juliasimd/triangularsolve.jl

rdiv!(::AbstractMatrix, ::UpperTriangular) and ldiv!(::LowerTriangular, ::AbstractMatrix)

triangularsolve.jl's Issues

Stuck on precompiling after upgrading to 1.8

Hello! After upgrading to Julia 1.8, one of my machines is refusing to precompile TriangularSolve. ] add TriangularSolve (or any other package that has this a dependency) results in the REPL getting stuck at

Precompiling project...
  Progress [>                                        ]  0/1
  ◑ TriangularSolve

I've let it sit for an hour, and nothing changes. On 1.7, I do not experience this issue. I've tried doing a fresh install of julia, to no avail. Interrupting with ctrl+c gives the following

ERROR: The following 1 direct dependency failed to precompile:

TriangularSolve [d5829a12-d9aa-46ab-831f-fb7c9ab06edf]

Failed to precompile TriangularSolve [d5829a12-d9aa-46ab-831f-fb7c9ab06edf] to /home/philip/.julia/compiled/v1.8/TriangularSolve/jl_uahkjT.

signal (2): Interrupt
in expression starting at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:535
rdiv_block_N! at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:380
rdiv_block_MandN! at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:416
div_dispatch! at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:312
rdiv! at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:323
unknown function (ip: 0x7f3459fe5b71)
_jl_invoke at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2367 [inlined]
ijl_apply_generic at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2549
macro expansion at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:543 [inlined]
macro expansion at /home/philip/.julia/packages/SnoopPrecompile/UWvXF/src/SnoopPrecompile.jl:51 [inlined]
macro expansion at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:540 [inlined]
macro expansion at /home/philip/.julia/packages/SnoopPrecompile/UWvXF/src/SnoopPrecompile.jl:107 [inlined]
top-level scope at /home/philip/.julia/packages/TriangularSolve/E948Y/src/TriangularSolve.jl:536
jl_toplevel_eval_flex at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:897
jl_eval_module_expr at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:203 [inlined]
jl_toplevel_eval_flex at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:709
jl_toplevel_eval_flex at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:850
ijl_toplevel_eval_in at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:965
eval at ./boot.jl:368 [inlined]
include_string at ./loading.jl:1428
_jl_invoke at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2367 [inlined]
ijl_apply_generic at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2549
_include at ./loading.jl:1488
include at ./Base.jl:419 [inlined]
include_package_for_output at ./loading.jl:1554
jfptr_include_package_for_output_37424.clone_1 at /home/philip/packages/julias/julia-1.8/lib/julia/sys.so (unknown line)
_jl_invoke at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2367 [inlined]
ijl_apply_generic at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2549
jl_apply at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/julia.h:1838 [inlined]
do_call at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/interpreter.c:126
eval_value at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/interpreter.c:215
eval_stmt_value at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/interpreter.c:166 [inlined]
eval_body at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/interpreter.c:612
jl_interpret_toplevel_thunk at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/interpreter.c:750
jl_toplevel_eval_flex at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:906
jl_toplevel_eval_flex at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:850
ijl_toplevel_eval_in at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/toplevel.c:965
eval at ./boot.jl:368 [inlined]
include_string at ./loading.jl:1428
include_string at ./loading.jl:1438
_jl_invoke at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2367 [inlined]
ijl_apply_generic at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2549
exec_options at ./client.jl:301
_start at ./client.jl:522
jfptr__start_30463.clone_1 at /home/philip/packages/julias/julia-1.8/lib/julia/sys.so (unknown line)
_jl_invoke at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2367 [inlined]
ijl_apply_generic at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/gf.c:2549
jl_apply at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/julia.h:1838 [inlined]
true_main at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/jlapi.c:575
jl_repl_entrypoint at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/src/jlapi.c:719
main at /cache/build/default-amdci4-3/julialang/julia-release-1-dot-8/cli/loader_exe.c:59
__libc_start_main at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
unknown function (ip: 0x401098)
unknown function (ip: (nil))
Allocations: 26634325 (Pool: 26620513; Big: 13812); GC: 27

Curiously, only one of my machines experiences this issue. The other could upgrade to 1.8 with no issues. Both are Ubuntu 20.04 VMs running on win10 hosts. versioninfo() gives

julia> versioninfo()
Julia Version 1.8.0
Commit 5544a0fab76 (2022-08-17 13:38 UTC)
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: 6 × AMD Ryzen 5 3600 6-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-13.0.1 (ORCJIT, znver2)
  Threads: 6 on 6 virtual cores
Environment:
  JULIA_EDITOR = code
  JULIA_NUM_THREADS = 6

In addition, I have the same issue with v0.1.12, although with a slightly different stacktrace upon ctrl+c.

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

Segfault on julia 1.6.5

Hi, I am getting a segfault on Julia 1.6.5 when using TriangularSolve:
https://gist.github.com/j-fu/03d66b7a48a592516c14001f3f132113

This doesn't happen neither with 1.6.4, nor with 1.7.0.

`using TriangularSolve` fails on Julia 1.8.0-beta1

On this system

julia> versioninfo()
Julia Version 1.8.0-beta1
Commit 7b711ce699 (2022-02-23 15:09 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
  CPU: 32 × AMD Ryzen Threadripper 2950X 16-Core Processor
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-13.0.1 (ORCJIT, znver1)
  Threads: 1 on 32 virtual cores
Environment:
  JULIA_EDITOR = code
  JULIA_NUM_THREADS = 

I get

julia> using TriangularSolve
[ Info: Precompiling TriangularSolve [d5829a12-d9aa-46ab-831f-fb7c9ab06edf]
ERROR: LoadError: DivideError: integer division error
Stacktrace:
  [1] vdiv
    @ ~/.julia/packages/VectorizationBase/yqq6V/src/llvm_intrin/binary_ops.jl:65 [inlined]
  [2] vdiv
    @ ~/.julia/packages/VectorizationBase/yqq6V/src/static.jl:56 [inlined]
  [3] vcld
    @ ~/.julia/packages/VectorizationBase/yqq6V/src/llvm_intrin/binary_ops.jl:79 [inlined]
  [4] rdiv_block_MandN!(spc::LayoutPointers.StridedPointer{Float64, 2, 1, 0, (1, 2), Tuple{Static.StaticInt{8}, Int64}, Tuple{Static.StaticInt{0}, Static.StaticInt{0}}}, spa::LayoutPointers.StridedPointer{Float64, 2, 1, 0, (1, 2), Tuple{Static.StaticInt{8}, Int64}, Tuple{Static.StaticInt{0}, Static.StaticInt{0}}}, spu::LayoutPointers.StridedPointer{Float64, 2, 1, 0, (1, 2), Tuple{Static.StaticInt{8}, Int64}, Tuple{Static.StaticInt{0}, Static.StaticInt{0}}}, M::Int64, N::Int64, #unused#::Val{false}, #unused#::Static.StaticInt{1})
    @ TriangularSolve ~/git/TriangularSolve.jl/src/TriangularSolve.jl:411
  [5] div_dispatch!(C::Matrix{Float64}, A::Matrix{Float64}, U::Matrix{Float64}, #unused#::Val{false}, #unused#::Val{false})
    @ TriangularSolve ~/git/TriangularSolve.jl/src/TriangularSolve.jl:312
  [6] rdiv!
    @ ~/git/TriangularSolve.jl/src/TriangularSolve.jl:323 [inlined]
  [7] top-level scope
    @ ~/git/TriangularSolve.jl/src/TriangularSolve.jl:542
  [8] include
    @ ./Base.jl:422 [inlined]
  [9] include_package_for_output(pkg::Base.PkgId, input::String, depot_path::Vector{String}, dl_load_path::Vector{String}, load_path::Vector{String}, concrete_deps::Vector{Pair{Base.PkgId, UInt64}}, source::Nothing)
    @ Base ./loading.jl:1399
 [10] top-level scope
    @ stdin:1

The problem comes in this section

@static if VERSION >= v"1.8.0-beta1"
  let
    while true
      A = rand(1, 1)
      B = rand(1, 1)
      res = similar(A)
      rdiv!(res, A, UpperTriangular(B))
      rdiv!(res, A, UnitUpperTriangular(B))
      rdiv!(res, A, UpperTriangular(B), Val(false))   # <----- THIS LINE
      rdiv!(res, A, UnitUpperTriangular(B), Val(false))

      __init__()
      ldiv!(res, LowerTriangular(B), A)
      ldiv!(res, UnitLowerTriangular(B), A)
      ldiv!(res, LowerTriangular(B), A, Val(false))
      ldiv!(res, UnitLowerTriangular(B), A, Val(false))
      break
    end
  end
end

I've tried commenting this out, loading the package, and then running the section manually. In this case I get a segfault.

stridedpointer_preserve no longer in VectorizationBase

In the latest codebase,

TriangularSolve.jl/src/TriangularSolve.jl

Line 4 in 50b01ec

using VectorizationBase: vfnmadd_fast, AbstractStridedPointer, AbstractMask, stridedpointer_preserve, zero_offsets, gesp, StridedPointer

the module imports stridedpointer_preserve from VectorizationBase, which does not seem to include this function anymore. Instead, it looks like it should be imported either directly from LayoutPointers or from LoopVectorization

I tried the latter approach and got an error

[ Info: Precompiling TriangularSolve [d5829a12-d9aa-46ab-831f-fb7c9ab06edf]
ERROR: LoadError: MethodError: no method matching (LayoutPointers.StridedPointer{Float64, 2, 1, 0, (1, 2), X, O} where {X, O})(::Ptr{Float64}, ::Tuple{Static.StaticInt{8}, Static.StaticInt{96}}, ::Tuple{Static.StaticInt{0}, Static.StaticInt{0}})

`ldiv!` invokes undefined behavior when calling `m_thread_block_size` with `nthreads == static(0)`.

TriangularSolve.jl/src/TriangularSolve.jl

Line 264 in e425ff5

div_dispatch!(transpose(C), transpose(A), transpose(parent(U)), static(0), Val(true))

and many more places ends up calling m_thread_block_size with nthreads == static(0).

The compiler can then prove that

TriangularSolve.jl/src/TriangularSolve.jl

Line 335 in e425ff5

min(M, VectorizationBase.vcld(M, nb*W)*W)

is called as VectorizationBase.vcld(M, 0) (which ends up calling Base.sdiv_int(x, 0) which is undefined behavior https://llvm.org/docs/LangRef.html#sdiv-instruction) and returns llvm's poison value.

Maybe this should just have been static(1)?

Segfault when using v0.1.3

Here is a representative stack trace:
https://buildkite.com/clima/climacore-ci/builds/183#73896f51-3ec9-4099-a226-3217264ed813/96-100

Culprit looks to be a mis-sized buffer?

TriangularSolve.jl/src/TriangularSolve.jl

Lines 288 to 290 in 08ca19b

	L > length(buff) && resize!(buff, L)
	ptr = Base.unsafe_convert(Ptr{T}, buff)
	StridedPointer{T,2,1,0,(1,2)}(ptr, (VectorizationBase.static_sizeof(T), RSUF), (StaticInt(0),StaticInt(0)))

This also might be resolved with the latest release that adjusted the compat bounds with VectorizationBase

Failure on latest version (0.1.4)

I received the following stacktrace:

ERROR: LoadError: MethodError: no method matching zero_offsets(::LayoutPointers.StridedPointer{Float64, 2, 2, -1, (2, 1), Tuple{Int64, Static.StaticInt{8}}, Tuple{Static.StaticInt{1}, Static.StaticInt{1}}})
Closest candidates are:
  zero_offsets(::Static.StaticInt{N}) where N at /Users/simon/.julia/packages/VectorizationBase/pTvQj/src/static.jl:126
  zero_offsets(::VectorizationBase.StridedPointer{T, N, C, B, R, X, O} where {X, O}) where {T, N, C, B, R} at /Users/simon/.julia/packages/VectorizationBase/pTvQj/src/strided_pointers/stridedpointers.jl:115
  zero_offsets(::VectorizationBase.FastRange{T, Static.StaticInt{0}, S, O} where {S, O}) where T<:Integer at /Users/simon/.julia/packages/VectorizationBase/pTvQj/src/strided_pointers/stridedpointers.jl:466
  ...
Stacktrace:
  [1] div_dispatch!
    @ ~/.julia/packages/TriangularSolve/pVj7v/src/TriangularSolve.jl:300 [inlined]
  [2] ldiv! (repeats 2 times)
    @ ~/.julia/packages/TriangularSolve/pVj7v/src/TriangularSolve.jl:339 [inlined]
  [3] reckernel!(A::StrideArraysCore.PtrArray{Tuple{Int64, Int64}, (true, true), Float64, 2, 1, 0, (1, 2), Tuple{Static.StaticInt{8}, Int64}, Tuple{Static.StaticInt{1}, Static.StaticInt{1}}}, pivot::Val{true}, m::Int64, n::Int64, ipiv::StrideArraysCore.PtrArray{Tuple{Int64}, (true,), Int64, 1, 1, 0, (1,), Tuple{Static.StaticInt{8}}, Tuple{Static.StaticInt{1}}}, info::Int64, blocksize::Int64, thread::Bool)
    @ RecursiveFactorization ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:165
  [4] reckernel!(A::StrideArraysCore.PtrArray{Tuple{Int64, Int64}, (true, true), Float64, 2, 1, 0, (1, 2), Tuple{Static.StaticInt{8}, Int64}, Tuple{Static.StaticInt{1}, Static.StaticInt{1}}}, pivot::Val{true}, m::Int64, n::Int64, ipiv::StrideArraysCore.PtrArray{Tuple{Int64}, (true,), Int64, 1, 1, 0, (1,), Tuple{Static.StaticInt{8}}, Tuple{Static.StaticInt{1}}}, info::Int64, blocksize::Int64, thread::Bool)
    @ RecursiveFactorization ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:159
  [5] recurse!
    @ ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:76 [inlined]
  [6] lu!(A::Matrix{Float64}, ipiv::Vector{Int64}, pivot::Val{true}; check::Bool, blocksize::Int64, threshold::Int64)
    @ RecursiveFactorization ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:62
  [7] #lu!#2
    @ ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:27 [inlined]
  [8] lu! (repeats 2 times)
    @ ~/.julia/packages/RecursiveFactorization/XJaVF/src/lu.jl:22 [inlined]

I didn't have time to get a reproducible example, but can try to put one together tomorrow if that would be helpful.