Re: [FSAIO][PATCH 7/8] Filesystem AIO read

From: Christoph Hellwig <hch@infradead.org>
To: Suparna Bhattacharya <suparna@in.ibm.com>
Cc: linux-aio@kvack.org, akpm@osdl.org, drepper@redhat.com,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
	jakub@redhat.com, mingo@elte.hu
Subject: Re: [FSAIO][PATCH 7/8] Filesystem AIO read
Date: Thu, 28 Dec 2006 11:57:47 +0000	[thread overview]
Message-ID: <20061228115747.GB25644@infradead.org> (raw)
In-Reply-To: <20061228084252.GG6971@in.ibm.com>

> +	if (in_aio()) {
> +		/* Avoid repeat readahead */
> +		if (kiocbTryRestart(io_wait_to_kiocb(current->io_wait)))
> +			next_index = last_index;
> +	}

Every place we use kiocbTryRestart in this and the next patch it's in
this from, so we should add a little helper for it:

int aio_try_restart(void)
{
	struct wait_queue_head_t *wq = current->io_wait;

	if (!is_sync_wait(wq) && kiocbTryRestart(io_wait_to_kiocb(wq)))
		return 1;
	return 0;
}

with a big kerneldoc comment explaining this idiom (and possible a better
name for the function ;-))

> +
> +		if ((error = __lock_page(page, current->io_wait))) {
> +			goto readpage_error;
> +		}

This should  be

		error = __lock_page(page, current->io_wait);
		if (error)
			goto readpage_error;

Pluse possible naming updates discussed in the last mail.  Also do we
really need to pass current->io_wait here?  Isn't the waitqueue in
the kiocb always guaranteed to be the same?  Now that all pagecache
I/O goes through the ->aio_read/->aio_write routines I'd prefer to
get rid of the task_struct field cludges and pass all this around in
the kiocb.  