DPDK patches and discussions
 help / color / mirror / Atom feed
From: Honnappa Nagarahalli <Honnappa.Nagarahalli@arm.com>
To: "Chautru, Nicolas" <nicolas.chautru@intel.com>,
	Phil Yang <Phil.Yang@arm.com>, "dev@dpdk.org" <dev@dpdk.org>,
	"david.marchand@redhat.com" <david.marchand@redhat.com>,
	"Hunt, David" <david.hunt@intel.com>
Cc: Ruifeng Wang <Ruifeng.Wang@arm.com>, nd <nd@arm.com>,
	Honnappa Nagarahalli <Honnappa.Nagarahalli@arm.com>,
	nd <nd@arm.com>
Subject: Re: [dpdk-dev] [PATCH v3 2/4] bbdev: use C11 atomic builtins for device processing counter
Date: Thu, 24 Sep 2020 22:44:40 +0000	[thread overview]
Message-ID: <DBAPR08MB581456F23F954A71C837F1EA98390@DBAPR08MB5814.eurprd08.prod.outlook.com> (raw)
In-Reply-To: <BY5PR11MB4451D4BEF802C005349C1941F8390@BY5PR11MB4451.namprd11.prod.outlook.com>

<snip>

> 
> Hi Phil,
> Naïve question but the deprecation document was stating that "DPDK will
> adopt C11 atomic operations semantics and provide wrappers using C11
> atomic built-ins."
At the time of writing the deprecation notice, that was the thinking. However, through further discussions [1] in the community, it was decided to use atomic built-ins.

[1] http://mails.dpdk.org/archives/dev/2020-May/167416.html

> Here you are using directly the C11 atomic built-ins and not providing and
> using a DPDK wrapper.
> Wasn't the intent to have a new rte_... wrapper here? Ie. the same way as
> the __sync_fetch_and_add() were called before behind the rte_atomicNN_XX
> wrapper.
> 
> Thanks
> Nic
> 
> 
> > -----Original Message-----
> > From: Phil Yang <phil.yang@arm.com>
> > Sent: Wednesday, September 23, 2020 10:39 PM
> > To: dev@dpdk.org; david.marchand@redhat.com; Chautru, Nicolas
> > <nicolas.chautru@intel.com>; Hunt, David <david.hunt@intel.com>
> > Cc: Ruifeng.Wang@arm.com; Honnappa.Nagarahalli@arm.com;
> nd@arm.com
> > Subject: [PATCH v3 2/4] bbdev: use C11 atomic builtins for device
> > processing counter
> >
> > Since rte_atomicXX APIs are not allowed to be used, use C11 atomic
> > builtins for device processing counter.
> >
> > Signed-off-by: Phil Yang <phil.yang@arm.com>
> > Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
> > Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> > ---
> >  lib/librte_bbdev/rte_bbdev.c | 5 +++--  lib/librte_bbdev/rte_bbdev.h
> > | 4 +---
> >  2 files changed, 4 insertions(+), 5 deletions(-)
> >
> > diff --git a/lib/librte_bbdev/rte_bbdev.c
> > b/lib/librte_bbdev/rte_bbdev.c index a4fdb69..5ba891c 100644
> > --- a/lib/librte_bbdev/rte_bbdev.c
> > +++ b/lib/librte_bbdev/rte_bbdev.c
> > @@ -210,7 +210,7 @@ rte_bbdev_allocate(const char *name)
> >  		return NULL;
> >  	}
> >
> > -	rte_atomic16_inc(&bbdev->data->process_cnt);
> > +	__atomic_add_fetch(&bbdev->data->process_cnt, 1,
> > __ATOMIC_RELAXED);
> >  	bbdev->data->dev_id = dev_id;
> >  	bbdev->state = RTE_BBDEV_INITIALIZED;
> >
> > @@ -252,7 +252,8 @@ rte_bbdev_release(struct rte_bbdev *bbdev)
> >  	}
> >
> >  	/* clear shared BBDev Data if no process is using the device anymore
> > */
> > -	if (rte_atomic16_dec_and_test(&bbdev->data->process_cnt))
> > +	if (__atomic_sub_fetch(&bbdev->data->process_cnt, 1,
> > +			      __ATOMIC_RELAXED) == 0)
> >  		memset(bbdev->data, 0, sizeof(*bbdev->data));
> >
> >  	memset(bbdev, 0, sizeof(*bbdev));
> > diff --git a/lib/librte_bbdev/rte_bbdev.h
> > b/lib/librte_bbdev/rte_bbdev.h index 5729137..7017124 100644
> > --- a/lib/librte_bbdev/rte_bbdev.h
> > +++ b/lib/librte_bbdev/rte_bbdev.h
> > @@ -33,7 +33,6 @@ extern "C" {
> >  #include <string.h>
> >
> >  #include <rte_compat.h>
> > -#include <rte_atomic.h>
> >  #include <rte_bus.h>
> >  #include <rte_cpuflags.h>
> >  #include <rte_memory.h>
> > @@ -426,8 +425,7 @@ struct rte_bbdev_data {
> >  	uint16_t dev_id;  /**< Device ID */
> >  	int socket_id;  /**< NUMA socket that device is on */
> >  	bool started;  /**< Device run-time state */
> > -	/** Counter of processes using the device */
> > -	rte_atomic16_t process_cnt;
> > +	uint16_t process_cnt;  /** Counter of processes using the device */
> >  };
> >
> >  /* Forward declarations */
> > --
> > 2.7.4


  reply	other threads:[~2020-09-24 22:44 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-11  3:29 [dpdk-dev] [PATCH 0/4] use C11 atomic builtins for libs Phil Yang
2020-09-11  3:29 ` [dpdk-dev] [PATCH 1/4] eal: use C11 atomic builtins for already initialized check Phil Yang
2020-09-11  3:29 ` [dpdk-dev] [PATCH 2/4] bbdev: use C11 atomic builtins for device processing counter Phil Yang
2020-09-11  3:29 ` [dpdk-dev] [PATCH 3/4] power: use C11 atomic builtins for power in use state update Phil Yang
2020-09-11  3:29 ` [dpdk-dev] [PATCH 4/4] ethdev: use C11 atomic builtins for link status update Phil Yang
2020-09-15 15:12 ` [dpdk-dev] [PATCH 0/4] use C11 atomic builtins for libs David Marchand
2020-09-16  7:32   ` Phil Yang
2020-09-16  8:23 ` [dpdk-dev] [PATCH v2 " Phil Yang
2020-09-16  8:23   ` [dpdk-dev] [PATCH v2 1/4] eal: use C11 atomic builtins for already initialized check Phil Yang
2020-09-23 13:06     ` David Marchand
2020-09-24  3:44       ` Phil Yang
2020-09-16  8:23   ` [dpdk-dev] [PATCH v2 2/4] bbdev: use C11 atomic builtins for device processing counter Phil Yang
2020-09-16  8:23   ` [dpdk-dev] [PATCH v2 3/4] power: use C11 atomic builtins for power in use state update Phil Yang
2020-09-16  8:23   ` [dpdk-dev] [PATCH v2 4/4] ethdev: use C11 atomic builtins for link status update Phil Yang
2020-09-17 16:08     ` Andrew Rybchenko
2020-09-23 13:18   ` [dpdk-dev] [PATCH v2 0/4] use C11 atomic builtins for libs David Marchand
2020-09-24  3:47     ` Phil Yang
2020-09-24  5:39   ` [dpdk-dev] [PATCH v3 " Phil Yang
2020-09-24  5:39     ` [dpdk-dev] [PATCH v3 1/4] eal: use C11 atomic builtins for already initialized check Phil Yang
2020-09-24  5:39     ` [dpdk-dev] [PATCH v3 2/4] bbdev: use C11 atomic builtins for device processing counter Phil Yang
2020-09-24 22:01       ` Chautru, Nicolas
2020-09-24 22:44         ` Honnappa Nagarahalli [this message]
2020-09-24 23:20           ` Chautru, Nicolas
2020-09-24  5:39     ` [dpdk-dev] [PATCH v3 3/4] power: use C11 atomic builtins for power in use state update Phil Yang
2020-09-24  8:34       ` David Hunt
2020-09-24  5:39     ` [dpdk-dev] [PATCH v3 4/4] ethdev: use C11 atomic builtins for link status update Phil Yang
2020-09-25 13:59     ` [dpdk-dev] [PATCH v3 0/4] use C11 atomic builtins for libs David Marchand

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=DBAPR08MB581456F23F954A71C837F1EA98390@DBAPR08MB5814.eurprd08.prod.outlook.com \
    --to=honnappa.nagarahalli@arm.com \
    --cc=Phil.Yang@arm.com \
    --cc=Ruifeng.Wang@arm.com \
    --cc=david.hunt@intel.com \
    --cc=david.marchand@redhat.com \
    --cc=dev@dpdk.org \
    --cc=nd@arm.com \
    --cc=nicolas.chautru@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).