# Use\_intel\_libm = 1

**URL:** <https://discourse.julialang.org/t/use-intel-libm-1/16817>\
**Category:** General Usage\
**Tags:** compilation\
**Created:** [October 26, 2018, 9:51am UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817 "2018-10-26T09:51:58Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![carstenbauer](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/carstenbauer/32/4981_2.png) [@carstenbauer](https://discourse.julialang.org/u/carstenbauer)\
**Post date:** [October 26, 2018, 9:51am UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/1 "2018-10-26T09:51:58Z")

</div>

Is anyone successfully using Intel Math Library (`USE_INTEL_LIBM = 1` in Make.user) in a self-compiled Julia 1.0.1?

Pre 1.0 I always used `USE_INTEL_LIBM` and `USE_INTEL_MKL` to replace all OpenBLAS by Intel pendants without any issues. However, for 1.0.1 things fail

```julia
  math: Error During Test at /projects/sborowsk/julia_build/julia/test/math.jl:841
  Test threw exception
  Expression: ≈(sinh(x), sinh(big(x)), rtol=eps(T))
  ccall: could not find function __ldexp_expf in library libimf

```

which seems to originate from `base/special/hyperbolic.jl` where in lines

```julia
 _ldexp_exp(x::Float64, i::T) where T = ccall(("__ldexp_exp", libm), Float64, (Float64, T), x, i)
 _ldexp_exp(x::Float32, i::T) where T = ccall(("__ldexp_expf",libm), Float32, (Float32, T), x, i)

```

libopenlibm specific functions are called which can’t be found in libimf.

If I replace those lines by

```julia
_ldexp_exp(x::Float64, i::T) where T = ccall(("ldexp", libm), Float64, (Float64, T), x, i)
_ldexp_exp(x::Float32, i::T) where T = ccall(("ldexpf",libm), Float32, (Float32, T), x, i)

```

Things work, but the test fails:

```julia
Error in testset math:
  Test Failed at /projects/sborowsk/julia_build/julia/test/math.jl:841
    Expression: ≈(sinh(x), sinh(big(x)), rtol=eps(T))
     Evaluated: 44.361416f0 ≈ 1.701399268705924328372129400192948964842537512956506377907196085335400360725516e+38 (rtol=1.1920929f-7)

```

Any related experiences anyone?

(I will write up a proper bug report if needed.)

---

<div class="post-metadata">

**Author:** ![carstenbauer](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/carstenbauer/32/4981_2.png) [@carstenbauer](https://discourse.julialang.org/u/carstenbauer)\
**Post date:** [October 26, 2018, 12:01pm UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/2 "2018-10-26T12:01:33Z")

</div>

Maybe as a motivation for why intel’s math library:

```julia
julia> @btime sinh.($(rand(100_000))); # LIBM: libopenlibm
  3.259 ms (2 allocations: 781.33 KiB)

```

```julia
julia> @btime sinh.($(rand(100_000))); # LIBM: libimf
  1.556 ms (2 allocations: 781.33 KiB)

```

---

<div class="post-metadata">

**Author:** ![shipengcheng1230](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/shipengcheng1230/32/4089_2.png) [@shipengcheng1230](https://discourse.julialang.org/u/shipengcheng1230)\
**Post date:** [December 7, 2018, 2:45pm UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/3 "2018-12-07T14:45:18Z")

</div>

Maybe a little bit irrelevant but I am on v1.0.2 and successfully compile it with intel-libm. In case anyone hits here, if it couldn’t find `libimf`, you may need to install [psxe\_runtime](https://software.intel.com/en-us/articles/installing-intel-parallel-studio-xe-runtime-2019-using-yum-repository):

```julia
export LD_LIBRARY_PATH=/opt/intel/psxe_runtime_2019.1.144/linux/compiler/lib/intel64_lin/:$LD_LIBRARY_PATH

```

And if you using `JLD` and get complaint about `version`CXXABI\_1.3.9’ not found`,

```julia
export LD_LIBRARY_PATH=/opt/anaconda3/lib/:$LD_LIBRARY_PATH

```

Or if you don’t want to mess up with the `$LD_LIBRARY_PATH`, you could edit `Make.user` similar like me:

```julia
USE_SYSTEM_XXX=1
LDFLAGS=-Wl,-rpath,/opt/intel/psxe_runtime/linux/compiler/lib/intel64_lin/
LDFLAGS+=-Wl,-rpath,/opt/gcc8/lib64/
LDFLAGS+=-Wl,-rpath,/opt/intel/compilers_and_libraries_2019.1.144/linux/mkl/lib/intel64_lin/
LDFLAGS+=-L/opt/gcc8/lib64/ -L/opt/intel/psxe_runtime/linux/compiler/lib/intel64_lin/
LDFLAGS+=-L/opt/intel/compilers_and_libraries_2019.1.144/linux/mkl/lib/intel64_lin/

```

---

<div class="post-metadata">

**Author:** ![carstenbauer](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/carstenbauer/32/4981_2.png) [@carstenbauer](https://discourse.julialang.org/u/carstenbauer)\
**Post date:** [December 7, 2018, 2:52pm UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/4 "2018-12-07T14:52:36Z")

</div>

Not irrelevant at all!

I also managed to use Intel’s libm by now. The relevant line mentioned above has been changed between 1.0.0 and 1.0.2 although unrelated to this issue I believe.

---

<div class="post-metadata">

**Author:** ![shipengcheng1230](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/shipengcheng1230/32/4089_2.png) [@shipengcheng1230](https://discourse.julialang.org/u/shipengcheng1230)\
**Post date:** [December 7, 2018, 2:54pm UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/5 "2018-12-07T14:54:11Z")

</div>

I also notice that the performance for those optimized functions like `sinh` are great!

---

<div class="post-metadata">

**Author:** ![einzigsue](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/einzigsue/32/7023_2.png) [@einzigsue](https://discourse.julialang.org/u/einzigsue)\
**Post date:** [February 8, 2019, 4:03am UTC](https://discourse.julialang.org/t/use-intel-libm-1/16817/6 "2019-02-08T04:03:09Z")

</div>

I encountered the same test error when updating to 1.0 but ignoring it all the time. I think it can be avoidable if and only if linked against openlibm.

These two libraries, \_\_ldexp\_exp and \_\_ldexp\_expf, do not exist in library libimf, they are openlibm specific, see openlibm/src/k\_exp.c and openlibm/src/k\_expf.c for details, and they have no equivalents in libimf.so.

The good thing is, in base/special/hyperbolic.jl it only calls these openlibm specific functions when the input of sinh/cosh is between Base.Math.H\_LARGE\_X(T) [inclusive] and Base.Math.H\_OVERFLOW\_X(T). That is [88.72283f0, 89.415985f0] for T=Float32 and [709.7822265633563,710.475860073944] for T=Float64.

I guess if you have to work within these two ranges using the corresponding type, you then have to implement something in, for example, myhyperbolic.jl to call directly from libimf to overwrite the ones in defined in base/special/hyperbolic.jl.