[R] LINUX SuSE15.4 (GNU & Intel) compiler problem in "configure" file

[Jannie Pretorius]

 Dear Dr Krylov,

 From our side, a sincere appreciation for the pointers you gave me.

a)  Apologies from my end:
        Our: */bin/sh* is a symlink to:
*sh -> /usr/bin/sh *

b) I further found that the specific HEADER files under:
       /usr/src/*linux-5.14.21-150400.2*2/include/linux
    are important (specific for some of the internal R-CRAN Compilations)
    These of course relate to the specific LINUX kernel revision:
*5.14.21-150400.22*

c) SuSe SLES 15.3/15.4 has the default environment PATH statement
            (under <*su*> status) as:
         /usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
      This order must be retained and is important.

d) And the [configure] script ran just fine (attached here)

e) You were correct, the host declaration was not sufficient:
         Now points to: x86_64-pc-linux

f) In some cases found embedded characters disrupting code sequences.
   This is typical of preparing and developing code across a local area
network,
       in our case found not to adhere to proper protocol support (under
traffic load).
   In which case, I normally strip code sections into a spreadsheet
workpage
       (widen one column) and store it as: CSV (MS-DOS) (*.csv)  --> SAVE
and open it again.
   In several instances found the character: *Ÿ*  embedded (not visible
under any STD text editors)

Very best Wishes and many, many thanks
   Jan Pretorius (UPta)


On Thu, 29 Jun 2023 at 16:45, Ivan Krylov <krylov.r00t using gmail.com> wrote:

> В Thu, 29 Jun 2023 16:17:50 +0500
> Jannie Pretorius <jannie.pretorius using up.ac.za> пишет:
>
> >    Our: */bin/sh* is a symlink to: ls -> /usr/bin/ls
>
> That sounds strange. Are you sure that /bin/sh (and not /bin/ls) points
> to /usr/bin/ls?
>
> >    The variables Rshlibpath_var and shlibparth_var are indeed empty.
>
> That's the reason why .../configure originally failed. It was asked to
> substitute a variable (which would normally look like
> ${LD_LIBRARY_PATH}), but with these variables empty the substitution
> looked like ${}, which is an error.
>
> >  The:  .../tools/config.guess file is attached
>
> Sorry, when I asked for tools/config.guess, I meant its output (what it
> prints on the console when run), not the contents.
>
> >  The:  ../config.log  file is attached
>
> config.log is indeed useful. It also must contain the output of
> tools/config.guess and tools/config.sub. I think that problem here is
> that for some reason it doesn't say "linux":
>
> > configure:4221: checking host system type
> > configure:4235: result: x86_64-pc-none
>
> Can you confirm that tools/config.guess prints x86_64-pc-linux-gnu but
> tools/config.sub $(tools/config.guess) prints x86_64-pc-none?
>
> When other people install other ./configure-based software on this
> computer, do they encounter similar problems?
>
> (It may be worth keeping the R-help mailing list in the Cc: header of
> your replies, because other people with more knowledge of GNU autotools
> may step in.)
>
> --
> Best regards,
> Ivan
>

-- 
This message and attachments are subject to a disclaimer.

Please refer to 
http://upnet.up.ac.za/services/it/documentation/docs/004167.pdf 
<http://upnet.up.ac.za/services/it/documentation/docs/004167.pdf> for
full 
details.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: COMPILE-script R-4.3.1.pdf
Type: application/pdf
Size: 143049 bytes
Desc: not available
URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20230704/462709c0/attachment.pdf>