | blob | history
test/CodeGen/AArch64/round-conv.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors