ath10k: implement NAPI support

Add NAPI support for rx and tx completion. NAPI poll is scheduled
from interrupt handler. The design is as below

 - on interrupt
     - schedule napi and mask interrupts
 - on poll
   - process all pipes (no actual Tx/Rx)
   - process Rx within budget
   - if quota exceeds budget reschedule napi poll by returning budget
   - process Tx completions and update budget if necessary
   - process Tx fetch indications (pull-push)
   - push any other pending Tx (if possible)
   - before resched or napi completion replenish htt rx ring buffer
   - if work done < budget, complete napi poll and unmask interrupts

This change also get rid of two tasklets (intr_tq and txrx_compl_task).

Measured peak throughput with NAPI on IPQ4019 platform in controlled
environment. No noticeable reduction in throughput is seen and also
observed improvements in CPU usage. Approx. 15% CPU usage got reduced
in UDP uplink case.

DL: AP DUT Tx
UL: AP DUT Rx

IPQ4019 (avg. cpu usage %)

========
                TOT              +NAPI
              ===========      =============
TCP DL       644 Mbps (42%)    645 Mbps (36%)
TCP UL       673 Mbps (30%)    675 Mbps (26%)
UDP DL       682 Mbps (49%)    680 Mbps (49%)
UDP UL       720 Mbps (28%)    717 Mbps (11%)

Signed-off-by: Rajkumar Manoharan <rmanohar@qti.qualcomm.com>
Signed-off-by: Kalle Valo <kvalo@qca.qualcomm.com>
This commit is contained in:
Rajkumar Manoharan
2016-09-02 19:46:09 +03:00
committed by Kalle Valo
parent c39265f72a
commit 3c97f5de1f
8 changed files with 159 additions and 96 deletions

View File

@@ -177,8 +177,6 @@ struct ath10k_pci {
/* Operating interrupt mode */
enum ath10k_pci_irq_mode oper_irq_mode;
struct tasklet_struct intr_tq;
struct ath10k_pci_pipe pipe_info[CE_COUNT_MAX];
/* Copy Engine used for Diagnostic Accesses */
@@ -294,8 +292,7 @@ void ath10k_pci_free_pipes(struct ath10k *ar);
void ath10k_pci_free_pipes(struct ath10k *ar);
void ath10k_pci_rx_replenish_retry(unsigned long ptr);
void ath10k_pci_ce_deinit(struct ath10k *ar);
void ath10k_pci_init_irq_tasklets(struct ath10k *ar);
void ath10k_pci_kill_tasklet(struct ath10k *ar);
void ath10k_pci_init_napi(struct ath10k *ar);
int ath10k_pci_init_pipes(struct ath10k *ar);
int ath10k_pci_init_config(struct ath10k *ar);
void ath10k_pci_rx_post(struct ath10k *ar);
@@ -303,6 +300,7 @@ void ath10k_pci_flush(struct ath10k *ar);
void ath10k_pci_enable_legacy_irq(struct ath10k *ar);
bool ath10k_pci_irq_pending(struct ath10k *ar);
void ath10k_pci_disable_and_clear_legacy_irq(struct ath10k *ar);
void ath10k_pci_irq_msi_fw_mask(struct ath10k *ar);
int ath10k_pci_wait_for_target_init(struct ath10k *ar);
int ath10k_pci_setup_resource(struct ath10k *ar);
void ath10k_pci_release_resource(struct ath10k *ar);