Error on large sparse matrices

[ranjanan]

Hey, if I do this:

a  = sprand(10^6, 10^6, 1e-6)
a = a + a' #So it's invertible 
a = a + speye(10^6) #So it's invertible
b = rand(10^6)
c = Array(Float64, 10^6)
@time CUSOLVER.csrlsvlu!(a, b, c, 1e-6, zero(Cint), 'O')

I get the following error:

WARNING: CUSOLVER error triggered from:

 in statuscheck at /home/ubuntu/.julia/v0.4/CUSOLVER/src/CUSOLVER.jl:44
 in csrlsvlu! at /home/ubuntu/.julia/v0.4/CUSOLVER/src/sparse.jl:27
ERROR: MethodError: `convert` has no method matching convert(::Type{AbstractString}, ::UInt32)
This may have arisen from a call to the constructor AbstractString(...),
since type constructors fall back to convert methods.
Closest candidates are:
  call{T}(::Type{T}, ::Any)
  convert{T<:AbstractString,S<:Union{Char,Int32,UInt32}}(::Type{T<:AbstractString}, ::AbstractArray{S<:Union{Char,Int32,UInt32},1})
  convert{S<:AbstractString}(::Type{S<:AbstractString}, ::Base.UTF8proc.GraphemeIterator{S<:AbstractString})
  ...
 in statuscheck at /home/ubuntu/.julia/v0.4/CUSOLVER/src/CUSOLVER.jl:46
 in csrlsvlu! at /home/ubuntu/.julia/v0.4/CUSOLVER/src/sparse.jl:27

This works for smaller sparse matrices. I got the error on the Cholesky kernel too. I didn't try the qr one yet. What do you think is happening?
My Julia setup is this:

Julia Version 0.4.2
Commit bb73f34 (2015-12-06 21:47 UTC)
Platform Info:
  System: Linux (x86_64-unknown-linux-gnu)
  CPU: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz
  WORD_SIZE: 64
  BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Sandybridge)
  LAPACK: libopenblas64_
  LIBM: libopenlibm
  LLVM: libLLVM-3.3

[kshyatt]

IIRC, the lu and chol solvers are host path only.

What's happening in that specific error is that my error handler is broken .

I suspect it's running out of memory.

[ranjanan]

Yes, I guess you're right. Inverses of those sparse matrices could be dense, which is why it's running out of memory, I suppose.