aio.h (a098fbc025095a0d98156ada7eac59de97c16499) | aio.h (54a16a63d021b2e28f08082abb20b9431d3c97d3) |
---|---|
1/* 2 * QEMU aio implementation 3 * 4 * Copyright IBM, Corp. 2008 5 * 6 * Authors: 7 * Anthony Liguori <aliguori@us.ibm.com> 8 * --- 60 unchanged lines hidden (view full) --- 69 * accessed with atomic primitives. If this field is 0, everything 70 * (file descriptors, bottom halves, timers) will be re-evaluated 71 * before the next blocking poll(), thus the event_notifier_set call 72 * can be skipped. If it is non-zero, you may need to wake up a 73 * concurrent aio_poll or the glib main event loop, making 74 * event_notifier_set necessary. 75 * 76 * Bit 0 is reserved for GSource usage of the AioContext, and is 1 | 1/* 2 * QEMU aio implementation 3 * 4 * Copyright IBM, Corp. 2008 5 * 6 * Authors: 7 * Anthony Liguori <aliguori@us.ibm.com> 8 * --- 60 unchanged lines hidden (view full) --- 69 * accessed with atomic primitives. If this field is 0, everything 70 * (file descriptors, bottom halves, timers) will be re-evaluated 71 * before the next blocking poll(), thus the event_notifier_set call 72 * can be skipped. If it is non-zero, you may need to wake up a 73 * concurrent aio_poll or the glib main event loop, making 74 * event_notifier_set necessary. 75 * 76 * Bit 0 is reserved for GSource usage of the AioContext, and is 1 |
77 * between a call to aio_ctx_check and the next call to aio_ctx_dispatch. | 77 * between a call to aio_ctx_prepare and the next call to aio_ctx_check. |
78 * Bits 1-31 simply count the number of active calls to aio_poll 79 * that are in the prepare or poll phase. 80 * 81 * The GSource and aio_poll must use a different mechanism because 82 * there is no certainty that a call to GSource's prepare callback 83 * (via g_main_context_prepare) is indeed followed by check and 84 * dispatch. It's not clear whether this would be a bug, but let's 85 * play safe and allow it---it will just cause extra calls to --- 372 unchanged lines hidden --- | 78 * Bits 1-31 simply count the number of active calls to aio_poll 79 * that are in the prepare or poll phase. 80 * 81 * The GSource and aio_poll must use a different mechanism because 82 * there is no certainty that a call to GSource's prepare callback 83 * (via g_main_context_prepare) is indeed followed by check and 84 * dispatch. It's not clear whether this would be a bug, but let's 85 * play safe and allow it---it will just cause extra calls to --- 372 unchanged lines hidden --- |