From: "Ananyev, Konstantin" <konstantin.ananyev@intel.com>
To: Thomas Monjalon <thomas.monjalon@6wind.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>,
Olivier Matz <olivier.matz@6wind.com>,
Adrien Mazarguil <adrien.mazarguil@6wind.com>,
"Zhang, Helin" <helin.zhang@intel.com>
Subject: Re: [dpdk-dev] [PATCH] mbuf: remove inconsistent assert statements
Date: Thu, 9 Jun 2016 15:45:23 +0000 [thread overview]
Message-ID: <2601191342CEEE43887BDE71AB97725836B6DA64@irsmsx105.ger.corp.intel.com> (raw)
In-Reply-To: <11387408.tmVAJa3yAF@xps13>
> -----Original Message-----
> From: Thomas Monjalon [mailto:thomas.monjalon@6wind.com]
> Sent: Thursday, June 09, 2016 4:28 PM
> To: Ananyev, Konstantin
> Cc: dev@dpdk.org; Olivier Matz; Adrien Mazarguil; Zhang, Helin
> Subject: Re: [dpdk-dev] [PATCH] mbuf: remove inconsistent assert statements
>
> 2016-06-09 13:21, Ananyev, Konstantin:
> > From: Olivier Matz [mailto:olivier.matz@6wind.com]
> > > Today:
> > >
> > > /* allowed */
> > > m = rte_pktmbuf_alloc();
> > > rte_pktmbuf_free(m);
> > >
> > > /* not allowed */
> > > m = rte_mbuf_raw_alloc();
> > > __rte_mbuf_raw_free(m);
> > >
> > > /* we should do instead (strange): */
> > > m = rte_mbuf_raw_alloc();
> > > rte_pktmbuf_free(m);
> > >
> > > What I suggest to have:
> > >
> > > /* allowed, no change */
> > > m = rte_pktmbuf_alloc();
> > > rte_pktmbuf_free(m);
> > >
> > > /* allowed, these functions would be symetrical */
> > > m = rte_mbuf_raw_alloc();
> > > rte_mbuf_raw_free(m);
> > >
> > > /* not allowed, m->refcnt is uninitialized */
> > > m = rte_mbuf_raw_alloc();
> > > rte_pktmbuf_free(m);
> >
> > Hmm, and what it will buy us (except of symmetry)?
>
> API consistency is important.
> It is a matter of making our users confident in DPDK.
>
> I would not like to use a library where I need to check in the doc
> for each function because I don't remember and I'm not confident
> that the function fish() do some fishing or hunting.
As you remember, the whole story started when people used
'internal mbuf function' inside their code and then
figured out that it doesn't work as they expect :)
But as I said, if everyone are that desperate about symmetry,
we can just create a new public function:
void
rte_mbuf_raw_free(stuct rte_mbuf *m)
{
if (rte_mbuf_refcnt_update(m, -1) == 0)
__rte_mbuf_raw_free(m);
}
That would be 'symmetric' to rte_mbuf_raw_alloc(),
and all three combinations above would be allowed.
Konstantin
next prev parent reply other threads:[~2016-06-09 15:45 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-08 8:31 Adrien Mazarguil
2016-06-08 10:34 ` Ananyev, Konstantin
2016-06-08 12:27 ` Adrien Mazarguil
2016-06-08 13:09 ` Ananyev, Konstantin
2016-06-08 13:57 ` Adrien Mazarguil
2016-06-08 14:11 ` Olivier Matz
2016-06-08 16:07 ` Ananyev, Konstantin
2016-06-09 7:46 ` Olivier Matz
2016-06-09 13:21 ` Ananyev, Konstantin
2016-06-09 14:19 ` Bruce Richardson
2016-06-09 15:27 ` Thomas Monjalon
2016-06-09 15:45 ` Ananyev, Konstantin [this message]
2016-06-09 18:42 ` Don Provan
2016-06-20 13:49 ` Adrien Mazarguil
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=2601191342CEEE43887BDE71AB97725836B6DA64@irsmsx105.ger.corp.intel.com \
--to=konstantin.ananyev@intel.com \
--cc=adrien.mazarguil@6wind.com \
--cc=dev@dpdk.org \
--cc=helin.zhang@intel.com \
--cc=olivier.matz@6wind.com \
--cc=thomas.monjalon@6wind.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).