X-Git-Url: https://gerrit.opnfv.org/gerrit/gitweb?a=blobdiff_plain;f=qemu%2Faio-posix.c;h=6006122e0b9a3841d901cb856bc45529364e5995;hb=a14b48d18a9ed03ec191cf16b162206998a895ce;hp=d4770336c5d5355758c3da207b8be5160f66fc5f;hpb=e44e3482bdb4d0ebde2d8b41830ac2cdb07948fb;p=kvmfornfv.git diff --git a/qemu/aio-posix.c b/qemu/aio-posix.c index d4770336c..6006122e0 100644 --- a/qemu/aio-posix.c +++ b/qemu/aio-posix.c @@ -13,10 +13,14 @@ * GNU GPL, version 2 or (at your option) any later version. */ +#include "qemu/osdep.h" #include "qemu-common.h" #include "block/block.h" #include "qemu/queue.h" #include "qemu/sockets.h" +#ifdef CONFIG_EPOLL_CREATE1 +#include +#endif struct AioHandler { @@ -25,9 +29,166 @@ struct AioHandler IOHandler *io_write; int deleted; void *opaque; + bool is_external; QLIST_ENTRY(AioHandler) node; }; +#ifdef CONFIG_EPOLL_CREATE1 + +/* The fd number threashold to switch to epoll */ +#define EPOLL_ENABLE_THRESHOLD 64 + +static void aio_epoll_disable(AioContext *ctx) +{ + ctx->epoll_available = false; + if (!ctx->epoll_enabled) { + return; + } + ctx->epoll_enabled = false; + close(ctx->epollfd); +} + +static inline int epoll_events_from_pfd(int pfd_events) +{ + return (pfd_events & G_IO_IN ? EPOLLIN : 0) | + (pfd_events & G_IO_OUT ? EPOLLOUT : 0) | + (pfd_events & G_IO_HUP ? EPOLLHUP : 0) | + (pfd_events & G_IO_ERR ? EPOLLERR : 0); +} + +static bool aio_epoll_try_enable(AioContext *ctx) +{ + AioHandler *node; + struct epoll_event event; + + QLIST_FOREACH(node, &ctx->aio_handlers, node) { + int r; + if (node->deleted || !node->pfd.events) { + continue; + } + event.events = epoll_events_from_pfd(node->pfd.events); + event.data.ptr = node; + r = epoll_ctl(ctx->epollfd, EPOLL_CTL_ADD, node->pfd.fd, &event); + if (r) { + return false; + } + } + ctx->epoll_enabled = true; + return true; +} + +static void aio_epoll_update(AioContext *ctx, AioHandler *node, bool is_new) +{ + struct epoll_event event; + int r; + + if (!ctx->epoll_enabled) { + return; + } + if (!node->pfd.events) { + r = epoll_ctl(ctx->epollfd, EPOLL_CTL_DEL, node->pfd.fd, &event); + if (r) { + aio_epoll_disable(ctx); + } + } else { + event.data.ptr = node; + event.events = epoll_events_from_pfd(node->pfd.events); + if (is_new) { + r = epoll_ctl(ctx->epollfd, EPOLL_CTL_ADD, node->pfd.fd, &event); + if (r) { + aio_epoll_disable(ctx); + } + } else { + r = epoll_ctl(ctx->epollfd, EPOLL_CTL_MOD, node->pfd.fd, &event); + if (r) { + aio_epoll_disable(ctx); + } + } + } +} + +static int aio_epoll(AioContext *ctx, GPollFD *pfds, + unsigned npfd, int64_t timeout) +{ + AioHandler *node; + int i, ret = 0; + struct epoll_event events[128]; + + assert(npfd == 1); + assert(pfds[0].fd == ctx->epollfd); + if (timeout > 0) { + ret = qemu_poll_ns(pfds, npfd, timeout); + } + if (timeout <= 0 || ret > 0) { + ret = epoll_wait(ctx->epollfd, events, + sizeof(events) / sizeof(events[0]), + timeout); + if (ret <= 0) { + goto out; + } + for (i = 0; i < ret; i++) { + int ev = events[i].events; + node = events[i].data.ptr; + node->pfd.revents = (ev & EPOLLIN ? G_IO_IN : 0) | + (ev & EPOLLOUT ? G_IO_OUT : 0) | + (ev & EPOLLHUP ? G_IO_HUP : 0) | + (ev & EPOLLERR ? G_IO_ERR : 0); + } + } +out: + return ret; +} + +static bool aio_epoll_enabled(AioContext *ctx) +{ + /* Fall back to ppoll when external clients are disabled. */ + return !aio_external_disabled(ctx) && ctx->epoll_enabled; +} + +static bool aio_epoll_check_poll(AioContext *ctx, GPollFD *pfds, + unsigned npfd, int64_t timeout) +{ + if (!ctx->epoll_available) { + return false; + } + if (aio_epoll_enabled(ctx)) { + return true; + } + if (npfd >= EPOLL_ENABLE_THRESHOLD) { + if (aio_epoll_try_enable(ctx)) { + return true; + } else { + aio_epoll_disable(ctx); + } + } + return false; +} + +#else + +static void aio_epoll_update(AioContext *ctx, AioHandler *node, bool is_new) +{ +} + +static int aio_epoll(AioContext *ctx, GPollFD *pfds, + unsigned npfd, int64_t timeout) +{ + assert(false); +} + +static bool aio_epoll_enabled(AioContext *ctx) +{ + return false; +} + +static bool aio_epoll_check_poll(AioContext *ctx, GPollFD *pfds, + unsigned npfd, int64_t timeout) +{ + return false; +} + +#endif + static AioHandler *find_aio_handler(AioContext *ctx, int fd) { AioHandler *node; @@ -43,11 +204,14 @@ static AioHandler *find_aio_handler(AioContext *ctx, int fd) void aio_set_fd_handler(AioContext *ctx, int fd, + bool is_external, IOHandler *io_read, IOHandler *io_write, void *opaque) { AioHandler *node; + bool is_new = false; + bool deleted = false; node = find_aio_handler(ctx, fd); @@ -66,7 +230,7 @@ void aio_set_fd_handler(AioContext *ctx, * releasing the walking_handlers lock. */ QLIST_REMOVE(node, node); - g_free(node); + deleted = true; } } } else { @@ -77,25 +241,32 @@ void aio_set_fd_handler(AioContext *ctx, QLIST_INSERT_HEAD(&ctx->aio_handlers, node, node); g_source_add_poll(&ctx->source, &node->pfd); + is_new = true; } /* Update handler with latest information */ node->io_read = io_read; node->io_write = io_write; node->opaque = opaque; + node->is_external = is_external; node->pfd.events = (io_read ? G_IO_IN | G_IO_HUP | G_IO_ERR : 0); node->pfd.events |= (io_write ? G_IO_OUT | G_IO_ERR : 0); } + aio_epoll_update(ctx, node, is_new); aio_notify(ctx); + if (deleted) { + g_free(node); + } } void aio_set_event_notifier(AioContext *ctx, EventNotifier *notifier, + bool is_external, EventNotifierHandler *io_read) { aio_set_fd_handler(ctx, event_notifier_get_fd(notifier), - (IOHandler *)io_read, NULL, notifier); + is_external, (IOHandler *)io_read, NULL, notifier); } bool aio_prepare(AioContext *ctx) @@ -111,10 +282,12 @@ bool aio_pending(AioContext *ctx) int revents; revents = node->pfd.revents & node->pfd.events; - if (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR) && node->io_read) { + if (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR) && node->io_read && + aio_node_check(ctx, node->is_external)) { return true; } - if (revents & (G_IO_OUT | G_IO_ERR) && node->io_write) { + if (revents & (G_IO_OUT | G_IO_ERR) && node->io_write && + aio_node_check(ctx, node->is_external)) { return true; } } @@ -152,6 +325,7 @@ bool aio_dispatch(AioContext *ctx) if (!node->deleted && (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) && + aio_node_check(ctx, node->is_external) && node->io_read) { node->io_read(node->opaque); @@ -162,6 +336,7 @@ bool aio_dispatch(AioContext *ctx) } if (!node->deleted && (revents & (G_IO_OUT | G_IO_ERR)) && + aio_node_check(ctx, node->is_external) && node->io_write) { node->io_write(node->opaque); progress = true; @@ -257,7 +432,9 @@ bool aio_poll(AioContext *ctx, bool blocking) /* fill pollfds */ QLIST_FOREACH(node, &ctx->aio_handlers, node) { - if (!node->deleted && node->pfd.events) { + if (!node->deleted && node->pfd.events + && !aio_epoll_enabled(ctx) + && aio_node_check(ctx, node->is_external)) { add_pollfd(node); } } @@ -268,7 +445,17 @@ bool aio_poll(AioContext *ctx, bool blocking) if (timeout) { aio_context_release(ctx); } - ret = qemu_poll_ns((GPollFD *)pollfds, npfd, timeout); + if (aio_epoll_check_poll(ctx, pollfds, npfd, timeout)) { + AioHandler epoll_handler; + + epoll_handler.pfd.fd = ctx->epollfd; + epoll_handler.pfd.events = G_IO_IN | G_IO_OUT | G_IO_HUP | G_IO_ERR; + npfd = 0; + add_pollfd(&epoll_handler); + ret = aio_epoll(ctx, pollfds, npfd, timeout); + } else { + ret = qemu_poll_ns(pollfds, npfd, timeout); + } if (blocking) { atomic_sub(&ctx->notify_me, 2); } @@ -297,3 +484,16 @@ bool aio_poll(AioContext *ctx, bool blocking) return progress; } + +void aio_context_setup(AioContext *ctx, Error **errp) +{ +#ifdef CONFIG_EPOLL_CREATE1 + assert(!ctx->epollfd); + ctx->epollfd = epoll_create1(EPOLL_CLOEXEC); + if (ctx->epollfd == -1) { + ctx->epoll_available = false; + } else { + ctx->epoll_available = true; + } +#endif +}