[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi #83143

paschalis-mpeis · 2024-02-27T15:18:50Z

TLI mappings for ArmPL modf, sincos, sincospi were removed due to having bugs by:

[TLI][AArch64] Adjust TLI mappings to vector functions taking linear pointers #80296

This patch re-introduces them.

ArmPL 24.04 Release fixed a bug concerning these methods, so they can
be re-introduced to LLVM mappings.

llvmbot · 2024-02-27T15:19:26Z

@llvm/pr-subscribers-llvm-analysis

@llvm/pr-subscribers-llvm-transforms

Author: Paschalis Mpeis (paschalis-mpeis)

Changes

TLI mappings for ArmPL modf, sincos, sincospi were removed due to having bugs by:

#80296

This patch re-introduces them.

NOTE: Merge will happen once ArmPL that fixes these methods is live.

Patch is 28.31 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/83143.diff

3 Files Affected:

(modified) llvm/include/llvm/Analysis/VecFuncs.def (+6)
(modified) llvm/test/Transforms/LoopVectorize/AArch64/veclib-function-calls.ll (+36-30)
(modified) llvm/test/Transforms/Util/add-TLI-mappings.ll (+25-7)

diff --git a/llvm/include/llvm/Analysis/VecFuncs.def b/llvm/include/llvm/Analysis/VecFuncs.def
index 394e4a05fbc0cf..10f1333cf8885c 100644
--- a/llvm/include/llvm/Analysis/VecFuncs.def
+++ b/llvm/include/llvm/Analysis/VecFuncs.def
@@ -1005,6 +1005,8 @@ TLI_DEFINE_VECFUNC("llvm.log2.f32", "armpl_svlog2_f32_x", SCALABLE(4), MASKED, "
 
 TLI_DEFINE_VECFUNC("modf", "armpl_vmodfq_f64", FIXED(2), NOMASK, "_ZGV_LLVM_N2vl8")
 TLI_DEFINE_VECFUNC("modff", "armpl_vmodfq_f32", FIXED(4), NOMASK, "_ZGV_LLVM_N4vl4")
+TLI_DEFINE_VECFUNC("modf", "armpl_svmodf_f64_x", SCALABLE(2), MASKED, "_ZGVsMxvl8")
+TLI_DEFINE_VECFUNC("modff", "armpl_svmodf_f32_x", SCALABLE(4), MASKED, "_ZGVsMxvl4")
 
 TLI_DEFINE_VECFUNC("nextafter", "armpl_vnextafterq_f64", FIXED(2), NOMASK, "_ZGV_LLVM_N2vv")
 TLI_DEFINE_VECFUNC("nextafterf", "armpl_vnextafterq_f32", FIXED(4), NOMASK, "_ZGV_LLVM_N4vv")
@@ -1033,9 +1035,13 @@ TLI_DEFINE_VECFUNC("llvm.sin.f32", "armpl_svsin_f32_x", SCALABLE(4), MASKED, "_Z
 
 TLI_DEFINE_VECFUNC("sincos", "armpl_vsincosq_f64", FIXED(2), NOMASK, "_ZGV_LLVM_N2vl8l8")
 TLI_DEFINE_VECFUNC("sincosf", "armpl_vsincosq_f32", FIXED(4), NOMASK, "_ZGV_LLVM_N4vl4l4")
+TLI_DEFINE_VECFUNC("sincos", "armpl_svsincos_f64_x", SCALABLE(2), MASKED, "_ZGVsMxvl8l8")
+TLI_DEFINE_VECFUNC("sincosf", "armpl_svsincos_f32_x", SCALABLE(4), MASKED, "_ZGVsMxvl4l4")
 
 TLI_DEFINE_VECFUNC("sincospi", "armpl_vsincospiq_f64", FIXED(2), NOMASK, "_ZGV_LLVM_N2vl8l8")
 TLI_DEFINE_VECFUNC("sincospif", "armpl_vsincospiq_f32", FIXED(4), NOMASK, "_ZGV_LLVM_N4vl4l4")
+TLI_DEFINE_VECFUNC("sincospi", "armpl_svsincospi_f64_x", SCALABLE(2), MASKED, "_ZGVsMxvl8l8")
+TLI_DEFINE_VECFUNC("sincospif", "armpl_svsincospi_f32_x", SCALABLE(4), MASKED, "_ZGVsMxvl4l4")
 
 TLI_DEFINE_VECFUNC("sinh", "armpl_vsinhq_f64", FIXED(2), NOMASK, "_ZGV_LLVM_N2v")
 TLI_DEFINE_VECFUNC("sinhf", "armpl_vsinhq_f32", FIXED(4), NOMASK, "_ZGV_LLVM_N4v")
diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/veclib-function-calls.ll b/llvm/test/Transforms/LoopVectorize/AArch64/veclib-function-calls.ll
index dd1495626eb984..d9cc630482fc80 100644
--- a/llvm/test/Transforms/LoopVectorize/AArch64/veclib-function-calls.ll
+++ b/llvm/test/Transforms/LoopVectorize/AArch64/veclib-function-calls.ll
@@ -2925,11 +2925,12 @@ define void @modf_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @modf_f64
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]]) #[[ATTR4:[0-9]+]]
+; ARMPL-SVE:    [[TMP23:%.*]] = call <vscale x 2 x double> @armpl_svmodf_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP22:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @modf_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    [[TMP5:%.*]] = call <2 x double> @armpl_vmodfq_f64(<2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP4:%.*]])
+; ARMPL-SVE-NOPRED:    [[TMP17:%.*]] = call <vscale x 2 x double> @armpl_svmodf_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP16:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    [[DATA:%.*]] = call double @modf(double [[NUM:%.*]], ptr [[GEPB:%.*]]) #[[ATTR64:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -2970,11 +2971,12 @@ define void @modf_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @modf_f32
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]]) #[[ATTR5:[0-9]+]]
+; ARMPL-SVE:    [[TMP23:%.*]] = call <vscale x 4 x float> @armpl_svmodf_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP22:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @modf_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    [[TMP5:%.*]] = call <4 x float> @armpl_vmodfq_f32(<4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP4:%.*]])
+; ARMPL-SVE-NOPRED:    [[TMP17:%.*]] = call <vscale x 4 x float> @armpl_svmodf_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP16:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    [[DATA:%.*]] = call float @modff(float [[NUM:%.*]], ptr [[GEPB:%.*]]) #[[ATTR65:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -3023,7 +3025,7 @@ define void @nextafter_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @nextafter_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svnextafter_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]]) #[[ATTR64:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @nextafter(double [[IN:%.*]], double [[IN]]) #[[ATTR66:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3068,7 +3070,7 @@ define void @nextafter_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @nextafter_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svnextafter_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]]) #[[ATTR65:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @nextafterf(float [[IN:%.*]], float [[IN]]) #[[ATTR67:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3116,7 +3118,7 @@ define void @pow_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @pow_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svpow_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x double> [[WIDE_LOAD]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @pow(double [[IN:%.*]], double [[IN]]) #[[ATTR66:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @pow(double [[IN:%.*]], double [[IN]]) #[[ATTR68:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3161,7 +3163,7 @@ define void @pow_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @pow_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svpow_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x float> [[WIDE_LOAD]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @powf(float [[IN:%.*]], float [[IN]]) #[[ATTR67:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @powf(float [[IN:%.*]], float [[IN]]) #[[ATTR69:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3209,7 +3211,7 @@ define void @sin_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sin_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svsin_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sin(double [[IN:%.*]]) #[[ATTR68:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sin(double [[IN:%.*]]) #[[ATTR70:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3254,7 +3256,7 @@ define void @sin_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sin_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svsin_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinf(float [[IN:%.*]]) #[[ATTR69:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinf(float [[IN:%.*]]) #[[ATTR71:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3297,11 +3299,12 @@ define void @sincos_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @sincos_f64
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR6:[0-9]+]]
+; ARMPL-SVE:    call void @armpl_svsincos_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP23:%.*]], ptr [[TMP24:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @sincos_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    call void @armpl_vsincosq_f64(<2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP5:%.*]], ptr [[TMP6:%.*]])
+; ARMPL-SVE-NOPRED:    call void @armpl_svsincos_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP17:%.*]], ptr [[TMP18:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    call void @sincos(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR72:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -3341,11 +3344,12 @@ define void @sincos_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @sincos_f32
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR7:[0-9]+]]
+; ARMPL-SVE:    call void @armpl_svsincos_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP23:%.*]], ptr [[TMP24:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @sincos_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    call void @armpl_vsincosq_f32(<4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP5:%.*]], ptr [[TMP6:%.*]])
+; ARMPL-SVE-NOPRED:    call void @armpl_svsincos_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP17:%.*]], ptr [[TMP18:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    call void @sincosf(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR73:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -3388,11 +3392,12 @@ define void @sincospi_f64(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @sincospi_f64
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR8:[0-9]+]]
+; ARMPL-SVE:    call void @armpl_svsincospi_f64_x(<vscale x 2 x double> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP23:%.*]], ptr [[TMP24:%.*]], <vscale x 2 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @sincospi_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    call void @armpl_vsincospiq_f64(<2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP5:%.*]], ptr [[TMP6:%.*]])
+; ARMPL-SVE-NOPRED:    call void @armpl_svsincospi_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], ptr [[TMP17:%.*]], ptr [[TMP18:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    call void @sincospi(double [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR74:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -3432,11 +3437,12 @@ define void @sincospi_f32(ptr noalias %a, ptr noalias %b, ptr noalias %c) {
 ;
 ; ARMPL-SVE-LABEL: define void @sincospi_f32
 ; ARMPL-SVE-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR9:[0-9]+]]
+; ARMPL-SVE:    call void @armpl_svsincospi_f32_x(<vscale x 4 x float> [[WIDE_MASKED_LOAD:%.*]], ptr [[TMP23:%.*]], ptr [[TMP24:%.*]], <vscale x 4 x i1> [[ACTIVE_LANE_MASK:%.*]])
 ;
 ; ARMPL-SVE-NOPRED-LABEL: define void @sincospi_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], ptr noalias [[C:%.*]]) #[[ATTR0]] {
-; ARMPL-SVE-NOPRED:    call void @armpl_vsincospiq_f32(<4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP5:%.*]], ptr [[TMP6:%.*]])
+; ARMPL-SVE-NOPRED:    call void @armpl_svsincospi_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], ptr [[TMP17:%.*]], ptr [[TMP18:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
+; ARMPL-SVE-NOPRED:    call void @sincospif(float [[NUM:%.*]], ptr [[GEPB:%.*]], ptr [[GEPC:%.*]]) #[[ATTR75:[0-9]+]]
 ;
 entry:
   br label %for.body
@@ -3484,7 +3490,7 @@ define void @sinh_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sinh_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svsinh_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinh(double [[IN:%.*]]) #[[ATTR70:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinh(double [[IN:%.*]]) #[[ATTR76:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3529,7 +3535,7 @@ define void @sinh_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sinh_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svsinh_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinhf(float [[IN:%.*]]) #[[ATTR71:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinhf(float [[IN:%.*]]) #[[ATTR77:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3577,7 +3583,7 @@ define void @sinpi_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sinpi_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svsinpi_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]]) #[[ATTR72:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sinpi(double [[IN:%.*]]) #[[ATTR78:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3622,7 +3628,7 @@ define void @sinpi_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sinpi_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svsinpi_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]]) #[[ATTR73:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sinpif(float [[IN:%.*]]) #[[ATTR79:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3670,7 +3676,7 @@ define void @sqrt_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sqrt_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svsqrt_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]]) #[[ATTR74:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @sqrt(double [[IN:%.*]]) #[[ATTR80:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3715,7 +3721,7 @@ define void @sqrt_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @sqrt_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 4 x float> @armpl_svsqrt_f32_x(<vscale x 4 x float> [[WIDE_LOAD:%.*]], <vscale x 4 x i1> shufflevector (<vscale x 4 x i1> insertelement (<vscale x 4 x i1> poison, i1 true, i64 0), <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]]) #[[ATTR75:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call float @sqrtf(float [[IN:%.*]]) #[[ATTR81:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3763,7 +3769,7 @@ define void @tan_f64(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @tan_f64
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9:%.*]] = call <vscale x 2 x double> @armpl_svtan_f64_x(<vscale x 2 x double> [[WIDE_LOAD:%.*]], <vscale x 2 x i1> shufflevector (<vscale x 2 x i1> insertelement (<vscale x 2 x i1> poison, i1 true, i64 0), <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer))
-; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tan(double [[IN:%.*]]) #[[ATTR76:[0-9]+]]
+; ARMPL-SVE-NOPRED:    [[CALL:%.*]] = tail call double @tan(double [[IN:%.*]]) #[[ATTR82:[0-9]+]]
 ;
   entry:
   br label %for.body
@@ -3808,7 +3814,7 @@ define void @tan_f32(ptr noalias %in.ptr, ptr noalias %out.ptr) {
 ; ARMPL-SVE-NOPRED-LABEL: define void @tan_f32
 ; ARMPL-SVE-NOPRED-SAME: (ptr noalias [[IN_PTR:%.*]], ptr noalias [[OUT_PTR:%.*]]) #[[ATTR0]] {
 ; ARMPL-SVE-NOPRED:    [[TMP9...
[truncated]

paschalis-mpeis · 2024-04-08T14:35:01Z

Rebased to latest main. This is now ready for review/merge.

mgabka

I think it is worth to be specific and mention that the bug has been fixed in the 24.04 release.

ArmPL 24.04 Release fixed a bug concerning these methods, so they can be re-introduced to LLVM mappings.

paschalis-mpeis · 2024-04-09T10:06:58Z

@mgabka agreed, thanks. Amended the commit message to now include this.

llvmbot added llvm:analysis llvm:transforms labels Feb 27, 2024

paschalis-mpeis requested a review from mgabka February 27, 2024 15:19

paschalis-mpeis mentioned this pull request Feb 27, 2024

[LV][LAA] Vectorize math lib calls with mem write-only attribute #78432

Draft

paschalis-mpeis requested review from paulwalker-arm and huntergr-arm February 28, 2024 09:08

paschalis-mpeis marked this pull request as draft March 8, 2024 09:58

paschalis-mpeis force-pushed the users/paschalis-mpeis/tli-armpl-modf-sincos-sincospi branch from a74ba11 to f28e739 Compare April 8, 2024 12:48

paschalis-mpeis marked this pull request as ready for review April 8, 2024 14:33

mgabka approved these changes Apr 9, 2024

View reviewed changes

[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi

36b0899

ArmPL 24.04 Release fixed a bug concerning these methods, so they can be re-introduced to LLVM mappings.

paschalis-mpeis force-pushed the users/paschalis-mpeis/tli-armpl-modf-sincos-sincospi branch from f28e739 to 36b0899 Compare April 9, 2024 10:04

paulwalker-arm approved these changes Apr 9, 2024

View reviewed changes

paschalis-mpeis merged commit e50c4c8 into main Apr 10, 2024
4 checks passed

paschalis-mpeis deleted the users/paschalis-mpeis/tli-armpl-modf-sincos-sincospi branch April 12, 2024 11:55

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi #83143

[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi #83143

paschalis-mpeis commented Feb 27, 2024 •

edited

Loading

llvmbot commented Feb 27, 2024 •

edited

Loading

paschalis-mpeis commented Apr 8, 2024

mgabka left a comment

paschalis-mpeis commented Apr 9, 2024

[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi #83143

[AArch64][TLI] Add TLI mappings for ArmPL modf, sincos, sincospi #83143

Conversation

paschalis-mpeis commented Feb 27, 2024 • edited Loading

llvmbot commented Feb 27, 2024 • edited Loading

paschalis-mpeis commented Apr 8, 2024

mgabka left a comment

Choose a reason for hiding this comment

paschalis-mpeis commented Apr 9, 2024

paschalis-mpeis commented Feb 27, 2024 •

edited

Loading

llvmbot commented Feb 27, 2024 •

edited

Loading