[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <87618083B2453E4A8714035B62D67992501E1DD6@FMSMSX105.amr.corp.intel.com>
Date: Thu, 15 Jan 2015 00:01:24 +0000
From: "Tantilov, Emil S" <emil.s.tantilov@...el.com>
To: Sowmini Varadhan <sowmini.varadhan@...cle.com>,
"Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>,
"Brandeburg, Jesse" <jesse.brandeburg@...el.com>,
"Allan, Bruce W" <bruce.w.allan@...el.com>,
"Wyborny, Carolyn" <carolyn.wyborny@...el.com>,
"Skidmore, Donald C" <donald.c.skidmore@...el.com>,
"Rose, Gregory V" <gregory.v.rose@...el.com>,
"Vick, Matthew" <matthew.vick@...el.com>,
"Ronciak, John" <john.ronciak@...el.com>,
"Williams, Mitch A" <mitch.a.williams@...el.com>
CC: Linux NICS <Linux-nics@...tope.jf.intel.com>,
"e1000-devel@...ts.sourceforge.net"
<e1000-devel@...ts.sourceforge.net>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"sparclinux@...r.kernel.org" <sparclinux@...r.kernel.org>
Subject: RE: [PATCH] ixgbe: Re-enable relaxed ordering as part of
init/restart sequence for non-DCA config
>-----Original Message-----
>From: Sowmini Varadhan [mailto:sowmini.varadhan@...cle.com]
>Relaxed ordering is disabled by default at driver initialization
>and re-enabled when DCA is used. The reason it is disabled was
>due to an issue on some chipsets (see comments in ixgbe_update_tx_dca()).
>But when DCA is not used, RO needs to be re-enabled, else we have
>a serialization bottleneck on platforms like SPARC.
>
>This patch eliminates the bottleneck for ixgbe when DCA is not configured.
>
>Signed-off-by: Sowmini Varadhan <sowmini.varadhan@...cle.com>
>Cc: Emil Tantilov <emil.s.tantilov@...el.com>
>---
> drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c | 1 +
> drivers/net/ethernet/intel/ixgbe/ixgbe_common.c | 20 ++++++++++++++++++++
> drivers/net/ethernet/intel/ixgbe/ixgbe_common.h | 1 +
> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 11 +++++++++++
> drivers/net/ethernet/intel/ixgbe/ixgbe_type.h | 1 +
> drivers/net/ethernet/intel/ixgbe/ixgbe_x540.c | 1 +
> 6 files changed, 35 insertions(+), 0 deletions(-)
>
>diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
>index c5c97b4..85c7a28 100644
>--- a/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
>+++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_82598.c
>@@ -1161,6 +1161,7 @@ static struct ixgbe_mac_operations mac_ops_82598 = {
> .clear_hw_cntrs = &ixgbe_clear_hw_cntrs_generic,
> .get_media_type = &ixgbe_get_media_type_82598,
> .enable_rx_dma = &ixgbe_enable_rx_dma_generic,
>+ .enable_relaxed_ordering = &ixgbe_enable_relaxed_ordering,
The IXGBE_DCA_TXCTRL register for 82598 is at a different offset. Also there is a limit of 16 registers. The function we have in our code for 82598 is as follows:
/**
* ixgbe_enable_relaxed_ordering_82598 - enable relaxed ordering
* @hw: pointer to hardware structure
*
**/
void ixgbe_enable_relaxed_ordering_82598(struct ixgbe_hw *hw)
{
u32 regval;
u32 i;
/* Enable relaxed ordering */
for (i = 0; ((i < hw->mac.max_tx_queues) &&
(i < IXGBE_DCA_MAX_QUEUES_82598)); i++) {
regval = IXGBE_READ_REG(hw, IXGBE_DCA_TXCTRL(i));
regval |= IXGBE_DCA_TXCTRL_DESC_WRO_EN;
IXGBE_WRITE_REG(hw, IXGBE_DCA_TXCTRL(i), regval);
}
for (i = 0; ((i < hw->mac.max_rx_queues) &&
(i < IXGBE_DCA_MAX_QUEUES_82598)); i++) {
regval = IXGBE_READ_REG(hw, IXGBE_DCA_RXCTRL(i));
regval |= IXGBE_DCA_RXCTRL_DATA_WRO_EN |
IXGBE_DCA_RXCTRL_HEAD_WRO_EN;
IXGBE_WRITE_REG(hw, IXGBE_DCA_RXCTRL(i), regval);
}
}
Thanks,
Emil
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists