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

Re: swap files cause vim 6.3 crash

Expand Messages
  • Jussi Hagman
    ... Well, this is almost identical to the crash you were experiencing before. Vim tries to show a dialog and there is a crash, intrestingly also this time not
    Message 1 of 4 , Sep 1 8:10 AM
    • 0 Attachment
      On 31.8.2005, at 19.07, Chris Fonnesbeck wrote:
      > I am using the latest 6.3 binary for OSX on Tiger, and am experiencing
      > regular crashes whenever vim encounters a file with an associated swap
      > file. The crash occurs every time. Here is the crash report:

      Well, this is almost identical to the crash you were experiencing
      before. Vim tries to show a dialog and there is a crash, intrestingly
      also this time not inside actual vim code but inside Apple's
      HIToolbox code. Maybe we are calling it wrongly but there definitely
      should not be a crash. So there is also a bug from Apple's part.

      I'll try to look a bit into this, but I probably can't do much given
      that I can't reproduce the bug :(

      Greetings,
      Jussi

      --
      Jussi Hagman, jhagman@..., iChat/AIM: jussihagman, ICQ: 54004113
      Studentbyn 4 D 33, 20540 Åbo, Finland +358 50 56 51 170
    • Stephen Riehm
      ... Hi, definitely check the way vim calls apple s API. While I was at Uni I had a similar effect with an X11 library. It took me weeks to find, but in the end
      Message 2 of 4 , Sep 1 9:57 AM
      • 0 Attachment
        On 01.09.2005, at 17:10, Jussi Hagman wrote:

        > Well, this is almost identical to the crash you were experiencing
        > before. Vim tries to show a dialog and there is a crash,
        > intrestingly also this time not inside actual vim code but inside
        > Apple's HIToolbox code. Maybe we are calling it wrongly but there
        > definitely should not be a crash. So there is also a bug from
        > Apple's part.

        Hi,

        definitely check the way vim calls apple's API.

        While I was at Uni I had a similar effect with an X11 library. It
        took me weeks to find, but in the end it was just a matter of swapped
        arguments. If I remember rightly, the API used one parameter to
        determine the structure of the other. Getting the parameters in the
        wrong order didn't cause an immediate error, but it blew up often
        enough that I knew something was going wrong. Once I swapped them
        over the problem was gone.

        I'm not suggesting that someone swapped parameters somewhere, but any
        data that gets past to an API does have to be Just Right™.
        It's tedious work, but that's where I would start, possibly using the
        checkin history to localise recent changes.

        Steve
      • Jussi Hagman
        ... I m gonna do that and search apple s carbon-dev list archives for possible similar issues ... Well, the thing that makes this hard is that I can t
        Message 3 of 4 , Sep 1 10:15 AM
        • 0 Attachment
          On 1.9.2005, at 19.57, Stephen Riehm wrote:

          >
          > On 01.09.2005, at 17:10, Jussi Hagman wrote:
          >
          >
          >> Well, this is almost identical to the crash you were experiencing
          >> before. Vim tries to show a dialog and there is a crash,
          >> intrestingly also this time not inside actual vim code but inside
          >> Apple's HIToolbox code. Maybe we are calling it wrongly but there
          >> definitely should not be a crash. So there is also a bug from
          >> Apple's part.
          >>
          >
          > Hi,
          >
          > definitely check the way vim calls apple's API.

          I'm gonna do that and search apple's carbon-dev list archives for
          possible similar issues


          > I'm not suggesting that someone swapped parameters somewhere, but
          > any data that gets past to an API does have to be Just Right™.

          Well, the thing that makes this hard is that I can't reproduce the
          problem is there anyone else than Chris seeing this problem? Any
          additional info of affected machines could help finding this bug.

          > It's tedious work, but that's where I would start, possibly using
          > the checkin history to localise recent changes.

          In vim 6.3 there aren't really recent changes. AFAIK the last change
          is from last year.

          To me it seems that something in Chris' system is making the bug
          visible. But I'll start with reading the code.

          Greetings,
          Jussi

          --
          Jussi Hagman, jhagman@..., iChat/AIM: jussihagman, ICQ: 54004113
          Studentbyn 4 D 33, 20540 Åbo, Finland +358 50 56 51 170
        Your message has been successfully submitted and would be delivered to recipients shortly.