DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Zhao1, Wei" <wei.zhao1@intel.com>
To: "Meunier,
	Julien (Nokia - FR/Paris-Saclay)" <julien.meunier@nokia.com>,
	"Zhang, Qi Z" <qi.z.zhang@intel.com>,
	"Ananyev, Konstantin" <konstantin.ananyev@intel.com>,
	"Lu, Wenzhuo" <wenzhuo.lu@intel.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH] net/ixgbe: add support of loopback for X540/X550
Date: Tue, 8 Jan 2019 10:10:05 +0000	[thread overview]
Message-ID: <A2573D2ACFCADC41BB3BE09C6DE313CA07E9ECFD@PGSMSX103.gar.corp.intel.com> (raw)
In-Reply-To: <6b2efbe0-94dc-750d-8a4f-b98d3f909260@nokia.com>

Hi,  Meunier, Julien

> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Meunier, Julien
> (Nokia - FR/Paris-Saclay)
> Sent: Monday, January 7, 2019 11:53 PM
> To: Zhang, Qi Z <qi.z.zhang@intel.com>; Ananyev, Konstantin
> <konstantin.ananyev@intel.com>; Lu, Wenzhuo <wenzhuo.lu@intel.com>
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] net/ixgbe: add support of loopback for
> X540/X550
> 
> Hi,
> 
> Mmm, you're right. However, as like for the 82599, the pmd skips the link
> configuration, so, it should not enable the autoneg.
> 
> During my tests, I had a 10G connectivity, and I didn't notice any problem. I
> used the DPDK application "test", with the test pmd_perf_autotest.

It seems you are right, autoneg is done in function of ixgbe_setup_link()->........ ()
which call function  ixgbe_setup_phy_link_speed_generic()->ixgbe_setup_phy_link ()->
ixgbe_setup_phy_link_generic (),  do the work for Auto-Negotiation,  but pmd code skip it.
But the bit C of Address 7.0 is initialized as 1 by default in datasheet,  what pmd code skip is just the 
Process of Auto-Negotiation, because we skip code of bellow In ixgbe_setup_phy_link_generic (), not disable bit 7.0.c.
So, you had better disable 7.0.c bit when mac loopback is enable.

"
	/* Restart PHY auto-negotiation. */
	hw->phy.ops.read_reg(hw, IXGBE_MDIO_AUTO_NEG_CONTROL,
			     IXGBE_MDIO_AUTO_NEG_DEV_TYPE, &autoneg_reg);

	autoneg_reg |= IXGBE_MII_RESTART;

	hw->phy.ops.write_reg(hw, IXGBE_MDIO_AUTO_NEG_CONTROL,
			      IXGBE_MDIO_AUTO_NEG_DEV_TYPE, autoneg_reg);
"


> 
> Should I need to modify my code to be sure that autoneg is disabled (and
> force it to 10G) ?
> 
> Thanks,
> Best regards,
> Julien Meunier
> 
> On 07/01/2019 07:53, Zhang, Qi Z wrote:
> >
> >
> >> -----Original Message-----
> >> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Julien Meunier
> >> Sent: Thursday, January 3, 2019 12:01 AM
> >> To: Ananyev, Konstantin <konstantin.ananyev@intel.com>; Lu, Wenzhuo
> >> <wenzhuo.lu@intel.com>
> >> Cc: dev@dpdk.org
> >> Subject: [dpdk-dev] [PATCH] net/ixgbe: add support of loopback for
> >> X540/X550
> >>
> >> Loopback mode is also supported on X540 and X550 NICs, according to
> >> their datasheet (section 15.2). The way to set it up is a little different of
> the 82599.
> >
> > Thanks for enable this.
> >
> > one question is, Datasheet also mentioned that auto negotiation should
> > be disabled but I didn't see any related change with it.
> >
> > Would you share more insight on this?
> >
> >>
> >> Signed-off-by: Julien Meunier <julien.meunier@nokia.com>
> >> ---
> >>   drivers/net/ixgbe/ixgbe_ethdev.c | 10 ++++++---
> >> drivers/net/ixgbe/ixgbe_ethdev.h |  5 ++---
> >>   drivers/net/ixgbe/ixgbe_rxtx.c   | 47
> >> ++++++++++++++++++++++++++++++++++------
> >>   3 files changed, 49 insertions(+), 13 deletions(-)
> >>
> >> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> >> b/drivers/net/ixgbe/ixgbe_ethdev.c
> >> index 7493110..7eb3303 100644
> >> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> >> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> >> @@ -2652,9 +2652,13 @@ ixgbe_dev_start(struct rte_eth_dev *dev)
> >>   		goto error;
> >>   	}
> >>
> >> -	/* Skip link setup if loopback mode is enabled for 82599. */
> >> -	if (hw->mac.type == ixgbe_mac_82599EB &&
> >> -			dev->data->dev_conf.lpbk_mode ==
> IXGBE_LPBK_82599_TX_RX)
> >> +	/* Skip link setup if loopback mode is enabled. */
> >> +	if ((hw->mac.type == ixgbe_mac_82599EB ||
> >> +	     hw->mac.type == ixgbe_mac_X540 ||
> >> +	     hw->mac.type == ixgbe_mac_X550 ||
> >> +	     hw->mac.type == ixgbe_mac_X550EM_x ||
> >> +	     hw->mac.type == ixgbe_mac_X550EM_a) &&
> >> +			dev->data->dev_conf.lpbk_mode ==
> IXGBE_LPBK_TX_RX)
> >>   		goto skip_link_setup;
> >>
> >>   	if (ixgbe_is_sfp(hw) && hw->phy.multispeed_fiber) { diff --git
> >> a/drivers/net/ixgbe/ixgbe_ethdev.h
> b/drivers/net/ixgbe/ixgbe_ethdev.h
> >> index 565c69c..c60a697 100644
> >> --- a/drivers/net/ixgbe/ixgbe_ethdev.h
> >> +++ b/drivers/net/ixgbe/ixgbe_ethdev.h
> >> @@ -65,9 +65,8 @@
> >>   #define IXGBE_QUEUE_ITR_INTERVAL_DEFAULT	500 /* 500us */
> >>
> >>   /* Loopback operation modes */
> >> -/* 82599 specific loopback operation types */
> >> -#define IXGBE_LPBK_82599_NONE   0x0 /* Default value. Loopback is
> disabled.
> >> */
> >> -#define IXGBE_LPBK_82599_TX_RX  0x1 /* Tx->Rx loopback operation is
> >> enabled. */
> >> +#define IXGBE_LPBK_NONE   0x0 /* Default value. Loopback is disabled.
> */
> >> +#define IXGBE_LPBK_TX_RX  0x1 /* Tx->Rx loopback operation is
> enabled.
> >> +*/
> >>
> >>   #define IXGBE_MAX_JUMBO_FRAME_SIZE      0x2600 /* Maximum
> Jumbo
> >> frame size. */
> >>
> >> diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c
> >> b/drivers/net/ixgbe/ixgbe_rxtx.c index 9a79d18..0ef7fdf 100644
> >> --- a/drivers/net/ixgbe/ixgbe_rxtx.c
> >> +++ b/drivers/net/ixgbe/ixgbe_rxtx.c
> >> @@ -4879,10 +4879,14 @@ ixgbe_dev_rx_init(struct rte_eth_dev *dev)
> >>   		hlreg0 &= ~IXGBE_HLREG0_JUMBOEN;
> >>
> >>   	/*
> >> -	 * If loopback mode is configured for 82599, set LPBK bit.
> >> +	 * If loopback mode is configured, set LPBK bit.
> >>   	 */
> >> -	if (hw->mac.type == ixgbe_mac_82599EB &&
> >> -			dev->data->dev_conf.lpbk_mode ==
> IXGBE_LPBK_82599_TX_RX)
> >> +	if ((hw->mac.type == ixgbe_mac_82599EB ||
> >> +	     hw->mac.type == ixgbe_mac_X540 ||
> >> +	     hw->mac.type == ixgbe_mac_X550 ||
> >> +	     hw->mac.type == ixgbe_mac_X550EM_x ||
> >> +	     hw->mac.type == ixgbe_mac_X550EM_a) &&
> >> +			dev->data->dev_conf.lpbk_mode ==
> IXGBE_LPBK_TX_RX)
> >>   		hlreg0 |= IXGBE_HLREG0_LPBK;
> >>   	else
> >>   		hlreg0 &= ~IXGBE_HLREG0_LPBK;
> >> @@ -5088,6 +5092,29 @@ ixgbe_setup_loopback_link_82599(struct
> >> ixgbe_hw
> >> *hw)
> >>   	msec_delay(50);
> >>   }
> >>
> >> +/*
> >> + * Set up link loopback for X540 / X550 mode Tx->Rx.
> >> + */
> >> +static inline void __attribute__((cold))
> >> +ixgbe_setup_loopback_link_x540_x550(struct ixgbe_hw *hw) {
> >> +	uint32_t macc;
> >> +	PMD_INIT_FUNC_TRACE();
> >> +
> >> +	/* datasheet 15.2.1: MACC.FLU = 1 (force link up) */
> >> +	macc = IXGBE_READ_REG(hw, IXGBE_MACC);
> >> +	macc |= IXGBE_MACC_FLU;
> >> +	IXGBE_WRITE_REG(hw, IXGBE_MACC, macc);
> >> +
> >> +	/* Restart link */
> >> +	IXGBE_WRITE_REG(hw,
> >> +			IXGBE_AUTOC,
> >> +			IXGBE_AUTOC_LMS_10G_LINK_NO_AN |
> IXGBE_AUTOC_FLU);
> >> +
> >> +	hw->mac.ops.release_swfw_sync(hw, IXGBE_GSSR_MAC_CSR_SM);
> >> +	msec_delay(50);
> >> +}
> >> +
> >>
> >>   /*
> >>    * Start Transmit and Receive Units.
> >> @@ -5148,10 +5175,16 @@ ixgbe_dev_rxtx_start(struct rte_eth_dev
> *dev)
> >>   	rxctrl |= IXGBE_RXCTRL_RXEN;
> >>   	hw->mac.ops.enable_rx_dma(hw, rxctrl);
> >>
> >> -	/* If loopback mode is enabled for 82599, set up the link accordingly
> */
> >> -	if (hw->mac.type == ixgbe_mac_82599EB &&
> >> -			dev->data->dev_conf.lpbk_mode ==
> IXGBE_LPBK_82599_TX_RX)
> >> -		ixgbe_setup_loopback_link_82599(hw);
> >> +	/* If loopback mode is enabled, set up the link accordingly */
> >> +	if (dev->data->dev_conf.lpbk_mode == IXGBE_LPBK_TX_RX) {
> >> +		if (hw->mac.type == ixgbe_mac_82599EB)
> >> +			ixgbe_setup_loopback_link_82599(hw);
> >> +		else if (hw->mac.type == ixgbe_mac_X540 ||
> >> +		     hw->mac.type == ixgbe_mac_X550 ||
> >> +		     hw->mac.type == ixgbe_mac_X550EM_x ||
> >> +		     hw->mac.type == ixgbe_mac_X550EM_a)
> >> +			ixgbe_setup_loopback_link_x540_x550(hw);
> >> +	}
> >>
> >>   #ifdef RTE_LIBRTE_SECURITY
> >>   	if ((dev->data->dev_conf.rxmode.offloads &
> >> --
> >> 2.10.2
> >

  reply	other threads:[~2019-01-08 10:10 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-01-02 16:00 Julien Meunier
2019-01-07  6:53 ` Zhang, Qi Z
2019-01-07 15:53   ` Meunier, Julien (Nokia - FR/Paris-Saclay)
2019-01-08 10:10     ` Zhao1, Wei [this message]
2019-01-08 12:39       ` Zhang, Qi Z
2019-01-08 20:18         ` Meunier, Julien (Nokia - FR/Paris-Saclay)
2019-01-09  1:30         ` Zhao1, Wei
2019-01-07 10:05 ` Zhao1, Wei
2019-01-07 16:04   ` Meunier, Julien (Nokia - FR/Paris-Saclay)
2019-01-08  3:10     ` Zhao1, Wei
2019-02-07 17:30 ` [dpdk-dev] [PATCH 1/2] net/ixgbe: do not start on unsupported loopback mode Julien Meunier
2019-02-07 17:30   ` [dpdk-dev] [PATCH 2/2] net/ixgbe: add support of loopback for X540/X550 Julien Meunier
2019-02-12  6:37     ` Zhao1, Wei
2019-02-18 11:20       ` Meunier, Julien (Nokia - FR/Paris-Saclay)
2019-02-12  6:36   ` [dpdk-dev] [PATCH 1/2] net/ixgbe: do not start on unsupported loopback mode Zhao1, Wei
2019-02-20 21:05 ` [dpdk-dev] [PATCH v3 " Julien Meunier
2019-02-20 21:05   ` [dpdk-dev] [PATCH v3 2/2] net/ixgbe: add support of loopback for X540/X550 Julien Meunier
2019-02-21  2:30     ` Zhao1, Wei
2019-03-01  7:46       ` Zhang, Qi Z
2019-03-01  7:45   ` [dpdk-dev] [PATCH v3 1/2] net/ixgbe: do not start on unsupported loopback mode 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=A2573D2ACFCADC41BB3BE09C6DE313CA07E9ECFD@PGSMSX103.gar.corp.intel.com \
    --to=wei.zhao1@intel.com \
    --cc=dev@dpdk.org \
    --cc=julien.meunier@nokia.com \
    --cc=konstantin.ananyev@intel.com \
    --cc=qi.z.zhang@intel.com \
    --cc=wenzhuo.lu@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).