Lines Matching full:polling
16 interrupts first (:ref:`busy polling<poll>`).
121 the NAPI instance - until NAPI polling finishes any further
155 mapped to queues and interrupts. NAPI is primarily a polling/processing
224 Busy polling
227 Busy polling allows a user process to check for incoming packets before
228 the device interrupt fires. As is the case with any busy polling it trades
229 off CPU cycles for lower latency (production uses of NAPI busy polling
232 Busy polling is enabled by either setting ``SO_BUSY_POLL`` on
234 ``net.core.busy_read`` sysctls. An io_uring API for NAPI busy polling
237 epoll-based busy polling
255 In order to enable busy polling, there are two choices:
280 While busy polling is supposed to be used by low latency applications,
289 polling operation periodically, and the driver should keep the device IRQs
293 busy polling applications, the ``prefer_busy_poll`` field of ``struct
297 The NAPI budget for busy polling is lower than the default (which makes
298 sense given the low latency intention of normal busy polling). This is
300 with the ``SO_BUSY_POLL_BUDGET`` socket option. For epoll-based busy polling
311 polling applications may be able to mitigate how much user processing happens
325 while busy polling (for example, because network traffic levels subsided), IRQ
369 IRQ suspension causes the system to alternate between polling mode and
371 overrides ``gro_flush_timeout`` and keeps the system busy polling, but when