projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ee868b2) | patch
[ARM] Match fminnan/fmaxnan for vector vmin/vmax instead of an intrinsic
authorJames Molloy <james.molloy@arm.com>
Tue, 11 Aug 2015 12:06:28 +0000 (12:06 +0000)
committerJames Molloy <james.molloy@arm.com>
Tue, 11 Aug 2015 12:06:28 +0000 (12:06 +0000)
commitd616c642bb02ceef5ae0abf7de74195c0457ab4d
treef8aae542546679389f31329181032692511c6865tree | snapshot
parentee868b2a3e5180d075e08a7466a501c1d3b98711commit | diff
[ARM] Match fminnan/fmaxnan for vector vmin/vmax instead of an intrinsic

Lower Intrinsic::arm_neon_vmins/vmaxs to fminnan/fmaxnan and match that instead. This is important because SDAG will soon be able to select FMINNAN itself, so we need a unified lowering path for intrinsics and SDAG.

NFCI.

llvm-svn: 244593
llvm/lib/Target/ARM/ARMISelLowering.cpp diff | blob | history
llvm/lib/Target/ARM/ARMInstrNEON.td diff | blob | history
Domain: System / Toolchain;