Commit 4fe81585 authored by Jason Xing's avatar Jason Xing Committed by David S. Miller

ixgbe: let the xdpdrv work with more than 64 cpus

Originally, ixgbe driver doesn't allow the mounting of xdpdrv if the
server is equipped with more than 64 cpus online. So it turns out that
the loading of xdpdrv causes the "NOMEM" failure.

Actually, we can adjust the algorithm and then make it work through
mapping the current cpu to some xdp ring with the protect of @tx_lock.

Here are some numbers before/after applying this patch with xdp-example
loaded on the eth0X:

As client (tx path):
                     Before    After
TCP_STREAM send-64   734.14    714.20
TCP_STREAM send-128  1401.91   1395.05
TCP_STREAM send-512  5311.67   5292.84
TCP_STREAM send-1k   9277.40   9356.22 (not stable)
TCP_RR     send-1    22559.75  21844.22
TCP_RR     send-128  23169.54  22725.13
TCP_RR     send-512  21670.91  21412.56

As server (rx path):
                     Before    After
TCP_STREAM send-64   1416.49   1383.12
TCP_STREAM send-128  3141.49   3055.50
TCP_STREAM send-512  9488.73   9487.44
TCP_STREAM send-1k   9491.17   9356.22 (not stable)
TCP_RR     send-1    23617.74  23601.60
...

Notice: the TCP_RR mode is unstable as the official document explains.

I tested many times with different parameters combined through netperf.
Though the result is not that accurate, I cannot see much influence on
this patch. The static key is places on the hot path, but it actually
shouldn't cause a huge regression theoretically.
Co-developed-by: default avatarShujin Li <lishujin@kuaishou.com>
Signed-off-by: default avatarShujin Li <lishujin@kuaishou.com>
Signed-off-by: default avatarJason Xing <xingwanli@kuaishou.com>
Tested-by: default avatarSandeep Penigalapati <sandeep.penigalapati@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent a3e4abac
...@@ -351,6 +351,7 @@ struct ixgbe_ring { ...@@ -351,6 +351,7 @@ struct ixgbe_ring {
}; };
u16 rx_offset; u16 rx_offset;
struct xdp_rxq_info xdp_rxq; struct xdp_rxq_info xdp_rxq;
spinlock_t tx_lock; /* used in XDP mode */
struct xsk_buff_pool *xsk_pool; struct xsk_buff_pool *xsk_pool;
u16 ring_idx; /* {rx,tx,xdp}_ring back reference idx */ u16 ring_idx; /* {rx,tx,xdp}_ring back reference idx */
u16 rx_buf_len; u16 rx_buf_len;
...@@ -375,11 +376,13 @@ enum ixgbe_ring_f_enum { ...@@ -375,11 +376,13 @@ enum ixgbe_ring_f_enum {
#define IXGBE_MAX_FCOE_INDICES 8 #define IXGBE_MAX_FCOE_INDICES 8
#define MAX_RX_QUEUES (IXGBE_MAX_FDIR_INDICES + 1) #define MAX_RX_QUEUES (IXGBE_MAX_FDIR_INDICES + 1)
#define MAX_TX_QUEUES (IXGBE_MAX_FDIR_INDICES + 1) #define MAX_TX_QUEUES (IXGBE_MAX_FDIR_INDICES + 1)
#define MAX_XDP_QUEUES (IXGBE_MAX_FDIR_INDICES + 1) #define IXGBE_MAX_XDP_QS (IXGBE_MAX_FDIR_INDICES + 1)
#define IXGBE_MAX_L2A_QUEUES 4 #define IXGBE_MAX_L2A_QUEUES 4
#define IXGBE_BAD_L2A_QUEUE 3 #define IXGBE_BAD_L2A_QUEUE 3
#define IXGBE_MAX_MACVLANS 63 #define IXGBE_MAX_MACVLANS 63
DECLARE_STATIC_KEY_FALSE(ixgbe_xdp_locking_key);
struct ixgbe_ring_feature { struct ixgbe_ring_feature {
u16 limit; /* upper limit on feature indices */ u16 limit; /* upper limit on feature indices */
u16 indices; /* current value of indices */ u16 indices; /* current value of indices */
...@@ -629,7 +632,7 @@ struct ixgbe_adapter { ...@@ -629,7 +632,7 @@ struct ixgbe_adapter {
/* XDP */ /* XDP */
int num_xdp_queues; int num_xdp_queues;
struct ixgbe_ring *xdp_ring[MAX_XDP_QUEUES]; struct ixgbe_ring *xdp_ring[IXGBE_MAX_XDP_QS];
unsigned long *af_xdp_zc_qps; /* tracks AF_XDP ZC enabled rings */ unsigned long *af_xdp_zc_qps; /* tracks AF_XDP ZC enabled rings */
/* TX */ /* TX */
...@@ -772,6 +775,22 @@ struct ixgbe_adapter { ...@@ -772,6 +775,22 @@ struct ixgbe_adapter {
#endif /* CONFIG_IXGBE_IPSEC */ #endif /* CONFIG_IXGBE_IPSEC */
}; };
static inline int ixgbe_determine_xdp_q_idx(int cpu)
{
if (static_key_enabled(&ixgbe_xdp_locking_key))
return cpu % IXGBE_MAX_XDP_QS;
else
return cpu;
}
static inline
struct ixgbe_ring *ixgbe_determine_xdp_ring(struct ixgbe_adapter *adapter)
{
int index = ixgbe_determine_xdp_q_idx(smp_processor_id());
return adapter->xdp_ring[index];
}
static inline u8 ixgbe_max_rss_indices(struct ixgbe_adapter *adapter) static inline u8 ixgbe_max_rss_indices(struct ixgbe_adapter *adapter)
{ {
switch (adapter->hw.mac.type) { switch (adapter->hw.mac.type) {
......
...@@ -299,7 +299,10 @@ static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter) ...@@ -299,7 +299,10 @@ static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
static int ixgbe_xdp_queues(struct ixgbe_adapter *adapter) static int ixgbe_xdp_queues(struct ixgbe_adapter *adapter)
{ {
return adapter->xdp_prog ? nr_cpu_ids : 0; int queues;
queues = min_t(int, IXGBE_MAX_XDP_QS, nr_cpu_ids);
return adapter->xdp_prog ? queues : 0;
} }
#define IXGBE_RSS_64Q_MASK 0x3F #define IXGBE_RSS_64Q_MASK 0x3F
...@@ -947,6 +950,7 @@ static int ixgbe_alloc_q_vector(struct ixgbe_adapter *adapter, ...@@ -947,6 +950,7 @@ static int ixgbe_alloc_q_vector(struct ixgbe_adapter *adapter,
ring->count = adapter->tx_ring_count; ring->count = adapter->tx_ring_count;
ring->queue_index = xdp_idx; ring->queue_index = xdp_idx;
set_ring_xdp(ring); set_ring_xdp(ring);
spin_lock_init(&ring->tx_lock);
/* assign ring to adapter */ /* assign ring to adapter */
WRITE_ONCE(adapter->xdp_ring[xdp_idx], ring); WRITE_ONCE(adapter->xdp_ring[xdp_idx], ring);
...@@ -1032,6 +1036,9 @@ static void ixgbe_free_q_vector(struct ixgbe_adapter *adapter, int v_idx) ...@@ -1032,6 +1036,9 @@ static void ixgbe_free_q_vector(struct ixgbe_adapter *adapter, int v_idx)
adapter->q_vector[v_idx] = NULL; adapter->q_vector[v_idx] = NULL;
__netif_napi_del(&q_vector->napi); __netif_napi_del(&q_vector->napi);
if (static_key_enabled(&ixgbe_xdp_locking_key))
static_branch_dec(&ixgbe_xdp_locking_key);
/* /*
* after a call to __netif_napi_del() napi may still be used and * after a call to __netif_napi_del() napi may still be used and
* ixgbe_get_stats64() might access the rings on this vector, * ixgbe_get_stats64() might access the rings on this vector,
......
...@@ -165,6 +165,9 @@ MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>"); ...@@ -165,6 +165,9 @@ MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
MODULE_DESCRIPTION("Intel(R) 10 Gigabit PCI Express Network Driver"); MODULE_DESCRIPTION("Intel(R) 10 Gigabit PCI Express Network Driver");
MODULE_LICENSE("GPL v2"); MODULE_LICENSE("GPL v2");
DEFINE_STATIC_KEY_FALSE(ixgbe_xdp_locking_key);
EXPORT_SYMBOL(ixgbe_xdp_locking_key);
static struct workqueue_struct *ixgbe_wq; static struct workqueue_struct *ixgbe_wq;
static bool ixgbe_check_cfg_remove(struct ixgbe_hw *hw, struct pci_dev *pdev); static bool ixgbe_check_cfg_remove(struct ixgbe_hw *hw, struct pci_dev *pdev);
...@@ -2197,6 +2200,7 @@ static struct sk_buff *ixgbe_run_xdp(struct ixgbe_adapter *adapter, ...@@ -2197,6 +2200,7 @@ static struct sk_buff *ixgbe_run_xdp(struct ixgbe_adapter *adapter,
{ {
int err, result = IXGBE_XDP_PASS; int err, result = IXGBE_XDP_PASS;
struct bpf_prog *xdp_prog; struct bpf_prog *xdp_prog;
struct ixgbe_ring *ring;
struct xdp_frame *xdpf; struct xdp_frame *xdpf;
u32 act; u32 act;
...@@ -2215,7 +2219,12 @@ static struct sk_buff *ixgbe_run_xdp(struct ixgbe_adapter *adapter, ...@@ -2215,7 +2219,12 @@ static struct sk_buff *ixgbe_run_xdp(struct ixgbe_adapter *adapter,
xdpf = xdp_convert_buff_to_frame(xdp); xdpf = xdp_convert_buff_to_frame(xdp);
if (unlikely(!xdpf)) if (unlikely(!xdpf))
goto out_failure; goto out_failure;
result = ixgbe_xmit_xdp_ring(adapter, xdpf); ring = ixgbe_determine_xdp_ring(adapter);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_lock(&ring->tx_lock);
result = ixgbe_xmit_xdp_ring(ring, xdpf);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_unlock(&ring->tx_lock);
if (result == IXGBE_XDP_CONSUMED) if (result == IXGBE_XDP_CONSUMED)
goto out_failure; goto out_failure;
break; break;
...@@ -2422,13 +2431,9 @@ static int ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector, ...@@ -2422,13 +2431,9 @@ static int ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
xdp_do_flush_map(); xdp_do_flush_map();
if (xdp_xmit & IXGBE_XDP_TX) { if (xdp_xmit & IXGBE_XDP_TX) {
struct ixgbe_ring *ring = adapter->xdp_ring[smp_processor_id()]; struct ixgbe_ring *ring = ixgbe_determine_xdp_ring(adapter);
/* Force memory writes to complete before letting h/w ixgbe_xdp_ring_update_tail_locked(ring);
* know there are new descriptors to fetch.
*/
wmb();
writel(ring->next_to_use, ring->tail);
} }
u64_stats_update_begin(&rx_ring->syncp); u64_stats_update_begin(&rx_ring->syncp);
...@@ -6320,7 +6325,7 @@ static int ixgbe_sw_init(struct ixgbe_adapter *adapter, ...@@ -6320,7 +6325,7 @@ static int ixgbe_sw_init(struct ixgbe_adapter *adapter,
if (ixgbe_init_rss_key(adapter)) if (ixgbe_init_rss_key(adapter))
return -ENOMEM; return -ENOMEM;
adapter->af_xdp_zc_qps = bitmap_zalloc(MAX_XDP_QUEUES, GFP_KERNEL); adapter->af_xdp_zc_qps = bitmap_zalloc(IXGBE_MAX_XDP_QS, GFP_KERNEL);
if (!adapter->af_xdp_zc_qps) if (!adapter->af_xdp_zc_qps)
return -ENOMEM; return -ENOMEM;
...@@ -8536,10 +8541,9 @@ static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb, ...@@ -8536,10 +8541,9 @@ static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb,
} }
#endif #endif
int ixgbe_xmit_xdp_ring(struct ixgbe_adapter *adapter, int ixgbe_xmit_xdp_ring(struct ixgbe_ring *ring,
struct xdp_frame *xdpf) struct xdp_frame *xdpf)
{ {
struct ixgbe_ring *ring = adapter->xdp_ring[smp_processor_id()];
struct ixgbe_tx_buffer *tx_buffer; struct ixgbe_tx_buffer *tx_buffer;
union ixgbe_adv_tx_desc *tx_desc; union ixgbe_adv_tx_desc *tx_desc;
u32 len, cmd_type; u32 len, cmd_type;
...@@ -10130,8 +10134,13 @@ static int ixgbe_xdp_setup(struct net_device *dev, struct bpf_prog *prog) ...@@ -10130,8 +10134,13 @@ static int ixgbe_xdp_setup(struct net_device *dev, struct bpf_prog *prog)
return -EINVAL; return -EINVAL;
} }
if (nr_cpu_ids > MAX_XDP_QUEUES) /* if the number of cpus is much larger than the maximum of queues,
* we should stop it and then return with ENOMEM like before.
*/
if (nr_cpu_ids > IXGBE_MAX_XDP_QS * 2)
return -ENOMEM; return -ENOMEM;
else if (nr_cpu_ids > IXGBE_MAX_XDP_QS)
static_branch_inc(&ixgbe_xdp_locking_key);
old_prog = xchg(&adapter->xdp_prog, prog); old_prog = xchg(&adapter->xdp_prog, prog);
need_reset = (!!prog != !!old_prog); need_reset = (!!prog != !!old_prog);
...@@ -10195,6 +10204,15 @@ void ixgbe_xdp_ring_update_tail(struct ixgbe_ring *ring) ...@@ -10195,6 +10204,15 @@ void ixgbe_xdp_ring_update_tail(struct ixgbe_ring *ring)
writel(ring->next_to_use, ring->tail); writel(ring->next_to_use, ring->tail);
} }
void ixgbe_xdp_ring_update_tail_locked(struct ixgbe_ring *ring)
{
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_lock(&ring->tx_lock);
ixgbe_xdp_ring_update_tail(ring);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_unlock(&ring->tx_lock);
}
static int ixgbe_xdp_xmit(struct net_device *dev, int n, static int ixgbe_xdp_xmit(struct net_device *dev, int n,
struct xdp_frame **frames, u32 flags) struct xdp_frame **frames, u32 flags)
{ {
...@@ -10212,18 +10230,21 @@ static int ixgbe_xdp_xmit(struct net_device *dev, int n, ...@@ -10212,18 +10230,21 @@ static int ixgbe_xdp_xmit(struct net_device *dev, int n,
/* During program transitions its possible adapter->xdp_prog is assigned /* During program transitions its possible adapter->xdp_prog is assigned
* but ring has not been configured yet. In this case simply abort xmit. * but ring has not been configured yet. In this case simply abort xmit.
*/ */
ring = adapter->xdp_prog ? adapter->xdp_ring[smp_processor_id()] : NULL; ring = adapter->xdp_prog ? ixgbe_determine_xdp_ring(adapter) : NULL;
if (unlikely(!ring)) if (unlikely(!ring))
return -ENXIO; return -ENXIO;
if (unlikely(test_bit(__IXGBE_TX_DISABLED, &ring->state))) if (unlikely(test_bit(__IXGBE_TX_DISABLED, &ring->state)))
return -ENXIO; return -ENXIO;
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_lock(&ring->tx_lock);
for (i = 0; i < n; i++) { for (i = 0; i < n; i++) {
struct xdp_frame *xdpf = frames[i]; struct xdp_frame *xdpf = frames[i];
int err; int err;
err = ixgbe_xmit_xdp_ring(adapter, xdpf); err = ixgbe_xmit_xdp_ring(ring, xdpf);
if (err != IXGBE_XDP_TX) if (err != IXGBE_XDP_TX)
break; break;
nxmit++; nxmit++;
...@@ -10232,6 +10253,9 @@ static int ixgbe_xdp_xmit(struct net_device *dev, int n, ...@@ -10232,6 +10253,9 @@ static int ixgbe_xdp_xmit(struct net_device *dev, int n,
if (unlikely(flags & XDP_XMIT_FLUSH)) if (unlikely(flags & XDP_XMIT_FLUSH))
ixgbe_xdp_ring_update_tail(ring); ixgbe_xdp_ring_update_tail(ring);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_unlock(&ring->tx_lock);
return nxmit; return nxmit;
} }
......
...@@ -12,7 +12,7 @@ ...@@ -12,7 +12,7 @@
#define IXGBE_TXD_CMD (IXGBE_TXD_CMD_EOP | \ #define IXGBE_TXD_CMD (IXGBE_TXD_CMD_EOP | \
IXGBE_TXD_CMD_RS) IXGBE_TXD_CMD_RS)
int ixgbe_xmit_xdp_ring(struct ixgbe_adapter *adapter, int ixgbe_xmit_xdp_ring(struct ixgbe_ring *ring,
struct xdp_frame *xdpf); struct xdp_frame *xdpf);
bool ixgbe_cleanup_headers(struct ixgbe_ring *rx_ring, bool ixgbe_cleanup_headers(struct ixgbe_ring *rx_ring,
union ixgbe_adv_rx_desc *rx_desc, union ixgbe_adv_rx_desc *rx_desc,
...@@ -23,6 +23,7 @@ void ixgbe_process_skb_fields(struct ixgbe_ring *rx_ring, ...@@ -23,6 +23,7 @@ void ixgbe_process_skb_fields(struct ixgbe_ring *rx_ring,
void ixgbe_rx_skb(struct ixgbe_q_vector *q_vector, void ixgbe_rx_skb(struct ixgbe_q_vector *q_vector,
struct sk_buff *skb); struct sk_buff *skb);
void ixgbe_xdp_ring_update_tail(struct ixgbe_ring *ring); void ixgbe_xdp_ring_update_tail(struct ixgbe_ring *ring);
void ixgbe_xdp_ring_update_tail_locked(struct ixgbe_ring *ring);
void ixgbe_irq_rearm_queues(struct ixgbe_adapter *adapter, u64 qmask); void ixgbe_irq_rearm_queues(struct ixgbe_adapter *adapter, u64 qmask);
void ixgbe_txrx_ring_disable(struct ixgbe_adapter *adapter, int ring); void ixgbe_txrx_ring_disable(struct ixgbe_adapter *adapter, int ring);
......
...@@ -100,6 +100,7 @@ static int ixgbe_run_xdp_zc(struct ixgbe_adapter *adapter, ...@@ -100,6 +100,7 @@ static int ixgbe_run_xdp_zc(struct ixgbe_adapter *adapter,
{ {
int err, result = IXGBE_XDP_PASS; int err, result = IXGBE_XDP_PASS;
struct bpf_prog *xdp_prog; struct bpf_prog *xdp_prog;
struct ixgbe_ring *ring;
struct xdp_frame *xdpf; struct xdp_frame *xdpf;
u32 act; u32 act;
...@@ -120,7 +121,12 @@ static int ixgbe_run_xdp_zc(struct ixgbe_adapter *adapter, ...@@ -120,7 +121,12 @@ static int ixgbe_run_xdp_zc(struct ixgbe_adapter *adapter,
xdpf = xdp_convert_buff_to_frame(xdp); xdpf = xdp_convert_buff_to_frame(xdp);
if (unlikely(!xdpf)) if (unlikely(!xdpf))
goto out_failure; goto out_failure;
result = ixgbe_xmit_xdp_ring(adapter, xdpf); ring = ixgbe_determine_xdp_ring(adapter);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_lock(&ring->tx_lock);
result = ixgbe_xmit_xdp_ring(ring, xdpf);
if (static_branch_unlikely(&ixgbe_xdp_locking_key))
spin_unlock(&ring->tx_lock);
if (result == IXGBE_XDP_CONSUMED) if (result == IXGBE_XDP_CONSUMED)
goto out_failure; goto out_failure;
break; break;
...@@ -334,13 +340,9 @@ int ixgbe_clean_rx_irq_zc(struct ixgbe_q_vector *q_vector, ...@@ -334,13 +340,9 @@ int ixgbe_clean_rx_irq_zc(struct ixgbe_q_vector *q_vector,
xdp_do_flush_map(); xdp_do_flush_map();
if (xdp_xmit & IXGBE_XDP_TX) { if (xdp_xmit & IXGBE_XDP_TX) {
struct ixgbe_ring *ring = adapter->xdp_ring[smp_processor_id()]; struct ixgbe_ring *ring = ixgbe_determine_xdp_ring(adapter);
/* Force memory writes to complete before letting h/w ixgbe_xdp_ring_update_tail_locked(ring);
* know there are new descriptors to fetch.
*/
wmb();
writel(ring->next_to_use, ring->tail);
} }
u64_stats_update_begin(&rx_ring->syncp); u64_stats_update_begin(&rx_ring->syncp);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment