[Bioc-devel] msPurity build fail on Mac OS X (morelia)

Hi All,

Thanks for the help with this. I have been using valgrind to try and
pinpoint in my code the problem and it seems to be from mzR::openMSfile().

library(mzR)

msmsPths <- list.files(system.file("extdata", "lcms", "mzML", package="msPurityData"), full.names = TRUE, pattern = "MSMS") 
msmsPths

tmp <- replicate(1000, mr <- mzR::openMSfile(msmsPths[1]))
all(sapply(tmp, validObject))

sessionInfo()

A simplified bit of code just using mzR I think will give the same results,
see below:

#-----------------------------------------------------------
------------------------------------------------------------
------------------------------------------------------------
library(mzR)
msmsPths <- list.files(system.file("extdata", "lcms", "mzML",
package="msPurityData"), full.names = TRUE, pattern = "MSMS")
mr <- mzR::openMSfile(msmsPths[1])
#-----------------------------------------------------------
------------------------------------------------------------
------------------------------------------------------------

This gives the  familiar message

mr <- mzR::openMSfile(msmsPths[1])

==38138== Mismatched free() / delete / delete []
==38138==    at 0x4C2EDEB: free (in /usr/lib/valgrind/vgpreload_
memcheck-amd64-linux.so)
..

However, I can prevent the  "Mismatched free() / delete / delete []"
message by forcing the backend to use the "pwiz" libraries (i think). e.g.

#-----------------------------------------------------------
------------------------------------------------------------
------------------------------------------------------------
mr <- mzR::openMSfile(msmsPths[1], backend="pwiz")
#-----------------------------------------------------------
------------------------------------------------------------
------------------------------------------------------------

I have updated to code so that a user can choose the backend for mzR with a
default option of "pwiz" and pushed the changes to svn.

Hopefully this will fix the problem on morelia, although I notice that the
server is down at the moment, when will the next build be?

Thanks for your help,
Tom

PS: Martin: I have removed the msPurity-vignette.R and
msPurity-vignette.html files from the vignette folder. Is this correct?





On Wed, Sep 21, 2016 at 7:38 PM, Kasper Daniel Hansen <
kasperdanielhansen at gmail.com> wrote:

So this seems like a bug in mzR which is the main backend / parser for
proteomics and metabolomics data.  Someone should try to fix this.

Unfortunately, mzR depends on several pieces of software which are written
by other people, but bundled with the package.  Martin's traceback suggests
that some of these files are (partially?) encrypted.  This might be hard to
address.

Best,
Kasper

On Wed, Sep 21, 2016 at 12:43 PM, Dan Tenenbaum <dtenenba at fredhutch.org>
wrote:

One thing I notice is that the crash does not happen every time. I have
successfully built the package on morelia by hand with "R CMD build".

Similarly I can source the stangled vignette without a crash sometimes.
But when it does crash, this is what I see:

source("msPurity-vignette.R", echo=TRUE, max=Inf)

## ------------------------------------------------------------

------------

library(msPurity)

Loading required package: Rcpp

msmsPths <- list.files(system.file("extdata", "lcms", "mzML",

package="msPurityData"), full.names = TRUE, pattern = "MSMS")

msPths <- list.files(system.file("extdata", "lcms", "mzML",

package="msPurityData"), full.names = TRUE, pattern = "LCMS_")

## ------------------------------------------------------------

------------

pa <- purityA(msmsPths)

 *** caught segfault ***
address 0x29800000297, cause 'memory not mapped'

Traceback:
 1: .External(list(name = "CppMethod__invoke_notvoid", address =
<pointer: 0x7fb56861cb10>,     dll = list(name = "Rcpp", path =
"/Library/Frameworks/R.framework/Versions/3.3/Resources/library/Rcpp/libs/Rcpp.so",
       dynamicLookup = TRUE, handle = <pointer: 0x7fb568606180>,
 info = <pointer: 0x1078f8000>), numParameters = -1L),     <pointer:
0x7fb5686ae9c0>, <pointer: 0x7fb56868cf80>, .pointer,     ...)
 2: object at backend$getPeakList(x)
 3: FUN(X[[i]], ...)
 4: lapply(X = X, FUN = FUN, ...)
 5: sapply(scans, function(x) object at backend$getPeakList(x)$peaks,
 simplify = FALSE)
 6: sapply(scans, function(x) object at backend$getPeakList(x)$peaks,
 simplify = FALSE)
 7: .local(object, ...)
 8: mzR::peaks(mr)
 9: mzR::peaks(mr)
10: getscans(filepth)
11: assessPuritySingle(filepth = pa at fileList[[i]], mostIntense =
mostIntense,     nearest = nearest, offsets = offsets, plotP = plotP,
plotdir = plotdir,     interpol = interpol, iwNorm = iwNorm, iwNormFun =
iwNormFun,     ilim = ilim)
12: eval(expr, envir, enclos)
13: eval(xpr, envir = envir)
14: doTryCatch(return(expr), name, parentenv, handler)
15: tryCatchOne(expr, names, parentenv, handlers[[1L]])
16: tryCatchList(expr, classes, parentenv, handlers)
17: tryCatch(eval(xpr, envir = envir), error = function(e) e)
18: doTryCatch(return(expr), name, parentenv, handler)
19: tryCatchOne(expr, names, parentenv, handlers[[1L]])
20: tryCatchList(expr, classes, parentenv, handlers)
21: tryCatch({    repeat {        args <- nextElem(it)        if
(obj$verbose) {            cat(sprintf("evaluation # %d:\n", i))
print(args)        }        for (a in names(args)) assign(a, args[[a]], pos
= envir,             inherits = FALSE)        r <- tryCatch(eval(xpr, envir
= envir), error = function(e) e)        if (obj$verbose) {
cat("result of evaluating expression:\n")            print(r)        }
  tryCatch(accumulator(list(r), i), error = function(e) {
cat("error calling combine function:\n")            print(e)
NULL        })        i <- i + 1    }}, error = function(e) {    if
(!identical(conditionMessage(e), "StopIteration"))
 stop(simpleError(conditionMessage(e), expr))})
22: e$fun(obj, substitute(ex), parent.frame(), e$data)
23: operator(foreach::foreach(i = 1:length(pa at fileList), .packages =
"mzR"),     assessPuritySingle(filepth = pa at fileList[[i]], mostIntense =
mostIntense,         nearest = nearest, offsets = offsets, plotP = plotP,
       plotdir = plotdir, interpol = interpol, iwNorm = iwNorm,
 iwNormFun = iwNormFun, ilim = ilim))
24: purityA(msmsPths)
25: eval(expr, envir, enclos)
26: eval(ei, envir)
27: withVisible(eval(ei, envir))
28: source("msPurity-vignette.R", echo = TRUE, max = Inf)

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:

sessionInfo()

R version 3.3.1 (2016-06-21)
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] parallel  stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
[1] xcms_1.49.6         Biobase_2.33.3      ProtGenerics_1.5.1
[4] BiocGenerics_0.19.2 mzR_2.7.4           msPurity_0.99.6
[7] Rcpp_0.12.7

loaded via a namespace (and not attached):
 [1] sapa_2.0-2             magrittr_1.5           ifultools_2.0-4
 [4] MASS_7.3-45            splines_3.3.1          BiocParallel_1.7.8
 [7] lattice_0.20-34        foreach_1.4.3          splus2R_1.2-2
[10] stringr_1.1.0          fastcluster_1.1.21     plyr_1.8.4
[13] tools_3.3.1            grid_3.3.1             snow_0.4-1
[16] iterators_1.0.8        survival_2.39-5        multtest_2.29.0
[19] doSNOW_1.0.14          Matrix_1.2-7.1         RColorBrewer_1.1-2
[22] reshape2_1.4.1         S4Vectors_0.11.16      codetools_0.2-14
[25] MassSpecWavelet_1.39.0 stringi_1.1.1          compiler_3.3.1
[28] stats4_3.3.1           RANN_2.5

----- Original Message -----

From: "Thomas Lawson" <thomas.nigel.lawson at gmail.com>
To: "Martin Morgan" <martin.morgan at roswellpark.org>
Cc: "bioc-devel" <Bioc-devel at r-project.org>
Sent: Wednesday, September 21, 2016 8:42:20 AM
Subject: Re: [Bioc-devel] msPurity build fail on Mac OS X (morelia)

Thanks for reply. Some of those errors are a bit cryptic for me also.

I have not heard of the valgrind functionality before in R. I will test

a

few things out with valgrind and hopefully I can pinpoint the error a

bit

more.

Thanks again.

Tom

On Wed, Sep 21, 2016 at 12:50 PM, Martin Morgan <
martin.morgan at roswellpark.org> wrote:

On 09/20/2016 05:18 AM, Thomas Lawson wrote:

Hi BioConductor community,

My package (msPurity) is passing the build on the Linux (*zin1*) and
Windows servers (*moscato1*) but failing on the Mac OS X server
(*morelia*).
Also I cannot seem to replicate the failure either on a local

installation

of Mac OS X (el captain) or with Travis CI.

I should probably note that for Travis CI I did have to install the
msPurityData dependency directly (without Bioconductor). See line 10
https://raw.githubusercontent.com/Viant-Metabolomics/msPurit
y/master/.travis.yml

The error I think is coming from a function I have that uses the
mzR::peaks() function but I am struggling to see why I am getting the
error.

Any help or suggestions would be really appreciated.

Hi Tom --

This might be fun!

You can see from


http://bioconductor.org/checkResults/3.4/bioc-LATEST/morelia
-R-instpkgs.html

(linked from http://bioconductor.org/checkR

esults/3.4/bioc-LATEST/index.

html '1633' installed packages) that in fact msPurityData is installed.
Also, segfaults are rarely the result of missing packages. Instead, it

is

likely due to errors in C code of one sort or another. On my linux, I

made

sure I was using the 'devel' version of Bioconductor, and that all of

my

packages were up-to-date via biocLite(). I then checked out msPurity

from

svn, changed into the msPurity directory and installed it

    R CMD INSTALL .

then I changed to the vignettes directory, Stangled the source code

   cd vignettes
   R CMD Stangle msPurity-vignette.Rmd

(by the way, the products of build the package / vignette,
msPurity-vignette.R should not be in svn).

I then ran the vignette under valgrind

    msPurity/vignettes$ R -d valgrind -f msPurity-vignette.R

leading to

pa <- purityA(msmsPths)

==19611== Mismatched free() / delete / delete []
==19611==    at 0x4C2EDEB: free (in /usr/lib/valgrind/vgpreload_me
mcheck-amd64-linux.so)
==19611==    by 0x11296DA5: cRamp::cRamp(char const*, bool)

(cramp.cpp:98)

==19611==    by 0x1129FC87: RcppRamp::open(char const*, bool)
(RcppRamp.cpp:23)
==19611==    by 0x112B49C4: Rcpp::CppMethod2<RcppRamp, void, char

const*,

bool>::operator()(RcppRamp*, SEXPREC**) (Module_generated_CppMethod.h:

215)

==19611==    by 0x112B0FBF: Rcpp::class_<RcppRamp>::invoke

_void(SEXPREC*,

SEXPREC*, SEXPREC**, int) (class.h:212)
==19611==    by 0xED73CA0: CppMethod__invoke_void(SEXPREC*)
(Module.cpp:200)
==19611==    by 0x4F0DFD0: do_External (dotcode.c:548)
==19611==    by 0x4F47F9E: Rf_eval (eval.c:713)
==19611==    by 0x4F4A6B7: do_begin (eval.c:1807)
==19611==    by 0x4F47D90: Rf_eval (eval.c:685)
==19611==    by 0x4F4964C: Rf_applyClosure (eval.c:1135)
==19611==    by 0x4F47B6C: Rf_eval (eval.c:732)
==19611==  Address 0x1b8a4220 is 0 bytes inside a block of size 400

alloc'd

==19611==    at 0x4C2E0EF: operator new(unsigned long) (in
/usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==19611==    by 0x11296949: cRamp::do_ramp(long, eWhatToRead)
(cramp.cpp:215)
==19611==    by 0x11296D9D: cRamp::cRamp(char const*, bool)

(cramp.cpp:97)

==19611==    by 0x1129FC87: RcppRamp::open(char const*, bool)
(RcppRamp.cpp:23)
==19611==    by 0x112B49C4: Rcpp::CppMethod2<RcppRamp, void, char

const*,

bool>::operator()(RcppRamp*, SEXPREC**) (Module_generated_CppMethod.h:

215)

==19611==    by 0x112B0FBF: Rcpp::class_<RcppRamp>::invoke

_void(SEXPREC*,

SEXPREC*, SEXPREC**, int) (class.h:212)
==19611==    by 0xED73CA0: CppMethod__invoke_void(SEXPREC*)
(Module.cpp:200)
==19611==    by 0x4F0DFD0: do_External (dotcode.c:548)
==19611==    by 0x4F47F9E: Rf_eval (eval.c:713)
==19611==    by 0x4F4A6B7: do_begin (eval.c:1807)
==19611==    by 0x4F47D90: Rf_eval (eval.c:685)
==19611==    by 0x4F4964C: Rf_applyClosure (eval.c:1135)
==19611==

which from http://valgrind.org/docs/manual/mc-manual.html#mc-manual.

rudefn

means that memory allocated with new[] is being deallocated with free
(rather than delete / delete[]

Remarkably, this change to mzR removes this particular valgind problem

Index: src/cramp.cpp
===================================================================
--- src/cramp.cpp       (revision 121179)
+++ src/cramp.cpp       (working copy)
@@ -95,7 +95,7 @@
         //      if (m_runInfo->m_data.scanCount < 0) { // undeclared

scan

count
         // this will provoke reading of index, which sets scan count
         rampScanInfo* tmp = getScanHeaderInfo ( 1 );
-        free(tmp);
+        delete(tmp);
         // }
         // END HENRY
     }

but doesn't get us out of the woods -- valgrind now complains

xset <- xcms::xcmsSet(msmsPths)

vex: the `impossible' happened:
   isZeroU
vex storage: T total 3029292920 bytes allocated
vex storage: P total 640 bytes allocated

valgrind: the 'impossible' happened:
   LibVEX called failure_exit().

host stacktrace:
==20822==    at 0x38083F48: ??? (in /usr/lib/valgrind/memcheck-amd
64-linux)
==20822==    by 0x38084064: ??? (in /usr/lib/valgrind/memcheck-amd
64-linux)
...
sched status:
  running_tid=1

Thread 1: status = VgTs_Runnable (lwpid 20822)
==20822==    at 0x25A294E0: ??? (in /lib/x86_64-linux-gnu/libcrypt
o.so.1.0.0)
==20822==    by 0x25A086FF: EC_POINT_mul (in

/lib/x86_64-linux-gnu/libcrypt

o.so.1.0.0)
==20822==    by 0xB67335F: ???
==20822==    by 0xCF7F76F: ???
==20822==    by 0x5EB461A205EFD6FF: ???
==20822==    by 0xC221D2F: ???
==20822==    by 0x25A10E47: EC_KEY_check_key (in
/lib/x86_64-linux-gnu/libcrypto.so.1.0.0)
==20822==    by 0x25A11260: EC_KEY_set_public_key_affine_coordinates

(in

/lib/x86_64-linux-gnu/libcrypto.so.1.0.0)
==20822==    by 0x25ACA882: ??? (in /lib/x86_64-linux-gnu/libcrypt
o.so.1.0.0)
==20822==    by 0x25AC637F: ??? (in /lib/x86_64-linux-gnu/libcrypt
o.so.1.0.0)
==20822==    by 0x25AC5A33: ??? (in /lib/x86_64-linux-gnu/libcrypt
o.so.1.0.0)
==20822==    by 0x2599970C: FIPS_mode_set (in
/lib/x86_64-linux-gnu/libcrypto.so.1.0.0)
==20822==    by 0x25995F89: OPENSSL_init_library (in
/lib/x86_64-linux-gnu/libcrypto.so.1.0.0)
==20822==    by 0x40104E9: call_init.part.0 (dl-init.c:72)
==20822==    by 0x40105FA: call_init (dl-init.c:30)
==20822==    by 0x40105FA: _dl_init (dl-init.c:120)
==20822==    by 0x4015711: dl_open_worker (dl-open.c:575)
==20822==    by 0x4010393: _dl_catch_error (dl-error.c:187)
==20822==    by 0x4014BD8: _dl_open (dl-open.c:660)
==20822==    by 0x6C80F08: dlopen_doit (dlopen.c:66)
==20822==    by 0x4010393: _dl_catch_error (dl-error.c:187)
==20822==    by 0x6C81570: _dlerror_run (dlerror.c:163)
==20822==    by 0x6C80FA0: dlopen@@GLIBC_2.2.5 (dlopen.c:87)
==20822==    by 0x4EA93F0: AddDLL (Rdynload.c:537)
==20822==    by 0x4EA996B: R_moduleCdynload (Rdynload.c:917)
==20822==    by 0x4F68979: internet_Init (internet.c:79)
==20822==    by 0x4F68AF2: R_newsock (internet.c:115)
==20822==    by 0x4EF39C0: do_sockconn (connections.c:3196)
==20822==    by 0x4F3B587: bcEval (eval.c:5658)
...

which frankly is too cryptic for me -- it seems perhaps like a call
opening a socket connection is going wrong, but I really don't know if

that

is cause or effect, or even relevant to your problem.

The bottom line is that this is likely a problem in the C code of one

of

the packages that you are using.

Do you or a colleague have the expertise to work through this?

Martin

Kind regards,
Tom


https://github.com/Bioconductor-mirror/msPurity
http://bioconductor.org/checkResults/devel/bioc-LATEST/
msPurity/morelia-buildsrc.html
https://bioconductor.org/packages/devel/bioc/html/msPurity.html

Travis Mac build report:
https://travis-ci.org/Viant-Metabolomics/msPurity/jobs/159938167


############################################################
##################
############################################################
##################
###
### Running command:
###
###   /Library/Frameworks/R.framework/Versions/Current/Resources/

bin/R

CMD build --keep-empty-dirs --no-resave-data msPurity
###
############################################################
##################
############################################################
##################


* checking for file ?msPurity/DESCRIPTION? ... OK
* preparing ?msPurity?:
* checking DESCRIPTION meta-information ... OK
* installing the package to build vignettes
* creating vignettes ...sh: line 1: 62889 Segmentation fault: 11
'/Library/Frameworks/R.framework/Resources/bin/Rscript' --vanilla
--default-packages= -e "tools::buildVignettes(dir = '.', tangle =
TRUE)" > '/tmp/RtmpCXdH7K/xshellf59d5efb2aea' 2>&1
 ERROR

 *** caught segfault ***
address 0x0, cause 'memory not mapped'

Traceback:
 1: .External(list(name = "CppMethod__invoke_notvoid", address =
<pointer: 0x7fb83b714be0>,     dll = list(name = "Rcpp", path =
"/Library/Frameworks/R.framework/Versions/3.3/Resources/
library/Rcpp/libs/Rcpp.so",
        dynamicLookup = TRUE, handle = <pointer: 0x7fb83b70ed10>,
   info = <pointer: 0x1052f30c0>), numParameters = -1L),     <pointer:
0x7fb83b590710>, <pointer: 0x7fb83b5af740>, .pointer,     ...)
 2: object at backend$getPeakList(x)
 3: FUN(X[[i]], ...)
 4: lapply(X = X, FUN = FUN, ...)
 5: sapply(scans, function(x) object at backend$getPeakList(x)$peaks,
simplify = FALSE)
 6: sapply(scans, function(x) object at backend$getPeakList(x)$peaks,
simplify = FALSE)
 7: .local(object, ...)
 8: mzR::peaks(mr)
 9: mzR::peaks(mr)
10: .fun(piece, ...)
11: (function (i) {    piece <- pieces[[i]]    if (.inform) {
res <- try(.fun(piece, ...))        if (inherits(res, "try-error")) {
          piece <- paste(utils::capture.output(print(piece)),
       collapse = "\n")            stop("with piece ", i, ": \n",
piece, call. = FALSE)        }    }    else {        res <-
.fun(piece, ...)    }    progress$step()    res})(2L)
12: .Call(loop_apply_, as.integer(n), f, env)
13: loop_apply(n, do.ply)
14: llply(.data = pieces, .fun = .fun, ..., .progress = .progress,
.inform = .inform, .parallel = .parallel, .paropts = .paropts)
15: plyr::alply(files, 1, function(x) {    mr <- mzR::openMSfile(x)
scan_peaks <- mzR::peaks(mr)    return(scan_peaks)})
16: getscans(filepths)
17: purityX(xset, offsets = c(0.5, 0.5), xgroups = c(1, 2))
18: eval(expr, envir, enclos)
19: eval(expr, envir, enclos)
20: withVisible(eval(expr, envir, enclos))
21: withCallingHandlers(withVisible(eval(expr, envir, enclos)),
warning = wHandler,     error = eHandler, message = mHandler)
22: handle(ev <- withCallingHandlers(withVisible(eval(expr, envir,
enclos)), warning = wHandler, error = eHandler, message = mHandler))
23: evaluate_call(expr, parsed$src[[i]], envir = envir, enclos =
enclos,     debug = debug, last = i == length(out), use_try =
stop_on_error !=         2L, keep_warning = keep_warning, keep_message
= keep_message,     output_handler = output_handler)
24: evaluate(code, envir = env, new_device = FALSE, keep_warning =
!isFALSE(options$warning),     keep_message =
!isFALSE(options$message), stop_on_error = if (options$error &&
 options$include) 0L else 2L, output_handler =
knit_handlers(options$render,         options))
25: in_dir(opts_knit$get("root.dir") %n% input_dir(), evaluate(code,
  envir = env, new_device = FALSE, keep_warning =
!isFALSE(options$warning),     keep_message =
!isFALSE(options$message), stop_on_error = if (options$error &&
 options$include) 0L else 2L, output_handler =
knit_handlers(options$render,         options)))
26: block_exec(params)
27: call_block(x)
28: process_group.block(group)
29: process_group(group)
30: withCallingHandlers(if (tangle) process_tangle(group) else
process_group(group),     error = function(e) {        setwd(wd)
 cat(res, sep = "\n", file = output %n% "")        message("Quitting
from lines ", paste(current_lines(i),             collapse = "-"), "
(", knit_concord$get("infile"),             ") ")    })
31: process_file(text, output)
32: knitr::knit(knit_input, knit_output, envir = envir, quiet = quiet,
    encoding = encoding)
33: rmarkdown::render(file, encoding = encoding, quiet = quiet, envir
= globalenv())
34: vweave_rmarkdown(...)
35: engine$weave(file, quiet = quiet, encoding = enc)
36: doTryCatch(return(expr), name, parentenv, handler)
37: tryCatchOne(expr, names, parentenv, handlers[[1L]])
38: tryCatchList(expr, classes, parentenv, handlers)
39: 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)})
40: tools::buildVignettes(dir = ".", tangle = TRUE)
An irrecoverable exception occurred. R is aborting now ...

        [[alternative HTML version deleted]]

This email message may contain legally privileged and/or confidential
information.  If you are not the intended recipient(s), or the

employee or

agent responsible for the delivery of this message to the intended
recipient(s), you are hereby notified that any disclosure, copying,
distribution, or use of this email message is prohibited.  If you have
received this message in error, please notify the sender immediately by
e-mail and delete this email message from your computer. Thank you.

      [[alternative HTML version deleted]]

	[[alternative HTML version deleted]]