Re: [PATCH 9/9] rsxx: Disallows DMA transfer during format.

["Philip J. Kelleher" <pjk1939@linux.vnet.ibm.com>]

Thanks for the prompt reply.

I was actually questioning myself if I should include this patch in
in my patch series.

This case really is only a solution if no data is running before
a format is issued. It is currently a know bug if data is running 
and a format is issued.

Go ahead and disregard this patch. Once I get sometime to fix
data and format for all cases is when I'll submitted it.


On Wed, May 01, 2013 at 01:43:26PM +0200, Jens Axboe wrote:
> On Tue, Apr 30 2013, Philip J. Kelleher wrote:
> > From: Philip J Kelleher <pjk1939@linux.vnet.ibm.com>
> > 
> > Verifies the card is not formatting before DMAs are
> > transferred.
> > 
> > Signed-off-by: Philip J Kelleher <pjk1939@linux.vnet.ibm.com>
> > -------------------------------------------------------------------------------
> > 
> > 
> > diff -uprN -X linux-block-vanilla/Documentation/dontdiff linux-block-vanilla/drivers/block/rsxx/dma.c linux-block/drivers/block/rsxx/dma.c
> > --- linux-block-vanilla/drivers/block/rsxx/dma.c	2013-04-29 16:06:49.997249172 -0500
> > +++ linux-block/drivers/block/rsxx/dma.c	2013-04-29 16:10:46.960184429 -0500
> > @@ -721,7 +721,8 @@ int rsxx_dma_queue_bio(struct rsxx_cardi
> >  		}
> >  	}
> >  
> > -	if (unlikely(card->halt))
> > +	if (unlikely(card->halt) ||
> > +	    unlikely(card->state == CARD_STATE_FORMATTING))
> >  		goto bvec_err;
> >  
> >  	for (i = 0; i < card->n_targets; i++) {
> 
> What happens if card->state is set to formatting right after this check?
> 
> -- 
> Jens Axboe
>