[OpenAFS-devel] Cache inconsistency in client 1.4.8 and above

[Derrick Brashear]

On Mon, Apr 20, 2009 at 4:25 AM, Felix Frank <Felix.Frank@desy.de> wrote:
> Last week, I posted a patch to RT #124627, only to notice this morning that
> I'd been testing with a large-ish cache.
> Also, there appears to be some regression in my own fiddlings, as #124627
> isn't
> solved at all. For what it's worth, the deadlock as reported in #120491 is
> apparently prevented, although mmap_test still gets blocked in some I/O-wait
> state.

The deadlock in 120491 only happens with a cache nearer the size of
the mmap'd file. I suspect a cache smaller simply won't work.

> Cache consistency with mmap'ed writes can apparently only be achieved when
> no afs_linux_writepage_sync's are aborted at all.
>
> I think I still haven't truly understood the exact idea of the antirecursion
> patch. To me, it seems to omit writing some precious data.

We had a backtrace which showed we were calling back into ourselves.
The sole goal was to prevent that. Anything beyond was incidental and
uneeded.