From: "Yan, Zhirun" <zhirun.yan@intel.com>
To: "Yang, Qiming" <qiming.yang@intel.com>, "dev@dpdk.org" <dev@dpdk.org>
Cc: "Zhang, Qi Z" <qi.z.zhang@intel.com>,
"Wang, Haiyue" <haiyue.wang@intel.com>
Subject: Re: [dpdk-dev] [PATCH v2] net/i40e: fix VF/PF port reset
Date: Fri, 7 Dec 2018 04:18:27 +0000 [thread overview]
Message-ID: <2F9759EBFF61914290A3BA61BDBE2C860B9D2EFA@SHSMSX101.ccr.corp.intel.com> (raw)
In-Reply-To: <F5DF4F0E3AFEF648ADC1C3C33AD4DBF17A444D70@SHSMSX101.ccr.corp.intel.com>
Hi Qiming,
> -----Original Message-----
> From: Yang, Qiming
> Sent: Thursday, December 6, 2018 10:57 AM
> To: dev@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; Wang, Haiyue
> <haiyue.wang@intel.com>
> Cc: Yan, Zhirun <zhirun.yan@intel.com>
> Subject: RE: [dpdk-dev] [PATCH v2] net/i40e: fix VF/PF port reset
>
> Hi, Zhirun
>
> > -----Original Message-----
> > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Zhirun Yan
> > Sent: Wednesday, December 5, 2018 8:46 PM
> > To: dev@dpdk.org; Zhang, Qi Z <qi.z.zhang@intel.com>; Wang, Haiyue
> > <haiyue.wang@intel.com>
> > Cc: Yan, Zhirun <zhirun.yan@intel.com>
> > Subject: [dpdk-dev] [PATCH v2] net/i40e: fix VF/PF port reset
> >
> > Port reset will call i40evf_uninit_vf() to release resource. It wants
> > to call
> > i40evf_dev_close() to do some clean work. Before this patch, port
> > reset will never call i40evf_dev_close() to shutdown adminq. So the
> > i40evf_dev_init() will failed.
> >
> > The status of stop/start should be flaged by dev->data->dev_started.
> > Close should be flaged by hw->adapter_stopped. This patch modify these
> > flags both in VF and PF.
> >
>
> For the fix patch need to add Fix line like:
> Fixes: commit number ("commit log"), just refer to the old patch.
Thanks, I will added it in the next version.
>
> > Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
> > Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
> > ---
> > drivers/net/i40e/i40e_ethdev.c | 13 ++++++-------
> > drivers/net/i40e/i40e_ethdev_vf.c | 17 ++++++++---------
> > 2 files changed, 14 insertions(+), 16 deletions(-)
> >
> > diff --git a/drivers/net/i40e/i40e_ethdev.c
> > b/drivers/net/i40e/i40e_ethdev.c index 7030eb1fa..8eaf6eba6 100644
> > --- a/drivers/net/i40e/i40e_ethdev.c
> > +++ b/drivers/net/i40e/i40e_ethdev.c
> > @@ -2391,15 +2391,11 @@ static void
> > i40e_dev_stop(struct rte_eth_dev *dev) {
> > struct i40e_pf *pf = I40E_DEV_PRIVATE_TO_PF(dev->data-
> > >dev_private);
> > - struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data-
> > >dev_private);
> > struct i40e_vsi *main_vsi = pf->main_vsi;
> > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
> > struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
> > int i;
> >
> > - if (hw->adapter_stopped == 1)
> > - return;
> > -
>
> adapter_stopped is used to avoid double call dev_stop, why you remove the
> check here?
hw->adapter_stopped try to control start/stop and close, but close will do more
than stop.
> Do you add another mechanism to do the same thing?
I guess dev->data->dev_started is the right flag to avoid the issue.
>
> > if (dev->data->dev_conf.intr_conf.rxq == 0) {
> > rte_eal_alarm_cancel(i40e_dev_alarm_handler, dev);
> > rte_intr_enable(intr_handle);
> > @@ -2442,8 +2438,6 @@ i40e_dev_stop(struct rte_eth_dev *dev)
> >
> > /* reset hierarchy commit */
> > pf->tm_conf.committed = false;
> > -
> > - hw->adapter_stopped = 1;
> > }
> >
> > static void
> > @@ -2460,7 +2454,10 @@ i40e_dev_close(struct rte_eth_dev *dev)
> >
> > PMD_INIT_FUNC_TRACE();
> >
> > - i40e_dev_stop(dev);
> > + if (dev->data->dev_started != 0) {
> > + dev->data->dev_started = 0;
> > + i40e_dev_stop(dev);
> > + }
> >
> > /* Remove all mirror rules */
> > while ((p_mirror = TAILQ_FIRST(&pf->mirror_list))) { @@ -2523,6
> > +2520,8 @@ i40e_dev_close(struct rte_eth_dev *dev)
> > I40E_WRITE_REG(hw, I40E_PFGEN_CTRL,
> > (reg | I40E_PFGEN_CTRL_PFSWR_MASK));
> > I40E_WRITE_FLUSH(hw);
> > +
> > + hw->adapter_stopped = 1;
>
> If you remove all the other hw->adapter_stopped, it's useless to add here.
>
I will use another flag like adapter_closed to mark the close status.
> > }
> >
> > /*
> > diff --git a/drivers/net/i40e/i40e_ethdev_vf.c
> > b/drivers/net/i40e/i40e_ethdev_vf.c
> > index ae55b9b18..cbe9bec3c 100644
> > --- a/drivers/net/i40e/i40e_ethdev_vf.c
> > +++ b/drivers/net/i40e/i40e_ethdev_vf.c
> > @@ -1924,15 +1924,12 @@ static int
> > i40evf_dev_start(struct rte_eth_dev *dev) {
> > struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data-
> > >dev_private);
> > - struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data-
> > >dev_private);
> > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
> > struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
> > uint32_t intr_vector = 0;
> >
> > PMD_INIT_FUNC_TRACE();
> >
> > - hw->adapter_stopped = 0;
> > -
>
> Question the same, will induce any problem?
I will check it.
>
> > vf->max_pkt_len = dev->data->dev_conf.rxmode.max_rx_pkt_len;
> > vf->num_queue_pairs = RTE_MAX(dev->data->nb_rx_queues,
> > dev->data->nb_tx_queues);
> > @@ -2004,7 +2001,6 @@ i40evf_dev_stop(struct rte_eth_dev *dev) {
> > struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
> > struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
> > - struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data-
> > >dev_private);
> > struct i40e_vf *vf = I40EVF_DEV_PRIVATE_TO_VF(dev->data-
> > >dev_private);
> >
> > PMD_INIT_FUNC_TRACE();
> > @@ -2012,8 +2008,6 @@ i40evf_dev_stop(struct rte_eth_dev *dev)
> > if (dev->data->dev_conf.intr_conf.rxq != 0)
> > rte_intr_disable(intr_handle);
> >
> > - if (hw->adapter_stopped == 1)
> > - return;
> > i40evf_stop_queues(dev);
> > i40evf_disable_queues_intr(dev);
> > i40e_dev_clear_queues(dev);
> > @@ -2029,7 +2023,6 @@ i40evf_dev_stop(struct rte_eth_dev *dev)
> > /* remove all multicast addresses */
> > i40evf_add_del_mc_addr_list(dev, vf->mc_addrs, vf->mc_addrs_num,
> > FALSE);
> > - hw->adapter_stopped = 1;
> >
> > }
> >
> > @@ -2246,7 +2239,11 @@ i40evf_dev_close(struct rte_eth_dev *dev) {
> > struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data-
> > >dev_private);
> >
> > - i40evf_dev_stop(dev);
> > + if (dev->data->dev_started != 0) {
> > + dev->data->dev_started = 0;
> > + i40evf_dev_stop(dev);
> > + }
> > +
> > i40e_dev_free_queues(dev);
> > /*
> > * disable promiscuous mode before reset vf @@ -2255,11 +2252,13
> @@
> > i40evf_dev_close(struct rte_eth_dev *dev)
> > */
> > i40evf_dev_promiscuous_disable(dev);
> > i40evf_dev_allmulticast_disable(dev);
> > -
> > i40evf_reset_vf(hw);
> > +
> > i40e_shutdown_adminq(hw);
> > i40evf_disable_irq0(hw);
> > rte_eal_alarm_cancel(i40evf_dev_alarm_handler, dev);
> > +
> > + hw->adapter_stopped = 1;
> > }
> >
> > /*
> > --
> > 2.17.1
next prev parent reply other threads:[~2018-12-07 4:18 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-12-05 12:46 Zhirun Yan
2018-12-05 12:51 ` Zhang, Qi Z
2018-12-06 2:56 ` Yang, Qiming
2018-12-07 4:18 ` Yan, Zhirun [this message]
2018-12-10 13:53 ` [dpdk-dev] [PATCH v3] net/i40e: fix VF/PF port close Zhirun Yan
2018-12-10 14:08 ` [dpdk-dev] [PATCH v4] " Zhirun Yan
2018-12-10 12:20 ` Zhang, Qi Z
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=2F9759EBFF61914290A3BA61BDBE2C860B9D2EFA@SHSMSX101.ccr.corp.intel.com \
--to=zhirun.yan@intel.com \
--cc=dev@dpdk.org \
--cc=haiyue.wang@intel.com \
--cc=qi.z.zhang@intel.com \
--cc=qiming.yang@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).