[Rd] FW: R-devel error

luke-tierney at uiowa.edu luke-tierney at uiowa.edu
Wed Dec 6 18:28:00 CET 2017


library() in R-devel checks that packages are built for the current
ABI and signals an error if they are not. A segfault is likely a
pre-existing bug unmasked by the changed memory layout. Several of
these have been resolved by recent package updates. I also cannot
reproduce the problem on current R-devel with fairly current packages
from CRAN.

Best,

luke

On Wed, 6 Dec 2017, Gabriel Becker wrote:

> Robert,
>
> To expand a bit on Goran's point, is it possible that you have an old
> library with packages installed that your R-devel is hitting. The R headers
> changes relatively recently, so any packages with compiled code that were
> built long enough ago (I don't have the exact date off the top of my head)
> need to be rebuilt before they can be safely used.
>
> If this is happening in a state where both the package in question AND all
> it's dependencies are freshly built under the exact version of R-devel
> you're trying to check httk with, then the above isn't the problem and
> something else must be going on.
>
> Best,
> ~G
>
> On Wed, Dec 6, 2017 at 8:14 AM, Göran Broström <goran.brostrom at umu.se>
> wrote:
>
>> Interesting, yesterday I had exactly this problem, but today I solved it
>> (see the R-pkg-devel list) by _installing_ R-devel and
>>
>> > update.packages(checkBuilt = TRUE)
>>
>> I also tried it on Martin's Fedora 26 example below. Worked fine (had to
>> install some packages...)
>>
>> On ubuntu 16.04.
>>
>> Göran Broström
>>
>>
>> On 2017-12-06 11:29, Martin Maechler wrote:
>>
>>> Pearce, Robert <Pearce.Robert at epa.gov>
>>>>>>>>      on Tue, 5 Dec 2017 21:31:09 +0000 writes:
>>>>>>>>
>>>>>>>
>>>      > I am resubmitting this bug report but with additional information.
>>> I am running this with windows 10:  w64-mingw32 with R Under development
>>> (unstable) (2017-12-04 r73829).
>>>
>>> There is no such message (as you cite below) in the R-devel archives:
>>>        https://stat.ethz.ch/pipermail/r-devel/2017-November/thread.html
>>>
>>> Hence it was never posted for some reason
>>> (You may have *sent* it, but it never arrived "safely")
>>>
>>>
>>>      > I build 'httk' from the command prompt using 'R CMD build httk'
>>> after installing the required packages. Then when the vignettes are being
>>> created, it crashes. Today I installed the latest versions of Rtools,
>>> MikTex and R-devel and still got the same error. We have also replicated
>>> this on another machine. I am able to successfully build the package in
>>> this way with R 3.4.3.
>>>
>>> Robert
>>>>
>>>
>>> To make this reproducible, indeed, I also get
>>> problem, possibly the same as you on my Linux machine (Fedora 26):
>>>
>>> 1. Get  httk_1.7.tar.gz from https://cran.r-project.org/package=httk
>>>
>>> 2. unpack (for me:)
>>>         tar xf httk_1.7.tar.gz
>>> 3.
>>>         R-devel CMD build httk
>>>
>>> I get the following.
>>> Notably after having waited several minutes at the prompt
>>> (line  of output):
>>>
>>>   '* creating vignettes '
>>>
>>>    -----------------------------------------------------------
>>> -------------
>>>    [nfslocal at lynne tmp]$ R-devel CMD build httk
>>>    * checking for file ‘httk/DESCRIPTION’ ... OK
>>>    * preparing ‘httk’:
>>>    * checking DESCRIPTION meta-information ... OK
>>>    * cleaning src
>>>    * installing the package to process help pages
>>>    * saving partial Rd database
>>>    * creating vignettes ...sh: line 1: 16259 Segmentation fault
>>> (core dumped) '/u/maechler/R/D/r-devel/inst/bin/Rscript' --vanilla
>>> --default-packages= -e "tools::buildVignettes(dir = '.', tangle = TRUE)" >
>>> '/tmp/RtmpV1grVw/xshell3e793c7d5258' 2>&1
>>>     ERROR
>>>
>>>     *** caught segfault ***
>>>    address 0x1, cause 'memory not mapped'
>>>
>>>    Traceback:
>>>     1: .Call(C_stri_locate_all_regex, str, pattern, omit_no_match,
>>> opts_regex)
>>>     2: stri_locate_all_regex(string, pattern, omit_no_match = TRUE,
>>>  opts_regex = opts)
>>>     3: stringr::str_locate_all(input, inline.code)
>>>     4: parse_inline(g, patterns)
>>>     5: FUN(X[[i]], ...)
>>>     6: lapply(groups, function(g) {    block = grepl(chunk.begin, g[1])
>>>   if (!set.preamble && !parent_mode()) {        return(if (block) "" else
>>> g)    }    if (block) {        n = length(g)        if (n >= 2 &&
>>> grepl(chunk.end, g[n]))             g = g[-n]        g = strip_block(g,
>>> patterns$chunk.code)        params.src = if (group_pattern(chunk.begin)) {
>>>           stringr::str_trim(gsub(chunk.begin, "\\1", g[1]))        }
>>>     else ""        parse_block(g[-1], g[1], params.src)    }    else
>>> parse_inline(g, patterns)})
>>>     7: split_file(lines = text)
>>>     8: process_file(text, output)
>>>     9: knitr::knit(knit_input, knit_output, envir = envir, quiet =
>>> quiet,     encoding = encoding)
>>>    10: rmarkdown::render(file, encoding = encoding, quiet = quiet, envir
>>> = globalenv())
>>>    11: vweave_rmarkdown(...)
>>>    12: engine$weave(file, quiet = quiet, encoding = enc)
>>>    13: doTryCatch(return(expr), name, parentenv, handler)
>>>    14: tryCatchOne(expr, names, parentenv, handlers[[1L]])
>>>    15: tryCatchList(expr, classes, parentenv, handlers)
>>>    16: tryCatch({    engine$weave(file, quiet = quiet, encoding = enc)
>>> setwd(startdir)    find_vignette_product(name, by = "weave", engine =
>>> engine)}, error = function(e) {    stop(gettextf("processing vignette '%s'
>>> failed with diagnostics:\n%s",         file, conditionMessage(e)), domain =
>>> NA, call. = FALSE)})
>>>    17: tools::buildVignettes(dir = ".", tangle = TRUE)
>>>    An irrecoverable exception occurred. R is aborting now ...
>>>    [nfslocal at lynne tmp]$
>>>    -----------------------------------------------------------
>>> -------------
>>>
>>>
>>> I've seen this (segfault, triggered from stringr stuff) before,
>>> occasionally and always thought it was a consequence of a messup
>>> (different R versions and package installations in libraries I
>>> use on the same computer....),  but now I have convinced myself
>>> that this is not the case: I can easily reproduce the above using
>>>
>>>         tools::buildVignettes(dir = "<mypath_to>/httk")
>>>
>>> ==> So I'd suggest there's a pretty low-level bug somewhere.
>>>
>>> - Either in 'stringr' or 'stringi' in their C code using
>>>    things they should not (because not in R's API),
>>>    which now shows in R-devel where many "inner workings" have been
>>>    modified (key "ALTREP")
>>>
>>> - or in R-devel .. in these new "inner workings".
>>>
>>>
>>> I don't have time to dig further at the moment.
>>>
>>> Thank you, Robert, for your report!
>>>
>>> Martin Maechler,
>>> ETH Zurich
>>>
>>>
>>>      > From: Pearce, Robert
>>>      > Sent: Monday, November 20, 2017 11:32 AM
>>>      > To: 'r-devel at r-project.org' <r-devel at r-project.org>
>>>      > Subject: R-devel error
>>>
>>>
>>>      > Hello,
>>>
>>>      > During the last week I've been getting an error with r-devel when
>>> building the source package for 'httk'. Everything works fine until:
>>> 'creating vignettes...' . Then I get a pop-up stating: 'R for Windows
>>> front-end has stopped working... close the program'. I have installed the
>>> latest version of R, and this error does not occur.
>>>
>>>      > Robert Pearce
>>>
>>> ______________________________________________
>>> R-devel at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>>
>>>
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>
>
>
>
> -- 
> Gabriel Becker, PhD
> Scientist (Bioinformatics)
> Genentech Research
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel

-- 
Luke Tierney
Ralph E. Wareham Professor of Mathematical Sciences
University of Iowa                  Phone:             319-335-3386
Department of Statistics and        Fax:               319-335-3017
    Actuarial Science
241 Schaeffer Hall                  email:   luke-tierney at uiowa.edu
Iowa City, IA 52242                 WWW:  http://www.stat.uiowa.edu


More information about the R-devel mailing list