diff options
Diffstat (limited to 'llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp')
| -rw-r--r-- | llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp | 15 |
1 files changed, 12 insertions, 3 deletions
diff --git a/llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp b/llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp index b2b7bf9af44..aaef1405157 100644 --- a/llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp +++ b/llvm/lib/Target/R600/AMDGPUAsmPrinter.cpp @@ -55,11 +55,20 @@ using namespace llvm; // We want to use these instructions, and using fp32 denormals also causes // instructions to run at the double precision rate for the device so it's // probably best to just report no single precision denormals. -static uint32_t getFPMode(const MachineFunction &) { +static uint32_t getFPMode(const MachineFunction &F) { + const AMDGPUSubtarget& ST = F.getTarget().getSubtarget<AMDGPUSubtarget>(); + // TODO: Is there any real use for the flush in only / flush out only modes? + + uint32_t FP32Denormals = + ST.hasFP32Denormals() ? FP_DENORM_FLUSH_NONE : FP_DENORM_FLUSH_IN_FLUSH_OUT; + + uint32_t FP64Denormals = + ST.hasFP64Denormals() ? FP_DENORM_FLUSH_NONE : FP_DENORM_FLUSH_IN_FLUSH_OUT; + return FP_ROUND_MODE_SP(FP_ROUND_ROUND_TO_NEAREST) | FP_ROUND_MODE_DP(FP_ROUND_ROUND_TO_NEAREST) | - FP_DENORM_MODE_SP(FP_DENORM_FLUSH_IN_FLUSH_OUT) | - FP_DENORM_MODE_DP(FP_DENORM_FLUSH_NONE); + FP_DENORM_MODE_SP(FP32Denormals) | + FP_DENORM_MODE_DP(FP64Denormals); } static AsmPrinter *createAMDGPUAsmPrinterPass(TargetMachine &tm, |
