Re: patches for PR 218, index corruption

[Milan Zamazal]

>>>>> "DB" == Dirk Bergstrom <address@hidden> writes:

    DB> *) second answer: oh, wait, i understand the problem.  mtime is
    DB> only to the second, so we could read the index, lock the db, and
    DB> check the mtime, and it would look fine, but we might miss that
    DB> someone had written a new index in the second half of the second
    DB> that we read it in.  mtime reads the same, but the index is
    DB> different.

Exactly.

    DB> i really don't want to force an index read *every* time we
    DB> access the index.  that could be a big hit.  i counted seven
    DB> getFirstPR() calls in an edit transaction,

I see.

    DB> ok, here's a solution:

    DB> put a call to getIndex() in lock_gnats(), so that after a
    DB> successful lock, we are guaranteed a clean index.  this adds one
    DB> extra index read, but it should solve the problem.

    DB> what do you think?

I think this would be a reasonable solution, thanks.

Regards,

Milan Zamazal

-- 
real programmer?  don't get me started.  if you need to hide your
pathetic excuse for a carreer behind super-macho languages like C, C++,
and/or Perl instead of writing clean, maintainable, efficient code, you
aren't much of a real programmer in my view.  -- Erik Naggum in comp.emacs