[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Sundials is way outdated



(Apologies for what is probably an HTML email - phone...)
> On 27 Jan 2017, at 21:24, Andreas Tille <andreas@an3as.eu> wrote:
> 
> Hi folks,
> 
> I just realised that our packaged sundials is lagging two releases
> behind upstream.  Its a bit bad to learn this right now - some days
> after changes can make it into Stretch. :-(
> 
> I think we should act anyway and thus we should act on this.  I tried
> `git svn clone` to migrate the packaging to Git but when trying to push
> I realised that Dima Kogan just had create a Git repository starting at
> the latest uploaded Debian revision.  While that's not optimal since we
> loose the history of SVN I decided that its better to remove SVN anyway
> to prevent others from uselessly doing the same work as I did.  Dima,
> it might be you have communicated what you did but it was obviously not
> heard and acted upon.  Please make more noise next time.
> 
> I kept on working with Dima's Git repository and tried to bring
> packaging up to date (DEP5 copyright with Files-Excluded by removing
> PDFs without source as the old package was done).
> 
> So while we somehow have a packaging attempt which can be uploaded to
> experimental (hey, did I said that its quite unfortunate that we act
> after freeze?) the package does not build:
> 
> ...
> /usr/bin/cmake -E cmake_link_script CMakeFiles/cmTC_798c0.dir/link.txt --verbose=1
> /usr/bin/cc  -g -O2 -fdebug-prefix-map=/build/sundials-2.7.0+dfsg=. -fstack-protector-strong -Wformat -Werror=format-security -Wdate-time -D_FORTIFY_SOURCE=2 -DCHECK_FUNCTION_EXISTS=pthread_create  -Wl,-z,relro  CMakeFiles/cmTC_798c0.dir/CheckFunctionExists.c.o  -o cmTC_798c0 -rdynamic -lpthreads 
> /usr/bin/ld: cannot find -lpthreads
> collect2: error: ld returned 1 exit status
> CMakeFiles/cmTC_798c0.dir/build.make:97: recipe for target 'cmTC_798c0' failed
> ...
> 
> I remember this kind of error but forgot the solution.
> 
> It would be cool if somebody could fix this.  I'd upload to experimental
> and keep on testing with my originl target libsbml-odesolver.

-lpthread, not -lpthreads, or better yet -pthread.

Regards,
James

Reply to: