diff --git a/lib/Target/ARM/ARMSchedule.td b/lib/Target/ARM/ARMSchedule.td
index 6e3b00e..a5ca773 100644
--- a/lib/Target/ARM/ARMSchedule.td
+++ b/lib/Target/ARM/ARMSchedule.td
@@ -10,11 +10,10 @@
 //===----------------------------------------------------------------------===//
 // Functional units across ARM processors
 //
-def FU_iALU   : FuncUnit; // Integer alu unit
-def FU_iLdSt  : FuncUnit; // Integer load / store unit
-def FU_FpALU  : FuncUnit; // FP alu unit
-def FU_FpLdSt : FuncUnit; // FP load / store unit
-def FU_Br     : FuncUnit; // Branch unit
+def FU_Pipe0   : FuncUnit; // pipeline 0 issue
+def FU_Pipe1   : FuncUnit; // pipeline 1 issue
+def FU_LdSt0   : FuncUnit; // pipeline 0 load/store
+def FU_LdSt1   : FuncUnit; // pipeline 1 load/store
 
 //===----------------------------------------------------------------------===//
 // Instruction Itinerary classes used for ARM
@@ -30,7 +29,16 @@
 //===----------------------------------------------------------------------===//
 // Processor instruction itineraries.
 
-def GenericItineraries : ProcessorItineraries<[]>;
+def GenericItineraries : ProcessorItineraries<[
+  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_iLoad   , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  InstrItinData<IIC_fpLoad  , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_fpALU   , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_Br      , [InstrStage<1, [FU_Pipe0]>]>
+]>;
+
 
 include "ARMScheduleV6.td"
 include "ARMScheduleV7.td"
diff --git a/lib/Target/ARM/ARMScheduleV6.td b/lib/Target/ARM/ARMScheduleV6.td
index 0cc5c36..f0b8116 100644
--- a/lib/Target/ARM/ARMScheduleV6.td
+++ b/lib/Target/ARM/ARMScheduleV6.td
@@ -11,12 +11,18 @@
 //
 //===----------------------------------------------------------------------===//
 
+// Single issue pipeline so every itinerary starts with FU_pipe0
 def V6Itineraries : ProcessorItineraries<[
-  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_iALU]>]>,
-  InstrItinData<IIC_iLoad   , [InstrStage<2, [FU_iLdSt]>]>,
-  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_iLdSt]>]>,
-  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_FpALU]>]>,
-  InstrItinData<IIC_fpLoad  , [InstrStage<2, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_Br      , [InstrStage<3, [FU_Br]>]>
+  // single-cycle integer ALU
+  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_Pipe0]>]>,
+  // loads have an extra cycle of latency, but are fully pipelined
+  InstrItinData<IIC_iLoad   , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  InstrItinData<IIC_fpLoad  , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  // fully-pipelined stores
+  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_Pipe0]>]>,
+  // fp ALU is not pipelined
+  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_Pipe0]>]>,
+  // no delay slots, so the latency of a branch is unimportant
+  InstrItinData<IIC_Br      , [InstrStage<1, [FU_Pipe0]>]>
 ]>;
diff --git a/lib/Target/ARM/ARMScheduleV7.td b/lib/Target/ARM/ARMScheduleV7.td
index 328bb98..30360bc 100644
--- a/lib/Target/ARM/ARMScheduleV7.td
+++ b/lib/Target/ARM/ARMScheduleV7.td
@@ -11,23 +11,34 @@
 //
 //===----------------------------------------------------------------------===//
 
+// Single issue pipeline so every itinerary starts with FU_Pipe0
 def V7Itineraries : ProcessorItineraries<[
-  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_iALU]>]>,
-  InstrItinData<IIC_iLoad   , [InstrStage<2, [FU_iLdSt]>]>,
-  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_iLdSt]>]>,
-  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_FpALU]>]>,
-  InstrItinData<IIC_fpLoad  , [InstrStage<2, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_Br      , [InstrStage<3, [FU_Br]>]>
+  // single-cycle integer ALU
+  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_Pipe0]>]>,
+  // loads have an extra cycle of latency, but are fully pipelined
+  InstrItinData<IIC_iLoad   , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  InstrItinData<IIC_fpLoad  , [InstrStage<1, [FU_Pipe0]>, InstrStage<1, [FU_LdSt0]>]>,
+  // fully-pipelined stores
+  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_Pipe0]>]>,
+  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_Pipe0]>]>,
+  // fp ALU is not pipelined
+  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_Pipe0]>]>,
+  // no delay slots, so the latency of a branch is unimportant
+  InstrItinData<IIC_Br      , [InstrStage<1, [FU_Pipe0]>]>
 ]>;
 
-
+// Dual issue pipeline so every itinerary starts with FU_Pipe0 | FU_Pipe1
 def CortexA8Itineraries : ProcessorItineraries<[
-  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_iALU]>]>,
-  InstrItinData<IIC_iLoad   , [InstrStage<2, [FU_iLdSt]>]>,
-  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_iLdSt]>]>,
-  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_FpALU]>]>,
-  InstrItinData<IIC_fpLoad  , [InstrStage<2, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_FpLdSt]>]>,
-  InstrItinData<IIC_Br      , [InstrStage<3, [FU_Br]>]>
+  // single-cycle integer ALU
+  InstrItinData<IIC_iALU    , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>]>,
+  // loads have an extra cycle of latency, but are fully pipelined
+  InstrItinData<IIC_iLoad   , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>, InstrStage<1, [FU_LdSt0]>]>,
+  InstrItinData<IIC_fpLoad  , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>, InstrStage<1, [FU_LdSt0]>]>,
+  // fully-pipelined stores
+  InstrItinData<IIC_iStore  , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>]>,
+  InstrItinData<IIC_fpStore , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>]>,
+  // fp ALU is not pipelined
+  InstrItinData<IIC_fpALU   , [InstrStage<6, [FU_Pipe0, FU_Pipe1]>]>,
+  // no delay slots, so the latency of a branch is unimportant
+  InstrItinData<IIC_Br      , [InstrStage<1, [FU_Pipe0, FU_Pipe1]>]>
 ]>;