package qs had non-zero exit status #38
Comments
|
Your CPU may be too old on that computer. What is the command you are using to install? |
|
Hello,
Thank you for your response. I am using a high performance cluster through
my university, so I am not sure how old the CPU is.
I have tried both install.packages("qs") and remotes::install_cran("qs",
type="source", configure.args="--with-simd=AVX2").
I get a lot more errors with the remotes::install_cran command.
Thanks,
Briana
…On Thu, May 14, 2020 at 1:10 PM Travers ***@***.***> wrote:
*Message sent from a system outside of UConn.*
Your CPU may be too old on that computer.
What is the command you are using to install?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#38 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/APLP6HBNU5VKBDLZ7YGLLDLRRQQW3ANCNFSM4NAYTGSA>
.
--
Briana Hennessy, Ph.D. Student
Research Methods, Measurement and Evaluation
Department of Educational Psychology
Neag School of Education
University of Connecticut
(860) 680-7469
|
|
The So I suspect that there is a misconfiguration in the compiler. Unfortunately, there's no way I can reproduce the error to investigate. Have you had any trouble installing other packages? |
|
No I haven't had any trouble with other packages.
I did initially have this issue on my old version of R before I updated to
R 4.0. I will see if I can reproduce it with an older version of R. I will
also pass your suggestion on to our tech people because maybe they will
know something about the misconfiguration in the compiler.
…On Thu, May 14, 2020 at 1:26 PM Travers ***@***.***> wrote:
*Message sent from a system outside of UConn.*
The shrx instruction call coming from the ZSTD compression library was an
instruction introduced in CPUs in 2014, but even still there is nothing in
the code that calls that instruction directly.
So I suspect that there is a misconfiguration in the compiler.
Unfortunately, there's no way I can reproduce the error to investigate.
Have you had any trouble installing other packages?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#38 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/APLP6HFIWESG5RQPSPDEAYLRRQSVJANCNFSM4NAYTGSA>
.
--
Briana Hennessy, Ph.D. Student
Research Methods, Measurement and Evaluation
Department of Educational Psychology
Neag School of Education
University of Connecticut
(860) 680-7469
|
|
More information; I was able to install qs with an older version of R, so
that was not the issue. When I load R on the HPC (where I am having the
issue) I also have to load gcc/9.2.0. I'm sorry I don't understand all of
this, but could that be the issue? Does it need to be a more recent version
of gcc?
Thanks,
Briana
On Thu, May 14, 2020 at 1:31 PM Briana Hennessy <briana.hennessy@uconn.edu>
wrote:
… No I haven't had any trouble with other packages.
I did initially have this issue on my old version of R before I updated to
R 4.0. I will see if I can reproduce it with an older version of R. I will
also pass your suggestion on to our tech people because maybe they will
know something about the misconfiguration in the compiler.
On Thu, May 14, 2020 at 1:26 PM Travers ***@***.***> wrote:
> *Message sent from a system outside of UConn.*
>
> The shrx instruction call coming from the ZSTD compression library was
> an instruction introduced in CPUs in 2014, but even still there is nothing
> in the code that calls that instruction directly.
>
> So I suspect that there is a misconfiguration in the compiler.
> Unfortunately, there's no way I can reproduce the error to investigate.
> Have you had any trouble installing other packages?
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <#38 (comment)>, or
> unsubscribe
> <https://github.com/notifications/unsubscribe-auth/APLP6HFIWESG5RQPSPDEAYLRRQSVJANCNFSM4NAYTGSA>
> .
>
--
Briana Hennessy, Ph.D. Student
Research Methods, Measurement and Evaluation
Department of Educational Psychology
Neag School of Education
University of Connecticut
(860) 680-7469
--
Briana Hennessy, Ph.D. Student
Research Methods, Measurement and Evaluation
Department of Educational Psychology
Neag School of Education
University of Connecticut
(860) 680-7469
|
|
The gcc version is fine, but there seems to be a mismatch between the CPU instructions gcc thinks is available and what's actually available. Because it's a cluster, it gets even more complicated (e.g., if you install on the head node CPU architecture is different than the rest of the nodes). I'm not sure if I can properly diagnose here, so I would have to suggest using some other approach to save data, as this could potentially waste a lot of your time. Sorry :( |
|
Hello @traversc, I seem to also have similar issues, but with a different set of error messages (see below). I'm sorry if this is redundant for you. Its a bit overwhelming to see all these errors, and I'm not sure where to begin. I am on an older R version 3.6, using centos6, and compiler gcc 7.2.0.
installing source package ‘qs’ ...
|
|
@ytakemon try it without the configure args: |
|
As an update - I did eventually get everything running when I only used the most up-to-date architecture on the cluster so you were correct in my case that the CPU was too old. |
|
Hi @traversc , I did try your suggestion first, but I get similar/ if note the same set of errors. I am trying to install qs in a server with 40 cores and doesn't have a job scheduler. Essentially anyone working here would be on the "head node". Would this complicate the installation process?
The downloaded source packages are in |
|
@ytakemon: I'm not sure, but the compiler is generating instructions not supported by your CPU. I don't know what the solution is other than running it on a newer CPU. |
|
@traersc -Do you have an idea what the minimum CPU requirements might be? Using |
|
A cpu from around ~2015 or later should have no problem(Yours is from 2014). However, as I mentioned to Briana, I believe this to be a server configuration issue. See below for similar issues in other places where you might find a solution: fstpackage/fst#167 |
|
@traversc I see - I'll look into your suggestions. Thanks for your help! |
* installing *source* package ‘stringfish’ ...
** package ‘stringfish’ successfully unpacked and MD5 sums checked
** using staged installation
checking for pkg-config... /usr/bin/pkg-config
stringfish configure script
PCRE2 10.43 library detected -- skipping PCRE2 compilation
-lpcre2-8
configure: creating ./config.status
config.status: creating src/Makevars
** libs
using C++ compiler: ‘g++ (GCC) 14.1.1 20240507’
g++ -std=gnu++17 -I"/usr/include/R/" -DNDEBUG -DRCPP_USE_UNWIND_PROTECT
-DRCPP_NO_RTTI -DPCRE2_CODE_UNIT_WIDTH=8 -DHAVE_CONFIG_H -I. -I'/home
/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include' -I'/home/pret
ender/R/x86_64-pc-linux-gnu-library/4.3/RcppParallel/include' -I/usr/loc
al/include -fpic -march=x86-64 -mtune=generic -O2 -pipe -fno-plt -fe
xceptions -Wp,-D_FORTIFY_SOURCE=2 -Wformat -Werror=format-securi
ty -fstack-clash-protection -fcf-protection -Wp,-D_GLIBCXX_ASSER
TIONS -g -ffile-prefix-map=/build/r/src=/usr/src/debug/r -flto=auto -ffa
t-lto-objects -c RcppExports.cpp -o RcppExports.o
In file included from /home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp
/routines.h:26,
from /home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp
Common.h:128,
from /home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp
.h:27,
from RcppExports.cpp:4:
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/iostream/Rstreambuf.h
: In member function ‘std::streamsize Rcpp::Rstreambuf<OUTPUT>::xsputn(const char*, std::
streamsize) [with bool OUTPUT = true; std::streamsize = long int]’:
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/iostream/Rstreambuf.h
:53:20: warning: field precision specifier ‘.*’ expects argument of type ‘int’, but argum
ent 2 has type ‘std::streamsize’ {aka ‘long int’} [-Wformat=]
53 | Rprintf("%.*s", num, s);
| ~~^~ ~~~
| | |
| int std::streamsize {aka long int}
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/iostream/Rstreambuf.h
: In member function ‘std::streamsize Rcpp::Rstreambuf<OUTPUT>::xsputn(const char*, std::
streamsize) [with bool OUTPUT = false; std::streamsize = long int]’:
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/iostream/Rstreambuf.h
:57:21: warning: field precision specifier ‘.*’ expects argument of type ‘int’, but argum
ent 2 has type ‘std::streamsize’ {aka ‘long int’} [-Wformat=]
57 | REprintf("%.*s", num, s);
| ~~^~ ~~~
| | |
| int std::streamsize {aka long int}
In file included from /home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp
Common.h:181:
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/print.h: In function
‘void Rcpp::warningcall(SEXP, const std::string&)’:
/home/pretender/R/x86_64-pc-linux-gnu-library/4.3/Rcpp/include/Rcpp/print.h:30:19: error:
format not a string literal and no format arguments [-Werror=format-security]
30 | Rf_warningcall(call, s.c_str());
| ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~
cc1plus: some warnings being treated as errors
make: *** [/usr/lib64/R/etc/Makeconf:204: RcppExports.o] Error 1
ERROR: compilation failed for package ‘stringfish’
* removing ‘/home/pretender/R/x86_64-pc-linux-gnu-library/4.4/stringfish’
[1] ".Rprofile loaded successfully"
ERROR: dependency ‘stringfish’ is not available for package ‘qs’
* removing ‘/home/pretender/R/x86_64-pc-linux-gnu-library/4.4/qs’
The downloaded source packages are in
‘/tmp/RtmpvFo91V/downloaded_packages’
Warning messages:
1: In i.p(...) :
installation of package ‘stringfish’ had non-zero exit status
2: In i.p(...) : installation of package ‘qs’ had non-zero exit status
>
I seem to be having the same issue as well, I have a AMD Ryzen 7 5700U so I'm pretty sure its not a cpu issue. Any idea what this issue could be? |
Hello,
I am trying to install the package "qs" using r/3.6.1 and I keep getting the error "installation of package 'qs' had non-zero exit status'. Prior to that error I see these messages:
/tmp/ccwSjoWV.s: Assembler messages:
/tmp/ccwSjoWV.s:1106: Error: no such instruction:
shlx %r13,%rax,%rax' /tmp/ccwSjoWV.s:1147: Error: no such instruction:shlx %rcx,%r13,%rcx'/tmp/ccwSjoWV.s:1170: Error: no such instruction:
shlx %r12,%rcx,%r12' /tmp/ccwSjoWV.s:1185: Error: no such instruction:shlx %rbx,%rcx,%rcx'/tmp/ccwSjoWV.s:1288: Error: no such instruction:
shrx %rdx,%rax,%rdx' /tmp/ccwSjoWV.s:1544: Error: no such instruction:shlx %rsi,%rdx,%rdx'/tmp/ccwSjoWV.s:1628: Error: no such instruction:
shrx %rax,%rdx,%rdx' /tmp/ccwSjoWV.s:1689: Error: no such instruction:shlx %rsi,%rcx,%rdx'make: *** [ZSTD/compress/huf_compress.o] Error 1
ERROR: compilation failed for package ‘qs’
the package installs correctly on another machine that uses R/4.0. Does this package only work on the updated version of R?
Thank you,
Briana
The text was updated successfully, but these errors were encountered: