Skip to content

Reproducible Segfault When Loading/Unloading data.table #990

Closed
#7451
Closed
Reproducible Segfault When Loading/Unloading data.table#990
#7451
Labels
bug

Description

@brodieG
brodieG
opened on Dec 18, 2014

In the R console:

library(data.table)
# data.table 1.9.5  For help type: ?data.table
# *** NB: by=.EACHI is now explicit. See README to restore previous behaviour.
dt.1 <- data.table(a=1:10, b=1:10)
detach("package:data.table", character.only=TRUE, unload=TRUE)

#  *** caught segfault ***
# address 0x10e79fbf0, cause 'memory not mapped'

This is the session info before running the above code:

sessionInfo()
# R version 3.1.1 (2014-07-10)
# Platform: x86_64-apple-darwin13.1.0 (64-bit)

# locale:
# [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

# attached base packages:
# [1] stats     graphics  grDevices utils     datasets  methods   base

This also happens in 1.9.4. In RStudio, segfault doesn't happen right away, but running code multiple times eventually causes it to happen.
Also reproduced with 1.9.4 in R3.1.2 on Win7.64.

Activity

  1. arunsrinivasan
    added
    bug
    on Jan 3, 2015

  2. arunsrinivasan commented on Jan 3, 2015

    @arunsrinivasan
    arunsrinivasan
    on Jan 3, 2015
    Member

    Thanks for the report. Reproducible. Please make sure you've read the contribution guidelines.

  3. brodieG commented on Jan 5, 2015

    @brodieG
    brodieG
    on Jan 5, 2015
    Author

    Arun, I did read the contribution guidelines. What did I miss?

  4. arunsrinivasan commented on Jan 5, 2015

    @arunsrinivasan
    arunsrinivasan
    on Jan 5, 2015
    Member

    Remove the > and comment the output, so that it's easy to copy/paste. I edited yours.

  5. Alectoria commented on Aug 17, 2015

    @Alectoria
    Alectoria
    on Aug 17, 2015

    It seems like the bug above was not fixed in 1.9.5:
    The code below creates a fatal crash in my fresh R session (after running it multiple times):

    library(data.table)
DT = data.table(x=rep(c("a","b","c"),each=3), y=c(1,3,6), v=1:9)
detach("package:data.table", unload=TRUE)

    Interestingly, this seems to be ok:

    library(data.table)
#DT = data.table(x=rep(c("a","b","c"),each=3), y=c(1,3,6), v=1:9)
detach("package:data.table", unload=TRUE)

    Here is my sessionInfo():

    R version 3.2.1 (2015-06-18)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.9.5 (Mavericks)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] data.table_1.9.5 devtools_1.8.0  

loaded via a namespace (and not attached):
 [1] httr_1.0.0      R6_2.1.0        magrittr_1.5    rversions_1.0.1 tools_3.2.1     curl_0.9.1          Rcpp_0.12.0    
 [8] memoise_0.2.1   xml2_0.1.1      stringi_0.5-5   knitr_1.10.5    git2r_0.10.1    stringr_1.0.0       digest_0.6.8   
[15] chron_2.3-47

    Same things happen on 1.9.4

  6. brodieG commented on Oct 7, 2015

    @brodieG
    brodieG
    on Oct 7, 2015
    Author

    Still reproducible in 1.9.6, though we need to add the additional library call at the end.

    library(data.table)
dt.1 <- data.table(a=1:10, b=1:10)
detach("package:data.table", character.only=TRUE, unload=TRUE)
library(data.table)

# *** caught segfault ***
# address 0x110eb2300, cause 'memory not mapped'
# 
# Possible actions:
# 1: abort (with core dump, if enabled)
# 2: normal R exit
# 3: exit R without saving workspace
# 4: exit R saving workspace

  7. jangorecki commented on Apr 3, 2020

    @jangorecki
    jangorecki
    on Apr 3, 2020
    Member

    Cannot reproduce it anymore, must have been fixed in the last 5 years. Checked on Linux. I would appriciate if someone on Windows and Mac could confirm as well if it is fixed.

    library(data.table)
#data.table 1.12.9 IN DEVELOPMENT built 2020-04-03 10:50:14 UTC; jan using 1 threads (see ?getDTthreads).  Latest news: r-datatable.com
dt.1 <- data.table(a=1:10, b=1:10)
detach("package:data.table", character.only=TRUE, unload=TRUE)
library(data.table)
#data.table 1.12.9 IN DEVELOPMENT built 2020-04-03 10:50:14 UTC; jan using 1 threads (see ?getDTthreads).  Latest news: r-datatable.com
  8. jangorecki
    closed this as completedon Apr 3, 2020

  9. brodieG commented on Apr 3, 2020

    @brodieG
    brodieG
    on Apr 3, 2020
    Author

    Still happening on OS X:

    R version 3.6.0 (2019-04-26) -- "Planting of a Tree"
Copyright (C) 2019 The R Foundation for Statistical Computing
Platform: x86_64-apple-darwin15.6.0 (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

  Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

library(data.table)
data.table 1.12.2 using 2 threads (see ?getDTthreads).  Latest news: r-datatable.com
dt.1 <- data.table(a=1:10, b=1:10)
detach("package:data.table", character.only=TRUE, unload=TRUE)
library(data.table)

 *** caught segfault ***
address 0x11147dfe0, cause 'memory not mapped'

Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace
Selection:

  10. MichaelChirico commented on Apr 3, 2020

    @MichaelChirico
    MichaelChirico
    on Apr 3, 2020
    Member

    Maybe R version-specific? not happening for me on OSx

    > sessionInfo()
R version 3.6.0 (2019-04-26)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Mojave 10.14.6

Matrix products: default
BLAS:   /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libRlapack.dylib

locale:
[1] C/UTF-8/C/C/C/C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
[1] compiler_3.6.0
> library(data.table)
data.table 1.12.9 IN DEVELOPMENT built 2020-02-20 09:16:40 UTC; travis using 4 threads (see ?getDTthreads).  Latest news: r-datatable.com
**********
This development version of data.table was built more than 4 weeks ago. Please update: data.table::update.dev.pkg()
**********
> data.table 1.12.2 using 2 threads (see ?getDTthreads).  Latest news: r-datatable.com
 エラー:  想定外の数値定数です  in "data.table 1.12"
> dt.1 <- data.table(a=1:10, b=1:10)
> detach("package:data.table", character.only=TRUE, unload=TRUE)
> library(data.table)
data.table 1.12.9 IN DEVELOPMENT built 2020-02-20 09:16:40 UTC; travis using 4 threads (see ?getDTthreads).  Latest news: r-datatable.com
**********
This development version of data.table was built more than 4 weeks ago. Please update: data.table::update.dev.pkg()
**********

  11. brodieG commented on Apr 3, 2020

    @brodieG
    brodieG
    on Apr 3, 2020
    Author

    I tried again with 1.12.8 and this appears to have gone away.

  12. MichaelChirico commented on Apr 3, 2020

    @MichaelChirico
    MichaelChirico
    on Apr 3, 2020
    Member

    Nothing particularly stands out to me as a fix for this issue, but anyway, this commit came up from git bisect as the first fixed commit

    04eeb30

  13. aitap commented on Jul 3, 2025

    @aitap
    aitap
    on Jul 3, 2025
    Member

    The issue still exists in current master:

    library(data.table)
x <- as.data.table(mtcars)
detach('package:data.table', unload = TRUE)
rm(x)
gc(full = TRUE)

    Thread 1 "R" received signal SIGSEGV, Segmentation fault.
0x00007fff71c91670 in ?? ()
(gdb) bt
#0  0x00007fff71c91670 in ?? ()
#1  0x00007ffff7b92d29 in R_RunWeakRefFinalizer (w=<optimized out>)
    at ../../../src/main/memory.c:1522
#2  0x00007ffff7b92faa in RunFinalizers () at ../../../src/main/memory.c:1589
#3  0x00007ffff7b93095 in R_RunPendingFinalizers ()
    at ../../../src/main/memory.c:1625
#4  0x00007ffff7b9679e in R_gc () at ../../../src/main/memory.c:3167

    The crash is due to the package DLL containing the finalizer having been unloaded from the address space despite the object still exists. On R < 3.4, the finalizer is an excellent hack, the only way to make sure that R counts memory correctly; otherwise, the over-allocated part of the list of pointers is not accounted for when the memory is released. On R ≥ 3.4, SET_GROWABLE_BIT could be used instead... if it was an API entry point.

  14. aitap
    reopened this on Jul 3, 2025
  15. aitap
    closed this as completedin #7451on Dec 12, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    bug

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions