This patchset covers optimizations in rx (first 7 patches)
and tx (remaining 5 patches) data path.
Running UDP DL/UL traffic on IPQ8074 5G radio showed an average 5-10%
improvement on a 4 core platform
---
v2:
- Addressed internal developer reported segfault and avoid lookup twice
by utilizing idr_remove (patch 12/12 and patch 2/12).
---
P Praneesh (12):
ath11k: disable unused CE8 interrupts for ipq8074
ath11k: allocate dst ring descriptors from cacheable memory
ath11k: modify dp_rx desc access wrapper calls inline
ath11k: avoid additional access to ath11k_hal_srng_dst_num_free
ath11k: avoid active pdev check for each msdu
ath11k: remove usage quota while processing rx packets
ath11k: add branch predictors in process_rx
ath11k: allocate HAL_WBM2SW_RELEASE ring from cacheable memory
ath11k: remove mod operator in dst ring processing
ath11k: avoid while loop in ring selection of tx completion interrupt
ath11k: add branch predictors in dp_tx path
ath11k: avoid unnecessary lock contention in tx_completion path
drivers/net/wireless/ath/ath11k/ce.c | 2 +-
drivers/net/wireless/ath/ath11k/core.c | 5 +
drivers/net/wireless/ath/ath11k/dp.c | 48 ++++++--
drivers/net/wireless/ath/ath11k/dp.h | 1 +
drivers/net/wireless/ath/ath11k/dp_rx.c | 205 ++++++++++++++++----------------
drivers/net/wireless/ath/ath11k/dp_tx.c | 76 ++++++------
drivers/net/wireless/ath/ath11k/hal.c | 35 +++++-
drivers/net/wireless/ath/ath11k/hal.h | 1 +
drivers/net/wireless/ath/ath11k/hw.h | 1 +
drivers/net/wireless/ath/ath11k/mac.c | 2 +-
10 files changed, 219 insertions(+), 157 deletions(-)
--
2.7.4
In ath11k_dp_process_rx(), after processing rx_desc from
ath11k_hal_srng_dst_get_next_entry(), ath11k_hal_srng_dst_num_free()
is accessed everytime because of done flag is not set.
To avoid this additional access to ath11k_hal_srng_dst_num_free(),
increment total_msdu_reaped only when continuation is not set and
update done flag correspondingly.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_rx.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
index b84c2db..994959b 100644
--- a/drivers/net/wireless/ath/ath11k/dp_rx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
@@ -2623,7 +2623,6 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
DMA_FROM_DEVICE);
num_buffs_reaped[mac_id]++;
- total_msdu_reaped++;
push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON,
desc.info0);
@@ -2646,10 +2645,15 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
__skb_queue_tail(&msdu_list, msdu);
- if (total_msdu_reaped >= quota && !rxcb->is_continuation) {
+ if (rxcb->is_continuation) {
+ done = false;
+ } else {
+ total_msdu_reaped++;
done = true;
- break;
}
+
+ if (total_msdu_reaped >= budget)
+ break;
}
/* Hw might have updated the head pointer after we cached it.
--
2.7.4
The Active Pdev and CAC check are done for each msdu in
ath11k_dp_rx_process_received_packets which is a overhead.
To avoid this overhead, collect all msdus in a per mac msdu
list and pass to function.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_rx.c | 66 ++++++++++++++++-----------------
1 file changed, 32 insertions(+), 34 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
index 994959b..1d85e10 100644
--- a/drivers/net/wireless/ath/ath11k/dp_rx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
@@ -2523,12 +2523,10 @@ static int ath11k_dp_rx_process_msdu(struct ath11k *ar,
static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
struct napi_struct *napi,
struct sk_buff_head *msdu_list,
- int *quota, int ring_id)
+ int *quota, int mac_id)
{
- struct ath11k_skb_rxcb *rxcb;
struct sk_buff *msdu;
struct ath11k *ar;
- u8 mac_id;
int ret;
if (skb_queue_empty(msdu_list))
@@ -2536,20 +2534,20 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
rcu_read_lock();
- while (*quota && (msdu = __skb_dequeue(msdu_list))) {
- rxcb = ATH11K_SKB_RXCB(msdu);
- mac_id = rxcb->mac_id;
- ar = ab->pdevs[mac_id].ar;
- if (!rcu_dereference(ab->pdevs_active[mac_id])) {
- dev_kfree_skb_any(msdu);
- continue;
- }
+ ar = ab->pdevs[mac_id].ar;
+ if (!rcu_dereference(ab->pdevs_active[mac_id])) {
+ __skb_queue_purge(msdu_list);
+ rcu_read_unlock();
+ return;
+ }
- if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
- dev_kfree_skb_any(msdu);
- continue;
- }
+ if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
+ __skb_queue_purge(msdu_list);
+ rcu_read_unlock();
+ return;
+ }
+ while ((msdu = __skb_dequeue(msdu_list))) {
ret = ath11k_dp_rx_process_msdu(ar, msdu, msdu_list);
if (ret) {
ath11k_dbg(ab, ATH11K_DBG_DATA,
@@ -2571,7 +2569,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
struct ath11k_dp *dp = &ab->dp;
struct dp_rxdma_ring *rx_ring;
int num_buffs_reaped[MAX_RADIOS] = {0};
- struct sk_buff_head msdu_list;
+ struct sk_buff_head msdu_list[MAX_RADIOS];
struct ath11k_skb_rxcb *rxcb;
int total_msdu_reaped = 0;
struct hal_srng *srng;
@@ -2580,10 +2578,13 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
bool done = false;
int buf_id, mac_id;
struct ath11k *ar;
- u32 *rx_desc;
+ struct hal_reo_dest_ring *desc;
+ enum hal_reo_dest_ring_push_reason push_reason;
+ u32 cookie;
int i;
- __skb_queue_head_init(&msdu_list);
+ for (i = 0; i < MAX_RADIOS; i++)
+ __skb_queue_head_init(&msdu_list[i]);
srng = &ab->hal.srng_list[dp->reo_dst_ring[ring_id].ring_id];
@@ -2592,13 +2593,11 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
ath11k_hal_srng_access_begin(ab, srng);
try_again:
- while ((rx_desc = ath11k_hal_srng_dst_get_next_entry(ab, srng))) {
- struct hal_reo_dest_ring desc = *(struct hal_reo_dest_ring *)rx_desc;
- enum hal_reo_dest_ring_push_reason push_reason;
- u32 cookie;
-
+ while (likely(desc =
+ (struct hal_reo_dest_ring *)ath11k_hal_srng_dst_get_next_entry(ab,
+ srng))) {
cookie = FIELD_GET(BUFFER_ADDR_INFO1_SW_COOKIE,
- desc.buf_addr_info.info1);
+ desc->buf_addr_info.info1);
buf_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_BUF_ID,
cookie);
mac_id = FIELD_GET(DP_RXDMA_BUF_COOKIE_PDEV_ID, cookie);
@@ -2625,7 +2624,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
num_buffs_reaped[mac_id]++;
push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON,
- desc.info0);
+ desc->info0);
if (push_reason !=
HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION) {
dev_kfree_skb_any(msdu);
@@ -2633,17 +2632,17 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
continue;
}
- rxcb->is_first_msdu = !!(desc.rx_msdu_info.info0 &
+ rxcb->is_first_msdu = !!(desc->rx_msdu_info.info0 &
RX_MSDU_DESC_INFO0_FIRST_MSDU_IN_MPDU);
- rxcb->is_last_msdu = !!(desc.rx_msdu_info.info0 &
+ rxcb->is_last_msdu = !!(desc->rx_msdu_info.info0 &
RX_MSDU_DESC_INFO0_LAST_MSDU_IN_MPDU);
- rxcb->is_continuation = !!(desc.rx_msdu_info.info0 &
+ rxcb->is_continuation = !!(desc->rx_msdu_info.info0 &
RX_MSDU_DESC_INFO0_MSDU_CONTINUATION);
rxcb->mac_id = mac_id;
rxcb->tid = FIELD_GET(HAL_REO_DEST_RING_INFO0_RX_QUEUE_NUM,
- desc.info0);
+ desc->info0);
- __skb_queue_tail(&msdu_list, msdu);
+ __skb_queue_tail(&msdu_list[mac_id], msdu);
if (rxcb->is_continuation) {
done = false;
@@ -2678,16 +2677,15 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
if (!num_buffs_reaped[i])
continue;
+ ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i],
+ "a, i);
+
ar = ab->pdevs[i].ar;
rx_ring = &ar->dp.rx_refill_buf_ring;
ath11k_dp_rxbufs_replenish(ab, i, rx_ring, num_buffs_reaped[i],
HAL_RX_BUF_RBM_SW3_BM);
}
-
- ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list,
- "a, ring_id);
-
exit:
return budget - quota;
}
--
2.7.4
In data path, to reduce the CPU cycles spending on descriptor access
wrapper function, changed those functions as static inline.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_rx.c | 114 +++++++++++++++++---------------
1 file changed, 59 insertions(+), 55 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
index 9a22481..b84c2db 100644
--- a/drivers/net/wireless/ath/ath11k/dp_rx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
@@ -20,13 +20,15 @@
#define ATH11K_DP_RX_FRAGMENT_TIMEOUT_MS (2 * HZ)
-static u8 *ath11k_dp_rx_h_80211_hdr(struct ath11k_base *ab, struct hal_rx_desc *desc)
+static inline
+u8 *ath11k_dp_rx_h_80211_hdr(struct ath11k_base *ab, struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_hdr_status(desc);
}
-static enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline
+enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
if (!ab->hw_params.hw_ops->rx_desc_encrypt_valid(desc))
return HAL_ENCRYPT_TYPE_OPEN;
@@ -34,32 +36,34 @@ static enum hal_encrypt_type ath11k_dp_rx_h_mpdu_start_enctype(struct ath11k_bas
return ab->hw_params.hw_ops->rx_desc_get_encrypt_type(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_decap_type(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_decap_type(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_decap_type(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_mesh_ctl_present(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline
+u8 ath11k_dp_rx_h_msdu_start_mesh_ctl_present(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mesh_ctl(desc);
}
-static bool ath11k_dp_rx_h_mpdu_start_seq_ctrl_valid(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline
+bool ath11k_dp_rx_h_mpdu_start_seq_ctrl_valid(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_seq_ctl_vld(desc);
}
-static bool ath11k_dp_rx_h_mpdu_start_fc_valid(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline bool ath11k_dp_rx_h_mpdu_start_fc_valid(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_fc_valid(desc);
}
-static bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab,
- struct sk_buff *skb)
+static inline bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab,
+ struct sk_buff *skb)
{
struct ieee80211_hdr *hdr;
@@ -67,8 +71,8 @@ static bool ath11k_dp_rx_h_mpdu_start_more_frags(struct ath11k_base *ab,
return ieee80211_has_morefrags(hdr->frame_control);
}
-static u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab,
- struct sk_buff *skb)
+static inline u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab,
+ struct sk_buff *skb)
{
struct ieee80211_hdr *hdr;
@@ -76,37 +80,37 @@ static u16 ath11k_dp_rx_h_mpdu_start_frag_no(struct ath11k_base *ab,
return le16_to_cpu(hdr->seq_ctrl) & IEEE80211_SCTL_FRAG;
}
-static u16 ath11k_dp_rx_h_mpdu_start_seq_no(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u16 ath11k_dp_rx_h_mpdu_start_seq_no(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_start_seq_no(desc);
}
-static void *ath11k_dp_rx_get_attention(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline void *ath11k_dp_rx_get_attention(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_attention(desc);
}
-static bool ath11k_dp_rx_h_attn_msdu_done(struct rx_attention *attn)
+static inline bool ath11k_dp_rx_h_attn_msdu_done(struct rx_attention *attn)
{
return !!FIELD_GET(RX_ATTENTION_INFO2_MSDU_DONE,
__le32_to_cpu(attn->info2));
}
-static bool ath11k_dp_rx_h_attn_l4_cksum_fail(struct rx_attention *attn)
+static inline bool ath11k_dp_rx_h_attn_l4_cksum_fail(struct rx_attention *attn)
{
return !!FIELD_GET(RX_ATTENTION_INFO1_TCP_UDP_CKSUM_FAIL,
__le32_to_cpu(attn->info1));
}
-static bool ath11k_dp_rx_h_attn_ip_cksum_fail(struct rx_attention *attn)
+static inline bool ath11k_dp_rx_h_attn_ip_cksum_fail(struct rx_attention *attn)
{
return !!FIELD_GET(RX_ATTENTION_INFO1_IP_CKSUM_FAIL,
__le32_to_cpu(attn->info1));
}
-static bool ath11k_dp_rx_h_attn_is_decrypted(struct rx_attention *attn)
+static inline bool ath11k_dp_rx_h_attn_is_decrypted(struct rx_attention *attn)
{
return (FIELD_GET(RX_ATTENTION_INFO2_DCRYPT_STATUS_CODE,
__le32_to_cpu(attn->info2)) ==
@@ -142,68 +146,68 @@ static u32 ath11k_dp_rx_h_attn_mpdu_err(struct rx_attention *attn)
return errmap;
}
-static u16 ath11k_dp_rx_h_msdu_start_msdu_len(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u16 ath11k_dp_rx_h_msdu_start_msdu_len(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_len(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_sgi(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_sgi(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_sgi(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_rate_mcs(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_rate_mcs(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_rate_mcs(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_rx_bw(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_rx_bw(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_rx_bw(desc);
}
-static u32 ath11k_dp_rx_h_msdu_start_freq(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u32 ath11k_dp_rx_h_msdu_start_freq(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_freq(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_pkt_type(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_pkt_type(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_msdu_pkt_type(desc);
}
-static u8 ath11k_dp_rx_h_msdu_start_nss(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_start_nss(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return hweight8(ab->hw_params.hw_ops->rx_desc_get_msdu_nss(desc));
}
-static u8 ath11k_dp_rx_h_mpdu_start_tid(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_mpdu_start_tid(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_tid(desc);
}
-static u16 ath11k_dp_rx_h_mpdu_start_peer_id(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u16 ath11k_dp_rx_h_mpdu_start_peer_id(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_peer_id(desc);
}
-static u8 ath11k_dp_rx_h_msdu_end_l3pad(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline u8 ath11k_dp_rx_h_msdu_end_l3pad(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_l3_pad_bytes(desc);
}
-static bool ath11k_dp_rx_h_msdu_end_first_msdu(struct ath11k_base *ab,
- struct hal_rx_desc *desc)
+static inline bool ath11k_dp_rx_h_msdu_end_first_msdu(struct ath11k_base *ab,
+ struct hal_rx_desc *desc)
{
return ab->hw_params.hw_ops->rx_desc_get_first_msdu(desc);
}
@@ -221,14 +225,14 @@ static void ath11k_dp_rx_desc_end_tlv_copy(struct ath11k_base *ab,
ab->hw_params.hw_ops->rx_desc_copy_attn_end_tlv(fdesc, ldesc);
}
-static u32 ath11k_dp_rxdesc_get_mpdulen_err(struct rx_attention *attn)
+static inline u32 ath11k_dp_rxdesc_get_mpdulen_err(struct rx_attention *attn)
{
return FIELD_GET(RX_ATTENTION_INFO1_MPDU_LEN_ERR,
__le32_to_cpu(attn->info1));
}
-static u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab,
- struct hal_rx_desc *rx_desc)
+static inline u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab,
+ struct hal_rx_desc *rx_desc)
{
u8 *rx_pkt_hdr;
@@ -237,8 +241,8 @@ static u8 *ath11k_dp_rxdesc_get_80211hdr(struct ath11k_base *ab,
return rx_pkt_hdr;
}
-static bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab,
- struct hal_rx_desc *rx_desc)
+static inline bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab,
+ struct hal_rx_desc *rx_desc)
{
u32 tlv_tag;
@@ -247,15 +251,15 @@ static bool ath11k_dp_rxdesc_mpdu_valid(struct ath11k_base *ab,
return tlv_tag == HAL_RX_MPDU_START;
}
-static u32 ath11k_dp_rxdesc_get_ppduid(struct ath11k_base *ab,
- struct hal_rx_desc *rx_desc)
+static inline u32 ath11k_dp_rxdesc_get_ppduid(struct ath11k_base *ab,
+ struct hal_rx_desc *rx_desc)
{
return ab->hw_params.hw_ops->rx_desc_get_mpdu_ppdu_id(rx_desc);
}
-static void ath11k_dp_rxdesc_set_msdu_len(struct ath11k_base *ab,
- struct hal_rx_desc *desc,
- u16 len)
+static inline void ath11k_dp_rxdesc_set_msdu_len(struct ath11k_base *ab,
+ struct hal_rx_desc *desc,
+ u16 len)
{
ab->hw_params.hw_ops->rx_desc_set_msdu_len(desc, len);
}
--
2.7.4
tcl_data and reo_dst rings are currently being allocated using
dma_allocate_coherent() which is non cacheable.
Allocating ring memory from cacheable memory area allows cached
descriptor access and prefetch next descriptors to optimize CPU
usage during descriptor processing on NAPI. Based on the hardware
param we can enable/disable this feature for the corresponding
platform.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Pradeep Kumar Chitrapu <[email protected]>
Signed-off-by: Pradeep Kumar Chitrapu <[email protected]>
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/core.c | 5 +++++
drivers/net/wireless/ath/ath11k/dp.c | 38 +++++++++++++++++++++++++++++-----
drivers/net/wireless/ath/ath11k/dp.h | 1 +
drivers/net/wireless/ath/ath11k/hal.c | 28 +++++++++++++++++++++++--
drivers/net/wireless/ath/ath11k/hal.h | 1 +
drivers/net/wireless/ath/ath11k/hw.h | 1 +
6 files changed, 67 insertions(+), 7 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/core.c b/drivers/net/wireless/ath/ath11k/core.c
index 969bf1a..298c4dc 100644
--- a/drivers/net/wireless/ath/ath11k/core.c
+++ b/drivers/net/wireless/ath/ath11k/core.c
@@ -71,6 +71,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
.supports_suspend = false,
.hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074),
.fix_l1ss = true,
+ .alloc_cacheable_memory = true,
},
{
.hw_rev = ATH11K_HW_IPQ6018_HW10,
@@ -112,6 +113,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
.supports_suspend = false,
.hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074),
.fix_l1ss = true,
+ .alloc_cacheable_memory = true,
},
{
.name = "qca6390 hw2.0",
@@ -152,6 +154,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
.supports_suspend = true,
.hal_desc_sz = sizeof(struct hal_rx_desc_ipq8074),
.fix_l1ss = true,
+ .alloc_cacheable_memory = false,
},
{
.name = "qcn9074 hw1.0",
@@ -190,6 +193,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
.supports_suspend = false,
.hal_desc_sz = sizeof(struct hal_rx_desc_qcn9074),
.fix_l1ss = true,
+ .alloc_cacheable_memory = true,
},
{
.name = "wcn6855 hw2.0",
@@ -230,6 +234,7 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
.supports_suspend = true,
.hal_desc_sz = sizeof(struct hal_rx_desc_wcn6855),
.fix_l1ss = false,
+ .alloc_cacheable_memory = false,
},
};
diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c
index b0c8f62..57059c6 100644
--- a/drivers/net/wireless/ath/ath11k/dp.c
+++ b/drivers/net/wireless/ath/ath11k/dp.c
@@ -101,8 +101,11 @@ void ath11k_dp_srng_cleanup(struct ath11k_base *ab, struct dp_srng *ring)
if (!ring->vaddr_unaligned)
return;
- dma_free_coherent(ab->dev, ring->size, ring->vaddr_unaligned,
- ring->paddr_unaligned);
+ if (ring->cached)
+ kfree(ring->vaddr_unaligned);
+ else
+ dma_free_coherent(ab->dev, ring->size, ring->vaddr_unaligned,
+ ring->paddr_unaligned);
ring->vaddr_unaligned = NULL;
}
@@ -222,6 +225,7 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring,
int entry_sz = ath11k_hal_srng_get_entrysize(ab, type);
int max_entries = ath11k_hal_srng_get_max_entries(ab, type);
int ret;
+ bool cached = false;
if (max_entries < 0 || entry_sz < 0)
return -EINVAL;
@@ -230,9 +234,28 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring,
num_entries = max_entries;
ring->size = (num_entries * entry_sz) + HAL_RING_BASE_ALIGN - 1;
- ring->vaddr_unaligned = dma_alloc_coherent(ab->dev, ring->size,
- &ring->paddr_unaligned,
- GFP_KERNEL);
+
+ if (ab->hw_params.alloc_cacheable_memory) {
+ /* Allocate the reo dst and tx completion rings from cacheable memory */
+ switch (type) {
+ case HAL_REO_DST:
+ cached = true;
+ break;
+ default:
+ cached = false;
+ }
+
+ if (cached) {
+ ring->vaddr_unaligned = kzalloc(ring->size, GFP_KERNEL);
+ ring->paddr_unaligned = virt_to_phys(ring->vaddr_unaligned);
+ }
+ }
+
+ if (!cached)
+ ring->vaddr_unaligned = dma_alloc_coherent(ab->dev, ring->size,
+ &ring->paddr_unaligned,
+ GFP_KERNEL);
+
if (!ring->vaddr_unaligned)
return -ENOMEM;
@@ -292,6 +315,11 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring,
return -EINVAL;
}
+ if (cached && ab->hw_params.alloc_cacheable_memory) {
+ params.flags |= HAL_SRNG_FLAGS_CACHED;
+ ring->cached = 1;
+ }
+
ret = ath11k_hal_srng_setup(ab, type, ring_num, mac_id, ¶ms);
if (ret < 0) {
ath11k_warn(ab, "failed to setup srng: %d ring_id %d\n",
diff --git a/drivers/net/wireless/ath/ath11k/dp.h b/drivers/net/wireless/ath/ath11k/dp.h
index ee768cc..e659148 100644
--- a/drivers/net/wireless/ath/ath11k/dp.h
+++ b/drivers/net/wireless/ath/ath11k/dp.h
@@ -64,6 +64,7 @@ struct dp_srng {
dma_addr_t paddr;
int size;
u32 ring_id;
+ u8 cached;
};
struct dp_rxdma_ring {
diff --git a/drivers/net/wireless/ath/ath11k/hal.c b/drivers/net/wireless/ath/ath11k/hal.c
index eaa0edc..f04edaf 100644
--- a/drivers/net/wireless/ath/ath11k/hal.c
+++ b/drivers/net/wireless/ath/ath11k/hal.c
@@ -627,6 +627,21 @@ u32 *ath11k_hal_srng_dst_peek(struct ath11k_base *ab, struct hal_srng *srng)
return NULL;
}
+static void ath11k_hal_srng_prefetch_desc(struct ath11k_base *ab,
+ struct hal_srng *srng)
+{
+ u32 *desc;
+
+ /* prefetch only if desc is available */
+ desc = ath11k_hal_srng_dst_peek(ab, srng);
+ if (likely(desc)) {
+ dma_sync_single_for_cpu(ab->dev, virt_to_phys(desc),
+ (srng->entry_size * sizeof(u32)),
+ DMA_FROM_DEVICE);
+ prefetch(desc);
+ }
+}
+
u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab,
struct hal_srng *srng)
{
@@ -642,6 +657,10 @@ u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab,
srng->u.dst_ring.tp = (srng->u.dst_ring.tp + srng->entry_size) %
srng->ring_size;
+ /* Try to prefetch the next descriptor in the ring */
+ if (srng->flags & HAL_SRNG_FLAGS_CACHED)
+ ath11k_hal_srng_prefetch_desc(ab, srng);
+
return desc;
}
@@ -775,11 +794,16 @@ void ath11k_hal_srng_access_begin(struct ath11k_base *ab, struct hal_srng *srng)
{
lockdep_assert_held(&srng->lock);
- if (srng->ring_dir == HAL_SRNG_DIR_SRC)
+ if (srng->ring_dir == HAL_SRNG_DIR_SRC) {
srng->u.src_ring.cached_tp =
*(volatile u32 *)srng->u.src_ring.tp_addr;
- else
+ } else {
srng->u.dst_ring.cached_hp = *srng->u.dst_ring.hp_addr;
+
+ /* Try to prefetch the next descriptor in the ring */
+ if (srng->flags & HAL_SRNG_FLAGS_CACHED)
+ ath11k_hal_srng_prefetch_desc(ab, srng);
+ }
}
/* Update cached ring head/tail pointers to HW. ath11k_hal_srng_access_begin()
diff --git a/drivers/net/wireless/ath/ath11k/hal.h b/drivers/net/wireless/ath/ath11k/hal.h
index 35ed3a1..0f4f9ce 100644
--- a/drivers/net/wireless/ath/ath11k/hal.h
+++ b/drivers/net/wireless/ath/ath11k/hal.h
@@ -513,6 +513,7 @@ enum hal_srng_dir {
#define HAL_SRNG_FLAGS_DATA_TLV_SWAP 0x00000020
#define HAL_SRNG_FLAGS_LOW_THRESH_INTR_EN 0x00010000
#define HAL_SRNG_FLAGS_MSI_INTR 0x00020000
+#define HAL_SRNG_FLAGS_CACHED 0x20000000
#define HAL_SRNG_FLAGS_LMAC_RING 0x80000000
#define HAL_SRNG_TLV_HDR_TAG GENMASK(9, 1)
diff --git a/drivers/net/wireless/ath/ath11k/hw.h b/drivers/net/wireless/ath/ath11k/hw.h
index 62f5978..7fe8edb 100644
--- a/drivers/net/wireless/ath/ath11k/hw.h
+++ b/drivers/net/wireless/ath/ath11k/hw.h
@@ -163,6 +163,7 @@ struct ath11k_hw_params {
bool supports_suspend;
u32 hal_desc_sz;
bool fix_l1ss;
+ bool alloc_cacheable_memory;
};
struct ath11k_hw_ops {
--
2.7.4
Host driver doesn't need to process CE8 interrupts (used
by target independently)
The volume of interrupts is huge within short interval,
CPU0 CPU1 CPU2 CPU3
14022188 0 0 0 GIC 71 Edge ce8
Hence disabling unused CE8 interrupt will improve CPU usage.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/ce.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath11k/ce.c b/drivers/net/wireless/ath/ath11k/ce.c
index de8b632..b6ffe03 100644
--- a/drivers/net/wireless/ath/ath11k/ce.c
+++ b/drivers/net/wireless/ath/ath11k/ce.c
@@ -77,7 +77,7 @@ const struct ce_attr ath11k_host_ce_config_ipq8074[] = {
/* CE8: target autonomous hif_memcpy */
{
- .flags = CE_ATTR_FLAGS,
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
.src_nentries = 0,
.src_sz_max = 0,
.dest_nentries = 0,
--
2.7.4
Similar to REO destination ring, also allocate HAL_WBM2SW_RELEASE
from cacheable memory so that descriptors could be prefetched during
tx completion handling.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c
index 57059c6..e4fc2ba 100644
--- a/drivers/net/wireless/ath/ath11k/dp.c
+++ b/drivers/net/wireless/ath/ath11k/dp.c
@@ -239,6 +239,7 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring,
/* Allocate the reo dst and tx completion rings from cacheable memory */
switch (type) {
case HAL_REO_DST:
+ case HAL_WBM2SW_RELEASE:
cached = true;
break;
default:
--
2.7.4
Currently while loop is used to find the tx completion ring number and
it is not required since the tx ring mask and the group id can be combined
to directly fetch the ring number. Hence remove the while loop
and directly get the ring number from tx mask and group id.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp.c | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp.c b/drivers/net/wireless/ath/ath11k/dp.c
index e4fc2ba..a39cbc7 100644
--- a/drivers/net/wireless/ath/ath11k/dp.c
+++ b/drivers/net/wireless/ath/ath11k/dp.c
@@ -770,13 +770,12 @@ int ath11k_dp_service_srng(struct ath11k_base *ab,
struct napi_struct *napi = &irq_grp->napi;
int grp_id = irq_grp->grp_id;
int work_done = 0;
- int i = 0, j;
+ int i, j;
int tot_work_done = 0;
- while (ab->hw_params.ring_mask->tx[grp_id] >> i) {
- if (ab->hw_params.ring_mask->tx[grp_id] & BIT(i))
- ath11k_dp_tx_completion_handler(ab, i);
- i++;
+ if (ab->hw_params.ring_mask->tx[grp_id]) {
+ i = __fls(ab->hw_params.ring_mask->tx[grp_id]);
+ ath11k_dp_tx_completion_handler(ab, i);
}
if (ab->hw_params.ring_mask->rx_err[grp_id]) {
--
2.7.4
Replace use of mod operator with a manual wrap around
to avoid additional cost of using mod operation.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/hal.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/hal.c b/drivers/net/wireless/ath/ath11k/hal.c
index f04edaf..7cf9e23 100644
--- a/drivers/net/wireless/ath/ath11k/hal.c
+++ b/drivers/net/wireless/ath/ath11k/hal.c
@@ -654,8 +654,11 @@ u32 *ath11k_hal_srng_dst_get_next_entry(struct ath11k_base *ab,
desc = srng->ring_base_vaddr + srng->u.dst_ring.tp;
- srng->u.dst_ring.tp = (srng->u.dst_ring.tp + srng->entry_size) %
- srng->ring_size;
+ srng->u.dst_ring.tp += srng->entry_size;
+
+ /* wrap around to start of ring*/
+ if (srng->u.dst_ring.tp == srng->ring_size)
+ srng->u.dst_ring.tp = 0;
/* Try to prefetch the next descriptor in the ring */
if (srng->flags & HAL_SRNG_FLAGS_CACHED)
--
2.7.4
Add branch prediction in dp_tx code path in
tx and tx completion handlers.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_tx.c | 45 +++++++++++++++++----------------
drivers/net/wireless/ath/ath11k/mac.c | 2 +-
2 files changed, 24 insertions(+), 23 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c
index 8bba523..ab9ccf0 100644
--- a/drivers/net/wireless/ath/ath11k/dp_tx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_tx.c
@@ -95,11 +95,11 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
u8 ring_selector = 0, ring_map = 0;
bool tcl_ring_retry;
- if (test_bit(ATH11K_FLAG_CRASH_FLUSH, &ar->ab->dev_flags))
+ if (unlikely(test_bit(ATH11K_FLAG_CRASH_FLUSH, &ar->ab->dev_flags)))
return -ESHUTDOWN;
- if (!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
- !ieee80211_is_data(hdr->frame_control))
+ if (unlikely(!(info->flags & IEEE80211_TX_CTL_HW_80211_ENCAP) &&
+ !ieee80211_is_data(hdr->frame_control)))
return -ENOTSUPP;
pool_id = skb_get_queue_mapping(skb) & (ATH11K_HW_MAX_QUEUES - 1);
@@ -130,7 +130,7 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
DP_TX_IDR_SIZE - 1, GFP_ATOMIC);
spin_unlock_bh(&tx_ring->tx_idr_lock);
- if (ret < 0) {
+ if (unlikely(ret < 0)) {
if (ring_map == (BIT(DP_TCL_NUM_RING_MAX) - 1)) {
atomic_inc(&ab->soc_stats.tx_err.misc_fail);
return -ENOSPC;
@@ -147,7 +147,7 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
ti.encap_type = ath11k_dp_tx_get_encap_type(arvif, skb);
ti.meta_data_flags = arvif->tcl_metadata;
- if (ti.encap_type == HAL_TCL_ENCAP_TYPE_RAW) {
+ if (unlikely(ti.encap_type == HAL_TCL_ENCAP_TYPE_RAW)) {
if (skb_cb->flags & ATH11K_SKB_CIPHER_SET) {
ti.encrypt_type =
ath11k_dp_tx_get_encrypt_type(skb_cb->cipher);
@@ -168,8 +168,8 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
ti.bss_ast_idx = arvif->ast_idx;
ti.dscp_tid_tbl_idx = 0;
- if (skb->ip_summed == CHECKSUM_PARTIAL &&
- ti.encap_type != HAL_TCL_ENCAP_TYPE_RAW) {
+ if (likely(skb->ip_summed == CHECKSUM_PARTIAL &&
+ ti.encap_type != HAL_TCL_ENCAP_TYPE_RAW)) {
ti.flags0 |= FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_IP4_CKSUM_EN, 1) |
FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_UDP4_CKSUM_EN, 1) |
FIELD_PREP(HAL_TCL_DATA_CMD_INFO1_UDP6_CKSUM_EN, 1) |
@@ -206,7 +206,7 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
}
ti.paddr = dma_map_single(ab->dev, skb->data, skb->len, DMA_TO_DEVICE);
- if (dma_mapping_error(ab->dev, ti.paddr)) {
+ if (unlikely(dma_mapping_error(ab->dev, ti.paddr))) {
atomic_inc(&ab->soc_stats.tx_err.misc_fail);
ath11k_warn(ab, "failed to DMA map data Tx buffer\n");
ret = -ENOMEM;
@@ -226,7 +226,7 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
ath11k_hal_srng_access_begin(ab, tcl_ring);
hal_tcl_desc = (void *)ath11k_hal_srng_src_get_next_entry(ab, tcl_ring);
- if (!hal_tcl_desc) {
+ if (unlikely(!hal_tcl_desc)) {
/* NOTE: It is highly unlikely we'll be running out of tcl_ring
* desc because the desc is directly enqueued onto hw queue.
*/
@@ -240,8 +240,8 @@ int ath11k_dp_tx(struct ath11k *ar, struct ath11k_vif *arvif,
* checking this ring earlier for each pkt tx.
* Restart ring selection if some rings are not checked yet.
*/
- if (ring_map != (BIT(DP_TCL_NUM_RING_MAX) - 1) &&
- !ar->ab->hw_params.tcl_0_only) {
+ if (unlikely(ring_map != (BIT(DP_TCL_NUM_RING_MAX) - 1) &&
+ !ar->ab->hw_params.tcl_0_only)) {
tcl_ring_retry = true;
ring_selector++;
}
@@ -322,7 +322,7 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab,
spin_lock_bh(&tx_ring->tx_idr_lock);
msdu = idr_find(&tx_ring->txbuf_idr, ts->msdu_id);
- if (!msdu) {
+ if (unlikely(!msdu)) {
ath11k_warn(ab, "htt tx completion for unknown msdu_id %d\n",
ts->msdu_id);
spin_unlock_bh(&tx_ring->tx_idr_lock);
@@ -432,12 +432,12 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar,
rcu_read_lock();
- if (!rcu_dereference(ab->pdevs_active[ar->pdev_idx])) {
+ if (unlikely(!rcu_dereference(ab->pdevs_active[ar->pdev_idx]))) {
dev_kfree_skb_any(msdu);
goto exit;
}
- if (!skb_cb->vif) {
+ if (unlikely(!skb_cb->vif)) {
dev_kfree_skb_any(msdu);
goto exit;
}
@@ -460,7 +460,7 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar,
(info->flags & IEEE80211_TX_CTL_NO_ACK))
info->flags |= IEEE80211_TX_STAT_NOACK_TRANSMITTED;
- if (ath11k_debugfs_is_extd_tx_stats_enabled(ar)) {
+ if (unlikely(ath11k_debugfs_is_extd_tx_stats_enabled(ar))) {
if (ts->flags & HAL_TX_STATUS_FLAGS_FIRST_MSDU) {
if (ar->last_ppdu_id == 0) {
ar->last_ppdu_id = ts->ppdu_id;
@@ -500,11 +500,11 @@ static inline void ath11k_dp_tx_status_parse(struct ath11k_base *ab,
{
ts->buf_rel_source =
FIELD_GET(HAL_WBM_RELEASE_INFO0_REL_SRC_MODULE, desc->info0);
- if (ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_FW &&
- ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_TQM)
+ if (unlikely(ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_FW &&
+ ts->buf_rel_source != HAL_WBM_REL_SRC_MODULE_TQM))
return;
- if (ts->buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW)
+ if (unlikely(ts->buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW))
return;
ts->status = FIELD_GET(HAL_WBM_RELEASE_INFO0_TQM_RELEASE_REASON,
@@ -551,8 +551,9 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head);
}
- if ((ath11k_hal_srng_dst_peek(ab, status_ring) != NULL) &&
- (ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head) == tx_ring->tx_status_tail)) {
+ if (unlikely((ath11k_hal_srng_dst_peek(ab, status_ring) != NULL) &&
+ (ATH11K_TX_COMPL_NEXT(tx_ring->tx_status_head) ==
+ tx_ring->tx_status_tail))) {
/* TODO: Process pending tx_status messages when kfifo_is_full() */
ath11k_warn(ab, "Unable to process some of the tx_status ring desc because status_fifo is full\n");
}
@@ -575,7 +576,7 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
mac_id = FIELD_GET(DP_TX_DESC_ID_MAC_ID, desc_id);
msdu_id = FIELD_GET(DP_TX_DESC_ID_MSDU_ID, desc_id);
- if (ts.buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW) {
+ if (unlikely(ts.buf_rel_source == HAL_WBM_REL_SRC_MODULE_FW)) {
ath11k_dp_tx_process_htt_tx_complete(ab,
(void *)tx_status,
mac_id, msdu_id,
@@ -585,7 +586,7 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
spin_lock_bh(&tx_ring->tx_idr_lock);
msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
- if (!msdu) {
+ if (unlikely(!msdu)) {
ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
msdu_id);
spin_unlock_bh(&tx_ring->tx_idr_lock);
diff --git a/drivers/net/wireless/ath/ath11k/mac.c b/drivers/net/wireless/ath/ath11k/mac.c
index e9b3689..7c4bf51 100644
--- a/drivers/net/wireless/ath/ath11k/mac.c
+++ b/drivers/net/wireless/ath/ath11k/mac.c
@@ -4339,7 +4339,7 @@ static void ath11k_mac_op_tx(struct ieee80211_hw *hw,
}
ret = ath11k_dp_tx(ar, arvif, skb);
- if (ret) {
+ if (unlikely(ret)) {
ath11k_warn(ar->ab, "failed to transmit frame %d\n", ret);
ieee80211_free_txskb(ar->hw, skb);
}
--
2.7.4
Avoid unnecessary idr_find calls before the idr_remove calls.
Because idr_remove gives the valid ptr if id is valid otherwise
return NULL ptr. So removed the idr_find before idr_remove in tx
completion path. Also no need to disable the bottom half preempt
if it is already in the bottom half context, so modify the
spin_lock_bh to spin_lock in the data tx completion path.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
Co-developed-by: Karthikeyan Periyasamy <[email protected]>
Signed-off-by: Karthikeyan Periyasamy <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_tx.c | 31 +++++++++++++------------------
1 file changed, 13 insertions(+), 18 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c
index ab9ccf0..3e0efad 100644
--- a/drivers/net/wireless/ath/ath11k/dp_tx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_tx.c
@@ -288,20 +288,18 @@ static void ath11k_dp_tx_free_txbuf(struct ath11k_base *ab, u8 mac_id,
struct sk_buff *msdu;
struct ath11k_skb_cb *skb_cb;
- spin_lock_bh(&tx_ring->tx_idr_lock);
- msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
- if (!msdu) {
+ spin_lock(&tx_ring->tx_idr_lock);
+ msdu = idr_remove(&tx_ring->txbuf_idr, msdu_id);
+ spin_unlock(&tx_ring->tx_idr_lock);
+
+ if (unlikely(!msdu)) {
ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
return;
}
skb_cb = ATH11K_SKB_CB(msdu);
- idr_remove(&tx_ring->txbuf_idr, msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
-
dma_unmap_single(ab->dev, skb_cb->paddr, msdu->len, DMA_TO_DEVICE);
dev_kfree_skb_any(msdu);
@@ -320,12 +318,13 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab,
struct ath11k_skb_cb *skb_cb;
struct ath11k *ar;
- spin_lock_bh(&tx_ring->tx_idr_lock);
- msdu = idr_find(&tx_ring->txbuf_idr, ts->msdu_id);
+ spin_lock(&tx_ring->tx_idr_lock);
+ msdu = idr_remove(&tx_ring->txbuf_idr, ts->msdu_id);
+ spin_unlock(&tx_ring->tx_idr_lock);
+
if (unlikely(!msdu)) {
ath11k_warn(ab, "htt tx completion for unknown msdu_id %d\n",
ts->msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
return;
}
@@ -334,9 +333,6 @@ ath11k_dp_tx_htt_tx_complete_buf(struct ath11k_base *ab,
ar = skb_cb->ar;
- idr_remove(&tx_ring->txbuf_idr, ts->msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
-
if (atomic_dec_and_test(&ar->dp.num_tx_pending))
wake_up(&ar->dp.tx_empty_waitq);
@@ -584,16 +580,15 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
continue;
}
- spin_lock_bh(&tx_ring->tx_idr_lock);
- msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
+ msdu = idr_remove(&tx_ring->txbuf_idr, msdu_id);
if (unlikely(!msdu)) {
ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
+ spin_unlock(&tx_ring->tx_idr_lock);
continue;
}
- idr_remove(&tx_ring->txbuf_idr, msdu_id);
- spin_unlock_bh(&tx_ring->tx_idr_lock);
+
+ spin_unlock(&tx_ring->tx_idr_lock);
ar = ab->pdevs[mac_id].ar;
--
2.7.4
In datapath, add branch predictors where required in the process rx().
This protects high value rx path without having performance overhead.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_rx.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
index e105bdc..5d805881 100644
--- a/drivers/net/wireless/ath/ath11k/dp_rx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
@@ -2535,13 +2535,13 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
rcu_read_lock();
ar = ab->pdevs[mac_id].ar;
- if (!rcu_dereference(ab->pdevs_active[mac_id])) {
+ if (unlikely(!rcu_dereference(ab->pdevs_active[mac_id]))) {
__skb_queue_purge(msdu_list);
rcu_read_unlock();
return;
}
- if (test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags)) {
+ if (unlikely(test_bit(ATH11K_CAC_RUNNING, &ar->dev_flags))) {
__skb_queue_purge(msdu_list);
rcu_read_unlock();
return;
@@ -2549,7 +2549,7 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
while ((msdu = __skb_dequeue(msdu_list))) {
ret = ath11k_dp_rx_process_msdu(ar, msdu, msdu_list);
- if (ret) {
+ if (unlikely(ret)) {
ath11k_dbg(ab, ATH11K_DBG_DATA,
"Unable to process msdu %d", ret);
dev_kfree_skb_any(msdu);
@@ -2604,7 +2604,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
rx_ring = &ar->dp.rx_refill_buf_ring;
spin_lock_bh(&rx_ring->idr_lock);
msdu = idr_find(&rx_ring->bufs_idr, buf_id);
- if (!msdu) {
+ if (unlikely(!msdu)) {
ath11k_warn(ab, "frame rx with invalid buf_id %d\n",
buf_id);
spin_unlock_bh(&rx_ring->idr_lock);
@@ -2623,8 +2623,8 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
push_reason = FIELD_GET(HAL_REO_DEST_RING_INFO0_PUSH_REASON,
desc->info0);
- if (push_reason !=
- HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION) {
+ if (unlikely(push_reason !=
+ HAL_REO_DEST_RING_PUSH_REASON_ROUTING_INSTRUCTION)) {
dev_kfree_skb_any(msdu);
ab->soc_stats.hal_reo_error[dp->reo_dst_ring[ring_id].ring_id]++;
continue;
@@ -2659,7 +2659,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
* head pointer so that we can reap complete MPDU in the current
* rx processing.
*/
- if (!done && ath11k_hal_srng_dst_num_free(ab, srng, true)) {
+ if (unlikely(!done && ath11k_hal_srng_dst_num_free(ab, srng, true))) {
ath11k_hal_srng_access_end(ab, srng);
goto try_again;
}
@@ -2668,7 +2668,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
spin_unlock_bh(&srng->lock);
- if (!total_msdu_reaped)
+ if (unlikely(!total_msdu_reaped))
goto exit;
for (i = 0; i < ab->num_radios; i++) {
--
2.7.4
The usage of quota variable inside
ath11k_dp_rx_process_received_packets() is redundant.
Since we would queue only max packets to the list before
calling this function so it would never exceed quota. Hence Removing
usage of quota variable.
Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
Co-developed-by: Sriram R <[email protected]>
Signed-off-by: Sriram R <[email protected]>
Signed-off-by: Jouni Malinen <[email protected]>
Signed-off-by: P Praneesh <[email protected]>
---
drivers/net/wireless/ath/ath11k/dp_rx.c | 9 +++------
1 file changed, 3 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
index 1d85e10..e105bdc 100644
--- a/drivers/net/wireless/ath/ath11k/dp_rx.c
+++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
@@ -2523,7 +2523,7 @@ static int ath11k_dp_rx_process_msdu(struct ath11k *ar,
static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
struct napi_struct *napi,
struct sk_buff_head *msdu_list,
- int *quota, int mac_id)
+ int mac_id)
{
struct sk_buff *msdu;
struct ath11k *ar;
@@ -2557,7 +2557,6 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
}
ath11k_dp_rx_deliver_msdu(ar, napi, msdu);
- (*quota)--;
}
rcu_read_unlock();
@@ -2574,7 +2573,6 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
int total_msdu_reaped = 0;
struct hal_srng *srng;
struct sk_buff *msdu;
- int quota = budget;
bool done = false;
int buf_id, mac_id;
struct ath11k *ar;
@@ -2677,8 +2675,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
if (!num_buffs_reaped[i])
continue;
- ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i],
- "a, i);
+ ath11k_dp_rx_process_received_packets(ab, napi, &msdu_list[i], i);
ar = ab->pdevs[i].ar;
rx_ring = &ar->dp.rx_refill_buf_ring;
@@ -2687,7 +2684,7 @@ int ath11k_dp_process_rx(struct ath11k_base *ab, int ring_id,
HAL_RX_BUF_RBM_SW3_BM);
}
exit:
- return budget - quota;
+ return total_msdu_reaped;
}
static void ath11k_dp_rx_update_peer_stats(struct ath11k_sta *arsta,
--
2.7.4
On 8/25/21 2:37 AM, P Praneesh wrote:
> @@ -292,6 +315,11 @@ int ath11k_dp_srng_setup(struct ath11k_base *ab, struct dp_srng *ring,
> return -EINVAL;
> }
>
> + if (cached && ab->hw_params.alloc_cacheable_memory) {
> + params.flags |= HAL_SRNG_FLAGS_CACHED;
> + ring->cached = 1;
> + }
"&& ab->hw_params.alloc_cacheable_memory" is redundant check.
Thanks,
Peter
On 8/25/21 2:37 AM, P Praneesh wrote:
> @@ -584,16 +580,15 @@ void ath11k_dp_tx_completion_handler(struct ath11k_base *ab, int ring_id)
> continue;
> }
>
> - spin_lock_bh(&tx_ring->tx_idr_lock);
> - msdu = idr_find(&tx_ring->txbuf_idr, msdu_id);
missing spin_lock(&tx_ring->tx_idr_lock);here?
> + msdu = idr_remove(&tx_ring->txbuf_idr, msdu_id);
> if (unlikely(!msdu)) {
> ath11k_warn(ab, "tx completion for unknown msdu_id %d\n",
> msdu_id);
> - spin_unlock_bh(&tx_ring->tx_idr_lock);
> + spin_unlock(&tx_ring->tx_idr_lock);
> continue;
> }
> - idr_remove(&tx_ring->txbuf_idr, msdu_id);
> - spin_unlock_bh(&tx_ring->tx_idr_lock);
> +
> + spin_unlock(&tx_ring->tx_idr_lock);
>
> ar = ab->pdevs[mac_id].ar;
>
On Wed, Aug 25, 2021 at 12:39 PM P Praneesh <[email protected]> wrote:
> In datapath, add branch predictors where required in the process rx().
> This protects high value rx path without having performance overhead.
>
> Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1.r2-00012-QCAHKSWPL_SILICONZ-1
> Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01695-QCAHKSWPL_SILICONZ-1
>
> Co-developed-by: Sriram R <[email protected]>
> Signed-off-by: Sriram R <[email protected]>
> Signed-off-by: Jouni Malinen <[email protected]>
> Signed-off-by: P Praneesh <[email protected]>
> ---
> drivers/net/wireless/ath/ath11k/dp_rx.c | 16 ++++++++--------
> 1 file changed, 8 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath11k/dp_rx.c b/drivers/net/wireless/ath/ath11k/dp_rx.c
> index e105bdc..5d805881 100644
> --- a/drivers/net/wireless/ath/ath11k/dp_rx.c
> +++ b/drivers/net/wireless/ath/ath11k/dp_rx.c
> @@ -2535,13 +2535,13 @@ static void ath11k_dp_rx_process_received_packets(struct ath11k_base *ab,
> rcu_read_lock();
>
> ar = ab->pdevs[mac_id].ar;
> - if (!rcu_dereference(ab->pdevs_active[mac_id])) {
> + if (unlikely(!rcu_dereference(ab->pdevs_active[mac_id]))) {
> __skb_queue_purge(msdu_list);
> rcu_read_unlock();
> return;
> }
The pointer that is returned by the rcu_dereference() is not
dereferenced, so it is preferable to use rcu_access_pointer() here.
And since rcu_access_pointer() did not require the RCU protection, the
whole block with check and list purging could be moved before the
rcu_read_lock() call.
--
Sergey
On Wed, Aug 25, 2021 at 12:39 PM P Praneesh <[email protected]> wrote:
> Add branch prediction in dp_tx code path in
> tx and tx completion handlers.
>
> Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.4.0.1-01734-QCAHKSWPL_SILICONZ-1 v2
>
> Co-developed-by: Sriram R <[email protected]>
> Signed-off-by: Sriram R <[email protected]>
> Signed-off-by: Jouni Malinen <[email protected]>
> Signed-off-by: P Praneesh <[email protected]>
> ---
> drivers/net/wireless/ath/ath11k/dp_tx.c | 45 +++++++++++++++++----------------
> drivers/net/wireless/ath/ath11k/mac.c | 2 +-
> 2 files changed, 24 insertions(+), 23 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath11k/dp_tx.c b/drivers/net/wireless/ath/ath11k/dp_tx.c
> index 8bba523..ab9ccf0 100644
> --- a/drivers/net/wireless/ath/ath11k/dp_tx.c
> +++ b/drivers/net/wireless/ath/ath11k/dp_tx.c
[skipped]
> @@ -432,12 +432,12 @@ static void ath11k_dp_tx_complete_msdu(struct ath11k *ar,
>
> rcu_read_lock();
>
> - if (!rcu_dereference(ab->pdevs_active[ar->pdev_idx])) {
> + if (unlikely(!rcu_dereference(ab->pdevs_active[ar->pdev_idx]))) {
> dev_kfree_skb_any(msdu);
> goto exit;
> }
This is another place where rcu_dereference() could be replaced by the
rcu_access_pointer() and the whole block could be moved above the
rcu_read_lock().
On the other hand this is an _unlikely_ path, so it looks like a
matter of taste which macro to use here.
--
Sergey