On Thu, Oct 14, 2010 at 09:49:31AM -0500, Alex Elder wrote:
> Dave's recent "dynamic speculative EOF preallocation" changes
> caused test 008 to no longer produce the kinds of holes it was
> expecting to see, so it got me looking at the program it uses,
> Since 008 is the only test using this program, it probably
> doesn't deserve the attention I gave it yesterday, but I've
> done the work so I might as well submit the series for
> review and get my changes in.
> The first patch in the series actually addresses two things
> that are bugs.
> The remaining 10 are just cleanup and refactoring changes.
> Other than two bug fix changes and one change in output
> when running in very-verbose mode, the patched code should
> be pretty much functionally identical to before.
I haven't really looked at the whole series closely, but nothing
stands out as a red flag. Seeing it's test code, it's probably
OK to go as is. The bug fixes look fine.