From: "Ouyang, Changchun" <changchun.ouyang@intel.com>
To: Vlad Zolotarov <vladz@cloudius-systems.com>,
"dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v5 4/6] ether: Check VMDq RSS mode
Date: Mon, 12 Jan 2015 03:41:04 +0000 [thread overview]
Message-ID: <F52918179C57134FAEC9EA62FA2F9625119593D6@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <54AFDC77.8040505@cloudius-systems.com>
From: Vlad Zolotarov [mailto:vladz@cloudius-systems.com]
Sent: Friday, January 09, 2015 9:50 PM
To: Ouyang, Changchun; dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v5 4/6] ether: Check VMDq RSS mode
On 01/09/15 07:54, Ouyang, Changchun wrote:
-----Original Message-----
From: Vlad Zolotarov [mailto:vladz@cloudius-systems.com]
Sent: Friday, January 9, 2015 2:49 AM
To: Ouyang, Changchun; dev@dpdk.org<mailto:dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v5 4/6] ether: Check VMDq RSS mode
On 01/08/15 11:19, Vlad Zolotarov wrote:
On 01/07/15 08:32, Ouyang Changchun wrote:
Check mq mode for VMDq RSS, handle it correctly instead of returning
an error; Also remove the limitation of per pool queue number has max
value of 1, because the per pool queue number could be 2 or 4 if it
is VMDq RSS mode;
The number of rxq specified in config will determine the mq mode for
VMDq RSS.
Signed-off-by: Changchun Ouyang <changchun.ouyang@intel.com><mailto:changchun.ouyang@intel.com>
changes in v5:
- Fix '<' issue, it should be '<=' to test rxq number;
- Extract a function to remove the embeded switch-case statement.
---
lib/librte_ether/rte_ethdev.c | 50
++++++++++++++++++++++++++++++++++++++-----
1 file changed, 45 insertions(+), 5 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c
b/lib/librte_ether/rte_ethdev.c index 95f2ceb..8363e26 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -503,6 +503,31 @@ rte_eth_dev_tx_queue_config(struct
rte_eth_dev
*dev, uint16_t nb_queues)
}
static int
+rte_eth_dev_check_vf_rss_rxq_num(uint8_t port_id, uint16_t nb_rx_q)
+{
+ struct rte_eth_dev *dev = &rte_eth_devices[port_id];
+ switch (nb_rx_q) {
+ case 1:
+ case 2:
+ RTE_ETH_DEV_SRIOV(dev).active =
+ ETH_64_POOLS;
+ break;
+ case 4:
+ RTE_ETH_DEV_SRIOV(dev).active =
+ ETH_32_POOLS;
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = nb_rx_q;
+ RTE_ETH_DEV_SRIOV(dev).def_pool_q_idx =
+ dev->pci_dev->max_vfs * nb_rx_q;
+
+ return 0;
+}
+
+static int
rte_eth_dev_check_mq_mode(uint8_t port_id, uint16_t nb_rx_q,
uint16_t nb_tx_q,
const struct rte_eth_conf *dev_conf)
{
@@ -510,8 +535,7 @@ rte_eth_dev_check_mq_mode(uint8_t port_id,
uint16_t nb_rx_q, uint16_t nb_tx_q,
if (RTE_ETH_DEV_SRIOV(dev).active != 0) {
/* check multi-queue mode */
- if ((dev_conf->rxmode.mq_mode == ETH_MQ_RX_RSS) ||
- (dev_conf->rxmode.mq_mode == ETH_MQ_RX_DCB) ||
+ if ((dev_conf->rxmode.mq_mode == ETH_MQ_RX_DCB) ||
(dev_conf->rxmode.mq_mode == ETH_MQ_RX_DCB_RSS) ||
(dev_conf->txmode.mq_mode == ETH_MQ_TX_DCB)) {
/* SRIOV only works in VMDq enable mode */ @@ -525,7
+549,6 @@ rte_eth_dev_check_mq_mode(uint8_t port_id, uint16_t
nb_rx_q, uint16_t nb_tx_q,
}
switch (dev_conf->rxmode.mq_mode) {
- case ETH_MQ_RX_VMDQ_RSS:
case ETH_MQ_RX_VMDQ_DCB:
case ETH_MQ_RX_VMDQ_DCB_RSS:
/* DCB/RSS VMDQ in SRIOV mode, not implement yet */ @@
-534,6 +557,25 @@ rte_eth_dev_check_mq_mode(uint8_t port_id,
uint16_t
nb_rx_q, uint16_t nb_tx_q,
"unsupported VMDQ mq_mode rx %u\n",
port_id, dev_conf->rxmode.mq_mode);
return (-EINVAL);
+ case ETH_MQ_RX_RSS:
+ PMD_DEBUG_TRACE("ethdev port_id=%" PRIu8
+ " SRIOV active, "
+ "Rx mq mode is changed from:"
+ "mq_mode %u into VMDQ mq_mode %u\n",
+ port_id,
+ dev_conf->rxmode.mq_mode,
+ dev->data->dev_conf.rxmode.mq_mode);
+ case ETH_MQ_RX_VMDQ_RSS:
+ dev->data->dev_conf.rxmode.mq_mode =
ETH_MQ_RX_VMDQ_RSS;
+ if (nb_rx_q <= RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool)
+ if (rte_eth_dev_check_vf_rss_rxq_num(port_id,
nb_rx_q) != 0) {
+ PMD_DEBUG_TRACE("ethdev port_id=%d"
+ " SRIOV active, invalid queue"
+ " number for VMDQ RSS\n",
+ port_id);
Some nitpicking here: I'd add the allowed values descriptions to the
error message. Something like: "invalid queue number for VMDQ RSS.
Allowed values are 1, 2 or 4\n".
+ return -EINVAL;
+ }
+ break;
default: /* ETH_MQ_RX_VMDQ_ONLY or ETH_MQ_RX_NONE */
/* if nothing mq mode configure, use default scheme */
dev->data->dev_conf.rxmode.mq_mode =
ETH_MQ_RX_VMDQ_ONLY; @@ -553,8 +595,6 @@
rte_eth_dev_check_mq_mode(uint8_t port_id, uint16_t nb_rx_q,
uint16_t nb_tx_q,
default: /* ETH_MQ_TX_VMDQ_ONLY or ETH_MQ_TX_NONE */
/* if nothing mq mode configure, use default scheme */
dev->data->dev_conf.txmode.mq_mode =
ETH_MQ_TX_VMDQ_ONLY;
- if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
- RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
I'm not sure u may just remove it. These lines originally belong to a
different flow. Are u sure u can remove them like that? What if the
mq_mode is ETH_MQ_RX_NONE and nb_q_per_pool has been initialized
to 4
or 8 in ixgbe_pf_host_init()?
I misread the patch - these lines belong to the txmode.mq_mode switch case.
I think it's ok to remove these really strange lines here. And when I look at it i
think for the similar reasons the similar lines should be removed in the Rx
case too: consider non-RSS case with MQ DCB Tx configuration.
I search code in this function, only one place has
" if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1)
RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;"
The only place is default branch, which is for rx_none, or vmdq_only mode,
Here is a snippet of an rte_eth_dev_check_mq_mode() from the current master:
switch (dev_conf->rxmode.mq_mode) {
case ETH_MQ_RX_VMDQ_RSS:
case ETH_MQ_RX_VMDQ_DCB:
case ETH_MQ_RX_VMDQ_DCB_RSS:
/* DCB/RSS VMDQ in SRIOV mode, not implement yet */
PMD_DEBUG_TRACE("ethdev port_id=%" PRIu8
" SRIOV active, "
"unsupported VMDQ mq_mode rx %u\n",
port_id, dev_conf->rxmode.mq_mode);
return (-EINVAL);
default: /* ETH_MQ_RX_VMDQ_ONLY or ETH_MQ_RX_NONE */
/* if nothing mq mode configure, use default scheme */
dev->data->dev_conf.rxmode.mq_mode = ETH_MQ_RX_VMDQ_ONLY;
if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1) <---- This is one
RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
break;
}
switch (dev_conf->txmode.mq_mode) {
case ETH_MQ_TX_VMDQ_DCB:
/* DCB VMDQ in SRIOV mode, not implement yet */
PMD_DEBUG_TRACE("ethdev port_id=%" PRIu8
" SRIOV active, "
"unsupported VMDQ mq_mode tx %u\n",
port_id, dev_conf->txmode.mq_mode);
return (-EINVAL);
default: /* ETH_MQ_TX_VMDQ_ONLY or ETH_MQ_TX_NONE */
/* if nothing mq mode configure, use default scheme */
dev->data->dev_conf.txmode.mq_mode = ETH_MQ_TX_VMDQ_ONLY;
if (RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool > 1) <------ This is two. This is what your patch is removing
RTE_ETH_DEV_SRIOV(dev).nb_q_per_pool = 1;
break;
}
Changchun: yes you are correct, what I mean in my last response is that only one place AFTER my removal, so there are 2 places before my removal.
no controversial here.
We don't need remove this, as it should assign as 1 because it did use 1 queue per pool.
And why is that? Just because RSS was not enabled? And what if a user wants multiple Tx queues? Mode 1100b of MRQE for instance?
Changchun: I can explain why I need this change(remove the second place) here,
In the txmode, when txmode is ETH_MQ_TX_NONE, but the rx mode could either be ETH_MQ_RX_NONE or
ETH_MQ_RX_VMDQ_RSS, so we could not forcedly set nb_q_per_pool into 1 just hit the condition of txmode is ETH_MQ_TX_NONE,
Because we need consider it is combination of rx mode is ETH_MQ_RX_VMDQ_RSS, and tx mode is ETH_MQ_TX_NONE,
In such a case, the queue number per pool could be 1, or 2, or 4.
In another hand, introducing ETH_MQ_TX_VMDQ_RSS for tx mode, seems very strange, because tx side has no rss feature.
thanks Changchun
next prev parent reply other threads:[~2015-01-12 3:41 UTC|newest]
Thread overview: 144+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-12-15 2:56 [dpdk-dev] [PATCH 0/6] Enable VF RSS for Niantic Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 1/6] ixgbe: Code cleanup Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 3/6] ixgbe: Get VF queue number Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 5/6] ixgbe: Config VF RSS Ouyang Changchun
2014-12-15 2:57 ` [dpdk-dev] [PATCH 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2014-12-15 10:55 ` [dpdk-dev] [PATCH 0/6] Enable VF RSS for Niantic Bruce Richardson
2014-12-16 0:58 ` Ouyang, Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 " Ouyang Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 1/6] ixgbe: Code cleanup Ouyang Changchun
2014-12-24 3:08 ` Zhang, Helin
2014-12-24 3:22 ` Ouyang, Changchun
2014-12-24 3:41 ` Zhang, Helin
2014-12-24 3:50 ` Ouyang, Changchun
2014-12-24 3:53 ` Zhang, Helin
2014-12-24 4:46 ` Ouyang, Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 3/6] ixgbe: Get VF queue number Ouyang Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 5/6] ixgbe: Config VF RSS Ouyang Changchun
2014-12-24 2:56 ` [dpdk-dev] [PATCH v2 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2014-12-24 5:22 ` [dpdk-dev] [PATCH v3 0/6] Enable VF RSS for Niantic Ouyang Changchun
2014-12-24 5:22 ` [dpdk-dev] [PATCH v3 1/6] ixgbe: Code cleanup Ouyang Changchun
2014-12-24 5:23 ` [dpdk-dev] [PATCH v3 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2014-12-24 5:23 ` [dpdk-dev] [PATCH v3 3/6] ixgbe: Get VF queue number Ouyang Changchun
2014-12-24 5:23 ` [dpdk-dev] [PATCH v3 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2014-12-24 5:23 ` [dpdk-dev] [PATCH v3 5/6] ixgbe: Config VF RSS Ouyang Changchun
2014-12-24 10:39 ` Vlad Zolotarov
2014-12-25 2:14 ` Ouyang, Changchun
2014-12-25 13:13 ` Vlad Zolotarov
2014-12-26 2:07 ` Ouyang, Changchun
2014-12-25 2:43 ` Ouyang, Changchun
2014-12-25 13:20 ` Vlad Zolotarov
2014-12-26 1:52 ` Ouyang, Changchun
2014-12-26 6:49 ` Vladislav Zolotarov
2014-12-26 7:26 ` Ouyang, Changchun
2014-12-26 7:37 ` Vladislav Zolotarov
2014-12-26 8:45 ` Ouyang, Changchun
2014-12-28 10:14 ` Vlad Zolotarov
2015-01-05 10:29 ` Bruce Richardson
2015-01-06 1:00 ` Ouyang, Changchun
2014-12-25 13:38 ` Vlad Zolotarov
2014-12-26 1:26 ` Ouyang, Changchun
2015-01-04 2:10 ` Liang, Cunming
2015-01-04 6:25 ` Ouyang, Changchun
2014-12-24 5:23 ` [dpdk-dev] [PATCH v3 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2014-12-24 9:59 ` [dpdk-dev] [PATCH v3 0/6] Enable VF RSS for Niantic Vlad Zolotarov
2014-12-25 1:46 ` Ouyang, Changchun
2015-01-05 10:38 ` Bruce Richardson
2015-01-05 13:02 ` Vlad Zolotarov
2015-01-06 1:11 ` Ouyang, Changchun
2015-01-06 11:18 ` Vlad Zolotarov
2015-01-06 11:18 ` Vlad Zolotarov
2015-01-06 1:04 ` Ouyang, Changchun
2014-12-24 10:49 ` Vlad Zolotarov
2014-12-25 2:26 ` Ouyang, Changchun
2014-12-25 12:46 ` Vlad Zolotarov
2014-12-26 2:37 ` Ouyang, Changchun
[not found] ` <CAOYyTHbrB-VinN5ZEd1tYTnS7_GhCT1jiHiZzNKkQUEJ1rG79w@mail.gmail.com>
2014-12-26 5:16 ` Vladislav Zolotarov
2014-12-26 5:25 ` Ouyang, Changchun
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 " Ouyang Changchun
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 1/6] ixgbe: Code cleanup Ouyang Changchun
2015-01-04 8:22 ` Vlad Zolotarov
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2015-01-04 8:26 ` Vlad Zolotarov
2015-01-04 8:30 ` Vlad Zolotarov
2015-01-04 8:37 ` Ouyang, Changchun
2015-01-04 8:40 ` Vlad Zolotarov
2015-01-04 8:51 ` Ouyang, Changchun
2015-01-04 9:37 ` Vlad Zolotarov
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 3/6] ixgbe: Get VF queue number Ouyang Changchun
2015-01-04 8:38 ` Vlad Zolotarov
2015-01-05 2:59 ` Ouyang, Changchun
2015-01-05 10:07 ` Vlad Zolotarov
2015-01-06 1:54 ` Ouyang, Changchun
2015-01-06 11:26 ` Vlad Zolotarov
2015-01-07 1:18 ` Ouyang, Changchun
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2015-01-04 8:45 ` Vlad Zolotarov
2015-01-04 8:58 ` Ouyang, Changchun
2015-01-04 9:45 ` Vlad Zolotarov
2015-01-05 1:00 ` Ouyang, Changchun
2015-01-05 10:09 ` Vlad Zolotarov
2015-01-06 1:56 ` Ouyang, Changchun
2015-01-06 19:56 ` Vlad Zolotarov
2015-01-07 2:28 ` Ouyang, Changchun
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 5/6] ixgbe: Config VF RSS Ouyang Changchun
2015-01-04 7:18 ` [dpdk-dev] [PATCH v4 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2015-01-04 8:49 ` Vlad Zolotarov
2015-01-04 9:01 ` Ouyang, Changchun
2015-01-04 9:46 ` Vlad Zolotarov
2015-01-05 2:38 ` Ouyang, Changchun
2015-01-05 10:12 ` Vlad Zolotarov
2015-01-06 2:01 ` Ouyang, Changchun
2015-01-06 12:53 ` Vlad Zolotarov
2015-01-07 1:50 ` Ouyang, Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 0/6] Enable VF RSS for Niantic Ouyang Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 1/6] ixgbe: Code cleanup Ouyang Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 3/6] ixgbe: Get VF queue number Ouyang Changchun
2015-01-08 9:01 ` Vlad Zolotarov
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2015-01-08 9:19 ` Vlad Zolotarov
2015-01-08 18:48 ` Vlad Zolotarov
2015-01-09 5:54 ` Ouyang, Changchun
2015-01-09 13:49 ` Vlad Zolotarov
2015-01-12 3:41 ` Ouyang, Changchun [this message]
2015-01-12 13:58 ` Vlad Zolotarov
2015-01-13 1:50 ` Ouyang, Changchun
2015-01-13 9:00 ` Vlad Zolotarov
2015-01-14 0:44 ` Ouyang, Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 5/6] ixgbe: Config VF RSS Ouyang Changchun
2015-01-08 9:43 ` Vlad Zolotarov
2015-01-09 6:07 ` Ouyang, Changchun
2015-01-09 14:01 ` Vlad Zolotarov
2015-01-12 5:11 ` Ouyang, Changchun
2015-01-07 6:32 ` [dpdk-dev] [PATCH v5 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2015-01-08 9:46 ` Vlad Zolotarov
2015-01-08 9:56 ` [dpdk-dev] [PATCH v5 0/6] Enable VF RSS for Niantic Vlad Zolotarov
2015-01-18 21:58 ` Thomas Monjalon
2015-01-19 9:40 ` Vlad Zolotarov
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 " Ouyang Changchun
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 1/6] ixgbe: Code cleanup Ouyang Changchun
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 2/6] ixgbe: Negotiate VF API version Ouyang Changchun
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 3/6] ixgbe: Get VF queue number Ouyang Changchun
2015-01-19 9:13 ` Wodkowski, PawelX
2015-01-20 0:54 ` Ouyang, Changchun
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 4/6] ether: Check VMDq RSS mode Ouyang Changchun
2015-01-12 14:06 ` Vlad Zolotarov
2015-01-18 22:04 ` Thomas Monjalon
2015-01-19 10:31 ` Wodkowski, PawelX
2015-01-20 1:03 ` Ouyang, Changchun
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 5/6] ixgbe: Config VF RSS Ouyang Changchun
2015-01-12 14:04 ` Vlad Zolotarov
2015-01-20 9:35 ` Wodkowski, PawelX
2015-01-21 2:43 ` Ouyang, Changchun
2015-01-21 8:44 ` Wodkowski, PawelX
2015-01-22 12:59 ` Vlad Zolotarov
2015-01-22 13:19 ` Wodkowski, PawelX
2015-01-12 5:59 ` [dpdk-dev] [PATCH v6 6/6] testpmd: Set Rx VMDq RSS mode Ouyang Changchun
2015-01-12 14:05 ` Vlad Zolotarov
2015-01-18 22:24 ` [dpdk-dev] [PATCH v6 0/6] Enable VF RSS for Niantic Thomas Monjalon
2015-01-19 4:51 ` Ouyang, Changchun
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=F52918179C57134FAEC9EA62FA2F9625119593D6@shsmsx102.ccr.corp.intel.com \
--to=changchun.ouyang@intel.com \
--cc=dev@dpdk.org \
--cc=vladz@cloudius-systems.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).