Re: ll_rw_block mark_buffer_clean without the big kernel lock

Linus Torvalds (torvalds@transmeta.com)
Wed, 23 Jun 1999 10:40:53 -0700 (PDT)

On Wed, 23 Jun 1999, Andrea Arcangeli wrote:
>
> ll_rw_block will refile the buffer, and currently it can do that without
> the big kernel lock held (at least in write-partial). I think we can't put
> the lock_kernel just around mark_buffer_clean (in a
> async_mark_buffer_clean style) because we would end waiting for the big
> kernel lock but with the io_request_lock just acquired while other part of
> the kernel gets the io_request_lock with the lock_kernel acquired.

Agreed. However, I think the easier solution is to move the
"mark_buffer_clean()" code into the early write case (search for the
comment "Hmmph! Nothing to write" ;)

> I am not sure if we can make the buffer clean outside the io_request_lock
> (looks not to be sure that the req->bh won't change under us).

We can. We have to do it _before_ rather than later, though - just in case
we race with somebody else marking it dirty (it's ok to do the IO twice,
but it's not ok to not do it at all).

> It also fixes a wrong param to async_mark_buffer_dirty.

Oops, right you are, I was going to change the rules but ended up
reverting those changes (except for this one case I missed). Thanks,

Linus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/