DPDK patches and discussions
 help / color / mirror / Atom feed
From: Olivier Matz <olivier.matz@6wind.com>
To: Bruce Richardson <bruce.richardson@intel.com>
Cc: "Ananyev, Konstantin" <konstantin.ananyev@intel.com>,
	"Verkamp, Daniel" <daniel.verkamp@intel.com>,
	"dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v2] ring: use aligned memzone allocation
Date: Thu, 8 Jun 2017 16:05:26 +0200	[thread overview]
Message-ID: <20170608160526.7953dd38@platinum> (raw)
In-Reply-To: <20170608132052.GA57628@bricha3-MOBL3.ger.corp.intel.com>

On Thu, 8 Jun 2017 14:20:52 +0100, Bruce Richardson <bruce.richardson@intel.com> wrote:
> On Thu, Jun 08, 2017 at 02:45:40PM +0200, Olivier Matz wrote:
> > On Tue, 6 Jun 2017 15:56:28 +0100, Bruce Richardson <bruce.richardson@intel.com> wrote:  
> > > On Tue, Jun 06, 2017 at 02:19:21PM +0100, Ananyev, Konstantin wrote:  
> > > > 
> > > >     
> > > > > -----Original Message-----
> > > > > From: Richardson, Bruce
> > > > > Sent: Tuesday, June 6, 2017 1:42 PM
> > > > > To: Ananyev, Konstantin <konstantin.ananyev@intel.com>
> > > > > Cc: Verkamp, Daniel <daniel.verkamp@intel.com>; dev@dpdk.org
> > > > > Subject: Re: [dpdk-dev] [PATCH v2] ring: use aligned memzone allocation
> > > > > 
> > > > > On Tue, Jun 06, 2017 at 10:59:59AM +0100, Ananyev, Konstantin wrote:    
> > > > > >    
> > > > > > > >
> > > > > > > >
> > > > > > > >    
> > > > > > > > >
> > > > > > > > > The PROD/CONS_ALIGN values on x86-64 are set to 2 cache lines, so members    
> > > > > > > > of struct rte_ring are 128 byte aligned,    
> > > > > > > > >and therefore the whole struct needs 128-byte alignment according to the ABI    
> > > > > > > > so that the 128-byte alignment of the fields can be guaranteed.
> > > > > > > >
> > > > > > > > Ah ok, missed the fact that rte_ring is 128B aligned these days.
> > > > > > > > BTW, I probably missed the initial discussion, but what was the reason for that?
> > > > > > > > Konstantin    
> > > > > > >
> > > > > > > I don't know why PROD_ALIGN/CONS_ALIGN use 128 byte alignment; it seems unnecessary if the cache line is only 64 bytes.  An    
> > > > > alternate    
> > > > > > > fix would be to just use cache line alignment for these fields (since memzones are already cache line aligned).    
> > > > > >
> > > > > > Yes, had the same thought.
> > > > > >    
> > > > > > > Maybe there is some deeper  reason for the >= 128-byte alignment logic in rte_ring.h?    
> > > > > >
> > > > > > Might be, would be good to hear opinion the author of that change.    
> > > > > 
> > > > > It gives improved performance for core-2-core transfer.    
> > > > 
> > > > You mean empty cache-line(s) after prod/cons, correct?
> > > > That's ok but why we can't keep them and whole rte_ring aligned on cache-line boundaries?
> > > > Something like that:
> > > > struct rte_ring {
> > > >    ...
> > > >    struct rte_ring_headtail prod __rte_cache_aligned;
> > > >    EMPTY_CACHE_LINE   __rte_cache_aligned;
> > > >    struct rte_ring_headtail cons __rte_cache_aligned;
> > > >    EMPTY_CACHE_LINE   __rte_cache_aligned;
> > > > };
> > > > 
> > > > Konstantin    
> > > 
> > > Sure. That should probably work too. 
> > > 
> > > /Bruce  
> > 
> > I also agree with Konstantin's proposal. One question though: since it
> > changes the alignment constraint of the rte_ring structure, I think it is
> > an ABI breakage: a structure including the rte_ring structure inherits
> > from this constraint.
> > 
> > How could we handle that, knowing this is probably a rare case?
> > 
> >  
> Is it an ABI break so long as we keep the resulting size and field
> placement of the structures the same? The alignment being reduced should
> not be a problem, as 128byte alignment is also valid as 64byte
> alignment, after all.

I'd say yes. Consider the following example:

---8<---
#include <stdio.h>
#include <stdlib.h>

#define ALIGN 64
/* #define ALIGN 128 */

/* dummy rte_ring struct */
struct rte_ring {
	char x[128];
} __attribute__((aligned(ALIGN)));

struct foo {
	struct rte_ring r;
	unsigned bar;
};

int main(void)
{
	struct foo array[2];

	printf("sizeof(ring)=%zu diff=%u\n",
		sizeof(struct rte_ring),
		(unsigned int)((char *)&array[1].r - (char *)array));

	return 0;
}
---8<---

The size of rte_ring is always 128.
diff is 192 or 256, depending on the value of ALIGN.



Olivier

  reply	other threads:[~2017-06-08 14:05 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-06-02 20:03 [dpdk-dev] [PATCH] " Daniel Verkamp
2017-06-02 20:12 ` [dpdk-dev] [PATCH v2] " Daniel Verkamp
2017-06-02 20:51   ` Ananyev, Konstantin
2017-06-02 22:24     ` Verkamp, Daniel
2017-06-03 10:00       ` Ananyev, Konstantin
2017-06-05 16:21         ` Verkamp, Daniel
2017-06-06  9:59           ` Ananyev, Konstantin
2017-06-06 12:42             ` Bruce Richardson
2017-06-06 13:19               ` Ananyev, Konstantin
2017-06-06 14:56                 ` Bruce Richardson
2017-06-08 12:45                   ` Olivier Matz
2017-06-08 13:20                     ` Bruce Richardson
2017-06-08 14:05                       ` Olivier Matz [this message]
2017-06-08 14:11                         ` Bruce Richardson
2017-06-08 14:50                           ` Ananyev, Konstantin
2017-06-08 15:24                             ` Bruce Richardson
2017-06-08 15:35                               ` Ananyev, Konstantin
2017-06-08 16:03                                 ` Bruce Richardson
2017-06-08 16:12                                   ` Ananyev, Konstantin
2017-06-08 16:20                                     ` Richardson, Bruce
2017-06-08 16:42                                       ` Ananyev, Konstantin
2017-06-09  9:02                                         ` Bruce Richardson
2017-06-12  9:02                                           ` Olivier Matz
2017-06-12  9:56                                             ` Bruce Richardson
2017-06-30 11:35                                               ` Olivier Matz
2017-06-09 12:47                 ` Yerden Zhumabekov
2017-06-09 17:16                   ` Stephen Hemminger
2017-06-09 17:28                     ` Jerin Jacob
2017-06-10  8:16                       ` Ananyev, Konstantin
2017-06-12  3:07                         ` Jerin Jacob
2017-06-12 10:18                           ` Ananyev, Konstantin
2017-06-12 10:34                             ` Jerin Jacob
2017-06-12 11:09                               ` Bruce Richardson
2017-06-12 11:41                                 ` Jerin Jacob
2017-06-12 12:17                                   ` Ananyev, Konstantin
2017-06-12 12:42                                     ` Jerin Jacob
2017-06-12 12:51                                       ` Ananyev, Konstantin
2017-06-12 13:06                                         ` Bruce Richardson
2017-06-12 13:20                                         ` Jerin Jacob
2017-06-30 11:36   ` Olivier Matz
2017-07-01 11:14     ` Thomas Monjalon
2017-07-01 11:25       ` Thomas Monjalon

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=20170608160526.7953dd38@platinum \
    --to=olivier.matz@6wind.com \
    --cc=bruce.richardson@intel.com \
    --cc=daniel.verkamp@intel.com \
    --cc=dev@dpdk.org \
    --cc=konstantin.ananyev@intel.com \
    /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;
as well as URLs for NNTP newsgroup(s).