Crash with summarize(n=n())

[momeara]

With the latest dplyr, d3ded01, executing this

library(dplyr)

z <- data.frame(
  x = c("a", "a"),
  y = c("1", "2"),
  z = c("b", "b"),
  stringsAsFactors=FALSE)  %>%
  dplyr::group_by(x, z) %>%
  dplyr::summarize(n=n())

gives

  *** caught segfault ***
 address (nil), cause 'unknown'

 Traceback:
  1: .Call(`_dplyr_summarise_impl`, df, dots)
  2: summarise_impl(.data, dots)
  3: summarise.tbl_df(., n = n())
  4: dplyr::summarize(., n = n())
  5: function_list[[1L]](value)
  6: freduce(value, `_function_list`)
  7: Recall(function_list[[1L]](value), function_list[-1L])
  8: freduce(value, `_function_list`)
  9: `_fseq`(`_lhs`)
 10: eval(quote(`_fseq`(`_lhs`)), env, env)
 11: eval(quote(`_fseq`(`_lhs`)), env, env)
 12: withVisible(eval(quote(`_fseq`(`_lhs`)), env, env))
 13: data.frame(x = c("a", "a"), y = c("1", "2"), z = c("b", "b"),     stringsAsFactors = FALSE) %>% dplyr::group_by(x, z) %>% dplyr::summarize(n = n())

 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: 2
!Save workspace image? [y/n/c]: n

This seg fault occurs on both linux,

> sessionInfo()
sessionInfo()
 R version 3.4.1 (2017-06-30)
 Platform: x86_64-pc-linux-gnu (64-bit)
 Running under: CentOS release 6.7 (Final)

 Matrix products: default
 BLAS: /mnt/nfs/home/momeara/opt/lib64/R/lib/libRblas.so
 LAPACK: /mnt/nfs/home/momeara/opt/lib64/R/lib/libRlapack.so

 locale:
  [1] LC_CTYPE=en_CA.UTF-8       LC_NUMERIC=C
  [3] LC_TIME=en_CA.UTF-8        LC_COLLATE=en_CA.UTF-8
  [5] LC_MONETARY=en_CA.UTF-8    LC_MESSAGES=en_CA.UTF-8
  [7] LC_PAPER=en_CA.UTF-8       LC_NAME=C
  [9] LC_ADDRESS=C               LC_TELEPHONE=C
 [11] LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C

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

 loaded via a namespace (and not attached):
 [1] compiler_3.4.1 tools_3.4.1

and macOS

> sessionInfo()
R version 3.4.0 (2017-04-21)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS  10.13.4

Matrix products: default
BLAS: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib

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

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

other attached packages:
[1] dplyr_0.7.99.9000

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.17       bindr_0.1.1        knitr_1.20         magrittr_1.5
 [5] devtools_1.13.5    tidyselect_0.2.4   munsell_0.4.3      colorspace_1.3-2
 [9] ape_5.1            lattice_0.20-35    R6_2.2.2           rlang_0.2.1.9000
[13] httr_1.3.1         plyr_1.8.4         tools_3.4.0        parallel_3.4.0
[17] grid_3.4.0         ggtree_1.8.2       gtable_0.2.0       nlme_3.1-137
[21] git2r_0.21.0       withr_2.1.2        assertthat_0.2.0   lazyeval_0.2.1
[25] digest_0.6.15      tibble_1.4.2       treeio_1.0.2       bindrcpp_0.2.2
[29] purrr_0.2.5        ggplot2_2.2.1.9000 tidyr_0.8.1        curl_3.2
[33] memoise_1.1.0      glue_1.3.0         compiler_3.4.0     pillar_1.2.3
[37] rvcheck_0.1.0      scales_0.5.0.9000  jsonlite_1.5       pkgconfig_2.0.1

[momeara]

I can confirm that the version currently in CRAN, 0.7.6 doesn't cause the seg fault

> sessionInfo()
 R version 3.4.1 (2017-06-30)
 Platform: x86_64-pc-linux-gnu (64-bit)
 Running under: CentOS release 6.7 (Final)

 Matrix products: default
 BLAS: /mnt/nfs/home/momeara/opt/lib64/R/lib/libRblas.so
 LAPACK: /mnt/nfs/home/momeara/opt/lib64/R/lib/libRlapack.so

 locale:
  [1] LC_CTYPE=en_CA.UTF-8       LC_NUMERIC=C
  [3] LC_TIME=en_CA.UTF-8        LC_COLLATE=en_CA.UTF-8
  [5] LC_MONETARY=en_CA.UTF-8    LC_MESSAGES=en_CA.UTF-8
  [7] LC_PAPER=en_CA.UTF-8       LC_NAME=C
  [9] LC_ADDRESS=C               LC_TELEPHONE=C
 [11] LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C

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

 other attached packages:
 [1] bindrcpp_0.2.2.9000 dplyr_0.7.6

 loaded via a namespace (and not attached):
  [1] tidyselect_0.2.4      compiler_3.4.1        magrittr_1.5.0.9000
  [4] assertthat_0.2.0.9000 R6_2.2.2              tools_3.4.1
  [7] pillar_1.3.0.9000     glue_1.3.0            tibble_1.4.2.9004
 [10] crayon_1.3.4          Rcpp_0.12.17.4        pkgconfig_2.0.1
 [13] rlang_0.2.1.9000      purrr_0.2.4.9000      bindr_0.1.1.9000

[krlmlr]

Thanks, confirmed. @romainfrancois: Could you please take a look?

[romainfrancois]

Sure. Will look it up in august.

[krlmlr]

Bisection identifies 072f050 as the first bad commit. I'll take a look in the context of the failures seen after #3610.

[krlmlr]

This seems to be related to string columns and grouping by more than one column:

library(tidyverse)

# Works
tibble(x = "a", z = "b") %>%
  count(x)

# Works
tibble(x = 1, z = 2) %>%
  count(x, z)

# Segfaults
tibble(x = "a", z = "b") %>%
  count(x, z)

[krlmlr]

Not sure if it's related, but with -O0 I'm seeing a protection error with the following code:

library(tidyverse)

d <- tibble(x = "a", z = "b") %>% group_by(x, z)

gc()
gctorture2(1)
z <- summarize(d, n())

[krlmlr]

The protection error also happens with:

library(tidyverse)

d <- tibble(x = "a", z = "b") %>% group_by(x)

gc()
gctorture2(1)
z <- summarize(d, n())

and

library(tidyverse)

d <- tibble(x = 1, z = 2) %>% group_by(x)

gc()
gctorture2(1)
z <- summarize(d, n())

it's unlikely to be related. Opening a new issue.

[lock]

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/