[BUG] Plot caused "stack smashing"

[anandijain]

This plot shouldn't work, but I'm not quite sure why it causes a SIGABRT, couldn't find an open issue with "stack smashing".

cs is a DataFrame, with the :symbol column of type String, and the :x1 column of type Int.

Details

julia> plot(cs.symbol[1:50], cs.x1[1:50], yscale=:log10, label=cs.symbol[1:50])
*** stack smashing detected ***: <unknown> terminated

signal (6): Aborted
in expression starting at REPL[17]:0
gsignal at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
abort at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
unknown function (ip: 0x7fa577016896)
unknown function (ip: 0x7fa5770c1cd0)
__stack_chk_fail at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
process_glyphs at /home//.julia/packages/GR/cRdXQ/src/../deps/gr/lib/libGR.so (unknown line)
unknown function (ip: 0x5600000048)
Allocations: 127171138 (Pool: 127073353; Big: 97785); GC: 86
fish: “/home//Downloads/Appli…” terminated by signal SIGABRT (Abort)

Backends

Backend	yes	no	untested
gr (default)	X

Versions

Plots version: 1.3.3
Julia Version 1.4.0
Commit b8e9a9ecc6 (2020-03-21 16:36 UTC)
Platform Info:
OS: Linux (x86_64-pc-linux-gnu)
CPU: Intel(R) Core(TM) i5-6400 CPU @ 2.70GHz
WORD_SIZE: 64
LIBM: libopenlibm
LLVM: libLLVM-8.0.1 (ORCJIT, skylake)

[BeastyBlacksmith]

Can you give some DataFrame data, that reproduces this bug?

[BeastyBlacksmith]

Also which version of DataFrames is this?
I can't reproduce on julia 1.4.2, Plots 1.3.5 and DataFrames 0.21.1

df = DataFrame([(a = i, b = Symbol(i)) for i in 1:10])
plot(df.b[1:3], df.a[1:3], yscale = :log10, label=df.b[1:3])
ERROR: Cannot convert Symbol to series data for plotting
Stacktrace:

[Seelengrab]

@BeastyBlacksmith , your code is not the original example - this crashes julia 1.4.2 and 1.5.0-dev:

julia> versioninfo()
Julia Version 1.5.0-DEV.876
Commit 0c388fc340* (2020-05-07 20:12 UTC)
Platform Info:
  OS: Linux (x86_64-linux-gnu)
  CPU: Intel(R) Core(TM) i7-6600U CPU @ 2.60GHz
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-9.0.1 (ORCJIT, skylake)
Environment:
  JULIA_NUM_THREADS = 4

(GRsmashing) pkg> st
Status `/c/Users/valentin/Documents/Projects/GRsmashing/Project.toml`
  [a93c6f00] DataFrames v0.21.1
  [91a5bcdd] Plots v1.3.5

julia> using Plots, DataFrames

julia> cs = DataFrame([(x1 = i, symbol = string(i)) for i in 1:50]);

julia> plot(cs.symbol[1:50], cs.x1[1:50], yscale=:log10, label=cs.symbol[1:50])
*** stack smashing detected ***: <unknown> terminated

signal (6): Aborted
in expression starting at none:0
gsignal at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
abort at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
unknown function (ip: 0x7f32d122038d)
unknown function (ip: 0x7f32d12c2e28)
__stack_chk_fail at /lib/x86_64-linux-gnu/libc.so.6 (unknown line)
process_glyphs at /home/vbogad/.julia/packages/GR/cRdXQ/src/../deps/gr/lib/libGR.so (unknown line)
unknown function (ip: 0x3400000021)
Allocations: 80212338 (Pool: 80182132; Big: 30206); GC: 75
Aborted (core dumped)

The GR version, according to the Manifest, is 0.49.1. I think this is a GR problem though, I'll see if I can step through and locate the error. Not having a line number makes this harder though...

[BeastyBlacksmith]

cc: @jheinen

[anandijain]

Also which version of DataFrames is this?
I can't reproduce on julia 1.4.2, Plots 1.3.5 and DataFrames 0.21.1

df = DataFrame([(a = i, b = Symbol(i)) for i in 1:10])
plot(df.b[1:3], df.a[1:3], yscale = :log10, label=df.b[1:3])
ERROR: Cannot convert Symbol to series data for plotting
Stacktrace:

Sorry, confusingly, the symbol column is type string, not Symbol.

[stev47]

another example that crashes GR in a similar way for me: plot(1.1.^(1:10000)) (on Julia 1.4.2, Plots 1.5.2, GR 0.50.1)

[jheinen]

1.1 .^ (1:10000)
10000-element Array{Float64,1}:
  1.1
  1.2100000000000002
  1.3310000000000004
  1.4641000000000004
  1.6105100000000006
  1.7715610000000008
  1.9487171000000012
  2.1435888100000016
  2.357947691000002
  2.5937424601000023
  ⋮
 Inf
 Inf
 Inf
 Inf
 Inf
 Inf
 Inf
 Inf
 Inf

What should happen with the Infs?

[stev47]

They should not lead to a core-dump:

julia> using Plots
julia> plot(1.1.^(1:10000))
┌ Warning: No strict ticks found
└ @ PlotUtils ~/.julia/packages/PlotUtils/nCtbM/src/ticks.jl:283
*** stack smashing detected ***: terminated

signal (6): Aborted
in expression starting at REPL[1]:0
gsignal at /usr/bin/../lib/libc.so.6 (unknown line)
abort at /usr/bin/../lib/libc.so.6 (unknown line)
__libc_message at /usr/bin/../lib/libc.so.6 (unknown line)
__fortify_fail at /usr/bin/../lib/libc.so.6 (unknown line)
__stack_chk_fail at /usr/bin/../lib/libc.so.6 (unknown line)
process_glyphs at /home/stev47/.julia/packages/GR/Atztx/src/../deps/gr/lib/libGR.so (unknown line)
unknown function (ip: 0x3500000032)
Allocations: 31575277 (Pool: 31568762; Big: 6515); GC: 27
Aborted (core dumped)

Even though the error is most probably in GR, I cannot reproduce this with the GR package, only when calling through Plots.

[BeastyBlacksmith]

Probably Plots converts this in many empty series, since it treats Infs the same as NaNs