CTRL-C in julia on windows 10

[fps]

Julia crashes quite often when interrupting loops with ctrl-c. This happens to me only in windows, linux is seemingly fine. The simplest NWE being this one:

julia> while true; end
WARNING: Force throwing a SIGINT

Please submit a bug report with steps to reproduce this fault, and any error messages that follow (in their entirety). Thanks.
Exception: EXCEPTION_ACCESS_VIOLATION at 0x7ffcd4ea0a65 -- jl_throw_in_ctx at C:/workdir/src\signals-win.c:130
in expression starting at REPL[1]:1
Allocations: 2998 (Pool: 2986; Big: 12); GC: 0

C:\Users\fschmidt\AppData\Local\Programs\Julia-1.9.1\bin>

But it also happens during more elaborate loops with plotting, optimizing a neural network and printing some terminal output. There's a ca. 50/50 chance of julia crashing. Here's one corresponding crash output:

fatal: error thrown and no exception handler available.
InterruptException()
sigatomic_end at .\c.jl:452 [inlined]
task_done_hook at .\task.jl:670
jfptr_task_done_hook_49669.clone_1 at C:\Users\fschmidt\AppData\Local\Programs\Julia-1.9.1\lib\julia\sys.dll (unknown line)
jl_apply at C:/workdir/src\julia.h:1879 [inlined]
jl_finish_task at C:/workdir/src\task.c:320
start_task at C:/workdir/src\task.c:1103

If you need more debugging info, please let me know..

julia> versioninfo()
Julia Version 1.9.1
Commit 147bdf428c (2023-06-07 08:27 UTC)
Platform Info:
  OS: Windows (x86_64-w64-mingw32)
  CPU: 4 × Intel(R) Core(TM) i5-7600 CPU @ 3.50GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-14.0.6 (ORCJIT, skylake)
  Threads: 2 on 4 virtual cores
Environment:
  JULIA_NUM_THREADS = 2

[ViralBShah]

Does it happen on older releases as well?

[fps]

Hi! Yeah, I don't actually know if there ever was a version where ctrl-c worked reliably without crashes from time to time. I started  with julia 0.9 iirc ;) I'm using julia in the windows cmd prompt exclusively. (edit: removed quote from original email)

[PallHaraldsson]

There was no Julia 0.9, only 0.7/1.0, but I believe CTRL-C worked at some point, better, on some 1.0+, then got worse. My recollection is it started not working, i.e. being ignored, I don't recall a segfault implicated.

[fps]

Here's 0.7 with the minimal example:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Julia-0.7.0\bin>

No segfault, but it exits... Here's a little more involved example:

julia> while true; x = randn(10,10); y = x * x; end
ERROR:
Stacktrace:
 [1] Type at .\boot.jl:396 [inlined]
 [2] Type at .\boot.jl:409 [inlined]
 [3] randn(::Random.MersenneTwister, ::Type{Float64}, ::Int64, ::Int64) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\Random\src\normal.jl:182
 [4] randn at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\Random\src\normal.jl:188 [inlined]
 [5] top-level scope at .\REPL[1]:1 [inlined]
 [6] top-level scope at .\none:0SYSTEM: show(lasterr) caused an error
InterruptException()

Stacktrace:
 [1] Type at .\boot.jl:396 [inlined]
 [2] Type at .\boot.jl:409 [inlined]
 [3] randn(::Random.MersenneTwister, ::Type{Float64}, ::Int64, ::Int64) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\Random\src\normal.jl:182
 [4] randn at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\Random\src\normal.jl:188 [inlined]
 [5] top-level scope at .\REPL[1]:1 [inlined]
 [6] top-level scope at .\none:0
 [7] eval(::Module, ::Any) at .\boot.jl:319
 [8] eval_user_input(::Any, ::REPL.REPLBackend) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:85
 [9] macro expansion at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:117 [inlined]
 [10] (::getfield(REPL, Symbol("##28#29")){REPL.REPLBackend})() at .\task.jl:262ERROR: InterruptException:
Stacktrace:
 [1] sigatomic_end at .\c.jl:425 [inlined]
 [2] print_response(::IO, ::Any, ::Any, ::Bool, ::Bool, ::Any) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:176
 [3] print_response(::REPL.AbstractREPL, ::Any, ::Any, ::Bool, ::Bool) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:139
 [4] (::getfield(REPL, Symbol("#do_respond#40")){Bool,getfield(REPL, Symbol("##50#59")){REPL.LineEditREPL,REPL.REPLHistoryProvider},REPL.LineEditREPL,REPL.LineEdit.Prompt})(::Any, ::Any, ::Any) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:708
 [5] #invokelatest#1 at .\essentials.jl:691 [inlined]
 [6] invokelatest at .\essentials.jl:690 [inlined]
 [7] run_interface(::REPL.Terminals.TextTerminal, ::REPL.LineEdit.ModalInterface, ::REPL.LineEdit.MIState) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\LineEdit.jl:2261
 [8] run_frontend(::REPL.LineEditREPL, ::REPL.REPLBackendRef) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:1029
 [9] run_repl(::REPL.AbstractREPL, ::Any) at C:\cygwin\home\Administrator\buildbot\worker\package_win64\build\usr\share\julia\stdlib\v0.7\REPL\src\REPL.jl:191
 [10] (::getfield(Base, Symbol("##831#833")){Bool,Bool,Bool,Bool})(::Module) at .\logging.jl:311
 [11] #invokelatest#1 at .\essentials.jl:691 [inlined]
 [12] invokelatest at .\essentials.jl:690 [inlined]
 [13] macro expansion at .\logging.jl:308 [inlined]
 [14] run_main_repl(::Bool, ::Bool, ::Bool, ::Bool, ::Bool) at .\client.jl:340
 [15] exec_options(::Base.JLOptions) at .\client.jl:252
 [16] _start() at .\client.jl:432

C:\Users\fschmidt\AppData\Local\Julia-0.7.0\bin>

Also no segfault, but exits sometimes instead of returning to the prompt. This exiting is somewhat hard to reproduce though. So I suspect a race condition somewhere..

[fps]

1.0.5:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Julia-1.0.5\bin>

I couldn't reproduce the second exit here in a couple of tries..

[fps]

1.1.1:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Julia-1.1.1\bin>

Same for the second example

[fps]

1.2.0:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Julia-1.2.0\bin>

Same for the second example

[fps]

1.3.0:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Julia-1.3.0\bin>

Same for the second example

[fps]

1.4.2:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Programs\Julia\Julia-1.4.2\bin>

Same for the second example

[fps]

Ok, coming from the other side :)

1.9.0:

julia> while true; end
WARNING: Force throwing a SIGINT

Please submit a bug report with steps to reproduce this fault, and any error messages that follow (in their entirety). Thanks.
Exception: EXCEPTION_ACCESS_VIOLATION at 0x7ffa094b09c5 -- jl_throw_in_ctx at C:/workdir/src\signals-win.c:130
in expression starting at REPL[1]:1
Allocations: 2997 (Pool: 2984; Big: 13); GC: 0

[fps]

1.8.5:

julia> while true; end
WARNING: Force throwing a SIGINT

C:\Users\fschmidt\AppData\Local\Programs\Julia-1.8.5\bin>

So the segfault was introduced somewhere between 1.8.5 and 1.9.0

[fps]

1.9.0-rc1 has it, too:

julia> while true; end
WARNING: Force throwing a SIGINT

Please submit a bug report with steps to reproduce this fault, and any error messages that follow (in their entirety). Thanks.
Exception: EXCEPTION_ACCESS_VIOLATION at 0x7ffa094beca5 -- jl_throw_in_ctx at C:/workdir/src\signals-win.c:130
in expression starting at REPL[1]:1
Allocations: 2995 (Pool: 2984; Big: 11); GC: 0

C:\Users\fschmidt\AppData\Local\Programs\Julia-1.9.0-rc1\bin>

[ViralBShah]

Related: #49541