Re: msync(2) bug(?), returns AOP_WRITEPAGE_ACTIVATE to userland

["Pekka Enberg" <penberg@cs.helsinki.fi>]

Hi Hugh,

On Sat, 13 Oct 2007, Pekka Enberg wrote:
> Doesn't msync(2) get to it via mm/page-writeback.c:write_cache_pages()
> without unionfs even?

On 10/14/07, Hugh Dickins <hugh@veritas.com> wrote:
> I believe not.  Please do double-check my assertions, I've always found
> the _writepages paths rather twisty; but my belief (supported by the
> fact that we've not hit shmem_writepage's BUG_ON(page_mapped(page))
> in five years) is that tmpfs/shmem opts out of all of that with its
>         .capabilities   = BDI_CAP_NO_ACCT_DIRTY | BDI_CAP_NO_WRITEBACK,
> in shmem_backing_dev_info, which avoids all those _writepages avenues
> (e.g. through bdi_cap_writeback_dirty tests), and write_cache_pages is
> just a subfunction of the _writepages.

Thanks for the explanation, you're obviously correct.

However, I don't think the mapping_cap_writeback_dirty() check in
__filemap_fdatawrite_range() works as expected when tmpfs is a lower
mount for an unionfs mount. There's no BDI_CAP_NO_WRITEBACK capability
for unionfs mappings so do_fsync() will call write_cache_pages() that
unconditionally invokes shmem_writepage() via unionfs_writepage().
Unless, of course, there's some other unionfs magic I am missing.

                                   Pekka