From: "Lu, Wenzhuo" <wenzhuo.lu@intel.com>
To: Olivier Matz <olivier.matz@6wind.com>,
Stephen Hemminger <stephen@networkplumber.org>
Cc: "dev@dpdk.org" <dev@dpdk.org>, "Tao, Zhe" <zhe.tao@intel.com>
Subject: Re: [dpdk-dev] [PATCH 2/8] lib/librte_ether: defind RX/TX lock mode
Date: Sun, 12 Jun 2016 05:25:41 +0000 [thread overview]
Message-ID: <6A0DE07E22DDAD4C9103DF62FEBC090903484BFF@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <57591FE1.9070908@6wind.com>
Hi Olivier,
> -----Original Message-----
> From: Olivier Matz [mailto:olivier.matz@6wind.com]
> Sent: Thursday, June 9, 2016 3:51 PM
> To: Lu, Wenzhuo; Stephen Hemminger
> Cc: dev@dpdk.org; Tao, Zhe
> Subject: Re: [dpdk-dev] [PATCH 2/8] lib/librte_ether: defind RX/TX lock mode
>
> Hi,
>
> On 06/08/2016 09:34 AM, Lu, Wenzhuo wrote:
> > Hi Stephen,
> >
> >
> >> -----Original Message-----
> >> From: Stephen Hemminger [mailto:stephen@networkplumber.org]
> >> Sent: Wednesday, June 8, 2016 10:16 AM
> >> To: Lu, Wenzhuo
> >> Cc: dev@dpdk.org; Tao, Zhe
> >> Subject: Re: [dpdk-dev] [PATCH 2/8] lib/librte_ether: defind RX/TX
> >> lock mode
> >>
> >> On Mon, 6 Jun 2016 13:40:47 +0800
> >> Wenzhuo Lu <wenzhuo.lu@intel.com> wrote:
> >>
> >>> Define lock mode for RX/TX queue. Because when resetting the device
> >>> we want the resetting thread to get the lock of the RX/TX queue to
> >>> make sure the RX/TX is stopped.
> >>>
> >>> Using next ABI macro for this ABI change as it has too much impact.
> >>> 7 APIs and 1 global variable are impacted.
> >>>
> >>> Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
> >>> Signed-off-by: Zhe Tao <zhe.tao@intel.com>
> >>
> >> Why does this patch set make a different assumption the rest of the DPDK?
> >>
> >> The rest of the DPDK operates on the principle that the application
> >> is smart enough to stop the device before making changes. There is no
> >> equivalent to the Linux kernel RTNL mutex. The API assumes
> >> application threads are well behaved and will not try and sabotage each
> other.
> >>
> >> If you restrict the reset operation to only being available when
> >> RX/TX is stopped, then no lock is needed.
> >>
> >> The fact that it requires lots more locking inside each device driver
> >> implies to me this is not correct way to architect this.
>
> +1
>
> I'm not sure adding locks is the proper way to do.
> This is the application responsibility to ensure that:
> - control functions are not called concurrently on the same port
> - rx/tx functions are not called when the device is stopped/reset/...
>
> However, I do think the usage paradigms of the ethdev api should be better
> documented in rte_ethdev.h (ex: which functions can be called concurrently).
> This would be a first step.
>
> If we really want a helper API to do that in DPDK, the _next_ step could be to
> add them in the ethdev api to achieve this. Maybe something like (the function
> names could be better):
>
> - to be called on one control thread:
>
> rte_eth_stop_rxtx(port)
> rte_eth_start_rxtx(port)
>
> rte_eth_get_rxtx_state(port)
> -> return "running" if at least one core is inside the rx/tx code
> -> return "stopped" if all cores are outside the rx/tx code
>
> - to be called on dataplane cores:
>
> /* same than rte_eth_rx_burst(), but checks if rx/tx is allowed
> * first, else do nothing */
> rte_eth_rx_burst_interruptible()
> rte_eth_tx_burst_interruptible()
>
>
> The code of control thread could be:
>
> rte_eth_stop_rxtx(port);
> /* wait that all dataplane cores finished their processing */
> while (rte_eth_get_rxtx_state(port) != stopped)
> ;
> rte_eth_some_control_operation(port);
> rte_eth_start_rxtx(port);
>
>
> I think this could be done without any lock, just with the proper memory barriers
> and a per-core status.
>
> But this API may impose a paradigm to the application, and I'm not sure the
> DPDK should do that.
I don't quite catch your point. Seems your solution still need the APP to change the code. I think it's more complex than just letting the APP to stop the rx/tx and reset the port. Our purpose of this patch set is to let APP do less as possible. It's not a good choice if we make it more complex.
And seems it's hard to stop and start rx/tx in rte layer. Normally APP should do that. To my opinion, we have to introduce lock in rte to achieve that.
>
> Regards,
> Olivier
next prev parent reply other threads:[~2016-06-12 5:25 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-06-06 5:40 [dpdk-dev] [PATCH 0/8] support reset of VF link Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 1/8] lib/librte_ether: support device reset Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 2/8] lib/librte_ether: defind RX/TX lock mode Wenzhuo Lu
2016-06-08 2:15 ` Stephen Hemminger
2016-06-08 7:34 ` Lu, Wenzhuo
2016-06-09 7:50 ` Olivier Matz
2016-06-12 5:25 ` Lu, Wenzhuo [this message]
2016-06-10 18:12 ` Stephen Hemminger
2016-06-12 5:27 ` Lu, Wenzhuo
2016-06-06 5:40 ` [dpdk-dev] [PATCH 3/8] ixgbe: RX/TX with lock on VF Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 4/8] ixgbe: implement device reset " Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 5/8] igb: RX/TX with lock " Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 6/8] igb: implement device reset " Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 7/8] i40e:RX/TX with lock " Wenzhuo Lu
2016-06-06 5:40 ` [dpdk-dev] [PATCH 8/8] i40e: implement device reset " Wenzhuo Lu
2016-06-15 3:03 ` [dpdk-dev] [PATCH v5 0/4] support reset of VF link Wenzhuo Lu
2016-06-15 3:03 ` [dpdk-dev] [PATCH v5 1/4] lib/librte_ether: support device reset Wenzhuo Lu
2016-06-16 15:31 ` Bruce Richardson
2016-06-16 15:36 ` Thomas Monjalon
2016-06-15 3:03 ` [dpdk-dev] [PATCH v5 2/4] ixgbe: implement device reset on VF Wenzhuo Lu
2016-06-15 3:03 ` [dpdk-dev] [PATCH v5 3/4] igb: " Wenzhuo Lu
2016-06-15 3:03 ` [dpdk-dev] [PATCH v5 4/4] i40e: " Wenzhuo Lu
2016-06-20 6:24 ` [dpdk-dev] [PATCH v6 0/4] support reset of VF link Wenzhuo Lu
2016-06-20 6:24 ` [dpdk-dev] [PATCH v6 1/4] lib/librte_ether: support device reset Wenzhuo Lu
2016-06-20 9:14 ` Jerin Jacob
2016-06-20 16:17 ` Stephen Hemminger
2016-06-21 3:51 ` Jerin Jacob
2016-06-21 6:14 ` Lu, Wenzhuo
2016-06-21 7:37 ` Jerin Jacob
2016-06-21 8:24 ` Lu, Wenzhuo
2016-06-21 8:55 ` Jerin Jacob
2016-06-21 9:26 ` Ananyev, Konstantin
2016-06-21 10:57 ` Jerin Jacob
2016-06-21 13:10 ` Ananyev, Konstantin
2016-06-21 13:30 ` Jerin Jacob
2016-06-21 14:03 ` Ananyev, Konstantin
2016-06-21 14:29 ` Jerin Jacob
2016-06-22 1:35 ` Lu, Wenzhuo
2016-06-22 2:37 ` Jerin Jacob
2016-06-22 3:32 ` Lu, Wenzhuo
2016-06-22 4:14 ` Jerin Jacob
2016-06-22 5:05 ` Lu, Wenzhuo
2016-06-22 6:10 ` Jerin Jacob
2016-06-22 6:42 ` Lu, Wenzhuo
2016-06-22 7:59 ` Jerin Jacob
2016-06-22 8:17 ` Thomas Monjalon
2016-06-22 8:25 ` Lu, Wenzhuo
2016-06-22 9:18 ` Thomas Monjalon
2016-06-22 11:06 ` Jerin Jacob
2016-06-23 0:45 ` Lu, Wenzhuo
2016-06-23 0:39 ` Lu, Wenzhuo
2016-06-21 0:51 ` Lu, Wenzhuo
2016-06-20 6:24 ` [dpdk-dev] [PATCH v6 2/4] ixgbe: implement device reset on VF Wenzhuo Lu
2016-06-20 6:24 ` [dpdk-dev] [PATCH v6 3/4] igb: " Wenzhuo Lu
2016-06-20 6:24 ` [dpdk-dev] [PATCH v6 4/4] i40e: " Wenzhuo Lu
2016-07-04 15:48 ` [dpdk-dev] [PATCH v6 0/4] support reset of VF link Luca Boccassi
2016-07-05 0:52 ` Lu, Wenzhuo
2016-07-05 9:52 ` Luca Boccassi
2016-07-06 0:45 ` Lu, Wenzhuo
2016-07-06 16:26 ` Luca Boccassi
[not found] ` <1467822182.32466.34.camel@brocade.com>
2016-07-07 1:09 ` Lu, Wenzhuo
2016-07-07 10:20 ` Luca Boccassi
2016-07-07 13:12 ` Lu, Wenzhuo
2016-07-07 16:19 ` Luca Boccassi
2016-07-08 0:14 ` Lu, Wenzhuo
2016-07-08 17:15 ` Luca Boccassi
2016-07-11 1:32 ` Lu, Wenzhuo
2016-07-11 12:02 ` Luca Boccassi
2016-07-11 15:43 ` Luca Boccassi
2016-07-12 1:19 ` Lu, Wenzhuo
2016-08-26 12:58 ` Luca Boccassi
2016-08-29 1:04 ` Lu, Wenzhuo
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=6A0DE07E22DDAD4C9103DF62FEBC090903484BFF@shsmsx102.ccr.corp.intel.com \
--to=wenzhuo.lu@intel.com \
--cc=dev@dpdk.org \
--cc=olivier.matz@6wind.com \
--cc=stephen@networkplumber.org \
--cc=zhe.tao@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).