Re: [RFC][PATCH] Fix for latent bug in vmtruncate()

[William Lee Irwin III]

On Tue, May 13, 2003 at 01:58:07PM -0700, Paul E. McKenney wrote:
> The vmtruncate() function shifts down by PAGE_CACHE_SHIFT, then
> calls vmtruncate_list(), which deals in terms of PAGE_SHIFT
> instead.  Currently, no harm done, since PAGE_CACHE_SHIFT and
> PAGE_SHIFT are identical.  Some day they might not be, hence
> this patch.
> I also took the liberty of modifying a hand-coded "if" that
> seems to optimize for files that are not mapped to instead
> use unlikely().

pgoff describes a file offset in the same units used to map files
with (the size of an area covered by a PTE), which is PAGE_SIZE (in
mainline; elsewhere it's called MMUPAGE_SIZE and I had to fix this
already for my tree). When they differ this would lose the offset into
the PAGE_CACHE_SIZE-sized file page; hence, well-spotted.

-- wli