[lustre-devel] [PATCH v3 0/2] iov_iter: allow iov_iter_get_pages_alloc to allocate more pages per call
hch at infradead.org
Fri Feb 3 03:09:30 PST 2017
On Fri, Feb 03, 2017 at 08:54:15AM +0000, Al Viro wrote:
> Hmm... Reuse part is really nasty ;-/ OTOH, it might make sense to have
> a "fill bio_vec array" as separate primitive - having that sucker come
> from bio looks like an artificial restriction.
Or just the only usecase :) But yes, it could be generalized to take a
bio_vec without too much effort.
> OK, next question, seeing that you've dealt with O_DIRECT guts more than
> I have. When we have iov_iter_get_pages() fail on do_direct_IO() write
> with some blocks already allocated, we pick zero page as data source.
> So far, so good, but:
> * should we bother zeroing anything unless buffer_new() is true?
I don't think so.
> * why, in case of more than a page worth of pending allocated
> blocks, do we bother with calling iov_iter_get_pages() again and again?
> We *do* take care not to allocate anything else after that point, but
> dio_get_page() will be calling iov_iter_get_pages() every time in that
> case - there's only one page in queue.
There shouldn't be a need for it.
But take it with a grain of salt - fs/direct-io.c is a hairy mess,
that's one of the reasons why I replaced it with the new iomap
code instead of trying to gradually move it to iomap API.
More information about the lustre-devel