Loading ...
Sorry, an error occurred while loading the content.
 

Re: runaway vim processes

Expand Messages
  • Daniel Elstner
    ... ;-) Please do test it if possible, it can t go into Vim until it has been confirmed that it works. Quick n dirty build instructions in case you never
    Message 1 of 16 , Mar 1 5:52 PM
      On Son, 2003-03-02 at 02:23, msoulier@... wrote:
      > Quoting Daniel Elstner <daniel.elstner@...>:
      >
      > > Just to be sure -- did you test the patch and verified that it works for
      > > you, or was this just a generic expression of joy about the rapid
      > > customer service in the Open Source world? :-)
      >
      > I haven't tried it yet, I just love the response time. :)

      ;-)

      Please do test it if possible, it can't go into Vim until it has been
      confirmed that it works. Quick'n'dirty build instructions in case you
      never built Vim from source before:

      1. get Vim + all patches (I think the easiest way is via CVS)
      2. apply the patch: patch -p0 < patchfile
      3. ./configure --with-features=big --enable-pythoninterp
      4. make
      5. no need to install, just run src/vim
      6. close the terminal window and pray :)

      Any other people on this list willing to give the patch a try?

      --Daniel
    • Sean Richards
      ... Vim 6.1 (patches 1-290) on Slackware 8.1. Patched successfully but still the same old problem appears :( Sean --
      Message 2 of 16 , Mar 2 12:54 AM
        Daniel Elstner wrote:

        > Any other people on this list willing to give the patch a try?

        Vim 6.1 (patches 1-290) on Slackware 8.1.
        Patched successfully but still the same old problem appears :(

        Sean

        --
        +---------------------------------------------------------------+
        | All spelling errors are intentional and are there to show new |
        | and improved ways of spelling old words. |
        +---------------------------------------------------------------+
      • Daniel Elstner
        ... Darn. Oh well. Could you please mail me the ./configure arguments you used, the output of ./configure, the arguments used for compiling and linking
        Message 3 of 16 , Mar 2 6:14 AM
          On Son, 2003-03-02 at 09:54, Sean Richards wrote:
          > Daniel Elstner wrote:
          >
          > > Any other people on this list willing to give the patch a try?
          >
          > Vim 6.1 (patches 1-290) on Slackware 8.1.
          > Patched successfully but still the same old problem appears :(

          Darn. Oh well. Could you please mail me the ./configure arguments you
          used, the output of ./configure, the arguments used for compiling and
          linking (:version tells you), plus the output of 'ldd src/vim'? Hmm
          that's a lot, but nonetheless it'd be really helpful :)

          Thanks in advance,
          --Daniel
        • Daniel Elstner
          Hi, thanks a lot for the feedback. It turned out to be indeed helpful, see below. ... (It s probably not the issue but nonetheless a good idea to rm
          Message 4 of 16 , Mar 3 1:02 AM
            Hi,

            thanks a lot for the feedback. It turned out to be indeed helpful, see
            below.

            On Mon, 2003-03-03 at 09:55, Sean Richards wrote:
            > Daniel Elstner wrote:
            > > Darn. Oh well. Could you please mail me the ./configure arguments you
            > > used, the output of ./configure, the arguments used for compiling and
            > > linking (:version tells you), plus the output of 'ldd src/vim'? Hmm
            > > that's a lot, but nonetheless it'd be really helpful :)
            >
            > As requested
            >
            > [1] ./configure arguments
            >
            > #!/bin/sh
            >
            > ./configure --enable-gui=gtk --with-features=huge --with-compiledby=sean@darkstar --enable-pythoninterp --enable-rubyinterp --enable-cscope
            >
            > [2] output from configure
            >
            > loading cache auto/config.cache

            (It's probably not the issue but nonetheless a good idea to
            rm src/auto/config.cache before reconfiguring. Just in case.)

            [...]

            > [3] output from :version
            >
            > :version
            > VIM - Vi IMproved 6.1 (2002 Mar 24, compiled Mar 3 2003 21:28:11)
            > Included patches: 1-290

            [...]

            > Compilation: gcc -c -I. -Iproto -DHAVE_CONFIG_H -DFEAT_GUI_GTK -I/usr/include/gtk-1.2
            > -I/usr/include/glib-1.2 -I/usr/lib/glib/include -I/usr/X11R6/include -g -O2 -I/usr/X
            > 11R6/include -I/usr/include/python2.2 -pthread -I/usr/local/lib/ruby/1.6/i686-linux

            That's fine, -pthread is there.

            > Linking: gcc -L/usr/X11R6/lib -rdynamic -L/usr/local/lib -o vim -L/usr/lib -L/usr/
            > X11R6/lib -lgtk -lgdk -rdynamic -lgmodule -lglib -ldl -lXext -lXt -lncurses -lgpm -ld
            > l -L/usr/lib/python2.2/config -lpython2.2 -ldl -lpthread -lutil -Xlinker -export-dyna
            > mic /usr/local/lib/ruby/1.6/i686-linux/libruby.a -ldl -lcrypt -lm

            There's the bugger: -lpthread should not be there. Do

            rm src/vim
            rm src/src/auto/link.sed

            and try again. link.sh should print messages like "Trying to remove the
            pthread library..." and either "we don't need the pthread library!" or
            "we DO need the pthread library". Hopefully the first.

            With some luck you just had a stale link.sed in the tree, otherwise this
            will need further investigation. Please mail me your results in either
            case :)

            Thanks,
            --Daniel
          • Sean Richards
            ... I had already done a make clean so that should have been taken care of. ... I am sad to report that the output from link.sh includes the lines ...
            Message 5 of 16 , Mar 3 1:35 AM
              Daniel Elstner wrote:
              > Hi,
              >
              > thanks a lot for the feedback. It turned out to be indeed helpful, see
              > below.

              > > [3] output from :version
              > >
              > > :version
              > > VIM - Vi IMproved 6.1 (2002 Mar 24, compiled Mar 3 2003 21:28:11)
              > > Included patches: 1-290
              >
              > [...]
              >
              > > Compilation: gcc -c -I. -Iproto -DHAVE_CONFIG_H -DFEAT_GUI_GTK -I/usr/include/gtk-1.2
              > > -I/usr/include/glib-1.2 -I/usr/lib/glib/include -I/usr/X11R6/include -g -O2 -I/usr/X
              > > 11R6/include -I/usr/include/python2.2 -pthread -I/usr/local/lib/ruby/1.6/i686-linux
              >
              > That's fine, -pthread is there.
              >
              > > Linking: gcc -L/usr/X11R6/lib -rdynamic -L/usr/local/lib -o vim -L/usr/lib -L/usr/
              > > X11R6/lib -lgtk -lgdk -rdynamic -lgmodule -lglib -ldl -lXext -lXt -lncurses -lgpm -ld
              > > l -L/usr/lib/python2.2/config -lpython2.2 -ldl -lpthread -lutil -Xlinker -export-dyna
              > > mic /usr/local/lib/ruby/1.6/i686-linux/libruby.a -ldl -lcrypt -lm
              >
              > There's the bugger: -lpthread should not be there. Do
              >
              > rm src/vim
              > rm src/src/auto/link.sed

              I had already done a 'make clean' so that should have been taken care
              of.

              > and try again. link.sh should print messages like "Trying to remove the
              > pthread library..." and either "we don't need the pthread library!" or
              > "we DO need the pthread library". Hopefully the first.

              I am sad to report that the output from link.sh includes the lines ...

              link.sh: Trying to remove the pthread library...
              link.sh: We DO need the pthread library.

              > With some luck you just had a stale link.sed in the tree, otherwise this
              > will need further investigation. Please mail me your results in either
              > case :)

              Cheers, Sean :)

              --
              +---------------------------------------------------------------+
              | All spelling errors are intentional and are there to show new |
              | and improved ways of spelling old words. |
              +---------------------------------------------------------------+
            • Daniel Elstner
              Hey, I found some interesting information on the issue. This Debian bug report explains what s going on:
              Message 6 of 16 , Mar 3 8:01 AM
                Hey,

                I found some interesting information on the issue. This Debian bug
                report explains what's going on:

                http://lists.debian.org/debian-glibc/2002/debian-glibc-200212/msg00347.html

                As mentioned in the mail, the problem with coroutines also applies to
                sigaltstack(). There seems to be only one way around the problem:
                simply don't use sigaltstack().

                The attached patch disables the alternative stack if compiling on Linux
                with pthreads, except for SIGSEGV. This is AFAIK the only signal for
                which the alternative stack is really necessary. Thus there shouldn't
                be a regression in functionality when switching to fixed-up pthreads
                some day.

                Regards,
                --Daniel
              Your message has been successfully submitted and would be delivered to recipients shortly.