Hardware
/
VexRiscv
mirror of https://github.com/SpinalHDL/VexRiscv.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

FpuTest document how to install berkley testfloat

This commit is contained in:
Charles Papon 2023-03-10 14:46:21 +08:00
parent 94f19032f0
commit 25eda80fee
1 changed files with 55 additions and 43 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

98
src/test/scala/vexriscv/ip/fpu/FpuTest.scala
View File

@ -19,7 +19,14 @@ import scala.collection.mutable.ArrayBuffer
import scala.sys.process.ProcessLogger
import scala.util.Random
import org.scalatest.funsuite.AnyFunSuite
/*
How to install :
git clone https://github.com/ucb-bar/berkeley-softfloat-3
(cd berkeley-softfloat-3/build/Linux-x86_64-GCC; make SPECIALIZE_TYPE=RISCV)
git clone https://github.com/ucb-bar/berkeley-testfloat-3
(cd berkeley-testfloat-3/build/Linux-x86_64-GCC; make SPECIALIZE_TYPE=RISCV)
Add the compiled binaries in the PATH
*/
//TODO Warning DataCache write aggregation will disable itself
class FpuTest extends AnyFunSuite{
@ -548,7 +555,7 @@ class FpuTest extends AnyFunSuite{
load(rs2, b)
op(rd,rs1,rs2, rounding, FpuFormat.FLOAT)
storeFloat(rd){v =>
assert(f2b(v) == f2b(ref), f"## ${a} ${opName} $b = $v, $ref $rounding")
assert(f2b(v) == f2b(ref), f"## ${a} ${opName} $b = $v, $ref $rounding, ${f2b(a)}%x ${f2b(v)}%x ${f2b(ref)}%x")
}
flagMatch(flag, ref, f"## ${opName} ${a} $b $ref $rounding")
@ -738,7 +745,7 @@ class FpuTest extends AnyFunSuite{
sqrt(rd,rs1, rounding, FpuFormat.DOUBLE)
store(rd){v =>
assert(d2b(v) == d2b(ref), f"## sqrt${a} = $v, $ref $rounding, ${d2b(a).toString(16)} ${d2b(ref).toString(16)}")
assert(d2b(v) == d2b(ref), f"## sqrt${a} = $v, $ref $rounding, ${d2b(a).toString(16)} ${d2b(ref).toString(16)} ${d2b(v).toString(16)}")
}
flagMatch(flag, ref, f"## sqrt${a} $ref $rounding")
@ -1328,7 +1335,11 @@ class FpuTest extends AnyFunSuite{
// testSqrtF64Exact(-1, -Double.NaN, 0, FpuRoundMode.RNE)
// for(_ <- 0 until 10000) testSgnjF32()
// for(_ <- 0 until 10000) testMulF64()
val runFactor = 1
//TODO test boxing
//TODO double <-> simple convertions
@ -1337,7 +1348,7 @@ class FpuTest extends AnyFunSuite{
testSqrtF64Exact(1.25*1.25, 1.25, 0, FpuRoundMode.RNE)
testSqrtF64Exact(1.5*1.5, 1.5, 0, FpuRoundMode.RNE)
for(_ <- 0 until 10000) testSqrtF64()
for(_ <- 0 until runFactor*10000) testSqrtF64()
println("f64 sqrt done")
// testDivF64Exact(1.0, 8.0, 0.125, 0, FpuRoundMode.RNE)
@ -1346,35 +1357,35 @@ class FpuTest extends AnyFunSuite{
// testDivF64Exact(1.5, 2.0, 0.75, 0, FpuRoundMode.RNE)
// testDivF64Exact(1.875, 1.5, 1.25, 0, FpuRoundMode.RNE)
for(_ <- 0 until 10000) testDivF64()
for(_ <- 0 until runFactor*10000) testDivF64()
println("f64 div done")
for(_ <- 0 until 10000) testSgnjF64()
for(_ <- 0 until runFactor*10000) testSgnjF64()
println("f64 sgnj done")
for(_ <- 0 until 10000) testSgnjF32()
for(_ <- 0 until runFactor*10000) testSgnjF32()
println("f32 sgnj done")
//380000000001ffef 5fffffffffff9ff 8000000000100000
// testBinaryOpF64(mul,-5.877471754282472E-39, 8.814425663400984E-280, -5.180654E-318 ,1, FpuRoundMode.RMM,"mul")
// 5.877471754282472E-39 8.814425663400984E-280 -5.180654E-318 RMM
for(_ <- 0 until 10000) testCvtF64F32() // 1 did not equal 3 Flag missmatch dut=1 ref=3 testCvtF64F32Raw 1.1754942807573643E-38 1.17549435E-38 RMM
for(_ <- 0 until runFactor*10000) testCvtF64F32() // 1 did not equal 3 Flag missmatch dut=1 ref=3 testCvtF64F32Raw 1.1754942807573643E-38 1.17549435E-38 RMM
println("FCVT_D_S done")
for(_ <- 0 until 10000) testCvtF32F64()
for(_ <- 0 until runFactor*10000) testCvtF32F64()
println("FCVT_S_D done")
for(_ <- 0 until 10000) testF2iF64()
for(_ <- 0 until runFactor*10000) testF2iF64()
println("f64 f2i done")
for(_ <- 0 until 10000) testF2uiF64()
for(_ <- 0 until runFactor*10000) testF2uiF64()
println("f64 f2ui done")
for(_ <- 0 until 10000) testMinF64()
for(_ <- 0 until 10000) testMaxF64()
for(_ <- 0 until runFactor*10000) testMinF64()
for(_ <- 0 until runFactor*10000) testMaxF64()
println("f64 minMax done")
@ -1384,13 +1395,13 @@ class FpuTest extends AnyFunSuite{
println("f64 fma done") //TODO
for(_ <- 0 until 10000) testLeF64()
for(_ <- 0 until 10000) testLtF64()
for(_ <- 0 until 10000) testEqF64()
for(_ <- 0 until runFactor*10000) testLeF64()
for(_ <- 0 until runFactor*10000) testLtF64()
for(_ <- 0 until runFactor*10000) testEqF64()
println("f64 Cmp done")
for(_ <- 0 until 10000) testClassF64()
for(_ <- 0 until runFactor*10000) testClassF64()
println("f64 class done")
//
@ -1401,14 +1412,14 @@ class FpuTest extends AnyFunSuite{
for(_ <- 0 until 10000) testAddF64()
for(_ <- 0 until 10000) testSubF64()
for(_ <- 0 until runFactor*10000) testAddF64()
for(_ <- 0 until runFactor*10000) testSubF64()
println("f64 Add done")
// testI2f64Exact(0x7FFFFFF5, 0x7FFFFFF5, 0, true, FpuRoundMode.RNE)
for(_ <- 0 until 10000) testUI2f64()
for(_ <- 0 until 10000) testI2f64()
for(_ <- 0 until runFactor*10000) testUI2f64()
for(_ <- 0 until runFactor*10000) testI2f64()
println("f64 i2f done")
@ -1424,7 +1435,7 @@ class FpuTest extends AnyFunSuite{
testBinaryOpF64(mul,1.0, 2.0, 2.0,0 , FpuRoundMode.RNE,"mul")
testBinaryOpF64(mul,2.5, 2.0, 5.0,0 , FpuRoundMode.RNE,"mul")
for(_ <- 0 until 10000) testMulF64()
for(_ <- 0 until runFactor*10000) testMulF64()
println("f64 Mul done")
testTransferF64Raw(1.0)
@ -1439,26 +1450,27 @@ class FpuTest extends AnyFunSuite{
testTransferF32F64Raw(b2f(0xFFFF1234), true)
testTransferF64F32Raw(b2d(0xFFF123498765463l << 4), true)
for (_ <- 0 until 10000) testTransferF64()
for (_ <- 0 until runFactor*10000) testTransferF64()
println("f64 load/store/rf transfer done")
for (_ <- 0 until 10000) testTransferF64F32()
for (_ <- 0 until runFactor*10000) testTransferF64F32()
println("f64 -> f32 load/store/rf transfer done")
for (_ <- 0 until 10000) testTransferF32F64()
for (_ <- 0 until runFactor*10000) testTransferF32F64()
println("f32 -> f64 load/store/rf transfer done")
}
for(_ <- 0 until 10000) testTransferF32()
for(_ <- 0 until runFactor*10000) testTransferF32()
println("f32 load/store/rf transfer done")
for(_ <- 0 until 10000) testMulF32()
for(_ <- 0 until runFactor*10000) testMulF32()
println("Mul done")
for(_ <- 0 until 10000) testUI2f32()
for(_ <- 0 until 10000) testI2f32()
for(_ <- 0 until runFactor*10000) testUI2f32()
for(_ <- 0 until runFactor*10000) testI2f32()
println("i2f done")
@ -1470,8 +1482,8 @@ class FpuTest extends AnyFunSuite{
for(_ <- 0 until 10000) testF2uiF32()
for(_ <- 0 until 10000) testF2iF32()
for(_ <- 0 until runFactor*10000) testF2uiF32()
for(_ <- 0 until runFactor*10000) testF2iF32()
println("f2i done")
@ -1492,28 +1504,28 @@ class FpuTest extends AnyFunSuite{
testEqRaw(Float.PositiveInfinity,Float.PositiveInfinity,1, 0)
testEqRaw(0f, 0f,1, 0)
for(_ <- 0 until 10000) testLeF32()
for(_ <- 0 until 10000) testLtF32()
for(_ <- 0 until 10000) testEqF32()
for(_ <- 0 until runFactor*10000) testLeF32()
for(_ <- 0 until runFactor*10000) testLtF32()
for(_ <- 0 until runFactor*10000) testEqF32()
println("Cmp done")
for(_ <- 0 until 10000) testDiv()
for(_ <- 0 until runFactor*10000) testDiv()
println("f32 div done")
for(_ <- 0 until 10000) testSqrt()
for(_ <- 0 until runFactor*10000) testSqrt()
println("f32 sqrt done")
for(_ <- 0 until 10000) testSgnjF32()
for(_ <- 0 until runFactor*10000) testSgnjF32()
println("f32 sgnj done")
for(_ <- 0 until 10000) testClassF32()
for(_ <- 0 until runFactor*10000) testClassF32()
println("f32 class done")
for(_ <- 0 until 10000) testMinF32()
for(_ <- 0 until 10000) testMaxF32()
for(_ <- 0 until runFactor*10000) testMinF32()
for(_ <- 0 until runFactor*10000) testMaxF32()
println("minMax done")
@ -1537,8 +1549,8 @@ class FpuTest extends AnyFunSuite{
for(_ <- 0 until 10000) testAddF32()
for(_ <- 0 until 10000) testSubF32()
for(_ <- 0 until runFactor*10000) testAddF32()
for(_ <- 0 until runFactor*10000) testSubF32()
println("Add done")
@ -1555,7 +1567,7 @@ class FpuTest extends AnyFunSuite{
// dut.clockDomain.waitSampling(1000)
// simSuccess()
for(i <- 0 until 100000) fxxTests.randomPick()()
for(i <- 0 until runFactor*100000) fxxTests.randomPick()()
waitUntil(cpu.rspQueue.isEmpty)
}