2013-12-21 13:12:51 +07:00
|
|
|
/*******************************************************************************
|
|
|
|
*
|
|
|
|
* Intel Ethernet Controller XL710 Family Linux Virtual Function Driver
|
2016-01-14 07:51:43 +07:00
|
|
|
* Copyright(c) 2013 - 2016 Intel Corporation.
|
2013-12-21 13:12:51 +07:00
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify it
|
|
|
|
* under the terms and conditions of the GNU General Public License,
|
|
|
|
* version 2, as published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope it will be useful, but WITHOUT
|
|
|
|
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
|
|
|
|
* more details.
|
|
|
|
*
|
2014-04-05 14:46:11 +07:00
|
|
|
* You should have received a copy of the GNU General Public License along
|
|
|
|
* with this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
*
|
2013-12-21 13:12:51 +07:00
|
|
|
* The full GNU General Public License is included in this distribution in
|
|
|
|
* the file called "COPYING".
|
|
|
|
*
|
|
|
|
* Contact Information:
|
|
|
|
* e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
|
|
|
|
* Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
|
|
|
|
*
|
|
|
|
******************************************************************************/
|
|
|
|
|
2014-01-11 11:00:31 +07:00
|
|
|
#include <linux/prefetch.h>
|
2015-01-24 16:58:35 +07:00
|
|
|
#include <net/busy_poll.h>
|
2014-01-11 11:00:31 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
#include "i40evf.h"
|
2017-04-13 15:45:44 +07:00
|
|
|
#include "i40e_trace.h"
|
2014-02-12 08:45:33 +07:00
|
|
|
#include "i40e_prototype.h"
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
static inline __le64 build_ctob(u32 td_cmd, u32 td_offset, unsigned int size,
|
|
|
|
u32 td_tag)
|
|
|
|
{
|
|
|
|
return cpu_to_le64(I40E_TX_DESC_DTYPE_DATA |
|
|
|
|
((u64)td_cmd << I40E_TXD_QW1_CMD_SHIFT) |
|
|
|
|
((u64)td_offset << I40E_TXD_QW1_OFFSET_SHIFT) |
|
|
|
|
((u64)size << I40E_TXD_QW1_TX_BUF_SZ_SHIFT) |
|
|
|
|
((u64)td_tag << I40E_TXD_QW1_L2TAG1_SHIFT));
|
|
|
|
}
|
|
|
|
|
|
|
|
#define I40E_TXD_CMD (I40E_TX_DESC_CMD_EOP | I40E_TX_DESC_CMD_RS)
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_unmap_and_free_tx_resource - Release a Tx buffer
|
|
|
|
* @ring: the ring that owns the buffer
|
|
|
|
* @tx_buffer: the buffer to free
|
|
|
|
**/
|
|
|
|
static void i40e_unmap_and_free_tx_resource(struct i40e_ring *ring,
|
|
|
|
struct i40e_tx_buffer *tx_buffer)
|
|
|
|
{
|
|
|
|
if (tx_buffer->skb) {
|
2016-09-13 04:18:39 +07:00
|
|
|
if (tx_buffer->tx_flags & I40E_TX_FLAGS_FD_SB)
|
|
|
|
kfree(tx_buffer->raw_buf);
|
|
|
|
else
|
|
|
|
dev_kfree_skb_any(tx_buffer->skb);
|
2013-12-21 13:12:51 +07:00
|
|
|
if (dma_unmap_len(tx_buffer, len))
|
|
|
|
dma_unmap_single(ring->dev,
|
|
|
|
dma_unmap_addr(tx_buffer, dma),
|
|
|
|
dma_unmap_len(tx_buffer, len),
|
|
|
|
DMA_TO_DEVICE);
|
|
|
|
} else if (dma_unmap_len(tx_buffer, len)) {
|
|
|
|
dma_unmap_page(ring->dev,
|
|
|
|
dma_unmap_addr(tx_buffer, dma),
|
|
|
|
dma_unmap_len(tx_buffer, len),
|
|
|
|
DMA_TO_DEVICE);
|
|
|
|
}
|
2015-11-07 06:26:03 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
tx_buffer->next_to_watch = NULL;
|
|
|
|
tx_buffer->skb = NULL;
|
|
|
|
dma_unmap_len_set(tx_buffer, len, 0);
|
|
|
|
/* tx_buffer must be completely set up in the transmit path */
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_clean_tx_ring - Free any empty Tx buffers
|
|
|
|
* @tx_ring: ring to be cleaned
|
|
|
|
**/
|
|
|
|
void i40evf_clean_tx_ring(struct i40e_ring *tx_ring)
|
|
|
|
{
|
|
|
|
unsigned long bi_size;
|
|
|
|
u16 i;
|
|
|
|
|
|
|
|
/* ring already cleared, nothing to do */
|
|
|
|
if (!tx_ring->tx_bi)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Free all the Tx ring sk_buffs */
|
|
|
|
for (i = 0; i < tx_ring->count; i++)
|
|
|
|
i40e_unmap_and_free_tx_resource(tx_ring, &tx_ring->tx_bi[i]);
|
|
|
|
|
|
|
|
bi_size = sizeof(struct i40e_tx_buffer) * tx_ring->count;
|
|
|
|
memset(tx_ring->tx_bi, 0, bi_size);
|
|
|
|
|
|
|
|
/* Zero out the descriptor ring */
|
|
|
|
memset(tx_ring->desc, 0, tx_ring->size);
|
|
|
|
|
|
|
|
tx_ring->next_to_use = 0;
|
|
|
|
tx_ring->next_to_clean = 0;
|
|
|
|
|
|
|
|
if (!tx_ring->netdev)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* cleanup Tx queue statistics */
|
2016-09-13 04:18:40 +07:00
|
|
|
netdev_tx_reset_queue(txring_txq(tx_ring));
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_free_tx_resources - Free Tx resources per queue
|
|
|
|
* @tx_ring: Tx descriptor ring for a specific queue
|
|
|
|
*
|
|
|
|
* Free all transmit software resources
|
|
|
|
**/
|
|
|
|
void i40evf_free_tx_resources(struct i40e_ring *tx_ring)
|
|
|
|
{
|
|
|
|
i40evf_clean_tx_ring(tx_ring);
|
|
|
|
kfree(tx_ring->tx_bi);
|
|
|
|
tx_ring->tx_bi = NULL;
|
|
|
|
|
|
|
|
if (tx_ring->desc) {
|
|
|
|
dma_free_coherent(tx_ring->dev, tx_ring->size,
|
|
|
|
tx_ring->desc, tx_ring->dma);
|
|
|
|
tx_ring->desc = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-02-24 12:26:03 +07:00
|
|
|
/**
|
2015-11-07 06:26:02 +07:00
|
|
|
* i40evf_get_tx_pending - how many Tx descriptors not processed
|
|
|
|
* @tx_ring: the ring of descriptors
|
2016-01-16 05:33:12 +07:00
|
|
|
* @in_sw: is tx_pending being checked in SW or HW
|
2015-02-24 12:26:03 +07:00
|
|
|
*
|
2015-11-07 06:26:02 +07:00
|
|
|
* Since there is no access to the ring head register
|
|
|
|
* in XL710, we need to use our local copies
|
2015-02-24 12:26:03 +07:00
|
|
|
**/
|
2016-01-16 05:33:12 +07:00
|
|
|
u32 i40evf_get_tx_pending(struct i40e_ring *ring, bool in_sw)
|
2015-02-24 12:26:03 +07:00
|
|
|
{
|
2015-11-07 06:26:02 +07:00
|
|
|
u32 head, tail;
|
2015-02-24 12:26:03 +07:00
|
|
|
|
2017-03-11 03:22:00 +07:00
|
|
|
head = ring->next_to_clean;
|
2015-11-07 06:26:02 +07:00
|
|
|
tail = readl(ring->tail);
|
|
|
|
|
|
|
|
if (head != tail)
|
|
|
|
return (head < tail) ?
|
|
|
|
tail - head : (tail + ring->count - head);
|
|
|
|
|
|
|
|
return 0;
|
2015-02-24 12:26:03 +07:00
|
|
|
}
|
|
|
|
|
2016-10-12 05:26:54 +07:00
|
|
|
#define WB_STRIDE 4
|
2015-01-10 08:07:19 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
|
|
|
* i40e_clean_tx_irq - Reclaim resources after transmit completes
|
2016-03-08 00:30:03 +07:00
|
|
|
* @vsi: the VSI we care about
|
|
|
|
* @tx_ring: Tx ring to clean
|
|
|
|
* @napi_budget: Used to determine if we are in netpoll
|
2013-12-21 13:12:51 +07:00
|
|
|
*
|
|
|
|
* Returns true if there's any budget left (e.g. the clean is finished)
|
|
|
|
**/
|
2016-03-08 00:30:03 +07:00
|
|
|
static bool i40e_clean_tx_irq(struct i40e_vsi *vsi,
|
|
|
|
struct i40e_ring *tx_ring, int napi_budget)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
|
|
|
u16 i = tx_ring->next_to_clean;
|
|
|
|
struct i40e_tx_buffer *tx_buf;
|
|
|
|
struct i40e_tx_desc *tx_desc;
|
2016-03-08 00:30:03 +07:00
|
|
|
unsigned int total_bytes = 0, total_packets = 0;
|
|
|
|
unsigned int budget = vsi->work_limit;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
tx_buf = &tx_ring->tx_bi[i];
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, i);
|
|
|
|
i -= tx_ring->count;
|
|
|
|
|
|
|
|
do {
|
|
|
|
struct i40e_tx_desc *eop_desc = tx_buf->next_to_watch;
|
|
|
|
|
|
|
|
/* if next_to_watch is not set then there is no work pending */
|
|
|
|
if (!eop_desc)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* prevent any other reads prior to eop_desc */
|
|
|
|
read_barrier_depends();
|
|
|
|
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(clean_tx_irq, tx_ring, tx_desc, tx_buf);
|
2017-03-11 03:22:00 +07:00
|
|
|
/* if the descriptor isn't done, no work yet to do */
|
|
|
|
if (!(eop_desc->cmd_type_offset_bsz &
|
|
|
|
cpu_to_le64(I40E_TX_DESC_DTYPE_DESC_DONE)))
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* clear next_to_watch to prevent false hangs */
|
|
|
|
tx_buf->next_to_watch = NULL;
|
|
|
|
|
|
|
|
/* update the statistics for this packet */
|
|
|
|
total_bytes += tx_buf->bytecount;
|
|
|
|
total_packets += tx_buf->gso_segs;
|
|
|
|
|
|
|
|
/* free the skb */
|
2016-03-08 00:30:03 +07:00
|
|
|
napi_consume_skb(tx_buf->skb, napi_budget);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* unmap skb header data */
|
|
|
|
dma_unmap_single(tx_ring->dev,
|
|
|
|
dma_unmap_addr(tx_buf, dma),
|
|
|
|
dma_unmap_len(tx_buf, len),
|
|
|
|
DMA_TO_DEVICE);
|
|
|
|
|
|
|
|
/* clear tx_buffer data */
|
|
|
|
tx_buf->skb = NULL;
|
|
|
|
dma_unmap_len_set(tx_buf, len, 0);
|
|
|
|
|
|
|
|
/* unmap remaining buffers */
|
|
|
|
while (tx_desc != eop_desc) {
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(clean_tx_irq_unmap,
|
|
|
|
tx_ring, tx_desc, tx_buf);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
tx_buf++;
|
|
|
|
tx_desc++;
|
|
|
|
i++;
|
|
|
|
if (unlikely(!i)) {
|
|
|
|
i -= tx_ring->count;
|
|
|
|
tx_buf = tx_ring->tx_bi;
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* unmap any remaining paged data */
|
|
|
|
if (dma_unmap_len(tx_buf, len)) {
|
|
|
|
dma_unmap_page(tx_ring->dev,
|
|
|
|
dma_unmap_addr(tx_buf, dma),
|
|
|
|
dma_unmap_len(tx_buf, len),
|
|
|
|
DMA_TO_DEVICE);
|
|
|
|
dma_unmap_len_set(tx_buf, len, 0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* move us one more past the eop_desc for start of next pkt */
|
|
|
|
tx_buf++;
|
|
|
|
tx_desc++;
|
|
|
|
i++;
|
|
|
|
if (unlikely(!i)) {
|
|
|
|
i -= tx_ring->count;
|
|
|
|
tx_buf = tx_ring->tx_bi;
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, 0);
|
|
|
|
}
|
|
|
|
|
2015-02-27 16:15:31 +07:00
|
|
|
prefetch(tx_desc);
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* update budget accounting */
|
|
|
|
budget--;
|
|
|
|
} while (likely(budget));
|
|
|
|
|
|
|
|
i += tx_ring->count;
|
|
|
|
tx_ring->next_to_clean = i;
|
|
|
|
u64_stats_update_begin(&tx_ring->syncp);
|
|
|
|
tx_ring->stats.bytes += total_bytes;
|
|
|
|
tx_ring->stats.packets += total_packets;
|
|
|
|
u64_stats_update_end(&tx_ring->syncp);
|
|
|
|
tx_ring->q_vector->tx.total_bytes += total_bytes;
|
|
|
|
tx_ring->q_vector->tx.total_packets += total_packets;
|
|
|
|
|
2015-12-23 05:25:07 +07:00
|
|
|
if (tx_ring->flags & I40E_TXR_FLAGS_WB_ON_ITR) {
|
|
|
|
/* check to see if there are < 4 descriptors
|
|
|
|
* waiting to be written back, then kick the hardware to force
|
|
|
|
* them to be written back in case we stay in NAPI.
|
|
|
|
* In this mode on X722 we do not enable Interrupt.
|
|
|
|
*/
|
2016-06-20 23:10:35 +07:00
|
|
|
unsigned int j = i40evf_get_tx_pending(tx_ring, false);
|
2015-12-23 05:25:07 +07:00
|
|
|
|
|
|
|
if (budget &&
|
2016-10-12 05:26:54 +07:00
|
|
|
((j / WB_STRIDE) == 0) && (j > 0) &&
|
2017-04-19 20:25:55 +07:00
|
|
|
!test_bit(__I40E_VSI_DOWN, vsi->state) &&
|
2015-12-23 05:25:07 +07:00
|
|
|
(I40E_DESC_UNUSED(tx_ring) != tx_ring->count))
|
|
|
|
tx_ring->arm_wb = true;
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:18:40 +07:00
|
|
|
/* notify netdev of completed buffers */
|
|
|
|
netdev_tx_completed_queue(txring_txq(tx_ring),
|
2013-12-21 13:12:51 +07:00
|
|
|
total_packets, total_bytes);
|
|
|
|
|
|
|
|
#define TX_WAKE_THRESHOLD (DESC_NEEDED * 2)
|
|
|
|
if (unlikely(total_packets && netif_carrier_ok(tx_ring->netdev) &&
|
|
|
|
(I40E_DESC_UNUSED(tx_ring) >= TX_WAKE_THRESHOLD))) {
|
|
|
|
/* Make sure that anybody stopping the queue after this
|
|
|
|
* sees the new next_to_clean.
|
|
|
|
*/
|
|
|
|
smp_mb();
|
|
|
|
if (__netif_subqueue_stopped(tx_ring->netdev,
|
|
|
|
tx_ring->queue_index) &&
|
2017-04-19 20:25:55 +07:00
|
|
|
!test_bit(__I40E_VSI_DOWN, vsi->state)) {
|
2013-12-21 13:12:51 +07:00
|
|
|
netif_wake_subqueue(tx_ring->netdev,
|
|
|
|
tx_ring->queue_index);
|
|
|
|
++tx_ring->tx_stats.restart_queue;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-09-25 05:13:15 +07:00
|
|
|
return !!budget;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
2015-01-10 08:07:19 +07:00
|
|
|
/**
|
2016-01-14 07:51:43 +07:00
|
|
|
* i40evf_enable_wb_on_itr - Arm hardware to do a wb, interrupts are not enabled
|
2015-01-10 08:07:19 +07:00
|
|
|
* @vsi: the VSI we care about
|
2016-01-14 07:51:43 +07:00
|
|
|
* @q_vector: the vector on which to enable writeback
|
2015-01-10 08:07:19 +07:00
|
|
|
*
|
|
|
|
**/
|
2016-01-14 07:51:43 +07:00
|
|
|
static void i40e_enable_wb_on_itr(struct i40e_vsi *vsi,
|
|
|
|
struct i40e_q_vector *q_vector)
|
2015-01-10 08:07:19 +07:00
|
|
|
{
|
2015-06-05 23:20:30 +07:00
|
|
|
u16 flags = q_vector->tx.ring[0].flags;
|
2016-01-14 07:51:43 +07:00
|
|
|
u32 val;
|
2015-06-05 23:20:30 +07:00
|
|
|
|
2016-01-14 07:51:43 +07:00
|
|
|
if (!(flags & I40E_TXR_FLAGS_WB_ON_ITR))
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (q_vector->arm_wb_state)
|
|
|
|
return;
|
|
|
|
|
|
|
|
val = I40E_VFINT_DYN_CTLN1_WB_ON_ITR_MASK |
|
|
|
|
I40E_VFINT_DYN_CTLN1_ITR_INDX_MASK; /* set noitr */
|
|
|
|
|
|
|
|
wr32(&vsi->back->hw,
|
|
|
|
I40E_VFINT_DYN_CTLN1(q_vector->v_idx +
|
|
|
|
vsi->base_vector - 1), val);
|
|
|
|
q_vector->arm_wb_state = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_force_wb - Issue SW Interrupt so HW does a wb
|
|
|
|
* @vsi: the VSI we care about
|
|
|
|
* @q_vector: the vector on which to force writeback
|
|
|
|
*
|
|
|
|
**/
|
|
|
|
void i40evf_force_wb(struct i40e_vsi *vsi, struct i40e_q_vector *q_vector)
|
|
|
|
{
|
|
|
|
u32 val = I40E_VFINT_DYN_CTLN1_INTENA_MASK |
|
|
|
|
I40E_VFINT_DYN_CTLN1_ITR_INDX_MASK | /* set noitr */
|
|
|
|
I40E_VFINT_DYN_CTLN1_SWINT_TRIG_MASK |
|
|
|
|
I40E_VFINT_DYN_CTLN1_SW_ITR_INDX_ENA_MASK
|
|
|
|
/* allow 00 to be written to the index */;
|
|
|
|
|
|
|
|
wr32(&vsi->back->hw,
|
|
|
|
I40E_VFINT_DYN_CTLN1(q_vector->v_idx + vsi->base_vector - 1),
|
|
|
|
val);
|
2015-01-10 08:07:19 +07:00
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
|
|
|
* i40e_set_new_dynamic_itr - Find new ITR level
|
|
|
|
* @rc: structure containing ring performance data
|
|
|
|
*
|
2015-09-29 01:16:51 +07:00
|
|
|
* Returns true if ITR changed, false if not
|
|
|
|
*
|
2013-12-21 13:12:51 +07:00
|
|
|
* Stores a new ITR value based on packets and byte counts during
|
|
|
|
* the last interrupt. The advantage of per interrupt computation
|
|
|
|
* is faster updates and more accurate ITR for the current traffic
|
|
|
|
* pattern. Constants in this function were computed based on
|
|
|
|
* theoretical maximum wire speed and thresholds were set based on
|
|
|
|
* testing data as well as attempting to minimize response time
|
|
|
|
* while increasing bulk throughput.
|
|
|
|
**/
|
2015-09-29 01:16:51 +07:00
|
|
|
static bool i40e_set_new_dynamic_itr(struct i40e_ring_container *rc)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
|
|
|
enum i40e_latency_range new_latency_range = rc->latency_range;
|
2015-09-29 01:16:53 +07:00
|
|
|
struct i40e_q_vector *qv = rc->ring->q_vector;
|
2013-12-21 13:12:51 +07:00
|
|
|
u32 new_itr = rc->itr;
|
|
|
|
int bytes_per_int;
|
2015-09-29 01:16:52 +07:00
|
|
|
int usecs;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
if (rc->total_packets == 0 || !rc->itr)
|
2015-09-29 01:16:51 +07:00
|
|
|
return false;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* simple throttlerate management
|
2015-09-29 01:16:53 +07:00
|
|
|
* 0-10MB/s lowest (50000 ints/s)
|
2013-12-21 13:12:51 +07:00
|
|
|
* 10-20MB/s low (20000 ints/s)
|
2015-09-29 01:16:53 +07:00
|
|
|
* 20-1249MB/s bulk (18000 ints/s)
|
|
|
|
* > 40000 Rx packets per second (8000 ints/s)
|
2015-09-29 01:16:52 +07:00
|
|
|
*
|
|
|
|
* The math works out because the divisor is in 10^(-6) which
|
|
|
|
* turns the bytes/us input value into MB/s values, but
|
|
|
|
* make sure to use usecs, as the register values written
|
2015-09-29 01:16:54 +07:00
|
|
|
* are in 2 usec increments in the ITR registers, and make sure
|
|
|
|
* to use the smoothed values that the countdown timer gives us.
|
2013-12-21 13:12:51 +07:00
|
|
|
*/
|
2015-09-29 01:16:54 +07:00
|
|
|
usecs = (rc->itr << 1) * ITR_COUNTDOWN_START;
|
2015-09-29 01:16:52 +07:00
|
|
|
bytes_per_int = rc->total_bytes / usecs;
|
2015-09-29 01:16:54 +07:00
|
|
|
|
2015-06-11 00:42:07 +07:00
|
|
|
switch (new_latency_range) {
|
2013-12-21 13:12:51 +07:00
|
|
|
case I40E_LOWEST_LATENCY:
|
|
|
|
if (bytes_per_int > 10)
|
|
|
|
new_latency_range = I40E_LOW_LATENCY;
|
|
|
|
break;
|
|
|
|
case I40E_LOW_LATENCY:
|
|
|
|
if (bytes_per_int > 20)
|
|
|
|
new_latency_range = I40E_BULK_LATENCY;
|
|
|
|
else if (bytes_per_int <= 10)
|
|
|
|
new_latency_range = I40E_LOWEST_LATENCY;
|
|
|
|
break;
|
|
|
|
case I40E_BULK_LATENCY:
|
2015-09-29 01:16:53 +07:00
|
|
|
case I40E_ULTRA_LATENCY:
|
2015-06-11 00:42:07 +07:00
|
|
|
default:
|
|
|
|
if (bytes_per_int <= 20)
|
|
|
|
new_latency_range = I40E_LOW_LATENCY;
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
}
|
2015-09-29 01:16:53 +07:00
|
|
|
|
|
|
|
/* this is to adjust RX more aggressively when streaming small
|
|
|
|
* packets. The value of 40000 was picked as it is just beyond
|
|
|
|
* what the hardware can receive per second if in low latency
|
|
|
|
* mode.
|
|
|
|
*/
|
|
|
|
#define RX_ULTRA_PACKET_RATE 40000
|
|
|
|
|
|
|
|
if ((((rc->total_packets * 1000000) / usecs) > RX_ULTRA_PACKET_RATE) &&
|
|
|
|
(&qv->rx == rc))
|
|
|
|
new_latency_range = I40E_ULTRA_LATENCY;
|
|
|
|
|
2015-06-11 00:42:07 +07:00
|
|
|
rc->latency_range = new_latency_range;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
switch (new_latency_range) {
|
|
|
|
case I40E_LOWEST_LATENCY:
|
2015-09-29 01:16:53 +07:00
|
|
|
new_itr = I40E_ITR_50K;
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
case I40E_LOW_LATENCY:
|
|
|
|
new_itr = I40E_ITR_20K;
|
|
|
|
break;
|
|
|
|
case I40E_BULK_LATENCY:
|
2015-09-29 01:16:53 +07:00
|
|
|
new_itr = I40E_ITR_18K;
|
|
|
|
break;
|
|
|
|
case I40E_ULTRA_LATENCY:
|
2013-12-21 13:12:51 +07:00
|
|
|
new_itr = I40E_ITR_8K;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
rc->total_bytes = 0;
|
|
|
|
rc->total_packets = 0;
|
2015-09-29 01:16:51 +07:00
|
|
|
|
|
|
|
if (new_itr != rc->itr) {
|
|
|
|
rc->itr = new_itr;
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
2015-11-19 08:35:42 +07:00
|
|
|
/**
|
2013-12-21 13:12:51 +07:00
|
|
|
* i40evf_setup_tx_descriptors - Allocate the Tx descriptors
|
|
|
|
* @tx_ring: the tx ring to set up
|
|
|
|
*
|
|
|
|
* Return 0 on success, negative on error
|
|
|
|
**/
|
|
|
|
int i40evf_setup_tx_descriptors(struct i40e_ring *tx_ring)
|
|
|
|
{
|
|
|
|
struct device *dev = tx_ring->dev;
|
|
|
|
int bi_size;
|
|
|
|
|
|
|
|
if (!dev)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2015-06-19 22:56:30 +07:00
|
|
|
/* warn if we are about to overwrite the pointer */
|
|
|
|
WARN_ON(tx_ring->tx_bi);
|
2013-12-21 13:12:51 +07:00
|
|
|
bi_size = sizeof(struct i40e_tx_buffer) * tx_ring->count;
|
|
|
|
tx_ring->tx_bi = kzalloc(bi_size, GFP_KERNEL);
|
|
|
|
if (!tx_ring->tx_bi)
|
|
|
|
goto err;
|
|
|
|
|
|
|
|
/* round up to nearest 4K */
|
|
|
|
tx_ring->size = tx_ring->count * sizeof(struct i40e_tx_desc);
|
|
|
|
tx_ring->size = ALIGN(tx_ring->size, 4096);
|
|
|
|
tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size,
|
|
|
|
&tx_ring->dma, GFP_KERNEL);
|
|
|
|
if (!tx_ring->desc) {
|
|
|
|
dev_info(dev, "Unable to allocate memory for the Tx descriptor ring, size=%d\n",
|
|
|
|
tx_ring->size);
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
tx_ring->next_to_use = 0;
|
|
|
|
tx_ring->next_to_clean = 0;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
err:
|
|
|
|
kfree(tx_ring->tx_bi);
|
|
|
|
tx_ring->tx_bi = NULL;
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_clean_rx_ring - Free Rx buffers
|
|
|
|
* @rx_ring: ring to be cleaned
|
|
|
|
**/
|
|
|
|
void i40evf_clean_rx_ring(struct i40e_ring *rx_ring)
|
|
|
|
{
|
|
|
|
unsigned long bi_size;
|
|
|
|
u16 i;
|
|
|
|
|
|
|
|
/* ring already cleared, nothing to do */
|
|
|
|
if (!rx_ring->rx_bi)
|
|
|
|
return;
|
|
|
|
|
2017-02-10 14:40:25 +07:00
|
|
|
if (rx_ring->skb) {
|
|
|
|
dev_kfree_skb(rx_ring->skb);
|
|
|
|
rx_ring->skb = NULL;
|
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* Free all the Rx ring sk_buffs */
|
|
|
|
for (i = 0; i < rx_ring->count; i++) {
|
2016-04-19 01:33:46 +07:00
|
|
|
struct i40e_rx_buffer *rx_bi = &rx_ring->rx_bi[i];
|
|
|
|
|
|
|
|
if (!rx_bi->page)
|
|
|
|
continue;
|
|
|
|
|
2017-01-31 03:29:35 +07:00
|
|
|
/* Invalidate cache lines that may have been written to by
|
|
|
|
* device so that we avoid corrupting memory.
|
|
|
|
*/
|
|
|
|
dma_sync_single_range_for_cpu(rx_ring->dev,
|
|
|
|
rx_bi->dma,
|
|
|
|
rx_bi->page_offset,
|
2017-04-05 18:51:01 +07:00
|
|
|
rx_ring->rx_buf_len,
|
2017-01-31 03:29:35 +07:00
|
|
|
DMA_FROM_DEVICE);
|
|
|
|
|
|
|
|
/* free resources associated with mapping */
|
|
|
|
dma_unmap_page_attrs(rx_ring->dev, rx_bi->dma,
|
2017-04-05 18:51:01 +07:00
|
|
|
i40e_rx_pg_size(rx_ring),
|
2017-01-31 03:29:35 +07:00
|
|
|
DMA_FROM_DEVICE,
|
|
|
|
I40E_RX_DMA_ATTR);
|
2017-04-05 18:51:01 +07:00
|
|
|
|
2017-02-22 06:55:39 +07:00
|
|
|
__page_frag_cache_drain(rx_bi->page, rx_bi->pagecnt_bias);
|
2016-04-19 01:33:46 +07:00
|
|
|
|
|
|
|
rx_bi->page = NULL;
|
|
|
|
rx_bi->page_offset = 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
bi_size = sizeof(struct i40e_rx_buffer) * rx_ring->count;
|
|
|
|
memset(rx_ring->rx_bi, 0, bi_size);
|
|
|
|
|
|
|
|
/* Zero out the descriptor ring */
|
|
|
|
memset(rx_ring->desc, 0, rx_ring->size);
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_ring->next_to_alloc = 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
rx_ring->next_to_clean = 0;
|
|
|
|
rx_ring->next_to_use = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_free_rx_resources - Free Rx resources
|
|
|
|
* @rx_ring: ring to clean the resources from
|
|
|
|
*
|
|
|
|
* Free all receive software resources
|
|
|
|
**/
|
|
|
|
void i40evf_free_rx_resources(struct i40e_ring *rx_ring)
|
|
|
|
{
|
|
|
|
i40evf_clean_rx_ring(rx_ring);
|
|
|
|
kfree(rx_ring->rx_bi);
|
|
|
|
rx_ring->rx_bi = NULL;
|
|
|
|
|
|
|
|
if (rx_ring->desc) {
|
|
|
|
dma_free_coherent(rx_ring->dev, rx_ring->size,
|
|
|
|
rx_ring->desc, rx_ring->dma);
|
|
|
|
rx_ring->desc = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_setup_rx_descriptors - Allocate Rx descriptors
|
|
|
|
* @rx_ring: Rx descriptor ring (for a specific queue) to setup
|
|
|
|
*
|
|
|
|
* Returns 0 on success, negative on failure
|
|
|
|
**/
|
|
|
|
int i40evf_setup_rx_descriptors(struct i40e_ring *rx_ring)
|
|
|
|
{
|
|
|
|
struct device *dev = rx_ring->dev;
|
|
|
|
int bi_size;
|
|
|
|
|
2015-06-19 22:56:30 +07:00
|
|
|
/* warn if we are about to overwrite the pointer */
|
|
|
|
WARN_ON(rx_ring->rx_bi);
|
2013-12-21 13:12:51 +07:00
|
|
|
bi_size = sizeof(struct i40e_rx_buffer) * rx_ring->count;
|
|
|
|
rx_ring->rx_bi = kzalloc(bi_size, GFP_KERNEL);
|
|
|
|
if (!rx_ring->rx_bi)
|
|
|
|
goto err;
|
|
|
|
|
2015-02-10 08:42:31 +07:00
|
|
|
u64_stats_init(&rx_ring->syncp);
|
2015-01-24 16:58:32 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* Round up to nearest 4K */
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_ring->size = rx_ring->count * sizeof(union i40e_32byte_rx_desc);
|
2013-12-21 13:12:51 +07:00
|
|
|
rx_ring->size = ALIGN(rx_ring->size, 4096);
|
|
|
|
rx_ring->desc = dma_alloc_coherent(dev, rx_ring->size,
|
|
|
|
&rx_ring->dma, GFP_KERNEL);
|
|
|
|
|
|
|
|
if (!rx_ring->desc) {
|
|
|
|
dev_info(dev, "Unable to allocate memory for the Rx descriptor ring, size=%d\n",
|
|
|
|
rx_ring->size);
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_ring->next_to_alloc = 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
rx_ring->next_to_clean = 0;
|
|
|
|
rx_ring->next_to_use = 0;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
err:
|
|
|
|
kfree(rx_ring->rx_bi);
|
|
|
|
rx_ring->rx_bi = NULL;
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_release_rx_desc - Store the new tail and head values
|
|
|
|
* @rx_ring: ring to bump
|
|
|
|
* @val: new head index
|
|
|
|
**/
|
|
|
|
static inline void i40e_release_rx_desc(struct i40e_ring *rx_ring, u32 val)
|
|
|
|
{
|
|
|
|
rx_ring->next_to_use = val;
|
2016-04-19 01:33:46 +07:00
|
|
|
|
|
|
|
/* update next to alloc since we have filled the ring */
|
|
|
|
rx_ring->next_to_alloc = val;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* Force memory writes to complete before letting h/w
|
|
|
|
* know there are new descriptors to fetch. (Only
|
|
|
|
* applicable for weak-ordered memory model archs,
|
|
|
|
* such as IA-64).
|
|
|
|
*/
|
|
|
|
wmb();
|
|
|
|
writel(val, rx_ring->tail);
|
|
|
|
}
|
|
|
|
|
2017-04-05 18:51:02 +07:00
|
|
|
/**
|
|
|
|
* i40e_rx_offset - Return expected offset into page to access data
|
|
|
|
* @rx_ring: Ring we are requesting offset of
|
|
|
|
*
|
|
|
|
* Returns the offset value for ring into the data buffer.
|
|
|
|
*/
|
|
|
|
static inline unsigned int i40e_rx_offset(struct i40e_ring *rx_ring)
|
|
|
|
{
|
|
|
|
return ring_uses_build_skb(rx_ring) ? I40E_SKB_PAD : 0;
|
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
2016-04-19 01:33:46 +07:00
|
|
|
* i40e_alloc_mapped_page - recycle or make a new page
|
|
|
|
* @rx_ring: ring to use
|
|
|
|
* @bi: rx_buffer struct to modify
|
2016-01-14 07:51:46 +07:00
|
|
|
*
|
2016-04-19 01:33:46 +07:00
|
|
|
* Returns true if the page was successfully allocated or
|
|
|
|
* reused.
|
2015-01-24 16:58:35 +07:00
|
|
|
**/
|
2016-04-19 01:33:46 +07:00
|
|
|
static bool i40e_alloc_mapped_page(struct i40e_ring *rx_ring,
|
|
|
|
struct i40e_rx_buffer *bi)
|
2015-01-24 16:58:35 +07:00
|
|
|
{
|
2016-04-19 01:33:46 +07:00
|
|
|
struct page *page = bi->page;
|
|
|
|
dma_addr_t dma;
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* since we are recycling buffers we should seldom need to alloc */
|
|
|
|
if (likely(page)) {
|
|
|
|
rx_ring->rx_stats.page_reuse_count++;
|
|
|
|
return true;
|
|
|
|
}
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* alloc new page for storage */
|
2017-04-05 18:51:01 +07:00
|
|
|
page = dev_alloc_pages(i40e_rx_pg_order(rx_ring));
|
2016-04-19 01:33:46 +07:00
|
|
|
if (unlikely(!page)) {
|
|
|
|
rx_ring->rx_stats.alloc_page_failed++;
|
|
|
|
return false;
|
|
|
|
}
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* map page for use */
|
2017-01-31 03:29:35 +07:00
|
|
|
dma = dma_map_page_attrs(rx_ring->dev, page, 0,
|
2017-04-05 18:51:01 +07:00
|
|
|
i40e_rx_pg_size(rx_ring),
|
2017-01-31 03:29:35 +07:00
|
|
|
DMA_FROM_DEVICE,
|
|
|
|
I40E_RX_DMA_ATTR);
|
2016-01-14 07:51:49 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* if mapping failed free memory back to system since
|
|
|
|
* there isn't much point in holding memory we can't use
|
2016-01-14 07:51:49 +07:00
|
|
|
*/
|
2016-04-19 01:33:46 +07:00
|
|
|
if (dma_mapping_error(rx_ring->dev, dma)) {
|
2017-04-05 18:51:01 +07:00
|
|
|
__free_pages(page, i40e_rx_pg_order(rx_ring));
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_ring->rx_stats.alloc_page_failed++;
|
|
|
|
return false;
|
2015-01-24 16:58:35 +07:00
|
|
|
}
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
bi->dma = dma;
|
|
|
|
bi->page = page;
|
2017-04-05 18:51:02 +07:00
|
|
|
bi->page_offset = i40e_rx_offset(rx_ring);
|
2017-03-15 00:15:24 +07:00
|
|
|
|
|
|
|
/* initialize pagecnt_bias to 1 representing we fully own page */
|
2017-02-22 06:55:39 +07:00
|
|
|
bi->pagecnt_bias = 1;
|
2016-01-14 07:51:46 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
return true;
|
|
|
|
}
|
2016-01-14 07:51:46 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/**
|
|
|
|
* i40e_receive_skb - Send a completed packet up the stack
|
|
|
|
* @rx_ring: rx ring in play
|
|
|
|
* @skb: packet to send up
|
|
|
|
* @vlan_tag: vlan tag for packet
|
|
|
|
**/
|
|
|
|
static void i40e_receive_skb(struct i40e_ring *rx_ring,
|
|
|
|
struct sk_buff *skb, u16 vlan_tag)
|
|
|
|
{
|
|
|
|
struct i40e_q_vector *q_vector = rx_ring->q_vector;
|
2016-01-14 07:51:46 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
if ((rx_ring->netdev->features & NETIF_F_HW_VLAN_CTAG_RX) &&
|
|
|
|
(vlan_tag & VLAN_VID_MASK))
|
|
|
|
__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tag);
|
|
|
|
|
|
|
|
napi_gro_receive(&q_vector->napi, skb);
|
2015-01-24 16:58:35 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2016-04-19 01:33:46 +07:00
|
|
|
* i40evf_alloc_rx_buffers - Replace used receive buffers
|
2013-12-21 13:12:51 +07:00
|
|
|
* @rx_ring: ring to place buffers on
|
|
|
|
* @cleaned_count: number of buffers to replace
|
2016-01-14 07:51:46 +07:00
|
|
|
*
|
2016-04-19 01:33:46 +07:00
|
|
|
* Returns false if all allocations were successful, true if any fail
|
2013-12-21 13:12:51 +07:00
|
|
|
**/
|
2016-04-19 01:33:46 +07:00
|
|
|
bool i40evf_alloc_rx_buffers(struct i40e_ring *rx_ring, u16 cleaned_count)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
2016-04-19 01:33:46 +07:00
|
|
|
u16 ntu = rx_ring->next_to_use;
|
2013-12-21 13:12:51 +07:00
|
|
|
union i40e_rx_desc *rx_desc;
|
|
|
|
struct i40e_rx_buffer *bi;
|
|
|
|
|
|
|
|
/* do nothing if no valid netdev defined */
|
|
|
|
if (!rx_ring->netdev || !cleaned_count)
|
2016-01-14 07:51:46 +07:00
|
|
|
return false;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_desc = I40E_RX_DESC(rx_ring, ntu);
|
|
|
|
bi = &rx_ring->rx_bi[ntu];
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
do {
|
|
|
|
if (!i40e_alloc_mapped_page(rx_ring, bi))
|
|
|
|
goto no_buffers;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2017-01-31 03:29:35 +07:00
|
|
|
/* sync the buffer for use by the device */
|
|
|
|
dma_sync_single_range_for_device(rx_ring->dev, bi->dma,
|
|
|
|
bi->page_offset,
|
2017-04-05 18:51:01 +07:00
|
|
|
rx_ring->rx_buf_len,
|
2017-01-31 03:29:35 +07:00
|
|
|
DMA_FROM_DEVICE);
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* Refresh the desc even if buffer_addrs didn't change
|
|
|
|
* because each write-back erases this info.
|
|
|
|
*/
|
|
|
|
rx_desc->read.pkt_addr = cpu_to_le64(bi->dma + bi->page_offset);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_desc++;
|
|
|
|
bi++;
|
|
|
|
ntu++;
|
|
|
|
if (unlikely(ntu == rx_ring->count)) {
|
|
|
|
rx_desc = I40E_RX_DESC(rx_ring, 0);
|
|
|
|
bi = rx_ring->rx_bi;
|
|
|
|
ntu = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* clear the status bits for the next_to_use descriptor */
|
|
|
|
rx_desc->wb.qword1.status_error_len = 0;
|
|
|
|
|
|
|
|
cleaned_count--;
|
|
|
|
} while (cleaned_count);
|
|
|
|
|
|
|
|
if (rx_ring->next_to_use != ntu)
|
|
|
|
i40e_release_rx_desc(rx_ring, ntu);
|
2016-01-14 07:51:46 +07:00
|
|
|
|
|
|
|
return false;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
no_buffers:
|
2016-04-19 01:33:46 +07:00
|
|
|
if (rx_ring->next_to_use != ntu)
|
|
|
|
i40e_release_rx_desc(rx_ring, ntu);
|
2016-01-14 07:51:46 +07:00
|
|
|
|
|
|
|
/* make sure to come back via polling to try again after
|
|
|
|
* allocation failure
|
|
|
|
*/
|
|
|
|
return true;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_rx_checksum - Indicate in skb if hw indicated a good cksum
|
|
|
|
* @vsi: the VSI we care about
|
|
|
|
* @skb: skb currently being received and modified
|
2016-04-19 01:33:46 +07:00
|
|
|
* @rx_desc: the receive descriptor
|
2013-12-21 13:12:51 +07:00
|
|
|
**/
|
|
|
|
static inline void i40e_rx_checksum(struct i40e_vsi *vsi,
|
|
|
|
struct sk_buff *skb,
|
2016-04-19 01:33:46 +07:00
|
|
|
union i40e_rx_desc *rx_desc)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
2016-04-19 01:33:46 +07:00
|
|
|
struct i40e_rx_ptype_decoded decoded;
|
|
|
|
u32 rx_error, rx_status;
|
2016-06-15 05:45:42 +07:00
|
|
|
bool ipv4, ipv6;
|
2016-04-19 01:33:46 +07:00
|
|
|
u8 ptype;
|
|
|
|
u64 qword;
|
|
|
|
|
|
|
|
qword = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
|
|
|
|
ptype = (qword & I40E_RXD_QW1_PTYPE_MASK) >> I40E_RXD_QW1_PTYPE_SHIFT;
|
|
|
|
rx_error = (qword & I40E_RXD_QW1_ERROR_MASK) >>
|
|
|
|
I40E_RXD_QW1_ERROR_SHIFT;
|
|
|
|
rx_status = (qword & I40E_RXD_QW1_STATUS_MASK) >>
|
|
|
|
I40E_RXD_QW1_STATUS_SHIFT;
|
|
|
|
decoded = decode_rx_desc_ptype(ptype);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
skb->ip_summed = CHECKSUM_NONE;
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
skb_checksum_none_assert(skb);
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* Rx csum enabled and ip headers found? */
|
2014-05-20 15:01:43 +07:00
|
|
|
if (!(vsi->netdev->features & NETIF_F_RXCSUM))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* did the hardware decode the packet and checksum? */
|
2015-06-05 03:24:02 +07:00
|
|
|
if (!(rx_status & BIT(I40E_RX_DESC_STATUS_L3L4P_SHIFT)))
|
2014-05-20 15:01:43 +07:00
|
|
|
return;
|
|
|
|
|
|
|
|
/* both known and outer_ip must be set for the below code to work */
|
|
|
|
if (!(decoded.known && decoded.outer_ip))
|
2013-12-21 13:12:51 +07:00
|
|
|
return;
|
|
|
|
|
2016-01-25 12:17:22 +07:00
|
|
|
ipv4 = (decoded.outer_ip == I40E_RX_PTYPE_OUTER_IP) &&
|
|
|
|
(decoded.outer_ip_ver == I40E_RX_PTYPE_OUTER_IPV4);
|
|
|
|
ipv6 = (decoded.outer_ip == I40E_RX_PTYPE_OUTER_IP) &&
|
|
|
|
(decoded.outer_ip_ver == I40E_RX_PTYPE_OUTER_IPV6);
|
2014-05-20 15:01:43 +07:00
|
|
|
|
|
|
|
if (ipv4 &&
|
2015-06-05 03:24:02 +07:00
|
|
|
(rx_error & (BIT(I40E_RX_DESC_ERROR_IPE_SHIFT) |
|
|
|
|
BIT(I40E_RX_DESC_ERROR_EIPE_SHIFT))))
|
2014-05-20 15:01:43 +07:00
|
|
|
goto checksum_fail;
|
|
|
|
|
2014-02-13 18:48:39 +07:00
|
|
|
/* likely incorrect csum if alternate IP extension headers found */
|
2014-05-20 15:01:43 +07:00
|
|
|
if (ipv6 &&
|
2015-06-05 03:24:02 +07:00
|
|
|
rx_status & BIT(I40E_RX_DESC_STATUS_IPV6EXADD_SHIFT))
|
2014-05-20 15:01:43 +07:00
|
|
|
/* don't increment checksum err here, non-fatal err */
|
2013-12-21 13:12:51 +07:00
|
|
|
return;
|
|
|
|
|
2014-05-20 15:01:43 +07:00
|
|
|
/* there was some L4 error, count error and punt packet to the stack */
|
2015-06-05 03:24:02 +07:00
|
|
|
if (rx_error & BIT(I40E_RX_DESC_ERROR_L4E_SHIFT))
|
2014-05-20 15:01:43 +07:00
|
|
|
goto checksum_fail;
|
|
|
|
|
|
|
|
/* handle packets that were not able to be checksummed due
|
|
|
|
* to arrival speed, in this case the stack can compute
|
|
|
|
* the csum.
|
|
|
|
*/
|
2015-06-05 03:24:02 +07:00
|
|
|
if (rx_error & BIT(I40E_RX_DESC_ERROR_PPRS_SHIFT))
|
2013-12-21 13:12:51 +07:00
|
|
|
return;
|
|
|
|
|
2016-06-15 05:45:42 +07:00
|
|
|
/* Only report checksum unnecessary for TCP, UDP, or SCTP */
|
|
|
|
switch (decoded.inner_prot) {
|
|
|
|
case I40E_RX_PTYPE_INNER_PROT_TCP:
|
|
|
|
case I40E_RX_PTYPE_INNER_PROT_UDP:
|
|
|
|
case I40E_RX_PTYPE_INNER_PROT_SCTP:
|
|
|
|
skb->ip_summed = CHECKSUM_UNNECESSARY;
|
|
|
|
/* fall though */
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
2014-05-20 15:01:43 +07:00
|
|
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
checksum_fail:
|
|
|
|
vsi->back->hw_csum_rx_error++;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2015-12-10 06:50:21 +07:00
|
|
|
* i40e_ptype_to_htype - get a hash type
|
2014-02-12 08:45:33 +07:00
|
|
|
* @ptype: the ptype value from the descriptor
|
|
|
|
*
|
|
|
|
* Returns a hash type to be used by skb_set_hash
|
|
|
|
**/
|
2016-04-19 01:33:46 +07:00
|
|
|
static inline int i40e_ptype_to_htype(u8 ptype)
|
2014-02-12 08:45:33 +07:00
|
|
|
{
|
|
|
|
struct i40e_rx_ptype_decoded decoded = decode_rx_desc_ptype(ptype);
|
|
|
|
|
|
|
|
if (!decoded.known)
|
|
|
|
return PKT_HASH_TYPE_NONE;
|
|
|
|
|
|
|
|
if (decoded.outer_ip == I40E_RX_PTYPE_OUTER_IP &&
|
|
|
|
decoded.payload_layer == I40E_RX_PTYPE_PAYLOAD_LAYER_PAY4)
|
|
|
|
return PKT_HASH_TYPE_L4;
|
|
|
|
else if (decoded.outer_ip == I40E_RX_PTYPE_OUTER_IP &&
|
|
|
|
decoded.payload_layer == I40E_RX_PTYPE_PAYLOAD_LAYER_PAY3)
|
|
|
|
return PKT_HASH_TYPE_L3;
|
|
|
|
else
|
|
|
|
return PKT_HASH_TYPE_L2;
|
|
|
|
}
|
|
|
|
|
2015-12-10 06:50:21 +07:00
|
|
|
/**
|
|
|
|
* i40e_rx_hash - set the hash value in the skb
|
|
|
|
* @ring: descriptor ring
|
|
|
|
* @rx_desc: specific descriptor
|
|
|
|
**/
|
|
|
|
static inline void i40e_rx_hash(struct i40e_ring *ring,
|
|
|
|
union i40e_rx_desc *rx_desc,
|
|
|
|
struct sk_buff *skb,
|
|
|
|
u8 rx_ptype)
|
|
|
|
{
|
|
|
|
u32 hash;
|
2016-04-19 01:33:46 +07:00
|
|
|
const __le64 rss_mask =
|
2015-12-10 06:50:21 +07:00
|
|
|
cpu_to_le64((u64)I40E_RX_DESC_FLTSTAT_RSS_HASH <<
|
|
|
|
I40E_RX_DESC_STATUS_FLTSTAT_SHIFT);
|
|
|
|
|
|
|
|
if (ring->netdev->features & NETIF_F_RXHASH)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if ((rx_desc->wb.qword1.status_error_len & rss_mask) == rss_mask) {
|
|
|
|
hash = le32_to_cpu(rx_desc->wb.qword0.hi_dword.rss);
|
|
|
|
skb_set_hash(skb, hash, i40e_ptype_to_htype(rx_ptype));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
2016-04-19 01:33:46 +07:00
|
|
|
* i40evf_process_skb_fields - Populate skb header fields from Rx descriptor
|
|
|
|
* @rx_ring: rx descriptor ring packet is being transacted on
|
|
|
|
* @rx_desc: pointer to the EOP Rx descriptor
|
|
|
|
* @skb: pointer to current skb being populated
|
|
|
|
* @rx_ptype: the packet type decoded by hardware
|
2013-12-21 13:12:51 +07:00
|
|
|
*
|
2016-04-19 01:33:46 +07:00
|
|
|
* This function checks the ring, descriptor, and packet information in
|
|
|
|
* order to populate the hash, checksum, VLAN, protocol, and
|
|
|
|
* other fields within the skb.
|
2013-12-21 13:12:51 +07:00
|
|
|
**/
|
2016-04-19 01:33:46 +07:00
|
|
|
static inline
|
|
|
|
void i40evf_process_skb_fields(struct i40e_ring *rx_ring,
|
|
|
|
union i40e_rx_desc *rx_desc, struct sk_buff *skb,
|
|
|
|
u8 rx_ptype)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
2016-04-19 01:33:46 +07:00
|
|
|
i40e_rx_hash(rx_ring, rx_desc, skb, rx_ptype);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
i40e_rx_checksum(rx_ring->vsi, skb, rx_desc);
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
skb_record_rx_queue(skb, rx_ring->queue_index);
|
2017-02-22 06:55:46 +07:00
|
|
|
|
|
|
|
/* modifies the skb - consumes the enet header */
|
|
|
|
skb->protocol = eth_type_trans(skb, rx_ring->netdev);
|
2016-04-19 01:33:46 +07:00
|
|
|
}
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/**
|
|
|
|
* i40e_cleanup_headers - Correct empty headers
|
|
|
|
* @rx_ring: rx descriptor ring packet is being transacted on
|
|
|
|
* @skb: pointer to current skb being fixed
|
|
|
|
*
|
|
|
|
* Also address the case where we are pulling data in on pages only
|
|
|
|
* and as such no data is present in the skb header.
|
|
|
|
*
|
|
|
|
* In addition if skb is not at least 60 bytes we need to pad it so that
|
|
|
|
* it is large enough to qualify as a valid Ethernet frame.
|
|
|
|
*
|
|
|
|
* Returns true if an error was encountered and skb was freed.
|
|
|
|
**/
|
|
|
|
static bool i40e_cleanup_headers(struct i40e_ring *rx_ring, struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
/* if eth_skb_pad returns an error the skb was freed */
|
|
|
|
if (eth_skb_pad(skb))
|
|
|
|
return true;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
return false;
|
|
|
|
}
|
2015-12-10 06:50:21 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/**
|
|
|
|
* i40e_reuse_rx_page - page flip buffer and store it back on the ring
|
|
|
|
* @rx_ring: rx descriptor ring to store buffers on
|
|
|
|
* @old_buff: donor buffer to have page reused
|
|
|
|
*
|
|
|
|
* Synchronizes page for reuse by the adapter
|
|
|
|
**/
|
|
|
|
static void i40e_reuse_rx_page(struct i40e_ring *rx_ring,
|
|
|
|
struct i40e_rx_buffer *old_buff)
|
|
|
|
{
|
|
|
|
struct i40e_rx_buffer *new_buff;
|
|
|
|
u16 nta = rx_ring->next_to_alloc;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
new_buff = &rx_ring->rx_bi[nta];
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* update, and store next to alloc */
|
|
|
|
nta++;
|
|
|
|
rx_ring->next_to_alloc = (nta < rx_ring->count) ? nta : 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* transfer page from old buffer to new buffer */
|
2017-02-22 06:55:39 +07:00
|
|
|
new_buff->dma = old_buff->dma;
|
|
|
|
new_buff->page = old_buff->page;
|
|
|
|
new_buff->page_offset = old_buff->page_offset;
|
|
|
|
new_buff->pagecnt_bias = old_buff->pagecnt_bias;
|
2016-04-19 01:33:46 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2017-02-10 14:43:30 +07:00
|
|
|
* i40e_page_is_reusable - check if any reuse is possible
|
2016-04-19 01:33:46 +07:00
|
|
|
* @page: page struct to check
|
2017-02-10 14:43:30 +07:00
|
|
|
*
|
|
|
|
* A page is not reusable if it was allocated under low memory
|
|
|
|
* conditions, or it's not in the same NUMA node as this CPU.
|
2016-04-19 01:33:46 +07:00
|
|
|
*/
|
2017-02-10 14:43:30 +07:00
|
|
|
static inline bool i40e_page_is_reusable(struct page *page)
|
2016-04-19 01:33:46 +07:00
|
|
|
{
|
2017-02-10 14:43:30 +07:00
|
|
|
return (page_to_nid(page) == numa_mem_id()) &&
|
|
|
|
!page_is_pfmemalloc(page);
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_can_reuse_rx_page - Determine if this page can be reused by
|
|
|
|
* the adapter for another receive
|
|
|
|
*
|
|
|
|
* @rx_buffer: buffer containing the page
|
|
|
|
*
|
|
|
|
* If page is reusable, rx_buffer->page_offset is adjusted to point to
|
|
|
|
* an unused region in the page.
|
|
|
|
*
|
|
|
|
* For small pages, @truesize will be a constant value, half the size
|
|
|
|
* of the memory at page. We'll attempt to alternate between high and
|
|
|
|
* low halves of the page, with one half ready for use by the hardware
|
|
|
|
* and the other half being consumed by the stack. We use the page
|
|
|
|
* ref count to determine whether the stack has finished consuming the
|
|
|
|
* portion of this page that was passed up with a previous packet. If
|
|
|
|
* the page ref count is >1, we'll assume the "other" half page is
|
|
|
|
* still busy, and this page cannot be reused.
|
|
|
|
*
|
|
|
|
* For larger pages, @truesize will be the actual space used by the
|
|
|
|
* received packet (adjusted upward to an even multiple of the cache
|
|
|
|
* line size). This will advance through the page by the amount
|
|
|
|
* actually consumed by the received packets while there is still
|
|
|
|
* space for a buffer. Each region of larger pages will be used at
|
|
|
|
* most once, after which the page will not be reused.
|
|
|
|
*
|
|
|
|
* In either case, if the page is reusable its refcount is increased.
|
|
|
|
**/
|
2017-03-15 00:15:24 +07:00
|
|
|
static bool i40e_can_reuse_rx_page(struct i40e_rx_buffer *rx_buffer)
|
2017-02-10 14:43:30 +07:00
|
|
|
{
|
2017-03-15 00:15:24 +07:00
|
|
|
unsigned int pagecnt_bias = rx_buffer->pagecnt_bias;
|
|
|
|
struct page *page = rx_buffer->page;
|
2017-02-10 14:43:30 +07:00
|
|
|
|
|
|
|
/* Is any reuse possible? */
|
|
|
|
if (unlikely(!i40e_page_is_reusable(page)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
#if (PAGE_SIZE < 8192)
|
|
|
|
/* if we are only owner of page we can reuse it */
|
2017-03-15 00:15:24 +07:00
|
|
|
if (unlikely((page_count(page) - pagecnt_bias) > 1))
|
2017-02-10 14:43:30 +07:00
|
|
|
return false;
|
|
|
|
#else
|
2017-04-05 18:51:01 +07:00
|
|
|
#define I40E_LAST_OFFSET \
|
|
|
|
(SKB_WITH_OVERHEAD(PAGE_SIZE) - I40E_RXBUFFER_2048)
|
|
|
|
if (rx_buffer->page_offset > I40E_LAST_OFFSET)
|
2017-02-10 14:43:30 +07:00
|
|
|
return false;
|
|
|
|
#endif
|
|
|
|
|
2017-02-22 06:55:39 +07:00
|
|
|
/* If we have drained the page fragment pool we need to update
|
|
|
|
* the pagecnt_bias and page count so that we fully restock the
|
|
|
|
* number of references the driver holds.
|
|
|
|
*/
|
2017-03-15 00:15:24 +07:00
|
|
|
if (unlikely(!pagecnt_bias)) {
|
2017-02-22 06:55:39 +07:00
|
|
|
page_ref_add(page, USHRT_MAX);
|
|
|
|
rx_buffer->pagecnt_bias = USHRT_MAX;
|
|
|
|
}
|
2017-02-10 14:43:30 +07:00
|
|
|
|
|
|
|
return true;
|
2016-04-19 01:33:46 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_add_rx_frag - Add contents of Rx buffer to sk_buff
|
|
|
|
* @rx_ring: rx descriptor ring to transact packets on
|
|
|
|
* @rx_buffer: buffer containing page to add
|
|
|
|
* @skb: sk_buff to place the data into
|
2017-03-15 00:15:24 +07:00
|
|
|
* @size: packet length from rx_desc
|
2016-04-19 01:33:46 +07:00
|
|
|
*
|
|
|
|
* This function will add the data contained in rx_buffer->page to the skb.
|
2017-03-15 00:15:25 +07:00
|
|
|
* It will just attach the page as a frag to the skb.
|
2016-04-19 01:33:46 +07:00
|
|
|
*
|
2017-03-15 00:15:25 +07:00
|
|
|
* The function will then update the page offset.
|
2016-04-19 01:33:46 +07:00
|
|
|
**/
|
2017-03-15 00:15:24 +07:00
|
|
|
static void i40e_add_rx_frag(struct i40e_ring *rx_ring,
|
2016-04-19 01:33:46 +07:00
|
|
|
struct i40e_rx_buffer *rx_buffer,
|
2017-03-15 00:15:24 +07:00
|
|
|
struct sk_buff *skb,
|
|
|
|
unsigned int size)
|
2016-04-19 01:33:46 +07:00
|
|
|
{
|
|
|
|
#if (PAGE_SIZE < 8192)
|
2017-04-05 18:51:01 +07:00
|
|
|
unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
|
2016-04-19 01:33:46 +07:00
|
|
|
#else
|
2017-04-05 18:51:02 +07:00
|
|
|
unsigned int truesize = SKB_DATA_ALIGN(size + i40e_rx_offset(rx_ring));
|
2015-01-24 16:58:35 +07:00
|
|
|
#endif
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2017-03-15 00:15:25 +07:00
|
|
|
skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, rx_buffer->page,
|
|
|
|
rx_buffer->page_offset, size, truesize);
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2017-03-15 00:15:24 +07:00
|
|
|
/* page is being used so we must update the page offset */
|
|
|
|
#if (PAGE_SIZE < 8192)
|
|
|
|
rx_buffer->page_offset ^= truesize;
|
|
|
|
#else
|
|
|
|
rx_buffer->page_offset += truesize;
|
|
|
|
#endif
|
2016-04-19 01:33:46 +07:00
|
|
|
}
|
|
|
|
|
2017-03-15 00:15:23 +07:00
|
|
|
/**
|
|
|
|
* i40e_get_rx_buffer - Fetch Rx buffer and synchronize data for use
|
|
|
|
* @rx_ring: rx descriptor ring to transact packets on
|
|
|
|
* @size: size of buffer to add to skb
|
|
|
|
*
|
|
|
|
* This function will pull an Rx buffer from the ring and synchronize it
|
|
|
|
* for use by the CPU.
|
|
|
|
*/
|
|
|
|
static struct i40e_rx_buffer *i40e_get_rx_buffer(struct i40e_ring *rx_ring,
|
|
|
|
const unsigned int size)
|
|
|
|
{
|
|
|
|
struct i40e_rx_buffer *rx_buffer;
|
|
|
|
|
|
|
|
rx_buffer = &rx_ring->rx_bi[rx_ring->next_to_clean];
|
|
|
|
prefetchw(rx_buffer->page);
|
|
|
|
|
|
|
|
/* we are reusing so sync this buffer for CPU use */
|
|
|
|
dma_sync_single_range_for_cpu(rx_ring->dev,
|
|
|
|
rx_buffer->dma,
|
|
|
|
rx_buffer->page_offset,
|
|
|
|
size,
|
|
|
|
DMA_FROM_DEVICE);
|
|
|
|
|
2017-03-15 00:15:24 +07:00
|
|
|
/* We have pulled a buffer for use, so decrement pagecnt_bias */
|
|
|
|
rx_buffer->pagecnt_bias--;
|
|
|
|
|
2017-03-15 00:15:23 +07:00
|
|
|
return rx_buffer;
|
|
|
|
}
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/**
|
2017-03-15 00:15:25 +07:00
|
|
|
* i40e_construct_skb - Allocate skb and populate it
|
2016-04-19 01:33:46 +07:00
|
|
|
* @rx_ring: rx descriptor ring to transact packets on
|
2017-03-15 00:15:23 +07:00
|
|
|
* @rx_buffer: rx buffer to pull data from
|
2017-03-15 00:15:22 +07:00
|
|
|
* @size: size of buffer to add to skb
|
2016-04-19 01:33:46 +07:00
|
|
|
*
|
2017-03-15 00:15:25 +07:00
|
|
|
* This function allocates an skb. It then populates it with the page
|
|
|
|
* data from the current receive descriptor, taking care to set up the
|
|
|
|
* skb correctly.
|
2016-04-19 01:33:46 +07:00
|
|
|
*/
|
2017-03-15 00:15:25 +07:00
|
|
|
static struct sk_buff *i40e_construct_skb(struct i40e_ring *rx_ring,
|
|
|
|
struct i40e_rx_buffer *rx_buffer,
|
|
|
|
unsigned int size)
|
2016-04-19 01:33:46 +07:00
|
|
|
{
|
2017-03-15 00:15:25 +07:00
|
|
|
void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
|
|
|
|
#if (PAGE_SIZE < 8192)
|
2017-04-05 18:51:01 +07:00
|
|
|
unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
|
2017-03-15 00:15:25 +07:00
|
|
|
#else
|
|
|
|
unsigned int truesize = SKB_DATA_ALIGN(size);
|
|
|
|
#endif
|
|
|
|
unsigned int headlen;
|
|
|
|
struct sk_buff *skb;
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2017-03-15 00:15:25 +07:00
|
|
|
/* prefetch first cache line of first page */
|
|
|
|
prefetch(va);
|
2016-04-19 01:33:46 +07:00
|
|
|
#if L1_CACHE_BYTES < 128
|
2017-03-15 00:15:25 +07:00
|
|
|
prefetch(va + L1_CACHE_BYTES);
|
2016-04-19 01:33:46 +07:00
|
|
|
#endif
|
|
|
|
|
2017-03-15 00:15:25 +07:00
|
|
|
/* allocate a skb to store the frags */
|
|
|
|
skb = __napi_alloc_skb(&rx_ring->q_vector->napi,
|
|
|
|
I40E_RX_HDR_SIZE,
|
|
|
|
GFP_ATOMIC | __GFP_NOWARN);
|
|
|
|
if (unlikely(!skb))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/* Determine available headroom for copy */
|
|
|
|
headlen = size;
|
|
|
|
if (headlen > I40E_RX_HDR_SIZE)
|
|
|
|
headlen = eth_get_headlen(va, I40E_RX_HDR_SIZE);
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2017-03-15 00:15:25 +07:00
|
|
|
/* align pull length to size of long to optimize memcpy performance */
|
|
|
|
memcpy(__skb_put(skb, headlen), va, ALIGN(headlen, sizeof(long)));
|
|
|
|
|
|
|
|
/* update all of the pointers */
|
|
|
|
size -= headlen;
|
|
|
|
if (size) {
|
|
|
|
skb_add_rx_frag(skb, 0, rx_buffer->page,
|
|
|
|
rx_buffer->page_offset + headlen,
|
|
|
|
size, truesize);
|
|
|
|
|
|
|
|
/* buffer is used by skb, update page_offset */
|
|
|
|
#if (PAGE_SIZE < 8192)
|
|
|
|
rx_buffer->page_offset ^= truesize;
|
|
|
|
#else
|
|
|
|
rx_buffer->page_offset += truesize;
|
|
|
|
#endif
|
|
|
|
} else {
|
|
|
|
/* buffer is unused, reset bias back to rx_buffer */
|
|
|
|
rx_buffer->pagecnt_bias++;
|
|
|
|
}
|
2017-03-15 00:15:24 +07:00
|
|
|
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
2017-04-05 18:51:03 +07:00
|
|
|
/**
|
|
|
|
* i40e_build_skb - Build skb around an existing buffer
|
|
|
|
* @rx_ring: Rx descriptor ring to transact packets on
|
|
|
|
* @rx_buffer: Rx buffer to pull data from
|
|
|
|
* @size: size of buffer to add to skb
|
|
|
|
*
|
|
|
|
* This function builds an skb around an existing Rx buffer, taking care
|
|
|
|
* to set up the skb correctly and avoid any memcpy overhead.
|
|
|
|
*/
|
|
|
|
static struct sk_buff *i40e_build_skb(struct i40e_ring *rx_ring,
|
|
|
|
struct i40e_rx_buffer *rx_buffer,
|
|
|
|
unsigned int size)
|
|
|
|
{
|
|
|
|
void *va = page_address(rx_buffer->page) + rx_buffer->page_offset;
|
|
|
|
#if (PAGE_SIZE < 8192)
|
|
|
|
unsigned int truesize = i40e_rx_pg_size(rx_ring) / 2;
|
|
|
|
#else
|
2017-05-15 11:52:00 +07:00
|
|
|
unsigned int truesize = SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) +
|
|
|
|
SKB_DATA_ALIGN(I40E_SKB_PAD + size);
|
2017-04-05 18:51:03 +07:00
|
|
|
#endif
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
|
|
/* prefetch first cache line of first page */
|
|
|
|
prefetch(va);
|
|
|
|
#if L1_CACHE_BYTES < 128
|
|
|
|
prefetch(va + L1_CACHE_BYTES);
|
|
|
|
#endif
|
|
|
|
/* build an skb around the page buffer */
|
|
|
|
skb = build_skb(va - I40E_SKB_PAD, truesize);
|
|
|
|
if (unlikely(!skb))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/* update pointers within the skb to store the data */
|
|
|
|
skb_reserve(skb, I40E_SKB_PAD);
|
|
|
|
__skb_put(skb, size);
|
|
|
|
|
|
|
|
/* buffer is used by skb, update page_offset */
|
|
|
|
#if (PAGE_SIZE < 8192)
|
|
|
|
rx_buffer->page_offset ^= truesize;
|
|
|
|
#else
|
|
|
|
rx_buffer->page_offset += truesize;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
2017-03-15 00:15:24 +07:00
|
|
|
/**
|
|
|
|
* i40e_put_rx_buffer - Clean up used buffer and either recycle or free
|
|
|
|
* @rx_ring: rx descriptor ring to transact packets on
|
|
|
|
* @rx_buffer: rx buffer to pull data from
|
|
|
|
*
|
|
|
|
* This function will clean up the contents of the rx_buffer. It will
|
|
|
|
* either recycle the bufer or unmap it and free the associated resources.
|
|
|
|
*/
|
|
|
|
static void i40e_put_rx_buffer(struct i40e_ring *rx_ring,
|
|
|
|
struct i40e_rx_buffer *rx_buffer)
|
|
|
|
{
|
|
|
|
if (i40e_can_reuse_rx_page(rx_buffer)) {
|
2016-04-19 01:33:46 +07:00
|
|
|
/* hand second half of page back to the ring */
|
|
|
|
i40e_reuse_rx_page(rx_ring, rx_buffer);
|
|
|
|
rx_ring->rx_stats.page_reuse_count++;
|
|
|
|
} else {
|
|
|
|
/* we are not reusing the buffer so unmap it */
|
2017-04-05 18:51:01 +07:00
|
|
|
dma_unmap_page_attrs(rx_ring->dev, rx_buffer->dma,
|
|
|
|
i40e_rx_pg_size(rx_ring),
|
2017-01-31 03:29:35 +07:00
|
|
|
DMA_FROM_DEVICE, I40E_RX_DMA_ATTR);
|
2017-02-22 06:55:39 +07:00
|
|
|
__page_frag_cache_drain(rx_buffer->page,
|
|
|
|
rx_buffer->pagecnt_bias);
|
2016-04-19 01:33:46 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* clear contents of buffer_info */
|
|
|
|
rx_buffer->page = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_is_non_eop - process handling of non-EOP buffers
|
|
|
|
* @rx_ring: Rx ring being processed
|
|
|
|
* @rx_desc: Rx descriptor for current buffer
|
|
|
|
* @skb: Current socket buffer containing buffer in progress
|
|
|
|
*
|
|
|
|
* This function updates next to clean. If the buffer is an EOP buffer
|
|
|
|
* this function exits returning false, otherwise it will place the
|
|
|
|
* sk_buff in the next buffer to be chained and return true indicating
|
|
|
|
* that this is in fact a non-EOP buffer.
|
|
|
|
**/
|
|
|
|
static bool i40e_is_non_eop(struct i40e_ring *rx_ring,
|
|
|
|
union i40e_rx_desc *rx_desc,
|
|
|
|
struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
u32 ntc = rx_ring->next_to_clean + 1;
|
|
|
|
|
|
|
|
/* fetch, update, and store next to clean */
|
|
|
|
ntc = (ntc < rx_ring->count) ? ntc : 0;
|
|
|
|
rx_ring->next_to_clean = ntc;
|
|
|
|
|
|
|
|
prefetch(I40E_RX_DESC(rx_ring, ntc));
|
|
|
|
|
|
|
|
/* if we are the last buffer then there is nothing else to do */
|
|
|
|
#define I40E_RXD_EOF BIT(I40E_RX_DESC_STATUS_EOF_SHIFT)
|
|
|
|
if (likely(i40e_test_staterr(rx_desc, I40E_RXD_EOF)))
|
|
|
|
return false;
|
|
|
|
|
|
|
|
rx_ring->rx_stats.non_eop_descs++;
|
|
|
|
|
|
|
|
return true;
|
2015-01-24 16:58:35 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2016-04-19 01:33:46 +07:00
|
|
|
* i40e_clean_rx_irq - Clean completed descriptors from Rx ring - bounce buf
|
|
|
|
* @rx_ring: rx descriptor ring to transact packets on
|
|
|
|
* @budget: Total limit on number of packets to process
|
|
|
|
*
|
|
|
|
* This function provides a "bounce buffer" approach to Rx interrupt
|
|
|
|
* processing. The advantage to this is that on systems that have
|
|
|
|
* expensive overhead for IOMMU access this provides a means of avoiding
|
|
|
|
* it by maintaining the mapping of the page to the system.
|
2015-01-24 16:58:35 +07:00
|
|
|
*
|
2016-04-19 01:33:46 +07:00
|
|
|
* Returns amount of work completed
|
2015-01-24 16:58:35 +07:00
|
|
|
**/
|
2016-04-19 01:33:46 +07:00
|
|
|
static int i40e_clean_rx_irq(struct i40e_ring *rx_ring, int budget)
|
2015-01-24 16:58:35 +07:00
|
|
|
{
|
|
|
|
unsigned int total_rx_bytes = 0, total_rx_packets = 0;
|
2017-02-10 14:40:25 +07:00
|
|
|
struct sk_buff *skb = rx_ring->skb;
|
2015-01-24 16:58:35 +07:00
|
|
|
u16 cleaned_count = I40E_DESC_UNUSED(rx_ring);
|
2016-01-14 07:51:46 +07:00
|
|
|
bool failure = false;
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
while (likely(total_rx_packets < budget)) {
|
2017-03-15 00:15:23 +07:00
|
|
|
struct i40e_rx_buffer *rx_buffer;
|
2016-04-19 01:33:46 +07:00
|
|
|
union i40e_rx_desc *rx_desc;
|
2017-03-15 00:15:22 +07:00
|
|
|
unsigned int size;
|
2015-01-24 16:58:35 +07:00
|
|
|
u16 vlan_tag;
|
2016-04-19 01:33:46 +07:00
|
|
|
u8 rx_ptype;
|
|
|
|
u64 qword;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* return some buffers to hardware, one at a time is too slow */
|
|
|
|
if (cleaned_count >= I40E_RX_BUFFER_WRITE) {
|
2016-01-14 07:51:46 +07:00
|
|
|
failure = failure ||
|
2016-04-19 01:33:46 +07:00
|
|
|
i40evf_alloc_rx_buffers(rx_ring, cleaned_count);
|
2013-12-21 13:12:51 +07:00
|
|
|
cleaned_count = 0;
|
|
|
|
}
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
rx_desc = I40E_RX_DESC(rx_ring, rx_ring->next_to_clean);
|
|
|
|
|
|
|
|
/* status_error_len will always be zero for unused descriptors
|
|
|
|
* because it's cleared in cleanup, and overlaps with hdr_addr
|
|
|
|
* which is always zero because packet split isn't used, if the
|
2017-03-15 00:15:22 +07:00
|
|
|
* hardware wrote DD then the length will be non-zero
|
2016-04-19 01:33:46 +07:00
|
|
|
*/
|
2017-03-15 00:15:22 +07:00
|
|
|
qword = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2015-01-24 16:58:35 +07:00
|
|
|
/* This memory barrier is needed to keep us from reading
|
2017-03-15 00:15:22 +07:00
|
|
|
* any other fields out of the rx_desc until we have
|
|
|
|
* verified the descriptor has been written back.
|
2015-01-24 16:58:35 +07:00
|
|
|
*/
|
2015-04-09 08:49:43 +07:00
|
|
|
dma_rmb();
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2017-04-10 16:18:43 +07:00
|
|
|
size = (qword & I40E_RXD_QW1_LENGTH_PBUF_MASK) >>
|
|
|
|
I40E_RXD_QW1_LENGTH_PBUF_SHIFT;
|
|
|
|
if (!size)
|
|
|
|
break;
|
|
|
|
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(clean_rx_irq, rx_ring, rx_desc, skb);
|
2017-03-15 00:15:23 +07:00
|
|
|
rx_buffer = i40e_get_rx_buffer(rx_ring, size);
|
|
|
|
|
2017-03-15 00:15:25 +07:00
|
|
|
/* retrieve a buffer from the ring */
|
|
|
|
if (skb)
|
|
|
|
i40e_add_rx_frag(rx_ring, rx_buffer, skb, size);
|
2017-04-05 18:51:03 +07:00
|
|
|
else if (ring_uses_build_skb(rx_ring))
|
|
|
|
skb = i40e_build_skb(rx_ring, rx_buffer, size);
|
2017-03-15 00:15:25 +07:00
|
|
|
else
|
|
|
|
skb = i40e_construct_skb(rx_ring, rx_buffer, size);
|
|
|
|
|
|
|
|
/* exit if we failed to retrieve a buffer */
|
|
|
|
if (!skb) {
|
|
|
|
rx_ring->rx_stats.alloc_buff_failed++;
|
|
|
|
rx_buffer->pagecnt_bias++;
|
2016-04-19 01:33:46 +07:00
|
|
|
break;
|
2017-03-15 00:15:25 +07:00
|
|
|
}
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2017-03-15 00:15:24 +07:00
|
|
|
i40e_put_rx_buffer(rx_ring, rx_buffer);
|
2015-01-24 16:58:35 +07:00
|
|
|
cleaned_count++;
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
if (i40e_is_non_eop(rx_ring, rx_desc, skb))
|
2015-01-24 16:58:35 +07:00
|
|
|
continue;
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* ERR_MASK will only have valid bits if EOP set, and
|
|
|
|
* what we are doing here is actually checking
|
|
|
|
* I40E_RX_DESC_ERROR_RXE_SHIFT, since it is the zeroth bit in
|
|
|
|
* the error field
|
|
|
|
*/
|
|
|
|
if (unlikely(i40e_test_staterr(rx_desc, BIT(I40E_RXD_QW1_ERROR_SHIFT)))) {
|
2015-01-24 16:58:35 +07:00
|
|
|
dev_kfree_skb_any(skb);
|
2017-02-22 06:55:41 +07:00
|
|
|
skb = NULL;
|
2015-01-24 16:58:35 +07:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
2017-02-10 14:40:25 +07:00
|
|
|
if (i40e_cleanup_headers(rx_ring, skb)) {
|
|
|
|
skb = NULL;
|
2016-04-19 01:33:46 +07:00
|
|
|
continue;
|
2017-02-10 14:40:25 +07:00
|
|
|
}
|
2016-04-19 01:33:46 +07:00
|
|
|
|
2015-01-24 16:58:35 +07:00
|
|
|
/* probably a little skewed due to removing CRC */
|
|
|
|
total_rx_bytes += skb->len;
|
|
|
|
|
2016-09-28 01:28:50 +07:00
|
|
|
qword = le64_to_cpu(rx_desc->wb.qword1.status_error_len);
|
|
|
|
rx_ptype = (qword & I40E_RXD_QW1_PTYPE_MASK) >>
|
|
|
|
I40E_RXD_QW1_PTYPE_SHIFT;
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* populate checksum, VLAN, and protocol */
|
|
|
|
i40evf_process_skb_fields(rx_ring, rx_desc, skb, rx_ptype);
|
2015-01-24 16:58:35 +07:00
|
|
|
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
vlan_tag = (qword & BIT(I40E_RX_DESC_STATUS_L2TAG1P_SHIFT)) ?
|
|
|
|
le16_to_cpu(rx_desc->wb.qword0.lo_dword.l2tag1) : 0;
|
|
|
|
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(clean_rx_irq_rx, rx_ring, rx_desc, skb);
|
2015-01-24 16:58:35 +07:00
|
|
|
i40e_receive_skb(rx_ring, skb, vlan_tag);
|
2017-02-10 14:40:25 +07:00
|
|
|
skb = NULL;
|
2015-01-24 16:58:35 +07:00
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* update budget accounting */
|
|
|
|
total_rx_packets++;
|
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2017-02-10 14:40:25 +07:00
|
|
|
rx_ring->skb = skb;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
u64_stats_update_begin(&rx_ring->syncp);
|
|
|
|
rx_ring->stats.packets += total_rx_packets;
|
|
|
|
rx_ring->stats.bytes += total_rx_bytes;
|
|
|
|
u64_stats_update_end(&rx_ring->syncp);
|
|
|
|
rx_ring->q_vector->rx.total_packets += total_rx_packets;
|
|
|
|
rx_ring->q_vector->rx.total_bytes += total_rx_bytes;
|
|
|
|
|
2016-04-19 01:33:46 +07:00
|
|
|
/* guarantee a trip back through this routine if there was a failure */
|
2016-01-14 07:51:46 +07:00
|
|
|
return failure ? budget : total_rx_packets;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
2015-09-29 01:16:51 +07:00
|
|
|
static u32 i40e_buildreg_itr(const int type, const u16 itr)
|
|
|
|
{
|
|
|
|
u32 val;
|
|
|
|
|
|
|
|
val = I40E_VFINT_DYN_CTLN1_INTENA_MASK |
|
2016-01-14 07:51:45 +07:00
|
|
|
/* Don't clear PBA because that can cause lost interrupts that
|
|
|
|
* came in while we were cleaning/polling
|
|
|
|
*/
|
2015-09-29 01:16:51 +07:00
|
|
|
(type << I40E_VFINT_DYN_CTLN1_ITR_INDX_SHIFT) |
|
|
|
|
(itr << I40E_VFINT_DYN_CTLN1_INTERVAL_SHIFT);
|
|
|
|
|
|
|
|
return val;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* a small macro to shorten up some long lines */
|
|
|
|
#define INTREG I40E_VFINT_DYN_CTLN1
|
2016-12-13 06:44:12 +07:00
|
|
|
static inline int get_rx_itr(struct i40e_vsi *vsi, int idx)
|
2016-09-13 04:18:44 +07:00
|
|
|
{
|
|
|
|
struct i40evf_adapter *adapter = vsi->back;
|
|
|
|
|
2016-12-13 06:44:12 +07:00
|
|
|
return adapter->rx_rings[idx].rx_itr_setting;
|
2016-09-13 04:18:44 +07:00
|
|
|
}
|
|
|
|
|
2016-12-13 06:44:12 +07:00
|
|
|
static inline int get_tx_itr(struct i40e_vsi *vsi, int idx)
|
2016-09-13 04:18:44 +07:00
|
|
|
{
|
|
|
|
struct i40evf_adapter *adapter = vsi->back;
|
|
|
|
|
2016-12-13 06:44:12 +07:00
|
|
|
return adapter->tx_rings[idx].tx_itr_setting;
|
2016-09-13 04:18:44 +07:00
|
|
|
}
|
2015-09-29 01:16:51 +07:00
|
|
|
|
2015-06-11 00:42:07 +07:00
|
|
|
/**
|
|
|
|
* i40e_update_enable_itr - Update itr and re-enable MSIX interrupt
|
|
|
|
* @vsi: the VSI we care about
|
|
|
|
* @q_vector: q_vector for which itr is being updated and interrupt enabled
|
|
|
|
*
|
|
|
|
**/
|
|
|
|
static inline void i40e_update_enable_itr(struct i40e_vsi *vsi,
|
|
|
|
struct i40e_q_vector *q_vector)
|
|
|
|
{
|
|
|
|
struct i40e_hw *hw = &vsi->back->hw;
|
2015-09-29 01:16:51 +07:00
|
|
|
bool rx = false, tx = false;
|
|
|
|
u32 rxval, txval;
|
2015-06-11 00:42:07 +07:00
|
|
|
int vector;
|
2016-09-13 04:18:44 +07:00
|
|
|
int idx = q_vector->v_idx;
|
|
|
|
int rx_itr_setting, tx_itr_setting;
|
2015-06-11 00:42:07 +07:00
|
|
|
|
|
|
|
vector = (q_vector->v_idx + vsi->base_vector);
|
2015-09-29 01:16:54 +07:00
|
|
|
|
|
|
|
/* avoid dynamic calculation if in countdown mode OR if
|
|
|
|
* all dynamic is disabled
|
|
|
|
*/
|
2015-09-29 01:16:51 +07:00
|
|
|
rxval = txval = i40e_buildreg_itr(I40E_ITR_NONE, 0);
|
|
|
|
|
2016-12-13 06:44:12 +07:00
|
|
|
rx_itr_setting = get_rx_itr(vsi, idx);
|
|
|
|
tx_itr_setting = get_tx_itr(vsi, idx);
|
2016-09-13 04:18:44 +07:00
|
|
|
|
2015-09-29 01:16:54 +07:00
|
|
|
if (q_vector->itr_countdown > 0 ||
|
2016-09-13 04:18:44 +07:00
|
|
|
(!ITR_IS_DYNAMIC(rx_itr_setting) &&
|
|
|
|
!ITR_IS_DYNAMIC(tx_itr_setting))) {
|
2015-09-29 01:16:54 +07:00
|
|
|
goto enable_int;
|
|
|
|
}
|
|
|
|
|
2016-09-13 04:18:44 +07:00
|
|
|
if (ITR_IS_DYNAMIC(rx_itr_setting)) {
|
2015-09-29 01:16:51 +07:00
|
|
|
rx = i40e_set_new_dynamic_itr(&q_vector->rx);
|
|
|
|
rxval = i40e_buildreg_itr(I40E_RX_ITR, q_vector->rx.itr);
|
2015-06-11 00:42:07 +07:00
|
|
|
}
|
2015-11-19 08:35:42 +07:00
|
|
|
|
2016-09-13 04:18:44 +07:00
|
|
|
if (ITR_IS_DYNAMIC(tx_itr_setting)) {
|
2015-09-29 01:16:51 +07:00
|
|
|
tx = i40e_set_new_dynamic_itr(&q_vector->tx);
|
|
|
|
txval = i40e_buildreg_itr(I40E_TX_ITR, q_vector->tx.itr);
|
|
|
|
}
|
2015-11-19 08:35:42 +07:00
|
|
|
|
2015-09-29 01:16:51 +07:00
|
|
|
if (rx || tx) {
|
|
|
|
/* get the higher of the two ITR adjustments and
|
|
|
|
* use the same value for both ITR registers
|
|
|
|
* when in adaptive mode (Rx and/or Tx)
|
|
|
|
*/
|
|
|
|
u16 itr = max(q_vector->tx.itr, q_vector->rx.itr);
|
|
|
|
|
|
|
|
q_vector->tx.itr = q_vector->rx.itr = itr;
|
|
|
|
txval = i40e_buildreg_itr(I40E_TX_ITR, itr);
|
|
|
|
tx = true;
|
|
|
|
rxval = i40e_buildreg_itr(I40E_RX_ITR, itr);
|
|
|
|
rx = true;
|
2015-06-11 00:42:07 +07:00
|
|
|
}
|
2015-09-29 01:16:51 +07:00
|
|
|
|
|
|
|
/* only need to enable the interrupt once, but need
|
|
|
|
* to possibly update both ITR values
|
|
|
|
*/
|
|
|
|
if (rx) {
|
|
|
|
/* set the INTENA_MSK_MASK so that this first write
|
|
|
|
* won't actually enable the interrupt, instead just
|
|
|
|
* updating the ITR (it's bit 31 PF and VF)
|
|
|
|
*/
|
|
|
|
rxval |= BIT(31);
|
|
|
|
/* don't check _DOWN because interrupt isn't being enabled */
|
|
|
|
wr32(hw, INTREG(vector - 1), rxval);
|
|
|
|
}
|
|
|
|
|
2015-09-29 01:16:54 +07:00
|
|
|
enable_int:
|
2017-04-19 20:25:55 +07:00
|
|
|
if (!test_bit(__I40E_VSI_DOWN, vsi->state))
|
2015-09-29 01:16:51 +07:00
|
|
|
wr32(hw, INTREG(vector - 1), txval);
|
2015-09-29 01:16:54 +07:00
|
|
|
|
|
|
|
if (q_vector->itr_countdown)
|
|
|
|
q_vector->itr_countdown--;
|
|
|
|
else
|
|
|
|
q_vector->itr_countdown = ITR_COUNTDOWN_START;
|
2015-06-11 00:42:07 +07:00
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
|
|
|
* i40evf_napi_poll - NAPI polling Rx/Tx cleanup routine
|
|
|
|
* @napi: napi struct with our devices info in it
|
|
|
|
* @budget: amount of work driver is allowed to do this pass, in packets
|
|
|
|
*
|
|
|
|
* This function will clean all queues associated with a q_vector.
|
|
|
|
*
|
|
|
|
* Returns the amount of work done
|
|
|
|
**/
|
|
|
|
int i40evf_napi_poll(struct napi_struct *napi, int budget)
|
|
|
|
{
|
|
|
|
struct i40e_q_vector *q_vector =
|
|
|
|
container_of(napi, struct i40e_q_vector, napi);
|
|
|
|
struct i40e_vsi *vsi = q_vector->vsi;
|
|
|
|
struct i40e_ring *ring;
|
|
|
|
bool clean_complete = true;
|
2015-01-10 08:07:19 +07:00
|
|
|
bool arm_wb = false;
|
2013-12-21 13:12:51 +07:00
|
|
|
int budget_per_ring;
|
2015-09-25 06:35:47 +07:00
|
|
|
int work_done = 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2017-04-19 20:25:55 +07:00
|
|
|
if (test_bit(__I40E_VSI_DOWN, vsi->state)) {
|
2013-12-21 13:12:51 +07:00
|
|
|
napi_complete(napi);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Since the actual Tx work is minimal, we can give the Tx a larger
|
|
|
|
* budget and be more aggressive about cleaning up the Tx descriptors.
|
|
|
|
*/
|
2015-01-10 08:07:19 +07:00
|
|
|
i40e_for_each_ring(ring, q_vector->tx) {
|
2016-03-08 00:30:03 +07:00
|
|
|
if (!i40e_clean_tx_irq(vsi, ring, budget)) {
|
2016-03-08 00:29:57 +07:00
|
|
|
clean_complete = false;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
arm_wb |= ring->arm_wb;
|
2015-07-24 03:54:34 +07:00
|
|
|
ring->arm_wb = false;
|
2015-01-10 08:07:19 +07:00
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2015-09-24 23:04:26 +07:00
|
|
|
/* Handle case where we are called by netpoll with a budget of 0 */
|
|
|
|
if (budget <= 0)
|
|
|
|
goto tx_only;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* We attempt to distribute budget to each Rx queue fairly, but don't
|
|
|
|
* allow the budget to go below 1 because that would exit polling early.
|
|
|
|
*/
|
|
|
|
budget_per_ring = max(budget/q_vector->num_ringpairs, 1);
|
|
|
|
|
2015-01-24 16:58:35 +07:00
|
|
|
i40e_for_each_ring(ring, q_vector->rx) {
|
2016-04-19 01:33:46 +07:00
|
|
|
int cleaned = i40e_clean_rx_irq(ring, budget_per_ring);
|
2015-09-25 06:35:47 +07:00
|
|
|
|
|
|
|
work_done += cleaned;
|
2016-03-08 00:29:57 +07:00
|
|
|
/* if we clean as many as budgeted, we must not be done */
|
|
|
|
if (cleaned >= budget_per_ring)
|
|
|
|
clean_complete = false;
|
2015-01-24 16:58:35 +07:00
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* If work not completed, return budget and polling will return */
|
2015-01-10 08:07:19 +07:00
|
|
|
if (!clean_complete) {
|
i40e/i40evf: fix interrupt affinity bug
There exists a bug in which a 'perfect storm' can occur and cause
interrupts to fail to be correctly affinitized. This causes unexpected
behavior and has a substantial impact on performance when it happens.
The bug occurs if there is heavy traffic, any number of CPUs that have
an i40e interrupt are pegged at 100%, and the interrupt afffinity for
those CPUs is changed. Instead of moving to the new CPU, the interrupt
continues to be polled while there is heavy traffic.
The bug is most readily realized as the driver is first brought up and
all interrupts start on CPU0. If there is heavy traffic and the
interrupt starts polling before the interrupt is affinitized, the
interrupt will be stuck on CPU0 until traffic stops. The bug, however,
can also be wrought out more simply by affinitizing all the interrupts
to a single CPU and then attempting to move any of those interrupts off
while there is heavy traffic.
This patch fixes the bug by registering for update notifications from
the kernel when the interrupt affinity changes. When that fires, we
cache the intended affinity mask. Then, while polling, if the cpu is
pegged at 100% and we failed to clean the rings, we check to make sure
we have the correct affinity and stop polling if we're firing on the
wrong CPU. When the kernel successfully moves the interrupt, it will
start polling on the correct CPU. The performance impact is minimal
since the only time this section gets executed is when performance is
already compromised by the CPU.
Change-ID: I4410a880159b9dba1f8297aa72bef36dca34e830
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2016-09-15 06:24:38 +07:00
|
|
|
const cpumask_t *aff_mask = &q_vector->affinity_mask;
|
|
|
|
int cpu_id = smp_processor_id();
|
|
|
|
|
|
|
|
/* It is possible that the interrupt affinity has changed but,
|
|
|
|
* if the cpu is pegged at 100%, polling will never exit while
|
|
|
|
* traffic continues and the interrupt will be stuck on this
|
|
|
|
* cpu. We check to make sure affinity is correct before we
|
|
|
|
* continue to poll, otherwise we must stop polling so the
|
|
|
|
* interrupt can move to the correct cpu.
|
|
|
|
*/
|
|
|
|
if (likely(cpumask_test_cpu(cpu_id, aff_mask))) {
|
2015-09-24 23:04:26 +07:00
|
|
|
tx_only:
|
i40e/i40evf: fix interrupt affinity bug
There exists a bug in which a 'perfect storm' can occur and cause
interrupts to fail to be correctly affinitized. This causes unexpected
behavior and has a substantial impact on performance when it happens.
The bug occurs if there is heavy traffic, any number of CPUs that have
an i40e interrupt are pegged at 100%, and the interrupt afffinity for
those CPUs is changed. Instead of moving to the new CPU, the interrupt
continues to be polled while there is heavy traffic.
The bug is most readily realized as the driver is first brought up and
all interrupts start on CPU0. If there is heavy traffic and the
interrupt starts polling before the interrupt is affinitized, the
interrupt will be stuck on CPU0 until traffic stops. The bug, however,
can also be wrought out more simply by affinitizing all the interrupts
to a single CPU and then attempting to move any of those interrupts off
while there is heavy traffic.
This patch fixes the bug by registering for update notifications from
the kernel when the interrupt affinity changes. When that fires, we
cache the intended affinity mask. Then, while polling, if the cpu is
pegged at 100% and we failed to clean the rings, we check to make sure
we have the correct affinity and stop polling if we're firing on the
wrong CPU. When the kernel successfully moves the interrupt, it will
start polling on the correct CPU. The performance impact is minimal
since the only time this section gets executed is when performance is
already compromised by the CPU.
Change-ID: I4410a880159b9dba1f8297aa72bef36dca34e830
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2016-09-15 06:24:38 +07:00
|
|
|
if (arm_wb) {
|
|
|
|
q_vector->tx.ring[0].tx_stats.tx_force_wb++;
|
|
|
|
i40e_enable_wb_on_itr(vsi, q_vector);
|
|
|
|
}
|
|
|
|
return budget;
|
2015-10-22 06:47:08 +07:00
|
|
|
}
|
2015-01-10 08:07:19 +07:00
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2015-06-05 23:20:30 +07:00
|
|
|
if (vsi->back->flags & I40E_TXR_FLAGS_WB_ON_ITR)
|
|
|
|
q_vector->arm_wb_state = false;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* Work is done so exit the polling mode and re-enable the interrupt */
|
2015-09-25 06:35:47 +07:00
|
|
|
napi_complete_done(napi, work_done);
|
i40e/i40evf: fix interrupt affinity bug
There exists a bug in which a 'perfect storm' can occur and cause
interrupts to fail to be correctly affinitized. This causes unexpected
behavior and has a substantial impact on performance when it happens.
The bug occurs if there is heavy traffic, any number of CPUs that have
an i40e interrupt are pegged at 100%, and the interrupt afffinity for
those CPUs is changed. Instead of moving to the new CPU, the interrupt
continues to be polled while there is heavy traffic.
The bug is most readily realized as the driver is first brought up and
all interrupts start on CPU0. If there is heavy traffic and the
interrupt starts polling before the interrupt is affinitized, the
interrupt will be stuck on CPU0 until traffic stops. The bug, however,
can also be wrought out more simply by affinitizing all the interrupts
to a single CPU and then attempting to move any of those interrupts off
while there is heavy traffic.
This patch fixes the bug by registering for update notifications from
the kernel when the interrupt affinity changes. When that fires, we
cache the intended affinity mask. Then, while polling, if the cpu is
pegged at 100% and we failed to clean the rings, we check to make sure
we have the correct affinity and stop polling if we're firing on the
wrong CPU. When the kernel successfully moves the interrupt, it will
start polling on the correct CPU. The performance impact is minimal
since the only time this section gets executed is when performance is
already compromised by the CPU.
Change-ID: I4410a880159b9dba1f8297aa72bef36dca34e830
Signed-off-by: Alan Brady <alan.brady@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
2016-09-15 06:24:38 +07:00
|
|
|
|
|
|
|
/* If we're prematurely stopping polling to fix the interrupt
|
|
|
|
* affinity we want to make sure polling starts back up so we
|
|
|
|
* issue a call to i40evf_force_wb which triggers a SW interrupt.
|
|
|
|
*/
|
|
|
|
if (!clean_complete)
|
|
|
|
i40evf_force_wb(vsi, q_vector);
|
|
|
|
else
|
|
|
|
i40e_update_enable_itr(vsi, q_vector);
|
|
|
|
|
2016-11-09 04:05:16 +07:00
|
|
|
return min(work_done, budget - 1);
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2015-04-17 07:06:10 +07:00
|
|
|
* i40evf_tx_prepare_vlan_flags - prepare generic TX VLAN tagging flags for HW
|
2013-12-21 13:12:51 +07:00
|
|
|
* @skb: send buffer
|
|
|
|
* @tx_ring: ring to send buffer on
|
|
|
|
* @flags: the tx flags to be set
|
|
|
|
*
|
|
|
|
* Checks the skb and set up correspondingly several generic transmit flags
|
|
|
|
* related to VLAN tagging for the HW, such as VLAN, DCB, etc.
|
|
|
|
*
|
|
|
|
* Returns error code indicate the frame should be dropped upon error and the
|
|
|
|
* otherwise returns 0 to indicate the flags has been set properly.
|
|
|
|
**/
|
2015-04-17 07:06:10 +07:00
|
|
|
static inline int i40evf_tx_prepare_vlan_flags(struct sk_buff *skb,
|
|
|
|
struct i40e_ring *tx_ring,
|
|
|
|
u32 *flags)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
|
|
|
__be16 protocol = skb->protocol;
|
|
|
|
u32 tx_flags = 0;
|
|
|
|
|
2015-03-31 14:45:03 +07:00
|
|
|
if (protocol == htons(ETH_P_8021Q) &&
|
|
|
|
!(tx_ring->netdev->features & NETIF_F_HW_VLAN_CTAG_TX)) {
|
|
|
|
/* When HW VLAN acceleration is turned off by the user the
|
|
|
|
* stack sets the protocol to 8021q so that the driver
|
|
|
|
* can take any steps required to support the SW only
|
|
|
|
* VLAN handling. In our case the driver doesn't need
|
|
|
|
* to take any further steps so just set the protocol
|
|
|
|
* to the encapsulated ethertype.
|
|
|
|
*/
|
|
|
|
skb->protocol = vlan_get_protocol(skb);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* if we have a HW VLAN tag being added, default to the HW one */
|
2015-01-13 23:13:44 +07:00
|
|
|
if (skb_vlan_tag_present(skb)) {
|
|
|
|
tx_flags |= skb_vlan_tag_get(skb) << I40E_TX_FLAGS_VLAN_SHIFT;
|
2013-12-21 13:12:51 +07:00
|
|
|
tx_flags |= I40E_TX_FLAGS_HW_VLAN;
|
|
|
|
/* else if it is a SW VLAN, check the next protocol and store the tag */
|
|
|
|
} else if (protocol == htons(ETH_P_8021Q)) {
|
|
|
|
struct vlan_hdr *vhdr, _vhdr;
|
2015-08-29 04:55:54 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
vhdr = skb_header_pointer(skb, ETH_HLEN, sizeof(_vhdr), &_vhdr);
|
|
|
|
if (!vhdr)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
protocol = vhdr->h_vlan_encapsulated_proto;
|
|
|
|
tx_flags |= ntohs(vhdr->h_vlan_TCI) << I40E_TX_FLAGS_VLAN_SHIFT;
|
|
|
|
tx_flags |= I40E_TX_FLAGS_SW_VLAN;
|
|
|
|
}
|
|
|
|
|
2015-03-31 14:45:03 +07:00
|
|
|
out:
|
2013-12-21 13:12:51 +07:00
|
|
|
*flags = tx_flags;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_tso - set up the tso context descriptor
|
2016-11-29 07:05:59 +07:00
|
|
|
* @first: pointer to first Tx buffer for xmit
|
2013-12-21 13:12:51 +07:00
|
|
|
* @hdr_len: ptr to the size of the packet header
|
2015-10-22 06:47:02 +07:00
|
|
|
* @cd_type_cmd_tso_mss: Quad Word 1
|
2013-12-21 13:12:51 +07:00
|
|
|
*
|
|
|
|
* Returns 0 if no TSO can happen, 1 if tso is going, or error
|
|
|
|
**/
|
2016-11-29 07:05:59 +07:00
|
|
|
static int i40e_tso(struct i40e_tx_buffer *first, u8 *hdr_len,
|
|
|
|
u64 *cd_type_cmd_tso_mss)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
2016-11-29 07:05:59 +07:00
|
|
|
struct sk_buff *skb = first->skb;
|
2016-01-25 12:16:20 +07:00
|
|
|
u64 cd_cmd, cd_tso_len, cd_mss;
|
2016-01-25 12:16:35 +07:00
|
|
|
union {
|
|
|
|
struct iphdr *v4;
|
|
|
|
struct ipv6hdr *v6;
|
|
|
|
unsigned char *hdr;
|
|
|
|
} ip;
|
2016-01-25 12:16:28 +07:00
|
|
|
union {
|
|
|
|
struct tcphdr *tcp;
|
2016-01-25 12:17:29 +07:00
|
|
|
struct udphdr *udp;
|
2016-01-25 12:16:28 +07:00
|
|
|
unsigned char *hdr;
|
|
|
|
} l4;
|
|
|
|
u32 paylen, l4_offset;
|
2016-11-29 07:05:59 +07:00
|
|
|
u16 gso_segs, gso_size;
|
2013-12-21 13:12:51 +07:00
|
|
|
int err;
|
|
|
|
|
2016-01-05 01:33:04 +07:00
|
|
|
if (skb->ip_summed != CHECKSUM_PARTIAL)
|
|
|
|
return 0;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
if (!skb_is_gso(skb))
|
|
|
|
return 0;
|
|
|
|
|
2014-03-30 10:14:53 +07:00
|
|
|
err = skb_cow_head(skb, 0);
|
|
|
|
if (err < 0)
|
|
|
|
return err;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-01-25 12:16:35 +07:00
|
|
|
ip.hdr = skb_network_header(skb);
|
|
|
|
l4.hdr = skb_transport_header(skb);
|
2015-02-21 13:44:16 +07:00
|
|
|
|
2016-01-25 12:16:35 +07:00
|
|
|
/* initialize outer IP header fields */
|
|
|
|
if (ip.v4->version == 4) {
|
|
|
|
ip.v4->tot_len = 0;
|
|
|
|
ip.v4->check = 0;
|
2016-01-25 12:16:28 +07:00
|
|
|
} else {
|
2016-01-25 12:16:35 +07:00
|
|
|
ip.v6->payload_len = 0;
|
|
|
|
}
|
|
|
|
|
2016-04-02 14:06:56 +07:00
|
|
|
if (skb_shinfo(skb)->gso_type & (SKB_GSO_GRE |
|
2016-04-15 04:19:25 +07:00
|
|
|
SKB_GSO_GRE_CSUM |
|
2016-05-18 23:06:10 +07:00
|
|
|
SKB_GSO_IPXIP4 |
|
2016-05-19 00:44:53 +07:00
|
|
|
SKB_GSO_IPXIP6 |
|
2016-04-02 14:06:56 +07:00
|
|
|
SKB_GSO_UDP_TUNNEL |
|
2016-01-25 12:17:29 +07:00
|
|
|
SKB_GSO_UDP_TUNNEL_CSUM)) {
|
2016-04-15 04:19:25 +07:00
|
|
|
if (!(skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL) &&
|
|
|
|
(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_TUNNEL_CSUM)) {
|
|
|
|
l4.udp->len = 0;
|
|
|
|
|
2016-01-25 12:17:29 +07:00
|
|
|
/* determine offset of outer transport header */
|
|
|
|
l4_offset = l4.hdr - skb->data;
|
|
|
|
|
|
|
|
/* remove payload length from outer checksum */
|
2016-03-19 06:06:47 +07:00
|
|
|
paylen = skb->len - l4_offset;
|
2016-12-13 06:44:17 +07:00
|
|
|
csum_replace_by_diff(&l4.udp->check,
|
|
|
|
(__force __wsum)htonl(paylen));
|
2016-01-25 12:17:29 +07:00
|
|
|
}
|
|
|
|
|
2016-01-25 12:16:35 +07:00
|
|
|
/* reset pointers to inner headers */
|
|
|
|
ip.hdr = skb_inner_network_header(skb);
|
|
|
|
l4.hdr = skb_inner_transport_header(skb);
|
|
|
|
|
|
|
|
/* initialize inner IP header fields */
|
|
|
|
if (ip.v4->version == 4) {
|
|
|
|
ip.v4->tot_len = 0;
|
|
|
|
ip.v4->check = 0;
|
|
|
|
} else {
|
|
|
|
ip.v6->payload_len = 0;
|
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
2016-01-25 12:16:28 +07:00
|
|
|
/* determine offset of inner transport header */
|
|
|
|
l4_offset = l4.hdr - skb->data;
|
|
|
|
|
|
|
|
/* remove payload length from inner checksum */
|
2016-03-19 06:06:47 +07:00
|
|
|
paylen = skb->len - l4_offset;
|
2016-12-13 06:44:17 +07:00
|
|
|
csum_replace_by_diff(&l4.tcp->check, (__force __wsum)htonl(paylen));
|
2016-01-25 12:16:28 +07:00
|
|
|
|
|
|
|
/* compute length of segmentation header */
|
|
|
|
*hdr_len = (l4.tcp->doff * 4) + l4_offset;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-11-29 07:05:59 +07:00
|
|
|
/* pull values out of skb_shinfo */
|
|
|
|
gso_size = skb_shinfo(skb)->gso_size;
|
|
|
|
gso_segs = skb_shinfo(skb)->gso_segs;
|
|
|
|
|
|
|
|
/* update GSO size and bytecount with header size */
|
|
|
|
first->gso_segs = gso_segs;
|
|
|
|
first->bytecount += (first->gso_segs - 1) * *hdr_len;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* find the field values */
|
|
|
|
cd_cmd = I40E_TX_CTX_DESC_TSO;
|
|
|
|
cd_tso_len = skb->len - *hdr_len;
|
2016-11-29 07:05:59 +07:00
|
|
|
cd_mss = gso_size;
|
2016-01-25 12:16:20 +07:00
|
|
|
*cd_type_cmd_tso_mss |= (cd_cmd << I40E_TXD_CTX_QW1_CMD_SHIFT) |
|
|
|
|
(cd_tso_len << I40E_TXD_CTX_QW1_TSO_LEN_SHIFT) |
|
|
|
|
(cd_mss << I40E_TXD_CTX_QW1_MSS_SHIFT);
|
2013-12-21 13:12:51 +07:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_tx_enable_csum - Enable Tx checksum offloads
|
|
|
|
* @skb: send buffer
|
2015-04-17 07:06:00 +07:00
|
|
|
* @tx_flags: pointer to Tx flags currently set
|
2013-12-21 13:12:51 +07:00
|
|
|
* @td_cmd: Tx descriptor command bits to set
|
|
|
|
* @td_offset: Tx descriptor header offsets to set
|
2016-01-25 12:17:10 +07:00
|
|
|
* @tx_ring: Tx descriptor ring
|
2013-12-21 13:12:51 +07:00
|
|
|
* @cd_tunneling: ptr to context desc bits
|
|
|
|
**/
|
2016-01-25 12:17:10 +07:00
|
|
|
static int i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
|
|
|
|
u32 *td_cmd, u32 *td_offset,
|
|
|
|
struct i40e_ring *tx_ring,
|
|
|
|
u32 *cd_tunneling)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
2016-01-25 12:16:42 +07:00
|
|
|
union {
|
|
|
|
struct iphdr *v4;
|
|
|
|
struct ipv6hdr *v6;
|
|
|
|
unsigned char *hdr;
|
|
|
|
} ip;
|
|
|
|
union {
|
|
|
|
struct tcphdr *tcp;
|
|
|
|
struct udphdr *udp;
|
|
|
|
unsigned char *hdr;
|
|
|
|
} l4;
|
2016-01-25 12:16:54 +07:00
|
|
|
unsigned char *exthdr;
|
2016-04-01 17:56:04 +07:00
|
|
|
u32 offset, cmd = 0;
|
2016-01-25 12:16:54 +07:00
|
|
|
__be16 frag_off;
|
2016-01-25 12:16:42 +07:00
|
|
|
u8 l4_proto = 0;
|
|
|
|
|
2016-01-25 12:17:10 +07:00
|
|
|
if (skb->ip_summed != CHECKSUM_PARTIAL)
|
|
|
|
return 0;
|
|
|
|
|
2016-01-25 12:16:42 +07:00
|
|
|
ip.hdr = skb_network_header(skb);
|
|
|
|
l4.hdr = skb_transport_header(skb);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-01-25 12:17:01 +07:00
|
|
|
/* compute outer L2 header size */
|
|
|
|
offset = ((ip.hdr - skb->data) / 2) << I40E_TX_DESC_LENGTH_MACLEN_SHIFT;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
if (skb->encapsulation) {
|
2016-04-01 17:56:04 +07:00
|
|
|
u32 tunnel = 0;
|
2016-01-25 12:16:48 +07:00
|
|
|
/* define outer network header type */
|
|
|
|
if (*tx_flags & I40E_TX_FLAGS_IPV4) {
|
2016-01-25 12:17:01 +07:00
|
|
|
tunnel |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
|
|
|
|
I40E_TX_CTX_EXT_IP_IPV4 :
|
|
|
|
I40E_TX_CTX_EXT_IP_IPV4_NO_CSUM;
|
|
|
|
|
2016-01-25 12:16:48 +07:00
|
|
|
l4_proto = ip.v4->protocol;
|
|
|
|
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
|
2016-01-25 12:17:01 +07:00
|
|
|
tunnel |= I40E_TX_CTX_EXT_IP_IPV6;
|
2016-01-25 12:16:54 +07:00
|
|
|
|
|
|
|
exthdr = ip.hdr + sizeof(*ip.v6);
|
2016-01-25 12:16:48 +07:00
|
|
|
l4_proto = ip.v6->nexthdr;
|
2016-01-25 12:16:54 +07:00
|
|
|
if (l4.hdr != exthdr)
|
|
|
|
ipv6_skip_exthdr(skb, exthdr - skb->data,
|
|
|
|
&l4_proto, &frag_off);
|
2016-01-25 12:16:48 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* define outer transport */
|
|
|
|
switch (l4_proto) {
|
2015-02-27 16:15:29 +07:00
|
|
|
case IPPROTO_UDP:
|
2016-01-25 12:17:01 +07:00
|
|
|
tunnel |= I40E_TXD_CTX_UDP_TUNNELING;
|
2015-04-17 07:06:00 +07:00
|
|
|
*tx_flags |= I40E_TX_FLAGS_VXLAN_TUNNEL;
|
2015-02-27 16:15:29 +07:00
|
|
|
break;
|
2016-01-25 12:16:48 +07:00
|
|
|
case IPPROTO_GRE:
|
2016-01-25 12:17:01 +07:00
|
|
|
tunnel |= I40E_TXD_CTX_GRE_TUNNELING;
|
2016-01-25 12:16:48 +07:00
|
|
|
*tx_flags |= I40E_TX_FLAGS_VXLAN_TUNNEL;
|
|
|
|
break;
|
2016-04-02 14:06:56 +07:00
|
|
|
case IPPROTO_IPIP:
|
|
|
|
case IPPROTO_IPV6:
|
|
|
|
*tx_flags |= I40E_TX_FLAGS_VXLAN_TUNNEL;
|
|
|
|
l4.hdr = skb_inner_network_header(skb);
|
|
|
|
break;
|
2015-02-27 16:15:29 +07:00
|
|
|
default:
|
2016-01-25 12:17:10 +07:00
|
|
|
if (*tx_flags & I40E_TX_FLAGS_TSO)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
skb_checksum_help(skb);
|
|
|
|
return 0;
|
2015-02-27 16:15:29 +07:00
|
|
|
}
|
2016-01-25 12:16:42 +07:00
|
|
|
|
2016-04-02 14:06:56 +07:00
|
|
|
/* compute outer L3 header size */
|
|
|
|
tunnel |= ((l4.hdr - ip.hdr) / 4) <<
|
|
|
|
I40E_TXD_CTX_QW0_EXT_IPLEN_SHIFT;
|
|
|
|
|
|
|
|
/* switch IP header pointer from outer to inner header */
|
|
|
|
ip.hdr = skb_inner_network_header(skb);
|
|
|
|
|
2016-01-25 12:17:01 +07:00
|
|
|
/* compute tunnel header size */
|
|
|
|
tunnel |= ((ip.hdr - l4.hdr) / 2) <<
|
|
|
|
I40E_TXD_CTX_QW0_NATLEN_SHIFT;
|
|
|
|
|
2016-01-25 12:17:29 +07:00
|
|
|
/* indicate if we need to offload outer UDP header */
|
|
|
|
if ((*tx_flags & I40E_TX_FLAGS_TSO) &&
|
2016-04-15 04:19:25 +07:00
|
|
|
!(skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL) &&
|
2016-01-25 12:17:29 +07:00
|
|
|
(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_TUNNEL_CSUM))
|
|
|
|
tunnel |= I40E_TXD_CTX_QW0_L4T_CS_MASK;
|
|
|
|
|
2016-01-25 12:17:01 +07:00
|
|
|
/* record tunnel offload values */
|
|
|
|
*cd_tunneling |= tunnel;
|
|
|
|
|
2016-01-25 12:16:42 +07:00
|
|
|
/* switch L4 header pointer from outer to inner */
|
|
|
|
l4.hdr = skb_inner_transport_header(skb);
|
2016-01-25 12:16:48 +07:00
|
|
|
l4_proto = 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-01-25 12:16:48 +07:00
|
|
|
/* reset type as we transition from outer to inner headers */
|
|
|
|
*tx_flags &= ~(I40E_TX_FLAGS_IPV4 | I40E_TX_FLAGS_IPV6);
|
|
|
|
if (ip.v4->version == 4)
|
|
|
|
*tx_flags |= I40E_TX_FLAGS_IPV4;
|
|
|
|
if (ip.v6->version == 6)
|
2015-04-17 07:06:00 +07:00
|
|
|
*tx_flags |= I40E_TX_FLAGS_IPV6;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Enable IP checksum offloads */
|
2015-04-17 07:06:00 +07:00
|
|
|
if (*tx_flags & I40E_TX_FLAGS_IPV4) {
|
2016-01-25 12:16:42 +07:00
|
|
|
l4_proto = ip.v4->protocol;
|
2013-12-21 13:12:51 +07:00
|
|
|
/* the stack computes the IP header already, the only time we
|
|
|
|
* need the hardware to recompute it is in the case of TSO.
|
|
|
|
*/
|
2016-01-25 12:17:01 +07:00
|
|
|
cmd |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
|
|
|
|
I40E_TX_DESC_CMD_IIPT_IPV4_CSUM :
|
|
|
|
I40E_TX_DESC_CMD_IIPT_IPV4;
|
2015-04-17 07:06:00 +07:00
|
|
|
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
|
2016-01-25 12:17:01 +07:00
|
|
|
cmd |= I40E_TX_DESC_CMD_IIPT_IPV6;
|
2016-01-25 12:16:54 +07:00
|
|
|
|
|
|
|
exthdr = ip.hdr + sizeof(*ip.v6);
|
|
|
|
l4_proto = ip.v6->nexthdr;
|
|
|
|
if (l4.hdr != exthdr)
|
|
|
|
ipv6_skip_exthdr(skb, exthdr - skb->data,
|
|
|
|
&l4_proto, &frag_off);
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
2016-01-25 12:16:42 +07:00
|
|
|
|
2016-01-25 12:17:01 +07:00
|
|
|
/* compute inner L3 header size */
|
|
|
|
offset |= ((l4.hdr - ip.hdr) / 4) << I40E_TX_DESC_LENGTH_IPLEN_SHIFT;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* Enable L4 checksum offloads */
|
2016-01-25 12:16:42 +07:00
|
|
|
switch (l4_proto) {
|
2013-12-21 13:12:51 +07:00
|
|
|
case IPPROTO_TCP:
|
|
|
|
/* enable checksum offloads */
|
2016-01-25 12:17:01 +07:00
|
|
|
cmd |= I40E_TX_DESC_CMD_L4T_EOFT_TCP;
|
|
|
|
offset |= l4.tcp->doff << I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
case IPPROTO_SCTP:
|
|
|
|
/* enable SCTP checksum offload */
|
2016-01-25 12:17:01 +07:00
|
|
|
cmd |= I40E_TX_DESC_CMD_L4T_EOFT_SCTP;
|
|
|
|
offset |= (sizeof(struct sctphdr) >> 2) <<
|
|
|
|
I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
case IPPROTO_UDP:
|
|
|
|
/* enable UDP checksum offload */
|
2016-01-25 12:17:01 +07:00
|
|
|
cmd |= I40E_TX_DESC_CMD_L4T_EOFT_UDP;
|
|
|
|
offset |= (sizeof(struct udphdr) >> 2) <<
|
|
|
|
I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
|
2013-12-21 13:12:51 +07:00
|
|
|
break;
|
|
|
|
default:
|
2016-01-25 12:17:10 +07:00
|
|
|
if (*tx_flags & I40E_TX_FLAGS_TSO)
|
|
|
|
return -1;
|
|
|
|
skb_checksum_help(skb);
|
|
|
|
return 0;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
2016-01-25 12:17:01 +07:00
|
|
|
|
|
|
|
*td_cmd |= cmd;
|
|
|
|
*td_offset |= offset;
|
2016-01-25 12:17:10 +07:00
|
|
|
|
|
|
|
return 1;
|
2013-12-21 13:12:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_create_tx_ctx Build the Tx context descriptor
|
|
|
|
* @tx_ring: ring to create the descriptor on
|
|
|
|
* @cd_type_cmd_tso_mss: Quad Word 1
|
|
|
|
* @cd_tunneling: Quad Word 0 - bits 0-31
|
|
|
|
* @cd_l2tag2: Quad Word 0 - bits 32-63
|
|
|
|
**/
|
|
|
|
static void i40e_create_tx_ctx(struct i40e_ring *tx_ring,
|
|
|
|
const u64 cd_type_cmd_tso_mss,
|
|
|
|
const u32 cd_tunneling, const u32 cd_l2tag2)
|
|
|
|
{
|
|
|
|
struct i40e_tx_context_desc *context_desc;
|
|
|
|
int i = tx_ring->next_to_use;
|
|
|
|
|
2014-02-14 09:14:41 +07:00
|
|
|
if ((cd_type_cmd_tso_mss == I40E_TX_DESC_DTYPE_CONTEXT) &&
|
|
|
|
!cd_tunneling && !cd_l2tag2)
|
2013-12-21 13:12:51 +07:00
|
|
|
return;
|
|
|
|
|
|
|
|
/* grab the next descriptor */
|
|
|
|
context_desc = I40E_TX_CTXTDESC(tx_ring, i);
|
|
|
|
|
|
|
|
i++;
|
|
|
|
tx_ring->next_to_use = (i < tx_ring->count) ? i : 0;
|
|
|
|
|
|
|
|
/* cpu_to_le32 and assign to struct fields */
|
|
|
|
context_desc->tunneling_params = cpu_to_le32(cd_tunneling);
|
|
|
|
context_desc->l2tag2 = cpu_to_le16(cd_l2tag2);
|
2014-06-05 03:41:54 +07:00
|
|
|
context_desc->rsvd = cpu_to_le16(0);
|
2013-12-21 13:12:51 +07:00
|
|
|
context_desc->type_cmd_tso_mss = cpu_to_le64(cd_type_cmd_tso_mss);
|
|
|
|
}
|
|
|
|
|
2015-11-19 08:35:42 +07:00
|
|
|
/**
|
2016-03-31 06:15:37 +07:00
|
|
|
* __i40evf_chk_linearize - Check if there are more than 8 buffers per packet
|
2015-02-21 13:42:35 +07:00
|
|
|
* @skb: send buffer
|
|
|
|
*
|
2016-03-31 06:15:37 +07:00
|
|
|
* Note: Our HW can't DMA more than 8 buffers to build a packet on the wire
|
|
|
|
* and so we need to figure out the cases where we need to linearize the skb.
|
|
|
|
*
|
|
|
|
* For TSO we need to count the TSO header and segment payload separately.
|
|
|
|
* As such we need to check cases where we have 7 fragments or more as we
|
|
|
|
* can potentially require 9 DMA transactions, 1 for the TSO header, 1 for
|
|
|
|
* the segment payload in the first descriptor, and another 7 for the
|
|
|
|
* fragments.
|
2015-02-21 13:42:35 +07:00
|
|
|
**/
|
2016-02-18 02:02:50 +07:00
|
|
|
bool __i40evf_chk_linearize(struct sk_buff *skb)
|
2015-02-21 13:42:35 +07:00
|
|
|
{
|
2016-02-18 02:02:50 +07:00
|
|
|
const struct skb_frag_struct *frag, *stale;
|
2016-03-31 06:15:37 +07:00
|
|
|
int nr_frags, sum;
|
2015-02-21 13:42:35 +07:00
|
|
|
|
2016-03-31 06:15:37 +07:00
|
|
|
/* no need to check if number of frags is less than 7 */
|
2016-02-18 02:02:50 +07:00
|
|
|
nr_frags = skb_shinfo(skb)->nr_frags;
|
2016-03-31 06:15:37 +07:00
|
|
|
if (nr_frags < (I40E_MAX_BUFFER_TXD - 1))
|
2016-02-18 02:02:50 +07:00
|
|
|
return false;
|
2015-02-21 13:42:35 +07:00
|
|
|
|
2016-02-18 02:02:50 +07:00
|
|
|
/* We need to walk through the list and validate that each group
|
2016-09-07 08:05:04 +07:00
|
|
|
* of 6 fragments totals at least gso_size.
|
2016-02-18 02:02:50 +07:00
|
|
|
*/
|
2016-03-31 06:15:37 +07:00
|
|
|
nr_frags -= I40E_MAX_BUFFER_TXD - 2;
|
2016-02-18 02:02:50 +07:00
|
|
|
frag = &skb_shinfo(skb)->frags[0];
|
|
|
|
|
|
|
|
/* Initialize size to the negative value of gso_size minus 1. We
|
|
|
|
* use this as the worst case scenerio in which the frag ahead
|
|
|
|
* of us only provides one byte which is why we are limited to 6
|
|
|
|
* descriptors for a single transmit as the header and previous
|
|
|
|
* fragment are already consuming 2 descriptors.
|
|
|
|
*/
|
2016-03-31 06:15:37 +07:00
|
|
|
sum = 1 - skb_shinfo(skb)->gso_size;
|
2016-02-18 02:02:50 +07:00
|
|
|
|
2016-03-31 06:15:37 +07:00
|
|
|
/* Add size of frags 0 through 4 to create our initial sum */
|
|
|
|
sum += skb_frag_size(frag++);
|
|
|
|
sum += skb_frag_size(frag++);
|
|
|
|
sum += skb_frag_size(frag++);
|
|
|
|
sum += skb_frag_size(frag++);
|
|
|
|
sum += skb_frag_size(frag++);
|
2016-02-18 02:02:50 +07:00
|
|
|
|
|
|
|
/* Walk through fragments adding latest fragment, testing it, and
|
|
|
|
* then removing stale fragments from the sum.
|
|
|
|
*/
|
|
|
|
stale = &skb_shinfo(skb)->frags[0];
|
|
|
|
for (;;) {
|
2016-03-31 06:15:37 +07:00
|
|
|
sum += skb_frag_size(frag++);
|
2016-02-18 02:02:50 +07:00
|
|
|
|
|
|
|
/* if sum is negative we failed to make sufficient progress */
|
|
|
|
if (sum < 0)
|
|
|
|
return true;
|
|
|
|
|
2016-09-07 08:05:04 +07:00
|
|
|
if (!nr_frags--)
|
2016-02-18 02:02:50 +07:00
|
|
|
break;
|
|
|
|
|
2016-03-31 06:15:37 +07:00
|
|
|
sum -= skb_frag_size(stale++);
|
2015-02-21 13:42:35 +07:00
|
|
|
}
|
|
|
|
|
2016-02-18 02:02:50 +07:00
|
|
|
return false;
|
2015-02-21 13:42:35 +07:00
|
|
|
}
|
|
|
|
|
2015-04-17 07:06:09 +07:00
|
|
|
/**
|
|
|
|
* __i40evf_maybe_stop_tx - 2nd level check for tx stop conditions
|
|
|
|
* @tx_ring: the ring to be checked
|
|
|
|
* @size: the size buffer we want to assure is available
|
|
|
|
*
|
|
|
|
* Returns -EBUSY if a stop is needed, else 0
|
|
|
|
**/
|
2016-02-18 02:02:43 +07:00
|
|
|
int __i40evf_maybe_stop_tx(struct i40e_ring *tx_ring, int size)
|
2015-04-17 07:06:09 +07:00
|
|
|
{
|
|
|
|
netif_stop_subqueue(tx_ring->netdev, tx_ring->queue_index);
|
|
|
|
/* Memory barrier before checking head and tail */
|
|
|
|
smp_mb();
|
|
|
|
|
|
|
|
/* Check again in a case another CPU has just made room available. */
|
|
|
|
if (likely(I40E_DESC_UNUSED(tx_ring) < size))
|
|
|
|
return -EBUSY;
|
|
|
|
|
|
|
|
/* A reprieve! - use start_queue because it doesn't call schedule */
|
|
|
|
netif_start_subqueue(tx_ring->netdev, tx_ring->queue_index);
|
|
|
|
++tx_ring->tx_stats.restart_queue;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/**
|
2015-04-17 07:06:10 +07:00
|
|
|
* i40evf_tx_map - Build the Tx descriptor
|
2013-12-21 13:12:51 +07:00
|
|
|
* @tx_ring: ring to send buffer on
|
|
|
|
* @skb: send buffer
|
|
|
|
* @first: first buffer info buffer to use
|
|
|
|
* @tx_flags: collected send information
|
|
|
|
* @hdr_len: size of the packet header
|
|
|
|
* @td_cmd: the command field in the descriptor
|
|
|
|
* @td_offset: offset for checksum or crc
|
|
|
|
**/
|
2015-04-17 07:06:10 +07:00
|
|
|
static inline void i40evf_tx_map(struct i40e_ring *tx_ring, struct sk_buff *skb,
|
|
|
|
struct i40e_tx_buffer *first, u32 tx_flags,
|
|
|
|
const u8 hdr_len, u32 td_cmd, u32 td_offset)
|
2013-12-21 13:12:51 +07:00
|
|
|
{
|
|
|
|
unsigned int data_len = skb->data_len;
|
|
|
|
unsigned int size = skb_headlen(skb);
|
|
|
|
struct skb_frag_struct *frag;
|
|
|
|
struct i40e_tx_buffer *tx_bi;
|
|
|
|
struct i40e_tx_desc *tx_desc;
|
|
|
|
u16 i = tx_ring->next_to_use;
|
|
|
|
u32 td_tag = 0;
|
|
|
|
dma_addr_t dma;
|
|
|
|
|
|
|
|
if (tx_flags & I40E_TX_FLAGS_HW_VLAN) {
|
|
|
|
td_cmd |= I40E_TX_DESC_CMD_IL2TAG1;
|
|
|
|
td_tag = (tx_flags & I40E_TX_FLAGS_VLAN_MASK) >>
|
|
|
|
I40E_TX_FLAGS_VLAN_SHIFT;
|
|
|
|
}
|
|
|
|
|
|
|
|
first->tx_flags = tx_flags;
|
|
|
|
|
|
|
|
dma = dma_map_single(tx_ring->dev, skb->data, size, DMA_TO_DEVICE);
|
|
|
|
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, i);
|
|
|
|
tx_bi = first;
|
|
|
|
|
|
|
|
for (frag = &skb_shinfo(skb)->frags[0];; frag++) {
|
2016-02-20 03:17:08 +07:00
|
|
|
unsigned int max_data = I40E_MAX_DATA_PER_TXD_ALIGNED;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
if (dma_mapping_error(tx_ring->dev, dma))
|
|
|
|
goto dma_error;
|
|
|
|
|
|
|
|
/* record length, and DMA address */
|
|
|
|
dma_unmap_len_set(tx_bi, len, size);
|
|
|
|
dma_unmap_addr_set(tx_bi, dma, dma);
|
|
|
|
|
2016-02-20 03:17:08 +07:00
|
|
|
/* align size to end of page */
|
|
|
|
max_data += -dma & (I40E_MAX_READ_REQ_SIZE - 1);
|
2013-12-21 13:12:51 +07:00
|
|
|
tx_desc->buffer_addr = cpu_to_le64(dma);
|
|
|
|
|
|
|
|
while (unlikely(size > I40E_MAX_DATA_PER_TXD)) {
|
|
|
|
tx_desc->cmd_type_offset_bsz =
|
|
|
|
build_ctob(td_cmd, td_offset,
|
2016-02-20 03:17:08 +07:00
|
|
|
max_data, td_tag);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
tx_desc++;
|
|
|
|
i++;
|
2015-10-27 06:44:29 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
if (i == tx_ring->count) {
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, 0);
|
|
|
|
i = 0;
|
|
|
|
}
|
|
|
|
|
2016-02-20 03:17:08 +07:00
|
|
|
dma += max_data;
|
|
|
|
size -= max_data;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-02-20 03:17:08 +07:00
|
|
|
max_data = I40E_MAX_DATA_PER_TXD_ALIGNED;
|
2013-12-21 13:12:51 +07:00
|
|
|
tx_desc->buffer_addr = cpu_to_le64(dma);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (likely(!data_len))
|
|
|
|
break;
|
|
|
|
|
|
|
|
tx_desc->cmd_type_offset_bsz = build_ctob(td_cmd, td_offset,
|
|
|
|
size, td_tag);
|
|
|
|
|
|
|
|
tx_desc++;
|
|
|
|
i++;
|
2015-10-27 06:44:29 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
if (i == tx_ring->count) {
|
|
|
|
tx_desc = I40E_TX_DESC(tx_ring, 0);
|
|
|
|
i = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
size = skb_frag_size(frag);
|
|
|
|
data_len -= size;
|
|
|
|
|
|
|
|
dma = skb_frag_dma_map(tx_ring->dev, frag, 0, size,
|
|
|
|
DMA_TO_DEVICE);
|
|
|
|
|
|
|
|
tx_bi = &tx_ring->tx_bi[i];
|
|
|
|
}
|
|
|
|
|
2016-10-12 05:26:54 +07:00
|
|
|
netdev_tx_sent_queue(txring_txq(tx_ring), first->bytecount);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
i++;
|
|
|
|
if (i == tx_ring->count)
|
|
|
|
i = 0;
|
|
|
|
|
|
|
|
tx_ring->next_to_use = i;
|
|
|
|
|
2016-02-18 02:02:43 +07:00
|
|
|
i40e_maybe_stop_tx(tx_ring, DESC_NEEDED);
|
2015-10-27 06:44:29 +07:00
|
|
|
|
2017-03-11 03:22:00 +07:00
|
|
|
/* write last descriptor with RS and EOP bits */
|
|
|
|
td_cmd |= I40E_TXD_CMD;
|
2015-10-27 06:44:29 +07:00
|
|
|
tx_desc->cmd_type_offset_bsz =
|
2016-10-12 05:26:54 +07:00
|
|
|
build_ctob(td_cmd, td_offset, size, td_tag);
|
|
|
|
|
|
|
|
/* Force memory writes to complete before letting h/w know there
|
|
|
|
* are new descriptors to fetch.
|
|
|
|
*
|
|
|
|
* We also use this memory barrier to make certain all of the
|
|
|
|
* status bits have been updated before next_to_watch is written.
|
|
|
|
*/
|
|
|
|
wmb();
|
|
|
|
|
|
|
|
/* set next_to_watch value indicating a packet is present */
|
|
|
|
first->next_to_watch = tx_desc;
|
2015-10-27 06:44:29 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* notify HW of packet */
|
2017-03-11 03:22:00 +07:00
|
|
|
if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
|
2015-10-27 06:44:29 +07:00
|
|
|
writel(i, tx_ring->tail);
|
2016-10-12 05:26:54 +07:00
|
|
|
|
|
|
|
/* we need this if more than one processor can write to our tail
|
|
|
|
* at a time, it synchronizes IO on IA64/Altix systems
|
|
|
|
*/
|
|
|
|
mmiowb();
|
2015-10-27 06:44:29 +07:00
|
|
|
}
|
2016-10-12 05:26:54 +07:00
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
return;
|
|
|
|
|
|
|
|
dma_error:
|
|
|
|
dev_info(tx_ring->dev, "TX DMA map failed\n");
|
|
|
|
|
|
|
|
/* clear dma mappings for failed tx_bi map */
|
|
|
|
for (;;) {
|
|
|
|
tx_bi = &tx_ring->tx_bi[i];
|
|
|
|
i40e_unmap_and_free_tx_resource(tx_ring, tx_bi);
|
|
|
|
if (tx_bi == first)
|
|
|
|
break;
|
|
|
|
if (i == 0)
|
|
|
|
i = tx_ring->count;
|
|
|
|
i--;
|
|
|
|
}
|
|
|
|
|
|
|
|
tx_ring->next_to_use = i;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40e_xmit_frame_ring - Sends buffer on Tx ring
|
|
|
|
* @skb: send buffer
|
|
|
|
* @tx_ring: ring to send buffer on
|
|
|
|
*
|
|
|
|
* Returns NETDEV_TX_OK if sent, else an error code
|
|
|
|
**/
|
|
|
|
static netdev_tx_t i40e_xmit_frame_ring(struct sk_buff *skb,
|
|
|
|
struct i40e_ring *tx_ring)
|
|
|
|
{
|
|
|
|
u64 cd_type_cmd_tso_mss = I40E_TX_DESC_DTYPE_CONTEXT;
|
|
|
|
u32 cd_tunneling = 0, cd_l2tag2 = 0;
|
|
|
|
struct i40e_tx_buffer *first;
|
|
|
|
u32 td_offset = 0;
|
|
|
|
u32 tx_flags = 0;
|
|
|
|
__be16 protocol;
|
|
|
|
u32 td_cmd = 0;
|
|
|
|
u8 hdr_len = 0;
|
2016-02-18 02:02:43 +07:00
|
|
|
int tso, count;
|
2015-08-29 04:55:54 +07:00
|
|
|
|
2015-10-27 06:44:30 +07:00
|
|
|
/* prefetch the data, we'll need it later */
|
|
|
|
prefetch(skb->data);
|
|
|
|
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(xmit_frame_ring, skb, tx_ring);
|
|
|
|
|
2016-02-18 02:02:43 +07:00
|
|
|
count = i40e_xmit_descriptor_count(skb);
|
2016-02-18 02:02:50 +07:00
|
|
|
if (i40e_chk_linearize(skb, count)) {
|
2016-11-29 07:05:59 +07:00
|
|
|
if (__skb_linearize(skb)) {
|
|
|
|
dev_kfree_skb_any(skb);
|
|
|
|
return NETDEV_TX_OK;
|
|
|
|
}
|
2016-02-20 03:17:08 +07:00
|
|
|
count = i40e_txd_use_count(skb->len);
|
2016-02-18 02:02:50 +07:00
|
|
|
tx_ring->tx_stats.tx_linearize++;
|
|
|
|
}
|
2016-02-18 02:02:43 +07:00
|
|
|
|
|
|
|
/* need: 1 descriptor per page * PAGE_SIZE/I40E_MAX_DATA_PER_TXD,
|
|
|
|
* + 1 desc for skb_head_len/I40E_MAX_DATA_PER_TXD,
|
|
|
|
* + 4 desc gap to avoid the cache line where head is,
|
|
|
|
* + 1 desc for context descriptor,
|
|
|
|
* otherwise try next time
|
|
|
|
*/
|
|
|
|
if (i40e_maybe_stop_tx(tx_ring, count + 4 + 1)) {
|
|
|
|
tx_ring->tx_stats.tx_busy++;
|
2013-12-21 13:12:51 +07:00
|
|
|
return NETDEV_TX_BUSY;
|
2016-02-18 02:02:43 +07:00
|
|
|
}
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-11-29 07:05:59 +07:00
|
|
|
/* record the location of the first descriptor for this packet */
|
|
|
|
first = &tx_ring->tx_bi[tx_ring->next_to_use];
|
|
|
|
first->skb = skb;
|
|
|
|
first->bytecount = skb->len;
|
|
|
|
first->gso_segs = 1;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
/* prepare the xmit flags */
|
2015-04-17 07:06:10 +07:00
|
|
|
if (i40evf_tx_prepare_vlan_flags(skb, tx_ring, &tx_flags))
|
2013-12-21 13:12:51 +07:00
|
|
|
goto out_drop;
|
|
|
|
|
|
|
|
/* obtain protocol of skb */
|
2014-08-25 21:34:53 +07:00
|
|
|
protocol = vlan_get_protocol(skb);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* setup IPv4/IPv6 offloads */
|
|
|
|
if (protocol == htons(ETH_P_IP))
|
|
|
|
tx_flags |= I40E_TX_FLAGS_IPV4;
|
|
|
|
else if (protocol == htons(ETH_P_IPV6))
|
|
|
|
tx_flags |= I40E_TX_FLAGS_IPV6;
|
|
|
|
|
2016-11-29 07:05:59 +07:00
|
|
|
tso = i40e_tso(first, &hdr_len, &cd_type_cmd_tso_mss);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
if (tso < 0)
|
|
|
|
goto out_drop;
|
|
|
|
else if (tso)
|
|
|
|
tx_flags |= I40E_TX_FLAGS_TSO;
|
|
|
|
|
|
|
|
/* Always offload the checksum, since it's in the data descriptor */
|
2016-01-25 12:17:10 +07:00
|
|
|
tso = i40e_tx_enable_csum(skb, &tx_flags, &td_cmd, &td_offset,
|
|
|
|
tx_ring, &cd_tunneling);
|
|
|
|
if (tso < 0)
|
|
|
|
goto out_drop;
|
2013-12-21 13:12:51 +07:00
|
|
|
|
2016-02-18 02:02:56 +07:00
|
|
|
skb_tx_timestamp(skb);
|
|
|
|
|
|
|
|
/* always enable CRC insertion offload */
|
|
|
|
td_cmd |= I40E_TX_DESC_CMD_ICRC;
|
|
|
|
|
2013-12-21 13:12:51 +07:00
|
|
|
i40e_create_tx_ctx(tx_ring, cd_type_cmd_tso_mss,
|
|
|
|
cd_tunneling, cd_l2tag2);
|
|
|
|
|
2015-04-17 07:06:10 +07:00
|
|
|
i40evf_tx_map(tx_ring, skb, first, tx_flags, hdr_len,
|
|
|
|
td_cmd, td_offset);
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
return NETDEV_TX_OK;
|
|
|
|
|
|
|
|
out_drop:
|
2017-04-13 15:45:44 +07:00
|
|
|
i40e_trace(xmit_frame_ring_drop, first->skb, tx_ring);
|
2016-11-29 07:05:59 +07:00
|
|
|
dev_kfree_skb_any(first->skb);
|
|
|
|
first->skb = NULL;
|
2013-12-21 13:12:51 +07:00
|
|
|
return NETDEV_TX_OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* i40evf_xmit_frame - Selects the correct VSI and Tx queue to send buffer
|
|
|
|
* @skb: send buffer
|
|
|
|
* @netdev: network interface device structure
|
|
|
|
*
|
|
|
|
* Returns NETDEV_TX_OK if sent, else an error code
|
|
|
|
**/
|
|
|
|
netdev_tx_t i40evf_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
|
|
|
|
{
|
|
|
|
struct i40evf_adapter *adapter = netdev_priv(netdev);
|
2015-10-27 06:44:40 +07:00
|
|
|
struct i40e_ring *tx_ring = &adapter->tx_rings[skb->queue_mapping];
|
2013-12-21 13:12:51 +07:00
|
|
|
|
|
|
|
/* hardware can't handle really short frames, hardware padding works
|
|
|
|
* beyond this point
|
|
|
|
*/
|
|
|
|
if (unlikely(skb->len < I40E_MIN_TX_LEN)) {
|
|
|
|
if (skb_pad(skb, I40E_MIN_TX_LEN - skb->len))
|
|
|
|
return NETDEV_TX_OK;
|
|
|
|
skb->len = I40E_MIN_TX_LEN;
|
|
|
|
skb_set_tail_pointer(skb, I40E_MIN_TX_LEN);
|
|
|
|
}
|
|
|
|
|
|
|
|
return i40e_xmit_frame_ring(skb, tx_ring);
|
|
|
|
}
|