[ARM] Thumb2: ConstantMaterializationCost

Constants can also be materialised using the negated value and a MVN, and this
case seem to have been missed for Thumb2. To check the constant materialisation
costs, we now call getT2SOImmVal twice, once for the original constant and then
also for its negated value, and this function checks if the constant can both
be splatted or rotated.

This was revealed by a test that optimises for minsize: instead of a LDR
literal pool load and having a literal pool entry, just a MVN with an immediate
is smaller (and also faster).

Differential Revision: https://reviews.llvm.org/D57327

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@352737 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/ARM/ARMISelDAGToDAG.cpp b/lib/Target/ARM/ARMISelDAGToDAG.cpp
index a51cbda4f1df7f864593233690de035e45bc7b1d..17a82e4c800334476b422f913472373ca278679b 100644 (file)
--- a/lib/Target/ARM/ARMISelDAGToDAG.cpp
+++ b/lib/Target/ARM/ARMISelDAGToDAG.cpp
@@ -452,8 +452,10 @@ unsigned ARMDAGToDAGISel::ConstantMaterializationCost(unsigned Val) const {
   if (Subtarget->isThumb()) {
     if (Val <= 255) return 1;                               // MOV
     if (Subtarget->hasV6T2Ops() &&
-        (Val <= 0xffff || ARM_AM::getT2SOImmValSplatVal(Val) != -1))
-      return 1; // MOVW
+        (Val <= 0xffff ||                                   // MOV
+         ARM_AM::getT2SOImmVal(Val) != -1 ||                // MOVW
+         ARM_AM::getT2SOImmVal(~Val) != -1))                // MVN
+      return 1;
     if (Val <= 510) return 2;                               // MOV + ADDi8
     if (~Val <= 255) return 2;                              // MOV + MVN
     if (ARM_AM::isThumbImmShiftedVal(Val)) return 2;        // MOV + LSL

diff --git a/test/CodeGen/ARM/subtarget-no-movt.ll b/test/CodeGen/ARM/subtarget-no-movt.ll
index fb0d5adc3f48bf388f1581273e846aedf457c26b..3fb84f00524f5c408821e17f68d5c3001d708ff9 100644 (file)
--- a/test/CodeGen/ARM/subtarget-no-movt.ll
+++ b/test/CodeGen/ARM/subtarget-no-movt.ll
@@ -1,99 +1,97 @@
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - | FileCheck -check-prefix=NO-OPTION %s
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -mattr=-no-movt | FileCheck -check-prefix=USE-MOVT %s
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -mattr=+no-movt | FileCheck -check-prefix=NO-USE-MOVT %s
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 | FileCheck -check-prefix=NO-OPTION-O0 %s
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 -mattr=-no-movt | FileCheck -check-prefix=USE-MOVT-O0 %s
-; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 -mattr=+no-movt | FileCheck -check-prefix=NO-USE-MOVT-O0 %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - | \
+; RUN:   FileCheck -check-prefixes=CHECK,NO-OPTION,NO-OPTION-COMMON %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -mattr=-no-movt | \
+; RUN:   FileCheck -check-prefixes=CHECK,USE-MOVT,USE-MOVT-COMMON %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -mattr=+no-movt | \
+; RUN:   FileCheck -check-prefixes=CHECK,NO-USE-MOVT,NO-USE-MOVT-COMMON %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 | \
+; RUN:   FileCheck -check-prefixes=CHECK,NO-OPTION-O0,NO-OPTION-COMMON %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 -mattr=-no-movt | \
+; RUN:  FileCheck -check-prefixes=CHECK,USE-MOVT-O0,USE-MOVT-COMMON %s
+; RUN: llc -mcpu=cortex-a8 -relocation-model=static %s -o - -O0 -mattr=+no-movt | \
+; RUN:   FileCheck -check-prefixes=CHECK,NO-USE-MOVT-O0,NO-USE-MOVT-COMMON %s
 
 target triple = "thumb-apple-darwin"
 
-; NO-OPTION-LABEL: {{_?}}foo0
-; NO-OPTION: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-OPTION: [[L0]]:
-; NO-OPTION: .long 2296237089
+; NO-OPTION-COMMON-LABEL: {{_?}}foo0
+; NO-OPTION-COMMON: ldr [[R0:r[0-9]+]], [[L0:.*]]
+; NO-OPTION-COMMON: [[L0]]:
+; NO-OPTION-COMMON: .long 2296237089
 
-; NO-OPTION-O0-LABEL: {{_?}}foo0
-; NO-OPTION-O0: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-OPTION-O0: [[L0]]:
-; NO-OPTION-O0: .long 2296237089
+; USE-MOVT-COMMON-LABEL: {{_?}}foo0
+; USE-MOVT-COMMON: movw [[R0:r[0-9]+]], #52257
+; USE-MOVT-COMMON: movt [[R0]], #35037
 
-; USE-MOVT-LABEL: {{_?}}foo0
-; USE-MOVT: movw [[R0:r[0-9]+]], #52257
-; USE-MOVT: movt [[R0]], #35037
-
-; USE-MOVT-O0-LABEL: {{_?}}foo0
-; USE-MOVT-O0: movw [[R0:r[0-9]+]], #52257
-; USE-MOVT-O0: movt [[R0]], #35037
-
-; NO-USE-MOVT-LABEL: {{_?}}foo0
-; NO-USE-MOVT: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-USE-MOVT: [[L0]]:
-; NO-USE-MOVT: .long 2296237089
-
-; NO-USE-MOVT-O0-LABEL: {{_?}}foo0
-; NO-USE-MOVT-O0: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-USE-MOVT-O0: [[L0]]:
-; NO-USE-MOVT-O0: .long 2296237089
+; NO-USE-MOVT-COMMON-LABEL: {{_?}}foo0
+; NO-USE-MOVT-COMMON: ldr [[R0:r[0-9]+]], [[L0:.*]]
+; NO-USE-MOVT-COMMON: [[L0]]:
+; NO-USE-MOVT-COMMON: .long 2296237089
 
 define i32 @foo0(i32 %a) #0 {
   %1 = xor i32 -1998730207, %a
   ret i32 %1
 }
 
-; NO-OPTION-LABEL: {{_?}}foo1
-; NO-OPTION: movw [[R0:r[0-9]+]], #52257
-; NO-OPTION: movt [[R0]], #35037
-
-; NO-OPTION-O0-LABEL: {{_?}}foo1
-; NO-OPTION-O0: movw [[R0:r[0-9]+]], #52257
-; NO-OPTION-O0: movt [[R0]], #35037
+; NO-OPTION-COMMON-LABEL: {{_?}}foo1
+; NO-OPTION-COMMON: movw [[R0:r[0-9]+]], #52257
+; NO-OPTION-COMMON: movt [[R0]], #35037
 
-; USE-MOVT-LABEL: {{_?}}foo1
-; USE-MOVT: movw [[R0:r[0-9]+]], #52257
-; USE-MOVT: movt [[R0]], #35037
+; USE-MOVT-COMMON-LABEL: {{_?}}foo1
+; USE-MOVT-COMMON: movw [[R0:r[0-9]+]], #52257
+; USE-MOVT-COMMON: movt [[R0]], #35037
 
-; USE-MOVT-O0-LABEL: {{_?}}foo1
-; USE-MOVT-O0: movw [[R0:r[0-9]+]], #52257
-; USE-MOVT-O0: movt [[R0]], #35037
-
-; NO-USE-MOVT-LABEL: {{_?}}foo1
-; NO-USE-MOVT: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-USE-MOVT: [[L0]]:
-; NO-USE-MOVT: .long 2296237089
-
-; NO-USE-MOVT-O0-LABEL: {{_?}}foo1
-; NO-USE-MOVT-O0: ldr [[R0:r[0-9]+]], [[L0:.*]]
-; NO-USE-MOVT-O0: [[L0]]:
-; NO-USE-MOVT-O0: .long 2296237089
+; NO-USE-MOVT-COMMON-LABEL: {{_?}}foo1
+; NO-USE-MOVT-COMMON: ldr [[R0:r[0-9]+]], [[L0:.*]]
+; NO-USE-MOVT-COMMON: [[L0]]:
+; NO-USE-MOVT-COMMON: .long 2296237089
 
 define i32 @foo1(i32 %a) {
   %1 = xor i32 -1998730207, %a
   ret i32 %1
 }
 
-; NO-OPTION-LABEL: {{_?}}foo2
+; NO-OPTION-COMMON-LABEL: {{_?}}foo2
 ; NO-OPTION:   mov.w   [[R0:r[0-9]+]], #-536813568
-
-; USE-MOVT-LABEL: {{_?}}foo2
-; USE-MOVT:    mov.w   [[R0:r[0-9]+]], #-536813568
-
-; NO-USE-MOVT-LABEL: {{_?}}foo2
-; NO-USE-MOVT: mov.w   [[R0:r[0-9]+]], #-536813568
-
-; NO-OPTION-O0-LABEL: {{_?}}foo2
 ; NO-OPTION-O0: movw   [[R0:r[0-9]+]], #57344
 ; NO-OPTION-O0: movt   [[R0]], #57344
 
-; USE-MOVT-O0-LABEL: {{_?}}foo2
-; USE-MOVT-O0:  movw   [[R0:r[0-9]+]], #57344
-; USE-MOVTT-O0: movt   [[R0]], #57344
+; USE-MOVT-COMMON-LABEL: {{_?}}foo2
+; USE-MOVT:     mov.w [[R0:r[0-9]+]], #-536813568
+; USE-MOVT-O0:  movw  [[R0:r[0-9]+]], #57344
+; USE-MOVT-O0:  movt  [[R0]], #57344
 
-; NO-USE-MOVT-O0-LABEL: {{_?}}foo2
+; NO-USE-MOVT-COMMON-LABEL: {{_?}}foo2
+; NO-USE-MOVT: mov.w   [[R0:r[0-9]+]], #-536813568
 ; NO-USE-MOVT-O0: ldr [[R0:r[0-9]+]], [[L0:.*]]
 ; NO-USE-MOVT-O0: [[L0]]:
 ; NO-USE-MOVT-O0: .long 3758153728     @ 0xe000e000
+
 define i32 @foo2() {
   %1 = load i32, i32* inttoptr (i32 -536813568 to i32*) ; load from 0xe000e000
   ret i32 %1
 }
 attributes #0 = { "target-features"="+no-movt" }
+
+define hidden i32 @no_litpool() minsize optsize {
+; CHECK-LABEL:  no_litpool:
+; CHECK:        mov.w r{{.}}, #65536
+; CHECK:        mov.w r{{.}}, #-134217728
+; CHECK:        mvn r{{.}}, #-134217728
+entry:
+  %call0 = tail call i32 @eat_const(i32 65536)
+  %call1 = tail call i32 @eat_const(i32 -134217728)
+  %call2 = tail call i32 @eat_const(i32 134217727)
+  ret i32 %call2
+}
+
+define hidden i32 @litpool() minsize optsize {
+; CHECK-LABEL:  litpool:
+; CHECK:        ldr r0, {{.*}}LCPI{{.*}}
+; CHECK-NEXT:   b.w {{.*}}eat_const
+entry:
+  %call1 = tail call i32 @eat_const(i32 8388601)
+  ret i32 %call1
+}
+
+declare dso_local i32 @eat_const(i32)
+