projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 66c0ebf) | patch
ARM: fix @llvm.convert.from.fp16 on softfloat targets.
authorTim Northover <tnorthover@apple.com>
Tue, 29 Jul 2014 09:56:38 +0000 (09:56 +0000)
committerTim Northover <tnorthover@apple.com>
Tue, 29 Jul 2014 09:56:38 +0000 (09:56 +0000)
commitb1692ad2c3e35a2d1f6c5ea0e121579445078ba6
tree276600ce641e65e682e30e449ff75f9a306ac1b4tree | snapshot
parent66c0ebf171f7f02e04c1ec8347a3a2ebbe5efef2commit | diff
ARM: fix @llvm.convert.from.fp16 on softfloat targets.

We need to make sure we use the softened version of all appropriate operands in
the libcall, or things go horribly wrong. This may entail actually executing a
1-stage softening.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@214175 91177308-0d34-0410-b5e6-96231b3b80d8
lib/CodeGen/SelectionDAG/LegalizeFloatTypes.cpp diff | blob | history
test/CodeGen/ARM/fp16.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors