Failure to precompile Plots.jl in Linux

Hi, I have trouble to precompile Plots.jl in Linux. I looked through a couple of similar questions but my problem still can not be solved. Any help will be appreciated. Below is the error msg and environment info:

julia> using Plots
[ Info: Precompiling Plots [91a5bcdd-55d7-5caf-9e0b-520d859cae80]
Invalid instruction at 0x7f9bc363e9e3: 0xc4, 0xc1, 0x7b, 0x92, 0xce, 0xc5, 0xf9, 0x28, 0xc4, 0x62, 0xf1, 0xff, 0x09, 0x10, 0xc5
[742125] signal (4.2): Illegal instruction
in expression starting at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/share/julia/packages/Plots/sxUvK/src/components.jl:103
_linspace at ./twiceprecision.jl:0
range_start_stop_length at ./twiceprecision.jl:668
_range at ./range.jl:161 [inlined]
#range#70 at ./range.jl:142 [inlined]
range at ./range.jl:142 [inlined]
partialcircle at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/share/julia/packages/Plots/sxUvK/src/components.jl:46 [inlined]
#makeshape#155 at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/share/julia/packages/Plots/sxUvK/src/components.jl:75 [inlined]
makeshape at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/share/julia/packages/Plots/sxUvK/src/components.jl:75
unknown function (ip: 0x7f9bc452c71a)
jl_apply at /usr/local/src/conda/julia-1.9.3/src/julia.h:1880 [inlined]
do_call at /usr/local/src/conda/julia-1.9.3/src/interpreter.c:126
eval_value at /usr/local/src/conda/julia-1.9.3/src/interpreter.c:226
eval_stmt_value at /usr/local/src/conda/julia-1.9.3/src/interpreter.c:177 [inlined]
eval_body at /usr/local/src/conda/julia-1.9.3/src/interpreter.c:606
jl_interpret_toplevel_thunk at /usr/local/src/conda/julia-1.9.3/src/interpreter.c:762
jl_toplevel_eval_flex at /usr/local/src/conda/julia-1.9.3/src/toplevel.c:912
jl_toplevel_eval_flex at /usr/local/src/conda/julia-1.9.3/src/toplevel.c:856
ijl_toplevel_eval_in at /usr/local/src/conda/julia-1.9.3/src/toplevel.c:971
eval at ./boot.jl:370 [inlined]
include_string at ./loading.jl:1903
include_string at ./loading.jl:1913 [inlined]
exec_options at ./client.jl:305
_start at ./client.jl:522
jfptr__start_32224.clone_1 at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/lib/julia/sys.so (unknown line)
jl_apply at /usr/local/src/conda/julia-1.9.3/src/julia.h:1880 [inlined]
true_main at /usr/local/src/conda/julia-1.9.3/src/jlapi.c:573
jl_repl_entrypoint at /usr/local/src/conda/julia-1.9.3/src/jlapi.c:717
main at /home/conda/feedstock_root/build_artifacts/julia_1693156704872/work/cli/loader_exe.c:59
__libc_start_main at /lib64/libc.so.6 (unknown line)
_start at /nas/longleaf/home/fengyu/.conda/envs/julia_env1/bin/julia (unknown line)
Allocations: 2791919 (Pool: 2790224; Big: 1695); GC: 4
ERROR: Failed to precompile Plots [91a5bcdd-55d7-5caf-9e0b-520d859cae80] to "/nas/longleaf/home/fengyu/.conda/envs/julia_env1/shar
e/julia/compiled/v1.9/Plots/jl_xp6ryK".
Stacktrace:
 [1] error(s::String)
   @ Base ./error.jl:35
 [2] compilecache(pkg::Base.PkgId, path::String, internal_stderr::IO, internal_stdout::IO, keep_loaded_modules::Bool)
   @ Base ./loading.jl:2300
 [3] compilecache
   @ ./loading.jl:2167 [inlined]
 [4] _require(pkg::Base.PkgId, env::String)
   @ Base ./loading.jl:1805
 [5] _require_prelocked(uuidkey::Base.PkgId, env::String)
   @ Base ./loading.jl:1660
 [6] macro expansion
   @ ./loading.jl:1648 [inlined]
 [7] macro expansion
   @ ./lock.jl:267 [inlined]
 [8] require(into::Module, mod::Symbol)
   @ Base ./loading.jl:1611

Environment info:
Julia Version 1.9.2
Plots v1.39.0
Commit e4ee485e90 (2023-07-05 09:39 UTC)
Platform Info:
OS: Linux (x86_64-conda-linux-gnu)
CPU: 14.0.6 (ORCJIT, znver3)
Threads: 1 on 64 virtual cores

1 Like

Could you please try downloading Julia from Download Julia or use juliaup via curl -fsSL https://install.julialang.org | sh?

While I have tried my best to make the julia in conda work, it is terribly complicated. As such the first thing to try is to download Julia from official sources.

3 Likes

Thanks for your quick reply! Do you know any way to install Julia without sudo permission? I am running codes on a cluster without sudo permission. That’s why I installed julia in conda

1 Like

Neither of the above methods require sudo. The first is just a tarball than you can untar. The second will install julia in your home folder.

2 Likes

Thank you! Now the issue is solved. Time to see farewell to conda.

1 Like