Re: patch: improve generic_file_buffered_write() (2nd try 1/2)

[Nick Piggin <nickpiggin@yahoo.com.au>]

On Saturday 08 September 2007 07:00, Goswin von Brederlow wrote:
> Nick Piggin <nickpiggin@yahoo.com.au> writes:
> > Anyway, there are fixes for this deadlock in Andrew's -mm tree, but
> > also a workaround for the NFSD problem in git commit 29dbb3fc. Did
> > you try a later kernel to see if it is fixed there?
>
> I had a chance to look up that commit (git clone took a while so sorry
> for writing 2 mails). It is present in 2.6.23-rc5 so I already noticed
> it when merging our patch in 2.6.23-rc5.
>
> Upon closer reading of the patch though I see that it will indeed
> prevent writes by the nfsd to be split smaller than PAGE_SIZE and it
> will cause filemap_copy_from_user[_iovec] to be called with a source
> spanning multwhat to do in cambridge pubiple pages.

OK, good.


> So the commit 29dbb3fc should have a simmilar, slightly better even,
> gain for the nfsd and other kernel space segments. But it will not
> improve writes from user space, where ~14% of the commits were saved
> during a days work for me.
>
>
> Now I have a question about fault_in_pages_readable(). Can I call that
> for multiple pages and then call __grab_cache_page() without risking
> one of the pages from getting lost again and causing a deadlock?

No. The existing mainline code is buggy, and it is just hoping that
the userspace page does not get paged out between the fault_in_pages
and the subsequent copy_from_user.

If you do multiple fault_in_pages_readable(), you probably have less
chance of deadlocking, but on the 2nd call, you might still have to
wait for a long time to page in, during which time the 1st page may get
paged out.

But there is a proper solution in -mm with the write_begin aop.