projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 202249a) | patch
Allow targets to specify the return type of libcalls that are generated for floating...
authorSanjiv Gupta <sanjiv.gupta@microchip.com>
Mon, 28 Dec 2009 02:40:33 +0000 (02:40 +0000)
committerSanjiv Gupta <sanjiv.gupta@microchip.com>
Mon, 28 Dec 2009 02:40:33 +0000 (02:40 +0000)
commit8f17a36d3107bdc4ffed53ce782c1724ef6460e7
tree68daae8400bd7b6a3faf6d0aaa7a7f9e6235f905tree | snapshot
parent202249a6984a1fae296f5a9e8e5d4292251a7ce6commit | diff
Allow targets to specify the return type of libcalls that are generated for floating point comparisons, rather than hard-coding them as i32.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92199 91177308-0d34-0410-b5e6-96231b3b80d8
include/llvm/Target/TargetLowering.h diff | blob | history
lib/CodeGen/SelectionDAG/LegalizeFloatTypes.cpp diff | blob | history
lib/CodeGen/SelectionDAG/TargetLowering.cpp diff | blob | history
lib/Target/PIC16/PIC16ISelLowering.cpp diff | blob | history
lib/Target/PIC16/PIC16ISelLowering.h diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors