From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S966047AbXDBWrK (ORCPT ); Mon, 2 Apr 2007 18:47:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S966033AbXDBWrI (ORCPT ); Mon, 2 Apr 2007 18:47:08 -0400 Received: from x35.xmailserver.org ([64.71.152.41]:50588 "EHLO x35.xmailserver.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S966052AbXDBWqt (ORCPT ); Mon, 2 Apr 2007 18:46:49 -0400 X-AuthUser: davidel@xmailserver.org From: Davide Libenzi To: Linux Kernel Mailing List Cc: Andrew Morton , Linus Torvalds , Ingo Molnar , Suparna Bhattacharya , Zach Brown , Benjamin LaHaise Date: Mon, 02 Apr 2007 15:46:31 -0700 Subject: [patch 13/13] signal/timer/event fds v10 - KAIO eventfd support example ... MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Message-ID: Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org ChangeLog: v10 - Added the "aio_flags" field (in place of the old "aio_reserved3") and introduced a new IOCB_FLAG_RESFD flag to tell that the "aio_resfd" field is valid. -- This is an example about how to add eventfd support to the current KAIO code, in order to enable KAIO to post readiness events to a pollable fd (hence compatible with POSIX select/poll). The KAIO code simply signals the eventfd fd when events are ready, and this triggers a POLLIN in the fd. This patch uses a reserved for future use member of the struct iocb to pass an eventfd file descriptor, that KAIO will use to post events every time a request completes. At that point, an aio_getevents() will return the completed result to a struct io_event. I made a quick test program to verify the patch, and it runs fine here: http://www.xmailserver.org/eventfd-aio-test.c The test program uses poll(2), but it'd, of course, work with select and epoll too. This can allow to schedule both block I/O and other poll-able devices requests, and wait for results using select/poll/epoll. In a typical scenario, an application would submit KAIO request using aio_submit(), and will also use epoll_ctl() on the whole other class of devices (that with the addition of signals, timers and user events, now it's pretty much complete), and then would: epoll_wait(...); for_each_event { if (curr_event_is_kaiofd) { aio_getevents(); dispatch_aio_events(); } else { dispatch_epoll_event(); } } Signed-off-by: Davide Libenzi - Davide Index: linux-2.6.21-rc5.fds/fs/aio.c =================================================================== --- linux-2.6.21-rc5.fds.orig/fs/aio.c 2007-04-02 15:06:11.000000000 -0700 +++ linux-2.6.21-rc5.fds/fs/aio.c 2007-04-02 15:06:47.000000000 -0700 @@ -30,6 +30,7 @@ #include #include #include +#include #include #include @@ -421,6 +422,7 @@ req->private = NULL; req->ki_iovec = NULL; INIT_LIST_HEAD(&req->ki_run_list); + req->ki_eventfd = ERR_PTR(-EINVAL); /* Check if the completion queue has enough free space to * accept an event from this io. @@ -462,6 +464,8 @@ { assert_spin_locked(&ctx->ctx_lock); + if (!IS_ERR(req->ki_eventfd)) + fput(req->ki_eventfd); if (req->ki_dtor) req->ki_dtor(req); if (req->ki_iovec != &req->ki_inline_vec) @@ -946,6 +950,14 @@ return 1; } + /* + * Check if the user asked us to deliver the result through an + * eventfd. The eventfd_signal() function is safe to be called + * from IRQ context. + */ + if (!IS_ERR(iocb->ki_eventfd)) + eventfd_signal(iocb->ki_eventfd, 1); + info = &ctx->ring_info; /* add a completion event to the ring buffer. @@ -1530,8 +1542,7 @@ ssize_t ret; /* enforce forwards compatibility on users */ - if (unlikely(iocb->aio_reserved1 || iocb->aio_reserved2 || - iocb->aio_reserved3)) { + if (unlikely(iocb->aio_reserved1 || iocb->aio_reserved2)) { pr_debug("EINVAL: io_submit: reserve field set\n"); return -EINVAL; } @@ -1555,6 +1566,19 @@ fput(file); return -EAGAIN; } + if (iocb->aio_flags & IOCB_FLAG_RESFD) { + /* + * If the IOCB_FLAG_RESFD flag of aio_flags is set, get an + * instance of the file* now. The file descriptor must be + * an eventfd() fd, and will be signaled for each completed + * event using the eventfd_signal() function. + */ + req->ki_eventfd = eventfd_fget((int) iocb->aio_resfd); + if (unlikely(IS_ERR(req->ki_eventfd))) { + ret = PTR_ERR(req->ki_eventfd); + goto out_put_req; + } + } req->ki_filp = file; ret = put_user(req->ki_key, &user_iocb->aio_key); Index: linux-2.6.21-rc5.fds/include/linux/aio.h =================================================================== --- linux-2.6.21-rc5.fds.orig/include/linux/aio.h 2007-04-02 15:06:11.000000000 -0700 +++ linux-2.6.21-rc5.fds/include/linux/aio.h 2007-04-02 15:06:47.000000000 -0700 @@ -119,6 +119,12 @@ struct list_head ki_list; /* the aio core uses this * for cancellation */ + + /* + * If the aio_resfd field of the userspace iocb is not zero, + * this is the underlying file* to deliver event to. + */ + struct file *ki_eventfd; }; #define is_sync_kiocb(iocb) ((iocb)->ki_key == KIOCB_SYNC_KEY) Index: linux-2.6.21-rc5.fds/include/linux/aio_abi.h =================================================================== --- linux-2.6.21-rc5.fds.orig/include/linux/aio_abi.h 2007-04-02 15:06:11.000000000 -0700 +++ linux-2.6.21-rc5.fds/include/linux/aio_abi.h 2007-04-02 15:06:47.000000000 -0700 @@ -45,6 +45,14 @@ IOCB_CMD_PWRITEV = 8, }; +/* + * Valid flags for the "aio_flags" member of the "struct iocb". + * + * IOCB_FLAG_RESFD - Set if the "aio_resfd" member of the "struct iocb" + * is valid. + */ +#define IOCB_FLAG_RESFD (1 << 0) + /* read() from /dev/aio returns these structures. */ struct io_event { __u64 data; /* the data field from the iocb */ @@ -84,7 +92,15 @@ /* extra parameters */ __u64 aio_reserved2; /* TODO: use this for a (struct sigevent *) */ - __u64 aio_reserved3; + + /* flags for the "struct iocb" */ + __u32 aio_flags; + + /* + * if the IOCB_FLAG_RESFD flag of "aio_flags" is set, this is an + * eventfd to signal AIO readiness to + */ + __u32 aio_resfd; }; /* 64 bytes */ #undef IFBIG