DPDK patches and discussions
 help / color / mirror / Atom feed
From: Andrew Rybchenko <arybchenko@solarflare.com>
To: <dev@dpdk.org>
Cc: Gautam Dawar <gdawar@solarflare.com>
Subject: [dpdk-dev] [PATCH 22/29] net/sfc/base: factor out upstream port vAdaptor allocation
Date: Mon, 10 Jun 2019 08:38:37 +0100	[thread overview]
Message-ID: <1560152324-20538-23-git-send-email-arybchenko@solarflare.com> (raw)
In-Reply-To: <1560152324-20538-1-git-send-email-arybchenko@solarflare.com>

From: Gautam Dawar <gdawar@solarflare.com>

Separate out vAdaptor allocation from ef10_nic_init() as it is not
required for SR-IOV use case. In case of SR-IOV, vAdaptor is allocated
early along with vSwitch creation and vPort configuration.

Signed-off-by: Gautam Dawar <gdawar@solarflare.com>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
---
 drivers/net/sfc/base/ef10_nic.c | 110 ++++++++++++++++++++++++++--------------
 1 file changed, 72 insertions(+), 38 deletions(-)

diff --git a/drivers/net/sfc/base/ef10_nic.c b/drivers/net/sfc/base/ef10_nic.c
index 7eada57..0cf9ddd 100644
--- a/drivers/net/sfc/base/ef10_nic.c
+++ b/drivers/net/sfc/base/ef10_nic.c
@@ -2220,6 +2220,58 @@
 	return (rc);
 }
 
+static	__checkReturn	efx_rc_t
+ef10_upstream_port_vadaptor_alloc(
+	__in		efx_nic_t *enp)
+{
+	uint32_t retry;
+	uint32_t delay_us;
+	efx_rc_t rc;
+
+	/*
+	 * On a VF, this may fail with MC_CMD_ERR_NO_EVB_PORT (ENOENT) if the PF
+	 * driver has yet to bring up the EVB port. See bug 56147. In this case,
+	 * retry the request several times after waiting a while. The wait time
+	 * between retries starts small (10ms) and exponentially increases.
+	 * Total wait time is a little over two seconds. Retry logic in the
+	 * client driver may mean this whole loop is repeated if it continues to
+	 * fail.
+	 */
+	retry = 0;
+	delay_us = 10000;
+	while ((rc = efx_mcdi_vadaptor_alloc(enp, EVB_PORT_ID_ASSIGNED)) != 0) {
+		if (EFX_PCI_FUNCTION_IS_PF(&enp->en_nic_cfg) ||
+		    (rc != ENOENT)) {
+			/*
+			 * Do not retry alloc for PF, or for other errors on
+			 * a VF.
+			 */
+			goto fail1;
+		}
+
+		/* VF startup before PF is ready. Retry allocation. */
+		if (retry > 5) {
+			/* Too many attempts */
+			rc = EINVAL;
+			goto fail2;
+		}
+		EFSYS_PROBE1(mcdi_no_evb_port_retry, int, retry);
+		EFSYS_SLEEP(delay_us);
+		retry++;
+		if (delay_us < 500000)
+			delay_us <<= 2;
+	}
+
+	return (0);
+
+fail2:
+	EFSYS_PROBE(fail2);
+fail1:
+	EFSYS_PROBE1(fail1, efx_rc_t, rc);
+
+	return (rc);
+}
+
 	__checkReturn	efx_rc_t
 ef10_nic_init(
 	__in		efx_nic_t *enp)
@@ -2228,12 +2280,13 @@
 	uint32_t min_vi_count, max_vi_count;
 	uint32_t vi_count, vi_base, vi_shift;
 	uint32_t i;
-	uint32_t retry;
-	uint32_t delay_us;
 	uint32_t vi_window_size;
 	efx_rc_t rc;
+	boolean_t alloc_vadaptor = B_TRUE;
 
-	EFSYS_ASSERT(EFX_FAMILY_IS_EF10(enp));
+	EFSYS_ASSERT(enp->en_family == EFX_FAMILY_HUNTINGTON ||
+	    enp->en_family == EFX_FAMILY_MEDFORD ||
+	    enp->en_family == EFX_FAMILY_MEDFORD2);
 
 	/* Enable reporting of some events (e.g. link change) */
 	if ((rc = efx_mcdi_log_ctrl(enp)) != 0)
@@ -2330,48 +2383,29 @@
 	}
 
 	/*
-	 * Allocate a vAdaptor attached to our upstream vPort/pPort.
-	 *
-	 * On a VF, this may fail with MC_CMD_ERR_NO_EVB_PORT (ENOENT) if the PF
-	 * driver has yet to bring up the EVB port. See bug 56147. In this case,
-	 * retry the request several times after waiting a while. The wait time
-	 * between retries starts small (10ms) and exponentially increases.
-	 * Total wait time is a little over two seconds. Retry logic in the
-	 * client driver may mean this whole loop is repeated if it continues to
-	 * fail.
+	 * For SR-IOV use case, vAdaptor is allocated for PF and associated VFs
+	 * during NIC initialization when vSwitch is created and vports are
+	 * allocated. Hence, skip vAdaptor allocation for EVB and update vport
+	 * id in NIC structure with the one allocated for PF.
 	 */
-	retry = 0;
-	delay_us = 10000;
-	while ((rc = efx_mcdi_vadaptor_alloc(enp, EVB_PORT_ID_ASSIGNED)) != 0) {
-		if (EFX_PCI_FUNCTION_IS_PF(&enp->en_nic_cfg) ||
-		    (rc != ENOENT)) {
-			/*
-			 * Do not retry alloc for PF, or for other errors on
-			 * a VF.
-			 */
-			goto fail5;
-		}
-
-		/* VF startup before PF is ready. Retry allocation. */
-		if (retry > 5) {
-			/* Too many attempts */
-			rc = EINVAL;
-			goto fail6;
-		}
-		EFSYS_PROBE1(mcdi_no_evb_port_retry, int, retry);
-		EFSYS_SLEEP(delay_us);
-		retry++;
-		if (delay_us < 500000)
-			delay_us <<= 2;
-	}
 
 	enp->en_vport_id = EVB_PORT_ID_ASSIGNED;
+#if EFSYS_OPT_EVB
+	if ((enp->en_vswitchp != NULL) && (enp->en_vswitchp->ev_evcp != NULL)) {
+		/* For EVB use vport allocated on vswitch */
+		enp->en_vport_id = enp->en_vswitchp->ev_evcp->evc_vport_id;
+		alloc_vadaptor = B_FALSE;
+	}
+#endif
+	if (alloc_vadaptor != B_FALSE) {
+		/* Allocate a vAdaptor attached to our upstream vPort/pPort */
+		if ((rc = ef10_upstream_port_vadaptor_alloc(enp)) != 0)
+			goto fail5;
+	}
 	enp->en_nic_cfg.enc_mcdi_max_payload_length = MCDI_CTL_SDU_LEN_MAX_V2;
 
 	return (0);
 
-fail6:
-	EFSYS_PROBE(fail6);
 fail5:
 	EFSYS_PROBE(fail5);
 fail4:
-- 
1.8.3.1


  parent reply	other threads:[~2019-06-10  7:40 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-06-10  7:38 [dpdk-dev] [PATCH 00/29] net/sfc/base: update base driver Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 01/29] net/sfc/base: enable chained multicast on all EF10 cards Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 02/29] net/sfc/base: fix signed/unsigned mismatch errors Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 03/29] net/sfc/base: fix shift by more bits than field width Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 04/29] net/sfc/base: improve code style in sensors decoding Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 05/29] net/sfc/base: do not rely on indirect header inclusion Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 06/29] net/sfc/base: add driver version string registration Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 07/29] net/sfc/base: add capabilities for bundle partition support Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 08/29] net/sfc/base: add extensible NVRAM info function Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 09/29] net/sfc/base: add NVRAM info to API Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 10/29] net/sfc/base: update MCDI headers Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 11/29] net/sfc/base: add firmware ID header Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 12/29] net/sfc/base: support direct FW update for bundle partitions Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 13/29] net/sfc/base: transition to the extensible NVRAM info API Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 14/29] net/sfc/base: add background mode firmware updating Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 15/29] net/sfc/base: add definition of bundle metadata partition Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 16/29] net/sfc/base: add definition of OEM TLV Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 17/29] net/sfc/base: export the zero-based MCDI port number Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 18/29] net/sfc/base: introduce of EVB module for SR-IOV support Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 19/29] net/sfc/base: add MCDI wrappers for vPort and vSwitch in EVB Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 20/29] net/sfc/base: add EVB module vSwitch/vPort/vAdaptor ops Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 21/29] net/sfc/base: implement vSwitch create/destroy Andrew Rybchenko
2019-06-10  7:38 ` Andrew Rybchenko [this message]
2019-06-10  7:38 ` [dpdk-dev] [PATCH 23/29] net/sfc/base: support data path with EVB module Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 24/29] net/sfc/base: support proxy auth operations for SR-IOV Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 25/29] net/sfc/base: implement proxy auth MCDI event handling Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 26/29] net/sfc/base: provide proxy APIs to client drivers Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 27/29] net/sfc/base: provide APIs to configure and reset vPort Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 28/29] net/sfc/base: provide API to fetch vPort statistics Andrew Rybchenko
2019-06-10  7:38 ` [dpdk-dev] [PATCH 29/29] net/sfc/base: add APIs for PTP privilege configuration Andrew Rybchenko
2019-06-18  7:52 ` [dpdk-dev] [PATCH 00/29] net/sfc/base: update base driver Ferruh Yigit
2019-06-23 22:37   ` Thomas Monjalon
2019-06-24  7:53     ` Andrew Rybchenko
2019-06-24 11:17       ` Ferruh Yigit
2019-06-24 13:07         ` Thomas Monjalon

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=1560152324-20538-23-git-send-email-arybchenko@solarflare.com \
    --to=arybchenko@solarflare.com \
    --cc=dev@dpdk.org \
    --cc=gdawar@solarflare.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).