On Thu, 2005-12-01 at 10:48 -0500, Charles P. Wright wrote:
> On Thu, 2005-12-01 at 00:39 -0500, Shaya Potter wrote:
> > I think without mmap support, this code is fairly useless if you run
> > programs, as it can't track usage by mmap.
> I disagree.  It may not prevent you from removing a branch that hosts a
> currently executing file, but it still prevents you from removing a
> branch that has a regularly opened file (which is going to be more
> common).

well, if you're running real applications, mmaped files are very
common.  

But my email was more of a rant, as my past few days have been suffering
on trying to figure out why I get a null pointer exception in
branchput_gen().  Thought I killed it (Which lead me to suffer from an
invalid page fault exception), but then I hit it again last night.

want to import all your changes into my changes, but need to make sure
my tests run first.  the bugs exists in the old code base, but for some
reason aren't hit.

<sigh>

_______________________________________________
unionfs mailing list
[email protected]
http://www.fsl.cs.sunysb.edu/mailman/listinfo/unionfs

Reply via email to