[llvm] r261675 - [AArch64] Generate csinv instruction more often
Geoff Berry via llvm-commits
llvm-commits at lists.llvm.org
Tue Feb 23 11:34:14 PST 2016
Author: gberry
Date: Tue Feb 23 13:34:13 2016
New Revision: 261675
URL: http://llvm.org/viewvc/llvm-project?rev=261675&view=rev
Log:
[AArch64] Generate csinv instruction more often
Reviewers: t.p.northover, jmolloy
Subscribers: aemerson, rengolin, mcrosier, llvm-commits
Differential Revision: http://reviews.llvm.org/D17546
Modified:
llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td
llvm/trunk/test/CodeGen/AArch64/cond-sel.ll
llvm/trunk/test/CodeGen/AArch64/fp16-v4-instructions.ll
Modified: llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td?rev=261675&r1=261674&r2=261675&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td (original)
+++ llvm/trunk/lib/Target/AArch64/AArch64InstrInfo.td Tue Feb 23 13:34:13 2016
@@ -1129,6 +1129,14 @@ def : Pat<(AArch64csel (i32 0), (i32 -1)
(CSINVWr WZR, WZR, (i32 imm:$cc))>;
def : Pat<(AArch64csel (i64 0), (i64 -1), (i32 imm:$cc), NZCV),
(CSINVXr XZR, XZR, (i32 imm:$cc))>;
+def : Pat<(AArch64csel GPR32:$tval, (i32 -1), (i32 imm:$cc), NZCV),
+ (CSINVWr GPR32:$tval, WZR, (i32 imm:$cc))>;
+def : Pat<(AArch64csel GPR64:$tval, (i64 -1), (i32 imm:$cc), NZCV),
+ (CSINVXr GPR64:$tval, XZR, (i32 imm:$cc))>;
+def : Pat<(AArch64csel (i32 -1), GPR32:$fval, (i32 imm:$cc), NZCV),
+ (CSINVWr GPR32:$fval, WZR, (i32 (inv_cond_XFORM imm:$cc)))>;
+def : Pat<(AArch64csel (i64 -1), GPR64:$fval, (i32 imm:$cc), NZCV),
+ (CSINVXr GPR64:$fval, XZR, (i32 (inv_cond_XFORM imm:$cc)))>;
// The inverse of the condition code from the alias instruction is what is used
// in the aliased instruction. The parser all ready inverts the condition code
Modified: llvm/trunk/test/CodeGen/AArch64/cond-sel.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/cond-sel.ll?rev=261675&r1=261674&r2=261675&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/AArch64/cond-sel.ll (original)
+++ llvm/trunk/test/CodeGen/AArch64/cond-sel.ll Tue Feb 23 13:34:13 2016
@@ -135,6 +135,34 @@ define void @test_csinv(i32 %lhs32, i32
; CHECK: ret
}
+define void @test_csinv0(i32 %lhs32, i32 %rhs32, i64 %lhs64, i64 %rhs64) minsize {
+; CHECK-LABEL: test_csinv0:
+
+ %tst1 = icmp ugt i32 %lhs32, %rhs32
+ %val1 = select i1 %tst1, i32 0, i32 -1
+ store volatile i32 %val1, i32* @var32
+; CHECK: cmp [[LHS:w[0-9]+]], [[RHS:w[0-9]+]]
+; CHECK: csetm {{w[0-9]+}}, ls
+
+ %rhs2 = add i32 %rhs32, 42
+ %tst2 = icmp sle i32 %lhs32, %rhs2
+ %val2 = select i1 %tst2, i32 -1, i32 %rhs2
+ store volatile i32 %val2, i32* @var32
+; CHECK: cmp [[LHS2:w[0-9]+]], [[RHS2:w[0-9]+]]
+; CHECK: csinv {{w[0-9]+}}, [[RHS2]], wzr, gt
+
+; Note that commuting rhs and lhs in the select changes ugt to ule (i.e. hi to ls).
+ %rhs3 = mul i64 %rhs64, 19
+ %tst3 = icmp ugt i64 %lhs64, %rhs3
+ %val3 = select i1 %tst3, i64 %rhs3, i64 -1
+ store volatile i64 %val3, i64* @var64
+; CHECK: cmp [[LHS3:x[0-9]+]], [[RHS3:x[0-9]+]]
+; CHECK: csinv {{x[0-9]+}}, [[RHS3]], xzr, hi
+
+ ret void
+; CHECK: ret
+}
+
define void @test_csneg(i32 %lhs32, i32 %rhs32, i64 %lhs64) minsize {
; CHECK-LABEL: test_csneg:
Modified: llvm/trunk/test/CodeGen/AArch64/fp16-v4-instructions.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AArch64/fp16-v4-instructions.ll?rev=261675&r1=261674&r2=261675&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/AArch64/fp16-v4-instructions.ll (original)
+++ llvm/trunk/test/CodeGen/AArch64/fp16-v4-instructions.ll Tue Feb 23 13:34:13 2016
@@ -277,10 +277,10 @@ define <4 x i16> @fptoui_i16(<4 x half>
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, ne
-; CHECK-DAG: csel {{.*}}, wzr, ne
-; CHECK-DAG: csel {{.*}}, wzr, ne
-; CHECK-DAG: csel {{.*}}, wzr, ne
+; CHECK-DAG: csetm {{.*}}, ne
+; CHECK-DAG: csetm {{.*}}, ne
+; CHECK-DAG: csetm {{.*}}, ne
+; CHECK-DAG: csetm {{.*}}, ne
define <4 x i1> @test_fcmp_une(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp une <4 x half> %a, %b
ret <4 x i1> %1
@@ -296,14 +296,14 @@ define <4 x i1> @test_fcmp_une(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, vs
-; CHECK-DAG: csel {{.*}}, vs
-; CHECK-DAG: csel {{.*}}, vs
-; CHECK-DAG: csel {{.*}}, vs
+; CHECK-DAG: csetm [[REG1:w[0-9]+]], eq
+; CHECK-DAG: csetm [[REG2:w[0-9]+]], eq
+; CHECK-DAG: csetm [[REG3:w[0-9]+]], eq
+; CHECK-DAG: csetm [[REG4:w[0-9]+]], eq
+; CHECK-DAG: csinv {{.*}}, [[REG1]], wzr, vc
+; CHECK-DAG: csinv {{.*}}, [[REG2]], wzr, vc
+; CHECK-DAG: csinv {{.*}}, [[REG3]], wzr, vc
+; CHECK-DAG: csinv {{.*}}, [[REG4]], wzr, vc
define <4 x i1> @test_fcmp_ueq(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ueq <4 x half> %a, %b
ret <4 x i1> %1
@@ -319,10 +319,10 @@ define <4 x i1> @test_fcmp_ueq(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, hi
-; CHECK-DAG: csel {{.*}}, wzr, hi
-; CHECK-DAG: csel {{.*}}, wzr, hi
-; CHECK-DAG: csel {{.*}}, wzr, hi
+; CHECK-DAG: csetm {{.*}}, hi
+; CHECK-DAG: csetm {{.*}}, hi
+; CHECK-DAG: csetm {{.*}}, hi
+; CHECK-DAG: csetm {{.*}}, hi
define <4 x i1> @test_fcmp_ugt(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ugt <4 x half> %a, %b
ret <4 x i1> %1
@@ -338,10 +338,10 @@ define <4 x i1> @test_fcmp_ugt(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, pl
-; CHECK-DAG: csel {{.*}}, wzr, pl
-; CHECK-DAG: csel {{.*}}, wzr, pl
-; CHECK-DAG: csel {{.*}}, wzr, pl
+; CHECK-DAG: csetm {{.*}}, pl
+; CHECK-DAG: csetm {{.*}}, pl
+; CHECK-DAG: csetm {{.*}}, pl
+; CHECK-DAG: csetm {{.*}}, pl
define <4 x i1> @test_fcmp_uge(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp uge <4 x half> %a, %b
ret <4 x i1> %1
@@ -357,10 +357,10 @@ define <4 x i1> @test_fcmp_uge(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, lt
-; CHECK-DAG: csel {{.*}}, wzr, lt
-; CHECK-DAG: csel {{.*}}, wzr, lt
-; CHECK-DAG: csel {{.*}}, wzr, lt
+; CHECK-DAG: csetm {{.*}}, lt
+; CHECK-DAG: csetm {{.*}}, lt
+; CHECK-DAG: csetm {{.*}}, lt
+; CHECK-DAG: csetm {{.*}}, lt
define <4 x i1> @test_fcmp_ult(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ult <4 x half> %a, %b
ret <4 x i1> %1
@@ -376,10 +376,10 @@ define <4 x i1> @test_fcmp_ult(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, le
-; CHECK-DAG: csel {{.*}}, wzr, le
-; CHECK-DAG: csel {{.*}}, wzr, le
-; CHECK-DAG: csel {{.*}}, wzr, le
+; CHECK-DAG: csetm {{.*}}, le
+; CHECK-DAG: csetm {{.*}}, le
+; CHECK-DAG: csetm {{.*}}, le
+; CHECK-DAG: csetm {{.*}}, le
define <4 x i1> @test_fcmp_ule(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ule <4 x half> %a, %b
ret <4 x i1> %1
@@ -395,10 +395,10 @@ define <4 x i1> @test_fcmp_ule(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, vs
-; CHECK-DAG: csel {{.*}}, wzr, vs
-; CHECK-DAG: csel {{.*}}, wzr, vs
-; CHECK-DAG: csel {{.*}}, wzr, vs
+; CHECK-DAG: csetm {{.*}}, vs
+; CHECK-DAG: csetm {{.*}}, vs
+; CHECK-DAG: csetm {{.*}}, vs
+; CHECK-DAG: csetm {{.*}}, vs
define <4 x i1> @test_fcmp_uno(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp uno <4 x half> %a, %b
ret <4 x i1> %1
@@ -414,14 +414,15 @@ define <4 x i1> @test_fcmp_uno(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, gt
-; CHECK-DAG: csel {{.*}}, gt
-; CHECK-DAG: csel {{.*}}, gt
-; CHECK-DAG: csel {{.*}}, gt
+; CHECK-DAG: csetm [[REG1:w[0-9]+]], mi
+; CHECK-DAG: csetm [[REG2:w[0-9]+]], mi
+; CHECK-DAG: csetm [[REG3:w[0-9]+]], mi
+; CHECK-DAG: csetm [[REG4:w[0-9]+]], mi
+; CHECK-DAG: csinv {{.*}}, [[REG1]], wzr, le
+; CHECK-DAG: csinv {{.*}}, [[REG2]], wzr, le
+; CHECK-DAG: csinv {{.*}}, [[REG3]], wzr, le
+; CHECK-DAG: csinv {{.*}}, [[REG4]], wzr, le
+
define <4 x i1> @test_fcmp_one(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp one <4 x half> %a, %b
ret <4 x i1> %1
@@ -437,10 +438,10 @@ define <4 x i1> @test_fcmp_one(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
-; CHECK-DAG: csel {{.*}}, wzr, eq
+; CHECK-DAG: csetm {{.*}}, eq
+; CHECK-DAG: csetm {{.*}}, eq
+; CHECK-DAG: csetm {{.*}}, eq
+; CHECK-DAG: csetm {{.*}}, eq
define <4 x i1> @test_fcmp_oeq(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp oeq <4 x half> %a, %b
ret <4 x i1> %1
@@ -456,10 +457,10 @@ define <4 x i1> @test_fcmp_oeq(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, gt
-; CHECK-DAG: csel {{.*}}, wzr, gt
-; CHECK-DAG: csel {{.*}}, wzr, gt
-; CHECK-DAG: csel {{.*}}, wzr, gt
+; CHECK-DAG: csetm {{.*}}, gt
+; CHECK-DAG: csetm {{.*}}, gt
+; CHECK-DAG: csetm {{.*}}, gt
+; CHECK-DAG: csetm {{.*}}, gt
define <4 x i1> @test_fcmp_ogt(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ogt <4 x half> %a, %b
ret <4 x i1> %1
@@ -475,10 +476,10 @@ define <4 x i1> @test_fcmp_ogt(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, ge
-; CHECK-DAG: csel {{.*}}, wzr, ge
-; CHECK-DAG: csel {{.*}}, wzr, ge
-; CHECK-DAG: csel {{.*}}, wzr, ge
+; CHECK-DAG: csetm {{.*}}, ge
+; CHECK-DAG: csetm {{.*}}, ge
+; CHECK-DAG: csetm {{.*}}, ge
+; CHECK-DAG: csetm {{.*}}, ge
define <4 x i1> @test_fcmp_oge(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp oge <4 x half> %a, %b
ret <4 x i1> %1
@@ -494,10 +495,10 @@ define <4 x i1> @test_fcmp_oge(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
-; CHECK-DAG: csel {{.*}}, wzr, mi
+; CHECK-DAG: csetm {{.*}}, mi
+; CHECK-DAG: csetm {{.*}}, mi
+; CHECK-DAG: csetm {{.*}}, mi
+; CHECK-DAG: csetm {{.*}}, mi
define <4 x i1> @test_fcmp_olt(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp olt <4 x half> %a, %b
ret <4 x i1> %1
@@ -513,10 +514,10 @@ define <4 x i1> @test_fcmp_olt(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, ls
-; CHECK-DAG: csel {{.*}}, wzr, ls
-; CHECK-DAG: csel {{.*}}, wzr, ls
-; CHECK-DAG: csel {{.*}}, wzr, ls
+; CHECK-DAG: csetm {{.*}}, ls
+; CHECK-DAG: csetm {{.*}}, ls
+; CHECK-DAG: csetm {{.*}}, ls
+; CHECK-DAG: csetm {{.*}}, ls
define <4 x i1> @test_fcmp_ole(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ole <4 x half> %a, %b
ret <4 x i1> %1
@@ -532,10 +533,10 @@ define <4 x i1> @test_fcmp_ole(<4 x half
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
; CHECK-DAG: fcvt
-; CHECK-DAG: csel {{.*}}, wzr, vc
-; CHECK-DAG: csel {{.*}}, wzr, vc
-; CHECK-DAG: csel {{.*}}, wzr, vc
-; CHECK-DAG: csel {{.*}}, wzr, vc
+; CHECK-DAG: csetm {{.*}}, vc
+; CHECK-DAG: csetm {{.*}}, vc
+; CHECK-DAG: csetm {{.*}}, vc
+; CHECK-DAG: csetm {{.*}}, vc
define <4 x i1> @test_fcmp_ord(<4 x half> %a, <4 x half> %b) #0 {
%1 = fcmp ord <4 x half> %a, %b
ret <4 x i1> %1
More information about the llvm-commits
mailing list