public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
To: Jan Beulich <JBeulich@suse.com>
Cc: Roger Pau Monne <roger.pau@citrix.com>,
	xen-devel@lists.xen.org, linux-kernel@vger.kernel.org
Subject: Re: [Xen-devel] [PATCH 1/5] xen-blkback: don't store dev_bus_addr
Date: Tue, 19 Mar 2013 11:10:00 -0400	[thread overview]
Message-ID: <20130319151000.GA8550@phenom.dumpdata.com> (raw)
In-Reply-To: <51488A8C02000078000C6C20@nat28.tlf.novell.com>

On Tue, Mar 19, 2013 at 02:55:56PM +0000, Jan Beulich wrote:
> >>> On 19.03.13 at 15:32, Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> wrote:
> > On Mon, Mar 18, 2013 at 05:49:32PM +0100, Roger Pau Monne wrote:
> >> dev_bus_addr returned in the grant ref map operation is the mfn of the
> >> passed page, there's no need to store it in the persistent grant
> >> entry, since we can always get it provided that we have the page.
> >> 
> >> This reduces the memory overhead of persistent grants in blkback.
> > 
> > I took this patch, but I redid it a bit:
> > 
> > commit 1d4cb410befdb8b373c6fad604b39e0200e0bee0
> > Author: Roger Pau Monne <roger.pau@citrix.com>
> > Date:   Mon Mar 18 17:49:32 2013 +0100
> > 
> >     xen-blkback: don't store dev_bus_addr
> >     
> >     dev_bus_addr returned in the grant ref map operation is the mfn of the
> >     passed page, there's no need to store it in the persistent grant
> >     entry, since we can always get it provided that we have the page.
> >     
> >     This reduces the memory overhead of persistent grants in blkback.
> >     
> >     While at it, rename the 'seg[i].buf' to be 'seg[i].offset' as
> >     it makes much more sense - as we use that value in bio_add_page
> >     which as the fourth argument expects the offset.
> >     
> >     We hadn't used the physical address as part of this at all.
> >     
> >     Signed-off-by: Roger Pau Monné <roger.pau@citrix.com>
> >     Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> >     Cc: xen-devel@lists.xen.org 
> >     [v1: s/buf/offset/]
> >     Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> > 
> > diff --git a/drivers/block/xen-blkback/blkback.c 
> > b/drivers/block/xen-blkback/blkback.c
> > index 2cf8381..061c202 100644
> > --- a/drivers/block/xen-blkback/blkback.c
> > +++ b/drivers/block/xen-blkback/blkback.c
> > @@ -442,7 +442,7 @@ int xen_blkif_schedule(void *arg)
> >  }
> >  
> >  struct seg_buf {
> > -	unsigned long buf;
> > +	unsigned long offset;
> 
> If you touch this anyway, why don't you reduce the type to
> "unsigned int", halving the overall structure size?
> 
> Even more, the field seems pointless to me altogether, since ...
> 
> >  	unsigned int nsec;
> >  };
> >  /*
> > @@ -621,30 +621,21 @@ static int xen_blkbk_map(struct blkif_request *req,
> >  				 * If this is a new persistent grant
> >  				 * save the handler
> >  				 */
> > -				persistent_gnts[i]->handle = map[j].handle;
> > -				persistent_gnts[i]->dev_bus_addr =
> > -					map[j++].dev_bus_addr;
> > +				persistent_gnts[i]->handle = map[j++].handle;
> >  			}
> >  			pending_handle(pending_req, i) =
> >  				persistent_gnts[i]->handle;
> >  
> >  			if (ret)
> >  				continue;
> > -
> > -			seg[i].buf = persistent_gnts[i]->dev_bus_addr |
> > -				(req->u.rw.seg[i].first_sect << 9);
> >  		} else {
> > -			pending_handle(pending_req, i) = map[j].handle;
> > +			pending_handle(pending_req, i) = map[j++].handle;
> >  			bitmap_set(pending_req->unmap_seg, i, 1);
> >  
> > -			if (ret) {
> > -				j++;
> > +			if (ret)
> >  				continue;
> > -			}
> > -
> > -			seg[i].buf = map[j++].dev_bus_addr |
> > -				(req->u.rw.seg[i].first_sect << 9);
> >  		}
> > +		seg[i].offset = (req->u.rw.seg[i].first_sect << 9);
> 
> ... this uses "i" as index on both sides, so ...
> 
> >  	}
> >  	return ret;
> >  }
> > @@ -971,7 +962,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif,
> >  		       (bio_add_page(bio,
> >  				     pages[i],
> >  				     seg[i].nsec << 9,
> > -				     seg[i].buf & ~PAGE_MASK) == 0)) {
> > +				     seg[i].offset & ~PAGE_MASK) == 0)) {
> 
> ... this one could as well use the original field.
> 
> And the masking with ~PAGE_MASK is not pointless in any case.

Good point. In which might as well make the 'struct seg_buf' be an
simple array of unsigned int.

> 
> Jan
> 
> >  
> >  			bio = bio_alloc(GFP_KERNEL, nseg-i);
> >  			if (unlikely(bio == NULL))
> > diff --git a/drivers/block/xen-blkback/common.h 
> > b/drivers/block/xen-blkback/common.h
> > index da78346..60103e2 100644
> > --- a/drivers/block/xen-blkback/common.h
> > +++ b/drivers/block/xen-blkback/common.h
> > @@ -187,7 +187,6 @@ struct persistent_gnt {
> >  	struct page *page;
> >  	grant_ref_t gnt;
> >  	grant_handle_t handle;
> > -	uint64_t dev_bus_addr;
> >  	struct rb_node node;
> >  };
> >  
> > 
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@lists.xen.org 
> > http://lists.xen.org/xen-devel 
> 
> 

  reply	other threads:[~2013-03-19 15:10 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-18 16:49 [PATCH 0/5] xen-block: cleanup and fixes Roger Pau Monne
2013-03-18 16:49 ` [PATCH 1/5] xen-blkback: don't store dev_bus_addr Roger Pau Monne
2013-03-19 14:32   ` Konrad Rzeszutek Wilk
2013-03-19 14:55     ` [Xen-devel] " Jan Beulich
2013-03-19 15:10       ` Konrad Rzeszutek Wilk [this message]
2013-03-19 15:24         ` Jan Beulich
2013-03-19 15:26         ` Roger Pau Monné
2013-03-19 16:56           ` Konrad Rzeszutek Wilk
2013-03-18 16:49 ` [PATCH 2/5] xen-blkback: fix foreach_grant_safe to handle empty lists Roger Pau Monne
2013-03-18 16:49 ` [PATCH 3/5] xen-blkfront: switch from llist to list Roger Pau Monne
2013-03-19 12:51   ` Konrad Rzeszutek Wilk
2013-03-19 12:57     ` Roger Pau Monné
2013-03-19 14:31   ` Konrad Rzeszutek Wilk
2013-03-18 16:49 ` [PATCH 4/5] xen-blkfront: pre-allocate pages for requests Roger Pau Monne
2013-03-18 16:49 ` [PATCH 5/5] xen-blkfront: remove frame list from blk_shadow Roger Pau Monne

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=20130319151000.GA8550@phenom.dumpdata.com \
    --to=konrad.wilk@oracle.com \
    --cc=JBeulich@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=roger.pau@citrix.com \
    --cc=xen-devel@lists.xen.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox