projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ade3075) | patch
ARM: implement some simple f64 materializations.
authorTim Northover <tnorthover@apple.com>
Tue, 20 Aug 2013 08:57:11 +0000 (08:57 +0000)
committerTim Northover <tnorthover@apple.com>
Tue, 20 Aug 2013 08:57:11 +0000 (08:57 +0000)
commit32c2bfda77d54ca6ad8e08d2de03daa7ae432305
treef5439b7924c93d46f23a17213b63e6cb49c47893tree | snapshot
parentade3075030ddd6db370649993a6da5e21e73daabcommit | diff
ARM: implement some simple f64 materializations.

Previously we used a const-pool load for virtually all 64-bit floating values.
Actually, we can get quite a few common values (including 0.0, 1.0) via "vmov"
instructions of one stripe or another.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@188773 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/ARM/ARMISelLowering.cpp diff | blob | history
test/CodeGen/ARM/constantfp.ll [new file with mode: 0644] blob
test/CodeGen/ARM/reg_sequence.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors