All of lore.kernel.org
 help / color / mirror / Atom feed
From: Andrew Morton <akpm@linux-foundation.org>
To: Tao Guo <glorioustao@gmail.com>
Cc: linux-kernel@vger.kernel.org, neilb@suse.de, axboe@fusionio.com,
	Tao Guo <Tao.Guo@emc.com>, Jens Axboe <axboe@kernel.dk>
Subject: Re: [PATCH] umem: fix up unplugging.
Date: Wed, 4 Apr 2012 14:46:56 -0700	[thread overview]
Message-ID: <20120404144656.7b2a06ec.akpm@linux-foundation.org> (raw)
In-Reply-To: <1325254651-15379-1-git-send-email-glorioustao@gmail.com>

On Fri, 30 Dec 2011 09:17:31 -0500

(Long delay...)

Tao Guo <glorioustao@gmail.com> wrote:

> We need to implement unplugging to make umem start to work, or I/O
> will never be triggered.
> 

I take it from this that umem simply doesn't work at all?

If so, for how long has this been the case and should we just delete the
thing?

> 
> --- a/drivers/block/umem.c
> +++ b/drivers/block/umem.c
> @@ -111,7 +111,8 @@ struct cardinfo {
>  	int		current_idx;
>  	sector_t	current_sector;
>  
> -	struct request_queue *queue;
> +	struct request_queue	*queue;
> +	struct blk_plug_cb	plug_cb;
>  
>  	struct mm_page {
>  		dma_addr_t		page_dma;
> @@ -513,6 +514,33 @@ static void process_page(unsigned long data)
>  	}
>  }
>  
> +static void mm_unplug(struct blk_plug_cb *cb)
> +{
> +	struct cardinfo *card = container_of(cb, struct cardinfo, plug_cb);
> +
> +	spin_lock_irq(&card->lock);
> +	activate(card);
> +	spin_unlock_irq(&card->lock);
> +}
> +
> +int mm_check_plugged(struct cardinfo *card)
> +{
> +	struct blk_plug *plug = current->plug;
> +	struct cardinfo *c;
> +
> +	if (!plug)
> +		return 0;
> +
> +	list_for_each_entry(c, &plug->cb_list, plug_cb.list) {
> +		if (c == card)
> +			return 1;
> +	}
> +
> +	/* set up unplug callback */
> +	list_add(&card->plug_cb.list, &plug->cb_list);
> +	return 1;
> +}
> +
>  static void mm_make_request(struct request_queue *q, struct bio *bio)
>  {
>  	struct cardinfo *card = q->queuedata;
> @@ -523,6 +551,8 @@ static void mm_make_request(struct request_queue *q, struct bio *bio)
>  	*card->biotail = bio;
>  	bio->bi_next = NULL;
>  	card->biotail = &bio->bi_next;
> +	if (bio->bi_rw & REQ_SYNC || !mm_check_plugged(card))
> +		activate(card);
>  	spin_unlock_irq(&card->lock);
>  
>  	return;
> @@ -884,6 +914,7 @@ static int __devinit mm_pci_probe(struct pci_dev *dev,
>  	blk_queue_make_request(card->queue, mm_make_request);
>  	card->queue->queue_lock = &card->lock;
>  	card->queue->queuedata = card;
> +	card->plug_cb.callback = mm_unplug;
>  
>  	tasklet_init(&card->tasklet, process_page, (unsigned long)card);


  reply	other threads:[~2012-04-04 21:47 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-12-30 14:17 [PATCH] umem: fix up unplugging Tao Guo
2012-04-04 21:46 ` Andrew Morton [this message]
2012-04-04 22:20   ` Tao Guo
2012-04-04 23:18     ` Jens Axboe
2012-04-04 23:58       ` Tao Guo
2012-04-05 23:51         ` Jens Axboe
2012-04-06  4:09           ` Tao Guo
2012-04-06  4:19             ` Jens Axboe
2012-04-06 16:31               ` Tao Guo

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120404144656.7b2a06ec.akpm@linux-foundation.org \
    --to=akpm@linux-foundation.org \
    --cc=Tao.Guo@emc.com \
    --cc=axboe@fusionio.com \
    --cc=axboe@kernel.dk \
    --cc=glorioustao@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=neilb@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.